Finding the balance between overtreatment and undertreatment of ductal carcinoma in situ Elshof, L.E.

Transcription

1 UvA-DARE (Digital Academic Repository) Finding the balance between overtreatment and undertreatment of ductal carcinoma in situ Elshof, L.E. Link to publication Citation for published version (APA): Elshof, L. E. (2017). Finding the balance between overtreatment and undertreatment of ductal carcinoma in situ General rights It is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), other than for strictly personal, individual use, unless the work is under an open content license (like Creative Commons). Disclaimer/Complaints regulations If you believe that digital publication of certain material infringes any of your rights or (privacy) interests, please let the Library know, stating your reasons. In case of a legitimate complaint, the Library will make the material inaccessible and/or remove it from the website. Please Ask the Library: or a letter to: Library of the University of Amsterdam, Secretariat, Singel 425, 1012 WP Amsterdam, The Netherlands. You will be contacted as soon as possible. UvA-DARE is a service provided by the library of the University of Amsterdam ( Download date: 26 Jan 2019

2

3 FINDING THE BALANCE BETWEEN OVERTREATMENT AND UNDERTREATMENT OF DUCTAL CARCINOMA IN SITU Lotte Elisabeth Elshof

4 The work described in this thesis was performed at the Netherlands Cancer Institute - Antoni van Leeuwenhoek, Amsterdam, The Netherlands. In cooperation with the Netherlands Comprehensive Cancer Organization (IKNL), the nationwide network and registry of histo- and cytopathology in The Netherlands (PALGA foundation), the Dutch breast cancer screening facilities (BOB), Statistics Netherlands (CBS) and the European Organisation for Research and Treatment of Cancer (EORTC), Brussels, Belgium. The research was funded by grants from Pink Ribbon, A Sister s Hope and the Dutch Cancer Society (KWF Kankerbestrijding). Unrestricted financial support for publication of this thesis was kindly provided by Human Capital Care, Chipsoft and EURIN B.V. Copyright 2017 Lotte Elshof, Amsterdam, The Netherlands Design cover: Inspired by Golden Cosmos, Berlin, Germany and La cerise sur le gateau, Mulhouse, France / Designed by Hellen van Berge Henegouwen, E&H, Leiden, The Netherlands and Lotte Elshof Design inside: Bregje Jaspers, ProefschriftOntwerp, Nijmegen, The Netherlands Printed by: Gildeprint, Enschede, The Netherlands ISBN:

5 FINDING THE BALANCE BETWEEN OVERTREATMENT AND UNDERTREATMENT OF DUCTAL CARCINOMA IN SITU ACADEMISCH PROEFSCHRIFT ter verkrijging van de graad van doctor aan de Universiteit van Amsterdam op gezag van de Rector Magnificus prof. dr. ir. K.I.J. Maex ten overstaan van een door het College voor Promoties ingestelde commissie, in het openbaar te verdedigen in de Agnietenkapel op vrijdag 3 november 2017, te uur door Lotte Elisabeth Elshof geboren te Leiden

6 Promotiecommissie: Promotores: prof. dr. E.J.T. Rutgers AMC-UvA prof. dr. ir. F.E. van Leeuwen Vrije Universiteit Amsterdam Copromotores: dr. J. Wesseling NKI-AvL dr. M. Schaapveld NKI-AvL Overige leden: prof. dr. M.J. van de Vijver AMC-UvA prof. dr. J.H.G. Klinkenbijl AMC-UvA prof. dr. R.M. Pijnappel Universiteit Utrecht prof. dr. L.V. van de Poll-Franse Universiteit van Tilburg dr. N. Bijker AMC-UvA dr. C.M. Ronckers AMC-UvA dr. H.A.O. Winter-Warnars NKI-AvL Faculteit der Geneeskunde

7 Voor mijn grootouders Robert van Zonneveld arts-epidemioloog Elisabeth Coco Limborgh Meijer Gerrit Elshof Anna Hulshof

8

9 Table of Contents Chapter 1 General introduction and outline of the thesis 9 Chapter 2 Finding the balance between overtreatment and undertreatment of ductal carcinoma in situ 21 Chapter 3 Ductaal carcinoma in situ: de balans tussen overbehandeling en onderbehandeling 47 Chapter 4 Subsequent risk of ipsilateral and contralateral invasive breast cancer after treatment for ductal carcinoma in situ: incidence and the effect of radiotherapy in a population-based cohort of 10,090 women 59 Chapter 5 Cause-specific mortality in a population-based cohort of 9,799 women treated for ductal carcinoma in situ 81 Chapter 6 The method of detection of ductal carcinoma in situ has no therapeutic implications: results of a population-based cohort study 109 Chapter 7 Feasibility of a prospective, randomised, open-label, international multicentre, phase III, non-inferiority trial to assess the safety of active surveillance for low risk ductal carcinoma in situ - The LORD study 137 Chapter 8 General discussion, ongoing research and future prospects 165 Chapter 9 Summary Samenvatting List of co-authors and their contributions PhD portfolio Dankwoord Curriculum Vitae

10

11 CHAPTER 1 General introduction and outline of the thesis

12 Ductal carcinoma in situ Ductal carcinoma in situ (DCIS) is considered a non-obligate precursor of invasive breast cancer. 1 Risk factors associated with DCIS are similar to those associated with invasive breast cancer and include positive family history of breast cancer, increased breast density, and nulliparity or late age at fi rst full-term pregnancy. 2-5 DCIS is characterized by an abnormal epithelial proliferation confi ned to the mammary ductal-lobular system. 6,7 In DCIS, the outer myoepithelial layer is still organized and the neoplastic epithelial cells do not grow beyond the basement membrane into the surrounding stroma (Figure 1). Therefore, DCIS has no metastatic capacity, unlike invasive carcinoma. The term DCIS encompasses a heterogeneous group of lesions that vary with regard to their histologic appearance and clinical behaviour There is very limited information on the natural history of DCIS because surgical removal of the lesions hinders evaluation of their evolution. The proportion of DCIS lesions that will progress to invasive disease if untreated is estimated to be 15-50%. 12 Figure 1. Crossing the border. Tissue abnormalities known as DCIS may stay confi ned within a milk duct for a lifetime; a minority break out to become invasive cancer. Credit: K. Sutliff, 10 General introduction

13 Breast cancer screening DCIS rarely produces symptoms or signs and most DCIS lesions are diagnosed following detection of suspiciously grouped, pleomorphic, or fine linear calcifications on mammograms Undergoing mammography is thereby the strongest risk factor for being diagnosed with DCIS. While DCIS was seldom detected before the advent of screening mammography (about 300 women with DCIS among a total Dutch female population of approximately 5 million women aged 25 years and older in the Netherlands in 1989), the widespread adoption and subsequent digitalization of mammography-based population-wide breast cancer screening programs have led to an impressive increase in incidence of DCIS (Figure 2a). 16 In 2015, DCIS was diagnosed in approximately 2400 women in the Netherlands (total 25 years female population approximately 6 million), accounting for 14% of all newly diagnosed cases of breast cancer. 1 Overdiagnosis Breast cancer screening aims to detect breast disease that will ultimately cause harm and that is more likely to be cured if detected early. However, after having treated many women with screen-detected DCIS, the incidence of invasive breast cancer has not declined (Figure 2b). Unfortunately, population-based mammographic screening programs also identify a reservoir of indolent DCIS and thus lead to overdiagnosis (Figure 3). Overdiagnosis of DCIS can be defined as the detection of DCIS that would not go on to cause symptoms or would not progress to symptomatic or ultimately lethal invasive breast cancer if it had remained undetected and can be explained by regressing, non-progressing, or slowly growing lesions (Figure 4). 17 Based on autopsy series it is estimated that 10-39% of middle-aged women harbour indolent DCIS. 18 General introduction 11

14 a. a b. b Figure 2. Incidence of a DCIS by age group and b DCIS and invasive breast cancer for all age groups, in the Netherlands from 1989 to The Dutch breast cancer screening program started in From 1989 to 1997 women aged 50 to 69 years were the target population. Full coverage for these women was achieved in In 1998 the program was extended to women aged 70 to 75 years. 16,19,20 12 General introduction

15 1 Figure 3. Overdiagnosis. The detection of indolent DCIS lesions in the screened group produces apparent increases in the number of cases of DCIS (three in the screened group in the fi gure and one in the control group) and in survival (two of three patients in the screened group were treated and died of natural causes, without evidence of disease [66% survival], and the one patient in the control group did not survive [0% survival]), with no effect on mortality (one death from breast cancer in each group). Two patients in the control group died with undiagnosed DCIS that did not affect their natural life span. Adapted from Patz et al. 21 Figure 4. Heterogeneity of DCIS progression. The arrow labeled fast represents a fast-growing DCIS, one that quickly progresses to invasive breast cancer and death. The arrow labeled slow represents a slowgrowing DCIS, one that leads to invasive breast cancer but only after many years. The subsequent invasive breast cancer may lead to death, or the patient may die of some other cause. Depending on the defi nition of overdiagnosis and on the cause of death this DCIS can be labeled as overdiagnosed. The arrow labeled very slow represents a DCIS that never causes problems because the patient will die of some other cause before the DCIS has progressed to invasive breast cancer or has caused symptoms. The arrows labeled non-progressive and regressive represent DCIS that never cause problems because they never progress or even regress. Adapted from Welch and Black. 17 General introduction 13

16 Guidelines for DCIS treatment The aim of treating women with DCIS is to prevent the development of invasive breast cancer and ultimately to prevent breast cancer mortality. In the past, DCIS was typically treated by mastectomy like invasive breast cancer, with no decisive evidence of necessity. When in the mideighties breast-conserving treatment proved to be safe for patients with invasive breast cancer, several randomized clinical trials were initiated to investigate the role of radiotherapy after breastconserving surgery in women with DCIS Since then, women diagnosed with DCIS in the Netherlands have been treated by mastectomy, breast-conserving surgery plus radiotherapy, or breast-conserving surgery alone. The current Dutch treatment guidelines recommend mastectomy or breast-conserving treatment, consisting of microscopic tumour excision and radiotherapy. Contraindications for breast-conserving treatment include multicentricity and residual disease. 26 The long-term outcomes in patients with DCIS treated by these conventional therapies are excellent with 10-year local (in situ or invasive) recurrence rates of 3% after mastectomy, 14% after breast-conserving surgery with radiotherapy, and 26% after breast-conserving surgery without radiotherapy, and very high overall survival, regardless of treatment Hormonal therapy is not recommended for DCIS patients in the Netherlands because of limited benefit in terms of local control and survival, and the side effects involved. 26,30-33 However, the national guidelines in the Unites States recommend consideration of this adjuvant treatment in women who are oestrogen receptor-positive or who undergo breast-conserving surgery alone. 34 In addition, a trend toward more aggressive treatment of DCIS is observed. While the proportion of women undergoing unilateral mastectomy for DCIS has declined over the years, increasingly more DCIS patients opt for contralateral prophylactic mastectomy Further, the use of neoadjuvant chemotherapeutic agents or other anticancer drugs such as trastuzumab in the treatment of DCIS has also been suggested. 38,39 Despite the favourable prognosis many DCIS patients overestimate their risk for future breast cancer and disease spread to other parts of the body They tend to view the disease very much like invasive breast cancer This inaccurate, heightened risk perception may reflect a lack of clarity and standardized information among health-care providers resulting in suboptimal communication. 47,48 On the other hand, anxiety among DCIS patients might impede accurate risk perception General introduction

17 The DCIS dilemma: overtreatment versus undertreatment Overdiagnosis and inaccurate risk perception can lead to unnecessary treatment among women diagnosed with DCIS. On the other hand, some DCIS have high malignant potential and undertreatment of these lesions should be prevented. The inability to accurately stratify DCIS patients into low-risk and high-risk groups results in a clinical DCIS dilemma (Figure 5). Finding the balance between overtreatment and undertreatment of DCIS requires an integrative and multidisciplinary approach. To optimize individualized risk prediction the compilation of representative DCIS cohorts is imperative. The nationwide Netherlands Cancer Registry and the possibility to link this registry to other nationwide registries provide clinicians and researchers with a wealth of clinical data and enables reliable and complete data collection. The Dutch DCIS cohort established within the scope of this thesis provides the foundation upon which future clinical, morphological and molecular studies can build. 1 Figure 5. DCIS dilemma: overtreatment versus undertreatment. These curves represent the population of patients with DCIS who benefi t from (adjuvant) treatment (i.e. they will develop invasive breast cancer without treatment) and who do not benefi t from (adjuvant) treatment (i.e. even without treatment they will never develop invasive breast cancer). Risk factors to accurately stratify low-risk and high-risk groups have to be identifi ed yet. Relaxing risk factor criteria for recommending treatment decreases overtreatment but risks increasing undertreatment. Restricting risk factor criteria decreases undertreatment but increases overtreatment. Adapted from Smith, General introduction 15

18 Outline This thesis aims to frame the DCIS dilemma by reviewing current knowledge and evaluating clinical outcome in a nationwide cohort of women treated for DCIS. Further, we developed a randomized non-inferiority clinical trial to assess safety of active surveillance in women with DCIS and explored international interest and feasibility of this study. Chapters 2 and 3 of this thesis review our current understanding of DCIS and identify current knowledge gaps. Chapter 2 proposes an integrative step-wise approach to solve the DCIS dilemma. Chapters 4, 5 and 6 provide the results of our population-based cohort studies for which we used data from the Netherlands Cancer Registry, the Dutch nationwide network and registry of histology and cytopathology (PALGA), the Dutch screening organization and Statistics Netherlands (cause of death). In chapter 4 we assessed the risk of subsequent ipsilateral and contralateral invasive breast cancer among 10,090 women diagnosed with primary pure DCIS between 1989 and 2004 in the Netherlands. Specifically, we evaluated the effect of different treatment strategies. In chapter 5 we compared cause-specific mortality between DCIS patients and women in the general population to evaluate excess mortality. Chapter 6 describes the prognostic role of method of detection among women with screen-detected, interval and nonscreening-related DCIS. Chapter 7 discusses the rationale and design of a randomized clinical trial between standard treatment and active surveillance for LOw-Risk DCIS (LORD). In addition, it provides the results of an explorative feasibility study among 53 European centres. In chapter 8 the ongoing research and future prospects of DCIS management and research are comprehensively discussed after a short discussion of the results. The thesis ends with a summary of the presented results in chapter Outline of the thesis

19 References 1. Cowell CF, Weigelt B, Sakr RA, et al. Progression from ductal carcinoma in situ to invasive breast cancer: revisited. Molecular Oncology. 2013;7(5): Kerlikowske K, Barclay J, Grady D, Sickles EA, Ernster V. Comparison of risk factors for ductal carcinoma in situ and invasive breast cancer. J Natl Cancer Inst. 1997;89(1): Claus EB, Stowe M, Carter D. Breast carcinoma in situ: risk factors and screening patterns. J Natl Cancer Inst. 2001;93(23): Claus EB, Stowe M, Carter D. Family history of breast and ovarian cancer and the risk of breast carcinoma in situ. Breast Cancer Res Treat. 2003;78(1): Claus EB, Petruzella S, Matloff E, Carter D. Prevalence of BRCA1 and BRCA2 mutations in women diagnosed with ductal carcinoma in situ. JAMA. 2005;293(8): Pinder SE, Ellis IO. The diagnosis and management of pre-invasive breast disease: ductal carcinoma in situ (DCIS) and atypical ductal hyperplasia (ADH)--current definitions and classification. Breast Cancer Res. 2003;5(5): Schnitt SJ, Collins LC. Biopsy Interpretation of the Breast. 2nd ed. Lippincott Williams & Wilkins, a Wolters Kluwer business; Leonard GD, Swain SM. Ductal carcinoma in situ, complexities and challenges. J Natl Cancer Inst. 2004;96(12): Allred DC. Biomarkers predicting recurrence and progression of ductal carcinoma in situ treated by lumpectomy alone. J Natl Cancer Inst. 2010;102(9): Bijker N, van Tienhoven G. Local and systemic outcomes in DCIS based on tumor and patient characteristics: the radiation oncologist s perspective. J Natl Cancer Inst Monographs. 2010;2010(41): Schnitt SJ. Local outcomes in ductal carcinoma in situ based on patient and tumor characteristics. J Natl Cancer Inst Monographs. 2010;2010(41): Erbas B, Provenzano E, Armes J, Gertig D. The natural history of ductal carcinoma in situ of the breast: a review. Breast Cancer Res Treat. 2006;97(2): Stomper PC, Connolly JL, Meyer JE, Harris JR. Clinically occult ductal carcinoma in situ detected with mammography: analysis of 100 cases with radiologic-pathologic correlation. Radiology. 1989;172(1): Sullivan DC. Ductal carcinoma in situ: atypical mammographic appearances. Radiology. 1990;175(1): Evans A. The diagnosis and management of pre-invasive breast disease: radiological diagnosis. Breast Cancer Res. 2003;5(5): IKNL CCCTN. Netherlands Cancer Registry Welch HG, Black WC. Overdiagnosis in cancer. J Natl Cancer Inst. 2010;102(9): Welch HG, Black WC. Using autopsy series to estimate the disease reservoir for ductal carcinoma in situ of the breast: how much more breast cancer can we find? Ann Intern Med. 1997;127(11): de Koning HJ, Fracheboud J, Boer R, et al. Nation-wide breast cancer screening in The Netherlands: support for breast-cancer mortality reduction. National Evaluation Team for Breast Cancer Screening (NETB). Int J Cancer. 1995;60(6): General introduction and outline of the thesis 17

20 19. Fracheboud J, de Koning HJ, Beemsterboer PM, et al. Nation-wide breast cancer screening in The Netherlands: results of initial and subsequent screening National Evaluation Team for Breast Cancer Screening. Int J Cancer. 1998;75(5): Patz EF, Goodman PC, Bepler G. Screening for lung cancer. N Engl J Med. 2000;343(22): Julien JP, Bijker N, Fentiman IS, et al. Radiotherapy in breast-conserving treatment for ductal carcinoma in situ: first results of the EORTC randomised phase III trial EORTC Breast Cancer Cooperative Group and EORTC Radiotherapy Group. Lancet. 2000;355(9203): Fisher B, Costantino J, Redmond C, et al. Lumpectomy compared with lumpectomy and radiation therapy for the treatment of intraductal breast cancer. N Engl J Med. 1993;328(22): Emdin SO, Granstrand B, Ringberg A, et al. SweDCIS: Radiotherapy after sector resection for ductal carcinoma in situ of the breast. Results of a randomised trial in a population offered mammography screening. Acta Oncol. 2006;45(5): Houghton J, George WD, Cuzick J, et al. Radiotherapy and tamoxifen in women with completely excised ductal carcinoma in situ of the breast in the UK, Australia, and New Zealand: randomised controlled trial. Lancet. 2003;362(9378): IKNL CCCTN, the Knowledge Institute of Medical Specialists KiMS. Guideline Breast Cancer Stuart KE, Houssami N, Taylor R, Hayen A, Boyages J. Long-term outcomes of ductal carcinoma in situ of the breast: a systematic review, meta-analysis and meta-regression analysis. BMC Cancer. 2015;15(1): Early Breast Cancer Trialists Collaborative Group (EBCTCG), Correa C, McGale P, et al. Overview of the randomized trials of radiotherapy in ductal carcinoma in situ of the breast. J Natl Cancer Inst Monographs. 2010;2010(41): Narod SA, Iqbal J, Giannakeas V, Sopik V, Sun P. Breast Cancer Mortality After a Diagnosis of Ductal Carcinoma In Situ. JAMA Oncol. JAMA Oncol. 2015;1: Cuzick J, Sestak I, Pinder SE, et al. Effect of tamoxifen and radiotherapy in women with locally excised ductal carcinoma in situ: long-term results from the UK/ANZ DCIS trial. Lancet Oncol. 2011;12(1): Allred DC, Anderson SJ, Paik S, et al. Adjuvant tamoxifen reduces subsequent breast cancer in women with estrogen receptor-positive ductal carcinoma in situ: a study based on NSABP protocol B-24. J Clin Oncol. 2012;30(12): Elshof LE, Tryfonidis K, Slaets L, et al. Feasibility of a prospective, randomised, open-label, international multicentre, phase III, non-inferiority trial to assess the safety of active surveillance for low risk ductal carcinoma in situ - The LORD study. Eur J Cancer. 2015;51(12): Staley H, McCallum I, Bruce J. Postoperative Tamoxifen for ductal carcinoma in situ: Cochrane systematic review and meta-analysis. Breast. July National Comprehensice Cancer Network. NCCN Guideline for Treatment of Breast Cancer. Accessed February 24, Tuttle TM, Jarosek S, Habermann EB, et al. Increasing rates of contralateral prophylactic mastectomy among patients with ductal carcinoma in situ. J Clin Oncol. 2009;27(9): Soran A, Kamali Polat A, Johnson R, McGuire KP. Increasing trend of contralateral prophylactic mastectomy: what are the factors behind this phenomenon? Surgeon. 2014;12(6): Rutter CE, Park HS, Killelea BK, Evans SB. Growing Use of Mastectomy for Ductal Carcinoma-In Situ of the Breast Among Young Women in the United States. Ann Surg Oncol. 2015;22(7): General introduction and outline of the thesis

21 37. Boughey JC, Gonzalez RJ, Bonner E, Kuerer HM. Current treatment and clinical trial developments for ductal carcinoma in situ of the breast. Oncologist. 2007;12(11): Kuerer HM, Albarracin CT, Yang WT, et al. Ductal carcinoma in situ: state of the science and roadmap to advance the field. J Clin Oncol. 2009;27(2): Partridge A, Adloff K, Blood E, et al. Risk perceptions and psychosocial outcomes of women with ductal carcinoma in situ: longitudinal results from a cohort study. J Natl Cancer Inst. 2008;100(4): Ruddy KJ, Meyer ME, Giobbie-Hurder A, et al. Long-term risk perceptions of women with ductal carcinoma in situ. Oncologist. 2013;18(4): De Morgan S, Redman S, D Este C, Rogers K. Knowledge, satisfaction with information, decisional conflict and psychological morbidity amongst women diagnosed with ductal carcinoma in situ (DCIS). Patient Educ Couns. 2011;84(1): Rakovitch E, Franssen E, Kim J, et al. A comparison of risk perception and psychological morbidity in women with ductal carcinoma in situ and early invasive breast cancer. Breast Cancer Res Treat. 2003;77(3): van Gestel YRBM, Voogd AC, Vingerhoets AJJM, et al. A comparison of quality of life, disease impact and risk perception in women with invasive breast cancer and ductal carcinoma in situ. Eur J Cancer. 2007;43(3): Liu Y, Pérez M, Aft RL, et al. Accuracy of perceived risk of recurrence among patients with early-stage breast cancer. Cancer Epidemiol Biomarkers Prev. 2010;19(3): Liu Y, Pérez M, Schootman M, et al. A longitudinal study of factors associated with perceived risk of recurrence in women with ductal carcinoma in situ and early-stage invasive breast cancer. Breast Cancer Res Treat. 2010;124(3): Partridge A, Winer JP, Golshan M, et al. Perceptions and management approaches of physicians who care for women with ductal carcinoma in situ. Clin Breast Cancer. 2008;8(3): Kennedy F, Harcourt D, Rumsey N. Perceptions of ductal carcinoma in situ (DCIS) among UK health professionals. Breast. 2009;18(2): Smith GL. Toward Minimizing Overtreatment and Undertreatment of Ductal Carcinoma In Situ in the United States. J Clin Oncol. 2016;34(11): General introduction and outline of the thesis 19

22 Emma J. Groen Lotte E. Elshof Lindy L. Visser Emiel J. Th. Rutgers Hillegonda A.O. Winter-Warnars Esther H. Lips Jelle Wesseling The Breast 2017; 31:

23 CHAPTER 2 Finding the balance between overtreatment and undertreatment of ductal carcinoma in situ

24 Abstract With the widespread adoption of population-based breast cancer screening, ductal carcinoma in situ (DCIS) has come to represent 20-25% of all breast neoplastic lesions diagnosed. Current treatment aims at preventing invasive breast cancer, but the majority of DCIS lesions will never progress to invasive disease. Still, DCIS is treated by surgical excision, followed by radiotherapy as part of breast conserving treatment, and/or endocrine therapy. This implies overtreatment of the majority of DCIS, as less than 1% of DCIS patients will go on to develop invasive breast cancer annually. If we are able to identify which DCIS is likely to progress or recur as invasive breast cancer and which DCIS would remain indolent, we can treat the first group intensively, while sparing the second group from such unnecessary treatment (surgery, radiotherapy, endocrine therapy) preserving the quality of life of these women. This review summarizes our current knowledge on DCIS and the risks involved regarding progression into invasive breast cancer. It also shows current knowledge gaps, areas where profound research is highly necessary for women with DCIS to prevent their over-treatment in case of a harmless DCIS, but provide optimal treatment for potentially hazardous DCIS. 22 Finding the balance between overtreatment and undertreatment of ductal carcinoma in situ

25 Introduction Since the introduction of population-based breast cancer screening and digital mammography, the incidence of precursor lesions has substantially increased in the Western world, without a decline in invasive breast cancer incidence. This suggests that overdiagnosis of such lesions exists. Most precursor lesions are ductal carcinoma in situ (DCIS) cases. DCIS of the breast represents a heterogeneous group of neoplastic lesions confined to the breast ducts and lobules that differ in histologic appearance and biological potential. The major gap in our current understanding of DCIS is, that we do not know yet which DCIS lesions will develop into invasive breast cancer and which will not. As a consequence, women with low risk DCIS face being harmed by intensive treatment without any benefit. If such overtreatment can be avoided without compromising the excellent outcomes presently achieved, this will safely spare many women with low risk DCIS intensive treatment and so preserve their quality of life. Here, we summarize our current understanding of DCIS and the challenges that lie ahead of us to find the balance between DCIS overtreatment and undertreatment. 2 DCIS incidence has increased over time In the United States (US), the incidence of DCIS markedly increased from 5.8 per 100,000 women in the 1970s to 32.5 per 100,000 women in 2004 and then reached a plateau [1]. Approximately 25% of breast neoplastic lesions diagnosed in the US are DCIS, i.e. over 51,000 women in the US alone in 2015 [2]. In the Netherlands and the United Kingdom (UK), similar rates apply ( This increase is attributed primarily to the widespread adoption of mammographic screening in the US, Europe and other high-income countries that has dramatically increased the number of DCIS cases, as more than 90% of all cases of DCIS are detected only on imaging studies [1]. DCIS is less common than invasive breast cancer. Like invasive breast cancer, the risk increases with age. DCIS is uncommon in women younger than 30. In the US, the rate of DCIS increases with age from 0.6 per 1000 screening examinations in women aged years to 1.3 per 1000 screening examinations in women aged years [4]. Risk of development of metastases and/or death in a patient diagnosed with pure DCIS is very low (<1%) [5]. The risk factors for DCIS and invasive breast cancer are similar, and include family history of breast cancer, increased breast density, obesity, and nulliparity or late age at first birth [6-9]. DCIS is also a component of the inherited breast-ovarian cancer syndrome defined by deleterious mutations in BRCA1 and BRCA2 genes; mutation rates, i.e. up to 5%, are similar to those for invasive breast cancer [9]. Finding the balance between overtreatment and undertreatment of ductal carcinoma in situ 23

26 Over-treatment of DCIS exists Increasing DCIS incidence is due mostly to introduction and uptake of population-based breast cancer screening [1,10-12] and use of digital mammography [13,14]. In the Netherlands, the incidence of in situ lesions has increased 5.6-fold between 1989 and 2011 ( nl). Higher screening sensitivity also labels more women as having disease, many of whom may never develop invasive cancer [15,16]. However, the incidence of advanced breast cancer has not decreased, despite screening [13,17]. In addition, there is strong evidence that treatment of DCIS in most women has no clear effect on mortality reduction [18]. This suggests overdiagnosis and hence overtreatment exists of DCIS in general, and of lowgrade DCIS in particular. The implication is that we could manage a subgroup of women with low-grade DCIS using active surveillance only [11,19,20]. The number of women eligible for this management strategy would be high, since 80% of all in situ carcinomas are DCIS lesions, and about 20% of all DCIS lesions is low-grade [21,22]. Fig. 1 illustrates the heterogeneous course of cancer, including its preliminary stages. Remarkably, a lesion with a similar risk of progression to invasive breast cancer is classic lobular carcinoma in situ [23,24]. If LCIS is the only finding, active surveillance is frequently offered. Somewhat incongruously, this risk is acceptable for both patients and clinicians. Most DCIS lesions go undetected Only 10% of DCIS cases are detected due to symptoms, such as nipple discharge, Paget s disease of the nipple, or a palpable mass [1]. As pointed out above, the majority of DCIS lesions found are detected by screening, as many DCIS lesions do not come with symptoms, but do contain calcifications that can be seen upon mammography. Obviously, DCIS lesions may be occult by mammography or the diameter of the area containing calcifications underestimates the extent of DCIS [25,26]. This is also illustrated by the much higher prevalence of DCIS (7-39%) found in autopsy studies concerning the age group for which population-based screening programs are in place, whereas in screening and clinical practice breast cancer was diagnosed in only 1% of women within a similar age range [21,27]. 24 Finding the balance between overtreatment and undertreatment of ductal carcinoma in situ

27 2 Figure 1. Heterogeneity of cancer progression in general. Fast growing cancers are likely to lead to symptoms, and even death, after a relatively short period of time, whereas slow growing cancers may lead to symptoms, and maybe death, after many years. The very slowly proliferating lesions most likely never lead to symptoms, implying that patients with such a lesion are likely to die due to other causes. This is also true for non-progressive lesions that might even regress. Adapted from Welch and Black [16]. The natural course of DCIS is poorly understood A multitude of factors have been implicated in the risk of developing an in situ or invasive recurrence [28]. It has been suggested that paracrine regulation is crucial for malignantly transformed luminal cells to become invasive [29]. By analysing stromal expression signatures in DCIS, it was shown that the microenvironment plays a role in the transition from pre-invasive to invasive growth [30,31]. The myoepithelium is considered as a factor preventing invasive growth by regulating luminal cell polarity, ductal morphogenesis, and basement membrane deposition. In DCIS, the myoepithelium shows decreased expression of e.g. thrombospondin, laminin, and oxytocin, promoting proliferation, migration, invasion, and angiogenesis [32,33]. It is uncertain whether changes in stroma and/or myoepithelium precede invasive growth or that the luminal DCIS cells can induce stromal and/or myoepithelial changes, and thereby paving the way for their own invasion. The pathology of DCIS provides limited prognostic value The pathology of DCIS aims to assess subtype and grade. Additionally, pathology will report on extent and margin status in case of surgical resection of DCIS. These aspects provide important prognostic information about the aggressiveness of a particular DCIS lesion. DCIS Finding the balance between overtreatment and undertreatment of ductal carcinoma in situ 25

28 is morphologically described by growth pattern, i.e. arrangement of the ductal cells, such as cribriform, solid, micropapillary, etc., cytoplasmic features, degree of nuclear pleomorphism, and degree of mitotic activity. Grading systems for DCIS are based on these cytonuclear features resulting in low (1), intermediate (2), or high (3) grade [34]. However, the accuracy of DCIS grading has some limitations, as diagnostic criteria are not always clear. Furthermore, poor to modest interobserver agreement exists, as has been reported in subgroups of in situ lesions, which is mainly due to differences in morphological interpretation and field selection in the often heterogeneous intraductal lesions [35-39]. Obviously, it is of utmost importance to classify the primary lesion reliably to be able to evaluate the natural course of DCIS or to interpret follow-up after treatment. Reliability studies are hard to compare as they often differ in study design. Also they are limited due to: mostly examining a small number of highly selected cases [35,40-45]; being assessed by expert breast pathologists only and; often after being giving instructions or tutorials beforehand [35,41,42,45,46]. Translation of these findings into daily practice is therefore questionable and, so far, has not reduced inter-observer variability. In addition, the interpretation of results and evaluation of potential bias is complicated by often inadequate reporting and missing information on important issues in reliability testing. In 2011, guidelines for reporting reliability and agreement studies were developed, highlighting key methodological issues that should be carefully thought through when reporting on reliability and agreement studies [47]. A recent study used exactly these guidelines to construct their study design [48]. In this study 115 pathologists each classified a set of 60 cases as either benign without atypia, atypia, DCIS, or invasive carcinoma. They found an overall concordance rate of 75%, and concordance rates of 48% and 84% for atypia and DCIS respectively, when compared with expert reference diagnoses. The concordance rate for invasive breast cancer was excellent (96%). Test cases were randomly selected, oversampling atypia and DCIS cases, and the participating pathologists had different geographic and clinical setting backgrounds leaving little room for selection bias. Unfortunately, as gold standard they used consensus-derived expert diagnoses without any information on follow up. Whether such results on concordance are biologically relevant, therefore remain unanswered. Although inter-observer variability may lead to overtreatment of DCIS, even with perfect (i.e. biologically relevant) definitions and classification systems, a 100% agreement will never be reached, as histological examination is not an absolute science. Hopefully the integration of various clinical, radiological, histological, and molecular markers will improve our ability to reliably distinguish between low- and high-risk lesions. 26 Finding the balance between overtreatment and undertreatment of ductal carcinoma in situ

29 A DCIS diagnosis comes with a chance missing invasive breast cancer Invasive breast cancer has been found in 8-43% of resection specimens from patients who were primarily diagnosed with DCIS based on a preoperative biopsy [49-58]. At least some of these highly variable numbers can be explained by differences in the size and quantity of biopsies taken as well as by the use of different imaging techniques. In addition, it is essential to be informed about why the biopsy was taken. This is illustrated by the fact that the risk of upgrade will be higher when mass lesions or architectural distortions are found on imaging compared to calcifications only [49,51,53,55,57,58]. Most studies also agree that larger lesions - based on the effect of size on imaging diagnosis - carry a higher upgrade risk than smaller ones [49-52,55,59]. Paradoxically, the upgrade risk for smaller tumours is higher, because the sensitivity of mammographically detecting tumours of only 0.5 cm is low (<30%) and high (>90%) for tumours of 1.0 cm [60]. In some studies a higher grade of DCIS was a significant predictor of upgrade with an upgrade occurring in only 6-7% of patients versus 13-31% for low and intermediate/high grade respectively [52-54,58]. Others found grade not to be predictive [49-51,53,55]. 2 Uncertainties about DCIS risks exist Among health care providers as well as among women with DCIS, considerable uncertainties exist regarding the degree of risk involved for developing invasive breast cancer. In general, DCIS has a relative risk of 8-11 for subsequent development of invasive carcinoma [22,34]. DCIS in itself has an excellent long-term breast cancer-specific survival exceeding 98% after 10 years of follow-up [1,61,62]. Strikingly, grade was not significantly associated with the risk of local recurrence. Factors associated with DCIS progression to invasive breast carcinoma remain poorly defined, because most patients are treated in order to completely eradicate the lesion [63]. Several studies have shown that high-grade DCIS has a higher probability of ipsilateral invasive breast cancer than low-grade DCIS. One of the largest studies is that conducted by The Eastern Cooperative Oncology Group (ECOG). This non-randomized prospective study included 670 patients with either low/intermediate grade DCIS or high-grade DCIS who underwent complete excision. At a median follow-up of 6.7 years, the low-intermediate group had a 10.5% risk of local relapse, whereas the high-grade group had an 18% recurrence rate, of which 35% were invasive breast cancers [13]. Our group analysed an unbiased, large population-based, nation-wide cohort, comprising 10,090 women with a primary diagnosis of DCIS between 1989 and 2004 (this thesis). In total, 5.8% developed ipsilateral invasive recurrence after treatment for DCIS (breast sparing or mastectomy) after a median follow-up of 11.6 years (Chapter 4). Narod and coworkers analysed the SEER database and showed that women younger than 35 and women of African-American descent have a higher risk of invasive recurrence and death [18]. A meta-analysis of four Finding the balance between overtreatment and undertreatment of ductal carcinoma in situ 27

30 randomized clinical trials to investigate the role of radiotherapy in BCT for DCIS after a complete local excision of the lesion showed a 50% reduction in the risk of local recurrence with half of these recurrences being invasive, but has no effect on breast cancer-specific mortality [15]. Taken together, these studies provide a generalized estimation of how large the risk is that DCIS progresses into invasive breast cancer, but without allowing individualized prediction. Current DCIS management is at the safe side In DCIS, prognosis is based on the risk of (invasive) local recurrence, although such risk estimations are far from precise as described above. If the lesion is not too extensive, breastconserving treatment for DCIS is frequently recommended, resulting in 60-70% of women being suitable for this therapy [64]. If the lesion is too extensive, a mastectomy with or without immediate reconstruction is generally advised. Radiotherapy after surgery is nowadays standard treatment for DCIS, as randomized controlled trials have demonstrated a 50% reduction in ipsilateral breast cancer risk [15]. For tamoxifen use there is no consensus if there is any absolute survival benefit that outweighs the harm of long-term endocrine treatment [65,66]. According to Dutch, English, and American guidelines and based on higher upgrade risks, indications for a sentinel lymph node biopsy (SLNB) in DCIS patients planning to undergo breast sparing surgery include a palpable mass, age below 55 years, intermediate or high grade DCIS, and a solid mass or a lesion larger than 25 mm or extensive calcifications on imaging (see e.g. As a SLNB is less reliable after mastectomy, it is also recommended for all patients treated by mastectomy. Strikingly, there is a tendency towards minimizing axillary surgery for invasive breast cancer [67,68]. There are now even trials investigating whether a SLNB can be left out of treatment of clinically node negative invasive breast cancer patients [69]. To date, no comparable trials have been undertaken for patients with only DCIS, while logically, risks seem even lower. It has been shown that even a positive SLNB in DCIS patients does not affect survival, although some patients did receive systemic treatment [70-72]. We need to await more definitive results indicating that omitting a SLNB for women with pure DCIS patients is likely to be safe. Distinguishing harmless from potentially hazardous DCIS is challenging Evidently, overtreatment of harmless DCIS should be prevented, without compromising the excellent outcomes presently achieved in DCIS management. This means being able to reliably distinguish harmless from potentially hazardous DCIS. Therefore, on-going research aims to find 28 Finding the balance between overtreatment and undertreatment of ductal carcinoma in situ

31 and validate much more accurate prognostic biomarkers, applying e.g. immunohistochemistry and genomic techniques, pursuing the Holy Grail in prediction will be described below. A multitude of markers have been implicated in identifying subgroups of DCIS by immunohistochemistry (IHC; see for a brief overview Table 1). The most commonly used markers are ER, PR, HER2, and Ki67. As in invasive breast cancer, they are sometimes used to determine the subtype and aggressiveness of DCIS. Expression of the hormone receptors, a low-grade, and a low percentage of Ki67-positive cells in DCIS are related to a lower rate of invasive recurrence and/or lower grade [4,73-75]. In general, overexpression of HER2 is associated with higher recurrence rates [74]. Besides the usual markers, expression of p16 and p53 is related to a higher local recurrence rate [4,76-79]. COX-2 is related to proliferation and as such risk on local recurrence [4,79]. Annexin A1 (ANXA1) might play dualistic roles being involved in variable mechanisms related to cancer development and progression. Loss of ANXA1 expression, as observed in the majority of breast cancers, seems to be related to early events of malignant transformation. However, overexpression was shown to be associated with poor relapse free survival [80,81]. Interestingly, intra-individual DCIS heterogeneity (high Ki67, mutant p53, and low p16) is associated with more aggressive DCIS [77]. This is relevant for the interpretation of further genomic profiling of DCIS. However, the impact of most of these studies is limited, as they involve small patient series relate to series with an adjacent invasive component and are therefore not pure DCIS, and information on follow up is also often lacking [82,83]. In recent years, several studies have also focused on finding molecular markers associated with aggressiveness in DCIS [28]. The use of laser capture microdissection to harvest defined cell populations has proven essential for the study of DCIS. Studies on DCIS and an adjacent invasive component have shown that molecular characteristics associated with invasiveness are already present in the DCIS lesion [84,85]. Petridis and co-workers showed that shared genetic susceptibility exists for DCIS and invasive ductal carcinoma (IDC) and that studies with larger numbers are needed to determine if IDC or DCIS specific loci exists [86]. Gene expression analysis has shown that pre-invasive lesions and invasive breast cancer display remarkable similar gene expression patterns [85]. Carraro et al. summarized differently expressed genes associated with aggressive behaviour of DCIS lesions [87]. Genes belonging to cell signalling (i.e. CDH1), cellular movement (MMPs), growth and proliferation are involved. Other studies focus on specific copy number alterations. 16q loss is found in the majority of low-grade DCIS lesions, while more complex karyotypes are observed in high-grade lesions. Specific copy number aberrations reported to be associated with DCIS are amplifications of MYC, FGFR1 and CCND1) [88]. Complicating factors in the studies employed so far are the low numbers of samples studied and the heterogeneity between lesions and within the lesions [89,90]. 2 Finding the balance between overtreatment and undertreatment of ductal carcinoma in situ 29

32 Table 1. Selected antigens reported to be related with aggressiveness of DCIS based on: (1) differential expression of the antigen between DCIS and IDC; (2) multivariable significance; (3) confirmation in more than 1 research paper. Antigens No. of cases Finding(s) Reference ER/PR 119 DCIS Presence of PR expression is associated with expression of ER and lack of comedo-necrosis in DCIS. Increasing tumour grade correlated with decrease in ER and PR positivity. Comedo-necrosis is associated with ER and PR negativity. 118 pure DCIS, 100 IBC Invasion is associated with a significant increase in Ki67 expression and decreases in ER, PR and Her-2 expression. 95 DCIS A direct positive relationship is observed for the expression of ER, PR and Bcl-2 negativity for the clinical recurrence of DCIS. HER2 180 DCIS HER2neu is regarded as an important prognostic and predictive marker, with its overexpression predicting local recurrence. 118 pure DCIS, 100 IBC Invasion is associated with a significant increase in Ki67 expression and decreases in ER, PR and Her-2 expression. 130 DCIS, 159 DCIS+IBC No significant differences between the gene amplification status of DCIS and invasive breast cancer concerning HER2, ESR1, CCND1, and MYC. Data suggest an early role of all analysed gene amplifications in breast cancer development but not in the initiation of invasive tumour growth. 226 DCIS cases Data suggests loss of RB can contribute to the function of ErbB2 (HER2) in driving disease progression. ErbB2 (HER2) alone is not sufficient to drive invasion into the surrounding matrix. RB deficiency potentially cooperates with ErbB2 loss and drive the phenotype towards EMT. AR Findings suggest that decreases in AR and androgenmetabolising enzymes (17βHSD5 and 5αR1) may be involved in the increased biological aggressiveness in triple-negative breast cancer. Also relating to triple-neg DCIS. Ki initial DCIS p16+ COX-2+ and Ki67+ in DCIS is prognostic for recurrence/invasive cancer and suggests that the biological correlation between COX-2 levels and proliferation may be significant. 36 DCIS+IBC Multiple DCIS lesions from the same patient frequently exhibit heterogeneity in the expression of clinically relevant markers: PR, HER2, Ki-67, and p16. Individuals with a heterogeneous DCIS cell population combined with high levels of Ki-67, increased mutant p53 and low p16 should be clinically managed more aggressively. p pure DCIS, 100 IBC Invasion is associated with a significant increase in Ki67 expression and decreases in ER, PR and Her-2 expression. P53 more frequent in high-grade DCIS. 103 DCIS Expression of mutant p53 is associated with high expression of VEGF and correlates with biological aggressiveness of DCIS lesions. 36 DCIS+IBC Multiple DCIS lesions from the same patient frequently exhibit heterogeneity in the expression of clinically relevant markers: PR, HER2, Ki-67, and p16. Individuals with a heterogeneous DCIS cell population combined with high levels of Ki-67, increased mutant p53 and low p16 should be clinically managed more aggressively. p initial DCIS p16+ COX-2+ and Ki67+ in DCIS is prognostic for recurrence of DCIS and/or invasive cancer. 50 DCIS, 50 IDC, 50 benign Luminal lesions of DCIS with high p16 are more likely to develop into aggressive breast cancer. p16 expression in luminal A breast cancer is associated with progression from DCIS to IDC. [107] [73] [75] [74] [73]] [108] [109] [110] [4] [77] [73] [111] [77] [4] [78] 30 Finding the balance between overtreatment and undertreatment of ductal carcinoma in situ

33 Antigens No. of cases Finding(s) Reference 40 UDH, 20 FEA, 40 ADH, 40 DCIS p16ink4a methylation is associated with DCIS, plays an important role in the initiation and progression of premalignant lesions and carcinomas and may be a crucial event in cell transformation. 36 DCIS+IBC Multiple DCIS lesions from the same patient frequently exhibit heterogeneity in the expression of clinically relevant markers: PR, HER2, Ki-67, and p16. Individuals with a heterogeneous DCIS cell population combined with high levels of Ki-67, increased mutant p53 and low p16 should be clinically managed more aggressively. MYC 141 DCIS, 18 DCIS+IBC High expression of c-myc in DCIS did not predict local recurrence, but still is of interest. Has to be confirmed in a larger trial. 130 DCIS, 159 DCIS+IBC No significant differences between the gene amplification status of DCIS and invasive breast cancer concerning HER2, ESR1, CCND1, and MYC. Data suggest an early role of all analysed gene amplifications in breast cancer development but not in the initiation of invasive tumour growth. COX-2 58 pure DCIS Findings suggest that COX-2 may be a predictive marker of early relapse in with DCIS 324 initial DCIS P16+ COX-2+ and Ki67+ in DCIS is prognostic for recurrence/ invasive cancer and suggests that the biological correlation between COX-2 levels and proliferation may be significant. ALDH1 236 DCIS Combination of EZH2 with ALDH1 within the DCIS epithelial compartment is associated with the prognosis for ipsilateral breast event and invasive progression. EZH2 236 DCIS Combination of EZH2 with ALDH1 within the DCIS epithelial compartment is associated with the prognosis for ipsilateral breast event and invasive progression. ANXA 82 IBC+LN metastasis and 21 DCIS+IBC Lack of ANXA1 expression in breast cancer and early loss of ANXA1 in DCIS, suggests a possible role for ANXA1 in early events of malignant transformation. 182 cases Significant loss of ANXA1 in DCIS and IBC as compared to normal. ANXA1 overexpression was correlated with poor RFS. [112] [77] [113] [108] [79] [4] [114] [114] [80] [81] 2 The Oncotype DX DCIS score is the first multi-gene assay that has been claimed to be validated in an independent study [91]. This score predicts both the risk of an in situ and invasive recurrence but still assumes that every DCIS should be treated by surgery, as the assay merely indicates patients having benefit from radiotherapy. Prospective validation of this assay has not been done yet. Taken together, a conclusive set of biomarkers suitable for implementation in routine clinical practice has not been identified yet. Campbell et al. therefore argued for the development of a PreCancer Genome Atlas to gain insight in the earliest molecular and cellular events associated with cancer initiation, which eventually will enable us to find biomarkers for risk stratification. Finding the balance between overtreatment and undertreatment of ductal carcinoma in situ 31

34 Solving the DCIS dilemma requires integrated and novel approaches Current pathology has limited additional value for more nuanced clinical practice when dealing with DCIS, its diagnosis and consequences for the women involved. We need to more seriously consider opportunities for integrated and novel approaches. To prioritize DCIS research, the US Patient-Centred Outcomes Research Institute commissioned a study to do so [92]. Stakeholders prioritized evidence gaps related to incorporation of patient-centred outcomes into future studies on DCIS, development of better methods to predict risk for invasive cancer, evaluation of a strategy of active surveillance, and testing of decisionmaking tools. First, individualized risk prediction should be optimized using well annotated retrospective data sets, enabling integration of clinical, morphological and molecular features. Strikingly, such an integrative approach is not available yet. Ultimately, such tools should be able to distinguish harmless from potentially hazardous screen-detected DCIS and help clinicians and women with DCIS to decide between management using active surveillance or more intensive treatment. For this, data from population-based screening, hospital records, cancer registries, pathology, current and upcoming molecular and biological techniques should be integrated in a stepwise manner: 1. Compile representative DCIS patient cohorts and collect all necessary data and material. Better methods to predict DCIS risk rely on large series of clinical data and tissue blocks for histopathologic and molecular analysis. Such studies have started in the Netherlands with the collection of a large nationwide, population-based, retrospective study (n = 10,090) (this thesis). Clinical, radiological and molecular data will be integrated and compared between women with DCIS who may or may not have developed an ipsilateral invasive recurrence after breast-conserving treatment, during a follow up period of more than 10 years. The excellent registration in the Netherlands at the Dutch Cancer Registry (NCR), the breast cancer screening and PALGA (Pathology National Automated Archive) is unique in the world and enables reliable and makes complete data collection possible. Another huge effort is the Sloane Project, a UK wide prospective audit of screen detected non-invasive and atypical hyperplasia of the breast. All UK NHS Breast Screening Units are invited to participate. It is a multi-disciplinary project involving radiologists, pathologists, surgeons and oncologists. Detailed follow up data of all DCIS detected by the NHS Breast Screening Program will be collected such as information on local recurrence, contralateral cancer, metastases, and death, as well as data on screening and treatment, and most importantly for biomarker research, tissue blocks will be collected, enabling molecular pathology studies ( 2. Find and validate molecular markers related to outcome. To obtain reliable, detailed results, DCIS should be analysed applying immunohistochemistry and genomic analysis on resection specimens, as the size of the biopsies is too small for these analyses. For these analyses, 32 Finding the balance between overtreatment and undertreatment of ductal carcinoma in situ

35 laser microdissection or alternative strategies should be used to capture the cells and tissue regions of interest at high specificity. Comprehensive genomic characterisation has to be done to understand the biological properties of DCIS that contribute to the evolution and aggressiveness of DCIS. This includes complete description of all drivers and mutation signatures in DCIS, exploring intralesional heterogeneity in DCIS, and finding putative associations between mutation signatures (see [93]) and the risk of progression into invasive breast cancer. By this means, clonal evolution, evolutionary pathways, and rare events in DCIS related to outcome (recurrence, progression to invasive disease) can be characterized. In addition, we can also test if genetic and microenvironmental diversity, including immune responses [94], provide universal biomarkers, helping to predict progression to invasive disease. This innovative approach could yield a universally applicable construct for understanding interactions between precancerous lesions and their environments. 3. Apply innovative molecular imaging technologies to understand the transition of DCIS into invasive breast cancer. The missing link in the full molecular picture can be obtained by analysing subregions of a DCIS lesion, e.g. by applying Mass Spectrometry Imaging (MSI), as this technique can be successfully applied on formalin-fixed, paraffin-embedded tissue [95]. Our preliminary evidence shows substantial intralesional heterogeneity of putative genomic markers in DCIS. Perhaps only a small part of the DCIS lesion has invasive potential, which means our tools need to be able to detect molecular differences within the lesion. Most likely, MSI has vital additional value in combination with advanced bioinformatics and statistical analysis, to characterize intralesional heterogeneity to determine phenotypes based on specific molecular signatures at different levels (e.g. metabolomics, lipidomics, and peptidomics). 4. Integrate clinical, morphological, and molecular data to build a robust risk stratification tool. Associations between clinical, morphological, and molecular data should be analysed to build a model accurately predicting subsequent risk for developing ipsilateral invasive breast cancer. Candidate risk stratification tools should then be thoroughly validated in independent retrospective DCIS series and prospective clinical trials. In order to communicate such a risk prediction model to patients and doctors, risk calculator software should be developed in analogy to existing calculators such as adjuvant online (www. adjuvantonline.com) and the breast cancer risk assessment tool ( These online tools have proven themselves to be very helpful and easy to use, which is essential when incorporated into daily practice. This ultimately will provide holistic integrative profiles per patient and an innovative multifactorial algorithm able to identify patients with very low-risk for invasive recurrence, i.e. indolent DCIS, that can be managed safely by active surveillance only. This can save many women from the potential physical and psychological harm of invasive treatment. Evidently, such an approach will only be successful if international collaboration between experienced dedicated researchers, clinicians, and patient partners are well established. 2 Finding the balance between overtreatment and undertreatment of ductal carcinoma in situ 33

36 Second, prospective studies on active surveillance should be conducted to deliver final proof that active surveillance is safe for DCIS already known to be low-risk. For example, the international LORD trial (LOw Risk Dcis; Chapter 7), which will start to recruit women with low grade DCIS in Europe in 2016 under the auspices of the European Organisation for Research and Treatment of Cancer (EORTC). In this study, women with pure low-grade DCIS detected at screening based on calcifications are randomized to either an active surveillance policy or standard therapy [19]. After inclusion, women will be followed for 10 years and main outcome measure is the risk of developing invasive breast cancer. If a relapse occurs, breast-conserving therapy with radiation therapy will still be an option. By contrast, when a recurrence develops after standard treatment for DCIS, an ablation is usually the only choice. Similar studies are the LORIS trial in the UK [96], the COMET trial in the US ( and the Australian LARRIKIN trial for which no detailed information is available yet (see Table 2). In moving forward, the following considerations are of paramount importance. First, low-grade hormone receptor-positive invasive breast cancer grows only a few millimetres per year and a delay in detection will not affect the excellent prognosis inherent to these tumours [97]. Second, there is convincing evidence that low-grade invasive breast tumours originate from low grade precursor lesions [84,98-103]. Third, women with low-grade lesions who meet these criteria for inclusion in the LORIS trial did not show any upgrade to invasive cancer [104]. This underlines again that active surveillance for women with low-grade screen-detected DCIS is likely to be a safe option, sparing these women the harms of ineffective treatment, preserving their quality of life. Adequate communication about DCIS risks involved is key In general, improving communication about the diagnosis and prognosis of DCIS patients will likely deliver the most essential improvements in the management of DCIS. This because there is much uncertainty about the long-term implications of the diagnosis of DCIS (including the risk of invasive breast cancer, therapeutic efficacy and safety), making it difficult for patients and health care providers to make well-informed decisions on treatment options. For a woman, it is difficult to understand that on the one hand DCIS is a breast cancer precursor but not yet an invasive disease, and on the other hand that intensive treatment is necessary. It is essential to better assess the risks involved and put these into perspective, taking into account the quality of life and competitive factors in terms of morbidity and mortality. Educating health care providers and developing a risk prediction model will contribute to this better understanding. It has been shown for prostate cancer, that such a strategy is well accepted [105,106]. 34 Finding the balance between overtreatment and undertreatment of ductal carcinoma in situ

37 Conclusion The incidence of DCIS has increased substantially. The rationale of DCIS treatment is mortality reduction as a result of invasive breast carcinoma. However, pure DCIS (without any invasive component) usually shows no symptoms and does not cause mortality. We know that a significant proportion of the DCIS lesions will never lead to invasive breast cancer. But right now we do not know which DCIS lesions will progress and which will not. The result of this knowledge gap is that every DCIS lesion is treated similar to invasive breast cancer. Risk stratification is therefore essential for making better-informed treatment decisions. In addition, large randomized clinical trials are necessary to investigate if active surveillance is an option for low grade DCIS. Last but not least, communicating in a correct and nuanced manner about the implications of the diagnosis of DCIS is essential for a realistic risk perception and optimal decision-making by the patient and the health care professionals involved. 2 Conflict of interest statement None of the authors have a conflict of interest. Acknowledgments Financial support for research in our team provided by Pink Ribbon Netherlands (2011.WO19. C88; ; ), Dutch Cancer Society/Alpe d Huzes (NKI ; NKI CT; NKI ALPE), and A Sister s Hope. We would like to thank Jonathan Watson for critically reading this manuscript and his helpful suggestions. Finding the balance between overtreatment and undertreatment of ductal carcinoma in situ 35

38 Table 2. Comparison of the designed and initiated prospective, randomised, open-label, phase III, non-inferiority trials to test whether less intensive treatment of low risk DCIS is safe. The information provided is based on literature for the LORIS and LORD trial [19,96] and on personal communication for the COMET and LARRIKIN trial. Trial name LORD LORIS COMET LARRIKIN Clarification acronym/ trial name Low risk DCIS Low risk DCIS Comparison of operative versus medical endocrine therapy for low risk DCIS The Australian slang word larrikin is associated with the Australian identity: a bloke who refuses to stand on ceremony. Trial status Recruitment will start in 2016 Recruiting from July 2014 Not yet recruiting Funding request submitted Setting and locations Mainland Europe (n > 30) United Kingdom (n > 20) United States (n = 100) Australia and New Zealand (n 12) Inclusion criteria Women 45 years with asymptomatic, pure low-grade DCIS based on representative vacuum-assisted biopsies (at least 6) of unilateral, calcifications only of any size detected by population-based or opportunistic screening mammography. Women 46 years with asymptomatic pure, non-high grade DCIS (e.g. low grade DCIS and intermediate grade DCIS with low grade features) based on vacuum assisted core biopsies of screen-detected or incidental calcifications only of any size (uni-/ bilateral). Women 40 years with pure, non-mass forming low-risk DCIS, e.g. ER + and/or PR + and HER- 2 receptor-negative grade I or II DCIS based on a core biopsy without evidence of other breast disease on physical examination and breast imaging within 6 months of registration. Women 55 years with pure, asymptomatic and low risk DCIS (low and intermediate grade) based on either a core biopsy and/or vacuum-assisted biopsy or open diagnostic surgical biopsy of screen detected or incidental calcifications (uni/bilateral but unifocal) 20 mm. Exclusion criteria No prior history of DCIS or invasive breast cancer, a BRCA 1/2 gene mutation present in family, no bilateral DCIS, synchronous contralateral invasive breast cancer, lobular carcinoma in situ, Paget s disease, or invasive breast disease on cytology/histology No prior history or current diagnosis of invasive breast cancer or ipsilateral DCIS and no high risk group for developing breast cancer Not known. No previous or current diagnosis of invasive cancer, previous ipsilateral DCIS, Paget s disease or LCIS, pregnancy/lactation or a known BRCA1/2 mutation Central review No central review of pathology. Real time central review of histological slides by expert DCIS pathologists. Not known. No central review planned. 36 Finding the balance between overtreatment and undertreatment of ductal carcinoma in situ

39 Randomisation between standard treatment according to physician and patient choice (surgery with/ without radiotherapy) and active surveillance. Patients in both groups are free to decide whether to opt for endocrine therapy for 5 years. Randomisation between standard treatment including surgery and radiation and active surveillance. Patients in both groups are free to decide whether to choose endocrine therapy. Randomisation between standard surgical and adjuvant treatment according to local policy and active surveillance, with specific notification that patients in the latter group should not receive anti-oestrogen treatment. Interventions Randomisation between standard treatment according to local policy (wide local excision with/without radiotherapy, mastectomy and possibly hormonal therapy by Tamoxifen) and active surveillance. Both study arms will be monitored with annual mammography for 10 years. Anti-oestrogen treatment is not allowed in the active surveillance arm. Both study arms will be monitored with annual digital mammography for 10 years. Both groups will be monitored with annual mammography for at least 10 years and regular clinical examinations or at patient s request for 5 years then annually. Both study arms will be carefully monitored with mammograms and physical exams every 6 months for 5 years. Randomisation Allocation ratio 1:1 Allocation ratio 1:1 Allocation ratio 1:1 Allocation ratio 1:1 Safety will be measured by ipsilateral breast cancer free survival at 5 and 10 years. Safety will be measured by assessing the invasive cancer rate in the affected breast at 2 and 5 years. Safety will be measured by ipsilateral invasive breast cancerfree survival time at 5 and 10 years. Primary end-points Safety will be measured by ipsilateral invasive breast cancerfree percentage at 5 and 10 years. -Rate of invasive disease and higher grade DCIS in final pathology specimen -Time to development of ipsilateral and any invasive breast cancer -Ipsilateral mastectomy rate at 5 years -Biopsy rate during follow-up -Overall survival -Time to failure of active surveillance strategy -Quality of Life -Cost Effectiveness -Mastectomy and breast conservation rate -Contralateral invasive cancer rate -Overall and disease specific survival -Breast MRI rate -Breast biopsy rate -Radiation rate -Chemotherapy rate -Psychosocial outcomes -Decision quality -Financial burden/employment -Time to development of ipsilateral, contralateral and any invasive breast cancer -Overall survival -Time to mastectomy or surgery -Quality of Life -Quality-adjusted life years -Translational exploratory assessment of predictive biomarkers -Patient reported outcomes -Cost-effectiveness Secondary end-points -Rate of invasive disease or DCIS grade 2/3 at final pathology specimen -Time to ipsilateral grade II or III DCIS and time to contralateral DCIS -Cumulative incidence of contralateral invasive breast cancer -Ipsilateral mastectomy rate -Biopsy rate during follow-up -Time to failure of active surveillance strategy -Distant metastases free interval -Overall survival -Central collection of imaging data and biosamples for translational research purposes -Patient reported outcomes -Cost-effectiveness Sample size needed 1240 patients 932 patients 1189 patients 550 patients 2 Finding the balance between overtreatment and undertreatment of ductal carcinoma in situ 37

40 References 1. Virnig BA, Tuttle TM, Shamliyan T, Kane RL. Ductal carcinoma in situ of the breast: a systematic review of incidence, treatment, and outcomes. JNCI J Natl Cancer Inst 2010;102: Siegel RL, Miller KD, Jemal A. Cancer statistics. CA A Cancer J Clin 2016;66(6): Kerlikowske K, Molinaro AM, Gauthier ML, Berman HK, Waldman F, Bennington J, et al. Biomarker expression and risk of subsequent tumors after initial ductal carcinoma in situ diagnosis. JNCI J Natl Cancer Inst 2010;102: Roses RE, Arun BK, Lari SA, Mittendorf EA, Lucci A, Hunt KK, et al. Ductal carcinoma-in-situ of the breast with subsequent distant metastasis and death. Ann Surg Oncol 2011;18: Kerlikowske K, Barclay J, Grady D, Sickles EA, Ernster V. Comparison of risk factors for ductal carcinoma in situ and invasive breast cancer. J Natl Cancer Inst 1997;89: Claus EB, Stowe M, Carter D. Breast carcinoma in situ: risk factors and screening patterns. J Natl Cancer Inst 2001;93: Claus EB, Stowe M, Carter D. Family history of breast and ovarian cancer and the risk of breast carcinoma in situ. Breast Cancer Res Treat 2003;78: Claus EB, Petruzella S, Matloff E, Carter D. Prevalence of BRCA1 and BRCA2 mutations in women diagnosed with ductal carcinoma in situ. JAMA J Am Med Assoc 2005;293: Ernster VL, Ballard-Barbash R, Barlow WE, Zheng Y, Weaver DL, Cutter G, et al. Detection of ductal carcinoma in situ in women undergoing screening mammography. JNCI J Natl Cancer Inst 2002;94: Esserman LJ, Thompson IM, Reid B. Overdiagnosis and overtreatment in cancer: an opportunity for improvement. JAMA J Am Med Assoc 2013;310: Marshall E. Breast cancer. Dare to do less. Science 2014;343: Hughes LL, Wang M, Page DL, Gray R, Solin LJ, Davidson NE, et al. Local excision alone without irradiation for ductal carcinoma in situ of the breast: a trial of the Eastern cooperative oncology group. J Clin Oncol 2009;27: Bluekens AMJ, Holland R, Karssemeijer N, Broeders MJM, Heeten den GJ. Comparison of digital screening mammography and screen-film mammography in the early detection of clinically relevant cancers: a multicenter study. Radiology 2012;265: Early Breast Cancer Trialists Collaborative Group (EBCTCG). Overview of the randomized trials of radiotherapy in ductal carcinoma in situ of the breast. J Natl Cancer Inst Monogr 2010;2010: Welch HG, Black WC. Overdiagnosis in Cancer. JNCI J Natl Cancer Inst 2010;102: Bleyer A, Welch HG. Effect of three decades of screening mammography on breast-cancer incidence. N Engl J Med 2012;367: Narod SA, Iqbal J, Giannakeas V, Sopik V, Sun P. Breast cancer mortality after a diagnosis of ductal carcinoma in situ. JAMA Oncol 2015;1: Elshof LE, Tryfonidis K, Slaets L, van Leeuwen-Stok AE, Skinner VP, Dif N, et al. Feasibility of a prospective, randomised, open-label, international multicentre, phase III, non-inferiority trial to assess the safety of active surveillance for low risk ductal carcinoma in situ e the LORD study. Eur J Cancer 2015;51: Wong JS. Prospective study of wide excision alone for ductal carcinoma in situ of the breast. J Clin Oncol 2006;24: Finding the balance between overtreatment and undertreatment of ductal carcinoma in situ

41 20. Kuerer HM, Albarracin CT, Yang WT, Cardiff RD, Brewster AM, Symmans WF, et al. Ductal carcinoma in situ: state of the science and roadmap to advance the field. J Clin Oncol 2009;27: Siziopikou KP. Ductal carcinoma in situ of the breast: current concepts and future directions. Arch Pathol Lab Med 2013;137: Lakhani SR, Audretsch W, Cleton-Jensen A-M, Cutuli B, Ellis I, Eusebi V, et al. The management of lobular carcinoma in situ (LCIS). Is LCIS the same as ductal carcinoma in situ (DCIS)? Eur J Cancer 2006;42: Ottesen GL, Graversen HP, Blichert-Toft M, Christensen IJ, Andersen JA. Carcinoma in situ of the female breast. 10 year follow-up results of a prospective nationwide study. Breast Cancer Res Treat 2000;62: Holland R, Hendriks JH, Vebeek AL, Mravunac M, Stekhoven JHS. Extent, distribution, and mammographic/histological correlations of breast ductal carcinoma in situ. Lancet 1990;335: Going JJ, Moffat DF. Escaping from Flatland: clinical and biological aspects of human mammary duct anatomy in three dimensions. J Pathol 2004;203: Welch HG, Black WC. Using autopsy series to estimate the disease reservoir for ductal carcinoma in situ of the breast: how much more breast cancer can we find? Ann Intern Med 1997;127: Lopez-Garcia MA, Geyer FC, Lacroix-Triki M, Marchio C, Reis-Filho JS. Breast cancer precursors revisited: molecular features and progression pathways. Histopathology 2010;57: Hanahan D, Weinberg RA. Hallmarks of Cancer: the next generation. Cell 2011;144: Sharma M, Beck AH, Webster JA, Espinosa I, Montgomery K, Varma S, et al. Analysis of stromal signatures in the tumor microenvironment of ductal carcinoma in situ. Breast Cancer Res Treat 2009;123: Ma X-J, Dahiya S, Richardson E, Erlander M, Sgroi DC. Gene expression profiling of the tumor microenvironment during breast cancer progression. Breast Cancer Res 2009;11:R Hu M, Yao J, Carroll DK, Weremowicz S, Chen H, Carrasco D, et al. Regulation of in situ to invasive breast carcinoma transition. Cancer Cell 2008;13: Cichon MA, Degnim AC, Visscher DW, Radisky DC. Microenvironmental influences that drive progression from Benign breast disease to invasive breast Cancer. J Mammary Gland Biol Neoplasia 2010;15: Bane A. Ductal carcinoma in situ: what the pathologist needs to know and why. Int J Breast Cancer 2013;2013: Elston CW, Sloane JP, Amendoeira I, Apostolikas N, Bellocq JP, Bianchi S, et al. Causes of inconsistency in diagnosing and classifying intraductal proliferations of the breast. European commission working group on breast screening pathology. Eur J Cancer 2000;36: O Malley FP, Mohsin SK, Badve S, Bose S, Collins LC, Ennis M, et al. Interobserver reproducibility in the diagnosis of flat epithelial atypia of the breast. Mod Pathol 2006;19: Jain RK, Mehta R, Dimitrov R, Larsson LG, Musto PM, Hodges KB, et al. Atypical ductal hyperplasia: interobserver and intraobserver variability. Mod Pathol 2011;24: Holland R, Peterse JL, Millis RR, Eusebi V, Faverly D, van de Vijver MJ, et al. Ductal carcinoma in situ: a proposal for a new classification. Semin Diagn Pathol 1994;11: van de Vijver MJ, Peterse H. The diagnosis and management of pre-invasive breast disease: pathological diagnosiseproblems with existing classifications. Breast Cancer Res 2003;5: Scott MA, Lagios MD, Axelsson K, Rogers LW, Anderson TJ, Page DL. Ductal carcinoma in situ of the breast: reproducibility of histological subtype analysis. Hum Pathol 1997;28: Finding the balance between overtreatment and undertreatment of ductal carcinoma in situ 39

42 40. Bethwaite P, Smith N, Delahunt B, Kenwright D. Reproducibility of new classification schemes for the pathology of ductal carcinoma in situ of the breast. J Clin Pathology 1998;51: Schnitt SJ, Connolly JL, Tavassoli FA, Fechner RE, Kempson RL, Gelman R, et al. Interobserver reproducibility in the diagnosis of ductal proliferative breast lesions using standardized criteria. Am J Surg Pathology 1992;16: Haupt B, Schwartz MR, Xu Q, Ro JY. Columnar cell lesions: a consensus study among pathology trainees. HumPathol 2010;41: Fitzgibbons PL. Atypical lobular hyperplasia of the breast: a study of pathologists responses in the college of American pathologists performance improvement program in surgical pathology. Arch Pathol Lab Med 2000;124: Schuh F, Biazús JV, Resetkova E, Benfica CZ, Edelweiss MIA. Reproducibility of three classification systems of ductal carcinoma in situ of the breast using a web-based survey. Pathology Res Pract 2010;206: Wells WA, Carney PA, Eliassen MS, Grove MR, Tosteson AN. Pathologists agreement with experts and reproducibility of breast ductal carcinoma-insitu classification schemes. Am J Surg Pathology 2000;24: Kottner J, Audige L, Brorson S, Donner A, Gajewski BJ, Hrobjartsson A, et al. Guidelines for reporting reliability and agreement studies (GRRAS) were proposed. Int J Nurs Stud 2011;48: Elmore JG, Longton GM, Carney PA, Geller BM, Onega T, Tosteson ANA, et al. Diagnostic concordance among pathologists interpreting breast biopsy specimens. Jama 2015;313: Kondo T, Hayashi N, Ohde S, Suzuki K, Yoshida A, Yagata H, et al. A model to predict upstaging to invasive carcinoma in patients preoperatively diagnosed with ductal carcinoma in situ of the breast. J Surg Oncol 2015;112: Diepstraten SCE, van de Ven SMWY, Pijnappel RM, Peeters PHM, van den Bosch MAAJ, Verkooijen HM, et al. Development and evaluation of a prediction model for underestimated invasive breast cancer in women with ductal carcinoma in situ at stereotactic large core needle biopsy. PLoS One 2013;8:e Han JS, Molberg KH, Sarode V. Predictors of invasion and axillary lymph node metastasis in patients with a core biopsy diagnosis of ductal carcinoma in situ: an analysis of 255 cases. Breast J 2011;17: Lee SK, Yang JH, Woo S-Y, Lee JE, Nam SJ. Nomogram for predicting invasion in patients with a preoperative diagnosis of ductal carcinoma in situ of the breast. Br J Surg 2013;100: Meijnen P, Oldenburg HSA, Loo CE, Nieweg OE, Peterse JL, Rutgers EJT. Risk of invasion and axillary lymph node metastasis in ductal carcinoma in situ diagnosed by core-needle biopsy. Br J Surg 2007;94: Houssami N, Ciatto S, Ellis I, Ambrogetti D. Underestimation of malignancy of breast core-needle biopsy: concepts and precise overall and categoryspecific estimates. Cancer 2007;109: Huo L, Sneige N, Hunt KK, Albarracin CT, Lopez A, Resetkova E. Predictors of invasion in patients with core-needle biopsy-diagnosed ductal carcinoma in situ and recommendations for a selective approach to sentinel lymph node biopsy in ductal carcinoma in situ. Cancer 2006;107: Mittendorf EA, Arciero CA, Gutchell V, Hooke J, Shriver CD. Core biopsy diagnosis of ductal carcinoma in situ: an indication for sentinel lymph node biopsy. Curr Surg 2005;62: Finding the balance between overtreatment and undertreatment of ductal carcinoma in situ

43 56. Park HS, Park S, Cho J, Park JM, Kim SI, Park B-W. Risk predictors of underestimation and the need for sentinel node biopsy in patients diagnosed with ductal carcinoma in situ by preoperative needle biopsy. J Surg Oncol 2013;107: Wilkie C, White L, Dupont E, Cantor A, Cox CE. An update of sentinel lymph node mapping in patients with ductal carcinoma in situ. Am J Surg 2005;190: Brennan ME, Turner RM, Ciatto S, Marinovich ML, French JR, Macaskill P, et al. Ductal carcinoma in situ at core-needle biopsy: meta-analysis of underestimation and predictors of invasive breast cancer. Radiology 2011;260: Weedon-Fekjaer H, Lindqvist BH, Vatten LJ, Aalen OO, Tretli S. Breast cancer tumor growth estimated through mammography screening data. Breast Cancer Res 2008; Ernster VL, Barclay J, Kerlikowske K, Wilkie H, Ballard-Barbash R. Mortality among women with ductal carcinoma in situ of the breast in the populationbased surveillance, epidemiology and end results program. Archives Intern Med 2000;160: Bijker N. Breast-conserving treatment with or without radiotherapy in ductal carcinoma-in-situ: tenyear results of European organisation for research and treatment of Cancer randomized phase III trial 10853ea study by the EORTC breast Cancer cooperative group and EORTC radiotherapy group. J Clin Oncol 2006;24: Collins LC, Tamimi RM, Baer HJ, Connolly JL, Colditz GA, Schnitt SJ. Outcome of patients with ductal carcinoma in situ untreated after diagnostic biopsy. Cancer 2005;103: Bijker N, Donker M, PhD MD, Wesseling J, Heeten den GJ, Rutgers EJT, et al. Is DCIS breast cancer, and how do I treat it? Curr Treat Options Oncol 2013;14: Wapnir IL, Dignam JJ, Fisher B, Mamounas EP, Anderson SJ, Julian TB, et al. Long-term outcomes of invasive ipsilateral breast tumor recurrences after lumpectomy in NSABP B-17 and B-24 randomized clinical trials for DCIS. JNCI J Natl Cancer Inst 2011;103: Cuzick J, Sestak I, Pinder SE, Ellis IO, Forsyth S, Bundred NJ, et al. Effect of tamoxifen and radiotherapy in women with locally excised ductal carcinoma in situ: long-term results from the UK/ ANZ DCIS trial. Lancet Oncol 2011;12: Giuliano AE, McCall L, Beitsch P, Whitworth PW, Blumencranz P, Leitch AM, et al. Locoregional recurrence after sentinel lymph node dissection with or without axillary dissection in patients with sentinel lymph node metastases: the American college of surgeons oncology group Z0011 randomized trial. Ann Surg 2010;252: Galimberti V, Cole BF, Zurrida S, Viale G, Luini A, Veronesi P, et al. Axillary dissection versus no axillary dissection in patients with sentinel-node micrometastases (IBCSG 23-01): a phase 3 randomised controlled trial. Lancet Oncol 2013;14: Gentilini O, Veronesi U. Abandoning sentinel lymph node biopsy in early breast cancer? A new trial in progress at the European Institute of Oncology of Milan (SOUND: sentinel node vs observation after axillary UltraSouND). Breast 2012;21: Broekhuizen LN, Wijsman JH, Peterse JL, Rutgers EJT. The incidence and significance of micrometastases in lymph nodes of patients with ductal carcinoma in situ and T1a carcinoma of the breast. Eur J Surg Oncol 2006;32: Francis AM, Haugen CE, Grimes LM, Crow JR, Yi M, Mittendorf EA, et al. Is sentinel lymph node dissection warranted for patients with a diagnosis of ductal carcinoma in situ? Ann Surg Oncol 2015;22:4270e9. 2 Finding the balance between overtreatment and undertreatment of ductal carcinoma in situ 41

44 71. van Roozendaal LM, Goorts B, Klinkert M, Keymeulen KBMI, De Vries B, Strobbe LJA, et al. Sentinel lymph node biopsy can be omitted in DCIS patients treated with breast conserving therapy. Breast Cancer Res Treat 2016;156: Sarode VR, Han JS, Morris DH, Peng Y, Rao R. A comparative analysis of biomarker expression and molecular subtypes of pure ductal carcinoma in situ and invasive breast carcinoma by image analysis: relationship of the subtypes with histologic grade, Ki67, p53 overexpression, and DNA ploidy. Int J Breast Cancer 2011;2011: Han K, Nofech-Mozes S, Narod S, Hanna W, Vesprini D, Saskin R, et al. Expression of HER2neu in ductal carcinoma in situ is associated with local recurrence. Clin Oncol R Coll Radiol 2012;24: Provenzano E, Hopper JL, Giles GG, Marr G, Venter DJ, Armes JE. Histological markers that predict clinical recurrence in ductal carcinoma in situ of the breast: an Australian population-based study. Pathology 2004;36: de Roos MA, de Bock GH, de Vries J, van der Vegt B, Wesseling J. p53 overexpression is a predictor of local recurrence after treatment for both in situ and invasive ductal carcinoma of the breast. J Surg Res 2007;140: Pape-Zambito D, Jiang Z, Wu H, Devarajan K, Slater CM, Cai KQ, et al. Identifying a highly-aggressive DCIS subgroup by studying intra-individual DCIS heterogeneity among invasive breast Cancer patients. PLoS One 2014;9: e Shan M, Zhang X, Liu X, Qin Y, Liu T, Liu Y, et al. P16 and p53 play distinct roles in different subtypes of breast cancer. PLoS One 2013;8:e Generali D, Buffa FM, Deb S, Cummings M, Reid LE, Taylor M, et al. COX-2 expression is predictive for early relapse and aromatase inhibitor resistance in patients with ductal carcinoma in situ of the breast, and is a target for treatment. Br J Cancer 2014;111: Cao Y, Li Y, Edelweiss M, Arun B, Rosen D, Resetkova E, et al. Loss of annexin A1 expression in breast cancer progression. Appl Immunohistochem Mol Morphol 2008;16: Yom CK, Han W, Kim S-W, Kim HS, Shin H-C, Chang JN, et al. Clinical significance of annexin A1 expression in breast Cancer. J Breast Cancer 2011;14: Hernandez L, Wilkerson PM, Lambros MB, Campion-Flora A, Rodrigues DN, Gauthier A, et al. Genomic and mutational profiling of ductal carcinomas in situ and matched adjacent invasive breast cancers reveals intra-tumour genetic heterogeneity and clonal selection. J Pathol 2012;227: Abba MC, Gong T, Lu Y, Lee J, Zhong Y, Lacunza E, et al. A molecular portrait of high-grade ductal carcinoma in situ. Cancer Res 2015;75: Ma X-J, Salunga R, Tuggle JT, Gaudet J, Enright E, McQuary P, et al. Gene expression profiles of human breast cancer progression. Proc Natl Acad Sci 2003;100: Vincent-Salomon A, Lucchesi C, Gruel N, Raynal V, Pierron G, Goudefroye R, et al. Integrated genomic and transcriptomic analysis of ductal carcinoma in situ of the breast. Clin Cancer Res 2008;14: Petridis C, Brook MN, Shah V, Kohut K, Gorman P, Caneppele M, et al. Genetic predisposition to ductal carcinoma in situ of the breast. Breast Cancer Res 2016;18: Carraro DM, Elias EV, Andrade VP. Ductal carcinoma in situ of the breast: morphological and molecular features implicated in progression. Biosci Rep 2014;34: Cowell CF, Weigelt B, Sakr RA, Ng CKY, Hicks J, King TA, et al. Progression from ductal carcinoma in situ to invasive breast cancer: revisited. Mol Oncol 2013;7: Finding the balance between overtreatment and undertreatment of ductal carcinoma in situ

45 88. Martelotto LG, Ng CK, Piscuoglio S, Weigelt B, Reis-Filho JS. Breast cancer intra-tumor heterogeneity. Breast Cancer Res 2014;16:R Wang Y, Waters J, Leung ML, Unruh A, Roh W, Shi X, et al. Clonal evolution in breast cancer revealed by single nucleus genome sequencing. Nature 2014;512: Rakovitch E, Nofech-Mozes S, Hanna W, Baehner FL, Saskin R, Butler SM, et al. A populationbased validation study of the DCIS Score predicting recurrence risk in individuals treated by breastconserving surgery alone. Breast Cancer Res Treat 2015;152: Gierisch JM, Myers ER, Schmit KM, Crowley MJ, McCrory DC, Chatterjee R, et al. Prioritization of research addressing management strategies for ductal carcinoma in situ. Ann Intern Med 2014;160: Nik-Zainal S, Davies H, Staaf J, Ramakrishna M, Glodzik D, Zou X, et al. Landscape of somatic mutations in 560 breast cancer whole-genome sequences. Nature Thompson E, Taube JM, Elwood H, Sharma R, Meeker A, Warzecha HN, et al. The immune microenvironment of breast ductal carcinoma in situ. Mod Pathol 2016;29: Mascini NE, Eijkel GB, Brugge Ter P, Jonkers J, Wesseling J, Heeren RMA. The use of mass spectrometry imaging to predict treatment response of patientderived xenograft models of triplenegative breast cancer. J Proteome Res 2015;14: Francis A, Thomas J, Fallowfield L, Wallis M, Bartlett JMS, Brookes C, et al. Addressing overtreatment of screen detected DCIS. LORIS trial Eur J Cancer Spratt JA, Fournier von D, Spratt JS, Weber EE. Mammographic assessment of human breast cancer growth and duration. Cancer 1993;71: Abdel-Fatah TM, Powe DG, Hodi Z, Lee AH, Reis-Filho JS, Ellis IO. High frequency of coexistence of columnar cell lesions, lobular neoplasia, and low grade ductal carcinoma in situ with invasive tubular carcinoma and invasive lobular carcinoma. Am J Surg Pathol 2007;31: Abdel-Fatah TM, Powe DG, Hodi Z, Reis-Filho JS, Lee AH, Ellis IO. Morphologic and molecular evolutionary pathways of low nuclear grade invasive breast cancers and their putative precursor lesions: further evidence to support the concept of low nuclear grade breast neoplasia family. Am J Surg Pathol 2008;32: Balleine RL, Webster LR, Davis S, Salisbury EL, Palazzo JP, Schwartz GF, et al. Molecular grading of ductal carcinoma in situ of the breast. Clin Cancer Res 2008;14: Roylance R, Gorman P, Harris W, Liebmann R, Barnes D, Hanby A, et al. Comparative genomic hybridization of breast tumors stratified by histological grade reveals new insights into the biological progression of breast cancer. Cancer Res 1999;59: Simpson PT, Gale T, Reis-Filho JS, Jones C, Parry S, Sloane JP, et al. Columnar cell lesions of the breast: the missing link in breast cancer progression? A morphological and molecular analysis. Am J Surg Pathol 2005;29: Simpson PT, Reis-Filho JS, Gale T, Lakhani SR. Molecular evolution of breast cancer. J Pathol 2005;205: Soumian S, Soumian S, Verghese ET, Verghese ET, Booth M, et al. Concordance between vacuum assisted biopsy and postoperative histology: implications for the proposed low risk DCIS Trial (LORIS). Eur J Surg Oncol 2013;39: Walsh PC, DeWeese TL, Eisenberger MA. Clinical practice. Localized prostate cancer. N Engl J Med 2007;357: Finding the balance between overtreatment and undertreatment of ductal carcinoma in situ 43

46 105. Bill-Axelson A, Holmberg L, Garmo H, Rider JR, Taari K, Busch C, et al. Radical prostatectomy or watchful waiting in early prostate Cancer. N Engl J Med 2014;370: Barnes NLP, Boland GP, Davenport A, Knox WF, Bundred NJ. Relationship between hormone receptor status and tumour size, grade and comedo necrosis in ductal carcinoma in situ. Br J Surg 2005;92: Burkhardt L, Grob TJ, Hermann I, Burandt E, Choschzick M, Janicke F, et al. Gene amplification in ductal carcinoma in situ of the breast. Breast Cancer Res Treat 2009;123: Witkiewicz AK, Cox DW, Rivadeneira D, Ertel AE, Fortina P, Schwartz GF, et al. The retinoblastoma tumor suppressor pathway modulates the invasiveness of ErbB2-positive breast cancer. Oncogene 2014;33: McNamara KM, Yoda T, Nurani AM, Shibahara Y, Miki Y, Wang L, et al. Androgenic pathways in the progression of triple-negative breast carcinoma: a comparison between aggressive and nonaggressive subtypes. Breast Cancer Res Treat 2014;145: Hieken TJ, Cheregi J, Farolan M, Kim J, Velasco JM. Predicting relapse in ductal carcinoma in situ patients: an analysis of biologic markers with long-term follow-up. Am J Surg 2007;194: Liu T, Niu Y, Feng Y, Niu R, Yu Y, Lv A, et al. Methylation of CpG islands of p16ink4a and cyclind1 overexpression associated with progression of intraductal proliferative lesions of the breast. Hum Pathol 2008;39: Altintas S, Lambein K, Huizing MT, Braems G, Asjoe FT, Hellemans H, et al. Prognostic significance of oncogenic markers in ductal carcinoma in situ of the breast: a clinicopathologic study. Breast J 2009;15: Knudsen ES, Dervishaj O, Kleer CG, Pajak T, Schwartz GF, Witkiewicz AK. EZH2 and ALDH1 expression in ductal carcinoma in situ: complex association with recurrence and progression to invasive breast cancer. Cell 2013;12: Finding the balance between overtreatment and undertreatment of ductal carcinoma in situ

47 Finding the balance between overtreatment and undertreatment of ductal carcinoma in situ 45 2

48 Emma J. Groen Lotte E. Elshof Emiel J.Th. Rutgers Hillegonda A.O. Winter-Warnars Esther H. Lips Jelle Wesseling Nederlands Tijdschrift voor Geneeskunde 2016; 160: A9773

49 CHAPTER 3 Ductaal carcinoma in situ: de balans tussen overbehandeling en onderbehandeling

50 Een 65-jarige vrouw komt op het spreekuur van de huisarts, omdat er bij het bevolkingsonderzoek naar borstkanker calcificaties op het mammogram werden gezien. Bij pathologisch onderzoek waren deze calcificaties gerelateerd aan ductaal carcinoma in situ (DCIS), een afwijking die wordt beschouwd als een potentieel voorstadium van mammacarcinoom. Patiënte zou zelf het liefst de hele borst laten amputeren, zodat ze voor altijd van het DCIS af is. Toch wil ze eerst haar opties en risico s bespreken. In de huidige praktijk krijgen veel vrouwen de diagnose DCIS. Omdat zij een risico hebben op het ontwikkelen van invasief mammacarcinoom worden ze behandeld. Er zijn echter meerdere aanwijzingen dat een deel van de DCIS-afwijkingen nooit invasief zal worden en dat deze patiënten dus overbehandeld worden. Maar juist omdat iedere patiënte met een DCIS-afwijking behandeld wordt en daardoor het natuurlijke beloop niet meer geëvalueerd kan worden, kunnen we op dit moment niet goed voorspellen welke DCIS-afwijking niet zal uitgroeien tot een mammacarcinoom. Wat is DCIS? De klierbuizen van de mamma worden aan de binnenzijde bekleed door luminaal epitheel, met daaromheen een laag myo-epitheliale cellen en een basaalmembraan. DCIS is een proliferatie van maligne, neoplastische luminale epitheelcellen die zich beperkt tot het klierbuissysteem van de borst, dus zonder invasie in het omliggende weefsel (figuur 1). Specifieke morfologische kenmerken, zoals atypie, monotonie, architecturale verstoring en proliferatieve activiteit, maken het mogelijk onderscheid te maken tussen een maligne en een benigne, hyperplastische epitheelproliferatie. Daarnaast wordt DCIS op basis van morfologische kenmerken ingedeeld in goed (graad 1), matig (graad 2) of slecht gedifferentieerd (graad 3). 1 DCIS geeft meestal geen klachten. Soms hebben patiënten een palpabele afwijking, veranderingen aan de tepel, zoals een eczeemachtig beeld (ziekte van Paget), of een heldere of bloederige tepeluitvloed. De diagnose DCIS wordt meestal gesteld op basis van calcificaties op het mammogram. De aanwezigheid van calcificaties duidt echter niet altijd op DCIS, omdat deze calcificaties ook worden gevonden bij een variëteit van benigne veranderingen van het klierweefsel. DCIS komt voor in combinatie met invasief mammacarcinoom meestal met massavorming of zonder mammacarcinoom; deze laatste vorm wordt puur DCIS genoemd. Als het gaat om een pure DCIS-afwijking, is de borstkankerspecifieke mortaliteit nihil, mits er geen invasief recidief optreedt. Als een pure DCIS-afwijking niet behandeld wordt, zal een deel ervan progressie naar invasief mammacarcinoom tonen. 2 De neoplastische proliferatie blijft dan niet langer beperkt tot het klierbuissysteem, maar vertoont invasie van het omringende weefsel, met risico op lymfogene en hematogene metastasering en mogelijk overlijden tot gevolg. 48 Ductaal carcinoma in situ: de balans tussen overbehandeling en onderbehandeling

51 Om progressie naar invasief mammacarcinoom te voorkomen worden patiënten met DCIS primair behandeld met een mammasparende operatie, meestal gevolgd door radiotherapie, of een ablatie. Als het risico op aanwezigheid van invasieve groei verhoogd is, dat wil zeggen: bij uitgebreider, slecht gedifferentieerd DCIS en aankleuring van de afwijking op een MRI-scan, wordt een schildwachtklierprocedure overwogen. De lokale behandeling is hiermee vrijwel identiek aan die voor patiënten met invasief mammacarcinoom stadium I of II. 3 Figuur 1. Ductaal carcinoma in situ (DCIS) dat gerelateerd is aan calcificaties. Links verwijde klierbuizen met DCIS en rechts twee niet-afwijkende klierbuislobjes. Toegenomen incidentie en overdiagnostiek Sinds de implementatie van het bevolkingsonderzoek naar borstkanker in 1989 is de incidentie van DCIS meer dan verzesvoudigd (bron: Nederlandse Kankerregistratie, nl) (figuur 2). Voorheen werd DCIS alleen gevonden bij verwijdering van een massa of als toevalsbevinding. Omdat het bevolkingsonderzoek juist in het leven is geroepen om kanker in een vroeg stadium op te sporen, lijken deze cijfers te wijzen op een succes. Voor een deel van de DCIS-afwijkingen geldt zeker dat vroege detectie erger kan voorkomen. Er zijn echter sterke aanwijzingen dat screening ook leidt tot overdiagnostiek, dat wil zeggen: er wordt een aandoening gediagnosticeerd die niet geleid zou hebben tot klachten of overlijden als er geen screening zou zijn verricht. 3 Figuur 3 illustreert het heterogene beloop van kanker, inclusief de voorstadia. 3 Ductaal carcinoma in situ: de balans tussen overbehandeling en onderbehandeling 49

52 Figuur 2. Incidentie van DCIS (geel) en invasief mammacarcinoom (blauw) bij vrouwen in Nederland. Bron: Nederlandse Kanker Registratie, De stijging van de incidentie van DCIS gaat niet gepaard met een afname van de incidentie van invasief mammacarcinoom (bron: Nederlandse Kankerregistratie, www. cijfersoverkanker.nl) (zie figuur 2). Daarnaast wordt bij obducties DCIS aangetroffen bij 10-39% van de vrouwen in de screeningsleeftijd, zonder dat zij kanker gerelateerde ziekte hadden bij leven. 4 Dit suggereert dat er een groot DCIS-reservoir bestaat zonder nadelig effect op de gezondheid. Ondanks de aanwijzingen dat overdiagnostiek een reëel probleem vormt en dat puur DCIS nooit fataal is, ervaren patiënten bij wie de diagnose DCIS wordt gesteld, evenveel angst als patiënten die zijn gediagnosticeerd met invasief mammacarcinoom. 5,6 Bovenstaande aspecten leiden tot het DCIS-dilemma: welke vrouwen hebben daadwerkelijk baat bij intensieve behandeling en welke niet? Ofwel: bij wie is er sprake van overdiagnostiek? Daarvoor is het essentieel onderscheid te maken tussen DCIS met een hoog en met een laag risico op het ontwikkelen van invasief mammacarcinoom. Hoe goed kunnen we dat nu? En hoe kan dat beter? 50 Ductaal carcinoma in situ: de balans tussen overbehandeling en onderbehandeling

53 3 Figuur 3. Heterogeniteit van kankerprogressie. De pijl met snel geeft een snel groeiende maligniteit weer, die snel tot klachten en overlijden leidt. De pijl met langzaam geeft een langzaam groeiende maligniteit weer, die pas na vele jaren zal leiden tot klachten en overlijden. De pijl met heel langzaam geeft een maligniteit weer die nooit problemen zal geven; de patiënt zal eerder overlijden door een andere oorzaak voordat de maligniteit groot genoeg is om klachten te geven. De pijl met niet progressief geeft een maligniteit weer die nooit zal groeien of die groeit maar daarna regressie zal tonen (gestreepte lijn); beide vormen zullen geen klachten geven. 3 Risico op invasief mammacarcinoom In een grote Nederlandse retrospectieve studie ontwikkelde 5,8% van de vrouwen met DCIS die een mammasparende behandeling of mastectomie hadden ondergaan, een ipsilateraal invasief recidief na een mediane follow-upduur van 10 jaar (Hoofdstuk 4). Belangrijk is hierbij te vermelden dat invasief mammacarcinoom dat zich ontwikkelt uit laaggradig DCIS altijd laaggradig en hormoonreceptorpositief zal zijn en een gering risico op metastasering heeft. Daardoor hebben deze patiënten een uitstekende prognose. Vrouwen jonger dan 35 jaar en vrouwen van Afro-Amerikaanse afkomst hebben wel een hoger risico op een invasief recidief en op overlijden. 7 Adjuvante radiotherapie na een radicale borstsparende behandeling zorgt voor een relatieve reductie van het aantal lokale recidieven na 10 jaar van 50%, waarvan de helft een invasief carcinoom betreft. 8 Radiotherapie heeft echter geen effect op de borstkanker specifieke mortaliteit na 10 jaar. 7 Ductaal carcinoma in situ: de balans tussen overbehandeling en onderbehandeling 51

54 Voorspellende markers Er zijn verschillende markers beschreven die progressie van DCIS naar invasief mammacarcinoom kunnen voorspellen. Maar deze zijn onvoldoende betrouwbaar om toe te kunnen passen in de praktijk. Histopathologisch Meerdere studies beschrijven een verhoogd risico op het ontwikkelen van een ipsilateraal recidief van een in-situcarcinoom of een invasief carcinoom bij patiënten met DCIS van een hogere graad. In een niet-gerandomiseerde prospectieve studie van 670 patiënten die alleen radicale excisie ondergingen was dit risico 10,5% voor patiënten met DCIS graad 1-2 vergeleken met 18% voor patiënten met DCIS graad 3 na een mediane follow-up duur van 6,3 jaar. 9 Een complicerende factor is dat met name laaggradige afwijkingen door interbeoordelaarsvariatie niet uniform en reproduceerbaar geclassificeerd worden, blijkbaar omdat de morfologische criteria onvoldoende eenduidig zijn. 10,11 Op dit moment heeft de differentiatiegraad geen invloed op het beleid, behalve voor het stellen van de indicatie voor een schildwachtklierprocedure. Immuunhistochemisch De bekendste immuunhistochemische markers, de oestrogeen- en progesteronreceptor, zijn gerelateerd aan DCIS van een lagere graad en aan een lager risico op een invasief mammacarcinoom of een recidief van DCIS. Overexpressie van HER2 daarentegen gaat gepaard met een hoger risico. 12,13 Expressie van p16 en p53 en de mate van heterogeniteit van de DCIS afwijking bij een patiënte zijn ook gerelateerd aan een hoger recidiefrisico. 14 Moleculair Er zijn genexpressieprofielen beschreven die gerelateerd zijn aan agressief biologisch gedrag van de tumor. 15 Deze bevindingen zijn echter onvoldoende gevalideerd om toe te kunnen passen in de praktijk. Beperkingen van huidig onderzoek Er zijn dus geen betrouwbare prognostische factoren die progressie kunnen voorspellen. Zoals gezegd leidt deze onzekerheid tot het behandelen van vrijwel iedere patiente met een DCISafwijking, waardoor het natuurlijke beloop niet meer te evalueren is. Er zijn dan ook vooral studies verricht waarin het risico op een recidief dat wil zeggen: een recidief van DCIS of een invasief mammacarcinoom na behandeling werd onderzocht. Robuuste observationele studies naar een beleid van actieve surveillance bij patiënten met aangetoond DCIS zijn er niet. De spaarzame retrospectieve studies die de follow-up van slechts een gering aantal onbehandelde patiënten met puur DCIS beschrijven, geven meestal een vertekend of onvolledig beeld. Zo ontbreken vaak radiologische gegevens, waardoor correlatie met histopathologische bevindingen niet mogelijk is. Ook is er veelal geen informatie over of het symptomatisch DCIS betreft of niet. Dit 52 Ductaal carcinoma in situ: de balans tussen overbehandeling en onderbehandeling

55 is relevant, want vrouwen met symptomatisch DCIS zouden een hoger risico hebben op invasief mammacarcinoom dan vrouwen met DCIS dat tijdens screening is gedetecteerd. 12 Vaak wordt gewezen op de kans van het missen van invasief carcinoom in een biopt als daarin slechts DCIS wordt aangetroffen. Adequate klinische en radiologische correlatie speelt hierbij een rol. Dat wil zeggen dat de klinische, radiologische en histopathologische bevindingen bij elkaar moeten passen en dat bij twijfel aanvullend onderzoek gedaan moet worden. Het is belangrijk te realiseren dat de kans van het missen van invasief carcinoom bij beeldvormend onderzoek of biopsie groter is naarmate de tumor kleiner is. 16 Een vertraging van bijvoorbeeld een jaar in de detectie van goed gedifferentieerd mammacarcinoom zal de overleving echter niet nadelig beïnvloeden, omdat de toename in omvang van deze carcinomen gering is. 17 Daarnaast zijn studies naar prognostische markers vrijwel altijd gericht op DCIS-afwijkingen met synchroon invasief mammacarcinoom, dus geen puur DCIS. Deze studies zijn methodologisch dan ook niet geschikt om markers te identificeren die de kans kunnen bepalen van het invasief worden van een bepaalde DCIS-afwijking in de toekomst. Tot slot is voor het aantonen van een significant verschil in het risico op een invasief recidief een lange follow-up tijd nodig, omdat het meer dan 10 jaar kan duren voordat een invasief carcinoom zich ontwikkelt. 18 Dit maakt het lastig om enerzijds alle andere verstorende variabelen constant te houden binnen de onderzoeksperiode, zoals de behandeling en detectiemethoden, en anderzijds de benodigde gegevens over een lange periode op een uniforme wijze te verkrijgen. 3 Behandeling en communicatie Het Amerikaanse Patient-Centered Outcomes Research Institute heeft een studie laten uitvoeren naar het prioriteren van onderzoek naar DCIS. Volgens een divers expertpanel, dat onder anderen bestaat uit patiënten, clinici, onderzoekers en beleidsmakers, heeft risicostratificatie de hoogste prioriteit. 19 Als strategie wordt het integreren van klinische, radiologische en moleculaire gegevens voor de ontwikkeling van een individueel toepasbaar risicopredictie model opgeworpen. Ook adviseert het panel een RCT voor het vergelijken van de veiligheid en effectiviteit van een actiefsurveillancebeleid met die van standaardtherapie. Op dit moment worden deze beide onderzoeksstrategieën uitgevoerd vanuit het AVL-NKI in samenwerking met andere ziekenhuizen binnen en buiten Nederland. Binnenkort zal de grote internationale LORD-studie (LORD staat voor LOw Risk Dcis; Hoofdstuk 7) beginnen in Europa onder auspiciën van de European Organisation for Research and Treatment of Cancer, waarbij de Borstkanker Onderzoek Groep de trial in Nederland coördineert. In deze studie zullen vrouwen met puur laaggradig DCIS, dat is gedetecteerd bij screening op basis van calcificaties, worden gerandomiseerd tussen een actief-surveillancebeleid en standaardtherapie (figuur 4). 20 Na inclusie zullen deze patiënten 10 jaar gevolgd worden en zal het risico op het Ductaal carcinoma in situ: de balans tussen overbehandeling en onderbehandeling 53

56 ontwikkelen van invasief mammacarcinoom bepaald worden. Als een recidief zich voordoet, is mammasparende therapie met radiotherapie nog een optie. Voor patiënten met een recidief na standaardbehandeling daarentegen is ablatie meestal de enige keuze. Een soortgelijke studie, de LORIS-studie (LORIS staat voor LOw RISk dcis ), is in het Verenigd Koninkrijk al in 2014 begonnen met het includeren van vrouwen met laagrisico-dcis. Complementair aan deze prospectieve studies worden in een groot Nederlands retrospectief onderzoek klinische, radiologische en moleculaire gegevens geïntegreerd en vergeleken van vrouwen met DCIS die wel of niet een ipsilateraal invasief recidief hebben ontwikkeld na een mammasparende behandeling (n = ). De uitstekende registratie in Nederland bij de Nederlandse Kankerregistratie, het bevolkingsonderzoek naar borstkanker en het Pathologisch- Anatomisch Landelijk Geautomatiseerd Archief (PALGA) is uniek in de wereld en maakt betrouwbare en complete gegevensverzameling mogelijk. Figuur 4. Stroomschema van de onderzoeksopzet van de LOw Risk Dcis (LORD)-trial. 54 Ductaal carcinoma in situ: de balans tussen overbehandeling en onderbehandeling

57 Ook het verbeteren van de communicatie over de diagnose en prognose van DCIS aan patiënten wordt aangemerkt als prioriteit. 19,21 Er bestaat veel onzekerheid over de lange termijn-implicaties van de diagnose DCIS, onder andere over het risico op invasief mammacarcinoom en over de therapeutische effectiviteit en veiligheid. Dit maakt het voor patiënten lastig om besluiten te nemen bij behandelkeuzes. Het is voor een vrouw bijvoorbeeld moeilijk te begrijpen dat zij een voorstadium van borstkanker heeft dus geen echte borstkanker maar dat een intensieve behandeling wel nodig wordt geacht. Het beter inschatten van de risico s en het in perspectief plaatsen daarvan is essentieel, waarbij de kwaliteit van leven en concurrerende factoren wat betreft morbiditeit en mortaliteit in ogenschouw worden genomen. Voorlichting aan zorgverleners en de ontwikkeling van een predictiemodel om het individuele risico op progressie te bepalen zullen hieraan bijdragen. 3 Epicrise Voor de huisarts van de patiënte in de casus aan het begin van dit artikel is het van belang haar uit te leggen dat ze een in-situ carcinoom heeft en geen invasief carcinoom, en dat er op dit moment geen sprake is van risico op uitzaaiingen en op overlijden. Daarnaast moet de huisarts haar vertellen dat de behandeling niet primair is gericht op het verwijderen van DCIS maar op het voorkomen van invasief mammacarcinoom. Omdat we momenteel niet goed weten bij wie eventuele progressie naar een invasief carcinoom zal optreden, worden alle patiënten met DCIS lokaal behandeld. Conclusie Door het bevolkingsonderzoek naar borstkanker is de incidentie van DCIS substantieel toegenomen. De rationale van behandeling van patiënten met DCIS is reductie van de mortaliteit door invasief mammacarcinoom. Puur DCIS geeft meestal geen klachten en leidt niet tot overlijden. Een deel van de DCIS afwijkingen zal nooit of pas laat leiden tot invasief mammacarcinoom, maar het is onbekend welke afwijking wél en welke geen progressie zal tonen. Daarom wordt iedere patiënte met een DCIS-afwijking uniform lokaal behandeld alsof zij invasief mammacarcinoom heeft. Om therapie op maat te kunnen aanbieden is risicostratificatie noodzakelijk. Deze aanpak wordt momenteel onderzocht in grote internationale studies en zal naar verwachting leiden tot de ontwikkeling van een individueel toepasbaar risicopredictiemodel. Juiste en genuanceerde communicatie over de implicaties van de diagnose DCIS is essentieel voor een reële risicoperceptie bij en optimale besluitvorming door de patiënte en de betrokken zorgprofessionals. Ductaal carcinoma in situ: de balans tussen overbehandeling en onderbehandeling 55

58 Literatuur 1. Holland R, Peterse JL, Millis RR, et al. Ductal carcinoma in situ: a proposal for a new classification. Semin Diagn Pathol. 1994;11: Lopez-Garcia MA, Geyer FC, Lacroix-Triki M, Marchió C, Reis-Filho JS. Breast cancer precursors revisited: molecular features and progression pathways. Histopathology. 2010;57: Welch HG, Black WC. Overdiagnosis in cancer. J Natl Cancer Inst. 2010;102: Welch HG, Black WC. Using autopsy series to estimate the disease reservoir for ductal carcinoma in situ of the breast: how much more breast cancer can we find? Ann Intern Med. 1997;127: Partridge A, Adloff K, Blood E, et al. Risk perceptions and psychosocial outcomes of women with ductal carcinoma in situ: longitudinal results from a cohort study. J Natl Cancer Inst. 2008;100: Van Gestel YRBM, Voogd AC, Vingerhoets AJJM, et al. A comparison of quality of life, disease impact and risk perception in women with invasive breast cancer and ductal carcinoma in situ. Eur J Cancer. 2007;43: Narod SA, Iqbal J, Giannakeas V, Sopik V, Sun P. Breast Cancer Mortality After a Diagnosis of Ductal Carcinoma In Situ. JAMA Oncol. 2015;1: Early Breast Cancer Trialists Collaborative Group (EBCTCG). Overview of the randomized trials of radiotherapy in ductal carcinoma in situ of the breast. JNCI Monographs. 2010;2010: Hughes LL, Wang M, Page DL, et al. Local excision alone without irradiation for ductal carcinoma in situ of the breast: a trial of the Eastern Cooperative Oncology Group. J Clin Oncol. 2009;27: Elmore JG, Longton GM, Carney PA, et al. Diagnostic concordance among pathologists interpreting breast biopsy specimens. JAMA. 2015;313: Elston CW, Sloane JP, Amendoeira I, et al; European Commission Working Group on Breast Screening Pathology. Causes of inconsistency in diagnosing and classifying intraductal proliferations of the breast. Eur J Cancer. 2000;36: Kerlikowske K, Molinaro AM, Gauthier ML, et al. Biomarker expression and risk of subsequent tumors after initial ductal carcinoma in situ diagnosis. J Natl Cancer Inst. 2010;102: Han K, Nofech-Mozes S, Narod S, et al. Expression of HER2neu in ductal carcinoma in situ is associated with local recurrence. Clin Oncol (R Coll Radiol). 2012;24: Pape-Zambito D, Jiang Z, Wu H, et al. Identifying a highly-aggressive DCIS subgroup by studying intra-individual DCIS heterogeneity among invasive breast cancer patients. PLOS ONE. 2014;9:e e Carraro DM, Elias EV, Andrade VP. Ductal carcinoma in situ of the breast: morphological and molecular features implicated in progression. Biosci Rep. 2014;34:e Weedon-Fekjaer H, Lindqvist BH, Vatten LJ, Aalen OO, Tretli S. Breast cancer tumor growth estimated through mammography screening data. Breast Cancer Res. 2008;10:R Spratt JA, von Fournier D, Spratt JS, Weber EE. Mammographic assessment of human breast cancer growth and duration. Cancer. 1993;71: Sanders ME, Schuyler PA, Simpson JF, Page DL, Dupont WD. Continued observation of the natural history of low-grade ductal carcinoma in situ reaffirms proclivity for local recurrence even after more than 30 years of follow-up. Mod Pathol. 2015;28: Gierisch JM, Myers ER, Schmit KM, et al. Prioritization of research addressing management strategies for ductal carcinoma in situ. Ann Intern Med. 2014;160: Ductaal carcinoma in situ: de balans tussen overbehandeling en onderbehandeling

59 20. Elshof LE, Tryfonidis K, Slaets L, et al. Feasibility of a prospective, randomised, open-label, international multicentre, phase III, noninferiority trial to assess the safety of active surveillance for low risk ductal carcinoma in situ The LORD study. Eur J Cancer. 2015;51: Ganz PA. Quality-of-life issues in patients with ductal carcinoma in situ. JNCI Monographs. 2010;2010: Ductaal carcinoma in situ: de balans tussen overbehandeling en onderbehandeling 57

60 Lotte E. Elshof Michael Schaapveld Marjanka K. Schmidt Emiel J. Rutgers Flora E. van Leeuwen Jelle Wesseling Breast Cancer Research and Treatment 2016; 159:

61 CHAPTER 4 Subsequent risk of ipsilateral and contralateral invasive breast cancer after treatment for ductal carcinoma in situ: Incidence and the effect of radiotherapy in a population-based cohort of 10,090 women

62 Abstract Purpose To assess the effect of different treatment strategies on the risk of subsequent invasive breast cancer (IBC) in women diagnosed with ductal carcinoma in situ (DCIS). Methods Up to 15-year cumulative incidences of ipsilateral IBC (iibc) and contralateral IBC (cibc) were assessed among a population-based cohort of 10,090 women treated for DCIS in the Netherlands between 1989 and Multivariable Cox regression analyses were used to evaluate associations of treatment with iibc risk. Results Fifteen years after DCIS diagnosis, cumulative incidence of iibc was 1.9 % after mastectomy, 8.8 % after breast-conserving surgery (BCS) and radiotherapy (RT), and 15.4 % after BCS alone. Patients treated with BCS alone had a higher iibc risk than those treated with BCS+RT during the first 5 years after treatment. This difference was less pronounced for patients <50 years [hazard ratio (HR) 2.11, 95 % confidence interval (CI) for women <50, and HR 4.44, 95 % CI for women 50, P interaction <0.0001]. Beyond 5 years of follow-up, iibc risk did not differ between patients treated with BCS+RT or BCS alone for women <50. Cumulative incidence of cibc at 15 years was 6.4 %, compared to 3.4 % in the general population. Conclusions We report an interaction of treatment with age and follow-up period on iibc risk, indicating that the benefit of RT seems to be smaller among younger women, and stressing the importance of clinical studies with long follow-up. Finally, the low cibc risk does not justify contralateral prophylactic mastectomies for many women with unilateral DCIS. 60 Subsequent ipsilateral and contralateral invasive breast cancer

63 Introduction Ductal carcinoma in situ (DCIS) is a potential precursor lesion of invasive breast cancer (IBC) [1]. Most women (80 85 %) diagnosed with DCIS present with a mammographic abnormality without clinical symptoms [2]. Since the introduction of population-based mammographic screening and, more recently, digital mammography, the incidence of DCIS has increased substantially [3 7]. In the Netherlands, the European standardized rate of in situ breast carcinoma of which DCIS is the most common type ( 80 %) increased fivefold since 1989, up to 25.1 per 100,000 women in 2013 [8]. In the United States, the incidence (age adjusted to the 2000 US standard population) increased even more: from 5.8 per 100,000 in 1975 to 33.8 per 100,000 women in 2010 [9]. The natural course of DCIS is not well known because DCIS has almost always been treated by mastectomy or breast-conserving surgery (BCS) with or without radiotherapy (RT). Between 1988 and 2011, only 2 % of women with DCIS were managed without surgery in the United States [10]. In the Netherlands, the percentage of non-operated DCIS between 1989 and 2004 was 0.8 % [11]. Women with DCIS are treated to prevent the development of IBC, assuming that this may lead to a reduction in breast cancer-specific deaths. Some women with unilateral DCIS even undergo contralateral prophylactic mastectomy. However, the long-term benefit of treating asymptomatic DCIS that may or may not progress to IBC is difficult to quantify [12]. Therefore, screening programs are criticized to be associated with overdiagnosis and resultant overtreatment of DCIS [13, 14]. Considerable uncertainty remains about the likelihood that a treatment strategy will prevent IBC, whether that likelihood will change based on specific patient and DCIS characteristics, and whether the reduction in risk is enough to justify the costs and the potential side effects of that treatment [12]. The effect of different treatment strategies on the risk of subsequent events in women diagnosed with DCIS has been addressed previously in both prospective trials and observational studies [15 27]. However, many of these studies focused on local recurrences, not discriminating between invasive and non-invasive events, or did not have complete information on treatment. Moreover, several studies have analyzed specific subgroups, such as favorable and good-risk DCIS, or focused on a specific treatment strategy. Gierisch et al. prioritized research needs for DCIS patients, and pointed out the assessment of the effect of treatment strategies on IBC, using existing observational data [12]. We assessed the effect of DCIS treatment strategies on risk of subsequent ipsilateral invasive breast cancer (iibc) using a large population-based cohort with complete information on treatment and followup. In addition, we analyzed the risk of contralateral IBC (cibc). 4 Subsequent ipsilateral and contralateral invasive breast cancer 61

64 Methods Patient selection All women diagnosed with breast carcinoma in situ in the Netherlands between January 1st 1989 and December 31st 2004 were selected from the Netherlands cancer registry (NCR) managed by the Netherlands Comprehensive Cancer Organization. Patients with previous malignancies, except for non-melanoma skin cancer, were not included. This cohort (n = 12,717) was linked to the nationwide network and registry of histology and cytopathology in the Netherlands (PALGA) [28]. The selection criteria for this study were a diagnosis of pure DCIS, i.e., no lobular or other subtype component, and only treated by surgery with or without RT. See Fig. 1 for a detailed list of the excluded cases (n = 2627). The study was approved by the review boards of the NCR and PALGA. DCIS treatment and other characteristics Information on treatment, age, date of diagnosis, and grade was derived from data provided by NCR. Guidelines for DCIS treatment in the Netherlands recommend mastectomy or BCS, consisting of microscopic complete tumor excision. From 1999, the addition of RT after BCS is included in the recommendation. Adjuvant (hormonal) treatment is not recommended. Primary DCIS treatment was categorized into (1) BCS+RT; (2) BCS alone; and (3) mastectomy. Initial treatment was defined as the final treatment for the ipsilateral breast within 3 months after DCIS diagnosis. For patients for whom surgery type was not coded by NCR, we retrieved this information from PALGA. We validated whether patients registered by NCR as treated with BCS had indeed undergone BCS using the conclusions of pathology reports within 3 months of DCIS diagnosis. Furthermore, we validated surgical treatment for women who developed subsequent iibc after mastectomy, using conclusion texts of all available pathology reports. Subsequently, we assessed whether women initially treated with BCS had undergone ipsilateral mastectomy during follow-up, using both NCR and PALGA data. Based on the gradual implementation of the national breast cancer screening program, we categorized year of DCIS diagnosis into two periods: (implementation phase) and (full coverage). Age was subdivided into two groups: <50 and 50 years. Grade was available for 53 % of the entire cohort. The grading system used in the Netherlands is based on the classification presented by Holland et al. [29]. 62 Subsequent ipsilateral and contralateral invasive breast cancer

65 4 Figure 1. Flow diagram for patient selection and median follow-up by initial treatment type. iibc ipsilateral invasive breast cancer, cibc contralateral invasive breast cancer. Follow-up data The occurrence of iibc and cibc was ascertained based on NCR data, and additionally, for patients treated with BCS, through evaluating pathology reports. Follow-up for subsequent IBC and vital status were complete until at least January 1, Statistical analyses Time at risk started at date of DCIS diagnosis and stopped at date of diagnosis of the event of interest (iibc or cibc), date of death or emigration, or January 1, 2011, whichever came first. We calculated cumulative incidence of iibc and cibc using death as competing risk. P values were based on competing risk regression [30], with time since DCIS diagnosis as time-scale Subsequent ipsilateral and contralateral invasive breast cancer 63

66 and adjusted for age (continuous). Further, we compared cumulative incidence of cibc with the expected cumulative incidence of IBC in the general population. Expected cumulative incidence was derived from ageand period-specific cancer incidence and overall mortality in the Dutch female population, estimated using the conditional method [31]. Cox proportional hazards analyses, using age as primary time-scale and time since DCIS diagnosis as secondary time-scale (0 5, 5 10, and 10 years), were used to quantify the effects of different treatments on iibc and cibc risks. Period of DCIS diagnosis and age group at DCIS diagnosis were added as covariables. Proportional hazard assumptions were verified using graphical and residual-based methods. To examine whether iibc risk differed by grade, we performed a subgroup analysis for women with a reported grade. Because the proportion of women with missing data on grade was more than 30 % up to 1998, we performed this subgroup analysis for women diagnosed between 1999 and Surgical treatment was either analyzed as initial DCIS treatment (cumulative incidence) or as a time-varying variable including subsequent mastectomies (Cox regression analysis). All statistical analyses were performed using STATA/ SE 13.1 (StataCorp LP, College Station, TX). A two-sided P value less than 0.05 was considered statistically significant. Results Patient characteristics Analyses included 10,090 women (Fig. 1), of whom 7931 (79 %) women were 50 years at DCIS diagnosis. Median age at DCIS diagnosis was 57.6 years (interquartile range years). Median follow-up was 10.7 years (interquartile range years). During follow-up, 1856 patients died. Table 1 shows characteristics, events and follow-up of the study population by treatment group. DCIS treatment Nearly 48 % (n = 4820) of DCIS patients were initially treated with mastectomy. Of all 5270 women initially treated with BCS, 50 % additionally received RT. Use of BCS increased over time in women <50 years (P trend = 0.010) and 50 years (P trend <0.001). The use of RT after BCS also increased over time in both groups (P trend <0.001) (Fig. 2). Fifteen years after initial DCIS treatment, cumulative incidence of subsequent ipsilateral mastectomy was 5.2 % in the BCS+RT group, versus 12.0 % in the BCS-alone group. Ipsilateral invasive breast cancer During follow-up, 588 women developed an iibc. The median time to iibc was 5.8 years (interquartile range years). Fifteen years after DCIS diagnosis, cumulative incidence of 64 Subsequent ipsilateral and contralateral invasive breast cancer

67 iibc was 1.9 % [95 % confidence interval (95 % CI) %] after mastectomy, 8.8 % (95 % CI %) after BCS+RT, and 15.4 % (95 % CI %) after BCS alone. When assessing the risk of iibc by treatment, the proportional hazards assumption was violated. We accounted for time dependency in the treatment effect by addition of an interaction term that involved time and treatment to the model (P interaction <0.001). Additionally, we found that the effect of treatment was different depending on age group (P interaction <0.0001). An extra interaction term that involved period of diagnosis and treatment was not significant (P interaction = 0.445). Therefore, Table 2 presents the effect of treatment on iibc risk by follow-up interval and age group. Women diagnosed with DCIS between 1999 and 2004 were less likely to develop iibc than women diagnosed between 1989 and 1998, regardless of treatment and age [hazard ratio (HR) 0.72, 95 % CI ]. After adjusting for treatment and period, women 50 years had lower iibc risk than <50 women years (HR 0.38, 95 % CI ). Figure 3 shows the cumulative incidence of iibc by treatment strategy stratified by period of DCIS diagnosis and age group at DCIS diagnosis. Both women <50 and 50 years treated with BCS alone had a higher risk of developing iibc than women treated with BCS+RT in the first 5 years after DCIS treatment. However, for women 50 years, the difference in iibc risk after BCS alone compared to BCS+RT was much larger than for women <50 years (HR 2.11, 95 % CI for women <50 years and HR 4.44, 95 % CI for women 50 years). While among patients <50 years at DCIS diagnosis, risk of iibc no longer differed after 5 years following BCS+RT or BCS alone (HR 1.01, 95 % CI for 5 10 years follow-up and HR 0.78, 95 % CI for 10 years follow-up), for women 50 years, iibc risk remained increased after BCS alone during subsequent follow-up intervals, although the difference in risks was smaller than in the first 5 years (HR 1.64, 95 % CI for 10 years follow-up). A trend in the proportional reduction with age was found when the data were subdivided into three groups according to age: <45, 45 55, and 55 years (data not shown). Women undergoing mastectomy were less likely to develop iibc compared to women undergoing BCS (Table 2). The highest absolute iibc risk after mastectomy was seen for women <50 years treated between 1989 and 1998 (10-year cumulative incidence: 2.9 %, 95 % CI %). For women 50 years diagnosed from 1999 to 2004 and treated with mastectomy, the 10-year cumulative incidence was lowest at 0.6 % (95 % CI %). In a subgroup analysis of women diagnosed with DCIS between 1999 and 2004, the Cox model including grade was comparable to the main model (data not shown). The difference in iibc risk after BCS alone and BCS+RT was of the same magnitude [e.g., for women 50 years in the first 5 years after DCIS treatment: HR 4.78, 95 % CI (model including grade) vs HR 4.57, 95 % CI (main model)]. Additionally, iibc risk did not differ by grade (adjusted estimate for intermediate vs low grade and high vs low grade: HR 1.25, 95 % CI and HR 1.19, 95 % CI , respectively). 4 Subsequent ipsilateral and contralateral invasive breast cancer 65

68 Table 1. Characteristics of the study population by treatment group Number of DCIS patients (%) Initial DCIS treatment BCS+RT BCS alone Mastectomy Total Age at DCIS diagnosis, years, median (interquartile range) 57.2 ( ) 58.9 ( ) 57.1 ( ) 57.6 ( ) Age at DCIS diagnosis (years) <40 91 (3.5) 108 (4.1) 360 (7.5) 559 (5.5) (14.1) 371 (14.0) 862 (17.9) 1600 (15.9) (41.6) 942 (35.4) 1553 (32.2) 3582 (35.5) (28.3) 785 (29.5) 1245 (25.8) 2769 (27.4) (11.8) 335 (12.6) 630 (13.1) 630 (13.1) (0.8) 117 (4.4) 170 (3.5) 170 (3.5) Period of DCIS diagnosis (implementation phase) 751 (28.8) 1677 (63.1) 2603 (54.0) 5031 (49.9) (full nationwide coverage) 1861 (71.3) 981 (36.9) 2217 (46.0) 5059 (50.1) DCIS grade ( a ) (13.6) 302 (40.8) 190 (10.2) 707 (16.9) (36.7) 235 (31.7) 554 (29.6) 1367 (32.6) (49.7) 204 (27.5) 1128 (60.3) 2115 (50.5) Subsequent ipsilateral mastectomy No 2497 (95.6) 2345 (88.2) NA 9662 (95.8) Yes 115 (4.4) 313 (11.8) NA 428 (4.2) Follow-up interval, years, median (interquartile range) Follow-up interval (years) 9.0 ( ) 12.0 ( ) 11.1 ( ) 10.7 ( ) 0 4 b 101 (3.9) 202 (7.6) 301 (6.2) 604 (6.0) (55.8) 656 (24.7) 1741 (36.1) 3855 (38.2) (40.3) 1800 (67.7) 2778 (57.6) 5631 (55.8) Subsequent invasive breast cancer c No 2351 (90.0) 2167 (81.5) 4501 (93.4) 9019 (89.4) Ipsilateral only 130 (5.0) 336 (12.6) 68 (1.4) 534 (5.3) Contralateral only 122 (4.7) 117 (4.4) 243 (5.0) 482 (4.8) Ipsilateral+contralateral 9 (0.3) 38 (1.4) 7 (0.15) 54 (0.5) Total a Data on grade is presented for cases diagnosed from Grade was not reported in 870 women (17.2 %). b Nine patients with follow-up time = 0 (BCS+RT n = 1, BCS alone n = 2, Mastectomy = 6). c One patient with unknown laterality of subsequent invasive breast cancer. 66 Subsequent ipsilateral and contralateral invasive breast cancer

69 a 150 BCS + RT BCS alone Mastectomy Number of women aged <50 years with DCIS b Year of of DCIS di diagnosis s 500 BCS + RT BCS alone Mastectomy 4 Number of women aged 50 years with DCIS Year of of DCIS di diagnosis s Figure 2. Treatment strategy by year of diagnosis for a women <50 years and b women 50 years. Subsequent ipsilateral and contralateral invasive breast cancer 67

70 a Overall P < Cumulative incidence of ipsilateral invasive breast cancer (%) BCS alone vs BCS+RT P =0.80 Mastectomy vs BCS+RT P <0.001 BCS+RT BCS alone Mastectomy Time (years) Number at risk BCS+RT BCS alone Mastectomy b Overall P< Cumulative incidence of ipsilateral invasive breast cancer (%) BCS alone vs BCS+RT Mastectomy vs BCS+RT BCS+RT BCS alone Mastectomy P<0.001 P< Time (years) Number at risk BCS+RT BCS alone Mastectomy Subsequent ipsilateral and contralateral invasive breast cancer

71 c Overall P=0.001 BCS alone vs BCS+RT P=0.11 Mastectomy vs BCS+RT P=0.05 Cumulative incidence of ipsilateral invasive breast cancer (%) BCS+RT BCS alone Mastectomy Time (years) Number at risk BCS+RT BCS alone Mastectomy d Overall P< Cumulative incidence of ipsilateral invasive breast cancer (%) BCS alone vs BCS+RT Mastectomy vs BCS+RT BCS+RT BCS alone Mastectomy Number at risk BCS+RT BCS alone Mastectomy P<0.001 P< Time (years) Figure 3. Cumulative incidence of iibc by treatment strategy for a women <50 years diagnosed between 1989 and 1998 b women 50 years diagnosed between 1989 and 1998 c women <50 years diagnosed between 1999 and 2004 d women 50 years diagnosed between 1999 and 2004, with death as competing risk. P values based on competing risk regression, adjusted for age (continuous) [30]. Subsequent ipsilateral and contralateral invasive breast cancer 69

72 Table 2. Multivariate Cox regression analysis* for iibc in women treated for DCIS Follow-up time Treatment Total Person-time HR (95 % CI) P value iibc (years) Age group at DCIS diagnosis <50 years 0 5 years BCS+RT Ref BCS alone ( ) Mastectomy ( ) < years BCS+RT Ref BCS alone ( ) 0.95 Mastectomy ( ) <0.001 >10 years BCS+RT Ref BCS alone ( ) 0.37 Mastectomy ( ) < years 0 5 years BCS+RT Ref BCS alone ( ) <0.001 Mastectomy ( ) < years BCS+RT Ref BCS alone ( ) <0.001 Mastectomy ( ) <0.001 >10 years BCS+RT Ref BCS alone ( ) 0.05 Mastectomy ( ) <0.001 Period of DCIS diagnosis Ref ( ) 0.03 Age group at DCIS diagnosis <50 years Ref 50 years ( ) <0.001 *With age as primary time-scale, and treatment as time-varying variable. 70 Subsequent ipsilateral and contralateral invasive breast cancer

73 Contralateral invasive breast cancer Contralateral IBC occurred in 536 women. The median time to cibc was 6.2 years (interquartile range years). Cumulative incidences of cibc at 15 and 20 years after DCIS diagnosis were 6.4 % (95 % CI %) and 8.9 % (95 % CI %), respectively, reaching a rate of % per annum. The risk of cibc did not differ by treatment, period of diagnosis, or age group (Supplemental Table 1). The cumulative risk of cibc is visualized in Fig. 4. The absolute risk of developing cibc in women treated for DCIS was slightly higher than the risk of IBC in the general population (3.4 % at 15 years). Supplemental table 1. Multivariate Cox regression analysis for contralateral invasive breast cancer in women treated for DCIS a Total cibc Person-time, years HR (95% CI) P-value Treatment BCS + RT ref BCS alone ( ) Mastectomy ( ) 0.38 Age group at DCIS diagnosis <50 years ref 50 years ( ) 0.59 Period of DCIS diagnosis ref ( ) 0.10 Follow-up interval 0-5 years ref 5-10 years ( ) years ( ) 0.02 a With age as primary time-scale and treatment as time-varying variable. Subsequent ipsilateral and contralateral invasive breast cancer 71

74 a Overall P=0.96 Cumulative incidence of (contralateral) invasive breast cancer (%) BCS alone vs BCS+RT Mastectomy vs BCS+RT BCS+RT BCS alone Mastectomy P=0.90 P= Time (years) Number at risk BCS+RT BCS only Mastectomy b Overall P=0.59 Cumulative incidence of (contralateral) invasive breast cancer (%) BCS alone vs BCS+RT Mastectomy vs BCS+RT BCS+RT BCS alone Mastectomy P=0.62 P= Time (years) Number at risk BCS+RT BCS only Mastectomy Figure 4. Cumulative incidence of cibc by treatment strategy compared with the expected cumulative incidence of IBC in the general population (dashed line) for a women <50 years, and b women 50 years, with death as competing risk. P values based on competing risk regression, adjusted for age (continuous) [30]. 72 Subsequent ipsilateral and contralateral invasive breast cancer

75 Discussion To the best of our knowledge, this is the largest population-based, nationwide cohort study with accurate and complete long-term outcome data of subsequent invasive breast cancer after DCIS treatment. For women treated with BCS, our study confirms the protective effect of RT with regard to iibc risk shown by randomized controlled trials (RCTs) [23 27, 32]. Importantly, the benefit of RT regarding iibc risk may differ by age and follow-up interval. It appeared that the use of RT after BCS in women <50 years reduced the risk of iibc only in the first years after treatment. In women 50 years, iibc risk remained increased during subsequent follow-up after BCS alone, compared to BCS+RT, but the difference became less pronounced with longer follow-up. Our results suggest that RT is effective in treating microscopic residual disease, but may not prevent de novo IBC in DCIS patients. One of the RCTs also found that the beneficial effect of RT seemed to be restricted to the first 5 years after treatment [24]. Interestingly, the results of our Cox regression analysis point towards less benefit from RT in women <50 years than in older women. This observation could be due to confounding if for example younger women treated with RT were more likely to have DCIS with unfavorable prognostic features. However, a meta-analysis of the RCTs also found age to modify the benefit of RT: women <50 years showed a smaller proportional risk reduction in the rate of local recurrence (either in situ or invasive) than women 50. A trend in the proportional reduction with age was also found when the data were subdivided into five age groups and was independent of histological grade, comedonecrosis, nuclear grade, or architecture [32]. Additionally, we found high iibc risks after BCS either with or without RT in women <50 years. Moreover, these young women treated with mastectomy had a higher cumulative iibc incidence than older women who received this treatment. Prior studies have also reported that local recurrences following mastectomy seem to occur particularly in younger women [33 35]. Data that may explain this higher risk in younger women are limited and inconsistent [35 38]. Despite the increased iibc risk, young age per se should not be considered a contraindication for BCS, especially because breast cancer-specific mortality has not been shown to differ between mastectomy and BCS [32, 39]. Another clinical relevant observation is that the absolute risk of cibc was low with a rate of % per annum. This result is comparable to the population-based study by Falk et al. (n = 3,163; median follow-up 5.2 years) [15]. Despite the low cibc risk, a marked increase in the use of contralateral prophylactic mastectomies among women with DCIS in recent years has been reported [40 42]. Because contralateral prophylactic mastectomies will not likely result in any survival advantage despite the minimization of cibc risk [43] and are not risk-free [43 45], we advocate that prophylactic contralateral mastectomies for DCIS in women without hereditary breast cancer risk should be discouraged. 4 Subsequent ipsilateral and contralateral invasive breast cancer 73

76 One of the strengths of our study was that we differentiated between invasive and non-invasive recurrences. Our 10-year estimates are in line with the 10-year absolute risks reported in other population-based cohort studies and RCTs [15, 17, 32]. However, direct comparison with previous studies, which focused most of their analyses on any local recurrence as outcome, is often difficult. Differences in study design, inclusion criteria, and statistical methods (e.g., cumulative incidence vs Kaplan Meier estimates) may for example play a role. Interestingly, the 10-year cumulative incidence and Kaplan Meier estimates in two, rather small, North American non-randomized prospective studies of women with favorable DCIS treated with BCS alone between 1995 and 2002, were only slightly lower than the 10-year cumulative incidence of iibc for women diagnosed between 1999 and 2004 and treated with BCS alone in our population-based cohort [21, 22]. On the other hand, the estimated 7-year iibc cumulative incidences in a fifth RCT between BCS+RT (n = 287) and BCS alone (n = 298) in a selected good-risk group of women were much lower [23]. Notably, in this RCT in which 62 % of women used tamoxifen, only eight iibcs occurred in the BCS-alone arm, and only one in the BCS+RT arm (median follow-up 7.2 years). The differences in risk estimates could be explained by differences in selection criteria, and utilization of tamoxifen, although the effect of tamoxifen on iibc seems to be minimal [46]. A limitation of our study is the potential of confounding by indication. As the allocation of DCIS treatment was not randomized and the indication for treatment may have been related to the risk of IBC, this could have introduced bias. It is plausible to assume that women with less favorable characteristics more often received adjuvant RT after BCS. Therefore, if confounding by indication plays a role, this will probably have resulted in an underestimation of the difference in iibc risk between BCS+RT and BCS alone. Although grade was associated with treatment strategy in our study, we found that grade was not a confounding factor in our subgroup analysis, as grade was not associated with iibc risk. We did not have information on several other risk factors associated with local recurrence, such as DCIS size and margin status after excision. However, it is still uncertain to what extent these factors are associated with subsequent invasive breast cancer risk [47, 48] and therefore whether these could be confounding factors in our study. A last issue concerns the applicability of our results to today s clinical practice. Our study shows that the risk of developing iibc was lower for women diagnosed between 1999 and 2004 than for women diagnosed between 1989 and 1998, while risk of cibc was similar for both periods. The decrease in iibc risk over the years was independent of treatment strategy and is likely the result of the detection of relatively more harmless DCIS lesions and improvements in preoperative assessment and surgical management. Most likely, the risk found for the latter period reflects the upper boundary of today s risk of iibc in women treated for DCIS, as patient evaluation and selection for treatment have evolved further since Subsequent ipsilateral and contralateral invasive breast cancer

77 It should be emphasized that the women in our cohort were not treated with tamoxifen for DCIS. In the Netherlands, hormonal treatment for DCIS is not recommended and its use is very limited in current clinical practice [49, 50]. A meta-analysis of RCTs assessing the effect of postoperative tamoxifen showed a reduced rate of cibc, but no impact on the risk of iibc or all-cause mortality [46]. The difference in absolute IBC risk between our cohort and a population in which tamoxifen was more common will therefore probably be limited. In summary, our finding that the reduction in iibc risk among women treated with BCS+RT, compared to BCS alone, diminishes with longer follow-up, emphasizes the importance of clinical studies with long-term follow-up. Furthermore, the beneficial effect of RT seems to be smaller among younger women and should be investigated further. Finally, the low risk of cibc does not justify contralateral prophylactic mastectomies for many women with unilateral DCIS. Acknowledgments The authors thank Otto Visser, Annemarie Eeltink and the registration teams of the Netherlands Comprehensive Cancer Organization for the collection of data for the Netherlands Cancer Registry. The authors also thank Lucy Overbeek and PALGA, the nationwide histopathology and cytopathology data network and archive, for providing pathology data. This work was supported by Pink Ribbon (Grant Number 2011.WO19.C88 to J.W.) and the Dutch Cancer Society (Grant Number NKI to M.K.S.). 4 Conflict of Interest LE Elshof, M Schaapveld, MK Schmidt, EJ Rutgers, FE van Leeuwen, and J Wesseling declare that they have no conflict of interest. Subsequent ipsilateral and contralateral invasive breast cancer 75

78 References 1. Van de Vijver MJ, Peterse H (2003) The diagnosis and management of pre-invasive breast disease: pathological diagnosis problems with existing classifications. Breast Cancer Res 5: Bartlett JMS, Nofech-Moses S, Rakovitch E (2014) Ductal carcinoma in situ of the breast: can biomarkers improve current management? Clin Chem 60: Ernster VL, Ballard-Barbash R, Barlow WE et al (2002) Detection of ductal carcinoma in situ in women undergoing screening mammography. J Natl Cancer Inst 94: Sørum R, Hofvind S, Skaane P, Haldorsen T (2010) Trends in incidence of ductal carcinoma in situ: the effect of a population based screening programme. Breast 19: van Steenbergen LN, Voogd AC, Roukema JA et al (2009) Screening caused rising incidence rates of ductal carcinoma in situ of the breast. Breast Cancer Res Treat 115: Vigeland E, Klaasen H, Klingen TA et al (2008) Full-field digital mammography compared to screen film mammography in the prevalent round of a population-based screening programme: the Vestfold County study. Eur Radiol 18: Bluekens AMJ, Holland R, Karssemeijer N et al (2012) Comparison of digital screening mammography and screen-film mammography in the early detection of clinically relevant cancers: a multicenter study. Radiology 265: Netherlands Comprehensive Cancer Organisation (IKNL) Howlader N, Noone AM, Krapcho M, et al. SEER cancer statistics review, , National Cancer Institute 10. Sagara Y, Mallory MA, Wong S et al (2015) Survival benefit of breast surgery for low-grade ductal carcinoma in situ: a population-based cohort study. JAMA Surg 150: doi: / jamasurg Boekel NB, Schaapveld M, Gietema JA et al (2014) Cardiovascular morbidity and mortality after treatment for ductal carcinoma in situ of the breast. J Natl Cancer Inst 106: Gierisch JM, Myers ER, Schmit KM et al (2014) Prioritization of research addressing management strategies for ductal carcinoma in situ. Ann Intern Med 160: Ripping TM, Verbeek ALM, Fracheboud J et al (2015) Overdiagnosis by mammographic screening for breast cancer studied in birth cohorts in The Netherlands. Int J Cancer 137: Harding C, Pompei F, Burmistrov D et al (2015) Breast cancer screening, incidence, and mortality across US counties. JAMA Intern Med. 15. Falk RS, Hofvind S, Skaane P, Haldorsen T (2011) Second events following ductal carcinoma in situ of the breast: a register-based cohort study. Breast Cancer Res Treat 129: Sprague BL, McLaughlin V, Hampton JM et al (2013) Disease free survival by treatment after a DCIS diagnosis in a population based cohort study. Breast Cancer Res Treat 141: Rakovitch E, Nofech-Mozes S, Narod SA et al (2013) Can we select individuals with low risk ductal carcinoma in situ (DCIS)? A population-based outcomes analysis. Breast Cancer Res Treat 138: Schouten van der Velden AP, van Vugt R, Van Dijck JAAM et al(2007) Local recurrences after different treatment strategies for ductal carcinoma in situ of the breast: a population-based study in the East Netherlands. Int J Radiat Oncol Biol Phys 69: Dick AW, Sorbero MS, Ahrendt GM et al (2011) Comparative effectiveness of ductal carcinoma in situ management and the roles of margins and surgeons. J Natl Cancer Inst 103: Subsequent ipsilateral and contralateral invasive breast cancer

79 20. Kane RL, Virnig BA, Shamliyan T et al (2010) The impact of surgery, radiation, and systemic treatment on outcomes in patients with ductal carcinoma in situ. J Natl Cancer Inst Monographs 2010: Wong JS, Chen Y-H, Gadd MA et al (2014) Eight-year update of a prospective study of wide excision alone for small low- or intermediate-grade ductal carcinoma in situ (DCIS). Breast Cancer Res Treat 143: Solin LJ, Gray R, Hughes LL et al (2015) Surgical excision without radiation for ductal carcinoma in situ of the breast: 12-year results from the ECOG-ACRIN E5194 study. J Clin Oncol 33(33): McCormick B, Winter K, Hudis C et al (2015) RTOG 9804: a prospective randomized trial for goodrisk ductal carcinoma in situ comparing radiotherapy with observation. J Clin Oncol 33: Donker M, Litie`re S, Werutsky G et al (2013) Breast-conserving treatment with or without radiotherapy in ductal carcinoma in situ: 15-year recurrence rates and outcome after a recurrence, from the EORTC randomized phase III trial. J Clin Oncol 31: Wapnir IL, Dignam JJ, Fisher B et al (2011) Long-term outcomes of invasive ipsilateral breast tumor recurrences after lumpectomy in NSABP B-17 and B-24 randomized clinical trials for DCIS. J Natl Cancer Inst 103: Wärnberg F, Garmo H, Emdin S et al (2014) Effect of radiotherapy after breast-conserving surgery for ductal carcinoma in situ: 20 years follow-up in the randomized SweDCIS trial. J Clin Oncol 32(32): Cuzick J, Sestak I, Pinder SE et al (2011) Effect of tamoxifen and radiotherapy in women with locally excised ductal carcinoma in situ: long-term results from the UK/ANZ DCIS trial. Lancet Oncol 12: Casparie M, Tiebosch ATMG, Burger G et al (2007) Pathology databanking and biobanking in The Netherlands, a central role for PALGA, the nationwide histopathology and cytopathology data network and archive. Cell Oncol 29: Holland R, Peterse JL, Millis RR et al (1994) Ductal carcinoma in situ: a proposal for a new classification. Semin Diagn Pathol 11: Fine JPG, Gray RJ (1999) A proportional hazards model for the subdistribution of a competing risk. J Am Stat Assoc 94: Ederer F, Heise H (1959) Instructions to Ibm 650 programmers in processing survival computations. Technical end results evaluation section. National Cancer Institute, Bethesda 32. Correa C, McGale P (2010) Overview of the randomized trials of radiotherapy in ductal carcinoma in situ of the breast early breast cancer trialists collaborative group (EBCTCG). J Natl Cancer Inst Monogr 2010: Rashtian A, Iganej S, Amy Liu I-L, Natarajan S (2008) Close or positive margins after mastectomy for DCIS: pattern of relapse and potential indications for radiotherapy. Int J Radiat Oncol Biol Phys 72: Carlson GW, Page A, Johnson E et al (2007) Local recurrence ofductal carcinoma in situ - after skinsparing mastectomy. J Am Coll Surg 204: Vicini FA, Recht A (2002) Age at diagnosis and outcome for women with ductal carcinoma-in situ of the breast: a critical review of the literature. J Clin Oncol 20: Bijker N, van Tienhoven G (2010) Local and systemic outcomes in DCIS based on tumor and patient characteristics: the radiation oncologist s perspective. J Natl Cancer Inst Monographs 2010: Pradier C, Cornuau M, Norca J et al (2011) Differences in breast carcinoma in situ between menopausal and premenopausal women. Anticancer Res 31: Subsequent ipsilateral and contralateral invasive breast cancer 77

80 38. Alvarado R, Lari SA, Roses RE et al (2012) Biology, treatment, and outcome in very young and older women with DCIS. Ann Surg Oncol 19: Narod SA, Iqbal J, Giannakeas V et al (2015) Breast cancer mortality after a diagnosis of ductal carcinoma in situ. JAMA Oncol. 40. Tuttle TM, Jarosek S, Habermann EB et al (2009) Increasing rates of contralateral prophylactic mastectomy among patients with ductal carcinoma in situ. J Clin Oncol 27: Soran A, Kamali Polat A, Johnson R, McGuire KP (2014) Increasing trend of contralateral prophylactic mastectomy: what are the factors behind this phenomenon? Surgeon 12: Rutter CE, Park HS, Killelea BK, Evans SB (2015) Growing use of mastectomy for ductal carcinoma-in situ of the breast among young women in the united states. Ann Surg Oncol 22: Goldflam K, Hunt KK, Gershenwald JE et al (2004) Contralateral prophylactic mastectomy. Predictors of significant histologic findings. Cancer 101: Montgomery LL, Tran KN, Heelan MC et al (1999) Issues of regret in women with contralateral prophylactic mastectomies. Ann Surg Oncol 6: Piot-Ziegler C, Sassi M-L, Raffoul W, Delaloye J-F (2010) Mastectomy, body deconstruction, and impact on identity: a qualitative study. Br J Health Psychol 15: Staley H, McCallum I, Bruce J (2014) Postoperative tamoxifen for ductal carcinoma in situ: cochrane systematic review and meta-analysis. Breast. doi: /j.breast Collins LC, Achacoso N, Haque R et al (2013) Risk factors for non-invasive and invasive local recurrence in patients with ductal carcinoma in situ. Breast Cancer Res Treat 139: Kerlikowske K, Molinaro A, Cha I et al (2003) Characteristics associated with recurrence among women with ductal carcinoma in situ treated by lumpectomy. J Natl Cancer Inst 95: Elshof LE, Tryfonidis K, Slaets L et al (2015) Feasibility of a prospective, randomised, open-label, international multicentre, phase III, non-inferiority trial to assess the safety of active surveillance for low risk ductal carcinoma in situ The LORD study. Eur J Cancer 51: Netherlands Comprehensive Cancer Organisation (IKNL) and the Knowledge institute of Medical Specialists (KiMS) guideline breast cancer (2012) Subsequent ipsilateral and contralateral invasive breast cancer

81 Subsequent ipsilateral and contralateral invasive breast cancer 79 4

82 Lotte E. Elshof Marjanka K. Schmidt Emiel J. Rutgers Flora E. van Leeuwen Jelle Wesseling Michael Schaapveld Annals of Surgery 2017; Epub ahead of print

83 CHAPTER 5 Cause-specific mortality in a population-based cohort of 9799 women treated for ductal carcinoma in situ

84 Abstract Objective To assess cause-specific mortality in women treated for ductal carcinoma in situ (DCIS). Background From screening and treatment perspective, it is relevant to weigh the low breast cancer mortality after DCIS against mortality from other causes and expected mortality in the general population. Methods: We conducted a population-based cohort study comprising 9799 Dutch women treated for primary DCIS between 1989 and 2004 and estimated standardized mortality ratios (SMRs). Results After a median follow up of 9.8 years, 1429 patients had died of whom 284 caused by breast cancer (2.9% of total cohort). DCIS patients <50 years experienced higher mortality compared with women in the general population (SMR 1.7; 95% confidence interval, CI: ), whereas patients >50 had significantly lower mortality (SMR 0.9; 95% CI: ). Overall, the risk of dying from general diseases and cancer other than breast cancer was lower than in the general population, whereas breast cancer mortality was increased. The SMR for breast cancer decreased from 7.5 (95% CI: ) to 2.8 (95% CI: ) for women aged <50 and >50 years, respectively. The cumulative breast cancer mortality 10 years after DCIS was 2.3% for women <50 years and 1.4% for women >50 years treated for DCIS between 1999 and Conclusions DCIS patients >50 years had lower risk of dying from all causes combined compared with the general female population, which may reflect differences in health behavior. Women with DCIS had higher risk of dying from breast cancer than the general population, but absolute 10-year risks were low. 82 Cause-specific mortality

85 Ductal carcinoma in situ (DCIS) is a proliferation of neoplastic cells confined to the ductolobular system that is, without invading the surrounding breast tissue. It is a heterogeneous disease entity ranging from indolent, harmless DCIS to aggressive lesions with high invasive potential. 1 As such, some DCIS lesions may progress into invasive breast cancer (IBC), and ultimately to fatal metastatic disease, whereas many DCIS lesions will never become invasive. 2 The low risk of death from breast cancer among DCIS patients 3,4 may be because of effective treatment, the potentially indolent and slow-growing nature of most DCIS, or both. Most DCIS cases are picked up by breast cancer screening mammograms. 5 Women with screen-detected breast lesions are generally in good health and do not experience any breast changes or symptoms. It is not surprising that inaccurate risk perceptions and anxiety in women diagnosed with DCIS are frequent. 6,7 DCIS patients are generally told that they do not have cancer and have normal life expectancy, but may experience short- and long-term morbidity from the invasive treatment, which is in many respects similar to that for women with IBC. Therefore, the treatment decision-making process in DCIS is complex and controversial. To fine-tune current practice and reduce confusing perceptions, there is an ongoing need to provide accurate information to DCIS patients and their health-care providers about the risks involved. Several population-based studies have studied breast cancer-specific mortality among DCIS patients, 4,8 10 but only a few assessed competing causes of death or comprehensively compared these with that of the general population. 3,11 Therefore, we assessed the likelihood of breast cancer-related death in DCIS patients and compared cause-specific mortality with rates expected based on mortality in the general population. 5 Methods Data Collection and Patient Selection The Netherlands Cancer Registry (NCR) identified all women who were diagnosed with noninvasive breast cancer as first primary neoplasm between 1 January 1989 and 31 December 2004 in the Netherlands. 12 The NCR provided date of birth, diagnosis and death, topography, morphology, grade, stage, type of surgery, and whether radiotherapy, chemotherapy, and/ or hormonal therapy were administered, and data on subsequent neoplasms. Follow up on subsequent malignancies and vital status was complete until at least 1 January Linkage with the nationwide network and registry of histology and cytopathology in the Netherlands (PALGA) was used to validate and complete missing data on surgery type and to exclude patients with Paget disease, which the NCR had not registered as such. 13 From an initial pool of 12,305 women we excluded those who were diagnosed at time of death (n = 4), whose DCIS was not histologically confirmed (n = 21), whose morphology was not pure DCIS (n = 2094), who were diagnosed with an IBC or second breast carcinoma in situ within 4 months of DCIS Cause-specific mortality 83

86 diagnosis (n = 146), who received chemotherapy or hormonal therapy as part of DCIS treatment (not recommended in the Netherlands) (n = 109), and who were not surgically treated or for whom surgery type remained unknown (n = 132). To obtain information on cause of death the cohort was linked with the nationwide cause of death registry at Statistics Netherlands. For analyses, age was categorized or subdivided into 2 groups of <50 and >50 years based on eligibility for the Dutch population-based breast cancer screening program. 14 Between 1989 and 1997, women >69 years, and between 1998 and 2004, women >75 years were not eligible for screening, but these women were added to the 50 years group in our analyses. Initial DCIS treatment was defined as the treatment strategy for the primary DCIS within 3 months of diagnosis and was subdivided into 3 categories: breast-conserving surgery (BCS) alone, BCS plus radiotherapy, and mastectomy. We categorized year of diagnosis into 2 periods, reflecting the gradual implementation of the screening program: 1989 to 1998 (implementation phase) and 1999 to 2004 (implementation completed), and the treatment guideline shift from mastectomy toward BCS plus radiotherapy for screendetected smaller DCIS. Grade was classified according to the method by Holland et al. 15 Information on grade was available for 24% of the women diagnosed between 1989 and 1998 and for 83% of thewomen diagnosed between 1999 and All data were coded and anonymous to the researchers. The study was approved by the review boards of the NCR, PALGA, and Statistics Netherlands. Statistical Analysis To estimate cause-specific excess mortality, we compared observed deaths in the study population with expected number of deaths in the Dutch female population, taking into account the person-years of observation in the study cohort. Expected numbers were calculated based on the corresponding sex-, age-, and calendar period-specific mortality rates in the general Dutch female population provided by Statistics Netherlands. We estimated standardized mortality ratios (SMRs) as ratios of observed and expected numbers of death. Absolute excess mortality (AEM) was calculated as the observed number of deaths minus the number expected, divided by the number of person-years at risk, and multiplied by 10, We stratified results for major causes of death by age group, treatment type, period of diagnosis, occurrence of subsequent IBC, and followup interval. In addition, we estimated SMRs for breast cancer by grade for women diagnosed between 1999 and Tests for homogeneity and trends of SMRs were performed within collapsed person-time Poisson regression models. We evaluated the likelihood of a model with a continuous variable or a variable representing the classes of a categorical variable as discrete value, respectively, against the likelihood of a model without that variable. We estimated the absolute risk of breast cancer mortality using death caused by other causes as a competing event. To quantify the effects of DCIS treatment, age at DCIS diagnosis, period 84 Cause-specific mortality

87 of DCIS diagnosis and DCIS grade on breast cancer mortality, within cohort comparisons were performed using competing risk regression models with death caused by causes other than breast cancer treated as competing risk. 17 We performed univariable and multivariable analyses to calculate unadjusted and adjusted subdistribution hazard ratios (SHR) and 95% confidence intervals (95% CI). In addition, we performed regression analysis by diagnostic period. To evaluate the impact of a subsequent ipsilateral or contralateral IBC on breast cancer mortality we added these as time-dependent variables in a second multivariable-adjusted model. Women who developed IBC contributed person-time to the no subsequent IBC group until IBC diagnosis, and subsequently to the subsequent IBC group. Time at risk started at DCIS diagnosis in all analyses and ended at date of death, emigration, or 1 January 2010, whichever occurred first, unless stated differently. Stata/Se 14.0 (StataCorp LP, College Station, TX) was used for statistical analysis, and a P < 0.05 was considered statistically significant. Results Our cohort comprised 9799 women treated for primary unilateral pure DCIS between 1989 and 2004 in the Netherlands. Median age at diagnosis was 57.4 years (interquartile range = years). Half of the women underwent mastectomy. Of the women treated by BCS, 50% received radiotherapy, mostly between 1999 and 2004 (Table 1). Median follow up time was 9.8 years (interquartile range = years). During follow up, 926 women were diagnosed with IBC and 1429 died. In total, 368 women died of cardiovascular diseases, 284 women of breast cancer, and 333 women of malignant disease other than breast cancer (respectively 26%, 20%, and 23% of all deaths) (Table 2). 5 Cause-specific mortality 85

88 Table 1. Characteristics of the study population by treatment DCIS treatment BCS alone BCS + RT Mastectomy Total Number of DCIS patients (%) Age at DCIS diagnosis, median (IQR), y 58.7 ( ) 57.2 ( ) 56.9 ( ) 57.4 ( ) Age at DCIS diagnosis, y < (4.1) 92 (3.6) 359 (7.7) 557 (5.7) (14.7) 371 (14.4) 857 (18.4) 1604 (16.4) (35.5) 1067 (41.5) 1508 (32.3) 3483 (35.5) (29.3) 730 (28.4) 1190 (25.5) 2670 (27.3) (9.8) 266 (10.3) 457 (9.8) 973 (9.9) > (6.6) 48 (1.9) 296 (6.3) 512 (5.2) Period of DCIS diagnosis (62.7) 736 (28.6) 2499 (53.6) 4840 (49.4) (37.3) 1838 (71.4) 2168 (46.5) 4959 (50.6) DCIS grade (16.1) 243 (9.4) 262 (5.6) 917 (9.4) (13.3) 660 (25.6) 684 (14.7) 1683 (17.2) (14.1) 881 (34.2) 1442 (30.9) 2659 (27.1) Unknown* 1471 (57.5) 790 (30.7) 2279 (48.8) 4540 (46.3) Follow-up time, median (IQR), y 11.1 ( ) 8.1 ( ) 10.2 ( ) 9.8 ( ) Follow-up time, y (7.0) 96 (3.7) 282 (6.0) 558 (5.7) (31.7) 1670 (64.9) 1988 (42.6) 4468 (45.6) (40.8) 588 (22.8) 1493 (32.0) 3125 (31.9) (20.5) 220 (8.6) 904 (19.4) 1648 (16.8) Subsequent invasive breast cancer No 2146 (83.9) 2355 (91.5) 4372 (93.7) 8873 (90.6) Yes 412 (16.1) 219 (8.5) 295 (6.3)** 926 (9.5) Ipsilateral only 274 (10.7) 99 (3.9) 62 (1.3) 435 (4.4) Contralateral only 104 (4.1) 113 (4.4) 226 (4.8) 443 (4.5) Ipsilateral and contralateral 34 (1.3) 7 (0.3) 6 (0.1) 47 (0.5) Subsequent invasive cancer*** No 1944 (76.0) 2190 (85.1) 4048 (86.7) 8182 (83.5) Yes 614 (24.0) 384 (14.9) 619 (13.3) 1617 (16.5) Only invasive breast cancer 373 (14.0) 208 (8.0) 281 (6.0) 862 (8.0) Any other subsequent invasive cancer and no IBC Any other subsequent invasive cancer and IBC 202 (7.9) 165 (6.4) 324 (6.9) 691 (7.1) 39 (1.5) 11 (0.4) 14 (0.3) 64 (0.7) 86 Cause-specific mortality

89 Vital status at end of follow-up Alive 1995 (78.0) 2308 (89.7) 3835 (82.2) 8138 (83.1) Dead 491 (19.2) 229 (8.9) 709 (15.2) 1429 (14.6) Emigrated 72 (2.8) 37 (1.4) 123 (2.6) 232 (2.4) BCS indicates breast conserving surgery; DCIS, ductal carcinoma in situ; IBC, invasive breast cancer; IQR, interquartile range; RT, radiotherapy; y, years. * : 76% unknown versus : 17% unknown ** 1 patient with unknown laterality of subsequent invasive breast cancer. *** Most common (n 40) subsequent invasive cancers other than invasive breast cancer: Malignant neoplasm of colon (C18, n = 125); Malignant neoplasm of bronchus and lung (C34, n = 104), Leukemia and related conditions (C42, n = 54); Other malignant neoplasms of skin (other than melanoma) (C44, n = 117); Malignant neoplasm of corpus uteri (C54, n = 58). Cause-specific Mortality: Comparison With the General Population DCIS patients had a significantly lower risk of dying than women in the general population (SMR 0.92; 95% CI: ) (Table 2). Specifically, they experienced significantly lower mortality from diseases of the circulatory (SMR 0.77; 95% CI: ), respiratory (SMR 0.73; 95% CI: ), and digestive system (SMR 0.74; 95% CI: ); mental and behavioral disorders (SMR 0.7; 95% CI: ); and endocrine, nutritional, and metabolic diseases (SMR 0.69; 95% CI: ). With regard to cancer mortality, DCIS patients had higher risk of breast cancer mortality (SMR 3.33; 95% CI: ), but lower risk of death from all other cancer combined (SMR 0.82; 95% CI: ) and from lung (SMR 0.74; 95% CI: ) and urogenital cancers (SMR 0.62; 95% CI: ) individually (Fig. 1). For most other categories the observed number of deaths was lower than the expected number; however, these differences were not statistically significant (See Table, Supplemental Content 1). 5 Cause-specific mortality 87

90 Table 2. Major causes of death and standardized mortality ratios in population-based cohort of DCIS patients Cause ICD-10 O E % SMR (95% CI) AEM All causes A00-Y ( ) Malignant neoplasm of breast C ( ) 19.6 Malignant neoplasm other than breast C00-97 (excl. C50) ( ) Malignant neoplasm of digestive tract & peritoneum C15-26, ( ) -0.7 Malignant neoplasm of lung, bronchus and trachea C ( ) -2.5 Diseases of circulatory system I ( ) Myocardial infarction I ( ) -1.3 Other heart disease I30-33, ( ) -4.5 Cerebrovascular disease I ( ) -2.2 Diseases of respiratory system J ( ) -3.7 AEM indicates absolute excess mortality per 10,000 patients per year; CI, confidence interval; E, expected number of deaths; O, observed number of deaths; SMR, standardized mortality ratio. 88 Cause-specific mortality

91 5 Figure 1. Observed and expected number of deaths from various disease categories in population-based cohort of 9799 DCIS patients (P < 0.05). Standardized mortality ratios for all causes and breast cancer differed by age at diagnosis, diagnostic period, and follow-up interval (Table 3; Supplemental Content 2). Compared with the general female population, women with DCIS <50 years had increased risk of dying from all causes combined (SMR 1.70; 95% CI: ), whereas DCIS patients >50 years had lower risk (SMR 0.88; 95% CI: ). With regard to breast cancer mortality women <50 years had higher risk of dying compared with the general population than women >50 years (SMR 7.46; 95% CI: and SMR 2.76; 95% CI: , respectively; P homogeneity < 0.001). When studying smaller age groups, we observed that the SMR for breast cancer decreased with increasing age (SMR 23.20; 95% CI: to SMR 1.91; 95% CI: for women aged <40 years and >75 years, respectively; P trend < 0.001). Compared with the general population women diagnosed between 1999 and 2004 were less likely to die from any cause (SMR 0.82; 95% CI: ), whereas women diagnosed earlier had a similar risk (SMR 0.96; 95% CI: ). The SMR for breast cancer was lower for women diagnosed with DCIS between 1999 and 2004 (SMR 2.09; 95% CI: ) than for Cause-specifi c mortality 89

92 those diagnosed between 1989 and 1998 (SMR 3.97; 95% CI: ; Phomogeneity < 0.001). Compared with the general population DCIS patients had lower risk of dying from all causes combined only in the first 5 years after diagnosis, whereas the relative risk of dying from breast cancer increased with later follow-up interval. The SMRs for breast cancer by DCIS grade for women diagnosed between 1999 and 2004 were 0.95 (95% CI: ) for DCIS grade 1, 1.91 (95% CI: ) for DCIS grade 2, and 2.94 (95% CI: ) for DCIS grade 3. Of all women who died from breast cancer, 122 (43%) had been diagnosed with subsequent IBC after DCIS (See Table, Supplemental Content 3). Patients who developed IBC had a 26.6-fold (95% CI: ) increased risk of dying from breast cancer compared with the general population (Supplemental Content 2). For women with subsequent ipsilateral IBC the SMR was (95% CI: ), whereas for women with contralateral IBC the SMR was (95% CI: ). Distribution of age, DCIS grade, and diagnostic period was similar between women who did versus did not have a registered IBC, and died from breast cancer. Distribution of DCIS treatment was different among patients who died from breast cancer: women without subsequent IBC had more often undergone mastectomy, whereas women who experienced IBC had more often received adjuvant radiotherapy. In a subgroup analysis in which women who had died from breast cancer without experiencing IBC were censored, the SMR for breast cancer was 1.43 (95% CI: ).When we stratified the subgroup analysis by diagnostic period, DCIS patients diagnosed between 1989 and 1998 had increased risk of breast cancer mortality compared with the general population (SMR 1.80; 95% CI: ), whereas women diagnosed between 1999 and 2004 had similar risk (SMR 0.72; 95% CI: ; P homogeneity < 0.001). Absolute Breast Cancer Mortality: Subgroups Comparisons Absolute breast cancer mortality was 1.0% at 5 years after DCIS diagnosis, 2.5% at 10 years, and 4.0% at 15 years (See Table, Supplemental Content 4). Women <50 years at DCIS diagnosis had an absolute risk of dying from breast cancer of 4.6% at 15 years, whereas the risk for women >50 years was 3.8% at 15 years. Stratified by treatment the absolute risk estimates at 15 years were 4.7% for BCS alone, 3.8% for BCS plus radiotherapy, and 3.6% for mastectomy (Fig. 2). Women diagnosed with DCIS between 1999 and 2004 had lower absolute risk than those diagnosed between 1989 and 1998 (10-year risk 1.5% vs. 3.1%, respectively; P < 0.001). The 10-year absolute risks for women treated between 1999 and 2004 stratified by agewere: 2.3% and 1.4% for women <50 and >50 years, respectively. Using multivariable-adjusted competing risk regression analysis, women <40 years at DCIS diagnosis were at higher risk for death from breast cancer (hazard ratio, HR 1.99; 95% CI: ) compared with DCIS patients aged 50 to 59 years, who in turn had the lowest risk of dying from breast cancer (See Table, Supplemental Content 5). Women diagnosed between 1999 and 2004 experienced lower breast cancer mortality than women diagnosed between 1989 and 1998 (HR 0.54; 95% CI: ). Women who developed ipsilateral or contralateral IBC had higher 90 Cause-specific mortality

93 risk of dying from breast cancer than women who did not (HR 14.65; 95% CI: and HR 6.26; 95% CI: , respectively, model 2). Between 1989 and 1998, DCIS treatment by mastectomy was associated with lower breast cancer mortality (HR 0.69; 95% CI: ), but between 1999 and 2004 women treated by mastectomy appeared to have similar risk compared with women treated by BCS alone (HR 1.06; 95% CI: ) (Table 4, Supplemental Content 6). However, there was no evidence for significant effect modification by period of diagnosis (P interaction = 0.098). When comparing BCS plus radiotherapy with BCS alone breast cancer mortality did not differ by period of diagnosis (HR 0.87; 95% CI: for and HR 1.06; 95%:CI for ; P interaction = 0.316). For women diagnosed with DCIS between 1999 and 2004 the risk of dying from breast cancer increased with higher grade (HR 1.99; 95% CI: for grade 2 vs. grade 1 and HR 2.81; 95% CI: for grade 3 vs. grade 1; P trend = 0.042). 5 Figure 2. Fifteen-year absolute risks of breast cancer mortality by treatment (P < 0.05). With death caused by other causes as a competing event. Cause-specific mortality 91

94 Table 3. Standardized mortality ratios of major causes of death by age, diagnostic period and treatment All causes Breast cancer O SMR (95% CI) AEM P O SMR (95% CI) AEM P Number of deaths ( ) ( ) 19.6 Age at diagnosis, y <0.001 <0.001 < ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) 20.3 > ( ) ( ) 21.8 Period of diagnosis < ( ) ( ) ( ) ( ) 8.5 DCIS treatment BCS + RT ( ) ( ) 15.6 BCS alone ( ) ( ) 25.5 Mastectomy ( ) ( ) 18.0 Other cancer Circulatory diseases O SMR (95% CI) AEM P O SMR (95% CI) AEM P Number of deaths Age at diagnosis, y < ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) > ( ) ( ) Period of diagnosis ( ) ( ) ( ) ( ) -9.2 DCIS treatment BCS + RT ( ) ( ) -9.4 BCS alone ( ) ( ) -9.2 Mastectomy ( ) ( ) AEM indicates absolute excess mortality per 10,000 patients per year; BCS, breast conserving surgery; CI, confidence interval; O, observed number of deaths; RT, radiotherapy; SMR, standardized mortality ratio. P-values (for homogeneity and trends of SMRs) based on within collapsed person-time Poisson regression models. 92 Cause-specific mortality

95 Table 4. Competing risk regression analysis for breast cancer mortality in women treated for DCIS Treatment Period of diagnosis BCS alone BCS + RT Mastectomy Adjusted SHR (95% CI) Model 1 ref 0.87 ( ) 0.75 ( ) P Adjusted SHR (95% CI) Model 2 ref 1.34 ( ) 1.45 ( ) P Adjusted SHR (95% CI) Model 1 ref 0.87 ( ) 0.69 ( ) P Adjusted SHR (95% CI) Model 2 ref 1.26 ( ) 1.30 ( ) P Adjusted SHR (95% CI) Model 1 ref 1.04 ( ) 1.06 ( ) P Adjusted SHR (95% CI) Model 2 ref 2.12 ( ) 2.58 ( ) P BCS indicates breast conserving surgery; CI, confidence interval; DCIS, ductal carcinoma in situ; NA, not applicable; RT, radiotherapy; SHR, subdistribution hazard ratio (with death due to other causes as a competing event). Model 1: adjusted for age at DCIS diagnosis, period of DCIS diagnosis and DCIS grade. Model 2: adjusted for age at DCIS diagnosis, period of DCIS diagnosis, DCIS grade and subsequent invasive breast cancer; with subsequent ipsilateral and contralateral invasive breast cancer as time-dependent variables. 5 Discussion In this large, nationwide study with 10 years of follow up, we evaluated cause-specific mortality in DCIS patients compared with the general population and we examined factors associated with mortality from specific causes. We observed that DCIS patients experienced lower mortality from diseases of the circulatory, respiratory and digestive system, mental and behavioral disorders, endocrine, nutritional and metabolic diseases, and cancer other than breast cancer. DCIS patients >50 years at diagnosis, which represent the majority of the DCIS population, had lower all-cause mortality than the general female population, whereas the relative risk of dying due to breast cancer was 2.8 times increased. Women <50 years at diagnosis had a 1.7 times increased risk of dying compared with women in the general population, which could be because of their increased breast cancer mortality. Only 6 percent of the DCIS study population was <40 years at DCIS diagnosis, but these women had a 23-times greater risk of dying from breast cancer than expected. Importantly, this highly increased relative risk results from a very low expected number of breast cancer deaths in the Cause-specific mortality 93

96 general population <40 years. Also the results from our within-cohort analysis show that age <40 years was associated with increased breast cancer mortality. An explanation for this finding might be that younger women possibly more often have a larger extent of symptomatic DCIS resulting in a higher risk of unrecognized invasive disease. Our results are in line with a study by Narod et al 4 and emphasize the importance of differential counselling of younger and older women diagnosed with DCIS. However, these young women and their DCIS may not optimally represent young patients and their noninvasive disease today (because of more opportunistic screening and awareness). Therefore, we want to stress that our results do not provide evidence that these women should be treated more intensively, for instance using hormonal treatment. Intuitively, it is very unlikely that the better life expectancy among DCIS patients >50 years is related to the DCIS in itself. A more plausible explanation may be differences in lifestyle characteristics as DCIS patients seem to represent a generally healthy subgroup of the general population. 3,11 In our study, women with DCIS had lower risk of dying caused by cardiovascular and respiratory disease and lung cancer, conditions that are largely caused by lifestyle factors. Notably, DCIS patients treated by radiotherapy were also at decreased risk for cardiovascular death compared with the general population, which has also been reported in a previous study in this cohort. 18 DCIS is mostly detected by screening, and it has been suggested that women who adhere to mammographic screening may be more health-conscious, more often belong to higher socioeconomic classes and have lower comorbidity, resulting in a healthy screenee effect. 3,18 20 This is in line with our finding that women diagnosed with DCIS between 1999 and 2004 (attendance of population-based screening 80% 14 ) and women >50 years (eligible for screening) experienced lower mortality, whereas women diagnosed between 1989 and 1998 (implementation phase) had equal risk to their general population counterparts. However, a previous study that tried to account for this bias by adjusting the models for previous mammography use, comorbidity, and health care utilization seem to contradict a healthy screenee effect. 21 They reported that women >66 years diagnosed with DCIS had similar comorbidity and visited primary care with similar frequency as their controls, and also concluded that a DCIS diagnosis in older women was associated with better survival. Nonetheless, in their discussion they state that If there is a mortality risk for DCIS [...], the risk is likely low and not strong enough to counterbalance a healthy user effect. Another hypothesis is that DCIS patients may seek more medical attention and might adopt a healthier lifestyle after their diagnosis, allowing for prevention or earlier diagnosis and treatment of other diseases. Similarly to previous studies, 3,4 we observed that the 10-year absolute risk of breast cancer mortality in DCIS patients was low and declined for DCIS patients diagnosed in more recent years. This decline in absolute mortality may be because of a decrease in unrecognized IBC at DCIS diagnosis, as radiological and pathological assessment, and treatment selection have improved. Further, in more recent years, with breast screening fully implemented, more indolent DCIS could have been detected, resulting in over diagnosis. 94 Cause-specific mortality

97 Treatment effects on breast cancer mortality found in our study should be interpreted with caution, as confounding by indication may play a significant role. Between 1989 and 1998, women treated by mastectomy had lower breast cancer-specific mortality than women treated by BCS alone, whereas between 1999 and 2004, no difference was found. The results were stratified because we assumed that the DCIS cases diagnosed between 1999 and 2004 better reflect current DCIS cases (more screen-detected) than our study population diagnosed between 1989 and However, there was no statistically significant effect modification by period of diagnosis. Our findings could be explained by improved surgical treatment planning in the latter period, in which the incidence of DCIS increased rapidly and results from randomized controlled trials focusing on BCS were published Moreover in a meta-analysis of the randomized controlled trials studying the effect of radiotherapy after BCS between 1985 and 1999, no difference between among groups was detected. 26 Furthermore, a large observational study, including both BCS and mastectomy between 1998 and 2011, did not report association between treatment and breast cancer mortality. 4 Importantly, the women in our study population were not treated with tamoxifen as part of DCIS treatment, because the clinical guidelines in the Netherlands do not recommend endocrine therapy for women with DCIS. Remarkable, women who developed ipsilateral IBC appeared to have higher risk of dying from breast cancer than women whowere diagnosed with contralateral IBC, a finding which is supported by a study from Narod et al. 4 The difference in outcome after subsequent ipsilateral versus contralateral invasive disease might be explained by potentially more aggressiveness of subsequent ipsilateral IBC after treatment than of new primary tumor in the untreated contralateral breast. In our study, 162 DCIS patients without subsequent IBC died from breast cancer. From all 8873 women with DCIS who did not develop IBC, this is only a small fraction (1.8%). When a woman with a history of DCIS dies from breast cancer, either the invasive component was unrecognized at the time of DCIS diagnosis, DCIS was left behind after treatment and progressed to IBC, or a new primary IBC developed. Therefore, among women with pure DCIS in our study, we hypothesize that true breast cancer deaths are the result of unrecognized, undetected, or unregistered IBC. Another possibility is that these women in fact did not die from breast cancer, but were registered on their death certificate as such. Also in 2 previous studies breast cancer deaths among women who did not have any IBC registered according to population-based cancer registry data were observed (9/2884 = 0.3% and 517/108,196 = 0.5%, respectively). 4,8 Our study has several limitations. The interpretation of mortality statistics is usually complicated by uncertainties about the degree of misclassification of causes of death. However, Harteloh et al 28 showed that for major causes of death, such as cancers, or acute myocardial infarction, reliability of cause of death statistics in the Netherlands was higher than 90%. Moreover, we could not rely on pathology and clinical record review with respect to the diagnoses of primary DCIS and subsequent IBC because of the extensive and anonymous dataset. For 5 Cause-specific mortality 95

98 example, some primary DCIS may have been unrecognized IBC during tumor sampling. Further, the Netherlands Cancer Registry might have missed some subsequent IBC, although their coverage is at least 96%. 12,29,30 In addition, we had no information on estrogen receptor status, comedonecrosis and lesion size, all factors that predicted breast cancer mortality in a study from Narod et al. 4 However, we were able to evaluate DCIS grade in women diagnosed between 1999 and 2004, and also found that the risk of dying from breast cancer increased with higher grade. Similar to the finding of Narod et al, 4 we previously observed that high grade DCIS was, however, not associated with an increased risk of subsequent ipsilateral invasive breast cancer compared with low grade DCIS. 31 Strengths of our study include its large size and population-based character. We were able to combine information on DCIS and subsequent IBC from the NCR with cause of death data from Statistics Netherlands. Furthermore, information from PALGA could be used to validate and complete treatment data. As a result, we had the unique opportunity to study a nationwide DCIS cohort with accurate and complete treatment information and follow-up. In conclusion, DCIS patients >50 years had lower risk of dying compared with women in the general population, which may reflect differences in health behavior. Women diagnosed with primary DCIS had higher risk of dying from breast cancer than women in the general population, but absolute risks were low: cumulative breast cancer mortality 10 years after DCIS was 2.3% for women <50 years and 1.4% for women >50 years treated for DCIS between 1999 and The relative and absolute risk estimates provided in this study are important input for health care providers when counselling women diagnosed with DCIS. Acknowledgments The authors would like to thank Naomi Boekel for performing the linkages, Statistics Netherlands for providing causes of death data and the registration teams of the Netherlands Comprehensive Cancer Organization for the collection of data for the Netherlands Cancer Registry. The authors also thank PALGA, the nationwide histopathology and cytopathology data network and archive, for providing pathology data. Disclosure Jelle Wesseling and Michael Schaapveld have joint last authorship of this study. The authors declared no conflict of interest. This work was supported by Pink Ribbon (grant No WO19. C88, WO29, WO54 to Jelle Wesseling) and the Dutch Cancer Society (grant No. NKI to Marjanka K. Schmidt). 96 Cause-specific mortality

99 Supplementary table 1. Causes of death and standardized mortality ratios in population-based cohort of DCIS patients Cause ICD-10 O E % SMR (95% CI) AEM All causes A00-Y ( ) Infectious and parasitic diseases A00-B ( ) -0.6 Malignant neoplasms C ( ) 12.3 Breast C ( ) 19.6 Other than breast C00-97 (excl. C50) ( ) -7.6 Digestive tract & peritoneum C15-26, ( ) -0.7 Stomach C ( ) 0.3 Colon C ( ) 0.7 Rectum, sigmoid & anus C ( ) -0.1 Pancreas C ( ) -0.4 Lung, bronchus and trachea C , ( ) -2.5 Urogenital tract C ( ) -2.6 Ovary C ( ) -0.4 Other & unspecified C76, ( ) -0.7 Blood, bone marrow & lymph C , ( ) -0.4 Leukemia C ( ) 0.2 Endocrine, nutritional and metabolic diseases E ( ) -1.7 Mental and behavioral disorders F ( ) -2.4 Diseases of central&peripheral nervous system G00-H ( Diseases of circulatory system I ( ) Myocardial infarction I ( ) -1.3 Other ischemic heart disease I20, ( ) -1.1 Other heart disease I30-33, ( ) -4.5 Cerebrovascular disease I ( ) -2.2 Other diseases of circulatory system I00-15, 26-28, ( ) , Diseases of respiratory system J ( ) -3.7 Pneumonia J ( ) -2.0 Chronic lower respiratory diseases J ( ) -1.5 Other diseases of respiratory system J00-06, 20-39, ( ) Diseases of digestive system K ( ) -1.7 Diseases of genitourinary system N ( ) -0.9 Symptoms, signs and abnormal clinical and R ( ) -0.6 laboratory findings, not elsewhere classified External causes of mortality V01-Y ( ) -0.4 Accidents W00-X ( ) -0.6 Suicide X ( ) Abbreviations: O = observed number of deaths; E = expected number of deaths; SMR = standardized mortality ratio; CI = confidence interval; AEM = absolute excess mortality per 10,000 patients per year. Causes of death presented if observed number of deaths is 10. Cause-specific mortality 97

100 Supplementary table 2. Standardized mortality ratios of major causes of death in DCIS patients by subsequent invasive breast cancer and follow-up interval All causes Breast cancer O SMR (95% CI) AEM P O SMR (95% CI) AEM P Number of deaths ( ) ( ) 19.6 Subsequent IBC NA NA No a ( ) ( ) 8.4 Yes b ( ) ( ) Ipsilateral only c ( ) ( ) Contralateral only d ( ) ( ) Ipsilateral&contralateral e ( ) ( ) Follow-up interval, y <0.001 < ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) 30.2 Other cancer Circulatory diseases O SMR (95% CI) AEM P O SMR (95% CI) AEM P Number of deaths Subsequent IBC NA NA No a ( ) ( ) Yes b ( ) ( ) -8.5 Follow-up interval, y < ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) Abbreviations: DCIS = ductal carcinoma in situ; IBC = invasive breast cancer; O = observed number of deaths; SMR = standardized mortality ratio; CI = confidence interval; AEM = absolute excess mortality per 10,000 patients per year; NA = not applicable. P-values (for homogeneity and trends of SMRs) based on within collapsed person-time Poisson regression models. With subsequent invasive breast cancers as time-dependent variables: a Time at risk ended at date of death, emigration, 31 December 2009, or invasive breast cancer diagnosis whichever occurred first. b Time at risk started at date of first invasive breast cancer diagnosis and ended at date of death, emigration, or 31 December 2009, whichever occurred first. c Time at risk started at date of ipsilateral invasive breast cancer diagnosis and ended at date of death, emigration, 31 December 2009, or contralateral invasive breast cancer diagnosis whichever occurred first. 98 Cause-specific mortality

101 d Time at risk started at date of contralateral invasive breast cancer diagnosis and ended at date of death, emigration, 31 December 2009, or ipsilateral invasive breast cancer diagnosis whichever occurred first. e Subgroup consists of women who developed both subsequent ipsilateral and contralateral invasive breast cancer. Time at risk started at date of second/last invasive breast cancer diagnosis and ended at date of death, emigration, or 31 December 2009, whichever occurred first. 5 Cause-specific mortality 99

102 Supplementary table 3. Characteristics of 284 DCIS patients who died of breast cancer stratified by the occurrence of subsequent invasive breast cancer Subsequent invasive breast cancer Yes No P value Number of DCIS patients (%) Age at DCIS diagnosis, y <40 15 (12.3) 15 (9.3) (16.4) 27 (16.7) (32.8) 47 (29.0) (23.8) 47 (29.0) (14.8) 26 (16.1) Period of DCIS diagnosis (82.8) 122 (75.3) (17.2) 40 (24.7) DCIS treatment <0.001 BCS + RT 28 (23.0) 26 (16.1) BCS alone 58 (47.5) 41 (25.3) Mastectomy 36 (29.5) 95 (58.6) DCIS grade (4.1) 8 (4.9) 2 4 (3.3) 17 (10.5) 3 22 (18.0) 34 (21.0) Unknown 91 (74.5) 103 (63.6) Abbreviations: DCIS = ductal carcinoma in situ; BCS = breast conserving surgery; RT = radiotherapy; y = years. 100 Cause-specific mortality

103 Supplementary table 4. Absolute breast cancer-specific mortality in women treated for DCIS Number of Events Patients at risk at 5-year (95% CI) 10-year (95%CI) 15-year (95%CI) patients 15 years Overall ( ) 2.50 ( ) 3.98 ( ) DCIS Treatment BCS alone ( ) 2.68 ( ) 4.72 ( ) BCS + RT ( ) 2.06 ( ) 3.84 ( ) Mastectomy ( ) 2.59 ( ) 3.64 ( ) Age at DCIS diagnosis < ( ) 5.12 ( ) 6.53 ( ) ( ) 2.14 ( ) 3.93 ( ) ( ) 2.16 ( ) 3.65 ( ) ( ) 2.38 ( ) 3.68 ( ) ( ) 3.04 ( ) 4.07 ( ) > ( ) 2.75 ( ) 4.65 ( ) Period of DCIS diagnosis ( ) 3.10 ( ) 4.59 ( ) ( ) 1.54 ( ). DCIS grade ( ) 1.42 ( ) 3.27 ( ) ( ) 1.40 ( ) 1.82 ( ) ( ) 2.40 ( ) 4.61 ( ) Unknown ( ) 2.99 ( ) 4.47 ( ) Abbreviations: DCIS = ductal carcinoma in situ; CI = confidence interval; BCS = breast conserving surgery; RT = radiotherapy. With death due to other causes as a competing event. Cause-specific mortality 101

104 Supplementary table 5. Competing risk regression analysis for breast cancer mortality in women treated for DCIS between 1989 and 2004 Person-years Events Unadjusted SHR (95% CI) P Adjusted SHR (95%CI) Model 1 P Adjusted SHR (95%CI) Model 2 P Treatment BCS alone ref ref ref BCS + RT ( ) ( ) ( ) Mastectomy ( ) ( ) ( ) Age < ( ) ( ) ( ) ( ) ( ) ( ) ref ref ref ( ) ( ) ( ) ( ) ( ) ( ) > ( ) ( ) ( ) Period ref ref ref ( ) < ( ) < ( ) DCIS grade ref ref ref ( ) ( ) ( ) ( ) ( ) ( ) Unknown ( ) ( ) ( ) Cause-specific mortality

105 Subsequent IBC No ipsilateral ref NA NA ref Ipsilateral a ( ) <0.001 NA NA ( ) <0.001 No contralateral ref NA NA ref Contralateral b ( ) <0.001 NA NA 6.26 ( ) <0.001 Abbreviations: DCIS = ductal carcinoma in situ; IBC = invasive breast cancer; SHR = subdistribution hazard ratio; CI = confidence interval; BCS = breast conserving surgery; RT = radiotherapy; NA = not applicable. With death due to other causes as a competing event. Model 1 includes DCIS treatment, age at DCIS diagnosis, period of DCIS diagnosis and DCIS grade. Model 2 includes DCIS treatment, age at DCIS diagnosis, period of DCIS diagnosis, DCIS grade and subsequent invasive breast cancer. With subsequent ipsilateral and contralateral invasive breast cancer as time-dependent variables. a Time at risk started at date of ipsilateral invasive breast cancer diagnosis and ended at date of death, emigration, or 31 December 2009, whichever occurred first. b Time at risk started at date of contralateral invasive breast cancer diagnosis and ended at date of death, emigration, or 31 December 2009, whichever occurred first. 5 Cause-specific mortality 103

106 Supplementary table 6. Competing risk regression analysis for breast cancer mortality in women treated for DCIS stratified by period of diagnosis Period of diagnosis Adjusted SHR (95%CI) Model 1 P value Adjusted SHR (95%CI) Model 1 P value Adjusted SHR (95%CI) Model 2 P value Adjusted SHR (95%CI) Model 2 P value Treatment BCS alone ref ref ref ref BCS + RT 0.87 ( ) ( ) ( ) ( ) Mastectomy 0.69 ( ) ( ) ( ) ( ) Age < ( ) ( ) ( ) ( ) ( ( ) ( ) ( ) ref ref ref ref ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) > ( ) ( ) ( ) ( ) DCIS grade* 1 ref ref ref ref ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) Unknown 1.27 ( ) ( ) ( ) ( ) Subsequent IBC No ipsilateral NA NA NA NA ref ref 104 Cause-specific mortality

107 Ipsilateral a NA NA NA NA ( ) < ( ) <0.001 No contralateral NA NA NA NA ref ref Contralateral b NA NA NA NA 6.47 ( ) < ( ) <0.001 Abbreviations: DCIS = ductal carcinoma in situ; IBC = invasive breast cancer; SHR = subdistribution hazard ratio; CI = confidence interval; BCS = breast conserving surgery; RT = radiotherapy. With death due to other causes as a competing event. Model 1 includes DCIS treatment, age at DCIS diagnosis, period of DCIS diagnosis and DCIS grade. Model 2 includes DCIS treatment, age at DCIS diagnosis, period of DCIS diagnosis, DCIS grade and subsequent invasive breast cancer. With subsequent ipsilateral and contralateral invasive breast cancer as time-dependent variables. a Time at risk started at date of ipsilateral invasive breast cancer diagnosis and ended at date of death, emigration, or 31 December 2009, whichever occurred first. b Time at risk started at date of contralateral invasive breast cancer diagnosis and ended at date of death, emigration, or 31 December 2009, whichever occurred first. * : 76% unknown versus : 17% unknown. 5 Cause-specific mortality 105

108 References 1. Lopez-Garcia MA, Geyer FC, Lacroix-Triki M, et al. Breast cancer precursorsrevisited: molecular features and progression pathways. Histopathology. 2010;57: Erbas B, Provenzano E, Armes J, et al. The natural history of ductalcarcinoma in situ of the breast: a review. Breast Cancer Res Treat. 2006;97: Ernster VL, Barclay J, Kerlikowske K, et al. Mortality among womenwith ductal carcinoma in situ of the breast in the population-based surveillance, epidemiology, and end results program. Arch Intern Med. 2000;160: Narod SA, Iqbal J, Giannakeas V, et al. Breast cancer mortality after adiagnosis of ductal carcinoma in situ. JAMA Oncol. 2015;1: Bartlett JMS, Nofech-Moses S, Rakovitch E. Ductal carcinoma in situ of thebreast: can biomarkers improve current management? Clin Chem. 2014;60: Partridge A, Adloff K, Blood E, et al. Risk perceptions and psychosocialoutcomes of women with ductal carcinoma in situ: longitudinal results from a cohort study. J Natl Cancer Inst. 2008;100: Ruddy KJ, Meyer ME, Giobbie-Hurder A, et al. Long-term risk perceptions of women with ductal carcinoma in situ. Oncologist. 2013;18: Falk RS, Hofvind S, Skaane P, et al. Second events following ductal carcinomain situ of the breast: a register-based cohort study. Breast Cancer Res Treat. 2011;129: Sagara Y, Mallory MA, Wong S, et al. Survival benefit of breast surgery forlow-grade ductal carcinoma in situ: a population-based cohort study. JAMA Surg. 2015;150: Wärnberg F, Bergh J, Holmberg L. Prognosis in women with a carcinoma in situ of the breast: a population-based study in Sweden. Cancer Epidemiol Biomarkers Prev. 1999;8: Worni M, Akushevich I, Greenup R, et al. Trends in treatment patterns andoutcomes for ductal carcinoma in situ. J Natl Cancer Inst. 2015;107: djv van der Sanden GA, Coebergh JW, Schouten LJ, et al. Cancer incidence in TheNetherlands in 1989 and 1990: first results of the nationwide Netherlands cancer registry. Coordinating Committee for Regional Cancer Registries. Eur J Cancer. 1995;31A: Casparie M, Tiebosch ATMG, Burger G, et al. Pathology databanking andbiobanking in The Netherlands, a central role for PALGA, the nationwide histopathology and cytopathology data network and archive. Cell Oncol. 2007;29: National Evaluation Team for Breast cancer screening NETB. Fracheboud J,van Luijt PA, Sankkatsing V, et al National evaluation of breast cancer screening in the Netherlands Available at: Accessed on January 5, Holland R, Peterse JL, Millis RR, et al. Ductal carcinoma in situ: a proposal fora new classification. Semin Diagn Pathol. 1994;11: Day NE, Breslow NE. Statistical Methods in Cancer Research Volume II The Design and Analysis of Cohort Studies. IARC Scientific Publications; Fine JPG, Gray RJ. A proportional hazards model for the subdistribution of acompeting risk. J Am Stat Assoc. 1999;94: Boekel NB, Schaapveld M, Gietema JA, et al. Cardiovascular morbidity andmortality after treatment for ductal carcinoma in situ of the breast. J Natl Cancer Inst. 2014;106:dju Cause-specific mortality

109 19. Lee JR, Vogel VG. Who uses screening mammography regularly? Cancer Epidemiol Biomarkers Prev. 1995;4: Hofer TP, Katz SJ. Healthy behaviors among women in the United States andontario: the effect on use of preventive care. Am J Public Health. 1996;86: Schonberg MA, Marcantonio ER, Ngo L, et al. Causes of death and relativesurvivalof older women after a breast cancer diagnosis. J Clin Oncol. 2011;29: Fisher B, Dignam J, Wolmark N, et al. Lumpectomy and radiation therapyfor the treatment of intraductal breast cancer: findings from National Surgical Adjuvant Breast and Bowel Project B-17. J Clin Oncol. 1998;16: Julien JP, Bijker N, Fentiman IS, et al. Radiotherapy in breast-conservingtreatment for ductal carcinoma in situ: first results of the EORTC randomised phase III trial EORTC Breast Cancer Cooperative Group and EORTC Radiotherapy Group. Lancet. 2000;355: Houghton J, George WD, Cuzick J, et al. Radiotherapy and tamoxifen inwomen with completely excised ductal carcinoma in situ of the breast in the UK, Australia, and New Zealand: randomised controlled trial. Lancet. 2003;362: Emdin SO, Granstrand B, Ringberg A, et al. SweDCIS: Radiotherapy aftersector resection for ductal carcinoma in situ of the breast. Results of a randomised trial in a population offered mammography screening. Acta Oncol. 2006;45: Early Breast Cancer Trialists Collaborative Group (EBCTCG), Correa C,McGale P, Taylor C, et al. Overview of the randomized trials of radiotherapy in ductal carcinoma in situ of the breast. J Natl Cancer Inst Monographs. 2010;2010: Mackenbach JP, Van Duyne WM, Kelson MC. Certification and coding of twounderlying causes of death in The Netherlands and other countries of the European Community. J Epidemiol Community Health. 1987;41: Harteloh P, de Bruin K, Kardaun J. The reliability of cause-of-death coding inthe Netherlands. Eur J Epidemiol. 2010;25: Berkel J. General practitioners and completeness of cancer registry. J Epidemiol Community Health. 1990;44: Schouten LJ, Straatman H, Kiemeney LA, et al. The capture-recapture methodfor estimation of cancer registry completeness: a useful tool? Int J Epidemiol. 1994;23: Elshof LE, Schaapveld M, Schmidt MK, et al. Subsequent risk of ipsilateral and contralateral invasive breast cancer after treatment for ductal carcinoma in situ: incidence and the effect of radiotherapy in a population-based cohort of 10,090 women. Breast Cancer Res Treat. 2016;159: Cause-specific mortality 107

110 Lotte E. Elshof Michael Schaapveld Emiel J. Rutgers Marjanka K. Schmidt Linda de Munck Flora E. van Leeuwen Jelle Wesseling Breast Cancer Research 2017; 19:26

111 CHAPTER 6 The method of detection of ductal carcinoma in situ has no therapeutic implications: Results of a population-based cohort study

112 Abstract Background Population screening with mammography has resulted in increased detection of ductal carcinoma in situ (DCIS). The aim of this population-based cohort study was to assess whether the method of detection should be considered when determining prognosis and treatment in women with DCIS. Methods This study includes 7042 women aged years, who were surgically treated for primary DCIS between 1989 and 2004 in the Netherlands. We calculated cumulative incidences of ipsilateral and contralateral invasive breast cancer and all-cause mortality among women with screendetected, interval, or non-screening-related DCIS, and assessed the association between method of detection and these outcomes, using multivariable Cox regression analyses. Results Compared with non-screening-related DCIS, women with screen-detected DCIS had a lower risk of developing ipsilateral invasive breast cancer (hazard ratio (HR) = 0.75, 95% CI = ), but a similar risk of contralateral invasive breast cancer (HR = 0.86, 95% CI = ). The absolute difference in risk of ipsilateral invasive breast cancer was 1% at 15 years. Screen detection was associated with lower all-cause mortality (HR = 0.85, 95% CI = ); when we additionally accounted for the occurrence of invasive breast cancer the magnitude of this effect remained similar (HR = 0. 86, 95% CI = ). Conclusions Screen detection was associated with lower risk of ipsilateral invasive breast cancer and allcause mortality. However, the absolute difference in risk of ipsilateral invasive breast cancer was very low and the lower all-cause mortality associated with screen-detected and interval DCIS might be explained by a healthy-user effect. Therefore, our findings do not justify different treatment strategies for women with screen-detected, interval, or non-screening-related DCIS. 110 Method of detection

113 Background Population-based breast cancer screening has been introduced on the basis of evidence that mammographic screening could reduce mortality in breast cancer. Yet, there is an ongoing debate about one of the major concerns of screening, overdiagnosis, which is the detection of an abnormality that would never have caused symptoms or death during one s lifetime if screening had been omitted [1 7]. Unfortunately, we are not able to distinguish between women who are overdiagnosed and women who do need treatment. Therefore, we tend to treat them all. This implies that overdiagnosis results in overtreatment [8]. Population screening with mammography has resulted in increased detection of ductal carcinoma in situ (DCIS), a non-obligate precursor lesion for invasive breast cancer. Several studies have indicated an association between mammographic screening and overdiagnosis and overtreatment of DCIS [9 12]. On the other hand, a recent ecological study reported that a higher rate of screen-detected DCIS is associated with a lower rate of invasive interval cancers, suggesting that detection and treatment of DCIS is worthwhile [13]. Method of detection has been shown to be an independent prognostic factor beyond stage migration in patients with invasive breast carcinoma [14, 15]. Women with invasive breast cancer detected at population-based screening have been shown to have better overall and breast cancer specific survival than those who have not participated in the screening program, with an absolute reduction in breast cancer-specific mortality of 7% at 10 years [14]. Previous studies suggest that the method of detection may also carry prognostic information in women treated for DCIS [16 19], but it is still unclear whether women with screen-detected DCIS have a clinically relevant better prognosis than women with non-screening-related DCIS and whether the method for detection should be used in the treatment decision-making process, such as the addition or omission of radiotherapy and antiestrogen treatment. To this end, we studied risk of subsequent ipsilateral and contralateral invasive breast cancer and all-cause mortality among a large population-based cohort of women with screendetected, interval, and non-screening-related DCIS. 6 Methods Patient selection All women who were diagnosed with DCIS in the Netherlands from 1989 through 2004, and who had no previous malignancies except for non-melanoma skin cancer, were selected from a cohort based on linked data from the Netherlands Cancer Registry (NCR) and the nationwide network and registry of histology and cytopathology in the Netherlands (PALGA) [20]. To be eligible for the current study we required: (1) a diagnosis of pure DCIS; (2) age years at Method of detection 111

114 DCIS diagnosis; (3) DCIS treatment consisting of breast-conserving surgery plus radiotherapy, breast-conserving surgery alone, or mastectomy; and (4) no invasive breast cancer or second breast carcinoma in situ within 4 months after initial DCIS diagnosis. Women who were diagnosed at autopsy (n = 1), women in whom the type of surgery could not be determined (n = 59), and women who received chemotherapy or hormonal therapy for DCIS (n = 47) were excluded. Breast cancer screening program The Dutch breast cancer screening program started in 1989 [21]. From 1989 to 1997 women aged years were the target population. Full coverage of these women was achieved in 1997 [22, 23]. In 1998 the program was extended to women aged years. In the Dutch screening program women receive an invitation for screening mammography once every 2 years starting in the year the women turn 50 or 51 years. Invitations for the next round are issued within months of the prior screening. For women who move to another local authority area the next screening could be delayed up to 6 months. Screening mammograms are performed in independent and mostly mobile screening units, and the images are interpreted double-blinded by trained radiologists. Information about screening mammography is recorded by the five regional screening facilities and collected in the database of the Dutch breast cancer screening organization. Method of detection To obtain information about the method of detection we linked our dataset with the database of the Dutch breast cancer screening organization. We classified three categories of DCIS on the basis of method of detection: (1) screen-detected DCIS, defined as DCIS that was detected <12 months after a first or subsequent positive screening examination in the Dutch breast cancer screening program; (2) interval DCIS, defined as DCIS diagnosed <30 months after a negative screening examination, or diagnosed months after a positive screening examination; and (3) DCIS detected outside the breast cancer screening program (non-screening-related), defined as DCIS diagnosed 30 months after the screening examination, diagnosed prior to the first screening examination, or diagnosed in women who never participated in the program. Women who had ever participated in the screening program, but for whom method of DCIS detection was unknown, were excluded from the analyses (n = 525). To examine possible confounding by excluding this group of women, we compared the risk of subsequent ipsilateral and contralateral invasive breast cancer and all-cause mortality between this group and the group of women with a known method of detection: no significant differences were found. Our study cohort included 7042 women in whom the method of detection was known. Because of the gradual implementation of the screening program in the Netherlands, we defined two periods of DCIS diagnosis, and In the first period the women with non-screening-related DCIS were more likely not to be invited for breast cancer screening, while 112 Method of detection

115 in the later period, the group of women with nonscreening-related DCIS could have comprised more women who chose not to participate in breast cancer screening. Treatment and grade Information about surgical DCIS treatment was obtained from the NCR database and further completed based on PALGA data. Further, the NCR provided information on whether radiotherapy was administered. Initial DCIS treatment was defined as the treatment strategy for the primary DCIS lesion within 3 months of diagnosis. Three categories were classified: breast-conserving surgery plus radiotherapy, breast-conserving surgery alone, and mastectomy. Additionally, using PALGA data we assessed whether women initially treated by breastconserving surgery had undergone ipsilateral mastectomy during follow up (due to any cause other than ipsilateral invasive breast cancer). Information on grade was derived from the NCR and was available for 56.4% of the analytical cohort. The grading system used in the Netherlands is based on the classification presented by Holland et al. [24]. Outcome data and statistical analyses Follow up of ipsilateral and contralateral invasive breast cancer and vital status was obtained from the NCR and PALGA databases and was complete up to 1 January The period of time at risk started at the date of DCIS diagnosis, and stopped at the date of the event of interest, emigration, or 31 December 2010, whichever came first. Up to 15-year cumulative incidences of ipsilateral and contralateral invasive breast cancer were calculated according to method of detection, using death as a competing risk. All-cause mortality was calculated according to method of detection and age group at DCIS diagnosis using the Kaplan-Meier method. P values were based either on competing risk regression [25] or Cox proportional hazards regression, with time since DCIS diagnosis as the time scale and adjusted for age (continuous). We used multivariable-adjusted Cox proportional hazards analyses to estimate relative differences in risk of ipsilateral and contralateral invasive breast cancer and all-cause mortality among women with screen-detected, interval, and non-screening-related DCIS. In these analyses we used age as the primary time scale, and time since DCIS diagnosis (0 5, 5 10, and 10 years) as secondary time scale. We adjusted for DCIS treatment (time-varying), DCIS grade and period of diagnosis. We performed a supplementary analysis restricted to women participating in the implemented Dutch screening program, in which we compared risk of ipsilateral and contralateral invasive breast cancer and mortality among women with screen-detected vs. interval DCIS between 1999 and Data on cause of death were not available in this study. In an attempt to correct for death due to invasive breast cancer we additionally included the occurrence of ipsilateral and contralateral 6 Method of detection 113

116 invasive breast cancer as timevarying covariables in the model, with all-cause mortality as outcome. The proportional hazard assumption was verified using graphical and residual-based methods. Furthermore, we assessed whether the effect of method of detection was modified by period of diagnosis by introducing an appropriate interaction term in the model. A P value <0.05 was considered statistically significant. All analyses were performed with STATA/ SE 13.1 (StataCorp LP, College Station, TX, USA). The study was approved by the review boards of the NCR, PALGA, and the Dutch breast cancer screening organization. Results Study population and method of detection The study cohort comprised 4814 women with screendetected DCIS, 651 with interval DCIS, and 1577 with non-screening-related DCIS. Among the women with screen-detected DCIS, 1622 (34%) were detected in the first screening round (i.e. prevalent DCIS), and 3192 (66%) in a subsequent screening round (i.e. incident DCIS). Most screen-detected DCIS was diagnosed within 2 months after the screening mammogram was performed (90th percentile) (Additional file 1). The proportion of screen-detected DCIS increased over time (P < 0.001), whereas the absolute number of non-screeningrelated DCIS remained stable (Fig. 1). The distribution of grade was dependent on screening status (P < 0.001) (Table 1). Of the women with screen-detected DCIS, 59% were treated by breastconserving surgery, compared to 51% of interval DCIS, and 46% of nonscreening-related DCIS (Table 1). The proportion of women undergoing mastectomy decreased over time in all detection groups (P < 0.001). Ipsilateral and contralateral invasive breast cancer With a median follow up of 10.5 years (interquartile range = ), 363 and 378 of 7042 women were diagnosed with ipsilateral and contralateral invasive breast cancer, respectively. Women with screen-detected DCIS had lower risk of ipsilateral invasive breast cancer than women with non-screening-related DCIS (adjusted hazard ratio (HR) = 0.75, 95% confidence interval (CI) = ) (Table 2, Additional file 2). The absolute difference in risk of ipsilateral invasive breast cancer at 15 years of follow up, between screen-detected and nonscreeningrelated DCIS was 1% (cumulative incidence = 6% vs. 7%, respectively) (Fig. 2). No statistically significant difference was observed between interval and nonscreening-related DCIS (HR = 1.02, 95% CI = ). Risk of contralateral invasive breast cancer was not associated with method of detection (screendetected vs. non-screening-related DCIS: HR = 0.86, 95% CI = ; interval versus non- 114 Method of detection

117 screening-related DCIS: HR = 0.83, 95% CI = ) (Table 2, Additional fi le 2). During the 15-year follow-up period, the cumulative incidence of contralateral invasive breast cancer was similar to the cumulative incidence of ipsilateral invasive breast cancer (Fig. 2). The association between method of detection and risk of ipsilateral invasive breast cancer did not differ by period of diagnosis (P interaction = 0.540), nor did it for contralateral invasive breast cancer risk (P interaction = 0.282). 6 Figure 1. Method of detection by year of diagnosis of DCIS. Method of detection 115

118 Table 1. Characteristics of the study population by method of detection Method of detection DCIS diagnosis Nonscreeningrelated Screendetected Interval Within 12 months after positive first screening, n (%) NA 1622 (33.7) NA Within 12 months after positive subsequent screening, n (%) NA 3192 (66.3) NA Within 30 months of negative screening, n (%) NA NA 577 (88.6) 13 to 30 months after positive screening, n (%) NA NA 74 (11.4) >30 months after screening participation, n (%) 159 (10.1) NA NA Prior to first screening participation, n (%) 115 (7.3) NA NA In women who never participated in the screening program, n (%) 1303 (82.6) NA NA Mammography type a Conventional (film) Digital Age at DCIS diagnosis, years Median (interquartile range) 58.1 ( ) 58.7 ( ) 59.8 ( ) 49-59, n (%) 868 (55.0) 2676 (55.6) 327 (50.2) 60-69, n (%) 450 (28.5) 1789 (37.2) 262 (40.3) 70-75, n (%) 259 (16.4) 349 (7.3) 62 (9.5) Period of DCIS diagnosis (implementation phase), n (%) 1044 (66.2) 2001 (41.6) 215 (33.0) (full nationwide coverage), n (%) 533 (33.8) 2813 (58.4) 436 (67.0) DCIS grade (8.0) 440 (9.1) 90 (13.8) (11.9) 926 (19.2) 127 (19.5) (18.3) 1577 (32.8) 208 (32.0) Unknown b 975 (61.8) 1871 (38.9) 226 (34.7) DCIS treatment within 3 months of diagnosis Breast-conserving surgery with radiotherapy, n (%) 276 (17.5) 1587 (33.0) 188 (28.9) Breast-conserving surgery without radiotherapy, n (%) 443 (28.1) 1256 (26.1) 144 (22.1) Mastectomy, n (%) 858 (54.4) 1971 (40.9) 319 (49.0) Follow-up interval, years Median (interquartile range) 12.2 ( ) 10.3 ( ) 9.9 ( ) 0-4 c, n (%) 115 (7.3) 179 (3.7) 29 (4.5) 5-9, n (%) 461 (29.2) 2117 (44.0) 305 (46.9) 10, n (%) 1001 (63.5) 2518 (52.3) 317 (48.7) 116 Method of detection

119 Subsequent invasive breast cancer d No, n (%) 1385 (87.8) 4357 (90.5) 592 (90.9) Ipsilateral only, n (%) 87 (5.5) 213 (4.4) 29 (4.5) Contralateral only, n (%) 94 (6.0) 225 (4.7) 25 (3.8) Ipsilateral + Contralateral, n (%) 11 (0.7) 18 (0.4) 5 (0.8) Vital Status at last follow-up Alive, n (%) 1153 (73.1) 4191 (87.1) 575 (88.3) Dead, n (%) 398 (25.2) 594 (12.3) 68 (10.5) Emigrated, n (%) 26 (1.7) 29 (0.6) 8 (1.2) Total a Last screening mammogram before diagnosis of ductal carcinoma in situ (DCIS). b : 75% unknown vs : 16% unknown. c Eight patients with follow-up time = 0. d One patient with unknown laterality of subsequent invasive breast cancer. NA = not applicable. All-cause mortality During follow up, 1060 of 7042 women died. In a multivariable-adjusted model, adjusted for treatment, grade and period of diagnosis, having a screen-detected or interval DCIS was associated with lower all-cause mortality compared to non-screening-related DCIS (for screendetected DCIS, HR = 0.85, 95% CI = ; for interval DCIS, HR = 0.73, 95% CI = ) (Table 2, Additional file 3). Additional adjustment for the occurrence of invasive breast cancer did not affect these risk estimates, while the confidence interval changed only slightly (HR = 0.86, 95% CI = for screen-detected DCIS; HR = 0.73, 95% CI = for interval DCIS). The association between method of detection and all-cause mortality did not change with period of diagnosis (P interaction = 0.531). Figure 3 shows the Kaplan-Meier curves for all-cause mortality by method of detection; there was no difference between screen-detected and interval DCIS. Differences in all cause-mortality between women attending screening (with screen-detected or interval DCIS) and women with non-screening-related DCIS were only detected in women aged years (P < 0.001) (Additional file 4). 6 In the supplementary analysis, we found that women with interval DCIS tended to have higher risk of ipsilateral invasive breast cancer than women with screen-detected DCIS (HR 1.64; 95% CI ), whereas the risk of contralateral invasive breast cancer was similar (HR 1.01; 95% CI ) (Additional file 5). In addition, the risk of dying was similar in women with screendetected DCIS and women with interval DCIS (HR 1.00; 95% CI ) (Additional file 6). Method of detection 117

120 Table 2. Multivariable-adjusted Cox regression analyses for different events in women aged years at DCIS diagnosis Method of detection Total number of events Person-time, years HR (95% CI) P value Ipsilateral invasive breast cancer Non-screening-related 98 18,710 ref Screen-detected , ( ) Interval ( ) Contralateral invasive breast cancer Non-screening-related ,825 ref Screen-detected , ( ) Interval ( ) All-cause mortality Non-screening-related ,361 ref Screen-detected , ( ) Interval ( ) Analysis was performed with age as the primary time scale and time since DCIS diagnosis (0 5, 5 10, and 10 years) as the secondary time scale. We adjusted for period of ductal carcinoma in situ (DCIS) diagnosis, DCIS grade and DCIS treatment (time-varying). Discussion This large nationwide cohort provides a unique opportunity to assess whether women with screen-detected DCIS have a clinically relevant better prognosis than women with nonscreening-related DCIS and whether the method of detection should be used in the treatment decision-making process. We found that women with screen-detected DCIS had lower risk of subsequent ipsilateral invasive breast cancer, irrespective of age and treatment, compared to women with DCIS that was not detected within the national screening program. However, the absolute difference in risk of ipsilateral invasive breast cancer was very small; 6% of patients with screendetected DCIS and 7% of patients with non-screeningrelated DCIS had developed ipsilateral invasive breast cancer at 15 years. 118 Method of detection

121 Overall Screen detected vs non-screening-related Interval vs non-screening-related P=0.351 P=0.176 P=0.906 a Cumulative incidence of ipsilateral IBC (%) Time (Years) Number at risk Screen-detected Interval Non-screening-related Overall Screen detected vs non-screening-related Interval vs non-screening-related P=0.421 P=0.207 P=0.401 b Cumulative incidence of contralateral IBC (%) 6 Time (Years) Number at risk Screen-detected Interval Non-screening-related Figure 2. Cumulative incidence of a ipsilateral and b contralateral invasive breast cancer by method of detection, with death analyzed as a competing risk. P values based on competing risk regression with time since diagnosis of ductal carcinoma in situ as the primary time scale, adjusted for age (continuous) [25]. Method of detection 119

122 Overall Screen-detected vs non-screening-related Interval vs non-screening-related P<0.001 P<0.001 P=0.002 All-cause mortality (%) Number at risk Time (Years) Screen-detected Interval Non-screening -related Figure 3. Kaplan-Meier curves for all-cause mortality by method of detection. P values based on Cox proportional hazards regression with time since diagnosis of ductal carcinoma in situ as the primary time scale, adjusted for age (continuous). Further, we observed that women with screen-detected and interval DCIS had lower all-cause mortality compared to women with non-screening-related DCIS. Timedependent adjustment for invasive breast cancer did not change the effect estimates, suggesting that the difference is caused by death due to causes other than invasive breast cancer. Importantly, this is circumstantial evidence and we cannot rule out that the difference may be partly explained by breast cancer mortality. However, breast cancer-specifi c mortality in women with DCIS is very low [26]. Therefore, if the difference in all-cause mortality is partly caused by breast cancer mortality, the absolute difference in breast cancer mortality between screen-detected, interval, and nonscreening-related DCIS will likely be very small and clinically not signifi cant. Another possible explanation for lower all-cause mortality in women who participate in mammographic screening is the healthy-user effect [27]. The healthy-user effect occurs when patients who choose to receive one preventive service also choose to receive other preventive services. These patients also may have less comorbidity and better functional status, increasing the likelihood of other healthy behaviors. This is supported by a Dutch study, in which the attendance rate for breast cancer screening was lower among women with a low socioeconomic status, and comorbidity was inversely associated with socioeconomic status [28]. A healthy-user effect among women who attend mammographic screening is also supported by other studies 120 Method of detection

123 [29 32]. Based on the available evidence, the healthy-user effect might explain the better overall survival in women with screen-detected DCIS. In earlier studies the association between method of detection and the development of subsequent invasive breast cancer was tested in patients with DCIS, but a recent meta-analysis showed that the classification of detection method varied and had not been uniformly defined [16]. As a consequence, interpreting the association between method of detection and outcome of treatment of DCIS is more difficult. A few studies have evaluated whether women with DCIS detected through a population-based screening program have better prognosis than women with non-screening-related DCIS [18, 19], and therefore can be compared to our findings. Cheung et al. [19] (n = 3930 subjects, length of follow up not reported) found that the difference in risk of ipsilateral invasive breast cancer between screen-detected and non-screening-related DCIS was somewhat larger than in our study (annual absolute risk 0.43% vs. 0.65%; HR = 0.32, P value <0.0001), possibly because of variation in selection criteria. Our study only included women who were eligible for participation in the population-based screening program based on age, whereas Cheung et al. analyzed all women, and thus also included younger women (<50 years) who have a higher risk of invasive breast cancer compared to older women [33]. Falk et al. [18] (n = 3163 subjects, median follow up 5.2 years) reported a lower risk of ipsilateral invasive breast cancer in screen-detected DCIS, which was comparable to our results (HR = 0.7, 95% CI = ), but did not report absolute risk estimates for groups according to different methods of detection. The meta-analysis by Zhang et al. included six studies; in five of these studies mammographically detected DCIS was compared with symptomatic or palpable DCIS [16]. Unfortunately, we had no information on whether the DCIS was detected by mammography or by clinical symptoms. In addition, the non-screening-related group might also include women who attended opportunistic screening or who were at higher risk of breast cancer and therefore attended a different screening program. However, Zhang et al. reported a higher risk of ipsilateral invasive breast cancer in symptomatic DCIS, which was comparable to the higher risk in non-screening-related DCIS in our study (HR = 1.38, 95% CI = and HR = 1.33, 95% CI = , respectively). It has been postulated that the difference in risk of ipsilateral invasive breast cancer between screen-detected and non-screening-related DCIS might be due to differences in the underlying biology of DCIS [34 36]. Symptomatic presentation has been shown to be associated with larger lesion size and higher risk of estrogen receptor negativity [34, 35]. Triple-negative, human epidermal growth factor receptor 2, and basal-like phenotypes were found to be more common in symptomatic DCIS, while the luminal A phenotype was more often observed in screen-detected DCIS [35]. Although there was a larger proportion of low-grade DCIS in the screen-detected group in one study, in others there was no difference in distribution by grade [34 36]. Also in a recent Dutch study, the distribution by grade was not dependent on mass screening status (screened population in the breast cancer screening program vs. population not subjected to/ 6 Method of detection 121

124 participating in mass screening) [37]. In our study the distribution by grade was dependent on screening status, but the difference in risk of ipsilateral invasive breast cancer between screendetected and non-screening-related DCIS was independent of grade. Unfortunately, there was no information available to explore other differences in the biology of DCIS. Breast density might be another explanation for the observed lower risk of ipsilateral invasive breast cancer among women with screen-detected DCIS. As high breast density is associated with decreased mammographic accuracy and increased risk of breast cancer [38, 39], it might be hypothesized that women with screen-detected DCIS have relatively low breast density and therefore confer a lower risk of subsequent breast cancer as compared with women with interval DCIS. The association between method of detection and mortality in women with DCIS has been studied less frequently. Our results are in line with the study by Koh et al. (n = 1202 subjects, median follow up 8.2 years), which showed that women with screen-detected DCIS had better overall survival than women with symptomatic DCIS [35]. The high overall survival rates (100% and 97.8% at 10 years, respectively) presented by the authors can be explained by the large proportion (45%) of women aged <50 years at diagnosis in their study. The prognostic value of the detection method seems to be different in patients with DCIS compared to patients with invasive breast cancer. As considerably better overall survival and breast cancer-specific survival has been observed in studies of women with screen-detected invasive breast cancer compared with non-screening-related invasive breast cancer, it has been suggested that method of detection should be used when selecting patients for adjuvant systemic therapy and that withholding chemotherapy for patients with screen-detected invasive cancer could be considered [14, 15, 40]. In contrast, our results do not support treatment decision-making, such as omitting radiation after breast-conserving surgery or using adjuvant hormonal therapy [41], based on the method of detection. A major strength of our study is that detailed information from the Dutch breast cancer screening program was available for this large population-based cohort with substantial follow up. On top of this, we had complete data on treatment type and therefore were able to adjust for this factor in the multivariable analyses. In addition, by using age as the primary time scale in these multivariable analyses we were able to correct for age in the most optimal way. A limitation is that the non-screening-related group represents a heterogeneous group, including both women who were not invited in the screening program and women who refused to participate. However, we found that the association between method of detection and all three outcomes was similar among women diagnosed between 1989 and 1998 (most likely random selection) and 1999 and 2004 (subject to selection bias). Another important question is to what extent our results reflect the current situation in which screen-film mammography has been replaced by full-field digital mammography in most Western breast cancer screening programs. Digital mammography seems to be associated 122 Method of detection

125 with a higher rate of detection of DCIS [42 44]. However, in a recent Dutch study to determine whether this transition has resulted in changes in performance indicators and characteristics of screen-detected and interval cancers, there was no increased incidence of DCIS [45]. Moreover, opposing effects have been reported on the distribution of grade. Both a higher proportion of lowgrade DCIS [46] and a higher proportion of high-grade DCIS [42] at digital mammography have been described. Further, others have found no differences in the distribution of grade after digital mammography compared with screen-film mammography [43, 45]. These discrepancies complicate the translation of our results to the current digital era, but also suggest that the truth lies somewhere in between. Therefore, our results are likely very relevant for patients diagnosed with DCIS today. Conclusions This study showed that having screen-detected DCIS was associated with lower risk of ipsilateral invasive breast cancer, but the absolute differences in risk of ipsilateral invasive breast cancer were not clinically significant. Women with screen-detected and interval DCIS had lower all-cause mortality compared with women with non-screening-related DCIS, which might be explained by the healthy-user effect based on the available evidence. Therefore, our findings indicate a limited prognostic role of the method of detection of DCIS, and do not justify different treatment strategies for women with screendetected, interval, or non-screening-related DCIS. Acknowledgements 6 The authors thank Jan Henk Sangers, Frank Yntema, and the Dutch screening organization for providing screening data. The authors also thank Otto Visser, Annemarie Eeltink, Ries Kranse, and the registration teams of the Netherlands Comprehensive Cancer Organization for the collection of data for the Netherlands Cancer Registry. Lucy Overbeek and PALGA, the nationwide histopathology and cytopathology data network and archive, are thanked for providing pathology data. This work was supported by Pink Ribbon (grant number 2011.WO19. C88); and the Dutch Cancer Society (grant number NKI to M.K.S.). Method of detection 123

126 Competing interests The authors declare that they have no competing interests. This work was supported by Pink Ribbon (grant number 2011.WO19.C88); and the Dutch Cancer Society (grant number NKI to MKS). The funding sources had no involvement in the study design, the collection, analysis, and interpretation of data, or in the writing of the manuscript. Additional file 1. Time (days) between a fi rst or subsequent screening examination by the Dutch breast cancer screening program and DCIS diagnosis in women with screen-detected DCIS (DCIS diagnostic period ). 124 Method of detection

127 Additional file 2a. Multivariable-adjusted Cox regression analysis of ipsilateral invasive breast cancer in women aged years at DCIS diagnosis (DCIS diagnostic period ). Age was the primary time scale, time since DCIS diagnosis (0 5, 5 10, and 10 years) the secondary time scale, and DCIS treatment a time-varying covariable. Total ipsilateral IBC Person-time, years HR (95% CI) p-value Method of detection Non-screening-related ref Screen-detected ( ) Interval ( ) Treatment BCS+RT ref BCS alone ( ) <0.001 Mastectomy ( ) <0.001 Year of diagnosis ref ( ) Grade ref ( ) ( ) Unknown ( ) Follow-up interval 0-5 years ref 5-10 years ( ) >10 years ( ) Method of detection 125

128 Additional file 2b. Multivariable-adjusted Cox regression analysis of contralateral invasive breast cancer in women aged years at DCIS diagnosis (DCIS diagnostic period ). Age was the primary time scale, time since DCIS diagnosis (0 5, 5 10, and 10 years) the secondary time scale, and DCIS treatment a time-varying covariable. Total contralateral IBC Person-time, years HR (95% CI) p-value Method of detection Non-screening-related ref Screen-detected ( ) Interval ( ) Treatment BCS+RT ref BCS alone ( ) Mastectomy ( ) Year of diagnosis ref ( ) Grade ref ( ) ( ) Unknown ( ) Follow-up interval 0-5 years ref 5-10 years ( ) >10 years ( ) Method of detection

129 Additional file 3. Multivariable-adjusted Cox regression analysis of overall mortality in women aged years at DCIS diagnosis (DCIS diagnostic period ). Age was the primary time scale and time since DCIS diagnosis (0 5, 5 10, and 10 years) the secondary time scale. Model 1 was adjusted for period of DCIS diagnosis, DCIS grade, and DCIS treatment (time-varying). Model 2 was adjusted for period of DCIS diagnosis, DCIS grade, DCIS treatment (time-varying), and the occurrence of ipsilateral and contralateral invasive breast cancer (time-varying). Total deaths Persontime, years HR (95% CI) p-value HR (95% CI) p-value Model 1 Model 2 Method of detection Non-screening-related ref ref Screen-detected ( ) ( ) Interval ( ) ( ) Treatment BCS+RT ref ref BCS alone ( ) ( ) Mastectomy ( ) ( ) Year of diagnosis ref ref ( ) ( ) Grade ref ref ( ) ( ) ( ) ( ) Unknown ( ) ( ) Follow-up interval 0-5 years ref ref 5-10 years ( ) ( ) >10 years ( ) ( ) Ipsilateral IBC No NA NA ref Yes NA NA 2.92 ( ) <0.001 Contralateral IBC No NA NA ref Yes NA NA 1.87 ( ) <0.001 Method of detection 127

130 Overall P =0.598 Screen-detected vs non-screening-related P =0.921 Interval vs non-screening-related P =0.336 a Y-as: All-cause mortality (%) Kaplan-Meier failure estimates analysis time Number at risk Screen-detected Interval Non-screening-related Overall P <0.001 Screen-detected vs non-screening-related P <0.001 Interval vs non-screening-related P =0.011 b Y-as: All-cause mortality (%) Kaplan-Meier failure estimates analysis time Number at risk Screen-detected Interval Non-screening-related Method of detection

131 Overall P =0.562 Screen-detected vs non-screening-related P =0.325 Interval vs non-screening-related P =0.491 Kaplan-Meier failure estimates c Y-as: All-cause mortality (%) analysis time Number at risk Number at risk Screen-detected Interval Non-screening-related Additional file 4. Kaplan-Meier curves for all-cause mortality by method of detection for a patients aged years at DCIS diagnosis, b patients aged years at DCIS diagnosis, and c patients aged years at DCIS diagnosis (DCIS diagnostic period ). P values based on Cox proportional hazards regression with time since DCIS diagnosis as the primary time scale, adjusted for age (continuous). 6 Method of detection 129

132 Additional file 5. Multivariable-adjusted Cox regression analysis of ipsilateral and contralateral invasive breast cancer in women aged years at DCIS diagnosis: comparison between screen-detected and interval DCIS (DCIS diagnostic period (screening implemented)). Age was the primary time scale and time since DCIS diagnosis (0 5, 5 10, and 10 years) the secondary time-scale. Total ipsilateral IBC Person-time, years HR (95% CI) p-value Method of detection Screen-detected ref Interval ( ) Treatment BCS+RT ref BCS alone ( ) <0.001 Mastectomy ( ) <0.001 Grade ref ( ) ( ) Unknown ( ) Follow-up interval 0-5 years ref 5-10 years ( ) years ( ) Total contralateral IBC Person-time, years HR (95% CI) p-value Method of detection Screen-detected ref Interval ( ) Treatment BCS+RT ref BCS alone ( ) Mastectomy ( ) Grade ref ( ) ( ) Unknown ( ) Follow-up interval 0-5 years ref 5-10 years ( ) years ( ) Method of detection

133 Additional file 6. Multivariable-adjusted Cox regression analysis of overall mortality in women aged years at DCIS diagnosis: comparison between screen-detected and interval DCIS (DCIS diagnostic period (screening implemented)). Age was the primary time scale and time since DCIS diagnosis (0 5, 5 10, and 10 years) was the secondary time scale. Model 1 was adjusted for period of DCIS diagnosis, DCIS grade, and DCIS treatment (time-varying). Model 2 was adjusted for period of DCIS diagnosis, DCIS grade, DCIS treatment (time-varying), and the occurrence of ipsilateral and contralateral invasive breast cancer (time-varying). Total deaths Person-time, years HR (95% CI) p-value HR (95% CI) p-value Model 1 Model 2 Method of detection Screen-detected ref ref Interval ( ) ( ) Treatment BCS+RT ref ref BCS alone ( ) ( ) Mastectomy ( ) ( ) Grade ref ref ( ) ( ) ( ) ( ) Unknown ( ) ( ) Follow-up interval 0-5 years ref ref 5-10 years ( ) ( ) years ( ) ( ) Ipsilateral IBC No NA NA ref Yes NA NA 3.32 ( ) <0.001 Contralateral IBC No NA NA ref Yes NA NA 1.79 ( ) Method of detection 131

134 References 1. Esserman LJ, Thompson IM, Reid B. Overdiagnosis and overtreatment in cancer: an opportunity for improvement. JAMA. 2013;310: Welch HG, Passow HJ. Quantifying the benefits and harms of screening mammography. JAMA Intern Med. 2014;174: Marmot MG, Altman DG, Cameron DA, Dewar JA, Thompson SG, Wilcox M. 4. The benefits and harms of breast cancer screening: an independent review. Br J Cancer. 2013;108: Carter JL, Coletti RJ, Harris RP. Quantifying and monitoring overdiagnosis in cancer screening: a systematic review of methods. BMJ. 2015;350:g Barratt A. Overdiagnosis in mammography screening: a 45 year journey from shadowy idea to acknowledged reality. BMJ. 2015;350:h Harding C, Pompei F, Burmistrov D, Welch HG, Abebe R, Wilson R. Breast cancer screening, incidence, and mortality across US counties. JAMA Intern Med. 2015;175(9): Independent UK Panel on Breast Cancer Screening. The benefits and harms of breast cancer screening: an independent review. Lancet. 2012;380: Welch HG. Overdiagnosis and mammography screening. BMJ. 2009;339:b Alvarado M, Ozanne E, Esserman L. Overdiagnosis and overtreatment of breast cancer. Am Soc Clin Oncol Educ Book. 2012:e Jørgensen KJ, Gøtzsche PC. Overdiagnosis in publicly organised mammography screening programmes: systematic review of incidence trends. BMJ. 2009;339:b Ernster VL, Barclay J, Kerlikowske K, Grady D, Henderson C. Incidence of and treatment for ductal carcinoma in situ of the breast. JAMA. 1996;275: Ripping TM, Verbeek ALM, Fracheboud J, de Koning HJ, van Ravesteyn NT, Broeders MJM. Overdiagnosis by mammographic screening for breast cancer studied in birth cohorts in The Netherlands. Int J Cancer. 2015;137: Duffy SW, Dibden A, Michalopoulos D, Offman J, Parmar D, Jenkins J, Collins B, Robson T, Scorfield S, Green K, Hall C, Liao X-H, Ryan M, Johnson F, Stevens G, Kearins O, Sellars S, Patnick J. Screen detection of ductal carcinoma in situ and subsequent incidence of invasive interval breast cancers: a retrospective population-based study. Lancet Oncol. 2016;17: Mook S, van t Veer LJ, Rutgers EJ, Ravdin PM, van de Velde AO, van Leeuwen FE, Visser O, Schmidt MK. Independent prognostic value of screen detection in invasive breast cancer. J Natl Cancer Inst. 2011;103: Wishart GC, Greenberg DC, Britton PD, Chou P, Brown CH, Purushotham AD, Duffy SW. Screendetected vs symptomatic breast cancer: is improved survival due to stage migration alone? Br J Cancer. 2008;98: Zhang X, Dai H, Liu B, Song F, Chen K. Predictors for local invasive recurrence of ductal carcinoma in situ of the breast: a meta-analysis. Eur J Cancer Prev. 2016;25: Wang S-Y, Shamliyan T, Virnig BA, Kane R. Tumor characteristics as predictors of local recurrence after treatment of ductal carcinoma in situ: a meta-analysis. Breast Cancer Res Treat. 2011;127: Falk RS, Hofvind S, Skaane P, Haldorsen T. Second events following ductal carcinoma in situ of the breast: a register-based cohort study. Breast Cancer Res Treat. 2011;129: Cheung S, Booth ME, Kearins O, Dodwell D. Risk of subsequent invasive breast cancer after a diagnosis of ductal carcinoma in situ (DCIS). Breast. 2014;23: Method of detection

135 20. Casparie M, Tiebosch ATMG, Burger G, Blauwgeers H, van de Pol A, van Krieken JHJM, Meijer GA. Pathology databanking and biobanking in The Netherlands, a central role for PALGA, the nationwide histopathology and cytopathology data network and archive. Cell Oncol. 2007;29: National Evaluation Team for Breast cancer screening, Fracheboud J, van 49. Luijt PA, Sankkatsing V, Ripping TM, Broeders M, Otten J, van Ineveld BM, Heijnsdijk E, Verbeek A, Holland R, den GJ H, de Bruijn AE, de Koning HJ. 50. National evaluation of breast cancer screening in the Netherlands / pdf/?lang=en. Accessed 1 Oct de Koning HJ, Fracheboud J, Boer R, Verbeek AL, Collette HJ, Hendriks JH, van Ineveld BM, de Bruyn AE, van der Maas PJ. Nation-wide breast cancer screening in The Netherlands: support for breast-cancer mortality reduction. 52. National Evaluation Team for Breast Cancer Screening (NETB). Int J Cancer. 1995;60: Fracheboud J, de Koning HJ, Beemsterboer PM, Boer R, Hendriks JH, Verbeek AL, van Ineveld BM, de Bruyn AE, van der Maas PJ. Nation-wide breast cancer screening in The Netherlands: results of initial and subsequent screening National Evaluation Team for Breast Cancer Screening. Int J Cancer. 1998;75: Holland R, Peterse JL, Millis RR, Eusebi V, Faverly D, van de Vijver MJ, Zafrani B. Ductal carcinoma in situ: a proposal for a new classification. Semin Diagn Pathol. 1994;11: Fine JPG, Gray RJ. A proportional hazards model for the subdistribution of a competing risk. J Am Stat Assoc. 1999;94: Narod SA, Iqbal J, Giannakeas V, Sopik V, Sun P. Breast cancer mortality after a diagnosis of ductal carcinoma in situ. JAMA Oncol. 2015;1: Silverman SL, Gold DT. Healthy users, healthy adherers, and healthy behaviors? J Bone Miner Res. 2011;26: Aarts MJ, Voogd AC, Duijm LEM, Coebergh JWW, Louwman WJ. Socioeconomic inequalities in attending the mass screening for breast cancer in the south of the Netherlands associations with stage at diagnosis and survival. Breast Cancer Res Treat. 2011;128: Lee JR, Vogel VG. Who uses screening mammography regularly? Cancer Epidemiol Biomarkers Prev. 1995;4: Hofer TP, Katz SJ. Healthy behaviors among women in the United States and Ontario: the effect on use of preventive care. Am J Public Health. 1996; 86: Boekel NB, Schaapveld M, Gietema JA, Rutgers EJT, Versteegh MIM, Visser O, Aleman BMP, van Leeuwen FE. Cardiovascular morbidity and mortality after treatment for ductal carcinoma in situ of the breast. J Natl Cancer Inst. 2014; 106:dju Ernster VL, Barclay J, Kerlikowske K, Wilkie H, Ballard-Barbash R. Mortality among women with ductal carcinoma in situ of the breast in the population-based surveillance, epidemiology and end results program. Arch Intern Med. 2000;160: Elshof LE, Schaapveld M, Schmidt MK, Rutgers EJ, van Leeuwen FE, Wesseling J. Subsequent risk of ipsilateral and contralateral invasive breast cancer after treatment for ductal carcinoma in situ: incidence and the effect of radiotherapy in a population-based cohort of 10,090 women. Breast Cancer Res Treat. 2016;159: Method of detection 133

136 64. Barnes NLP, Dimopoulos N, Williams KE, Howe M, Bundred NJ. The frequency of presentation and clinico-pathological characteristics of symptomatic versus screen detected ductal carcinoma in situ of the breast. Eur J Surg Oncol. 2014;40: Koh VCY, Lim JCT, Thike AA, Cheok PY, Thu MMM, Tan VKM, Tan BKT, Ong KW, Ho GH, Tan WJ, Tan Y, Salahuddin AS, Busmanis I, Chong APY, Iqbal J, Thilagaratnam S, Wong JSL, Tan PH. Characteristics and behaviour of screen-detected ductal carcinoma in situ of the breast: comparison with symptomatic patients. Breast Cancer Res Treat. 2015;152: Evans AJ, Pinder SE, Ellis IO, Wilson AR. Screen detected ductal carcinoma in situ (DCIS): overdiagnosis or an obligate precursor of invasive disease? J Med Screen. 2001;8: van Luijt PA, Heijnsdijk EAM, Fracheboud J, Overbeek LIH, Broeders MJM, Wesseling J, den Heeten GJ, de Koning HJ. The distribution of ductal carcinoma in situ (DCIS) grade in 4232 women and its impact on overdiagnosis in breast cancer screening. Breast Cancer Res. 2016;18: Boyd NF, Guo H, Martin LJ, Sun L, Stone J, Fishell E, Jong RA, Hislop G, Chiarelli A, Minkin S, Yaffe MJ. Mammographic density and the risk and detection of breast cancer. N Engl J Med. 2007;356: Carney PA, Miglioretti DL, Yankaskas BC, Kerlikowske K, Rosenberg R, Rutter CM, Geller BM, Abraham LA, Taplin SH, Dignan M, Cutter G, Ballard-Barbash R. Individual and combined effects of age, breast density, and hormone replacement therapy use on the accuracy of screening mammography. Ann Intern Med. 2003;138: Shen Y, Yang Y, Inoue LYT, Munsell MF, Miller AB, Berry DA. Role of detection method in predicting breast cancer survival: analysis of randomized screening trials. J Natl Cancer Inst. 2005;97: Senkus E, Kyriakides S, Ohno S, Penault-Llorca F, Poortmans P, Rutgers E, Zackrisson S, Cardoso F, ESMO Guidelines Committee. Primary breast cancer: ESMO Clinical Practice Guidelines for diagnosis, treatment and follow-up. Ann Oncol. 2015;26 Suppl 5:v Vigeland E, Klaasen H, Klingen TA, Hofvind S, Skaane P. Full-field digital mammography compared to screen film mammography in the prevalent round of a population-based screening programme: the Vestfold County Study. Eur Radiol. 2008;18: Bluekens AMJ, Holland R, Karssemeijer N, Broeders MJM, den GJ H. Comparison of digital screening mammography and screen-film mammography in the early detection of clinically relevant cancers: a multicenter study. Radiology. 2012;265: van Luijt PA, Fracheboud J, Heijnsdijk EAM, den GJ H, de Koning HJ. National Evaluation Team for Breast Cancer Screening in Netherlands Study Group (NETB). Nation-wide data on screening performance during the transition to digital mammography: observations in 6 million screens. Eur J Cancer. 2013;49: de Munck L, de Bock GH, Otter R, Reiding D, Broeders MJ, Willemse PH, Siesling S. Digital vs screen-film mammography in population-based breast cancer screening: performance indicators and tumour characteristics of screen-detected and interval cancers. Br J Cancer. 2016;115: Nederend J, Duijm LEM, Louwman MWJ, Groenewoud JH, Donkers-van Rossum AB, Voogd AC. Impact of transition from analog screening mammography to digital screening mammography on screening outcome in The Netherlands: a population-based study. Ann Oncol. 2012;23: Method of detection

137 Method of detection 135 6

138 Lotte E. Elshof Konstantinos Tryfonidis Leen Slaets A. Elise van Leeuwen-Stok Victoria P. Skinner Nicolas Dif Ruud M. Pijnappel Nina Bijker Emiel J.T. Rutgers Jelle Wesseling European Journal of Cancer 2015; 51:

139 CHAPTER 7 Feasibility of a prospective, randomised, open-label, international multicentre, phase III, non-inferiority trial to assess the safety of active surveillance for low risk ductal carcinoma in situ - The LORD study

140 Abstract Background The current debate on overdiagnosis and overtreatment of screen-detected ductal carcinoma in situ (DCIS) urges the need for prospective studies to address this issue. A substantial number of DCIS lesions will never form a health hazard, particularly if it concerns non- to slow-growing low-grade DCIS. The LORD study aims to evaluate the safety of active surveillance in women with LOw-Risk Dcis. Design This is a randomised, international multicentre, open-label, phase III non-inferiority trial, led by the Dutch Breast Cancer Research Group (BOOG ) and the European Organization for Research and Treatment of Cancer (EORTC-BCG 1401). Standard treatment will be compared to active surveillance in 1240 women aged 45 years with asymptomatic, screen-detected, pure low-grade DCIS based on vacuum-assisted biopsies of microcalcifications only. Both study arms will be monitored with annual digital mammography for a period of 10 years. The primary endpoint is 10-year ipsilateral invasive breast cancer free percentage. Secondary end-points include patient reported outcomes, diagnostic biopsy rate during follow-up, ipsilateral mastectomy rate and translational research. Feasibility To explore interest in and feasibility of the LORD study we conducted a survey among EORTC and BOOG centres. A vast majority of EORTC and BOOG responding centres expressed interest in participation in the LORD study. The proposed study design is endorsed by nearly all centres. 138 The LORD study

141 1. Background The introduction of population-based mammographic breast cancer screening programmes and the implementation of digital mammography have led to impressive increase in incidence of ductal carcinoma in situ (DCIS) [1 7]. This increase is not clearly associated with a substantial decrease in incidence of advanced invasive breast cancer (ibc), and since a proportion of DCIS will either not progress to ibc or progress at a much later stage in life, it is suggested that overdiagnosis and resultant overtreatment exist [8 13]. Currently, in the United States, approximately one woman is diagnosed with DCIS for every four women diagnosed with ibc [7]. In the Netherlands in 2012, 14,296 women were diagnosed with ibc and 2245 women were diagnosed with non-invasive pre-stage breast cancer. Of these, approximately 80% were DCIS. The European standardised rate of non-invasive, pre-stage breast cancer has increased almost fivefold from 1989 to 2012 in the Netherlands 4.9 per 100,000 to 22.3 per 100,000. About one in six of all DCIS cases was a low-grade disorder [14]. The goal in treating patients with DCIS is to prevent development of ibc. The conventional treatment of DCIS is similar to that of early-stage ibc, i.e. wide local excision (WLE) often followed by radiotherapy (RT), or mastectomy, and possibly hormonal therapy (HT). The outcomes in patients with DCIS treated by these conventional therapies are excellent, but the best way to manage DCIS is still subject to debate. The natural history of DCIS is not clear, since traditionally, patients diagnosed with DCIS are treated. Therefore, series of patients not treated for DCIS are not available [15]. However, there exists strong circumstantial evidence that DCIS is a non-obligate precursor lesion of ibc [16]. Studies where DCIS was initially misdiagnosed as benign and treated by biopsy alone suggest that between 50% and 85% of all DCIS will never progress into ibc [15]. From autopsy series we have learned that the proportion of middle-aged women who harboured undetected DCIS ranged from 10% to 39% [17]. A range of proliferative breast lesions like DCIS have been considered risk indicators or precursor lesions and carry a relative risk of ibc ranging from 1.2 to 10 [16]. Lobular carcinoma in situ (LCIS) is generally considered a risk indicator, inferring an increased rate of development of ibc, either in the ipsilateral or contralateral breast, of about 1 2% per year, and a relative risk similar to that of DCIS (8 10) [18 23]. Strikingly, low-grade DCIS is intensively treated, and coincidentallyfound classical LCIS is managed by active surveillance [18]. Unfortunately, accurate prognostic factors to distinguish potentially life-threatening DCIS from non-hazardous pre-invasive lesions are lacking. The aggressive subtypes require intensive treatment, but apparently indolent DCIS might be managed by active surveillance [24]. Progression from DCIS to ibc constitutes a complex biological phenomenon [16,25]. It has been hypothesised that breast cancer evolution can be classified into two groups: a low- and highgrade breast neoplasia family [16,26 30]. The low- and high-grade multistep model of breast 7 The LORD study 139

142 cancer progression based on morphological, immunophenotypical and molecular features described by Lopez-Garcia et al. suggests that if low-grade DCIS progresses to invasive disease this will be low-grade ibc with favourable characteristics in most cases and survival rates after treatment of this invasive cancer will still be excellent. This model is supported by other studies [25,31]. Furthermore, it has been hypothesised that progression from low-grade pre-stage breast cancer to high-grade ibc is an uncommon biological phenomenon [16,31,32]. Four randomised controlled trials that begun between 1985 and 1990 showed that adjuvant RT after WLE reduced local failure rates, but did not reduce the risk of metastases or breast cancer death for patients with DCIS as a whole. A subgroup of patients in whom RT can be omitted could not be identified [33 36]. Few studies prospectively evaluated the outcomes after WLE only in a subgroup of DCIS patients. The Eastern Cooperative Oncology Group (ECOG) 5194 trial (lowor intermediate-grade DCIS 62.5 cm or high-grade DCIS 61.0 cm excised with final margins P3 mm) found that after a median follow-up time of 6.7 years, 26 invasive ipsilateral breast events occurred in 565 women with low/intermediate-grade DCIS treated by WLE only [37]. Wong et al. performed a phase II, single-arm, prospective trial in patients with small low- or intermediate grade DCIS, and wide excision with final microscopic margins P1 cm. Among 143 patients treated by WLE only, only six developed an invasive local recurrence as first event within 8 years [38]. Strong imaging markers to guide therapy decisions in DCIS have been lacking. Studies have shown that magnetic resonance imaging (MRI) is more sensitive than mammography for identifying ibc, and in particular high and intermediate grade DCIS [39 44]. The potential role of MRI in the routine diagnostic workup of DCIS is controversial [45,46]. It might be a useful tool to diminish the overtreatment of indolent screen-detected lesions, and prevent the undertreatment of aggressive DCIS and underestimated invasive disease Rationale The question has been raised whether intensive treatment for low-risk DCIS might be considered overtreatment [47]. A substantial number of DCIS lesions will never form a health hazard, particularly if it concerns non- to slow-growing low-grade DCIS. This implies that many women might be unnecessarily going through intensive treatment resulting in a decrease in quality of life and an increase in health care costs, without any survival benefit. The LORD (LOw Risk Dcis) study is a randomised, international, multicentre, open-label, phase III non-inferiority trial that assesses the safety of active surveillance in women with low-risk DCIS. The leading groups are the Dutch Breast Cancer Research Group (BOOG) and the European Organisation for Research and Treatment of Cancer (EORTC). To explore interest in and feasibility of the LORD study, to map standard approach of low-risk DCIS in different countries and to investigate the possible role of MRI in the LORD study, we conducted a survey among EORTC and BOOG centres. Here we present the LORD study protocol and discuss the results of the explorative feasibility survey. 140 The LORD study

143 2. Design 2.1. Objectives of the trial Primary objective The primary objective of the LORD study is to determine whether low-grade DCIS can safely be managed by an active surveillance strategy or that conventional treatment, being either WLE alone, WLE + RT, or mastectomy, and possibly HT, followed by active surveillance, will remain standard of care. Safety will be measured by ipsilateral invasive breast cancer-free percentage at 10 years (10-year iibc-free%). Details on the primary end-point will be discussed in the statistical section Secondary objectives Secondary objectives/end-points of the LORD include: - Patient reported outcomes - Cost-effectiveness - Central collection of imaging data and bio samples for future translational research purposes By integrating clinical, imaging, morphological and molecular data of both retrospective DCIS series and the prospective LORD study, we aim at developing a tool, reliably distinguishing harmless from aggressive screen-detected DCIS, that may help clinicians and women with DCIS to decide between management by active surveillance or more intensive treatment. - Rate of invasive disease at final pathology specimen in the conventional treatment arm - Rate of DCIS grade II/III at final pathology specimen in the conventional treatment arm - Biopsy rate during follow-up - Ipsilateral mastectomy rate - Time to ipsilateral DCIS grade II/III - Time to contralateral DCIS grade I/II/III - Cumulative incidence of contralateral invasive breast cancer - Time to failure of active surveillance strategy in the experimental arm, i.e. time to crossover to conventional treatment, due to any cause - Distant metastases free interval - Overall survival Patient selection criteria Inclusion criteria Women aged 45 years or older with asymptomatic, pure and low-grade DCIS based on vacuum-assisted biopsies (VACB) of microcalcifications only, detected by population-based or opportunistic screening mammography, are eligible if they have an American Society of Anaesthesiologists (ASA) score of 1 2 and life expectancy of more than 5 years. Prior surgery The LORD study 141

144 of the ipsilateral breast because of benign lesions is allowed. Before randomisation, written informed consent must be given according to ICH/GCP and national/local regulations Exclusion criteria Individuals will be excluded from the trial if they have personal history of DCIS or ibc, or if a BRCA1 or BRCA2 mutation was previously identified in their family. Other exclusion criteria are symptomatic DCIS, i.e. DCIS detected by palpation or nipple discharge, bilateral DCIS, synchronous contralateral ibc, LCIS, Paget s disease, or invasive breast disease on cytology or histology and serious disease that precludes definitive surgical treatment Participant recruitment Patients will be recruited at participating trial sites across different countries, i.e. breast centres, highly specialised gynaecologic departments or gynaecological and oncological outpatient units. Patients whose diagnostic core biopsy as standard of care of screen-detected microcalcifications shows histologically confirmed unilateral primary low-grade DCIS of the breast by local pathology will be able to participate in the LORD study. Only if the informed consent form is obtained, and inclusion and exclusion criteria are compliant, the patient can be enrolled into the study. Patients who are not registered prior to any trial-related procedure cannot be accepted for the trial at a later time Diagnostic workup before randomisation (Fig. 1) Population-based or opportunistic screening mammography should reveal lesions consisting of microcalcifications only. A minimum of six cores with VACB of the microcalcifications must be taken, and a marker should be placed at the biopsy site. The biopsy specimen will be considered representative if a substantial amount of microcalcifications is found in two or more of the biopsies or if more than 75% of the microcalcifications is removed in one of the cores, and this should be proven by specimen radiography. The definition of grade I DCIS described by Schnitt and Collins will be adhered to [48]. Key features of low-grade DCIS include: Cytologic features: - Monotonous, uniform, rounded cell population - Subtle increase in nuclear-cytoplasmic ratio - Equidistant or highly organised nuclear distribution - Rounded nuclei with inconspicuous nucleoli - Hyperchromasia may or may not be present 142 The LORD study

145 Architectural features: - Cribiform, micropapillary or solid patterns most frequent - Bridges and arcades, when present, of uniform thickness - Cells polarise around extracellular lumens - Comedo necrosis rare 7 Figure 1. Flow chart of diagnostic workup prior to randomisation. In case of extended (>4 cm) DCIS grade I, VACB from the centre of the lesion and a peripheral part, or two peripheral parts of the lesion must be taken. In case of multicentricity two locations of clusters in different quadrants must be biopsied. Radiological and pathological findings should correlate, i.e. both findings should confirm low-grade DCIS and no suspicion of intermediategrade or high-grade DCIS or ibc should exist. The LORD study 143

146 2.5. Study design The LORD trial will be an international multicentre, phase III, open-label, randomised non-inferiority study. Individuals will be randomised in equal numbers to one of the following arms: active surveillance or standard treatment according to local policy, followed by active surveillance. Study centre will be included in the stratification of this trial. Randomisation will take place within 12 weeks after histologically proven low-grade DCIS on VACB. A computer-based allocation using the minimisation method with a random component will be used for the randomisation procedure. Surgery in the standard arm will take place within 8 weeks after randomisation Study arms The trial will compare standard treatment according to local policy and an active surveillance strategy in patients with low-grade screen-detected DCIS (see Fig. 2): Experimental arm Active surveillance, i.e. monitoring by annual digital mammography for a period of 10 years, and treatment if there is any sign of progression to higher-grade DCIS or invasive cancer Standard arm Standard treatment according to local policy. This can be either WLE alone, WLE + RT, or mastectomy, and HT by Tamoxifen will be allowed. In case invasive disease is found at final pathology after surgery, patients should be treated and followed up according to local policy. The same follow-up scheme will be applied in both study arms, i.e. annual mammography for a period of 10 years Clinical evaluation and follow-up Timing of follow-up visits will be based on the date of registration. Follow-up mammograms will be scheduled annually from the date of registration for 10 years in both study groups. Fig. 3 shows a decision tree to be used for additional diagnostics during follow-up. Biopsy during follow-up will be recommended in case of an increase of >30% of the largest diameter of the index lesion on mammography as compared to the mammography performed one year after the initial VACB. The lesion to act upon must be at least 1 cm in diameter. Biopsy will also be strongly recommended in case of suspicion of malignancy according to the BIRADS criteria of the American College of Radiology [49]. 144 The LORD study

147 Low-risk DCIS R Standard Arm n=620 Experimental Arm n=620 Surgery (WLE or MX) +/- Radiotherapy +/- Hormonal therapy Active Surveillance Annual Mammography for a period of 10 years Annual Mammography for a period of 10 years Figure 2. Flow chart of study design. R = randomisation. WLE = wide local excision. MX = mastectomy Statistical considerations Primary end-point Safety will be measured by ipsilateral invasive breast cancer free percentage at 10 years (10- year iibc-free%). Regional and distant metastases and death from breast cancer in the absence of iibc will also count as events in this definition. This definition is in line with the definition of recurrence free-interval in the adjuvant setting [50]. If the primary DCIS is excised and final pathology concludes invasive disease, this incidence will not count as an event for the primary end-point. These patients will not be excluded from the study and/or analysis and will be followed for events from surgery onwards Hypotheses We have applied the following null and alternative hypotheses (H0 and H1 respectively): H0 (inferiority of active surveillance): 10-year iibc-free% active surveillance arm 10-year iibcfree% standard arm - 10%. H1 ( non-inferiority of active surveillance): 10-year iibc-free% active surveillance arm 10-year iibc-free% standard arm - 5%. The LORD study 145

148 We have re-formulated the design assumptions based on 5-year iibc-free%, presuming a fairly constant event rate over time [33 38,51 58]. We have assumed that the mastectomy rate in the standard treatment arm is 20% and that the 5-year iibc-free% equals 100% for this treatment. Furthermore, we have assumed that the breast-conserving treatment rate in the standard treatment arm is 80% and that the 5-year iibc-free% equals 97% for this treatment. Given these percentages it is presumed that the 5-year iibc-free% in the standard treatment arm equals 97.5%, and the 10-year iibc-free% equals 95% Primary test The 10 year iibc-free% will be estimated based on the Weibull survival model for the time to iibc in each treatment arm separately. The difference between these two estimates will be compared with a critical value of The primary test will be conducted in the per protocol population, consisting of all eligible patients who effectively started in the randomised treatment arm Sample size With a one-sided test for non-inferiority with alpha = and a power of 80%, 930 patients need to be assessed for the primary end-point, and hence need to fulfil a follow-up period of at least 10 years. We anticipate that 25% of all randomised patients will be excluded from the per protocol population, or will have dropped out before the end of the 10-year follow-up period, and therefore 1240 low-grade DCIS patients need to be randomised. Accrual is expected to take 4 years, resulting in an estimated total study duration of 14 years. The power for this trial design fluctuates slightly for different plausible 10-year iibc-free percentages in the standard treatment arm as shown in Table 1. Table 1 also takes into account a non-binding interim analysis for futility (see below) and a parametric Weibull estimation of the 10-year iibc-free% Interim analysis One non-binding interim look for futility is planned when 742 patients have fulfilled a follow-up period of 5 years. This is estimated to occur 7.6 years after accrual of the first patient. The 5-year iibc-free% will be estimated based on the Weibull survival model for the time to iibc in each treatment arm separately. The difference between these two estimates will be compared with a critical value of Simulated operating characteristics of the interim look are illustrated in Table The LORD study

149 Statistical analysis The primary analysis will be based on the per protocol population (see primary test section), and will be supplemented with an intention-to-treat analysis. A multivariate analysis will be performed using a cox proportional hazards regression model. Table 1. Simulated design characteristics (30,000 simulation runs for each setting). As per planned design in grey. Five scenarios for the 10-year ipsilateral invasive breast cancer-free percentage in the standard arm are considered. True 10-yr iibc-free % Null Hypothesis 10-yr iibc-free % (-10%) Alternative Hypothesis 10-yr iibc-free % (-5%) Simulated Power Simulated Significance Level (1-sided) Standard Treatment Active surveillance Active surveillance 99% 89% 94% 89% 0.6% 97% 87% 92% 84% 1.4% 95% 85% 90% 80% 2.5% 93% 83% 88% 77% 3.5% 91% 81% 86% 74% 4.6% The table contains the simulated power and significance level under the planned design for each scenario, taking into account a Weibull estimate of the 10-year ipsilateral invasive breast cancer-free percentage and a non-binding interim test for futility. 7 The LORD study 147

150 Table 2. Simulated operating characteristics of the interim look (30,000 simulation runs for each setting). As per planned design marked in grey. Probability to stop at interim for futility under the alternative under the null under major inferiority Defined as a difference of 13% in 10-yr iibc-free % True 5-yr iibc-free % 5-yr iibc-free % 5-yr iibc-free % 5-yr iibc-free % Experimental arm Experimental arm Experimental arm Standard arm under the alternative under the null under major inferiority 99.5% 0.6% (97.0%) 58% (94.3%) 93% (92.7%) 98.5% 2% (95.9%) 59% (93.3%) 91% (91.7%) 97.5% 4% (94.9%) 60% (92.2%) 90% (90.6%) 94.4% 6% (93.8%) 61% (91.1%) 89% (89.4%) 95.4% 8% (92.7%) 61% (90.0%) 88% (88.3%) Five different scenarios are considered for the 5-year ipsilateral invasive breast cancer-free percentage in the standard arm (first column). The five scenarios in this column correspond to those in Table 1 given a constant event rate over time. Columns 2 4 contain the operating characteristics of the interim look under three different scenario s regarding the difference in 5-year ipsilateral invasive breast cancer-free percentage between the standard and experimental arm. 5-yr iibc-free% = 5-year ipsilateral invasive breast cancer-free percentage. 10-yr iibc-free% = 10-year ipsilateral invasive breast cancer-free percentage. 148 The LORD study

151 3. Feasibility In May 2014 a survey was released to the EORTC and BOOG networks. This survey aimed to explore interest in and feasibility of the LORD trial, to map standard approach of low-risk DCIS in different countries, and to investigate the possible role of MRI in this trial Methods A team of clinical breast cancer experts, which consisted of a surgeon, pathologist, radiologist, radiotherapist, medical oncologist, nurse practitioner, and statistician, developed the survey items Results A total of 53 centres from 14 different countries responded to the survey. Most responding centres were from the Netherlands (36%), Belgium (25%), and France (11%). Other responding centres were from Egypt, Greece, Israel, Italy, Poland, Portugal, Serbia, Slovenia, Spain, Sweden and Turkey Interest and feasibility More than three quarters of the responders were interested in participation in the study (77%). Twelve centres that reported not to be interested were from Belgium (6/13 responding centres), France (3/6), Greece (1/2), Israel (1/1) and Sweden (1/1). Reasons for no interest could be subdivided into three main issues: accrual difficulties (n = 5), not convinced by study design or rationale (5) and administrative issues (1) (1 missing). Further presentation of the survey results is based on the answers of centres that expressed interest (n = 41) and shown in Table 3. The reported numbers resulted in an estimated total of around 400 low-grade screen-detected DCIS patients per year in the interested centres. 90% of the centres believed that low-grade DCIS patients would be willing to participate in the LORD study. Three centres, from Greece, the Netherlands and Turkey, that thought that patients would not be willing to participate, expected that most patients would demand surgery Current management of DCIS Screening Most screening programmes apply a lower age limit of 50 years. In Italy and Portugal a lower age limit of 45 years was reported (four centres), and one centre in Turkey reported a limit of 40 years. The LORD study 149

152 Table 3. Survey results based on the answers of centres that expressed interest (n = 41). n % Interest and feasibility How many patients with screen-detected DCIS do you see in your hospital per year? > What percentage of the screen-detected DCIS lesions are low-grade (grade 1)? 5% % % >25% 9 22 Missing 4 10 Do you think that these low-grade DCIS patients will be willing to participate in this trial? Yes No 3 7 Missing 1 2 Current management of DCIS Is there a population-based screening program in your country? Yes No 4 10 Missing 1 2 If yes, How is the screening program organized? (n = 36) Frequency Biannual Missing 1 3 Location Screening unit Hospital 4 11 Other 2 6 Missing 1 3 Participation rate estimate < 50% % % Missing 7 19 Lower age limit 40 years years years Missing The LORD study

153 Table 3. Continued n % Upper age limit 65 years years years How are patients with DCIS treated in your hospital? (several options are possible) No treatment 0 0 Excision biopsy 8 20 Wide Local Excision Wide Local Excision + Radiotherapy Mastectomy Hormonal therapy Does the work-up scheme which is described in the outline reflect the standard of practice in your hospital? Yes No 5 5 Missing 1 1 Is the work-up of DCIS in your hospital considered standard in your entire country? Yes No 3 7 Are you comfortable to comply to the follow-up flow-chart described in the outline? Yes No 4 10 MRI Are you using MRI in the work-up of low-grade DCIS? Yes, routinely 2 5 Yes, in special cases No, never Is MRI in the work-up of DCIS reimbursed in your country? Yes No Missing 2 5 Would you prefer to add a MRI to the work-up in this trial? Yes No If an additional MRI in the work-up of DCIS leads to a diagnosis of histologically proven high grade DCIS or invasive breast cancer (upgrading) would you then include MRI in your standard work-up? Yes, if MRI leads to an upgrade in at least 5% of patients Yes, if MRI leads to an upgrade in at least 10% of patients Missing The LORD study 151

154 Treatment To assess the current treatment of DCIS, centres were allowed to give multiple answers. The most frequently offered treatment approach of DCIS was WLE followed by RT: 95% of the centres reported to offer this approach as one of the treatment options, which means that 5% of the centres reported not to offer WLE + RT to DCIS patients. A minority of the centres reported to offer HT (all Tamoxifen) (29%). Of the centres that reported to offer HT as one of the treatment options, five were from Belgium, two from Italy, and one from Poland, Portugal, Serbia, Spain and Turkey each Diagnostic workup and follow-up The proposed diagnostic workup and follow-up strategy in the LORD study was endorsed by most centres (85% and 90%, respectively). Centres that reported to employ different diagnostic workup of DCIS (n = 5) explained that they would take less biopsies in the diagnostic process. Centres that reported not to feel comfortable to comply with the follow-up scheme (n = 4), would like to add a mammography 6 months after the first diagnosis, or every 6 months, and would recommend attenuating the definition of progression MRI The majority of the interested centres preferred not to add an MRI to the diagnostic workup in the trial (66%). 4. Discussion leading to rationale and design of the LORD trial 4.1. Considerations and arguments The current debate on overdiagnosis and overtreatment of screen-detected breast cancer and pre-invasive breast lesions urges the need for prospective studies to address these issues. With the increased use of screening mammography, the likelihood of detecting DCIS is considerable with 20% of all screen-detected breast cancer cases. No randomised trial has compared WLE + RT with mastectomy for the treatment of DCIS, but four randomised controlled trials have addressed the value of adding RT to WLE [59 62]. Controversy exists primarily because of the absence of randomised trials documenting that treatment improves survival and quality of life. Because almost all DCIS is excised when detected, uncertainty about the natural history exists, and the clinical significance of screen-detected DCIS is debated. It is sufficiently clear that DCIS represents a heterogeneous group of diseases, and it is time to address the safety of active surveillance in women with screen-detected low-risk DCIS in order to reduce the negative impact of identifying and treating non-lethal disease. 152 The LORD study

155 The LORD study aims to contribute to solve a substantial clinical dilemma in breast cancer diagnosis and treatment: finding the balance between overdiagnosis and undertreatment in individuals with low-risk DCIS of the breast. We believe it is reasonable to assume that active surveillance is nearly as safe as the conventional management of screen-detected low-grade DCIS in terms of the 10-year ipsilateral invasive breast cancer free rate and consequently the related survival. For our sample size calculation we have assumed a slightly decreased 5-year iibc-free rate in the experimental arm, as opposed to the standard treatment arm. Data on the natural history of DCIS are not well established, and therefore we also based our assumptions on data on similar low-risk non-obligate precursor lesions, like LCIS and atypical hyperplasia. We hold the position that our assumed iibc-free rate in the active surveillance arm, in relation to the expected iibc-free rate in the standard treatment arm, is clinically acceptable with broad clinical support [63] Ipsilateral invasive breast cancer-free rate When discussing the trade-off of an acceptable iibc-free rate, the potential benefits of not treating should also be taken into account. Active surveillance does not remove DCIS, and may therefore miss an opportunity to end or delay disease progression. However, despite a slightly higher chance of developing iibc, we concur that active surveillance may save a substantial number of women from going through surgery and radiotherapy, avoiding possible harmful side-effects of treatment, and as a consequence, this may lead to an improved quality of life (cosmetic outcomes, body self-image, self-esteem, etc.). Of course, the potential distress caused by no treatment will be actively evaluated in the LORD study in order to gain a well-balanced insight into patient reported outcomes. Furthermore, the evaluation of health economics data might show that an active surveillance strategy will lead to a decrease in health care costs. Costeffectiveness research will allow patients, health care providers and payers to better understand the true value of each disease-management strategy. One other important argument to assume that the expected iibc-free rates in this low-risk DCIS population are clinically acceptable is provided by the circumstantial evidence that there is a difference in natural history between high-grade and low-grade DCIS. If low-grade DCIS progresses to ibc, this is most likely to be low-grade ibc [16,25,31]. After an active surveillance strategy of a low-grade DCIS lesion, multiple treatment options of the subsequent low-grade ibc will still be feasible and excellent long-term survival outcomes can be preserved. Subsequent high-grade ibc is unlikely to develop from low-grade DCIS, and will more likely be a new primary tumour [16,31,32]. Women managed by an active surveillance strategy, as well as women treated by WLE for DCIS will have a chance of developing new primary breast cancer, but the breast-conserving options might be more limited in the latter group. 7 The LORD study 153

156 4.3. Standard treatment arm The LORD study includes all current DCIS treatment strategies within the standard treatment arm. A multi-arm design was not considered feasible because this would lead to a non-feasible increase in sample size. Furthermore, we concluded from our discussions that the study should reflect the current standard clinical practice as much as possible. We will stratify by centre to strive for an equal contribution of all preferred current practice approaches in both study arms, and moreover, the stratification accounts for small differences in the diagnostic workup and follow-up of the study individuals. The option to exclude mastectomy from the standard treatment arm, because of its deviant iibc-free rate, was abandoned. Either way, some women will end up undergoing a mastectomy in this treatment arm as a result of positive margins. The use of HT in the treatment of DCIS differs among and within the anticipated participating countries. In the Netherlands for example, HT is generally not offered to patients with pure DCIS. Because of the international multi-centre design of the LORD the use of HT is allowed in the standard arm. However, its use is not allowed in patients within the active surveillance arm Age The lower age limit for inclusion in this trial was set at 45 years. In most screening programmes women from the age of 50 years are targeted. However, in our survey, four centres reported a lower age limit of 45 years to be applicable in their screening programme. Women with a history of any grade DCIS younger than 40 or 50 years have an increased risk of developing ibc than older women [7]. We discussed not to exclude potential trial participants between 45 and 50 years, because the division between young and old is arbitrary and variable between studies, and it is believed that women with screen-detected, asymptomatic low-grade DCIS have a favourable long-term outcome as opposed to women with symptomatic disease Central review The LORD trial does not apply pre-randomisation central pathology or imaging review. Literature shows that classifying low-risk in situ lesions is associated with significant interobserver variation [64 68]. This suggests that some true low-grade DCIS lesions are missed, whereas other lesions are classified as DCIS incorrectly. This could be an indication to use pre-randomisation central pathology review by expert pathologists. However, in daily practice treatment decisions are based on the pathology reports assessed by local pathologists. Both downgraded and upgraded lesions will appear to some extent in the study population as well as in the real population. Furthermore, central pathology review involves complicated and expensive logistics owing to enrolment of large numbers of patients in multiple countries and requires long turnaround time for slide review at the central location. In the LORD study we pursue the definition of low-grade DCIS as described in the book Biopsy Interpretation of the Breast written by Schnitt and Collins [48]. Tissue blocks of pre-randomisation 154 The LORD study

157 VACB and resection specimens, as well as digital mammography data will be collected for retrospective central review and future translational research purposes MRI Mammography is the mainstay for diagnosing DCIS. However, MRI of the breast has the ability to enhance additional findings in women when performed after initial clinical evaluation. Yet, the clinical relevance of these additional findings is uncertain [46,69,70]. During protocol development, the possibility of adding pre-randomisation MRI to the diagnostic workup has been discussed. The advantages of pre-randomisation MRI include improved sensitivity and therefore increased ability to detect intermediate-grade or high-grade DCIS, and ibc, which are considered exclusion criteria in the LORD trial. In spite of improved sensitivity a retrospective cohort study showed no association between the perioperative use of MRI and local regional recurrence rate in women undergoing breast-conserving surgery for DCIS [71]. Currently, breast MRI is not routinely performed in patients with DCIS, which was also reflected in the results of our explorative survey. An MRI side-study could give more insights into the role of MRI in the diagnostic workup of screen-detected DCIS. However, the majority of centres that showed interest in participation in the LORD trial were not in favour of adding MRI to the workup in the study. Main reasons were a foreseen high rate in false-positive findings resulting in increased number of diagnostic procedures, psychological distress and costs. In the LORD study, the use of MRI is allowed in the workup of the study individuals before randomisation at the discretion of the physician. The use of MRI after diagnosis of low-grade DCIS on core biopsy is not recommended Statistical considerations The LORD study comprises some non-conventional choices in the statistical design. First, we have applied a special type of non-inferiority design, where the alternative hypothesis corresponds to minor inferiority. A classical non-inferiority design would have had the following H1: H1 classical (non-inferiority of active surveillance): 10-year iibc-free% active surveillance arm 10- year iibc-free% standard arm. 7 However, we expect slightly more events with an active surveillance approach as opposed to excision with or without RT and/or HT in patients with low-grade DCIS, and therefore H1 classical was not considered realistic in this study setting. Second, the primary test in the LORD study will not be based on the relative risk (hazard ratio), but on a fixed difference in absolute risk (10-year iibc-free%). In this trial we expect a low event rate for the primary end-point, and therefore the absolute difference in 10-year event-free% is considered to be of more clinical significance than the relative difference. When the 10-year The LORD study 155

158 iibc-free% in the standard arm would be higher than anticipated the hazard ratio would increase exponentially, rendering a statistical design based on the hazard ratio, e.g. the logrank test, majorly underpowered. The outcome of the trial would then not be in line with the actual difference in 10- year iibc-free% (see Fig. 4). The power of the current proposed statistical test is less sensitive to deviations from the hypothesised 10-year iibc-free% in the standard arm (see Table 1). Third, we chose not to conduct an event-driven trial, because of the expected low primary event rate in both study arms. Instead we will use a follow-up time-driven design. As a result we will greatly depend on achieving the pre-specifi ed follow-up for 75% of patients. Furthermore, we will primarily test the difference between iibc-free% between both study arms at 10-year follow-up instead of 5-year follow-up. Because low-grade DCIS has a long natural history, it is necessary to follow up these patients beyond 5 years for late recurrences, and draw conclusions based on longer follow-up. Figure 3. Decision tree to be used during follow-up. ACR = American College of Radiology. 156 The LORD study

159 Figure 4. Relation between the hazard ratio and the 10-year ipsilateral invasive breast cancer-free percentage in the standard treatment arm, for a fixed difference of 10% in 10-year ipsilateral invasive breast cancer-free percentage between the standard and experimental arm. HR = hazard ratio. 10-yr iibc-free% = 10-year ipsilateral invasive breast cancer-free percentage. 5. Trial status The start of active recruitment is estimated to be in the fourth quarter of Recruitment is estimated to take 4 years. The projected date of study completion will be after an additional follow-up period of 10 years. Conflict of interest statement None declared. The study is funded by project grants from Pink Ribbon ( ) and Alpe d Huzes/KWF (NKI ). The funding sources had no involvement in study design; in the collection, analysis and interpretation of data; in the writing of the report; or in the decision to submit the article for publication. 7 Acknowledgements We would like to thank Gonneke Winter-Warnars, Claudette Loo, Flora van Leeuwen, Michael Schaapveld, Marjanka Schmidt, Eveline Bleiker, Esther Lips, and the ECCO-AACR-EORTC-ESMO Workshop on Methods in Clinical Cancer Research 2013 for their contribution to the study design. The LORD study 157

160 References 1. Ernster VL, Ballard-Barbash R, Barlow WE, Zheng Y, Weaver DL, Cutter G, et al. Detection of ductal carcinoma in situ in women undergoing screening mammography. J Natl Cancer Inst 2002;94: Sørum R, Hofvind S, Skaane P, Haldorsen T. Trends in incidence of ductal carcinoma in situ: the effect of a population-based screening programme. Breast 2010;19: van Steenbergen LN, Voogd AC, Roukema JA, Louwman WJ, Duijm LEM, Coebergh JWW, et al. Screening caused rising incidence rates of ductal carcinoma in situ of the breast. Breast Cancer Res Treat 2009;115: van Luijt PA, Fracheboud J, Heijnsdijk EAM, Heeten den GJ, de Koning HJ, National Evaluation Team for Breast Cancer Screening in Netherlands Study Group (NETB). Nation-wide data on screening performance during the transition to digital mammography: observations in 6 million screens. Eur J Cancer 2013;49: Bluekens AMJ, Holland R, Karssemeijer N, Broeders MJM, den Heeten GJ. Comparison of digital screening mammography and screen-film mammography in the early detection of clinically relevant cancers: a multicenter study. Radiology 2012;265: Vigeland E, Klaasen H, Klingen TA, Hofvind S, Skaane P. Full field digital mammography compared to screen film mammogra phy in the prevalent round of a population-based screening programme: the Vestfold County Study. Eur Radiol 2008;18: Virnig BA, Tuttle TM, Shamliyan T, Kane RL. Ductal carcinoma in situ of the breast: a systematic review of incidence, treatment, and outcomes. J Natl Cancer Inst 2010;102: Bleyer A, Welch HG. Effect of three decades of screening mammography on breast-cancer incidence. N Engl J Med 2012;367: Chu KC, Kramer BS, Smart CR. Analysis of the role of cancer prevention and control measures in reducing cancer mortality. J Natl Cancer Inst 1991;83: de Gelder R, Fracheboud J, Heijnsdijk EAM, Heeten den G, Verbeek ALM, Broeders MJM, et al. Digital mammography screening: weighing reduced mortality against increased overdiag nosis. Prev Med 2011;53: Verbeek ALM, Broeders MJM, Otto SJ, Frachebou J, Otten JDMH, Holland R, et al. Effects of the population screening into breast cancer. Ned Tijdschr Geneeskd 2013;157:A Welch HG, Black WC. Overdiagnosis in cancer. J Natl Cancer Inst 2010;102: Welch HG. Overdiagnosis and mammography screening. BMJ 2009;339, b IKNL CCCTN. Netherlands Cancer Registry. 15. Erbas B, Provenzano E, Armes J, Gertig D. The natural history of ductal carcinoma in situ of the breast: a review. Breast Cancer Res Treat 2006;97: Lopez-Garcia MA, Geyer FC, Lacroix-Triki M, Marchio C, Reis-Filho JS. Breast cancer precursors revisited: molecular features and progression pathways. Histopathology 2010;57: Welch HG, Black WC. Using autopsy series to estimate the disease reservoir for ductal carcinoma in situ of the breast: how much more breast cancer can we find?. Ann Intern Med 1997;127: Lakhani SR, Audretsch W, Cleton-Jensen A-M, Cutuli B, Ellis I, Eusebi V, et al. The management of lobular carcinoma in situ (LCIS). Is LCIS the same as ductal carcinoma in situ (DCIS)? Eur J Cancer 2006;42: The LORD study

161 19. Page DL, Dupont WD. Proliferative breast disease: diagnosis and implications. Science 1991;253: Fitzgibbons PL, Henson DE, Hutter RV. Benign breast changes and the risk for subsequent breast cancer: an update of the 1985 consensus statement. Cancer Committee of the College of American Pathologists. Arch Pathol Lab Med 1998;122: Chen Y-Y, Hwang E-SS, Roy R, DeVries S, Anderson J, Wa C, et al. Genetic and phenotypic characteristics of pleomorphic lobular carcinoma in situ of the breast. Am J Surg Pathol 2009;33: Arpino G, Laucirica R, Elledge RM. Premalignant and in situ breast disease: biology and clinical implications. Ann Intern Med 2005;143: Sanders ME, Schuyler PA, Dupont WD, Page DL. The natural history of low-grade ductal carcinoma in situ of the breast in women treated by biopsy only revealed over 30 years of long-term follow-up. Cancer 2005;103: Van de Vijver MJ, Peterse H. The diagnosis and management of pre-invasive breast disease: pathological diagnosis-problems with existing classifications. Breast Cancer Res 2003;5: Hernandez L, Wilkerson PM, Lambros MB, Campion-Flora A, Nava Rodrigues D, Gauthier A, et al. Genomic and mutational profiling of ductal carcinomas in situ and matched adjacent invasive breast cancers reveals intra-tumour genetic heterogeneity and clonal selection. J Pathol 2012;227: Lagios MD. Heterogeneity of duct carcinoma in situ (DCIS): relationship of grade and subtype analysis to local recurrence and risk of invasive transformation. Cancer Lett 1995;90: Buerger H, Otterbach F, Simon R, Scha fer KL, Poremba C, Diallo R, et al. Different genetic pathways in the evolution of invasive breast cancer are associated with distinct morphological subtypes. J Pathol 1999;189: Roylance R, Gorman P, Hanby A, Tomlinson I. Allelic imbalance analysis of chromosome 16q shows that grade I and grade III invasive ductal breast cancers follow different genetic pathways. J Pathol 2002;196: O Connell P, Pekkel V, Fuqua SA, Osborne CK, Clark GM, Allred DC. Analysis of loss of heterozygosity in 399 premalignant breast lesions at 15 genetic loci. J Natl Cancer Inst 1998;90: Farabegoli F, Champeme M-H, Bieche I, Santini D, Ceccarelli C, Derenzini M, et al. Genetic pathways in the evolution of breast ductal carcinoma in situ. J Pathol 2002;196: Bijker N, Peterse H, Duchateau L, Robanus-Maandag EC, Bosch CAJ, Duval C, et al. Histological type and marker expression of the primary tumour compared with its local recurrence after breastconserving therapy for ductal carcinoma in situ. Br J Cancer 2001;84: Millis RR, Barnes DM, Lampejo OT, Egan MK, Smith P. Tumour grade does not change between primary and recurrent mammary carcinoma. Eur J Cancer 1998;34: Donker M, Litière S, Werutsky G, Julien J-P, Fentiman IS, Agresti R, et al. Breast-conserving treatment with or without radiotherapy in ductal carcinoma in situ: 15-year recurrence rates and outcome after a recurrence, from the EORTC randomized phase III trial. J Clin Oncol 2013;31: Wapnir IL, Dignam JJ, Fisher B, Mamounas EP, Anderson SJ, Julian TB, et al. Long-term outcomes of invasive ipsilateral breast tumor recurrences after lumpectomy in NSABP B-17 and B-24 randomized clinical trials for DCIS. J Natl Cancer Inst 2011;103: Wärnberg F, Garmo H, Emdin S, Hedberg V, Adwall L, Sandelin K, et al. Effect of radiotherapy after breast-conserving surgery for ductal carcinoma in situ: 20 years follow-up in the randomized SweDCIS trial. J Clin Oncol 2014, JCO The LORD study 159

162 36. Cuzick J, Sestak I, Pinder SE, Ellis IO, Forsyth S, Bundred NJ, et al. Effect of tamoxifen and radiotherapy in women with locally excised ductal carcinoma in situ: long-term results from the UK/ ANZ DCIS trial. Lancet Oncol 2011;12: Hughes LL, Wang M, Page DL, Gray R, Solin LJ, Davidson NE, et al. Local excision alone without irradiation for ductal carcinoma in situ of the breast: a trial of the Eastern Cooperative Oncology Group. J Clin Oncol 2009;27: Wong JS, Chen Y-H, Gadd MA, Gelman R, Lester SC, Schnitt SJ, et al. Eight-year update of a prospective study of wide excision alone for small low- or intermediate-grade ductal carcinoma in situ (DCIS). Breast Cancer Res Treat 2014;143: Berg WA, Gutierrez L, NessAiver MS, Carter WB, Bhargavan M, Lewis RS, et al. Diagnostic accuracy of mammography, clinical examination, US, and MR imaging in preoperative assessment of breast cancer. Radiology 2004;233: Boetes C, Mus RD, Holland R, Barentsz JO, Strijk SP, Wobbes T, et al. Breast tumors: comparative accuracy of MR imaging relative to mammography and US for demonstrating extent. Radiology 1995;197: Hata T, Takahashi H, Watanabe K, Takahashi M, Taguchi K, Itoh T, et al. Magnetic resonance imaging for preoperative evaluation of breast cancer: a comparative study with mammography and ultrasonography. J Am Coll Surg 2004;198: Menell JH, Morris EA, Dershaw DD, Abramson AF, Brogi E, Liberman L. Determination of the presence and extent of pure ductal carcinoma in situ by mammography and magnetic resonance imaging. Breast J 2005;11: Kuhl CK, Schrading S, Bieling HB, Wardelmann E, Leutner CC, Koenig R, et al. MRI for diagnosis of pure ductal carcinoma in situ: a prospective observational study. Lancet 2007;370: Lehman CD. Magnetic resonance imaging in the evaluation of ductal carcinoma in situ. J Natl Cancer Inst Monographs 2010; 2010: van der Velden AP Schouten, Schlooz-Vries MS, Boetes C, Wobbes T. Magnetic resonance imaging of ductal carcinoma in situ: what is its clinical application? A review. Am J Surg 2009;198: Margenthaler JA. Magnetic resonance imaging in patients with ductal carcinoma in situ: routine, selective, or not at all? Ann Surg Oncol 2014;21: Alvarado M, Ozanne E, Esserman L. Overdiagnosis and overtreatment of breast cancer. Am Soc Clin Oncol Educ Book 2012, e Schnitt SJ, Collins LC. Biopsy interpretation of the breast. 2nd ed. Lippincott Williams & Wilkins; 2013, a Wolters Kluwer business. 49. D Orsi CJ, Sickles EA, Mendelson EB, Morris EA. ACR BIRADS Atlas, Breast Imaging Reporting and Data System. Reston, VA, American College of Radiology; Hudis CA, Barlow WE, Costantino JP, Gray RJ, Pritchard KI, Chapman J-AW, et al. Proposal for standardized definitions for efficacy end points in adjuvant breast cancer trials: the STEEP system. J Clin Oncol 2007;25: Early Breast Cancer Trialists Collaborative Group, Correa C, McGale P, Taylor C, Wang Y, Clarke M, et al. Overview of the randomized trials of radiotherapy in ductal carcinoma in situ of the breast. J Natl Cancer Inst Monographs 2010;2010: Fisher B, Costantino JP, Wickerham DL, Cecchini RS, Cronin WM, Robidoux A, et al. Tamoxifen for the prevention of breast cancer: current status of the National Surgical Adjuvant Breast and Bowel Project P-1 study. J Natl Cancer Inst 2005;97: The LORD study

163 53. Goodwin A, Parker S, Ghersi D, Wilcken N. Post-operative radiotherapy for ductal carcinoma in situ of the breast. Cochrane Database Syst Rev 2013;11, CD Benson JR, Wishart GC. Predictors of recurrence for ductal carcinoma in situ after breast-conserving surgery. Lancet Oncol 2013;14:e Wang S-Y, Shamliyan T, Virnig BA, Kane R. Tumor characteristics as predictors of local recurrence after treatment of ductal carcinoma in situ: a meta-analysis. Breast Cancer Res Treat 2011;127: Boyages J, Delaney G, Taylor R. Predictors of local recurrence after treatment of ductal carcinoma in situ: a meta-analysis. Cancer 1999;85: Kuerer HM, Albarracin CT, Yang WT, Cardiff RD, Brewster AM, Symmans WF, et al. Ductal carcinoma in situ: state of the science and roadmap to advance the field. J Clin Oncol 2009;27: Morrow M, Strom EA, Bassett LW, Dershaw DD, Fowble B, Harris JAYR, et al. Standard for the management of ductal carcinoma in situ of the breast (DCIS). CA Cancer J Clin 2002;52: Fisher B, Costantino J, Redmond C, Fisher E, Margolese R, Dimitrov N, et al. Lumpectomy compared with lumpectomy and radiation therapy for the treatment of intraductal breast cancer. N Engl J Med 1993;328: Julien JP, Bijker N, Fentiman IS, Peterse JL, Delledonne V, Rouanet P, et al. Radiotherapy in breastconserving treatment for ductal carcinoma in situ: first results of the EORTC randomised phase III trial EORTC Breast Cancer Cooperative Group and EORTC Radiotherapy Group. Lancet 2000;355: Emdin SO, Granstrand B, Ringberg A, Sandelin K, Arnesson LG, Nordgren H, et al. SweDCIS: Radiotherapy after sector resection for ductal carcinoma in situ of the breast. Results of a randomised trial in a population offered mammography screening. Acta Oncol 2006;45: Houghton J, George WD, Cuzick J, Duggan C, Fentiman IS, Spittle M, et al. Radiotherapy and tamoxifen in women with completely excised ductal carcinoma in situ of the breast in the UK, Australia, and New Zealand: randomised controlled trial. Lancet 2003;362: Jenks S. Downgrading cancer definitions: overdiagnosis fuels the discussion. J Natl Cancer Inst 2014;106:dju Page DL, Rogers LW. Combined histologic and cytologic criteria for the diagnosis of mammary atypical ductal hyperplasia. Hum Pathol 1992;23: Sloane JP, Amendoeira I, Apostolikas N, Bellocq JP, Bianchi S, Boecker W, et al. Consistency achieved by 23 European pathologists from 12 countries in diagnosing breast disease and reporting prognostic features of carcinomas. European Commission Working Group on Breast Screening Pathology. Virchows Arch 1999;434: Elston CW, Sloane JP, Amendoeira I, Apostolikas N, Bellocq JP, Bianchi S, et al. Causes of inconsistency in diagnosing and classifying intraductal proliferations of the breast. European Commission Working Group on Breast Screening Pathology. Eur J Cancer 2000;36: O Malley FP, Mohsin SK, Badve S, Bose S, Collins LC, Ennis M, et al. Interobserver reproducibility in the diagnosis of flat epithelial atypia of the breast. Mod Pathol 2006;19: Jain RK, Mehta R, Dimitrov R, Larsson LG, Musto PM, Hodges KB, et al. Atypical ductal hyperplasia: interobserver and intraobserver variability. Mod Pathol 2011;24: Morrow M. Magnetic resonance imaging in the breast cancer patient: curb your enthusiasm. J Clin Oncol 2008;26: The LORD study 161

164 70. Elshof LE, Rutgers EJT, Deurloo EE, Loo CE, Wesseling J, Pengel KE, et al. A practical approach to manage additional lesions at preoperative breast MRI in patients eligible for breast conserving therapy: results. Breast Cancer Res Treat 2010;124: Pilewskie M, Olcese C, Eaton A, Patil S, Morris E, Morrow M, et al. Perioperative breast MRI is not associated with lower locoregional recurrence rates in DCIS patients treated with or without radiation. Ann Surg Oncol 2014;21: The LORD study

165 The LORD study 163 7

166

167 CHAPTER 8 General discussion, ongoing research and future prospects

168 General discussion and clinical implications The goal in treating women with ductal carcinoma in situ (DCIS) is to prevent the development of invasive breast cancer and ultimately to prevent breast cancer mortality. After having treated many women with DCIS for the past decades, opportunity to evaluate the long-term follow-up outcomes in this patient population has become widely present. Large epidemiological studies, including the studies presented in this thesis (Chapters 4, 5 and 6), have contributed to our knowledge of the risk of disease recurrence and breast cancer mortality among women treated for DCIS In the Netherlands, we are in a unique position to combine accurate information from the Netherlands Cancer Registry, 17 the nationwide network and registry of pathology (PALGA), 18 the National Breast Cancer Screening Programme, 19 and Statistics Netherlands. Because of these well-defined and well-organised registries with high and accurate coverage of data, we were able to study a large population-based cohort of women diagnosed with DCIS. The inclusion of women treated for DCIS between 1989 and 2004 may be a limitation in terms of applicability to today s patients, but is necessary to study long-term outcome. Further, one of the strengths of our studies is that we provided accurate risk estimates for subsequent invasive breast cancer, while many previous reports did not discriminate between subsequent invasive and non-invasive breast events. In the Netherlands, DCIS has always been treated by breast-conserving surgery with or without adjuvant radiotherapy, or by mastectomy, while hormonal therapy was never recommended. This implicates that in our studies we were able to evaluate the effect of local treatment on risk of subsequent ipsilateral and contralateral invasive breast cancer and cause-specific mortality without confounding by endocrine treatment. This may be different from the clinical practice in other countries, for instance the United States, where tamoxifen is approved as part of standard DCIS treatment and where information about the application of this treatment is often missing in large studies. 7,12,13 Another strength of our studies is that information about surgical treatment is complete as a result of our extensive examination of the data provided by the Netherlands Cancer Registry and PALGA. Further, in our analysis of subsequent invasive breast cancer risk we were able to take into account that some women initially treated by breast-conserving surgery underwent mastectomy during follow-up. An important limitation with regard to our treatment information is the potential of confounding by indication, because the allocation of DCIS treatment was not randomized in our studies. Based on the available evidence we estimate the 10-year risk of developing ipsilateral invasive breast cancer to be approximately 10% after breast-conserving surgery alone, 5% after breastconserving surgery followed by radiotherapy, and 1% after mastectomy (Chapter 4). 1,3,5-11,20 Another clinical relevant observation was that the absolute risk of contralateral invasive breast cancer among women treated for DCIS was low, i.e. approximately 0.5% per annum (Chapter 4) General discussion

169 This implies that prophylactic contralateral mastectomies for unilateral DCIS in women without hereditary breast cancer risk should be discouraged. The difference in ipsilateral invasive breast cancer risk between the various treatment strategies does not significantly affect breast cancer mortality (Chapter 5). 12,20 We estimate the 10-year risk of dying from breast cancer after DCIS treatment to be 2% for women younger than 50 at diagnosis, and 1% for women aged 50 years or older (Chapter 5). 12 DCIS patients aged 50 years or older at diagnosis had lower all-cause mortality than the general female population, which likely reflects differences in health-related behaviour (Chapter 5). 15,16,21-23 Also a recent study implicates that DCIS patients should be aware that their health-related behaviour is associated with mortality outcome. 24 Despite large efforts in the field there still remain a lot of uncertainties about the long-term implications of a diagnosis of DCIS. Consequently, the road to the optimal management strategy for women with DCIS is paved with controversies. As accurate predictive markers to distinguish indolent from aggressive DCIS are lacking current DCIS management guidelines recommend treating all DCIS patients. Clinical factors associated with DCIS progression to invasive breast cancer and breast cancer mortality have not been consistently studied and identified. However, younger age (younger than years) is consistently associated with increased subsequent invasive breast cancer risk and breast cancer mortality (Chapter 4 and 5). 12,20 Also black ethnicity was found to be an important risk factor for subsequent invasive disease and death from breast cancer in American studies. 12,25,26 Method of detection, on the other hand, seems to have a very limited independent prognostic role (Chapter 6). 1,27-29 Further, it is uncertain to what extent DCIS grade, lesion size and margin status after excision are associated with risk of subsequent invasive breast cancer and breast cancer mortality (Chapter 4, Chapter 5). 30,31,32 Recently two prioritized research agendas were developed listing the priorities in DCIS management based on different stakeholders perspectives (Table 1). 33,34 The main challenges remain a better understanding of the biology of the initiation and progression of DCIS, to develop clinical tools to distinguish harmless from hazardous DCIS and guide management decisions, and to effectuate a paradigm shift in how patients and providers perceive a diagnosis of cancer precursors. In the following section, we will discuss preliminary results of our ongoing research and consider future steps towards finding the balance between overtreatment and undertreatment of DCIS. 8 General discussion 167

170 Table 1. Most critical research questions Topic Multisite databases Risk-stratification Prioritized future research need Collection of consistent and detailed data on the clinical, pathological, radiological, and molecular characteristics of DCIS through the creation of multisite databases of DCIS that would include annotated specimen and imaging repositories. 33 Investigation and validation of combinations of new and existing clinical, radiological, pathological, and molecular factors to improve risk stratification of DCIS patients and thus to identify the optimal therapy for each individual. Ease of use, predictive ability, reproducibility, and generalizability are important components of prognostic model development. 33 Develop and validate risk-stratification models. 34 Imaging Communication and decision-making Comparative sensitivity and specificity of breast MRI, mammography, and other preoperative imaging evaluations for detecting occult invasive breast cancer. 34 Development and use of standardized reporting methods and terminology for DCIS detection and diagnosis across all disciplines. 33 Comparative effectiveness of different approaches to communicating the diagnosis of DCIS to the patient. 34 Research on patient provider communication, informed consent (at the time of screening), patient preferences, and decision-making concerning the diagnosis and treatment of DCIS. Decision aids should be further developed, evaluated for their impact on quality of care, and integrated into clinical practice. 33 Comparative effectiveness of decision-making tools compared with usual care. 34 Patient-centred outcomes Investigations of the impact a diagnosis and treatment of DCIS has on the quality of life. 33 Identify most important patient-centred outcomes for women diagnosed with DCIS. 34 Management strategies Investigations into the comparative effectiveness of the methods of treatment for DCIS. 33 Assess effect of DCIS management strategies on rates of invasive cancer. 34 Compare safety and effectiveness of a management strategy involving no immediate treatment (i.e., observation/active surveillance) vs. immediate treatment with surgery, RT, and/or medical therapy. 34 Determine whether safety and effectiveness of DCIS management strategies differ depending on variations in clinical, pathologic, and genomic presentations of DCIS. 34 Compare safety and effectiveness of partial-breast RT vs. whole-breast RT. 34 Assess effect of DCIS management strategies on comorbid conditions General discussion

171 Ongoing research DCIS progression DCIS is generally regarded as a (non-obligate) precursor of invasive breast cancer. Several arguments support this view: 1) invasive breast cancer is often found in areas of DCIS, 2) the risk factors for developing DCIS are similar to those for developing invasive breast cancer, and 3) genetic markers found in DCIS are similar to those detected in invasive breast cancer. 35,36 Nevertheless, clinical behaviour of DCIS and the biology of progression to invasive breast cancer are still poorly understood, because there is only limited information on the long-term natural history of DCIS. Estimates of the proportion of women diagnosed with DCIS who might develop invasive cancer without treatment range from 15% to 50%. 37 Considerable effort has been put into defining characteristics of low-risk and high-risk DCIS in treated DCIS patients, but several challenges in this quest should be acknowledged. First, a well-defined population-based DCIS cohort is necessary to establish a dataset of clinically informative samples with tumour, immune microenvironment and phenotyping data. Second, long-term clinical follow-up is imperative because the progression to invasive breast cancer can take a long time. 38 Additionally, the low event rate of ipsilateral invasive breast cancer after current treatment requires a large sample size. Third, integration of genomic, epigenomic, and transcriptional data with clinicopathological features is needed because each entity alone might not be sufficient to predict progression to invasive breast cancer. Fourth, as most DCIS samples are stored as formalin-fixed paraffin-embedded (FFPE) tissue blocks and need to be micro-dissected for RNA and DNA extraction, state-of-the-art technology is required to analyse low inputs of poor quality FFPE extracted DNA and RNA. Case-control studies offer a unique and efficient approach to address these challenges. 39 Two studies on DCIS patients treated by breast-conserving surgery alone Within our population-based DCIS cohort (this thesis) we have initiated pathological and molecular studies to improve our understanding of DCIS progression. Our aims were 1) to evaluate the likelihood that a subsequent ipsilateral invasive breast cancer originates from a primary DCIS (case-case comparison) and 2) to evaluate associations of patient and DCIS characteristics with risk of developing subsequent ipsilateral invasive breast cancer (nested case-control comparison). Because there are too few cases of untreated DCIS, we have focused on the subgroup of patients who were treated with breast-conserving surgery alone. In this group of patients the remaining breast tissue has not been affected by radiation. We previously found that 8 Ongoing research 169

172 the cumulative incidence of ipsilateral invasive breast cancer is higher than that of contralateral invasive breast cancer in women treated with breast-conserving surgery alone (Chapter 4). This may be caused by the presence of remnants of DCIS or by an unknown early developmental factor also causative to the primary lesion. Alternatively, it cannot be excluded that surgery alone causes grafting by displacement of neoplastic in situ cells resulting in an increased risk of developing ipsilateral invasive breast cancer. Nested case-case and case-control comparisons We conducted detailed case-case and case-control studies nested in the cohort described in this thesis. First, we selected all DCIS patients treated by breast-conserving surgery alone between 1989 and 2004 in the Netherlands based on data from the nationwide Netherlands Cancer Registry. In addition, we used information from PALGA to verify treatment and followup, and eventually to request tumour samples. From an initial group of 2,654 women with DCIS treated by breast-conserving surgery, 330 women developed an ipsilateral invasive breast cancer and 155 women developed a contralateral invasive breast cancer (median follow-up 12.9 years; Figure 1). We identified 316 women who were diagnosed with a subsequent ipsilateral invasive breast cancer as first invasive breast cancer and we considered these women as cases in our studies. For our nested-case control study we defined controls as DCIS patients, treated by breast-conserving surgery alone, who had survived, without being diagnosed with ipsilateral or contralateral invasive breast cancer, for at least as long as the interval between the DCIS diagnosis and the date of diagnosis of the ipsilateral invasive breast cancer of the corresponding case. The matching criteria used were age at DCIS diagnosis and follow-up interval. For each case at least two controls with available pathology reports were identified. To achieve this, a total number of 1,292 controls were selected and screened. During this process we also identified additional subsequent ipsilateral invasive cases (as first invasive event) and DCIS recurrences. Eventually, we requested 58 laboratories to provide DCIS samples of 1,101 women and invasive breast cancer samples of 320 women. The selection and collection of this case-control group has lasted more than two years and has resulted in a collection of 1,063 DCIS samples (886 unique patients) and 252 invasive breast cancer samples (244 unique patients). A huge effort has been put into the review of all these samples by dedicated breast pathologists within the Netherlands Cancer Institute and this has resulted in the collection of 158 complete case-case pairs (defined as a pure DCIS sample and a subsequent ipsilateral invasive breast cancer sample of the same patient) and 214 complete case-control sets (defined as at least one pure DCIS sample of a patient without subsequent invasive breast cancer [control] for each pure DCIS sample of a patient with subsequent invasive breast cancer [case]) (Figure 1). 170 Ongoing research

173 Case-case comparison: Comparison of primary pure DCIS and subsequent invasive breast cancer Previous studies have claimed that DCIS and invasive breast cancer are very much alike. 36,40,41 However, almost all of these previous studies compared DCIS with synchronous invasive breast cancer in the same patient and this lesion has already proven to possess the tool-kit to become invasive. Since the majority of DCIS lesions are indolent and will never progress to invasive breast cancer, we feel it is crucial to focus on the beginning of the progression chain and take into account a progression phase between pure DCIS and the development of invasive breast cancer (Figure 2). Therefore, to understand what is needed for progression, studies should focus on the comparison of primary pure DCIS and subsequent ipsilateral invasive breast cancer. In our case-case comparison we aim to examine whether subsequent invasive breast cancers originate from primary pure DCIS lesions. Further, we want to explore which characteristics differ between primary DCIS and subsequent invasive breast cancer. In addition, we aim to assess which dissimilarities can be found when studying pure DCIS and subsequent invasive breast cancer versus synchronous DCIS and invasive breast cancer (Figure 2). To this end, we completed pathology review and immunohistochemistry (IHC) assessment of 158 primary DCIS and 158 matched subsequent ipsilateral invasive breast cancers (Figure 1), as well as of any synchronous DCIS surrounding this invasive breast cancer. 8 Ongoing research 171

174 Population-based DCIS cohort Treatment by breast conserving surgery alone N=2,654 Contralateral invasive breast cancer N=155 Ipsilateral invasive breast cancer (iibc) N=330 DCIS Cases DCIS Controls First case-control selection iibc as first invasive breast cancer No subsequent invasive breast cancer N=316 N=1,292 (no unique patients) Exclusions: Inclusion case-control study: Complete matched casecontrol sets DCIS Cases N=214 DCIS Controls N=510 - Not selected as control in case-control study - Mastectomy during follow-up - Invasive breast cancer during follow-up - Material not available - No pure DCIS after pathological review - No invasive breast cancer after pathological review Inclusion case-case study: Complete primary DCISsubsequent invasive breast cancer sets (same patient) DCIS Cases iibc Cases N=158 N=158 Figure 1. Set up of nested case-case and case-control comparisons within Dutch population-based DCIS cohort. Cases were defined as DCIS patients, treated by breast-conserving surgery alone, who were diagnosed with a subsequent ipsilateral invasive breast cancer as first invasive breast cancer. Controls were defined as DCIS patients, treated by breast-conserving surgery alone, who had survived, without being diagnosed with ipsilateral or contralateral invasive breast cancer, for at least as long as the interval between the DCIS diagnosis and the date of diagnosis of the ipsilateral invasive breast cancer of the corresponding case. 172 Ongoing research

175 Figure 2. Primary pure DCIS and subsequent invasive breast cancer versus synchronous DCIS and invasive breast cancer. We assessed histological grade and we selected 6 IHC markers for the assessment of protein expression based on the literature: ER, PR, HER2, Ki67, p16, and p53. IHC stained slides were fi rst scanned and subsequently the digital images were scored by a team of 7 observers, including 5 pathologists. Due to unreliable staining results we had to exclude Ki67 from the analysis. Probably these unreliable staining results were caused by degradation of expression and detection of the marker with time (loss of antigenicity within the paraffi n block). 42 Further, as tissue handling is more critical for the assessment of Ki-67 compared with ER and HER2, variation in fi xative, time to fi xation and duration of fi xation may have played a major role. 43 Preliminary results include data of 158 DCIS samples with 158 subsequent invasive breast cancer samples of the same patient. These 158 DCIS cases appeared to be a good representation of the initial selection of cases (Table 2). Of these 158 cases, 83 patients had subsequent invasive breast cancer embedded in DCIS (synchronous DCIS). Almost all (97%) subsequent ipsilateral invasive breast cancers were located in the same region as the initial pure DCIS lesion. In addition, the concordance of histological grade between primary pure DCIS and subsequent ipsilateral invasive breast cancer was 94%. 8 Ongoing research 173

176 Table 2. Clinical variables of the DCIS patients included in the case-case comparison Inclusion case-case comparison Initial case selection p-value n= 158 n = 316 Age at DCIS diagnosis 57 (32-87) 58 (30-89) 0.25 a Year of DCIS diagnosis 1996 ( ) 1996 ( ) 0.92 a Time to invasive breast cancer 0-5 years 75 (47%) 158 (50%) 6-10 years 59 (37%) 112 (35%) years 18 (11%) 37 (12%) >15 years 6 (4%) 9 (9%) 0.85 b Method of DCIS detection Screen-detected 69 (44%) 135 (43%) Not screen-detected 73 (46%) 155 (49%) Unknown 16 (10%) 26 (8%) 0.49 b a p-value calculated by unpaired T-test. b p-value calculated by Fisher s exact test. For 145 DCIS patients with subsequent invasive breast cancer, IHC data of five markers were collected (Table 3). Grade and IHC marker expression identified two different aggressive primary DCIS subgroups (Figure 3): a group that consisted of DCIS lesions that were high grade, hormone receptor negative, HER2 positive, and had a high expression of p16 and p53, and a group that included DCIS lesions that were low grade, hormone receptor positive, HER2 negative, and had a low expression of p16 and p53. It has been hypothesized that low grade DCIS may progress to low grade invasive carcinoma over a fairly long time period, whereas high grade DCIS may progress to high grade invasive carcinoma over a shorter period of time. 38,44,45 There was, however, no correlation between DCIS subgroup and time to progression in our preliminary analysis (Figure 3). We noticed that primary DCIS was more often HER2 positive than subsequent invasive breast cancer (Table 3). Sixteen of 44 (36%) HER2 positive DCIS cases had a HER2 negative invasive recurrence. In addition, 16% of the subsequent invasive breast cancers had a different surrogate intrinsic subtype than the primary DCIS lesion, which was mainly caused by the loss of HER2 (Figure 4). When comparing synchronous DCIS with invasive breast cancer we did not detect the same changes in biology (Table 4). The concordance of HER2 status between primary DCIS and subsequent invasive breast cancer was 87%, whereas for synchronous DCIS and invasive breast cancer the concordance was 99%. Further, no differences in the surrogate intrinsic subtype were observed when comparing synchronous DCIS with invasive breast cancer (Figure 4). 174 Ongoing research

177 Table 3. Comparison of grade, IHC marker expression and surrogate intrinsic subtypes between primary pure DCIS and subsequent invasive breast cancer Number of patients (%) Primary pure DCIS Subsequent IBC Agreement p-value a Total 145 (100) 145 (100) Nuclear grade Low (Grade 1 and 2) 111 (77) 95 (66) High (Grade 3) 34 (23) 47 (32) N/A 0 3 (2) 71% Surrogate intrinsic subtype Lum A-like 89 (61) 79 (54) Lum B/ HER2 neg-like 6 (4) 24 (17) Lum B/ HER2 pos-like 20 (14) 16 (11) HER2-like 24 (17) 15 (10) Triple negative 5 (3) 10 (7) N/A 1 (1) 1 (1) 64% Surrogate intrinsic subtype ER+ HER2-95 (66) 103 (71) HER2+ (ER- PR-) 44 (30) 31 (21) Triple negative 5 (3) 10 (7) N/A 1 (1) 1 (1) 84% ER Negative 29 (20) 25 (17) Positive 116 (80) 120 (83) N/A % PR Negative 56 (39) 59 (41) Positive 87 (60) 86 (59) N/A 2 (1) 0 73% HER2 Negative 100 (69) 114 (79) Positive 44 (30) 31 (21) N/A 1 (1) 0 87% p16 Low 75 (52) 84 (58) High 69 (48) 60 (41) N/A 1 (1) 1 (1) 54% p53 WT p53 90 (62) 85 (59) 8 mut p53 53 (37) 60 (41) N/A 2 (1) 0 67% a Comparisons between primary pure DCIS and subsequent invasive breast cancer performed by marginal homogeneity tests. IBC = invasive breast cancer; N/A = not applicable; WT = wild-type; mut = mutant. Ongoing research 175

178 Figure 3. Primary DCIS subgroups based on grade and IHC marker expression. Figure 4. Comparison of surrogate intrinsic subtype a of pure DCIS with subsequent invasive breast cancer and synchronous DCIS with invasive breast cancer. Example of the fi rst 30 patients. a Immunohistochemical surrogate for intrinsic subtypes based on gene expression data. 176 Ongoing research

179 Table 4. Comparison of grade, IHC marker expression and surrogate intrinsic subtypes between synchronous DCIS and invasive breast cancer Number of patients (%) Synchronous DCIS IBC Agreement p-value a Total 83 (100) 83 (100) Nuclear grade Low (Grade 1 and 2) 61 (74) 59 (71) High (Grade 3) 22 (27) 24 (29) 81% Surrogate intrinsic subtype Lum A-like 50 (60) 48 (58) Lum B/ HER2 neg-like 6 (7) 10 (12) Lum B/ HER2 pos-like 8 (10) 11 (13) HER2-like 13 (16) 8 (10) Triple negative 5 (6) 5 (6) N/A 1 (1) 1 (1) 83% Surrogate intrinsic subtype ER+ HER2-56 (68) 58 (70) HER2+ (ER- PR-) 21 (25) 19 (23) Triple negative 5 (6) 5 (6) N/A 1 (1) 1 (1) 98% ER Negative 18 (22) 13 (16) Positive 63 (76) 69 (83) N/A 2 (2) 1 (1) 95% PR Negative 33 (40) 34 (41) Positive 49 (59) 48 (58) N/A 1 (1) 1 (1) 80% HER2 Negative 62 (75) 63 (76) Positive 21 (25) 19 (23) N/A 0 1 (1) 99% p16 Low 45 (54) 48 (58) High 38 (46) 34 (41) N/A 0 1 (1) 68% p53 WT p53 90 (62) 85 (59) mut p53 53 (37) 60 (41) N/A 2 (1) 0 67% a Comparisons between synchronous DCIS and invasive breast cancer performed by marginal homogeneity 8 tests. IBC = invasive breast cancer; N/A = not applicable; WT = wild-type; mut = mutant. Ongoing research 177

180 Intralesional heterogeneity During assessment of the IHC stainings, we frequently observed variable marker expression within DCIS lesions (Figure 5a). To explore whether the change in surrogate intrinsic subtype between primary DCIS and subsequent invasive breast cancer could be due to the existence of multiple variants within one primary DCIS lesion, we identifi ed 91 DCIS lesions consisting of more than 10 ducts. For each of these extensive DCIS lesions, 10 randomly chosen ducts were assessed individually. a b Figure 5. Assessment of DCIS intralesional heterogeneity. a Representative example of variation in HER2 between two DCIS ducts. Ten individual DCIS ducts were selected within every DCIS lesion (see B). An IHC marker was considered variably expressed when at least 20% of the DCIS lesion had different expression levels compared to the majority of the lesion; b Matrix table of the IHC scores of a random sample of 10 patients for the analysis of intralesional heterogeneity involving 10 individual DCIS ducts. Blue is low protein expression and red is high protein expression. 178 Ongoing research

181 ER expression was homogeneous within the DCIS lesions and variable expression was seen in only 2% of the patients. With respect to PR and HER2 expression, variable expression was seen in 19% and 22% of the patients, respectively (Figure 5b). Most DCIS lesions consisted of a single surrogate intrinsic subtype (92%; Figure 6). Among the few cases with DCIS consisting of multiple subtypes we found that the subtype of most subsequent invasive breast cancers could be explained by a specifi c DCIS variant (Figure 7). Figure 6. Assessment of DCIS intralesional heterogeneity - surrogate intrinsic subtype. 8 Figure 7. Comparison of surrogate intrinsic subtype between heterogeneous primary pure DCIS and subsequent invasive breast cancer. Ongoing research 179

182 The preliminary results of our case-case comparison suggest that most subsequent ipsilateral invasive breast cancers (>90%) were recurrences, instead of second primary tumours, and that most recurrences represented regrowth of residual disease. However, we are awaiting molecular data to confirm this finding. Further, invasive breast cancer and synchronous DCIS seem to be more alike than primary pure DCIS and subsequent IBC, which is supported by previous studies. 46,47 Moreover, we detected differences between pure DCIS and subsequent invasive breast cancer, which were not observed by comparison of invasive breast cancer and synchronous DCIS, and vice versa. Specifically, we found loss of HER2 expression during progression from DCIS to invasive breast cancer in more than one third of HER2 positive DCIS. This suggests that HER2 expression is important for DCIS development, but not so much for invasive outgrowth. For 16% of the DCIS patients who developed subsequent invasive breast cancer the surrogate intrinsic subtype of the subsequent invasive breast cancer differed from the subtype of the DCIS lesion. This was mainly caused by HER2 loss. One might think that this reflects the presence of multiple DCIS variants with different surrogate subtypes (intralesional heterogeneity). However, we demonstrated that relatively few (<10%) of primary DCIS lesions consisted of different subtypes, suggesting that, based on IHC data, intralesional heterogeneity does not play a major role in the progression to invasive disease. In the near future we will assess whether a primary DCIS has similar genomic alterations as the subsequent ipsilateral invasive cancer by using low-coverage copy number sequencing and mutation analysis, and evaluate whether clinical, histological and immunohistochemical DCIS characteristics (cytonuclear features, grade, ER, PR, Ki-67, HER2 status) predict similarity of copy number aberration profile and mutation profile between DCIS and subsequent invasive breast cancer. Nested case-control study Over the past few years, several studies have tried to find markers that could predict local recurrence or progression in DCIS patients. Several histopathological DCIS characteristics such as lesion size, surgical margins, nuclear grade, architectural subtype, and presence of necrosis have been identified as possible predictors of recurrence. However, other studies reported contradictory results and most studies did not differentiate between in situ and invasive recurrences because of low event rates. 30,31,48 Furthermore, a multitude of markers have been studied to identify low-risk vs. high-risk subgroups of DCIS by IHC. 49,50 As most of these studies involved relatively small patient series of women treated by breast conserving surgery, and because progression of DCIS is a relatively rare event, the power of the studies to detect markers associated with disease progression was limited. Hence, this asks for analysis of large, well characterized DCIS series. To this end, we will evaluate the association of clinical and centrally revised histopathological and immunohistochemical DCIS characteristics, as well as molecular markers, with the risk of developing invasive breast cancer. While it is unfeasible to measure 180 Ongoing research

183 these markers on our entire population-based cohort, we are conducting a nested case-control study because this design offers a unique and efficient approach to study such markers. Ongoing analysis of our nested case-control study includes 214 DCIS patients who developed subsequent ipsilateral invasive breast cancer (cases) and 510 matched DCIS patients who were of the same age and similarly treated but did not develop subsequent invasive breast cancer (controls) (Figure 1). Pathology review consisted of the assessment of morphological characteristics, including growth pattern, nuclear grade, necrosis and calcifications, and microenvironmental characteristics, including periductal fibrosis and the presence of lymphocytes. For the assessment of protein expression the same IHC markers were selected as for the case-case comparison: ER, PR, HER2, p16, and p53. Currently, we are systematically reviewing the literature for markers that have been associated with DCIS progression. This systematic review will serve as a starting point for our molecular studies. Subsequently, we will perform conditional logistic regression analysis to determine univariable- and multivariable-adjusted odds ratios for the various clinicopathological and molecular characteristics among cases compared with controls. Genomic characterization of DCIS To further discriminate between indolent and aggressive DCIS and to provide possible targets for tailored interventions, molecular studies are needed. While for invasive breast cancer most common key driver mutations (those that are causally implicated in carcinogenesis) have been identified, for DCIS mutations that drive prognosis have not been identified yet. Further, unique combinations of mutation types generated by different mutational processes (mutational signatures) might differentiate between aggressive and indolent DCIS. The comparison of driver and mutational signature data between invasive breast cancer and DCIS may also provide better understanding of the biology and trajectory of DCIS. In addition, studying methylation and transcriptomic profiles of non-progressing and progressing DCIS may help to understand whether there are differences in genes and pathways involved in conferring disease aggression in DCIS breast tumorigenesis. Within the scope of our case-case and case-control study we extracted both DNA and RNA from laser micro-dissected tissue fragments of our formalin-fixed paraffin-embedded (FFPE) samples (Figure 8). In several explorative studies we showed that the quality of our obtained DNA and RNA is sufficient to perform reliable molecular profiling. At present, 200 DCIS samples (100 of our cases and 100 of our controls ; Figure 1) and 100 subsequent invasive breast cancer samples (i.e. ipsilateral invasive breast cancer cases; Figure 1) are included in our molecular studies. The obtained DNA will be used to run a targeted breast cancer panel including genes most frequently mutated in breast cancer. Furthermore, we will assess copy number variations and methylation by performing low coverage DNA sequencing and multiplex ligation-dependent probe amplification, respectively. Additionally, the obtained RNA will be used for transcriptome analysis to identify genes differentially expressed between high-risk and low-risk DCIS. 8 Ongoing research 181

184 Figure 8. Laser micro-dissection of DCIS tissue samples. Credit: Lindy Visser. Immune microenvironment The immune microenvironment plays a critical role in invasive breast cancer progression and response to therapy. 51,52 Yet, little is known about the host immune response in DCIS and how this response evolves during progression from DCIS to invasive breast cancer. It is of interest whether the extent and composition of the immune microenvironment of DCIS predict which patients will develop invasive breast cancer. Further, if high-risk DCIS can be identifi ed via immune microenvironment signatures, this may offer an opportunity to fi nd immune-oncology targets for therapeutic intervention. To characterize the role of the immune microenvironment as determinant of DCIS clinical biology, we need to fi rst explore which immune cell subsets are present in DCIS patients by using conventional IHC and multiplex imaging. The most promising and technically suitable immune markers should be identifi ed and multiplex infl ammation marker panels should be optimized. Next, the potential key distinguishing immune microenvironment markers and signatures need to be confi rmed in validation studies. 182 Ongoing research

185 Future challenges and prospects Risk stratification For many years classification of DCIS lesions was essentially an academic exercise because all patients diagnosed with DCIS were treated with mastectomy. Today, the number of intervention strategies has increased and include active surveillance, breast-conserving surgery alone, breast-conserving surgery with radiotherapy, and mastectomy. Moreover, in the United States, endocrine therapy is offered to hormone-receptor positive DCIS patients, unless bilateral mastectomy is performed. With our current concerns regarding overtreatment of many DCIS, and infrequent (but equally concerning) potential of undertreatment at the other extreme, the development and validation of risk-stratification tools has become one of the most critical DCIS research priorities (Table 1). Importantly, such tools should be easy to use, applicable to a well-defined group of DCIS patients and able to accurately predict long-term risks of negative outcomes (invasive breast cancer, breast cancer-related mortality and overall mortality; Figure 9a). The most promising molecular markers should be combined in an easy to use clinical assay, as deep sequencing is not feasible for daily clinical routine. Ultimately, the risk-stratification tool should aid DCIS patients and health professionals in discussing the optimal intervention strategy for each individual, analogous to current tools for invasive breast cancer (e.g. Adjuvant! Online; Figure 9b). To date, no reliable clinical tool is available. Its development requires integration of clinical, radiological, pathological and molecular data as well as long-term follow-up outcome data of many DCIS patients. Our nationwide DCIS cohort, presented in this thesis, provides unique opportunities for such an integrative approach. It is based on well-organized registries (i.e. the Netherlands Cancer Registry, the national pathology network PALGA, the Dutch screening organizations, and Statistics Netherlands) in the Netherlands and the possibility to link these datasets. As discussed before, a main obstacle in DCIS research is that the long-term natural history of DCIS cannot be studied because of surgical and radiotherapeutic interventions. As such, risk stratification tools will be developed based on data from DCIS patients treated by breastconserving surgery with or without radiotherapy. As a result, the tool will hopefully aid patients treated by breast-conserving surgery in discussing the risks and benefits of receiving radiotherapy after surgery. Moreover, validation of these tools will also be possible in current prospective clinical trials studying the safety of active surveillance in low-risk DCIS (see information on LORD, LORIS and COMET below). Therefore, such tools will hopefully also help to predict negative outcomes in non-treated DCIS patients. 8 Future prospects 183

186 a b Figure 9. a Example of a future clinical risk-stratification DCIS tool. b Example of Adjuvant! Online tool for invasive breast cancer (adjuvantonline.com). 184 Future prospects

187 The issue of interobserver variability in the pathological classification of DCIS Risk prediction is hampered by interobserver variation among pathologists in classifying and grading DCIS Diagnostic inconsistencies may impact treatment decision-making and consequently affect prognosis. Unfortunately, experienced pathologists may reach different diagnoses of one DCIS lesion. This is mainly due to differences in morphological interpretation and field selection. Although 100% agreement will never be reached, the endorsement of uniform definitions and a single classification system may improve uniformness in the interpretation and reporting of DCIS. Furthermore, classification should be related to outcome, to have clinical significance. In the near future, we will hopefully be able to develop and adopt a new classification scheme, which can be reliably used for risk stratification. A first step to achieve this is to study interobserver variation among general and dedicated pathologists from different countries and determine which features are easy versus difficult to score. Next, specific features and classification methods that are related to clinical outcome need to be identified and their relation to other prognostic factors (such as imaging and molecular markers) needs to be assessed. Imaging features distinguishing benign tissue, DCIS and invasive breast cancer Preliminary results showed that the biochemical composition of calcifications is associated with benign breast lesions versus DCIS versus invasive breast cancer. 57 It will be of interest to study whether radiographic and physicochemical characteristics (e.g. carbonate content) of DCIS and adjacent tissues can accurately distinguish high-risk from low-risk DCIS. Future studies are needed to explore novel imaging biomarkers by linking these imaging findings with DCIS biology. Patient communication Communication and decision-making are important issues in DCIS management (Table 1). Specific topics include nomenclature, risk perception and the efficacy of decision aids. 58 There is little consensus regarding how best to explain DCIS to patients. 59 It has been proposed to change the nomenclature of DCIS by removing the term carcinoma. The term ductal intraepithelial neoplasia (DIN) has been recommended instead. 60 However, there are no data to support that this change would reduce risk misperceptions or anxiety, and, on the other hand, improve inter-observer agreement or decision-making. Instead, it is believed that focus on accurate and consistent reporting and terminology is more important at this time Future prospects 185

188 Active surveillance The current concerns about overdiagnosis and overtreatment of screen-detected DCIS urge the need to address the safety of active surveillance in women with this disease. The LORD (Chapter 7), LORIS and COMET studies have a common aim of assessing which low or intermediate grade DCIS require primary surgical management and whether regular monitoring for disease progression by mammography is a safe strategy, with intervention only in those women with evidence of progression to high-risk DCIS or invasive breast cancer. 61,62 The common research objectives of the three trials are: To evaluate the safety, effectiveness, cost effectiveness and acceptability of no surgical intervention in patients with newly diagnosed, mammogram detected asymptomatic, low or low intermediate grade DCIS. To define the natural history of low risk DCIS and to identify those patients who require surgery because their DCIS is at risk of progression to invasive disease. All of these trials are prospectively randomising patients with screen-detected or incidental low grade DCIS to standard surgical treatment or active monitoring. LORIS has recently opened to recruitment and is expected to recruit over the next 5 years and LORD and COMET will shortly open to recruitment. The large numbers of patients treated with active monitoring alone in the relatively safe confines of the three prospective randomised trials present a unique opportunity to address the translational drivers of progression to invasion. However, the results of these studies will only be available after many years rendering the ongoing studies described previously in this chapter very important. Finding the balance between overtreatment and undertreatment To guide treatment decision-making in DCIS we need better predictors of which DCIS lesions will become life-threatening invasive disease and which will not. Further, we need to effectuate a paradigm shift in how patients and health care providers perceive a diagnosis of cancer precursors. To address these key challenges a multidisciplinary, comprehensive approach is needed. Therefore, in an international (UK, USA, NL) collaborative effort we will unite a unique mix of scientific, clinical and patient expertise (PREvent ductal Carcinoma In Situ Invasive Overtreatment Now; PRECISION). Combining the information from several large well-defined datasets will allow us to increase our understanding of DCIS progression and develop a clinical DCIS risk test. Hopefully, this approach will reframe risk perceptions and avoid overtreatment and undertreatment of DCIS. 186 Future prospects

189 References 1. Falk RS, Hofvind S, Skaane P, Haldorsen T. Second events following ductal carcinoma in situ of the breast: a register-based cohort study. Breast Cancer Res Treat. 2011;129(3): Sprague BL, McLaughlin V, Hampton JM, Newcomb PA, Trentham-Dietz A. Disease-free survival by treatment after a DCIS diagnosis in a population-based cohort study. Breast Cancer Res Treat. 2013;141(1): Rakovitch E, Nofech-Mozes S, Narod SA, et al. Can we select individuals with low risk ductal carcinoma in situ (DCIS)? A population-based outcomes analysis. Breast Cancer Res Treat. 2013;138(2): Kane RL, Virnig BA, Shamliyan T, Wang S-Y, Tuttle TM, Wilt TJ. The impact of surgery, radiation, and systemic treatment on outcomes in patients with ductal carcinoma in situ. J Natl Cancer Inst Monographs. 2010;2010(41): Wong JS, Chen Y-H, Gadd MA, et al. Eight-year update of a prospective study of wide excision alone for small low- or intermediate-grade ductal carcinoma in situ (DCIS). Breast Cancer Res Treat. 2014;143(2): Solin LJ, Gray R, Hughes LL, et al. Surgical Excision Without Radiation for Ductal Carcinoma in Situ of the Breast: 12-Year Results From the ECOG-ACRIN E5194 Study. J Clin Oncol. September 2015:JCO McCormick B, Winter K, Hudis C, et al. RTOG 9804: a prospective randomized trial for good-risk ductal carcinoma in situ comparing radiotherapy with observation. J Clin Oncol. 2015;33(7): Donker M, Litière S, Werutsky G, et al. Breast-conserving treatment with or without radiotherapy in ductal carcinoma In Situ: 15-year recurrence rates and outcome after a recurrence, from the EORTC randomized phase III trial. J Clin Oncol. 2013;31(32): Wapnir IL, Dignam JJ, Fisher B, et al. Long-term outcomes of invasive ipsilateral breast tumor recurrences after lumpectomy in NSABP B-17 and B-24 randomized clinical trials for DCIS. J Natl Cancer Inst. 2011;103(6): Wärnberg F, Garmo H, Emdin S, et al. Effect of Radiotherapy After Breast-Conserving Surgery for Ductal Carcinoma in Situ: 20 Years Follow-Up in the Randomized SweDCIS Trial. J Clin Oncol. October 2014:JCO Cuzick J, Sestak I, Pinder SE, et al. Effect of tamoxifen and radiotherapy in women with locally excised ductal carcinoma in situ: long-term results from the UK/ANZ DCIS trial. Lancet Oncol. 2011;12(1): Narod SA, Iqbal J, Giannakeas V, Sopik V, Sun P. Breast Cancer Mortality After a Diagnosis of Ductal Carcinoma In Situ. JAMA Oncol. August doi: /jamaoncol Sagara Y, Mallory MA, Wong S, et al. Survival Benefit of Breast Surgery for Low-Grade Ductal Carcinoma In Situ: A Population-Based Cohort Study. JAMA Surg. 2015;150(8): Wärnberg F, Bergh J, Holmberg L. Prognosis in women with a carcinoma in situ of the breast: a population-based study in Sweden. Cancer Epidemiol Biomarkers Prev. 1999;8(9): Ernster VL, Barclay J, Kerlikowske K, Wilkie H, Ballard-Barbash R. Mortality among women with ductal carcinoma in situ of the breast in the population-based surveillance, epidemiology and end results program. Arch Intern Med. 2000;160(7): Worni M, Akushevich I, Greenup R, et al. Trends in Treatment Patterns and Outcomes for Ductal Carcinoma In Situ. J Natl Cancer Inst. 2015;107(12):djv General discussion, ongoing research and future prospects 187

190 17. van der Sanden GA, Coebergh JW, Schouten LJ, Visser O, van Leeuwen FE. Cancer incidence in The Netherlands in 1989 and 1990: first results of the nationwide Netherlands cancer registry. Coordinating Committee for Regional Cancer Registries. Eur J Cancer. 1995;31A(11): Casparie M, Tiebosch ATMG, Burger G, et al. Pathology databanking and biobanking in The Netherlands, a central role for PALGA, the nationwide histopathology and cytopathology data network and archive. Cell Oncol. 2007;29(1): National Evaluation Team for Breast cancer screening NETB, Fracheboud J, van Luijt PA, et al. National Evaluation of Breast Cancer Screening in the Netherlands / erasmusmc.nl/mage/publications-collaborations/pub/national-evaluation-breastcancer?lang=en. 20. Early Breast Cancer Trialists Collaborative Group (EBCTCG), Correa C, McGale P, et al. Overview of the randomized trials of radiotherapy in ductal carcinoma in situ of the breast. J Natl Cancer Inst Monographs. 2010;2010(41): Lee JR, Vogel VG. Who uses screening mammography regularly? Cancer Epidemiol Biomarkers Prev. 1995;4(8): Hofer TP, Katz SJ. Healthy behaviors among women in the United States and Ontario: the effect on use of preventive care. Am J Public Health. 1996;86(12): Boekel NB, Schaapveld M, Gietema JA, et al. Cardiovascular morbidity and mortality after treatment for ductal carcinoma in situ of the breast. J Natl Cancer Inst. 2014;106(8):dju Veal CT, Hart V, Lakoski SG, et al. Health-related behaviors and mortality outcomes in women diagnosed with ductal carcinoma in situ. J Cancer Surviv. January 2017: Berkman A, F Cole B, Ades PA, et al. Racial differences in breast cancer, cardiovascular disease, and all-cause mortality among women with ductal carcinoma in situ of the breast. Breast Cancer Res Treat. October 2014: Liu Y, Colditz GA, Gehlert S, Goodman M. Racial disparities in risk of second breast tumors after ductal carcinoma in situ. Breast Cancer Res Treat. September 2014: Zhang X, Dai H, Liu B, Song F, Chen K. Predictors for local invasive recurrence of ductal carcinoma in situ of the breast: a meta-analysis. Eur J Cancer Prev. 2016;25(1): Wang S-Y, Shamliyan T, Virnig BA, Kane R. Tumor characteristics as predictors of local recurrence after treatment of ductal carcinoma in situ: a meta-analysis. Breast Cancer Res Treat. 2011;127(1): Cheung S, Booth ME, Kearins O, Dodwell D. Risk of subsequent invasive breast cancer after a diagnosis of ductal carcinoma in situ (DCIS). Breast. September doi: /j.breast Collins LC, Achacoso N, Haque R, et al. Risk factors for non-invasive and invasive local recurrence in patients with ductal carcinoma in situ. Breast Cancer Res Treat. 2013;139(2): Kerlikowske K, Molinaro A, Cha I, et al. Characteristics associated with recurrence among women with ductal carcinoma in situ treated by lumpectomy. J Natl Cancer Inst. 2003;95(22): Hughes LL, Wang M, Page DL, et al. Local excision alone without irradiation for ductal carcinoma in situ of the breast: a trial of the Eastern Cooperative Oncology Group. J Clin Oncol. 2009;27(32): Allegra CJ, Aberle DR, Ganschow P, et al. National Institutes of Health State-of-the-Science Conference statement: Diagnosis and Management of Ductal Carcinoma In Situ September 22-24, In: Vol 102. Oxford University Press; 2010: Gierisch JM, Myers ER, Schmit KM, et al. Prioritization of research addressing management strategies for ductal carcinoma in situ. Ann Intern Med. 2014;160(7): To T, Wall C, Baines CJ, Miller AB. Is carcinoma in situ a precursor lesion of invasive breast cancer? Int J Cancer. 2014;135(7): General discussion, ongoing research and future prospects

191 36. Cowell CF, Weigelt B, Sakr RA, et al. Progression from ductal carcinoma in situ to invasive breast cancer: revisited. Molecular Oncology. 2013;7(5): Erbas B, Provenzano E, Armes J, Gertig D. The natural history of ductal carcinoma in situ of the breast: a review. Breast Cancer Res Treat. 2006;97(2): Sanders ME, Schuyler PA, Dupont WD, Page DL. The natural history of low-grade ductal carcinoma in situ of the breast in women treated by biopsy only revealed over 30 years of long-term follow-up. Cancer. 2005;103(12): Ernster VL. Nested case-control studies. Prev Med. 1994;23(5): Lampejo OT, Barnes DM, Smith P, Millis RR. Evaluation of infiltrating ductal carcinomas with a DCIS component: correlation of the histologic type of the in situ component with grade of the infiltrating component. Semin Diagn Pathol. 1994;11(3): Goldstein NS, Murphy T. Intraductal carcinoma associated with invasive carcinoma of the breast. A comparison of the two lesions with implications for intraductal carcinoma classification systems. Am J Clin Pathol. 1996;106(3): Combs SE, Han G, Mani N, Beruti S, Nerenberg M, Rimm DL. Loss of antigenicity with tissue age in breast cancer. Lab Invest. 2016;96(3): Arima N, Nishimura R, Osako T, et al. The importance of tissue handling of surgically removed breast cancer for an accurate assessment of the Ki-67 index. J Clin Pathol. 2016;69(3): Wallis MG, Clements K, Kearins O, Ball G, Macartney J, Lawrence GM. The effect of DCIS grade on rate, type and time to recurrence after 15 years of follow-up of screen-detected DCIS. Br J Cancer. 2012;106(10): Tsikitis VL, Chung MA. Biology of ductal carcinoma in situ classification based on biologic potential. Am J Clin Oncol. 2006;29(3): Iakovlev VV, Arneson NCR, Wong V, et al. Genomic differences between pure ductal carcinoma in situ of the breast and that associated with invasive disease: a calibrated acgh study. Clin Cancer Res. 2008;14(14): Kim SY, Jung S-H, Kim MS, et al. Genomic differences between pure ductal carcinoma in situ and synchronous ductal carcinoma in situ with invasive breast cancer. Oncotarget. 2015;6(10): Ringberg A, Nordgren H, Thorstensson S, et al. Histopathological risk factors for ipsilateral breast events after breast conserving treatment for ductal carcinoma in situ of the breast--results from the Swedish randomised trial. Eur J Cancer. 2007;43(2): Lari SA, Kuerer HM. Biological Markers in DCIS and Risk of Breast Recurrence: A Systematic Review. J Cancer. 2011;2: Kerlikowske K, Molinaro AM, Gauthier ML, et al. Biomarker expression and risk of subsequent tumors after initial ductal carcinoma in situ diagnosis. J Natl Cancer Inst. 2010;102(9): Emens LA. Breast cancer immunobiology driving immunotherapy: vaccines and immune checkpoint blockade. Expert Rev Anticancer Ther. 2012;12(12): Cimino-Mathews A, Foote JB, Emens LA. Immune targeting in breast cancer. Oncology (Williston Park, NY). 2015;29(5): Elston CW, Sloane JP, Amendoeira I, et al. Causes of inconsistency in diagnosing and classifying intraductal proliferations of the breast. European Commission Working Group on Breast Screening Pathology. Eur J Cancer. 2000;36(14): O Malley FP, Mohsin SK, Badve S, et al. Interobserver reproducibility in the diagnosis of flat epithelial atypia of the breast. Mod Pathol. 2006;19(2): General discussion, ongoing research and future prospects 189

192 55. Jain RK, Mehta R, Dimitrov R, et al. Atypical ductal hyperplasia: interobserver and intraobserver variability. Mod Pathol. 2011;24(7): Van de Vijver MJ, Peterse H. The diagnosis and management of pre-invasive breast disease: pathological diagnosis--problems with existing classifications. Breast Cancer Res. 2003;5(5): Baker R, Rogers KD, Shepherd N, Stone N. New relationships between breast microcalcifications and cancer. Br J Cancer. 2010;103(7): Partridge AH, Elmore JG, Saslow D, McCaskill-Stevens W, Schnitt SJ. Challenges in ductal carcinoma in situ risk communication and decision-making: report from an American Cancer Society and National Cancer Institute workshop. CA Cancer J Clin. 2012;62(3): Fallowfield L, Matthews L, Francis A, Jenkins V, Rea D. Low grade Ductal Carcinoma in situ (DCIS): How best to describe it? Breast. June doi: /j.breast Veronesi U, Zurrida S, Goldhirsch A, Rotmensz N, Viale G. Breast cancer classification: time for a change. J Clin Oncol. 2009;27(15): Francis A, Thomas J, Fallowfield L, et al. Addressing overtreatment of screen detected DCIS; the LORIS trial. Eur J Cancer. 2015;51(16): Grimm LJ, Shelley Hwang E. Active Surveillance for DCIS: The Importance of Selection Criteria and Monitoring. Ann Surg Oncol. 2016;23(13): General discussion, ongoing research and future prospects

193 General discussion, ongoing research and future prospects 191 8

194

195 CHAPTER 9 Summary Samenvatting List of co-authors PhD portfolio Dankwoord Curriculum Vitae

196 This thesis describes clinical outcomes of a large nationwide cohort of women treated for ductal carcinoma in situ (DCIS) and discusses steps towards finding the balance between overtreatment and undertreatment of DCIS. Chapter 1 provides a general discussion into the epidemiology, detection and treatment of DCIS. DCIS is a precursor lesion of invasive breast cancer. However, not all DCIS will progress into invasive disease. While DCIS was seldom detected before the advent of screening mammography, the widespread adoption and the subsequent digitalization of mammographybased population-wide breast cancer screening programs have led to an impressive increase in incidence of DCIS. Therefore, screening programs are criticized for being associated with substantial overdiagnosis of DCIS. DCIS treatment aims at preventing invasive breast cancer, assuming that this will reduce deaths due to breast cancer. Standard treatment of DCIS consists of mastectomy, or breast-conserving surgery, which is often followed by radiotherapy as part of breast-conserving treatment. Some women with unilateral DCIS even undergo contralateral prophylactic mastectomy. In the Netherlands, endocrine therapy is not recommended, whereas in the United States, it is considered part of standard treatment. The question has been raised whether intensive treatment for low-risk DCIS might be considered overtreatment. Chapters 2 and 3 summarize our current understanding of DCIS and identify the challenges in finding the balance between overtreatment and undertreatment. Unfortunately, currently, we are not able to distinguish DCIS lesions that will progress to invasive disease from lesions that will not. The result of this knowledge gap is that every DCIS lesion is treated similarly to earlystage invasive breast cancer. Risk stratification is therefore essential for making better-informed treatment decisions. In addition, large randomized clinical trials are necessary to investigate if active surveillance is a safe option for low-grade DCIS. Last but not least, clear communication between health care professionals and patients about the implications of a DCIS diagnosis is essential for accurate risk perception and optimal decision-making. Chapters 4, 5 and 6 discuss the results of a large population-based cohort study. All women diagnosed with DCIS in the Netherlands between January 1 st 1989 and December 31 st 2004 were selected from the Netherlands Cancer Registry. This cohort was linked to the nationwide network and registry of histology and cytopathology in the Netherlands (PALGA) and their information was used to validate and complete missing data on surgery type and morphology. In Chapter 4, we assessed the effect of different treatment strategies on the risk of subsequent invasive breast cancer in 10,090 women treated for unilateral primary pure DCIS. Fifteen years after DCIS diagnosis, cumulative incidence of ipsilateral invasive breast cancer was 1.9% after mastectomy, 8.8% after breast-conserving surgery with radiotherapy, and 15.4% after breast- 194 Summary

197 conserving surgery alone. Cumulative incidence of contralateral invasive breast cancer at 15 years was 6.4 %, which was twice as high as the expected breast cancer incidence based on incidence rates in the general Dutch population (3.4%). The association of DCIS treatment with ipsilateral invasive breast cancer risk was modified by age at DCIS diagnosis and follow-up duration. Our results indicate that the benefit of radiotherapy may be smaller among younger women, and stress the importance of clinical studies with long follow-up. Finally, the low contralateral invasive breast cancer risk does not justify contralateral prophylactic mastectomies for many women with unilateral DCIS. In Chapter 5, we aimed to weigh breast cancer mortality after DCIS treatment against mortality from other causes and expected mortality in the general population. To obtain information on cause of death, our DCIS cohort was linked with the nationwide cause of death registry at Statistics Netherlands. Our study comprised 9,799 DCIS patients and we estimated standardized mortality ratios as ratios of observed and expected numbers of death. DCIS patients aged 50 years or older had lower risk of dying compared with the general female population, which may reflect differences in health behaviour. Women with DCIS had higher risk of dying from breast cancer than the general population, but absolute risks were low: the cumulative breast cancer mortality 10 years after DCIS treatment was 2.3% for women younger than 50 years and 1.4% for women aged 50 years or older treated for DCIS between In Chapter 6, we assessed whether the method of detection should be considered when determining prognosis and treatment in women with DCIS. To obtain information about the method of detection we linked our cohort with the database of the Dutch breast cancer screening organization. We studied 7,042 women aged with screen-detected, interval, or nonscreening-related DCIS. Screen detection was associated with lower risk of ipsilateral invasive breast cancer and all-cause mortality. However, the absolute difference in risk of ipsilateral invasive breast cancer was very small (1%) and the lower all-cause mortality associated with screen-detected and interval DCIS might be explained by a healthy-screenee effect. Therefore, our findings do not justify different treatment strategies for women with screen-detected, interval, or non-screening-related DCIS. The LORD study aims to evaluate the safety of active surveillance in women with LOw-Risk Dcis. Chapter 7 discusses the rationale and design of this randomised, international multicentre, open-label, phase III non-inferiority trial, led by the Dutch Breast Cancer Research Group (BOOG ) and the European Organization for Research and Treatment of Cancer (EORTC-BCG 1401). Standard treatment will be compared to active surveillance in 1,240 women aged 45 years or older with asymptomatic, screen-detected, pure low-grade DCIS based on vacuumassisted biopsies of calcifications only. Both study arms will be monitored with annual digital 9 Summary 195

198 mammography for a period of 10 years. The primary end-point is 10-year ipsilateral invasive breast cancer free percentage. Secondary end-points include patient reported outcomes, diagnostic biopsy rate during follow-up, ipsilateral mastectomy rate and translational research. To explore interest in and feasibility of the LORD study we conducted a survey among EORTC and BOOG centres. A vast majority of responding EORTC and BOOG centres expressed interest in participation in the LORD study. The proposed study design is endorsed by nearly all centres. In Chapter 8, preliminary results of our ongoing research are discussed and future steps towards finding the balance between overtreatment and undertreatment of DCIS are considered. To guide treatment decision-making in DCIS we need to have better predictors of which DCIS lesions will become life-threatening invasive disease and which will not. Further, we need to effectuate a paradigm shift in how patients and health care providers perceive a diagnosis of cancer precursors. To address these key challenges a multidisciplinary, comprehensive approach is needed. 196 Summary

199 Summary 197 9

200

201 CHAPTER 9 Summary Samenvatting List of co-authors PhD portfolio Dankwoord Curriculum Vitae

202 Dit proefschrift richt zich op het dilemma van overdiagnostiek en overbehandeling van ductaal carcinoma in situ van de borst (DCIS). Het behandelt de klinische resultaten van een observationele landelijke cohort studie en bespreekt het traject dat zich richt op het vinden van de balans tussen overbehandeling en onderbehandeling van DCIS. Hoofdstuk 1 geeft een algemene inleiding over de detectie en behandeling van DCIS. DCIS is een voorstadium van invasief borstkanker. Echter, niet alle vrouwen met DCIS zullen invasief borstkanker ontwikkelen. Sinds de introductie van het bevolkingsonderzoek naar borstkanker en door de verbeterde mammografietechnieken is de incidentie van DCIS aanzienlijk toegenomen. Uit de literatuur blijkt dat DCIS altijd al veel voorkwam, maar dat dit vroeger, vóór het bevolkingsonderzoek, slechts sporadisch gedetecteerd werd. Er zijn aanwijzingen dat overdiagnostiek van DCIS in reële mate voorkomt. De huidige behandeling van DCIS is chirurgie (borstsparende operatie of mastectomie) met of zonder radiotherapie en heeft als doel invasief borstkanker te voorkomen, om zo uiteindelijk sterfte aan borstkanker te reduceren. Sommige vrouwen met eenzijdig DCIS ondergaan zelfs een contralaterale profylactische mastectomie. Adjuvante hormonale behandeling voor DCIS wordt in Nederland niet geadviseerd, terwijl in Noord-Amerika deze therapie deel uit maakt van de standaard behandelingsrichtlijn. Omdat niet alle vrouwen met DCIS invasief borstkanker zullen ontwikkelen zijn er toenemende zorgen dat vrouwen met laag-risico DCIS geen baat hebben van behandeling en dus overbehandeld worden. Hoofdstukken 2 en 3 geven een overzicht van de huidige wetenschappelijke kennis over DCIS en identificeren de uitdagingen van het onderzoek dat zich richt op het vinden van de balans tussen overbehandeling en onderbehandeling van DCIS. Op dit moment kunnen we helaas niet goed voorspellen welke DCIS afwijkingen invasief zullen worden. Dit heeft als gevolg dat alle DCIS patiënten dezelfde behandeling ondergaan als vrouwen met een vroeg stadium van invasieve borstkanker. Om therapie op maat te kunnen aanbieden is het noodzakelijk dat we DCIS beter kunnen indelen in laag-risico en hoog-risico DCIS. Tevens zijn er gerandomiseerde klinische studies nodig om te onderzoeken of een actief afwachtend beleid voor laag-risico DCIS een veilige optie is. Ten slotte is juiste en genuanceerde communicatie over de implicaties van de diagnose DCIS essentieel voor een reële risicoperceptie bij en optimale besluitvorming door de patiënte en de betrokken zorgprofessionals. Hoofdstukken 4, 5 en 6 beschrijven de resultaten van een grote landelijke DCIS cohort studie. Met behulp van de Nederlandse Kanker Registratie werden alle vrouwen die primair gediagnosticeerd zijn met een DCIS in Nederland tussen 1 januari 1989 en 31 december 2004 geselecteerd. Dit cohort werd gekoppeld aan het Pathologisch Anatomisch Landelijk 200 Samenvatting

203 Geautomatiseerd Archief (PALGA) en met behulp van deze informatie konden gegevens over behandeling en morfologie geverifieerd en toegevoegd worden. In Hoofdstuk 4 onderzochten wij het effect van behandelstrategie op het risico van het ontwikkelen van een invasief mammacarcinoom in de dezelfde borst in een cohort van vrouwen die behandeld werden voor een enkelzijdig puur DCIS. Vijftien jaar na de DCIS diagnose was de cumulatieve incidentie van ipsilateraal invasief borstkanker 1,9% na een mastectomie, 8,8% na een borstsparende operatie met adjuvante radiotherapie en 15.4% na een borstsparende operatie zonder radiotherapie. De cumulatieve incidentie van contralateraal invasief borstkanker na 15 jaar was 6,4% (ter vergelijking: het risico op invasief borstkanker in de algemene bevolking was 3,4%). Onze resultaten suggereren dat de risicoreductie ten gunste van radiotherapie na borstsparende chirurgie kleiner is bij jonge vrouwen (jonger dan 50 jaar ten tijde van DCIS diagnose) en benadrukken het belang van klinische studies met lange follow-up. Tevens concludeerden wij dat een contralaterale profylactische mastectomie niet moet worden geadviseerd aan vrouwen met een unilateraal DCIS zonder familiare belasting, gezien het lage risico op het ontwikkelen van contralateraal invasief mammacarcinoom. In Hoofdstuk 5 hebben wij het risico om te overlijden aan borstkanker na behandeling van DCIS geanalyseerd en afgewogen tegen het risico om te overlijden aan andere oorzaken. Een belangrijk aspect van deze studie was de vergelijking van mortaliteit met de algemene bevolking door het berekenen van gestandaardiseerde mortaliteit ratio s. Met behulp van een bestaande koppeling met het Centraal Bureau voor de Statistiek werd informatie met betrekking tot doodsoorzaken verkregen en konden wij 9799 DCIS patiënten includeren in deze studie. DCIS patiënten van 50 jaar en ouder hadden een betere levensverwachting vergeleken met de algemene bevolking. Dit kan mogelijk verklaard worden door verschillen in gezondheidsgedrag. Vrouwen met DCIS hadden weliswaar een hogere kans om te overlijden aan borstkanker dan vrouwen in de algemene bevolking, maar de absolute risico s waren desalniettemin zeer laag: de cumulatieve borstkanker sterftekans 10 jaar na DCIS behandeling tussen 1999 en 2004 was 2,3% voor vrouwen jonger dan 50 ten tijde van diagnose en 1,4% voor vrouwen van 50 jaar of ouder. In Hoofdstuk 6 hebben we onderzocht of de methode van DCIS detectie onafhankelijke prognostische waarde heeft. Door middel van een koppeling met het Bevolkingsonderzoek Borstkanker was het mogelijk om 7042 vrouwen van 49 tot en met 75 jaar oud met screening gedetecteerde DCIS, interval DCIS en niet-screening gerelateerde DCIS te vergelijken. Screening gedetecteerde DCIS afwijkingen waren geassocieerd met betere overleving en lager risico op ipsilateraal invasief mammacarcinoom. Het absolute verschil in risico op ipsilateraal invasief borstkanker was echter klinisch verwaarloosbaar (1%) en de betere overleving kan waarschijnlijk 9 Samenvatting 201

204 worden toegeschreven aan een healthy-user effect. Wij concludeerden daarom dat onze bevindingen geen bewijs leveren dat screening gedetecteerde DCIS, interval DCIS en nietscreening gerelateerde DCIS op een andere wijze behandeld moeten worden. De LORD studie heeft als doel om te onderzoeken of een actief afwachtend beleid voor laagrisico DCIS (LOw Risk Dcis) veilig is. Hoofdstuk 7 behandelt de overwegingen en het design van deze gerandomiseerde, internationale multicenter, open-label, fase III, non-inferioriteitsstudie. De studie wordt onder de hoede van de Borstkanker Onderzoek Groep (BOOG ) en de European Organisation for Research and Treatment of Cancer (EORTC 1401) uitgevoerd. In totaal zullen 1240 vrouwen van 45 jaar of ouder met asymptomatisch laaggradig DCIS op basis van biopten van calcificaties gerandomiseerd worden tussen standaard behandeling (mastectomie of borstsparende chirurgie, eventueel gevolgd door radiotherapie en/of hormonale therapie) en een actief afwachtend beleid. Vrouwen in beide studie armen zullen jaarlijks worden gecontroleerd met behulp van een mammogram gedurende een periode van 10 jaar. Het primaire eindpunt is het 10-jaar ipsilateraal invasief mammacarcinoom vrije-percentage. Secundaire eindpunten betreffen onder andere patiënt-gerelateerde uitkomsten, het aantal diagnostische biopten tijdens follow-up, het aantal ipsilaterale mastectomieën tijdens followup en translationeel onderzoek. Om de interesse in en de haalbaarheid van de LORD studie te onderzoeken hebben we een vragenlijst uitgestuurd naar EORTC en BOOG-gerelateerde ziekenhuizen. De resultaten van deze vragenlijst waren positief: de opzet van de LORD studie werd internationaal zeer goed ontvangen en de meerderheid van de respondenten gaf aan deel te willen nemen aan de studie. In Hoofdstuk 8 bespreken we de voorlopige resultaten van onze lopende studies en schetsen we de vervolgstappen die hopelijk zullen leiden tot een betere balans tussen overbehandeling en onderbehandeling van DCIS. Het accuraat inschatten van de risico s en het in perspectief plaatsen daarvan is essentieel, waarbij de kwaliteit van leven en concurrerende factoren met betrekking tot morbiditeit en mortaliteit in ogenschouw moeten worden genomen. Voorlichting aan zorgverleners en de ontwikkeling van een predictiemodel om het individuele risico op progressie te bepalen zullen hieraan bijdragen. 202 Samenvatting

205 Samenvatting 203 9

206

207 CHAPTER 9 Summary Samenvatting List of co-authors PhD portfolio Dankwoord Curriculum Vitae

208 List of co-authors Nina Bijker Department of Radiotherapy, Academic Medical Center, Amsterdam Nicolas Dif Department of Clinical Operations, European Organisation for Research and Treatment of Cancer, Brussels Emma J. Groen Department of Pathology, Netherlands Cancer Institute/Antoni van Leeuwenhoek, Amsterdam Flora E. van Leeuwen Division of Psychosocial research and Epidemiology, Netherlands Cancer Institute/Antoni van Leeuwenhoek, Amsterdam A. Elise van Leeuwen-Stok BOOG Study Center, Amsterdam Esther H. Lips Division of Molecular Pathology, Netherlands Cancer Institute/Antoni van Leeuwenhoek, Amsterdam Linda de Munck Department of Research, Netherlands Comprehensive Cancer Organisation, Utrecht Ruud M. Pijnappel Department of Radiology, University Medical Centre Utrecht, Utrecht Emiel J. T. Rutgers Division of Surgical Oncology, Netherlands Cancer Institute/Antoni van Leeuwenhoek, Amsterdam Michael Schaapveld Division of Psychosocial research and Epidemiology,Netherlands Cancer Institute/Antoni van Leeuwenhoek, Amsterdam Marjanka K. Schmidt Division of Molecular Pathology, Division of Psychosocial research and Epidemiology, Netherlands Cancer Institute/Antoni van Leeuwenhoek, Amsterdam Victoria P. Skinner Division of Surgical Oncology, Netherlands Cancer Institute/Antoni van Leeuwenhoek, Amsterdam 206 List of co-authors and their contributions

209 Leen Slaets Konstantinos Tryfonidis Lindy L. Visser Jelle Wesseling Hillegonda A.O. Winter-Warnars Department of Statistics, European Organisation for Research and Treatment of Cancer, Brussels Medical Department, European Organisation for Research and Treatment of Cancer, Brussels Division of Molecular Pathology, Netherlands Cancer Institute/Antoni van Leeuwenhoek, Amsterdam Department of Pathology, Division of Molecular Pathology, Netherlands Cancer Institute/Antoni van Leeuwenhoek, Amsterdam Department of Radiology, Netherlands Cancer Institute/Antoni van Leeuwenhoek, Amsterdam 9 List of co-authors and their contributions 207

210 Author contributions Chapter 2 EG, LE, LV, ER, HW, EL, and JW designed the study; EG, LE, and LV collected the data; EG and JW wrote the first draft of the manuscript; all authors (EG, LE, LV, ER, HW, EL, and JW) contributed to the interpretation and critically reviewed the first draft of the manuscript; they all read and approved the final version of the manuscript. Chapter 3 EG, LE, ER, HW, EL, and JW designed the study; EG and LE collected the data; EG wrote the first draft of the manuscript; all authors (EG, LE, ER, HW, EL, and JW) contributed to the interpretation and critically reviewed the first draft of the manuscript; they all read and approved the final version of the manuscript. Chapter 4 LE, MS, MKS, FL, and JW designed the study; LE, MS, and MKS collected the data; LE, MS, and MKS performed the statistical analysis; LE wrote the first draft of the manuscript; all authors (LE, MS, ER, MKS, FvL, and JW) contributed to the interpretation of the results and critically reviewed the first draft of the manuscript; they all read and approved the final version of the manuscript. Chapter 5 LE, MS, MKS, ER, FL, and JW designed the study; LE, MS, and MKS collected the data; LE, MS, and MKS performed the statistical analysis; LE wrote the first draft of the manuscript; all authors (LE, MS, ER, MKS, FvL, and JW) contributed to the interpretation of the results and critically reviewed the first draft of the manuscript; they all read and approved the final version of the manuscript. Chapter 6 LE, MS, MKS, FL, and JW designed the study; LE, MS, MKS, and LM collected the data; LE, MS, and MKS performed the statistical analysis; LE wrote the first draft of the manuscript; all authors (LE, MS, ER, MKS, LdM, FvL, and JW) contributed to the interpretation of the results and critically reviewed the first draft of the manuscript; they all read and approved the final version of the manuscript. Chapter 7 LE, ER, LS, KT, RM, NB and JW designed the study; LE, LS, KT, and ND collected the data; LE and LS performed the statistical analysis; LE wrote the first draft of the manuscript; all authors (LE, ER, LS, KT, AL, VS, ND, RM, NB and JW) contributed to the interpretation of the results and critically reviewed the first draft of the manuscript; they all read and approved the final version of the manuscript. 208 List of co-authors and their contributions

211 List of co-authors and their contributions 209 9

212

213 CHAPTER 9 Summary Samenvatting List of co-authors PhD portfolio Dankwoord Curriculum Vitae

214 PhD portfolio Graduate school Oncology Amsterdam (OOA) PhD student L.E. Elshof PhD period September 2012 January 2017 PhD supervisors E.J.T. Rutgers F.E. van Leeuwen J. Wesseling M. Schaapveld M.K. Schmidt Year General courses OOA course Basic Medical Statistics 2012 Training course Good Clinical Practice 2012 OOA workshop How to write high impact papers and what to do when your manuscript is rejected 2013 VUmc CCA PhD meeting workshop Trigger a broad audience 2013 EORTC course Clinical Trial Statistics for Non Statisticians 2014 EORTC course A One-Day Journey through EORTC Activities 2014 EORTC-PHARMED-PHARMATRAIN course in Modern Developments of Oncological Treatments 2014 Specific courses Epidm cursus Epidemiologisch onderzoek: opzet en interpretatie 2012 Epidm cursus Principes van epidemiologische data-analyse 2012 Epidm cursus Regressietechnieken 2013 Epidm cursus Systematische reviews en meta-analyse 2013 Epidm cursus Klinimetrie 2013 Doorlopende Nascholingscursus Medische Oncologie Mammacarcinoom 2014 ECCO-AACR-EORTC-ESMO Flims Workshop: Methods in Clinical Cancer Research 2013 Nederlandse vereniging voor oncologie cursus Basiscursus Oncologie 2013 Epidm cursus Epidemiologie van ziekten 2013 Epidm cursus Praktische epidemiologie: opzetten van een onderzoek 2013 Epidm Winter course Clinical Prediction Models 2014 Epidm Doelmatigheidsonderzoek: methoden en principes 2014 Epidm Stage Praktisch Werk 2015 Workload (ECTS) PhD portfolio

215 Seminars, workshops and master classes Seminar lunch with C. Swanton Division seminars Epidm workshop Omgaan met de pers Epidm workshop Successful poster design Epidm workshop Het schrijven van een peer review Year Workload (ECTS) Other Annual Graduate Student Retreat OOA BOOG vergadering EORTC Breast Cancer Group meeting BIG Scientific Meeting (Inter)national conferences AVL Mammacarcinoom symposium EGAM EORTC meeting San Antonio Breast Cancer Conference European Cancer Congress Bossche Mamma Congres NABON / BOOG Symposium European Breast Cancer Conference 2012, , , , , , PhD portfolio 213

216 Presentations Year LORD trial, oral presentations, BOOG vergadering, Utrecht Identification of risk factors associated with subsequent invasive ipsilateral breast cancer following a first diagnosis of primary DCIS, elevator pitch & poster presentation, Pathologendagen, Zeist LORD trial, oral presentations, EORTC Breast Cancer Group meeting, various EU Risk of subsequent ipsilateral invasive breast cancer after a primary diagnosis of ductal carcinoma in situ, poster presentation, SABCS, San Antonio DCIS, oral presentatation, A Sister s Hope, Amsterdam Management of low risk ductal carcinoma in situ, oral presentation, Flims Alumni Club, EBCC, Glasgow LORD trial Low Risk DCIS, oral presentation, French Breast Cancer Intergroupe Unicancer meeting, Parijs Finding the balance between overdiagnosis and undertreatment of ductal carcinoma in situ, oral presentation, EORTC BCG scientific meeting, Amsterdam Gelukkig laaggradig - Hoe gaat het terugkijkend?, oral presentation, AvL-NKI Mammasymposium, Amsterdam LORD trial Low Risk DCIS, oral presentation, Roche Breast Cancer Masterclass, Milaan The LORD trial - A randomized non-inferiority trial between active surveillance versus standard treatment in patients with LOw-Risk Ductal Carcinoma In Situ, poster presentation, SABCS, San Antonio LORD trial, oral presentation, EORTC Radiation Oncology Group meeting, EGAM, Brussel Screen detected DCIS: een aparte categorie, oral presentation, AvL-NKI Mammasymposium, Amsterdam Prognostic value of method of detection in primary pure DCIS, poster presentation, SABCS, San Antonio How to Communicate the Diagnosis and Treatment Strategy, oral case presentation, EBCC, Amsterdam Low cause-specific mortality in women treated for ductal carcinoma in situ of the breast, poster discussion, ECCO, Amsterdam PhD portfolio

217 List of publications Elshof LE, Schmidt MK, Rutgers EJ, van Leeuwen FE, Wesseling J, Schaapveld M. Causespecific mortality in a population-based cohort of 9,799 women treated for ductal carcinoma in situ. Ann Surg Apr. Epub ahead of print. Elshof LE, Schaapveld M, Rutgers EJ, Schmidt MK, de Munck L, van Leeuwen FE, Wesseling J. The method of detection of ductal carcinoma in situ has no therapeutic implications: results of a population-based cohort study. Breast Cancer Res Mar;19(1):26. Groen EJ, Elshof LE, Visser LL, Rutgers EJ, Winter-Warnars HA, Lips EH, Wesseling J. Finding the balance between over- and under-treatment of ductal carcinoma in situ (DCIS). Breast Feb;31: Elshof LE, Schaapveld M, Schmidt MK, Rutgers EJ, van Leeuwen FE, Wesseling J. Subsequent risk of ipsilateral and contralateral invasive breast cancer after treatment for ductal carcinoma in situ: incidence and the effect of radiotherapy in a population-based cohort of 10,090 women. Breast Cancer Res Treat Oct;159(3): Groen EJ, Elshof LE, Rutgers EJ, Winter-Warnars HA, Lips EH, Wesseling J. Ductal carcinoma in situ: the balance between over- and undertreatment. Ned Tijdschr Geneeskd. 2016;160:A9773. Elshof LE, Tryfonidis K, Slaets L, van Leeuwen-Stok AE, Skinner VP, Dif N, Pijnappel RM, Bijker N, Rutgers EJ, Wesseling J. Feasibility of a prospective, randomised, open-label, international multicentre, phase III, non- inferiority trial to assess the safety of active surveillance for low risk ductal carcinoma in situ - The LORD study. Eur J Cancer Aug;51(12): Koolen BB, Elshof LE, Loo CE, Wesseling J, Vrancken Peeters MJ, Vogel WV, Rutgers EJ, Valdés Olmos RA. Does the pretreatment tumor sampling location correspond with metabolic activity on 18F-FDG PET/CT in breast cancer patients scheduled for neoadjuvant chemotherapy? Eur J Radiol Dec;82(12): Elshof LE, Rutgers EJ, Deurloo EE, Loo CE, Wesseling J, Pengel KE, Gilhuijs KGA. A practical approach to manage additional lesions at preoperative breast MRI in patients eligible for breast conserving therapy: results. Breast Cancer Res Treat Dec;124(3): PhD portfolio 215

218

219 CHAPTER 9 Summary Samenvatting List of co-authors PhD portfolio Dankwoord Curriculum Vitae

220 Augustus 2017, op de grens van Tarn- en Lot et Garonne in Frankrijk, blik ik terug op de afgelopen jaren in het Nederlands Kanker Instituut Antoni van Leeuwenhoek (NKI-AVL). Ik kwam als 22-jarige Geneeskunde student binnen en vertrok als 30-jarige arts-assistent in opleiding, bijna klaar met mijn promotieonderzoek. Graag wil ik degenen die mij de afgelopen jaren hebben begeleid en ondersteund bedanken. Prof. dr. E.J.T. Rutgers, lieve Emiel, met een telefoongesprek met jou in mijn derde studiejaar is mijn hele NKI-AVL avontuur begonnen. Ik kan me nog goed herinneren dat ik in de hal met het indrukwekkende Regenboog plafond wachtte op Kenneth Gilhuijs. Hij zou mij introduceren tot ons onderzoek naar de Unidentified Breast Objects, beter bekend als UBO s. Jullie hebben mij toen gegrepen voor het borstkankeronderzoek. Tijdens mijn co-schappen in Rotterdam was ik het Amsterdamse Mamma-Meisje en de terugkeer naar het NKI-AVL voor mijn oudste co-schap en eerste baan in de heelkundige kliniek verbaasde niemand. Door jou heb ik zoveel kansen gekregen en gepakt. Ik vind het heel bijzonder dat jij mijn promotor bent. Prof. dr. ir. F.E. van Leeuwen, lieve Floor, als jonge clinicus vond ik het aanvankelijk spannend om onder jouw supervisie epidemiologisch onderzoek te mogen doen. Tijdens de studie Geneeskunde was er weinig onderwijstijd voor methodologie van onderzoek, maar ik ben blij dat ik dat de afgelopen jaren heb kunnen inhalen. Van jouw feedback heb ik erg veel geleerd. Ik ben erg trots dat jij mijn promotor bent. Dr. J. Wesseling, lieve Jelle, op 1 oktober 2010 in de trein terug van de Scholingscursus Mammacarcinoom in Tiel spraken wij elkaar voor het eerst. Anderhalf jaar later, terwijl ik op de 5 e en 6 e verdieping van het NKI-AVL werkte, heb jij mij als eerste promovendus een klinisch onderzoeker op het DCIS project binnen gehaald. En nu, bijna 5 jaar nadat we het project begonnen, is het onder jouw leiding uitgegroeid tot een Grand Challenge. Op één van mijn jaargesprekken vroeg je mij hoe ik mijn werk vond op een schaal van 1 tot 100. Ik antwoordde 100 en dat meende ik ook. Dat gevoel komt voor een groot deel door jouw vertrouwen in mij en de vele mogelijkheden tot ontwikkeling die jij mij gegeven hebt. Daarnaast was er ook altijd veel ruimte en aandacht voor de belangrijkste dingen in het leven. Dr. M. Schaapveld, lieve Michael, jij leerde mij Stata, één van mijn favoriete bezigheden tijdens mijn onderzoekersbestaan. Ook epidemiologie in de praktijk heb ik van jou geleerd. Jij begeleidde mij bij alle analyses en was mijn stagebegeleider tijdens mijn masteropleiding Epidemiologie. Met dezelfde aandacht en tijd die je voor mij had hoop ik later jonge onderzoekers te kunnen begeleiden. 218 Dankwoord

221 Dr. M.K. Schmidt, lieve Marjanka, terugblikkend was ik je halverwege mijn onderzoek een beetje kwijt. Maar gelukkig heb ik je weer teruggevonden, want jij hebt mij door de laatste periode geleid, aangemoedigd, ogen geopend, gekletst en geadviseerd. Voor mij bracht jij precies de juiste balans tussen statistische analyses, klinische relevantie en pragmatische planning. De leden van mijn promotie commissie; prof. dr. M.J. van de Vijver, prof. dr. J.H.G. Klinkenbijl, prof. dr. R.M. Pijnappel, prof. dr. L.V. van de Poll-Franse, dr. N. Bijker, dr. C.M. Ronckers en dr. H.A.O. Winter-Warnars wil ik graag bedanken voor het beoordelen van mijn manuscript. Alle medeauteurs en onderzoekers. Dank voor jullie waardevolle input en kritische blik, zonder welke dit proefschrift niet tot stand was gekomen. Jos Jonkers en Hester Oldenburg, veel dank voor het zitting nemen in mijn OOA commissie en het van een afstand toezicht houden op mijn promotietraject. Lieve Olga Klok, heel veel dankjewels voor heel veel organisatorische hulp en lieve woorden. Ik kon zelfs Maurits met gerust hart bij jou achterlaten, terwijl ik met de laatste loodjes bezig was. De Wesseling groep, dank voor alle leermomenten tijdens de Wesseling meetings. Jullie maakten de wereld van DNA isolatie, next generation sequencing en xenograft modellen iets beter te begrijpen. Emilie, ik was blij verrast toen jij als jonge, nieuwe mamma-patholoog in het NKI-AVL kwam werken. Ik wens je veel succes met de vervolg DCIS studies en ik ben er als je vragen hebt. Esther, dank voor het meedenken tijdens onze DCIS besprekingen. Dank analisten en studenten voor al jullie werk met betrekking tot invoer, verwerking en bewerking van al die DCIS en IBC blokjes. Michiel in het bijzonder, met een snel naderende deadline voor het opvragen van die blokjes kwam jij mij in het O-gebouw helpen met de laatste PALGA excerpten. Mathilde, Maartje & Sina, succes met het micromilieu, de DCIS recidieven en de LORD, en alles wat er bij komt kijken. Lindy Visser, ik ben erg blij dat jij, met een flinke portie nieuwsgierigheid en ijver, mij bent komen versterken op het DCIS project. Gigantische bergen werk heb je verzet en de vruchten werpen zich af. Dank collega s van de HOD, PSOE en C2. Zowel de Afdeling Pathologie als de Divisie Moleculaire Pathologie, waaronder ook de Core Facility Molecular Pathology and Biobanking. Koen van de Vijver, dank voor je hulp bij de DCIS studies, waaronder de PA revisies. Lieve Mirna Ekelschot-van Diermen, de beste office manager die ik ken, dank voor alle organisatorische hulp en de tijd die je voor me nam in de afgelopen jaren. Philip en Gwen, net nieuw op C2 werd 9 Dankwoord 219

222 ik door jullie wegwijs gemaakt. Dank voor jullie openheid en gezelligheid. Datzelfde geldt voor Renske Fles. Dank leden van de mammatumorenwerkgroep, waaronder Claudette Loo, Marie-Jeanne Vrancken Peeters, Astrid Scholten, Nicola Russell, Frederique van Duijnhoven, Sabine Linn, Gabe Sonke, Paula Elkhuizen en Jolanda Remmelzwaal. Elise van Leeuwen-Stok en BOOG medewerkers, tijdens mijn eerste BOOG vergadering mocht ik de eerste aanzet tot de LORD studie presenteren. Bloed, zweet en tranen en jaren later is de studie open! Dank voor de fijne samenwerking. Ruud Pijnappel en Nina Bijker, met jullie als co-principal investigators heb ik met veel plezier aan de LORD gewerkt. Victoria Skinner, jouw betrokkenheid en de gezellige gesprekken blijven mij bij. Dank alle anderen die meegewerkt hebben aan de opzet van de LORD studie. Dear Leen Slaets and Kostas Tryfonidis, thank you for your help during protocol development, fine-tuning and approval of the LORD. My days at the EORTC headquarters in Brussels were fun because of you. Together with you, Berta Sousa, Fei Fei and Caroline Drukker, I enjoyed being a first baby dinosaur within the EORTC Breast Cancer Group. Dear EORTC Breast Cancer Group, especially Jan Bogaerts, Fatima Cardoso, David Cameron, Hervé Bonnefoi, Etienne Brain, Michail Ignatiadis, Elżbieta Senkus-Konefka, Suzette Delaloge, thank you for supporting the Young BCG and the entertaining meetings. Lieve Hester Oldenburg, Jacqueline Stouthard, Elsken van der Wall en Gonneke Winter-Warnars, in de afgelopen jaren heb ik met jullie gesproken over mijn vervolgtraject. Jullie zijn en blijven voor mij een voorbeeld. Lieve Kamer Flamingo; Rosa Djajadiningrat en Anke Kuijpers, en lieve Kamer Harvey Specter; Rosa van der Kaaij, Ravi Vermeulen en Anna Stiekema, geen onderzoekskamer was ooit meer over de top versierd en met vintage design ingericht dan de kamer waar ik in zat, I loved it. Dank voor al het lachen, de muziekjes, de theaterbezoeken, het huilen, het geshop, de PAN, de chocola, de koffie, de dansjes, de GT s en zoveel meer. Lieve mamma-onderzoekers; Mila en Bas, jullie waren eervolle voorgangers uit de Rutgers dynastie. Caroline Drukker, als ervaren kracht introduceerde jij me in de mamma-onderzoekers-wereld en wij werden een perfect mamma-duo (die zelfs een enerverende taxirit door een ingesneeuwd 220 Dankwoord

223 New York overleefden). Ik dank je voor je adviezen, tips & tricks en de bijpraatsessies waarin het combineren van opvoeden, opleiding en onderzoek ook altijd even besproken kan worden. Matthijs Nijenhuis, goede herinneringen aan: met jouw blauwe duivel naar de Veluwe, peking eend in Chinatown NYC, Virgil van Dijk aanmoedigen in Glasgow, Howlen at the moon in San Antonio en zoveel meer. Ik ben blij dat ik een groot deel van mijn promotietijd samen met jou kon doen. Mette van Ramshorst en Sophie Bosma, we vormden een mooi clubje drie klinische musketiers in de Wesseling groep. Dank voor de koffie breaks waarin we weer even de stand van zaken doorspraken. Mette, promoveren in dezelfde week nadat we een parallel pad aflegden op de MOD en HOD is bijzonder. Ik ben blij dat ik het samen met jou kan vieren! Onderzoekers van het O-gebouw; Sanne, Lisanne, Juliette, Minke, Max, Hanneke, Elies, Sarah, Marnix, Maarten, Simone, Merijn, Daphne, Fleur, Marieke, Roelien, Lisette, Tessa, Sheima, Noortje, Ronni, Jurrien, Laura, Sophie, Liset, Sharon, Jacqueline, dank voor alle gezelligheid, verjaardags- en publicatietaart en champagne op de gang als er opleidingsnieuws te vieren was. Jakob Kist, Gijs KleinJan en Ruud Wortel, de forenzen, blij dat jullie met wat mannelijk geweld ook het overpad naar het bloemige en pastelkleurige O wisten te vinden. Collega s en radiologen van het UMCU. Dank voor de goed georganiseerde start van mijn opleiding in Utrecht. Mede hierdoor is het mij gelukt mijn proefschrift dit jaar af te ronden. Lieve T4b, jullie zijn onze extended family. Hellen, ontwerpster van de Young BCG en LORD logo s, dankjewel voor de gezellige cover design avondjes aan onze ontplofte ronde keukentafel waarin wij geïnspireerd door theedoeken dolgelukkig waren met een tot knot opgestoken hairdo en ons tureluurs tuurden op een vergrootglas. Lieve vriendinnetjes en plus ones van Beer, HG74, hockey en Indonesië, ik prijs mij gelukkig met zulke dierbare vriendschappen. Lieve Yvonne, Gerrit, Kim, Gerolf, Isabel, Emilia, Gwen en Bjørn, dank voor alle belangstelling, maar ook zeker voor de oppasmomenten, uitgebreide familie tandartsbezoeken en gezellige en vertrouwde familievieringen in binnen- en buitenland over de jaren heen. Lieve Anne, mijn lieve vriendinnetje die mij zo goed kent en met wie ik lief en leed kan delen. Ik ben heel blij dat jij mijn paranimf bent. 9 Dankwoord 221

224 Lieve Maurits, onze band, samen met Dré, is goud waard. Ik kan mij een ondeugend broertje herinneren (Kan iemand Mauwits weghalen), die mij altijd in vertrouwen nam. Dat vertrouwen is wederzijds en nu sta jij naast mij als paranimf. Je bent mijn aller leukste liefste stoutste broertje. Lieve Andrea, klein ben je niet meer, maar wel mijn kleine zusje. Mijn kleine zusje die iedereen aftroeft, of het nou om voetbalweetjes, ingewikkelde formules, of vuurtjes stoken gaat. Ik ben zo trots op jou. Lieve papa en mama, jullie thuis is een thuis dat voor ons alle drie thuis blijft. Door jullie weet ik dat heel veel mogelijk is en dat je moet doorzetten. Ik voel mij door jullie onvoorwaardelijk gesteund en geliefd. Jullie zijn er altijd (Mam, met mij). Lieve mama, toen jij iets ouder was dan ik nu ging oma dood, aan de gevolgen van borstkanker. Ik was toen bijna 6 jaar, maar ik kan me nu beter voorstellen hoe erg je haar mist. Dit proefschrift is voor haar, maar ook voor jou, omdat jij altijd het beste voor mij wil. Lieve papa, jij bent mijn beste en hoogst gewaardeerde adviseur. Als ik de huislijn bel, weet ik dat er een grote kans bestaat dat ik met jou kan praten. Lieve Gregor en Maurits, Out with the boys werd de wish come true. Jullie zijn mijn grootste geluk. 222 Dankwoord

225 Dankwoord 223 9

226

227 CHAPTER 9 Summary Samenvatting List of co-authors and their contributions PhD portfolio Dankwoord Curriculum Vitae

228 226 Curriculum Vitae

229 Lotte Elisabeth Elshof was born on the 27 th of October 1986 in Leiden, the Netherlands, to Renée van Zonneveld and Albert Elshof. She spent her early childhood in Zambia and Indonesia. At the age of four, her family moved to Oegstgeest, the Netherlands, where Lotte grew up together with her brother Maurits and sister Andrea. She graduated from secondary education at the Rijnlands Lyceum in Oegstgeest in 2004 and continued with medical school at the Erasmus University in Rotterdam. For her scientific internship Lotte performed a research project on additional findings at pre-operative breast MRI at the Netherlands Cancer Institute Antoni van Leeuwenhoek, which she presented at the European Breast Cancer Conference in Barcelona in For this work Lotte was awarded the Nijbakker-Morra award. After her graduation in 2012 Lotte started as a surgical resident in the Netherlands Cancer Institute. She continued as a research physician to study clinical outcomes of a nationwide cohort of women treated for ductal carcinoma in situ (DCIS). She worked under the supervision of dr. Jelle Wesseling at the Division of Molecular Pathology of the Netherlands Cancer Institute in close collaboration with prof. dr. Emiel Rutgers (Division of Surgical Oncology), dr. Marjanka Schmidt (Division of Molecular Pathology), dr. Michael Schaapveld and prof. dr. ir. Flora van Leeuwen (both Division of Psychosocial Research and Epidemiology). During her PhD she attended the 2013 ECCO-AACR-EORTC-ESMO workshop on Methods in Clinical Cancer Research in Flims, Switzerland, to write the LORD study protocol. Since then, she has been involved as study co-coordinator in this randomized clinical trial that aims to evaluate the safety of active surveillance in women with LOw-Risk Dcis. To further develop her epidemiological and biostatistical skills she attended the EpidM Master s Programme in Epidemiology at the VU University Medical Centre in Amsterdam, which she completed in As of the 1 st of February 2017, Lotte started her specialisation in Radiology at the University Medical Centre in Utrecht under the supervision of dr. R.A.J. Nievelstein. She lives with Gregor Noltes and their son Maurits in Utrecht. 9 Curriculum Vitae 227