Citation for published version (APA): Laan, H. P. V. D. (2010). Optimising CT guided radiotherapy for breast cancer Groningen: s.n.

Transcription

1 University of Groningen Optimising CT guided radiotherapy for breast cancer Laan, Hans Paul van der IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite from it. Please check the document version below. Document Version Publisher's PDF, also known as Version of record Publication date: 2010 Link to publication in University of Groningen/UMCG research database Citation for published version (APA): Laan, H. P. V. D. (2010). Optimising CT guided radiotherapy for breast cancer Groningen: s.n. Copyright Other than for strictly personal use, 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), unless the work is under an open content license (like Creative Commons). Take-down policy If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim. Downloaded from the University of Groningen/UMCG research database (Pure): For technical reasons the number of authors shown on this cover page is limited to 10 maximum. Download date:

2

3 H.P. van der Laan Optimising CT guided radiotherapy for breast cancer ISBN (book) ISBN (file) Copyright 2010 H.P. van der Laan, Groningen NL All rights are reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic or mechanical, including photocopying and recording, or otherwise, without the written permission of the author. Cover The birth of Venus (detail), Sandro Botticelli ( ) Printed by Drukkerij van Denderen, Groningen NL Publication of this thesis was financially supported by Stichting Onderwijs en Onderzoek Radiotherapie UMC Groningen, Groningen University Institute for Drug Exploration (GUIDE), Integraal Kankercentrum Noord Oost, Cablon Medical B.V., Philips Radiation Oncology Systems and Elekta B.V.

4 Optimising CT guided radiotherapy for breast cancer Proefschrift ter verkrijging van het doctoraat in de Medische Wetenschappen aan de Rijksuniversiteit Groningen op gezag van de Rector Magnificus, dr. F. Zwarts, in het openbaar te verdedigen op woensdag 3 maart 2010 om uur door Hans Paul van der Laan geboren op 22 juli 1970 te Groningen

5 Promotor: Copromotor: Beoordelingscommissie: Prof. dr. J.A. Langendijk Dr. W.V. Dolsma Prof. dr. J.J.W. Lagendijk Prof. dr. J.T.M. Plukker Prof. dr. M. Verheij

6 Contents Chapter 1. General introduction 7 Chapter 2. Current technological clinical practise in breast radiotherapy; results of a survey in EORTC-Radiation Oncology Group affiliated institutions 17 Chapter 3. Dosimetric consequences of the shift towards CT guided target definition and planning for breast conserving radiotherapy 33 Chapter 4. Three-dimensional conformal simultaneously integrated boost technique for breast-conserving radiotherapy 51 Chapter 5. Limited benefit of inversely optimised intensity modulation in breast conserving radiotherapy with simultaneously integrated boost 69 Chapter 6. Comparison of normal tissue dose with three-dimensional conformal techniques for breast cancer irradiation including the internal mammary nodes 87 Chapter 7. Minimising contralateral breast dose in post-mastectomy intensity-modulated radiotherapy by incorporating conformal electron irradiation 109 Chapter 8. Summary and general discussion 127 Samenvatting en algemene discussie 143 Dankwoord / Acknowledgement 164 Curriculum Vitae 167 List of publications 168

7 - 6 -

8 Chapter 1 General introduction - 7 -

9 Chapter 1 General introduction Much has changed in the treatment of breast cancer since William Halsted ( ) in the early 1900 s first described the procedure for radical mastectomy [1]. At that time, surgery included the removal of the whole breast gland together with draining lymph nodes and pectoral muscles. The procedure was highly mutilating and patients often suffered from severe post-operative morbidity. Moreover, breast cancer very often recurred, causing a majority of patients to eventually die from their disease [2]. As time progressed, and better patient selection resulted in improved survival rates, the introduction of more conservative surgical techniques could limit the severity of post-operative morbidity [3]. Further improvements in breast cancer survival were obtained in the 1960 s, with the introduction of adjuvant chemotherapy and hormonal therapy [4,5]. More recently, the risk of a recurrence could be reduced with the introduction of new systemic treatments, such as targeted therapy [6]. Treatment outcomes could also be improved after the start of population-based screening programmes in the late 1980 s, new imaging modalities and new diagnostic tools, that enabled breast cancer to be detected in an earlier stage [7]. More recently, the introduction of new surgical methods, such as the sentinel node biopsy procedure, allowed for omitting axillary surgery in many patients, and thereby also contributed to a reduction of treatment-related morbidity [8]. The use of radiation in the treatment of breast cancer dates back from the early 1900 s. At that time, it was primarily given to inoperable patients with locally advanced disease, and treatment was limited to the use of low-energy orthovoltage irradiation that had rather unfavorable dose characteristics. It was not until the 1970 s that radiotherapy techniques improved, and randomized clinical trials demonstrated that conservative mastectomy combined with post-operative radiotherapy had survival rates similar to that of radical mastectomy [9]. Subsequent studies established that breast conserving surgery (lumpectomy) with post-operative radiotherapy had survival rates similar to that of mastectomy, and reduced the risk of a local recurrence [10]. More recently, it became clear that the - 8 -

10 General introduction risk of a local recurrence could be further reduced by delivering an additional dose of 16 Gy to the lumpectomy cavity after administration of 50 Gy to the whole breast [11]. A recent overview of the randomised breast cancer trails confirmed that for patients receiving either breast conserving surgery or mastectomy, the addition of radiotherapy could significantly decrease the breast cancer mortality risk after a follow-up of 15 years [12]. After breast conserving treatment for early stage breast cancer, recurrences were mostly located in the conserved breast, leaving the options open for salvage surgery. After mastectomy, in the case of more advanced breast cancer, recurrences were mainly located in the chest wall or regional lymph nodes. It appeared that in both patient groups, the absolute gain of radiotherapy in terms of breast cancer survival and overall survival was comparable [12]. After many years of progress, we have learned that the combination of surgery, adjuvant radiotherapy and systemic therapy, can reduce treatment-related morbidity and improve treatment outcome. However, it also became apparent that, especially in patients receiving post-mastectomy radiotherapy, the 15-year survival gain was somewhat smaller for overall mortality than for breast cancer mortality. This may be explained by the increase of radiation side effects in patients that received postmastectomy radiotherapy; i.e., while radiotherapy successfully decreased the risk of a local recurrence it also increased the risk of long-term radiotherapy-related mortality. Of all late complications, cardiovascular disease and second malignancies are considered the most serious since they do not only cause substantial morbidity but also considerable mortality [13]. A number of studies have demonstrated that patients with left-sided breast cancer who received radiotherapy after mastectomy had an increased risk of late cardiac mortality [14,15]. Although limited data is currently available on the precise risk of cardiac complications in relation to the irradiated volume and the dose delivered to particular regions of the heart, there is growing evidence that the risk of fatal cardiac complications can be reduced by limiting the irradiated volume of the heart during radiotherapy [16]. Other reports recently pointed at the incidence of second primary malignancies in the contralateral breast after radiotherapy [17]. Although it was already demonstrated - 9 -

11 Chapter 1 that the risk of breast cancer is associated with breast tissue dose [18], and even very low doses could increase this risk, there is now growing evidence from new studies confirming that the risk of second breast cancer is a matter of concern and efforts should be made during radiotherapy planning to reduce the volumes of the contralateral breast receiving a dose of 0.05 to 2 Gy [19,20]. While efforts to reduce the dose delivered to heart and contralateral breast could further improve overall survival for breast cancer patients, reducing excess dose to lungs, skin and ipsilateral breast tissue could substantially reduce treatment related morbidity and improve quality of life. It has been demonstrated that irradiation of lung tissue may result in radiation pneumonitis that could eventually lead to lung fibrosis and lung function loss [21]. A reduction of irradiated lung volume could therefore reduce the risk of such complications. The skin is also sensitive to radiation and radiotherapy may inflict acute effects such as erythema, moist desquamation and in some cases oedema [22]. Although these effects may be transient, they can be highly discomforting during a course of irradiation. In addition to the acute skin reactions, late effects such as hyperpigmentation and telangiectasia have also been reported frequently, particularly in patients receiving electron irradiation [23,24]. As breast conserving therapy has become common practise in early-stage breast cancer, cosmetic results are becoming increasingly important. Late fibrosis in irradiated breast tissue may manifest itself with symptoms such as increased breast tissue density and firmness, and also with retraction of the skin or nipple. Studies have shown that the risk of breast fibrosis could be reduced by limiting the volumes of breast tissue receiving a high dose [25]. Ongoing research in the field of breast cancer radiotherapy has resulted in the introduction of new techniques that improved dose uniformity in the designated target volumes, while reducing the dose delivered to organs at risk. In the last decade, breast radiotherapy planning has shifted from conventional fluoroscopy guided treatment simulation and two-dimensional (2D) dose planning to computer tomography (CT) guided treatment simulation and three-dimensional (3D) dose planning. CT image information can now be used for a 3D definition of the clinical

12 General introduction target volumes and organs at risk, while providing for a more accurate calculation of the dose distributions. New radiotherapy equipment allows for more efficient use of protective shielding in radiation beams, enabling the use of 3D-conformal radiotherapy (3D-CRT), that can adequately minimise the dose in organs at risk, while dose uniformity can be improved by means of intensity-modulated radiotherapy (IMRT) and computer-assisted dose optimisation. Although new radiotherapy technology has the potential to improve treatment outcomes in breast cancer patients, most of the actual benefits, drawbacks and limitations are still unknown. New treatment methods may help to further reduce the radiation dose in normal tissues and thereby decrease the risk of late radiationinduced morbidity and mortality, but they might also increase normal tissue volumes receiving a low to intermediate dose. New imaging modalities may provide for better identification of the clinical target volumes, but they could also lead to larger clinical target volumes, larger volumes of normal tissue irradiated and a greater risk of long term complications. This highlights the importance to study the benefits and harms of new radiotherapy technology. It appears that new technology can be applied in many different ways, and it may not be the new technology itself, rather its appropriate application, that could potentially contribute to longer complication free survival of breast cancer patients. It is therefore the aim of this research to investigate and optimise different applications of advanced technology in breast cancer radiotherapy, and to assess and improve the capability of these applications to reduce the radiation dose in normal tissues. The main focus of this thesis is on three different technological advances that all had a major impact on radiotherapy in the last decade: CT image guidance for clinical target volume delineation and radiotherapy treatment planning; 3D-CRT; and computer assisted treatment plan optimisation with IMRT. In the various chapters it was investigated to what extend applications of these technological advances contribute to a reduction of dose delivered to organs at risk. In chapters 2 to 5 the focus is on breast conserving radiotherapy, and in chapters 6 and 7 the focus is on loco-regional post-mastectomy radiotherapy. In chapter 2, a survey was conducted to determine to what extend CT image guidance, 3D-CRT and IMRT

13 Chapter 1 with corresponding methods and procedures are currently used in breast conserving radiotherapy throughout Europe. Because the dosimetric consequences of the introduction of CT image guided breast conserving radiotherapy have not been clearly assessed, it was investigated in chapter 3 how the shift to CT guided radiotherapy affected the delineation of clinical target volumes and the radiation dose in organs at risk. Traditionally, breast conserving radiotherapy comprises 25 daily fractions of breast irradiation followed by 5-10 fractions of irradiation (a boost) to the lumpectomy cavity. This so-called sequential boost irradiation has a number of practical and dosimetric disadvantages because with this standard method, breast and boost planning are performed separately. In chapter 4, a new method of incorporating breast and boost radiotherapy in one single 3D-CRT treatment plan was examined. The benefits of this new procedure, the simultaneously integrated boost (SIB) procedure, were compared to the conventional 3D-CRT method of separate breast and boost planning. In chapter 5, it was examined whether in the case of SIB, computer assisted treatment plan optimisation with IMRT (IMRT-SIB) might result in even lower doses in organs at risk than with 3D-CRT-SIB. In chapters 6 and 7 it was investigated whether applications of 3D-CRT and IMRT might reduce dose to organs at risk in loco-regional radiotherapy. With this particular treatment, not only the breast or chest wall, but also regional lymph nodes are irradiated. As in most cases this treatment requires relatively large irradiation fields, it can be quite a challenge to limit the dose in organs at risk, and the choice of radiotherapy technique can have a considerable impact. In chapter 6, a new 3D-CRT technique, developed to reduce the dose to the heart and contralateral breast, was compared to 3D-CRT techniques that are commonly used for loco-regional irradiation. Although it appeared that the new developed 3D-CRT technique was the overall best performing technique, it was examined in chapter 7 whether in loco-regional post-mastectomy radiotherapy, the application of IMRT might reduce the dose in organs at risk even further, below the values obtained with the new developed 3D-CRT technique

14 General introduction The specific aims of this thesis are: 1) To get insight in the current technological clinical practise in Europe and to determine to what extend new breast radiotherapy technology has been introduced. Chapter 2 2) To determine the consequences of the shift from 2D conventional target definition and treatment planning to CT guided target definition and 3D-CRT treatment planning for breast cancer by evaluating dose-volume parameters of planning target volumes and organs at risk. Chapter 3 3) To determine the potential benefit of a 3D-CRT simultaneously integrated boost (SIB) when compared to a conventional 3D-CRT sequential boost by evaluating dose-volume parameters of planning target volumes and organs at risk. Chapter 4 4) To determine the value of inversely planned IMRT in the application of SIB when compared to 3D-CRT-SIB with regard to the ability to reduce the radiation dose delivered to organs at risk. Chapter 5 5) To compare the dose delivered to organs at risk with different 3D-CRT techniques for loco-regional breast cancer and to determine the capability of a new 3D-CRT technique with conformal photon and electron irradiation to limit the dose delivered to organs at risk. Chapter 6 6) To determine the value of incorporating conformal electron irradiation in photon IMRT when compared to photon-only IMRT and photon-electron 3D-CRT, by evaluating dose-volume parameters of planning target volumes and organs at risk. Chapter

15 Chapter 1 References 1. Halsted WS. The Results of Radical Operations for the Cure of Carcinoma of the Breast. Ann Surg 1907;46: Lewis D, Rienhoff WF. Results of Operations at the Johns Hopkins Hospital for Cancer of the Breast: Performed at the Johns Hopkins Hospital from 1889 to Ann Surg 1932;95: Scanlon EF, Caprini JA. Modified radical mastectomy. Cancer 1975;35: Bonadonna G, Brusamolino E, Valagussa P, et al. Combination chemotherapy as an adjuvant treatment in operable breast cancer. N Engl J Med 1976;294: Early Breast Cancer Trialists' Collaborative Group Effects of chemotherapy and hormonal therapy for early breast cancer on recurrence and 15-year survival: an overview of the randomised trials. Lancet 2005;365: Viani GA, Afonso SL, Stefano EJ, et al. Adjuvant trastuzumab in the treatment of her- 2-positive early breast cancer: a meta-analysis of published randomized trials. BMC Cancer 2007;7: Nystrom L, Andersson I, Bjurstam N, et al. Long-term effects of mammography screening: updated overview of the Swedish randomised trials. Lancet 2002;359: Kelly AM, Dwamena B, Cronin P, et al. Breast cancer sentinel node identification and classification after neoadjuvant chemotherapy-systematic review and meta analysis. Acad Radiol 2009;16: Fisher B, Redmond C, Fisher ER, et al. Ten-year results of a randomized clinical trial comparing radical mastectomy and total mastectomy with or without radiation. N Engl J Med 1985;312: Fisher B, Redmond C, Poisson R, et al. Eight-year results of a randomized clinical trial comparing total mastectomy and lumpectomy with or without irradiation in the treatment of breast cancer. N Engl J Med 1989;320: Bartelink H, Horiot JC, Poortmans PM, et al. Impact of a higher radiation dose on local control and survival in breast-conserving therapy of early breast cancer: 10-year results of the randomized boost versus no boost EORTC trial. J Clin Oncol 2007;25: Clarke M, Collins R, Darby S, et al. Effects of radiotherapy and of differences in the extent of surgery for early breast cancer on local recurrence and 15-year survival: an overview of the randomised trials. Lancet 2005;366:

16 General introduction 13. Hooning MJ, Aleman BM, van Rosmalen AJ, et al. Cause-specific mortality in longterm survivors of breast cancer: A 25-year follow-up study. Int J Radiat Oncol Biol Phys 2006;64: Rutqvist LE, Lax I, Fornander T, et al. Cardiovascular mortality in a randomized trial of adjuvant radiation therapy versus surgery alone in primary breast cancer. Int J Radiat Oncol Biol Phys 1992;22: Gagliardi G, Lax I, Ottolenghi A, et al. Long-term cardiac mortality after radiotherapy of breast cancer--application of the relative seriality model. Br J Radiol 1996;69: Nixon AJ, Manola J, Gelman R, et al. No long-term increase in cardiac-related mortality after breast-conserving surgery and radiation therapy using modern techniques. J Clin Oncol 1998;16: Tubiana M. Can we reduce the incidence of second primary malignancies occurring after radiotherapy? A critical review. Radiother Oncol 2009;91: Carmichael A, Sami AS, Dixon JM, Breast cancer risk among the survivors of atomic bomb and patients exposed to therapeutic ionising radiation. Eur J Surg Oncol 2003;29: Hooning MJ, Aleman BM, Hauptmann M, et al. Roles of radiotherapy and chemotherapy in the development of contralateral breast cancer. J Clin Oncol 2008;26: Stovall M, Smith SA, Langholz BM, et al. Dose to the contralateral breast from radiotherapy and risk of second primary breast cancer in the WECARE study. Int J Radiat Oncol Biol Phys 2008;72: Semenenko VA, Li XA. Lyman-Kutcher-Burman NTCP model parameters for radiation pneumonitis and xerostomia based on combined analysis of published clinical data. Phys Med Biol 2008;53: Freedman GM, Anderson PR, Li J, et al. Intensity modulated radiation therapy (IMRT) decreases acute skin toxicity for women receiving radiation for breast cancer. Am J Clin Oncol 2006;29: Johansen J, Overgaard J, Rose C, et al. Cosmetic outcome and breast morbidity in breast-conserving treatment; results from the Danish DBCG-82TM national randomized trial in breast cancer. Acta Oncol 2002;41: Huang EY, Chen HC, Wang CJ, et al. Predictive factors for skin telangiectasia following post-mastectomy electron beam irradiation. Br J Radiol 2002;75:

17 Chapter Borger JH, Kemperman H, Smitt HS, et al. Dose and volume effects on fibrosis after breast conservation therapy. Int J Radiat Oncol Biol Phys 1994;30:

18 Chapter 2 Current technological clinical practise in breast radiotherapy; results of a survey in EORTC- Radiation Oncology Group affiliated institutions Hans Paul van der Laan, Coen W. Hurkmans, Abraham Kuten, Helen A. Westenberg, on behalf of the EORTC-ROG Breast Working Party Accepted for publication in Radiotherapy & Oncology

19 Chapter 2 Abstract Purpose: To determine the current technological clinical practise of radiation therapy of the breast in Europe. Materials and Methods: A survey was conducted between August 2008 and January 2009 on behalf of the Breast Working Party within the EORTC Radiation Oncology Group. The questionnaire comprised 32 questions on 4 main topics: fractionation schedules, treatment planning methods, volume definitions and position verification procedures. Results: Sixty-eight institutions out of 16 countries responded (a response rate of 47%). The standard fraction dose was generally 2 Gy for both whole breast and lumpectomy cavity (boost) treatment, although a 2.67 Gy boost fraction dose is routinely given in the United Kingdom. A simultaneously integrated boost fractionation is implemented in 23% of the institutions and is the standard choice of fractionation in a third of these institutions. The main boost modality was electrons in 55%, photons in 47% and brachytherapy in 3% of the institutions (equal use of photon and electron irradiation in 5% of the institutions). All institutions used computed tomography guided treatment planning. Wide variations are seen in the definition of the breast and boost target volumes, with margins around the lumpectomy cavity ranging from 0-30 mm. Inverse planned intensity modulated radiotherapy (IMRT) is available in 27% and breath-hold techniques in 19% of the institutions. The number of patients treated with IMRT and breath-hold varied per institution. Electronic portal imaging for patient set-up is used by 92% of the institutions. Conclusion: This survey has established precise details of radiotherapy techniques currently implemented for breast irradiation in Europe

20 Survey of technological practice in breast radiotherapy Introduction Randomised controlled trials are regarded to be the foundation for evidencebased medicine. They have shown to improve the various standards of care. With respect to the treatment of cancer, improving treatment standards by conducting clinical trials is an important goal of the European Organisation for Research and Treatment of Cancer (EORTC). To date, many successful trials have been conducted by the EORTC-Radiation Oncology Group (ROG), including trials in the field of radiotherapy for breast cancer. Based on these and other trials, adjuvant radiotherapy to the breast is now considered part of the standard of care in breast conserving therapy. In the past years there is a growing awareness of the necessity of homogeneity in radiation treatment across institutions, especially when a particular treatment is being evaluated in clinical trials. Not only should radiotherapy be applied according to international standards, but also the various components within the radiotherapy process, such as target volume and organs at risk definitions, dose-fractionation schedule, overall treatment time, applied techniques, etc., should be described and performed in a consistent and thus comparable manner. These details can potentially have an important influence on trial outcome and accounts for the fact that nowadays quality assurance is part of any radiotherapy trial [1-3]. Breast cancer radiotherapy techniques have evolved considerably over the last years, due to the wider availability of computed tomography (CT), the introduction of intensity modulated radiotherapy (IMRT) and the use of image guided radiotherapy (IGRT) techniques. When designing new breast cancer radiotherapy trials it is important to know to what extend participating radiotherapy institutions have implemented these techniques. Only a few surveys focussed specifically on the technological aspects of breast cancer irradiation [4-9]. Although some of these articles provide interesting data from either a specific European country or on a specific breast cancer radiotherapy technique, no general overview exists of breast irradiation techniques currently used in Europe. To generate such an overview, a

21 Chapter 2 survey was conducted on behalf of the Breast Working Party within the EORTC- ROG. Materials and Methods A questionnaire was developed jointly by the radiotherapy departments of the University Medical Center Groningen (UMCG) and the Catharina Hospital in Eindhoven, The Netherlands. The questionnaire comprised 32 questions on 4 main topics: fractionation schedules, treatment planning methods, volume definitions and position verification procedures. The majority of the questions (23) were closed questions. Open questions were used to retrieve detailed information on the closed questions. For example, to determine the reason why certain clinical target volume (CTV) to planning target volume (PTV) margins were chosen. The closed questions included quantitative questions and multiple choice questions. When relevant, questions concentrated on whole breast irradiation, lumpectomy cavity (boost) irradiation, or both. The questionnaire was developed as a web-based application within the freeware tool Thesistools ( Using this web-based system, respondents were enabled to type in data and to select the appropriate answer from a list of predefined answers. Furthermore, this method allowed for easy access to the questionnaire and enabled convenient analysis of the collected data. The questionnaire was first tested by colleagues with expertise in the field of breast radiotherapy and adjustments were made based on their comments before it was distributed. By using the EORTC-ROG membership mailing list, s were sent with the request to complete the questionnaire. The invitation was successfully delivered to representatives of 145 EORTC-ROG institutions spread over 26 countries. Although the questionnaire was also attached as a Microsoft Word document, respondents were encouraged to use the web link provided in the to complete the questionnaire online on the web. A personal code was needed to gain access to the questionnaire as well as adding data on later occasions. The questionnaire was first distributed in August 2008 and reminders were sent in

22 Survey of technological practice in breast radiotherapy September and December of the same year. Results were analysed per institution and statistics were calculated for all institutions together. Results Response The response rate was 47% (68 / 145 institutions). It included responses from 16 countries: Austria (2); Belgium (6); France (12); Germany (4); Hungary (1); Israel (1); Italy (5); Lithuania (1); The Netherlands (17); Poland (1); Portugal (1); Slovenia (1); Spain (3); Sweden (1); Switzerland (7); and United Kingdom (5). The main results of the survey are summarised in Table 1. Table 1. Condensed overview of percentages of institutions applying specific techniques Application Percentages of institutes Standard fractionation schedule Breast 25 x 2 Gy: 72%; Boost 5-10 x 2 Gy: 83% CT-based treatment plan Breast: 100%; Photon boost: 100% Conformal technique based on CT defined targets Breast: 90%; Photon boost: 96% Inverse planned IMRT Breast: 27%; Photon boost: 14% Boost delivery modality Sequential: 98%; Concomitant: 14%; SIB: 23% Main boost modality a Electrons: 55%; Photons: 47%; Brachytherapy: 3% Patient setup verification 92% Partial breast irradiation b 21% (<5% of patients) Prone position irradiation b 12% (<1% of patients c ) Breath-hold b 19% (1-30% of patients) Abbreviations: CT = computer tomography; IMRT = intensity modulated radiation therapy; SIB = simultaneously integrated boost. a Equal use of photon and electron irradiation in 5% of the institutions. b The response on these questions was limited (see paragraph on new technologies). c One institution treats 8% of their patients in prone position. Fractionation schedules For whole breast radiotherapy, the most common institutions standard fractionation schedule is 25 fractions of 2 Gy (72% of the institutions). Four institutions customary prescribe 15 fractions of 2.67 Gy (4/5 institutions in the United Kingdom). A higher dose per fraction was also customary prescribed by

23 Chapter 2 three other institutions: 22 times 2.3 Gy, 18 times 2.5 Gy or 17 times 2.5 Gy. All four responding institutions from Germany prescribe 28 fractions of 1.8 Gy. Eight institutions in The Netherlands have implemented a simultaneously integrated boost (SIB) fractionation schedule, five of which use this schedule as their standard schedule [10,11]. Seven of these institutions prescribe 28 fractions with a daily dose of 2.3 Gy delivered to the boost volume and 1.8 Gy to the remainder of the breast. The remaining institution prescribes 25 fractions of 2.75 Gy for the boost and 25 fractions of 2.0 Gy to the remainder of the breast. Seven non-dutch institutions also apply a SIB, but only in a limited proportion of their patients. As for boost irradiation, most institutions (83%) use a sequential boost delivered with a standard daily fraction size of 2 Gy for 5 to 10 fractions (Figure 1). Figure 1. Fractionation schedules Distribution of standard fraction sizes (in Gy) for breast (A) and boost (B) irradiation over institutions. Number of institutions is given in brackets. Simultaneously integrated boost (SIB) is planned and delivered simultaneously with whole breast plan. Treatment planning methods CT guided treatment planning was used in all responding institutions, with 4 institutions indicating to use body-outline contours in some of their patients. Specific questions were directed to the different treatment planning methods. With respect to CT guided boost planning, questions were restricted to irradiation by photon beams. CT is used for electron density-based dose calculations as well as

24 Survey of technological practice in breast radiotherapy conformal planning based on the target volumes drawn on CT in the vast majority of institutions. In 10% and 4% of the institutions, use of CT is limited to electron density-based dose calculations for the breast and boost plans, respectively. Manual optimisation of dose uniformity (forward planning) for breast and photon boost plans is performed in 89% and 85% of the institutions. Inverse planning (objective-based IMRT) is performed in 27% and 14% of the institutions, respectively, of which only 5 institutions use IMRT in more than 20% of their patients. Boost treatment methods Sequential delivery of the boost (after whole breast irradiation), is performed for some or all patients in 98% of the institutions, while a concomitant boost (a separately planned boost plan delivered on the same day as the whole breast irradiation) is used in 14% of the institutions. SIB is used in 23% of the institutions. More specifically, 8 Dutch institutions treat on average 72% of their patients with SIB, while 7 non-dutch institutions treat on average 5% of their patients with SIB. The boost delivery method most commonly used (used in 50% of the institutions patients) is electron irradiation in 55%, photon irradiation in 47% and brachytherapy in 3% of the institutions (equal use of photon and electron irradiation in 5% of the institutions). When photon beams are used for boost delivery, the common number of different boost gantry angles is 2 (in 56% of the institutions) and 3 (in 35% of the institutions). Photon boost beam directions are tangential only (37%), non-tangential only (7%) or both tangential and nontangential (in 56% of the institutions). Target volume delineation When breast target volume delineation is performed, various references and landmarks are used: radiopaque wires visible on CT are used in 59%, glandular breast tissue as visible on CT is used in 69% and bony structures as visible on CT are used in 28% of the institutions. For the purpose of boost target volume definition, surgical clips, when available, are used in 95% of the institutions, while

25 Chapter 2 hematoma and seroma visible on CT are used in 49% of the institutions. Availability of surgical clips varies largely, between different countries but also within countries. No institution reported never to have clips available. On average, surgical clips are available in 56% of the patients. Large variations among institutions and countries are also observed with regard to the various margins used in breast and boost target volume definition. Zero, 5 and 10 mm are the most commonly used breast CTV to PTV margins (Table 2). Few institutions apply a larger margin of 15 mm (3 institutions) or 20 mm (2 institutions). The variation in margin from lumpectomy cavity to boost PTV is even more widespread, ranging from 0 to 30 mm. Twelve institutions (22%) take the resection free margin as stated in the pathology report into account in their margin. Ten out of these 12 institutions, all from The Netherlands, have the policy to use a margin of 20 mm minus the resection free margin when available. Table 2. Margins applied for breast and boost planning target volumes Breast CTV to breast PTV Lumpectomy cavity to boost PTV a Margin n=51 Institutions (%) Margin 0 mm 22 0 mm 7 4 mm 2 5 mm 4 5 mm mm 13 7 mm 4 15 mm mm mm mm 6 25 mm 7 20 mm 4 30 mm mm 2 15 mm minus 5 mm free margin mm 2 20 mm minus free margin mm 2 25 mm minus free margin 2 n=55 Institutions (%) Abbreviations: CTV = clinical target volume; PTV = planning target volume. a Lumpectomy cavity to boost PTV margin was binned using 5 mm increments and based on the summation of lumpectomy cavity to boost CTV margin and boost CTV to boost PTV margin. Organs at risk delineation Three-dimensional (3D) delineation of organs at risk (OAR) is performed in 95% of responding institutions. This involves delineation of the heart in 78%, ipsilateral lung in 92%, contralateral lung in 52%, and contralateral breast in 23% of the institutions that perform OAR delineation. Dose-volume histograms of

26 Survey of technological practice in breast radiotherapy delineated OARs are used to decide on plan acceptance in relation to specific criteria in 88% of the responding institutions for all or selected patients. In addition, the Central Lung Distance (CLD) and Maximum Heart Distance (MHD [12]) are used as a criterion for treatment plan acceptance in 71% (max CLD 2-3 cm) and 59% (max MHD cm) of responding institutions, respectively, in all or selected patients. Position verification procedures Some of the questions regarding position verification procedures had limited response. Eleven out of 28 institutions (39%) reported to use X-ray film for position verification, while 48 out of 52 institutions (92%) reported to use electronic portal imaging (EPI). In 4 institutions, both X-ray film and EPI are used, each in approximately 50% of the patients. Cone-beam CT is used in 7 out of the 23 institutions that responded to the corresponding question. New technologies / strategies Specific questions were added to identify the use of partial breast irradiation (PBI), irradiation in prone position, and breath-hold techniques. It appeared that PBI is used in 14 institutions spread over 13 countries, mostly in selected patients i.e., patients treated within clinical trials (generally <5% of the local population). Irradiation in prone position is rarely used. Eight institutions in 6 countries treat <1% of their patients with this technique, with the exception of one of these institutions, where 8% of patients are treated in prone position. Prone irradiation is used to irradiate pendulous breasts as a means to reduce lung exposure. Breathhold techniques are used in 13 institutions spread over 8 countries in 1-30% of the patients. Patients selected for treatment with breath-hold techniques are mostly young patients that have a higher probability of cardiac complications due to involvement of the heart in the radiation fields

27 Chapter 2 Discussion Fractionation schedules The predominant dose per fraction size for breast irradiation is still 2 Gy in Europe, with only 3 institutions outside the UK using a higher fraction size. It is interesting to note that 4 out of the 5 UK institutions that responded, use a fractionation schedule of 15 fractions of 2.67 Gy, which is the fractionation schedule used in the UK START B trial [13]. Hypofractionation as such, despite the published results of clinical trials [14,15], does not seem to have been implemented in daily practise. Very recently, a few institutions in the Netherlands have introduced hypofractionation for selected patients and the discussion to incorporate the results into the national treatment guidelines has started (source: personal communication). The prevalent fraction size for boost irradiation is also 2 Gy, although a wider variation in both fraction size and total dose exists compared to the whole breast fractionation schedule. The SIB technique has been remarkably rapid put into clinical practise in The Netherlands [10,11,16]. Eight out of 17 institutions perform this technique for some or all patients since it was first put into clinical use in The Netherlands at the UMCG in March Treatment planning methods The use of CT has increased rapidly in the last years and all institutions now routinely use CT scans for treatment planning. This is a major difference compared to the results found in e.g., a survey in the United Kingdom performed between 1997 and 1999 where only 2 out of 46 institutions used CT [5], and in an Australian survey published in 1999 where only 3 out of 11 institutions used CT [17]. CT is now used for delineation of target volumes and organs at risk, density corrections, shielding definition and manual or inverse plan optimisation. There are data available in the literature that suggest an advantage of (inversely optimised) IMRT over conventional (non-optimised) treatment of the breast with regard to a reduction in acute and late breast and skin toxicity [18-20]. It is therefore expected

28 Survey of technological practice in breast radiotherapy that the use of IMRT will increase over the coming years as a means to improve dose homogeneity in the breast and to standardise treatment planning procedures. Partial breast irradiation could potentially show even more advantageous than IMRT in selected patients, since it allows much smaller volumes of the breast and surrounding normal tissues to be irradiated. Boost treatment methods The rapid introduction of SIB by Dutch institutions might be explained by the fact that in The Netherlands the majority of institutions participate in a prospectively randomised multi-centre trial, investigating the value of a 26 Gy versus the standard 16 Gy boost dose in patients 50 years of age (the Young Boost Trial ). In that trial, the margins for CTV and PTV are carefully described and the SIB technique is proposed as one of the standard techniques. In the case of an electron boost, (conformal) CT guided target definition and planning are rarely performed. This might be because a SIB using electron boost fields is much more challenging concerning treatment planning. Furthermore, incorporating both photons and electrons in every treatment session is quite labour intensive. Also target coverage, particularly at the deeper parts of the target volume, might be less adequate with electron beams than with CT-based photon techniques [21]. However, sub-analysis of the EORTC boost no-boost trial has not shown a difference in local recurrence rates between electron and photon boosts [22]. Thus, one might argue that electron irradiation is just as effective in preventing local recurrences. It should however be noted that also photon boosts in the aforementioned study were not CT-based in most cases. Target volume delineation Although CT seems to be the current standard for treatment planning, there is a very large variation in the definition of the target volumes; especially CTV to PTV margins applied for breast and boost vary significantly. Obviously, CT images do not provide good contrast for breast and boost delineation. Therefore, identification of the mammary gland by palpation for the purpose of CTV definition

29 Chapter 2 remains an important reference for many radiation oncologists. Furthermore, the survey did not include questions regarding the rationale of the margins used, e.g., whether or not they are derived from studies on local position verification measurements. Still, it is observed that large variations in margins are seen in institutions that can be assumed to participate in EORTC-ROG clinical trials, in which margins are often prescribed as part of the trial protocol. Conversely, in The Netherlands, it seems that all institutions participating in the aforementioned Young Boost Trial use the same lumpectomy cavity to boost PTV margin. Particularly in the case of highly conformal irradiation, such as advanced 3D conformal radiotherapy (3D-CRT)-SIB or IMRT-SIB, margin selection is becoming increasingly important. The introduction of CT guided treatment planning has shown to generally result in an increased boost target volume [22], stressing the importance even more to limit the CTV to PTV boost margins as much as possible. Furthermore, it has been shown previously that the boost volume may change during a course of radiotherapy [16]. As a result, adaptive treatment planning techniques aimed at minimising the boost margins while maintaining target coverage will probably become more important in the near future. Position verification procedures The majority of the institutions indicated to use EPI for patient set-up verification. This is in substantial contrast with results from a UK survey published in 2002, where only half of the institutions performed set-up verification [5]. New technologies and strategies The increased use of breath-hold techniques, together with the high frequency of delineation of the heart as organ at risk and the frequent use of the maximum heart distance as criterion for plan acceptance indicates that the reduction of heart dose is a matter of concern in the treatment of many patients, and that methods to prevent late cardiac complications are used more frequently than before. Irradiation in prone position can also be used to reduce the dose to critical structures such as the heart. However, probably due to the limited applicability

30 Survey of technological practice in breast radiotherapy (pendulous breasts) and practical limitations of this technique it is not yet used on a large scale. There also seems to be a reluctance to prescribe PBI as routine treatment to patients and only patients in trials receive such irradiation. However, the outcome of these clinical trials may eventually lead to an increased use of this technique. Conclusions This survey among European radiotherapy institutions has established that recent advances in radiotherapy technology are currently widely adopted for the treatment of breast cancer. All responding institutions reported to use CT guided treatment planning. 3D-CRT and EPI-based patient set-up verification are now in mainstream use, with IMRT techniques being used by 27% of the institutions. This indicates that new radiotherapy techniques, when being addressed in clinical trials, are feasible in the network of EORTC-ROG institutions. The boost is applied sequentially in 98% of the responding institutions. The SIB technique is used in 8 Dutch institutions that treat on average 72% of their patients with SIB, while 7 non-dutch institutions treat on average 5% of their patients with SIB. Our survey also reveals considerable variations between institutions, especially in boost delineation and applied margins. When designing new radiotherapy trials, quality assurance should focus on these issues because we found these to have the most variability compared to other radiotherapy details. For Dutch institutions participating in the Young Boost Trial, we found that trial participation increases consistency among institutions with respect to the use of treatment methodology, ensuring high quality radiation treatment available for patients outside clinical trials

31 Chapter 2 References 1. Bolla M, Bartelink H, Garavaglia G, et al. EORTC guidelines for writing protocols for clinical trials of radiotherapy. Radiother Oncol 1995;36: Poortmans PM, Venselaar JL, Struikmans H, et al. The potential impact of treatment variations on the results of radiotherapy of the internal mammary lymph node chain: a quality-assurance report on the dummy run of EORTC Phase III randomized trial 22922/10925 in Stage I-III breast cancer. Int J Radiat Oncol Biol Phys 2001;49: Peters L, O'Sullivan B, Girald J, et al. Critical impact of radiotherapy protocol compliance and plan quality in treatment of advanced head and neck cancer (Abstr.). 2nd ICHNO conference abstract book 2009:s6. 4. Delaney G, Blakey D, Drummond R, et al. Breast radiotherapy: an Australasian survey of current treatment techniques. Australas Radiol 2001;45: Winfield E, Deighton A, Venables K, et al. Survey of UK breast radiotherapy techniques: background prior to the introduction of the quality assurance programme for the START (standardisation of radiotherapy) trial in breast cancer. Clin Oncol (R Coll Radiol ) 2002;14: Jalali R, Singh S, Budrukkar A. Techniques of tumour bed boost irradiation in breast conserving therapy: current evidence and suggested guidelines. Acta Oncol 2007;46: Vu TT, Pignol JP, Rakovitch E, et al. Variability in radiation oncologists' opinion on the indication of a bolus in post-mastectomy radiotherapy: an international survey. Clin Oncol (R Coll Radiol ) 2007;19: Thomsen MS, Berg M, Nielsen HM, et al. Post-mastectomy radiotherapy in Denmark: from 2D to 3D treatment planning guidelines of The Danish Breast Cancer Cooperative Group. Acta Oncol 2008;47: Morgia M, Lamoury G, Morgan G. Survey of radiotherapy planning and treatment of the supraclavicular fossa in breast cancer. J Med Imaging Radiat Oncol 2009;53: Hurkmans CW, Meijer GJ, van Vliet-Vroegindeweij C, et al. High-dose simultaneously integrated breast boost using intensity-modulated radiotherapy and inverse optimization. Int J Radiat Oncol Biol Phys 2006;66:

32 Survey of technological practice in breast radiotherapy 11. van der Laan HP, Dolsma WV, Maduro JH, et al. Three-dimensional conformal simultaneously integrated boost technique for breast-conserving radiotherapy. Int J Radiat Oncol Biol Phys 2007;68: Hurkmans CW, Borger JH, Bos LJ, et al. Cardiac and lung complication probabilities after breast cancer irradiation. Radiother Oncol 2000;55: Bentzen SM, Agrawal RK, Aird EG, et al. The UK Standardisation of Breast Radiotherapy (START) Trial B of radiotherapy hypofractionation for treatment of early breast cancer: a randomised trial. Lancet 2008;371: Bentzen SM, Agrawal RK, Aird EG, et al. The UK Standardisation of Breast Radiotherapy (START) Trial A of radiotherapy hypofractionation for treatment of early breast cancer: a randomised trial. Lancet Oncol 2008;9: Whelan T, MacKenzie R, Julian J, et al. Randomized trial of breast irradiation schedules after lumpectomy for women with lymph node-negative breast cancer. J Natl Cancer Inst 2002;94: Hurkmans CW, Admiraal M, van der Sangen M, et al. Significance of breast boost volume changes during radiotherapy in relation to current clinical interobserver variations. Radiother Oncol 2009;90: Veness MJ, Delaney G. Variations in breast tangent radiotherapy: a survey of practice in New South Wales and the Australian Capital Territory. Australas Radiol 1999;43: Pignol JP, Olivotto I, Rakovitch E, et al. A multicenter randomized trial of breast intensity-modulated radiation therapy to reduce acute radiation dermatitis. J Clin Oncol 2008;26: Donovan E, Bleakley N, Denholm E, et al. Randomised trial of standard 2D radiotherapy (RT) versus intensity modulated radiotherapy (IMRT) in patients prescribed breast radiotherapy. Radiother Oncol 2007;82: Harsolia A, Kestin L, Grills I, et al. Intensity-modulated radiotherapy results in significant decrease in clinical toxicities compared with conventional wedge-based breast radiotherapy. Int J Radiat Oncol Biol Phys 2007;68: Benda RK, Yasuda G, Sethi A, et al. Breast boost: are we missing the target? Cancer 2003;97: van der Laan HP, Dolsma WV, Maduro JH, et al. Dosimetric consequences of the shift towards computed tomography guided target definition and planning for breast conserving radiotherapy. Radiat Oncol 2008;3:6-31 -

33 - 32 -

34 Chapter 3 Dosimetric consequences of the shift towards computed tomography guided target definition and planning for breast conserving radiotherapy Hans Paul van der Laan, Wil V. Dolsma, John H. Maduro, Erik W. Korevaar, Johannes A. Langendijk Radiation Oncology 2008; 3:

35 Chapter 3 Abstract Purpose: The shift from conventional two-dimensional (2D) to threedimensional (3D)-conformal target definition and treatment planning seems to have introduced volumetric as well as geometric changes. The purpose of this study was to compare coverage of computed tomography (CT)-based breast and boost planning target volumes (PTV), absolute volumes irradiated, and dose delivered to the organs at risk with conventional 2D and 3D-conformal breast conserving radiotherapy (3D-CRT). Materials and Methods: Twenty-five patients with left-sided breast cancer were subject of CT guided target definition and 3D-CRT treatment planning, and conventionally defined target volumes and treatment plans were reconstructed on the planning CT. Accumulated dose distributions were calculated for the conventional and 3D-CRT treatment plans, taking into account a prescribed dose of 50 Gy for the breast plans and 16 Gy for the boost plans. Results: With conventional treatment plans, CT-based breast and boost PTVs received the intended dose in 78% and 32% of the patients, respectively, and smaller volumes received the prescribed breast and boost doses compared with 3D- CRT treatment planning. The mean lung dose, the volume of the lungs receiving 20 Gy, the mean heart dose, and volume of the heart receiving 30 Gy were significantly less with conventional treatment plans. Specific areas within the breast and boost PTVs systematically received a lower than intended dose with conventional treatment plans. Conclusion: The shift towards CT guided target definition and planning as the golden standard for breast conserving radiotherapy has resulted in improved target coverage at the cost of larger irradiated volumes and an increased dose delivered to organs at risk. Tissue is now included into the breast and boost target volumes that was never explicitly defined or included with conventional treatment. Therefore, a coherent definition of the breast and boost target volumes is needed, based on clinical data confirming tumour control probability and normal tissue complication probability with the use of 3D-CRT

36 The shift towards CT guided breast radiotherapy Introduction Ever since the early days of breast cancer radiotherapy, irradiation was performed by means of tangential beams directed to treat the whole breast or chest wall [1]. With the use of tangential beams, non-target thoracic structures were avoided as much as possible. To ensure that all breast parenchyma was included into the target volume, one relied upon visible or palpable anatomy as assessed by physical examination and/or fluoroscopy [2]. Standard field borders were usually placed within a certain range outside the palpable breast, while field projections and collimator angles were verified and adapted by means of radiographic examination. To enable computed dose calculation and optimisation of wedgefractions, one or more body-outline contours were provided on which treatment planning, with or without lung-density correction, was performed [3]. However, the breast clinical target volume (CTV), i.e., the glandular breast tissue, was never explicitly defined. Currently, breast cancer radiotherapy has gradually shifted towards computed tomography (CT) guided treatment planning. This enabled the application of new techniques such as three-dimensional (3D)-conformal radiotherapy (3D-CRT) and intensity modulated radiotherapy (IMRT) [4,5]. With these techniques, an accurate delineation of the target volume is critical because its size and shape directly affects the amount of normal tissue irradiated. However, with regard to the definition of the breast CTV, there is still no general consensus, and target volume delineation is subject to a large interobserver variability [6,7]. This may be explained by the fact that it can be difficult to distinguish the glandular breast tissue from the surrounding fatty tissue. In an effort to solve this problem, the palpable breast is often marked with a radiopaque wire during the CT scan [6]. The breast CTV is then defined within the CT images, guided by this radiopaque wire. Subsequently, a planning target volume (PTV) can be defined and 3D-CRT breast beams can be constructed. It appeared that large discrepancies exist between a CT-based beam set-up and beams defined during the conventional process of direct simulation [8]

37 Chapter 3 The introduction of CT guided treatment planning also seems to have influenced the way the lumpectomy cavity with corresponding CTV and PTV are defined [9]. Nowadays, surgical clips, hematoma, seroma and other surgical changes are used to define the lumpectomy cavity (boost) target volume in 3D, while in the conventional setting, information was limited to the location of the scar and, when available, the position of surgical clips. Although several investigators drew attention to the volumetric and geometric changes introduced with CT guided treatment planning in breast conserving radiotherapy, the dosimetric consequences, i.e., target coverage and dose delivered to normal tissues, have not been clearly assessed. Therefore, the purpose of this study was to compare coverage of CT-based breast and boost PTVs, absolute volumes irradiated, and dose delivered to the organs at risk with conventional treatment plans and CT guided 3D-conformal breast conserving radiotherapy. Materials and Methods Patients and CT scanning Twenty-five patients with early-stage left-sided breast cancer that underwent radiotherapy after breast-conserving surgery were included in this study. A planning CT scan in treatment position was made for each patient. Before the CT scan, skin marks were placed to locate the boost-volume isocenter and enable patient repositioning during treatment. Radiopaque wires and markers were placed to locate palpable breasts, scars, and skin marks on the CT images. In addition, markers were placed to represent the conventional field borders (a mid-sternal marker, representing the medial field border, and a marker placed mm dorsally from the lateral palpable breast representing the lateral field border). The cranial and caudal field borders were marked 15 mm beyond the palpable breast. Patients were scanned with CT from the level of the larynx to the level of the upper abdomen, including both lungs, with a scan thickness and index of 5 mm. The CT data for all patients were transferred to the Helax-TMS 3D treatment planning system, version 6.1B (Nucletron, Veenendaal, The Netherlands). All patients

38 The shift towards CT guided breast radiotherapy provided informed consent before starting therapy, and the ethics committee at the University Medical Center Groningen approved the procedures followed. Reconstruction of conventional treatment plans The markers representing the conventional field borders were used to construct two opposing tangential beams by means of virtual simulation, similar to the conventional procedure by direct simulation as performed in the past at our department. Wedge fractions were defined by evaluating dose distributions limited to a slice situated in the centre of the breast, and slices at 50 mm superior and inferior to this central slice. To enable definition of a conventional boost PTV (PTV CON ), a body-outline contour of the slice containing the boost-volume isocenter was derived from the CT data set. All density information was erased. The body-outline contour only contained the boost-volume isocenter, a two dimensional (2D) reconstruction of all surgical clips, and the marked location of the scar. On the basis of the position of the clips and the available pre-operative information, the assumed lumpectomy cavity was defined within the 2D body-outline contour. Subsequently, the conventional boost CTV (CTV CON ) and the boost PTV CON were created by adding margins of 10 mm and 5 mm, respectively. The resulting boost PTV CON was then transferred into the CT data-set. The field length of the boost beams was prescribed on the basis of the surgical clips, as visualised by means of digitally reconstructed radiographs. The conventional boost plan consisted of three equally weighted photon beams with manually optimised gantry angles. Beam widths and wedge fractions were selected in such a way that the 95%-isodose closely encompassed the boost PTV CON in the boost central slice. Dose distributions in slices other than the boost central slice were not evaluated and no additional shielding was used. For all beams 6-MV photons were used, and an energy fluence based pencil beam algorithm was used for all dose calculations. Eventually, an accumulated dose distribution was calculated, taking into account 50 Gy for the breast plan and an additional 16 Gy for the boost plan

39 Chapter 3 CT guided definition of target volumes and organs at risk The breast CT-based CTV (CTV CT ) included the glandular breast tissue of the ipsilateral breast. In practise, the breast CTV CT was delineated within the extent of the radiopaque wires marking the palpable breast. The breast CTV CT did not extend into the pectoralis major or the ribs and did not include the skin. The breast CTbased PTV (PTV CT ) was generated by adding a 3D-margin of 5 mm around the breast CTV CT. Definition of the lumpectomy cavity was guided by the position of the surgical clips and pre-operative information, but also by hematoma, seroma, and/or other surgery-induced changes, that were considered to be part of the lumpectomy cavity. The boost CTV CT was generated by adding a 3D-margin of 10 mm around the lumpectomy cavity. The boost PTV CT was generated accordingly by adding an additional margin of 5 mm. Both breast and boost PTV CT were restricted to 5 mm within the skin surface. The heart was contoured to the level of the pulmonary trunk superiorly, including the pericardium, excluding the major vessels. Both lungs were contoured as a single organ at risk with the automatic contouring tool of the Helax-TMS planning system, and the right breast was contoured as an organ at risk similar to the left breast CTV CT. 3D-CRT treatment planning Conformal to the breast PTV CT, two opposing tangential beams were constructed. With the use of beam s-eye-view projections, gantry angles were determined to achieve maximum avoidance of the heart, ipsilateral lung and right breast. Shielding was adapted with use of a multileaf collimator (MLC). Wedges and/or a maximum of three MLC segments were added by means of forward planning to obtain a homogeneous dose distribution. Subsequently, a boost plan was created conformal to the boost PTV CT. It consisted of three equally weighted photon beams with gantry angles identical to those that were used with the conventional boost plan. Wedges and MLC shielding were applied in such a way that the 95%-isodose closely encompassed the boost PTV CT in three dimensions, and a uniform dose distribution was obtained. Eventually, an accumulated dose distribution was calculated incorporating both the 3D-CRT breast and boost plans,

40 The shift towards CT guided breast radiotherapy taking into account 50 Gy for the breast plan and an additional 16 Gy for the boost plan. Analyses of target coverage and normal tissue dose Target coverage was determined for both the conventional and 3D-CRT treatment plans by evaluating the relative volumes of the breast PTV CT and the boost PTV CT receiving at least 95% of the prescribed dose (the CT-based PTVs were regarded as the golden standard). For each of the accumulated treatment plans, the total volume and the volume outside the CT-based PTVs receiving at least 95% of the prescribed breast and boost doses were determined. In addition, the relative volumes of the heart receiving 30 Gy (V30), the mean heart dose, the relative total volume of both lungs receiving 20 Gy (V20), the mean lung dose, the relative volume of the right breast receiving 10 Gy (V10) and the right breast mean dose were derived from the dose-volume histograms (DVH). Statistical analysis For comparison of the DVH parameters of the accumulated treatment plans, the mean values were analysed with the Wilcoxon signed ranks test or the pairedsamples t-test on statistical significance whenever appropriate. All tests were twotailed, and differences were considered statistically significant at p Results PTV CT coverage and absolute volumes irradiated With conventional breast beams, coverage of the breast PTV CT was adequate in 72% of the patients (in these patients, 95% of the prescribed breast dose was delivered to 95% of the breast PTV CT ). With 3D-CRT, coverage of the breast PTV CT was adequate for all patients (Table 1). The volume outside the breast PTV CT that received 95% of the prescribed breast dose was significantly smaller when conventional breast beams were used (427 cm 3 vs. 529 cm 3 with 3D-CRT)

41 Chapter 3 Table 1. Target coverage and irradiated volumes Accumulated dose plans CT-based Conventional p-values Target coverage (%) Breast PTV CT 95% 99.2 ( ) 95.3 ( ) < Boost PTV CT 95% 99.6 ( ) 90.1 ( ) < Irradiated volumes (cm 3 ) Volume 95% * 50 Gy 1276 ( ) 1142 ( ) 0.01 Volume 95% * 66 Gy 241 ( ) 187 (84-376) < Excess volumes (cm 3 ) 95% * 50 Gy outside breast PTV CT 529 ( ) 427 ( ) % * 66 Gy outside boost PTV CT 124 (65-260) 82 (31-144) < Abbreviations: PTV = planning target volume; PTVCT = computed tomography (CT)-based PTV. Data presented as mean values, with ranges in parentheses. Table 2. Mean dose and percentage of volume of heart, lungs and right breast irradiated Accumulated dose plans Organs at risk CT-based Conventional p-values Heart Volume 30 Gy (%) 3.6 ( ) 1.4 ( ) < Volume 20 Gy (%) 5.1 ( ) 1.9 ( ) < Volume 10 Gy (%) 9.0 ( ) 4.9 ( ) < Volume 5 Gy (%) 27.4 ( ) 20.5 ( ) < Mean dose (Gy) 5.5 ( ) 4.0 ( ) < Lungs Volume 30 Gy (%) 4.7 ( ) 3.5 ( ) Volume 20 Gy (%) 5.6 ( ) 4.2 ( ) Volume 10 Gy (%) 8.2 ( ) 6.7 ( ) Volume 5 Gy (%) 16.6 ( ) 14.7 ( ) Mean dose (Gy) 4.7 ( ) 4.0 ( ) Right breast Volume 30 Gy (%) 0.1 ( ) 0.0 ( ) ns Volume 20 Gy (%) 0.2 ( ) 0.0 ( ) ns Volume 10 Gy (%) 0.3 ( ) 0.0 ( ) ns Volume 5 Gy (%) 0.8 ( ) 0.2 ( ) ns Mean dose (Gy) 0.9 ( ) 0.9 ( ) ns Data presented as mean values, with ranges in parentheses. With conventional boost beams, coverage of the boost PTV CT was adequate in only 32% of the patients, while coverage was adequate for all patients with 3D-CRT. The volume outside the boost PTV CT that received 95% of the prescribed boost

42 The shift towards CT guided breast radiotherapy dose was significantly less when conventional beams were used (82 cm 3 vs. 124 cm 3 with 3D-CRT). Organs at Risk The mean heart dose and the heart V5-V30 were significantly larger with 3D- CRT (Table 2). Similar results were observed with regard to the mean lung dose and the lung V5-V30. The right breast mean dose and right breast V5-V30 were minimal and similar for the accumulated conventional and 3D-CRT treatment plans. Conventional field borders in relation to PTV CT Conventional breast beams resulted in poor coverage of the medio-dorsal and latero-dorsal areas of the breast PTV CT in the majority of the patients (Fig. 1). Particularly the latero-dorsal areas of the breast PTV CT significantly extended beyond conventional field borders (Table 3). Table 3. PTV volumes and dimensions CT-based target definition Conventional target definition p-values Absolute volumes (cm 3 ) Breast PTV 753 ( ) - - Boost PTV 117 (40-243) - - Breast PTV CT extending beyond conventional field borders (cm) Medial 0.03 ( ) ns Lateral 0.37 ( ) 0.01 Boost PTV CT extending beyond PTV CON (cm) Right 0.18 ( ) 0.03 Left 0.36 ( ) 0.01 Ventral 0.22 ( ) ns Dorsal 0.23 ( ) ns Boost field length 7.9 ( ) 6.7 ( ) < Abbreviations: PTV = planning target volume; PTVCT = computed tomography (CT)-based PTV; PTVCON = conventional PTV. Data presented as mean values, with ranges in parentheses

43 Chapter 3 Figure 1. Conventional breast beams relative to CT-based breast target volumes Representation of computed tomography (CT)-based clinical target volume (CTV) and planning target volume (PTV) and the 95%-isodose (green) resulting from conventional breast beams. Note the areas of PTV (red) not covered by the 95%-isodose when conventional beams are used. Under-dosage of the PTV is caused by including additional tissue (marked yellow-wash areas) into the CTV. The boost PTV CT generally extended beyond the boost PTV CON in the medial and lateral directions (Fig. 2) and Table 3). In the ventral and dorsal directions, the dimensions of PTV CT and PTV CON differed in most cases, without one extending consistently beyond the other. The cranial and caudal borders of the 3D-CRT boost beams extended beyond the conventional boost beams in the majority of patients

44 The shift towards CT guided breast radiotherapy Figure 2. Conventional and CT-based boost planning target volume Transversal (left) and sagittal (right) cross-sections of conventional and computed tomography (CT)- based boost planning target volume (PTV). Discussion On the basis of the current analysis we conclude that CT guided target definition and planning for breast conserving radiotherapy results in improved target coverage at the cost of an increased dose delivered to organs at risk. It seems that when CT densities are used to define the breast CTV, tissue is included that would not have been specifically targeted with conventional breast beams. However, it is uncertain whether or not the additional included tissue is really breast tissue at risk. The dosimetric results with 3D-CRT strongly depend on institutional guidelines used for delineation of the breast target volumes. Various methods have been used in the past to delineate the breast CTV: anatomic references have been used as a guide [10,11], but also radiopaque wires marking the palpable breast [6,12]. In some studies, a conventional beam set-up was used even when CT data were available [13,14]. In these studies, CT data were used for dose calculation and evaluation of the dose to organs at risk, while the breast CTV was not explicitly

45 Chapter 3 defined. In addition, no margins for position uncertainties or penumbra were specified, while in other studies, a 5-7 mm margin for position uncertainties was used together with a margin for penumbra [12,15]. This illustrates that consensus is needed on how the breast target volumes should be defined within the CT images. The delineation method used in the present study, resulted in relatively consistent results because the information used was threefold: 1) palpable breast tissue marked by a radiopaque wire; 2) glandular breast tissue as visible in the CT images; and 3) the use of anatomic references. Therefore, we consider this method to be the current golden standard for CT guided target definition in breast conserving RT. Patient selection was started more than one year after the introduction of CT guided target definition and planning as standard procedure for breast conserving RT at our institution. Therefore, all involved physicians had at least one year of experience, while there were regular interobserver consultations to discuss the delineation of the target volumes. In this way, the effect of a learning curve was eliminated as much as possible. In the present study, the tangential beams of the conventional and 3D-CRT plans were not adjusted when they included more normal tissue than expected. However, in our clinical practise, the gantry angles of the tangential beams are adjusted when the contralateral breast is partially included or when the central lung distance exceeds 30 mm. In some patients, avoidance of the contralateral breast is not possible without a significant increase of the dose delivered to the lungs. In these cases, inadequate coverage of the medial and lateral aspects of the breast PTV is accepted as long as adequate coverage of the boost PTV is maintained. The position of the conventional breast beams was evaluated in relation to the breast PTV CT. Although 3D-CRT field sizes were predominantly larger than conventional field sizes, in some cases the resulting 3D-CRT fields were actually smaller than the conventional fields. As shown in Table 3, the medial aspect of the breast PTV CT was in some cases positioned as far as 1.5 cm within the conventional field borders, while the lateral aspect of the breast PTV CT was in some cases positioned as far as 1.1 cm within the conventional field borders

46 The shift towards CT guided breast radiotherapy In the present study, CT guided target definition and planning resulted in larger boost PTVs that were inadequately covered in 68% of the cases when conventional boost beams were used. Although the volume increase can be partly explained by the additional density information provided by CT, it also appeared that with CT guided planning, the margins for penumbra needed in the cranial and caudal directions could measure up to 10 mm. We conclude that margins for position uncertainties and penumbra were not fully taken into account when the field lengths were prescribed for the conventional boost beams. In most cases, the boost PTV CT extended beyond PTV CON, resulting in larger boost volumes with 3D-CRT. In some patients, however, the CT-based lumpectomy cavity was defined to (marginally) exclude one or more of the surgical clips when these appeared remote from the lumpectomy cavity. In these patients, the PTV CON extended beyond the PTV CT in one or more directions. Equally weighted boost beams were used in the current study. In our clinical practise, however, the boost-beam weights are optimised for each individual patient. For methodological reasons, optimisation of the boost-beam weights was not performed separately for the two different treatments in this study. While photon beams were used for boost irradiation in the present study, others reported on the dosimetric results with an electron boost. It was demonstrated by Benda et al. [16] that target coverage with electron beams, determined without the use of CT data, resulted in very poor target coverage (with on average only 51% of the CT-based boost PTV receiving 90% or more of the prescribed dose). It is likely that such inadequate coverage of the boost volume has also been the case in the boost vs. no boost trial [17,18]. This trial showed that an additional boost dose of 16 Gy, delivered with the use of conventional photon or electron techniques, significantly reduced the risk of a local recurrence. Because it has been demonstrated that in most cases, local recurrences occur close to the primary tumour site [19], it may be possible that CT guided target definition in conjunction with 3D-CRT treatment planning will further reduce the risk of local recurrence as the dose distribution to the lumpectomy cavity is more adequate

47 Chapter 3 CT guided target definition and planning resulted in higher doses delivered to the heart and lungs because larger tangential beams were needed to include the breast PTV CT. The largest increase was observed with the heart V30. Although the absolute increase in normal tissue dose seems to be relatively small, clinical consequences can never be ruled out and attempts should always be made to minimise the dose delivered to organs at risk. A number of studies pointed out that patients who received partial irradiation of the heart had an increased risk of dying from cardiac disease [20-22]. In these studies, conventional radiotherapy techniques were used. The present study demonstrates that the introduction of CT guided target definition and planning may result in an increase of the dose delivered to the heart in some cases. Table 4. Heart dose and target coverage with and without conformal heart shielding Heart Breast PTV CT Mean dose (Gy) Volume 30 Gy (%) Volume receiving 95% of prescribed dose 3D-CRT 3D-CRT heart shielding 3D-CRT 3D-CRT heart shielding 3D-CRT 3D-CRT heart shielding Patient Patient Patient Abbreviations: 3D-CRT = three-dimensional conformal radiotherapy; PTV = planning target volume; PTVCT = computed tomography (CT)-based PTV. Dosimetric results with and without deliberate multileaf collimator shielding of the heart in the tangential breast beams. Results are based on accumulated treatment plans (breast plan 50 Gy + boost plan 16 Gy). Patients had upper-quadrant tumour sites. Boost PTVCT target coverage was not compromised. Other authors already reported on restricting the 3D-CRT field edges in the vicinity of the heart and the application of cardiac shielding to reduce the heart dose [23]. We also tested this method at our institute in three patients that had upperquadrant tumour sites. It appeared that the heart V30 could be reduced to 0% at the cost of reduced coverage of the breast PTV CT (Table 4). Although the use of

48 The shift towards CT guided breast radiotherapy cardiac shielding was not specifically analysed as a part of the current study, it could be regarded as a first and rather safe step towards partial breast irradiation in selected patients who have early-stage disease at locations remote from the heart. In this way, it may be possible to reduce the heart dose with 3D-CRT even below the levels resulting from conventional treatment. A large randomized trial would be necessary to determine tumour control probability and normal tissue complication probability with the different applications of 3D-CRT in breast conserving radiotherapy. Conclusions The shift towards CT guided target definition and planning as the golden standard for breast conserving radiotherapy has resulted in improved target coverage at the cost of larger irradiated volumes and an increased dose delivered to organs at risk. Tissue is now included into the breast and boost target volumes that was never explicitly defined or included with conventional treatment. Therefore, a coherent definition of the breast and boost target volumes is needed, based on clinical data confirming tumour control probability and normal tissue complication probability with the use of 3D-CRT

49 Chapter 3 References 1. Clarke KH. A system of dosage estimation for the tangential irradiation of the breast without bolus. Br J Radiol 1950;23: Veronesi U, Zucali R, Luini A. Local control and survival in early breast cancer: the Milan trial. Int J Radiat Oncol Biol Phys 1986;12: Pierce LJ, Strawderman MH, Douglas KR, et al. Conservative surgery and radiotherapy for early-stage breast cancer using a lung density correction: the University of Michigan experience. Int J Radiat Oncol Biol Phys 1997;39: Zackrisson B, Arevarn M, Karlsson M. Optimized MLC-beam arrangements for tangential breast irradiation. Radiother Oncol 2000;54: Vicini FA, Sharpe M, Kestin L, et al. Optimizing breast cancer treatment efficacy with intensity-modulated radiotherapy. Int J Radiat Oncol Biol Phys 2002;54: Hurkmans CW, Borger JH, Pieters BR, et al. Variability in target volume delineation on CT scans of the breast. Int J Radiat Oncol Biol Phys 2001;50: Struikmans H, Warlam-Rodenhuis C, Stam T, et al. Interobserver variability of clinical target volume delineation of glandular breast tissue and of boost volume in tangential breast irradiation. Radiother Oncol 2005;76: Bentel G, Marks LB, Hardenbergh P, et al. Variability of the location of internal mammary vessels and glandular breast tissue in breast cancer patients undergoing routine CT-based treatment planning. Int J Radiat Oncol Biol Phys 1999;44: Goldberg H, Prosnitz RG, Olson JA, et al. Definition of postlumpectomy tumor bed for radiotherapy boost field planning: CT versus surgical clips. Int J Radiat Oncol Biol Phys 2005;63: Gonzalez VJ, Buchholz DJ, Langen KM, et al. Evaluation of two tomotherapy-based techniques for the delivery of whole-breast intensity-modulated radiation therapy. Int J Radiat Oncol Biol Phys 2006;65: Mayo CS, Urie MM, Fitzgerald TJ. Hybrid IMRT plans--concurrently treating conventional and IMRT beams for improved breast irradiation and reduced planning time. Int J Radiat Oncol Biol Phys 2005;61: van der Laan HP, Dolsma WV, Maduro JH, et al. Three-dimensional conformal simultaneously integrated boost technique for breast-conserving radiotherapy. Int J Radiat Oncol Biol Phys 2007;68:

50 The shift towards CT guided breast radiotherapy 13. van Asselen B, Schwarz M, van Vliet-Vroegindeweij C, et al. Intensity-modulated radiotherapy of breast cancer using direct aperture optimization. Radiother Oncol 2006;79: van Vaerenberg K, De Gersem W, Vakaet L, et al. Automatic generation of a plan optimization volume for tangential field breast cancer radiation therapy. Strahlenther Onkol 2005;181: Cho BC, Hurkmans CW, Damen EM, et al. Intensity modulated versus non-intensity modulated radiotherapy in the treatment of the left breast and upper internal mammary lymph node chain: a comparative planning study. Radiother Oncol 2002;62: Benda RK, Yasuda G, Sethi A, et al. Breast boost: are we missing the target? Cancer 2003; 97: Bartelink H, Horiot JC, Poortmans P, et al. Recurrence rates after treatment of breast cancer with standard radiotherapy with or without additional radiation. N Engl J Med 2001;345: Poortmans P, Bartelink H, Horiot JC, et al. The influence of the boost technique on local control in breast conserving treatment in the EORTC 'boost versus no boost' randomised trial. Radiother Oncol 2004;72: Holland R, Veling SH, Mravunac M, et al. Histologic multifocality of Tis, T1-2 breast carcinomas. Implications for clinical trials of breast-conserving surgery. Cancer 1985;56: Rutqvist LE, Lax I, Fornander T, et al. Cardiovascular mortality in a randomized trial of adjuvant radiation therapy versus surgery alone in primary breast cancer. Int J Radiat Oncol Biol Phys 1992;22: Cuzick J, Stewart H, Rutqvist L, et al. Cause-specific mortality in long-term survivors of breast cancer who participated in trials of radiotherapy. J Clin Oncol 1994;12: Gagliardi G, Lax I, Ottolenghi A, et al. Long-term cardiac mortality after radiotherapy of breast cancer--application of the relative seriality model. Br J Radiol 1996;69: Raj KA, Evans ES, Prosnitz RG, et al. Is there an increased risk of local recurrence under the heart block in patients with left-sided breast cancer? Cancer J 2006;12:

51 - 50 -

52 Chapter 4 Tree-dimensional conformal simultaneously integrated boost technique for breast-conserving radiotherapy Hans Paul van der Laan, Wil V. Dolsma, John H. Maduro, Erik W. Korevaar, Miranda Hollander, Johannes A. Langendijk International Journal of Radiation Oncology Biology Physics 2007; 68:

53 Chapter 4 Abstract Purpose: To compare the target coverage and normal tissue dose with the simultaneously integrated boost (SIB) and the sequential boost technique in breast cancer, and to evaluate the incidence of acute skin toxicity in patients treated with the SIB technique. Materials and Methods: Thirty patients with early-stage left-sided breast cancer underwent breast-conserving radiotherapy using the SIB technique. The breast and boost planning target volumes (PTVs) were treated simultaneously (during each fraction, the breast and boost PTVs received 1.81 Gy and 2.3 Gy, respectively). Three-dimensional conformal beams with wedges were shaped and weighted using forward planning. Dose volume histograms of the PTVs and organs at risk with the SIB technique, 28x ( Gy), were compared with those for the sequential boost technique, 25x2 Gy + 8x2 Gy. Acute skin toxicity was evaluated for 90 patients treated with the SIB technique according to Common Terminology Criteria for Adverse Events, version 3.0. Results: PTV coverage was adequate with both techniques. With SIB, more efficiently shaped boost beams resulted in smaller irradiated volumes. The mean volume receiving 107% of the breast dose was reduced by 20%, the mean volume outside the boost PTV receiving 95% of the boost dose was reduced by 54%, and the mean heart and lung dose were reduced by 10%. Of the evaluated patients, 32.2% had Grade 2 or worse toxicity. Conclusion: The SIB technique is proposed for standard use in breastconserving radiotherapy because of its dose-limiting capabilities, easy implementation, reduced number of treatment fractions, and relatively low incidence of acute skin toxicity

54 3D-CRT-SIB for breast conserving radiotherapy Introduction Breast-conserving radiotherapy (RT) has become the standard treatment for early stage breast cancer since the survival rates proved to be similar to that with radical surgery [1,2]. Moreover, recent data have shown that local control can be improved by an additional boost of 16 Gy to the lumpectomy cavity after administration of 50 Gy to the whole breast [3]. Traditionally, delivery of this boost dose has been performed sequentially (after completion of the whole breast RT) [4,5]. However, with the implementation of intensity-modulated RT (IMRT) in breast cancer, the so called simultaneously integrated boost (SIB) technique has also been introduced for breast-conserving RT [6]. With this method, the initial planning target volume (PTV) encompassing the whole breast and the boost PTV are integrated in a single treatment plan. With the SIB technique, patients are treated with the same treatment plan for each fraction throughout the treatment course. In general, a greater dose per fraction is delivered to the boost PTV and the number of treatment fractions is reduced. To date, the SIB technique has only been presented in combination with IMRT. It seems that with IMRT-SIB, a high level of dose-to-target conformity can be obtained [7]. However, in the studies that have reported on IMRT-SIB to date, the attention has been focused on comparing IMRT-SIB with conventional RT with a sequential boost [8]. Therefore, it remains unclear whether the improved conformity results from the IMRT itself or the SIB. Because the use of a SIB technique has practical advantages (a reduced number of fractions) and, theoretically, a radiobiological advantage with respect to tumour control (greater dose per fraction and a reduction of the overall treatment time), we decided to perform a planning study to compare conventional three-dimensional (3D)- conformal RT (3D-CRT) with a sequential boost technique (SBT) vs. 3D-CRT using the SIB technique. We compared the dose distribution in both target volumes and relevant organs at risk. In addition, acute toxicity was evaluated in the first 90 consecutive patients treated with the SIB technique at our institution

55 Chapter 4 Materials and Methods Patients and computed tomography Thirty consecutive patients with left-sided breast cancer (Stage T1-T2 N0-N1a M0) scheduled to undergo RT after breast-conserving surgery were included in this comparative planning study. A planning computed tomography (CT) scan was made for each patient. The patients were positioned on a breast board with both arms abducted alongside the head. Before the CT scan, skin marks were placed to locate the boost volume isocenter and to enable patient repositioning during treatment. Radiopaque catheters and markers were placed to locate both palpable breasts, scars, and skin marks on the CT images. Patients were scanned from the level of the larynx to the level of the upper abdomen, including both lungs, with a scan thickness and index of 5 mm. The CT data for all patients were transferred to the Helax-TMS 3D treatment planning system, version 6.1B (Nucletron, Veenendaal, The Netherlands). All patients provided informed consent before starting therapy, and the ethics committee at the University Medical Center Groningen approved the procedures followed. Definition of target volumes and organs at risk The breast clinical target volume (CTV) included the glandular breast tissue of the ipsilateral breast. In practise, the breast CTV was delineated within the extent of the radiopaque catheters marking the palpable breast. The breast CTV did not extend into the pectoralis major or the ribs and did not include the skin. The breast PTV was generated by adding a 3D margin of 5 mm around the breast CTV. A margin of 10 mm was used in the cranial and caudal directions. The definition of the lumpectomy cavity was guided by the presence of the surgical clips, as well as by hematoma, seroma, and/or other surgery induced changes considered to be a part of the lumpectomy cavity. The boost CTV was generated by adding a 3D margin of 10 mm around the lumpectomy cavity. The boost PTV was generated accordingly by adding an additional margin of 5 mm. The breast and boost PTVs were restricted to 5 mm within the skin surface. The heart was contoured to the

56 3D-CRT-SIB for breast conserving radiotherapy level of the pulmonary trunk superiorly, including the pericardium, and excluding the major vessels. Both lungs were contoured as a single organ at risk using the automatic contouring tool of the Helax-TMS planning system, edited when needed, and then visually verified. The right breast was contoured as an organ at risk similar to the left breast CTV. Sequential boost treatment planning Conformal to the breast PTV, two opposing tangential beams were constructed. With the use of beam s-eye-view projections, the gantry angles were determined to achieve maximal avoidance of the heart, ipsilateral lung, and right breast. Shielding was adapted with use of a multileaf collimator (MLC). Wedges and/or a maximum of three additional MLC segments were used to obtain a homogeneous dose distribution. The wedge fractions and relative weights of any additional segments were weighted manually, i.e., by forward planning. Subsequently, a boost plan was created conformal to the boost PTV. It consisted of three equally weighted photon beams with manually selected gantry angles. Wedges and MLC shielding were selected in such a way that the 95%- isodose encompassed the boost PTV in three dimensions, and a uniform dose distribution was obtained according to the recommendations of the International Commission on Radiation Units and Measurements [9]. The isocenter and dose-normalization point of the breast plan were identical to that of the boost plan. Both were placed centrally in a slice representative of the boost PTV. Eventually, a cumulative dose plan was calculated incorporating both the breast plan and the boost plan, taking into account 25 fractions of 2 Gy for the breast plan and an additional eight fractions of 2 Gy for the boost plan, to a cumulative dose of 66 Gy

57 Chapter 4 SIB fractionation schedule With the SIB technique, the breast and boost beams are combined into an integrated treatment plan, i.e., patients are treated with the same plan for each fraction throughout the entire treatment. Therefore, an alternative fractionation schedule is necessary. Using the linear-quadratic cell survival model, we calculated fraction sizes and total doses for the breast and boost PTVs that were biologically equivalent to the total dose delivered to the PTVs in 2-Gy fractions with the SBT [10]. For this purpose, an α/β ratio of 10 Gy for tumour response and an α/β ratio of 3 Gy for late-responding normal tissues were used. It appeared that the breast and boost fraction sizes depended on the selected number of fractions (Table 1). Although the fractionation schedules were isoeffective for tumour response, a greater biologically effective total dose was delivered to the glandular breast tissue in the boost area when the number of fractions decreased. At our institute, a schedule of 28 fractions was applied with a daily dose of 1.81 Gy delivered to the breast PTV and 2.3 Gy to the boost PTV. In this way, the total treatment time was reduced by 1 week. Table 1. SIB fractionation schedules with tumour control probability equivalent to conventional fractionation Dose per fraction (Gy) Total dose (Gy) Equivalent total dose normal tissue (Gy) Fractions Breast PTV Boost PTV Breast PTV Boost PTV Breast PTV Boost PTV Abbreviations: SIB = simultaneously integrated boost; PTV = planning target volume; conventional fractionation = breast PTV 25 x 2 Gy followed by boost PTV 8 x 2 Gy. SIB fractionation schedules were calculated using α/β = 10 Gy for tumour response and α/β = 3 Gy for late-responding normal tissue

58 3D-CRT-SIB for breast conserving radiotherapy SIB treatment planning The SIB treatment plan was created by copying the sequentially planned breast and boost beams into an integrated treatment plan. The same isocenter and dose-normalization point were used. The breast beams were set to contribute a daily dose of 1.81 Gy to the dose-normalization point, and the boost beams were set to contribute a daily dose of 0.49 Gy. Subsequently, the wedge fractions and MLC settings of the boost beams were adapted in such a way that the 95% isodose closely encompassed the boost PTV in three dimensions, and a uniform dose distribution was obtained according to the recommendations of the International Commission on Radiation Units and Measurements. All other planning parameters were left unchanged compared with the SBT. Eventually, a cumulative dose plan was calculated, taking into account 28 fractions of 2.3 Gy for the SIB plan. Analyses of target coverage and normal tissue dose Target coverage was determined for the cumulative SBT and SIB plans by evaluating the relative volumes of the breast PTV and boost PTV receiving at least 95% of the prescribed dose. In addition, the total volume and undesired excess volume receiving at least 95% of the prescribed breast and boost doses were determined for each of the cumulative dose plans. For both cumulative dose plans, the relative volume of the heart receiving 30 Gy, mean heart dose, relative total volume of both lungs receiving 20 Gy, mean lung dose, relative volume of the right breast receiving 10 Gy, and right breast mean dose were obtained from the dose volume histograms. Acute toxicity evaluation To determine the severity and incidence of acute toxicity with the SIB technique, 90 patients were evaluated by their physician during the last weeks of RT. The maximal skin toxicity was scored using the Common Terminology Criteria for Adverse Events, version 3.0, using dermatitis associated with radiation as an adverse event [11]. This resulted in a toxicity grade for each patient

59 Chapter 4 Statistical analysis For comparison of the dose volume histogram parameters of the cumulative dose plans, the mean values were analyzed with the Wilcoxon signed ranks test or the paired-samples t-test on statistical significance, whenever appropriate. All tests were two-tailed, and differences were considered statistically significant at p Table 2. Target coverage and irradiated volumes accumulated dose plans Accumulated dose plans SBT a SIB b p-values Target coverage Breast PTV 95% (%) 99.2 ( ) 99.0 ( ) Boost PTV 95% (%) 99.7 ( ) 99.1 ( ) < Volumes irradiated Volume 95% breast dose (cm 3 ) 1289 ( ) 1259 ( ) < Volume 107% breast dose (cm 3 ) 643 ( ) 515 ( ) < Volume 95% boost dose (cm 3 ) 249 ( ) 177 (53-385) < Excess volumes irradiated Volume outside breast PTV 95% breast dose (cm 3 ) 535 ( ) 506 ( ) < Volume outside boost PTV 95% boost dose (cm 3 ) 129 (65-260) 59 (14-143) < Abbreviations: PTV = planning target volume; SBT = sequential boost technique; SIB = simultaneously integrated boost. Data presented as mean values, with ranges in parentheses. a Cumulative dose plan SBT: 25 x 2-Gy breast plan + 8 x 2-Gy boost plan. b Cumulative dose plan SIB: 28 x 1.81-Gy breast beams + 28 x 0.49-Gy boost beams. Results PTV coverage and absolute volumes irradiated At least 95% of the prescribed dose was delivered to 99% of the breast and boost PTVs for the SBT and SIB plans (Table 2). However, the absolute irradiated volumes were always smaller with the SIB plans. Although only a minimal difference was observed with regard to the mean volume receiving 95% of the

60 3D-CRT-SIB for breast conserving radiotherapy prescribed breast dose, the mean volume receiving 107% of the prescribed breast dose was 20% smaller with the SIB technique. Figure 1. Reconstructed radiographs from boost beam s-eye-view With the simultaneously integrated boost technique, multileaf collimator shielding (short dotted lines) can be applied without use of margins around boost planning target volume (white solid line), resulting in substantial reduction of excess volumes irradiated. When the SBT beams were constructed, MLC shielding was applied with a margin of 5 10 mm outside the boost PTV to obtain adequate coverage (Fig. 1). Consequently, after adding the breast and boost dose distributions of the SBT plans, the mean volume outside the boost PTV that received 95% of the prescribed boost dose increased to 129 cm 3 (Table 2). When the SIB plans were constructed, MLC shielding could be applied without any margin outside the boost PTV, significantly lowering the mean volume outside the boost PTV to 59 cm 3 (Fig. 2)

61 Chapter 4 Figure 2. Axial representation of lumpectomy cavity (LC) and boost planning target volume (PTV) The black isodose represents the 50% breast dose, and the white-wash area represents the volume outside the boost PTV that received 95% of the prescribed boost dose with sequential boost and simultaneously integrated boost techniques. Normal tissue Because irradiated volumes were always smaller with the SIB plans, the dose delivered to the heart and lungs was significantly reduced compared with the dose in the SBT plans (Table 3). Although the absolute differences in the relative volume of the heart receiving 30 Gy and relative total volume of both lungs receiving 20 Gy were minimal, the mean heart dose and mean lung dose were both reduced by approximately 10%. The greater dose-per-fraction delivered to the boost area did not result in the heart and lungs being treated at a higher dose per fraction compared with a breast beams-only plan delivered with 2-Gy fractions (Fig. 3), with the exception of a slightly larger low-dose volume resulting from the nontangential boost beams. The SIB fractionation schedule resulted in a reduced dose per fraction in approximately 60% of the breast PTV. No differences were found with regard to the dose delivered to the contralateral breast

62 3D-CRT-SIB for breast conserving radiotherapy Table 3. Mean dose and percentage of volume of heart, lungs, and right breast irradiated Accumulated dose plans Organs at risk SBT a SIB b p-values Heart V30 (%) 3.5 ( ) 3.4 ( ) 0.02 Mean dose (Gy) 5.3 ( ) 4.8 ( ) < Lungs V20 (%) 5.6 ( ) 5.5 ( ) < Mean dose (Gy) 4.7 ( ) 4.3 ( ) < Right breast V10 (%) 0.3 ( ) 0.3 ( ) ns Mean dose (Gy) 0.9 ( ) 0.9 ( ) ns Abbreviations: V30 = volume of heart receiving 30 Gy; V20 = volume of lungs receiving 20 Gy; V10 = volume of right breast receiving 10 Gy; SBT = sequential boost technique; SIB = simultaneously integrated boost. Data presented as mean values, with ranges in parentheses. a Cumulative dose plan SBT: 25 x 2 Gy breast plan + 8 x 2 Gy boost plan. b Cumulative dose plan SIB: 28 x 1.81 Gy breast beams + 28 x 0.49 Gy boost beams. Acute toxicity Grade 0, 1, 2, and 3 acute toxicity was assigned to 7.8%, 60.0%, 31.1%, and 1.1% of the 90 evaluated patients, respectively. Discussion With the SIB technique, a higher dose per fraction can be delivered to the boost area in combination with a reduced number of treatment fractions. Because MLC shielding can be applied without any margins around the boost PTV, the volumes receiving undesired excess doses are substantially reduced. The heart and lungs receive a lower total dose without receiving a higher dose per fraction. A lower dose per fraction was delivered to large proportions of the ipsilateral breast, which could explain the relatively low incidence of acute toxicity we observed in the patients treated with the SIB technique. Furthermore, the SIB technique can be easily implemented to replace the SBT without increasing the complexity and time involved with treatment planning and delivery

63 Chapter 4 Figure 3. Cumulative average dose volume histograms of heart, lungs, and breast planning target volume (PTV) (n = 30) Histograms on the left represent dose per volume resulting from single fraction with simultaneously integrated boost (SIB) plan (breast beams 1.81 Gy, boost beams 0.49 Gy) and breast beams-only plan (2 Gy). Histograms on the right represent dose per volume resulting from complete treatment with simultaneously integrated boost plan (breast beams Gy, boost beams Gy) and sequential boost plan (breast plan 50 Gy, boost plan 16 Gy)

64 3D-CRT-SIB for breast conserving radiotherapy Others have proposed the use of SIB in combination with inversely planned IMRT [6-8]. Although the dose-to-target conformity might be increased further using IMRT techniques, it seems that, compared with our results, similar reductions were obtained for the volumes outside the boost PTV receiving 95% of the prescribed dose (129 cm 3 reduced to 55 cm 3 vs. 129 cm 3 reduced to 59 cm 3 in the present study) [7]. Furthermore, IMRT is used as the standard treatment for breast cancer patients in a relatively small proportion of institutes worldwide. In contrast, the proposed 3D-CRT-SIB technique is a straightforward technique that can be easily implemented to replace 3D-CRT-SBT. The 3D-CRT-SIB technique seems to have results equal to that of an IMRT-SIB technique but does not involve the complexity and time required to implement, plan and deliver IMRT-SIB. The SIB technique has the advantage of providing treatment flexibility, allowing for individual fractionation schedules that are isoeffective for tumour control but have different dose-per-fraction sizes (Table 1). For example, using the 33-fraction schedule, the boost dose is delivered in 2-Gy fractions, similar to conventional fractionation, but the breast dose is delivered with a lower dose per fraction, which could further reduce the normal tissue complication probability. Even when schedules with more fractions are used, the SIB technique has the advantage of reducing the excess dose volumes because the boost beams can be constructed more efficiently. We evaluated acute toxicity during the last weeks of RT in the 90 patients treated with the SIB technique. We found that 32.2% of the patients had Grade 2 or worse toxicity. Other studies that prospectively evaluated acute side effects with the use of modern RT techniques reported a greater incidence of Grade 2 or worse acute toxicity. Freedman et al. [12] scored acute toxicity in 73 patients who had received whole breast IMRT to a total dose of Gy in 2-Gy fractions, followed by a Gy electron boost in 2-Gy fractions. They found that 70% of the patients had Grade 2 or worse toxicity according to the Common Terminology Criteria for Adverse Events, version 3.0. Vicini et al. [13] scored acute toxicity in 262 patients who underwent whole breast IMRT to a total dose of 45 Gy in 1.8-Gy fractions, followed by a 16-Gy electron boost in 2-Gy fractions. They found that

65 Chapter 4 44% of the patients had Grade 2 or worse toxicity according to Radiation Therapy Oncology Group criteria, which resemble the Common Terminology Criteria for Adverse Events, version 3.0 for assigning Grade 2 or worse acute skin toxicity [14]. Although the difference in outcomes could also be explained by factors that are not related to treatment technique or a given dose, we believe our toxicity evaluation demonstrates that the SIB technique results in a relatively low incidence of acute skin toxicity. We are currently investigating the incidence of late toxicity in patients treated with the SIB technique, because the higher dose per fraction delivered to the boost area could increase the risk of late fibrosis. Although the impact of fraction size on late fibrosis was mainly found at large fraction sizes ( 3 Gy), delivered to the whole breast [15], others have found that adding a boost or changing the boost delivery technique did not have an impact on treatment induced fibrosis [3,16]. Because an increased risk of late fibrosis may still exist, some will hesitate to use the SIB technique; therefore, this issue has to be clarified. The use of the initial planning CT scan for the treatment of the boost throughout the entire treatment course has a disadvantage. It has been shown that the lumpectomy cavity shrinks with the time elapsed since surgery [17]. Therefore, the position and shape of the breast and boost PTVs should be monitored and verified during the RT session. In the case of changes in position, shape, or volume that could affect the anticipated dose distributions, a new planning CT scan should be made to adapt the treatment plan to the new situation. In this report, we assumed an α/β ratio of 10 Gy for tumour response. However, recent reports have suggested that the sensitivity of breast cancer to a fraction size, i.e., the α/β value, may be as low as 4 Gy, which is almost similar to the fractionation sensitivity of the dose-limiting response of healthy tissue [18]. If this is the case, the use of an α/β ratio of 10 Gy would result in an underestimation of the tumour response. Therefore, the use of the SIB technique in combination with the proposed shortened fractionation schedule might even increase the tumour control probability

66 3D-CRT-SIB for breast conserving radiotherapy Conclusion The SIB technique is proposed for standard use in breast-conserving RT, because it can be easily implemented to reduce excess volumes of normal tissue irradiated, shorten the treatment course, decrease the dose per fraction for the breast, and increase the dose per fraction for the boost, with a relatively low incidence of acute skin toxicity

67 Chapter 4 References 1. Early Breast Cancer Trialists Collaborative Group. Effects of radiotherapy and surgery in early breast cancer: An overview of the randomized trials. N Engl J Med 1995;333: Fisher B, Anderson S, Bryant J, et al. Twenty-year follow-up of a randomized trial comparing total mastectomy, lumpectomy, and lumpectomy plus irradiation for the treatment of invasive breast cancer. N Engl J Med 2002;347: Bartelink H, Horiot JC, Poortmans P, et al. Recurrence rates after treatment of breast cancer with standard radiotherapy with or without additional radiation. N Engl J Med 2001;345: Veronesi U, Luini A, Del Vecchio M, et al. Radiotherapy after breast-preserving surgery in women with localized cancer of the breast. N Engl J Med 1993;328: Omlin A, Amichetti M, Azria D, et al. Boost radiotherapy in young women with ductal carcinoma in situ: A multicentre, retrospective study of the Rare Cancer Network. Lancet Oncol 2006;7: Guerrero M, Li XA, Earl MA, et al. Simultaneous integrated boost for breast cancer using IMRT: A radiobiological and treatment planning study. Int J Radiat Oncol Biol Phys 2004;59: Hurkmans CW, Meijer GJ, van Vliet-Vroegindeweij C, et al. High-dose simultaneously integrated breast boost using intensity-modulated radiotherapy and inverse optimization. Int J Radiat Oncol Biol Phys 2006;66: Singla R, King S, Albuquerque K, et al. Simultaneous-integrated boost intensitymodulated radiation therapy (SIBIMRT) in the treatment of early-stage left-sided breast carcinoma. Med Dosim 2006;31: International Commission on Radiation Units and Measurements (ICRU). ICRU report 50: Prescribing, recording, and reporting photon beam therapy. Bethesda: ICRU Publications; Joiner MC, van der Kogel AJ. The linear-quadratic approach to fractionation and calculation of isoeffect relationships. In: Steel GG, editor. Basic clinical radiobiology. 2 nd ed. London: Arnold;

68 3D-CRT-SIB for breast conserving radiotherapy 11. National Cancer Institute Cancer Therapy Evaluation Program. Common Terminology Criteria for Adverse Events, version 3.0. Available at: Accessed September 23, Freedman GM, Anderson PR, Li J, et al. Intensity modulated radiation therapy (IMRT) decreases acute skin toxicity for women receiving radiation for breast cancer. Am J Clin Oncol 2006;29: Vicini FA, Sharpe M, Kestin L, et al. Optimizing breast cancer treatment efficacy with intensity-modulated radiotherapy. Int J Radiat Oncol Biol Phys 2002;54: Cox JD, Stetz J, Pajak TF. Toxicity criteria of the Radiation Therapy Oncology Group (RTOG) and the European Organization for Research and Treatment of Cancer (EORTC). Int J Radiat Oncol Biol Phys 1995;31: Yarnold J, Ashton A, Bliss J, et al. Fractionation sensitivity and dose response of late adverse effects in the breast after radiotherapy for early breast cancer: Long-term results of a randomised trial. Radiother Oncol 2005;75: Poortmans P, Bartelink H, Horiot JC, et al. The influence of the boost technique on local control in breast conserving treatment in the EORTC boost versus no boost randomised trial. Radiother Oncol 2004;72: Jacobson G, Betts V, Smith B. Change in volume of lumpectomy cavity during external-beam irradiation of the intact breast. Int J Radiat Oncol Biol Phys 2006;65: Owen JR, Ashton A, Bliss JM, et al. Effect of radiotherapy fraction size on tumour control in patients with early-stage breast cancer after local tumour excision: Longterm results of a randomised trial. Lancet Oncol 2006;7:

69 - 68 -

70 Chapter 5 Limited benefit of inversely optimised intensity modulation in breast conserving radiotherapy with simultaneously integrated boost Hans Paul van der Laan, Wil V. Dolsma, Cornelis Schilstra, Erik W. Korevaar, Geertruida H. de Bock, John H. Maduro, Johannes A. Langendijk Submitted to Radiotherapy & Oncology

71 Chapter 5 Abstract Purpose: To examine whether in breast conserving radiotherapy (RT) with simultaneously integrated boost (SIB), application of inversely planned intensity modulated radiotherapy (IMRT-SIB) instead of three-dimensional RT (3D-CRT- SIB) has benefits that justify the additional costs, and to evaluate whether a potential benefit of IMRT-SIB depends on specific patient characteristics. Materials and Methods: 3D-CRT-SIB and various IMRT-SIB treatment plans were constructed and optimised for 30 patients with early stage left-sided breast cancer. Coverage of planning target volumes (PTV) and dose delivered to organs at risk (OAR) were determined for each plan. Overlap between heart and breast PTV (OHB), size of breast and boost PTVs and boost location were examined in their ability to identify patients that might benefit from IMRT-SIB. Results: All plans had adequate PTV coverage. IMRT-SIB generally reduced dose levels delivered to heart, lungs, and normal breast tissue relative to 3D-CRT- SIB. However, IMRT-SIB benefit differed per patient. For many patients, comparable results were obtained with 3D-CRT-SIB, while patients with OHB >1.4 cm and a relatively large boost PTV volume (>125 cm 3 ) gained most from the use of IMRT-SIB. Conclusion: In breast conserving RT, comparable results are obtained with 3D-CRT-SIB and IMRT-SIB. Patient characteristics could be used to identify patients that are most likely to benefit from IMRT-SIB

72 Strategies for simultaneously integrated breast boost Introduction Breast-conserving therapy with the adjuvant use of radiotherapy (RT) has been acknowledged as standard treatment for early stage breast cancer since survival rates proved to be similar to those obtained with radical surgery [1,2]. Moreover, recent data confirmed that local control can be improved by an additional boost of 16 Gy to the lumpectomy cavity after administration of 50 Gy to the whole breast [3]. Traditionally, delivery of this boost dose has been performed sequentially, i.e., after completion of whole-breast RT [4]. However, with the clinical introduction of intensity-modulated RT (IMRT) in breast cancer, the socalled simultaneously integrated boost (SIB) has also been introduced for breastconserving RT [5]. With the SIB method, the initial planning target volume (PTV) includes breast PTV and boost PTV, integrated in a single treatment plan that is applied during each fraction throughout the course of treatment. In general, a higher dose per fraction is delivered to the boost PTV allowing for a reduction of the number of treatment fractions. Recently, we reported on the use of a more advanced three-dimensional conformal technique using SIB (3D-CRT-SIB) [6]. This 3D-CRT-SIB technique allowed for a significantly improved dose-to-boost-target conformity as compared to the classical sequential 3D-CRT technique. Direct comparisons between whole breast IMRT and classical 3D-CRT techniques showed favourable results for IMRT regarding the dose delivered to normal tissues [7,8]. As the results obtained with our 3D-CRT-SIB technique with regard to the amount of normal breast tissue that received excess boost dose outside the boost PTV appeared to be similar to those described with IMRT-SIB [9], the question arises as to whether the dose distributions obtained with 3D-CRT-SIB are comparable to those obtained with IMRT-SIB. While breast IMRT may generally reduce dose to organs at risk (OAR) below the levels obtained with conventional 3D-CRT, the actual reductions that can be obtained with manually optimised SIB are not yet clear. In addition, when a potential benefit of IMRT-SIB exists, it may be observed in specific subgroups of

73 Chapter 5 patients only. This could be of particular interest for institutes that do not (yet) have the capability to prescribe IMRT to all of their patients, i.e., implementation of 3D-CRT-SIB instead of the conventional sequential 3D-CRT boost technique could be a very cost effective solution in countries where treatment with IMRT requires substantial investments and may increase the costs of radiotherapy [10,11]. Therefore, a planning comparative study was performed to examine different strategies for breast IMRT-SIB and to determine whether the dose delivered to OAR can be reduced compared to 3D-CRT-SIB. In addition, we examined specific patient characteristics in relation to potential reductions of dose delivered to OARs with IMRT-SIB. Materials and Methods Patients Thirty patients with early stage left-sided breast cancer who had previously undergone RT after breast-conserving surgery were selected for this study. Patients were selected in such a way that an equal distribution of breast shapes, breast sizes, boost locations and cardiac anatomy was obtained. Of each patient, a planning computed tomography (CT) scan was available, acquired in treatment position with a slice thickness and index of 3 mm. All patients had provided informed consent before starting RT, and the ethics committee at the University Medical Center Groningen approved the procedures followed. Regions of interest and 3D-CRT-SIB treatment planning Definition of regions of interest and the 3D-CRT-SIB technique have been described previously in more detail [6]. Briefly, breast and boost clinical target volumes (CTVs) were delineated and expanded with a margin of 5 mm to generate the breast and boost PTVs. The breast and boost PTVs were restricted to 6 mm within the skin surface. Radiotherapy treatment planning was performed with the Pinnacle 3 treatment planning system (TPS), using 6 MV photons, and the adaptive convolve algorithm for all dose calculations. Dose distributions were calculated

74 Strategies for simultaneously integrated breast boost taking into account a SIB schedule of 28 fractions, with a daily dose of 1.81 Gy delivered to the breast PTV and 2.30 Gy delivered to the boost PTV. The 3D-CRT-SIB treatment plans were constructed by using tangential breast beams with multileaf collimator (MLC) shielding conformal to the breast PTV. Gantry angles for the breast beams were chosen such that maximum avoidance of the heart, ipsilateral lung, and contralateral breast was achieved. Three boost beams were added with MLC shielding conformal to the boost PTV. Boost beam gantry angles were selected on the basis of patient anatomy and boost location. In general, one or more boost beams had non-tangential beam directions, avoiding the heart as much as possible and excluding the contralateral breast at all times. Manually shaped MLC segments (up to a maximum number of 3 for each plan) were added to enable dose plan optimisation. All beams had the same isocenter and the same dose-normalization point, that were both placed centrally in the boost PTV. Manual optimisation of the SIB dose plan was performed by adjusting beam weights, wedge fractions and MLC settings for all beams in such a way that the 95%-isodose closely encompassed the PTVs in three dimensions, and volumes receiving 107% of the dose prescribed to the PTVs were minimised (Table 1). 100%IMRT-SIB and 25%IMRT-SIB treatment planning Four direct aperture optimisation (DAO)-based IMRT treatment plans were constructed for each patient: two fully segmented plans (100%IMRT-SIB) and two plans combining conformal beams with skin flash and IMRT segments (25%IMRT- SIB). In the 25%IMRT-SIB plans, 65% of the prescribed total dose was delivered with beams conformal to the breast PTV, 10% with beams conformal to the boost PTV and 25% with IMRT segments. In each patient, beam directions used in the IMRT-SIB plans were always similar to those used in the 3D-CRT-SIB plan. Conformal beams included a skin flash of 3 cm whenever appropriate. DAOsettings were the same for each IMRT-SIB plan: a minimal segment size of 4 cm 2, at least 4 monitor units per segment, and a maximum of 10 segments for each beam direction. The IMRT-SIB optimisation process was performed on the basis of a series of structures that were created in addition to the OARs and PTVs (Fig. 1)

75 Chapter 5 Objectives and objective values were entered for each structure and adjusted during treatment plan optimisation to comply best with the criteria for dose plan acceptance (Table 1). Table 1. Criteria of acceptance for 3D-CRT-SIB and IMRT-SIB dose plans Objects 3D-CRT-SIB Class solution IMRT-SIB Optimised solution IMRT-SIB Planning target volumes 98% of volume 95% of dose 98% of volume 95% of dose 98% of volume 95% of dose Volume 108% boost dose 2 cm 3 2 cm 3 2 cm 3 Volume 95% breast dose minimized a 3D-CRT-SIB 3D-CRT-SIB Volume 107% breast dose minimized a 3D-CRT-SIB 3D-CRT-SIB / minimized priority 3 Volume 95% boost dose minimized a 3D-CRT-SIB 3D-CRT-SIB Heart V30 / mean heart dose minimized a 3D-CRT-SIB 3D-CRT-SIB / minimized priority 1 Lungs V20 / mean lung dose minimized a 3D-CRT-SIB 3D-CRT-SIB / minimized priority 2 Contralateral breast avoided b avoided b avoided b Abbreviations: 3D-CRT-SIB = three-dimensional conformal radiotherapy including simultaneously integrated boost; IMRT-SIB = intensity modulated radiotherapy including simultaneously integrated boost; Vx = proportion of organ at risk receiving x Gy. a Optimisation was performed manually by using a maximum total of 3 forward planned sub-beams and effective use of multileaf collimator shielding. b Beam directions were chosen on the basis of patient anatomy to exclude the right breast, on condition that adequate coverage of the breast planning target volume was maintained. priority 1-3 Optimisation was performed without harming other objectives and objectives with higher priority. IMRT-SIB objectives For each patient, two optimisation methods were used for both the 25%IMRT- SIB and the 100%IMRT-SIB plans: 1) an optimised solution (OS), and; 2) a class solution (CS) optimisation method. First, OS plans were created for all patients. The aim was to create plans with (in order of priority) minimised heart dose, minimised lung dose and a minimised volume receiving 107% of the prescribed breast dose, without allowing deterioration of the dose distribution as obtained with 3D-CRT-SIB (Table 1). Objective values and weights were entered for the various structures and adjusted each sequence while optimising OS plans for all

76 Strategies for simultaneously integrated breast boost patients until a result was reached that optimally complied with the OS plan objectives. After completing all OS plans for all patients, the final objective values and weights entered for all OS plans were evaluated and compiled into a standard set of inverse planning parameters (class solution) meant to serve as a starting point for the CS optimisation process. The aim of the CS optimisation process was to produce plans that complied at least with the criteria for 3D-CRT-SIB while requiring a minimum number of optimisation sequences, so they could be planned in a time span similar to that needed to plan 3D-CRT-SIB (Table 1). The CS optimisation procedure ended as soon as an acceptable, not necessarily optimal, solution was reached. Figure 1. Structures created in addition to organs at risk and planning target volumes Additional structures were created to enable direct aperture optimisation-based intensity modulated radiotherapy including simultaneously integrated boost: structure A) extending 4 mm inside boost planning target volume (PTV); structure B) surrounding the boost PTV with a margin of 10 mm; structure C) including the breast PTV, yet excluding the boost PTV with surrounding structure; structure D) extending 4 mm inside structure C, and; structure E) extending from the breast PTV in the mediodorsal direction with a margin of 3 cm

77 Chapter 5 Target coverage, irradiated volumes and OAR dose Target coverage was determined for all plans by evaluating proportions of the breast and boost PTVs receiving 95% of the prescribed dose. In addition, irradiated volumes were determined receiving 95% and 107% of the prescribed breast dose and 95% of the prescribed boost dose. For each plan, the proportion of heart receiving 30 Gy (V30), mean heart dose, proportion of both lungs combined receiving 20 Gy (V20), mean dose of both lungs combined, proportion of contralateral breast receiving 1 Gy (V1), 5 Gy (V5) and contralateral breast mean dose were obtained from the dose volume histograms (DVH). In order to get insight into what extend differences in lung dose distributions translate into different NTCP-values, NTCP calculations were performed using parameter values derived by Seppenwoolde et al. [12], and Semenenko et al. [13]. Predictive patient characteristics Patient characteristics were collected that might relate to a potential benefit of IMRT-SIB: the overlap between heart and breast PTV (OHB), measured as the maximum distance of the heart contour projected within the breast PTV on a beam s-eye-view of the tangential beams (Fig. 2), the absolute volume of the breast and boost PTVs, and the location of the boost volume. The boost volume location was sub-divided in the cranio-caudal direction (cranial, central and caudal), and in the medio-lateral direction (medial, central and lateral). Statistical analyses Correlation statistics were performed for the aforementioned patient characteristics in relation to relevant effect variables, i.e., reductions with IMRT- SIB when compared with 3D-CRT-SIB with respect to the total volume receiving 107% of the prescribed breast dose (V54.23), heart V30, mean heart dose, lung V20 and mean lung dose. Receiver operating characteristics (ROC) were used to evaluate the predictive value of the individual patient characteristics and to determine optimum threshold values (cut-off values). On the basis of these threshold values and for each of the patient characteristics, patients were divided

78 Strategies for simultaneously integrated breast boost into two groups, patients with a lower value and patients with a higher value. Mean OAR dose reductions with IMRT-SIB were determined for each group and twosided non-parametric rank tests were used to determine if values in patient groups were statistically different (p 0.05). Figure 2. Overlap between heart and breast planning target volume The overlap between heart and breast planning target volume (PTV) was measured as the maximum distance of the heart contour projected within the breast PTV on a tangential beam s-eye-view. Results Treatment planning Treatment planning time was similar for CS IMRT-SIB plans that were generally completed after 1 3 optimisation sequences, and 3D-CRT-SIB. Optimising OS IMRT-SIB plans could take an additional 4 hours and included multiple optimisation sequences

79 Chapter 5 PTV coverage and irradiated volumes At least 98% of the volume of the breast and boost PTVs received at least 95% of the prescribed dose in all plans and all patients. However, the absolute irradiated volumes receiving 95% and 107% of the prescribed breast dose were smaller with the various IMRT-SIB strategies when compared to 3D-CRT-SIB (Table 2). IMRT-SIB resulted in volumes receiving excess boost dose outside the boost PTV ( 95% boost dose) that were comparable to those with 3D-CRT-SIB (Table 2). Table 2. Irradiated volumes and dose delivered to organs at risk with 3D-CRT-SIB and IMRT-SIB 3D-CRT-SIB Class solution IMRT-SIB Optimised solution IMRT-SIB 25%IMRT 100%IMRT 25%IMRT 100%IMRT Irradiated volumes 95% breast dose (cm 3 ) 1076 ( ) 1052 ( ) 988 ( ) 1037 ( ) 989 ( ) 107% breast dose (cm 3 ) 405 ( ) 350 ( ) 345 ( ) 370 ( ) 357 ( ) 95% boost dose (cm 3 ) 166 (51-470) 170 (55-456) 170 (58-451) 167 (51-467) 167 (54-455) Heart V30 (%) 2.1 ( ) 1.6 ( ) 1.1 ( ) 1.6 ( ) 1.1 ( ) Mean dose (Gy) 4.1 ( ) 3.5 ( ) 3.0 ( ) 3.4 ( ) 3.0 ( ) Lungs V20 (%) 5.0 ( ) 4.4 ( ) 3.8 ( ) 4.3 ( ) 3.8 ( ) Mean dose (Gy) 3.7 ( ) 3.4 ( ) 3.2 ( ) 3.3 ( ) 3.0 ( ) Contralateral breast V1 (%) 15.0 ( ) 16.8 ( ) 17.5 ( ) 16.9 ( ) 16.9 ( ) V5 (%) 0.2 ( ) 0.1 ( ) 0.2 ( ) 0.1 ( ) 0.2 ( ) Mean dose (Gy) 0.5 ( ) 0.6 ( ) 0.6 ( ) 0.5 ( ) 0.5 ( ) Abbreviations: 3D-CRT-SIB = three-dimensional conformal radiotherapy including simultaneously integrated boost; 25%IMRT = 75% conformal beams with skin flash and 25% IMRT segments; 100%IMRT = fully segmented IMRT; IMRT-SIB = intensity modulated radiotherapy including simultaneously integrated boost; Vx = proportion of organ at risk receiving x Gy. Data presented as mean values, with ranges in parenthesis. Dose to OARs With the results of all studied patients combined, the mean parameter values for heart and lung dose with 3D-CRT-SIB were somewhat lower with IMRT-SIB (Table 2). The average reduction of heart V30 with the IMRT-SIB strategies ranged from 0.5% - 1.0% relative to 3D-CRT-SIB and the average reduction of mean heart dose ranged from 0.6 Gy to 1.1 Gy. The average reduction of lung V20 with the various IMRT-SIB strategies ranged from 0.6% - 1.2% relative to 3D-CRT-SIB and

80 Strategies for simultaneously integrated breast boost the average reduction of mean lung dose ranged from 0.3 Gy to 0.7 Gy. Only marginal differences were found between the different IMRT-SIB strategies, with a slight benefit for the 100%IMRT-SIB methods. NTCP values for radiation pneumonitis Grade 2 or higher were zero for both IMRT-SIB and IMRT-SIB when using the model parameters derived by Seppenwoolde et al., whereas it ranged from 1.5% with IMRT-SIB to 1.7% with 3D-CRT-SIB when using the model parameters derived by Semenenko et al.. Similar dose was delivered to the contralateral breast with 3D-CRT-SIB and the various IMRT-SIB strategies, except for a slight increase in the low dose values with IMRT-SIB. Predictive patient characteristics In the second part of the analysis, we investigated whether OAR dose reductions with IMRT-SIB, observed in the group of patients as a whole, would also apply to particular subgroups of patients. It appeared that significant correlations were present between the OHB and reductions in heart V30 with IMRT-SIB (r = 0.38, p <0.001), the size of the boost PTV and reductions in V54.23 (r = 0.54, p <0.001), the size of the breast PTV and reductions in V54.23 (r = 0.24, p = 0.01), the boost location in cranio-caudal direction and reductions in lung V20 (r = 0.51, p <0.001), the boost location in cranio-caudal direction and reductions in mean lung dose (r = 0.60, p <0.001), the boost location in medio-lateral direction and reductions in lung V20 (r = 0.24, p = 0.01) and the boost location in medio-lateral direction and reductions in mean lung dose (r = 0.24, p = 0.01). We examined the various patient characteristics in their ability to predict whether IMRT-SIB would result in a more than average reduction of the correlated dose parameters. We found that the OHB factor was a good predictor for a more than average reduction in heart V30 (>0.7%). The ROC area under the curve (AUC)-value was 0.75 (95% CI: ), and the optimal cut-off value was 1.4 cm (Fig. 3). Patients with an OHB >1.4 cm had a significantly larger mean reduction in heart V30 with IMRT-SIB (a mean reduction of 1.3%) than patients with OHB 1.4 cm (a mean reduction of 0.5%, p <0.001). Similarly, we found that the size of the boost PTV was a good predictor for a more than average reduction in

81 Chapter 5 V54.23 (>50 cm 3 ). The ROC AUC-value was 0.77 (95% CI: ), and the best cut-off value was a boost PTV of 125 cm 3. Patients with a larger boost PTV also had a significantly larger mean reduction in V54.23 (a mean reduction of 83 cm 3 ) than patients with a smaller boost PTV (a mean reduction of 21 cm 3, p <0.001). The size of the breast PTV appeared not to be a good predictor for a reduction in V The ROC AUC-value was 0.60 (95% CI: ) and no cut-off value could discriminate two groups that had significantly different reductions in V54.23 with IMRT-SIB. Figure 3. Receiver operating characteristic curves The receiver operating characteristic (ROC) curves and the associated areas under the curve for A: the overlap between heart and breast planning target volume (OHB) (Fig. 2) as a predictor of a more than average reduction in the proportion of the heart receiving 30 Gy or more (V30) with intensity modulated radiotherapy including simultaneously integrated boost (IMRT-SIB) when compared to threedimensional conformal radiotherapy (3D-CRT)-SIB, and B: the size of the boost planning target volume (PTV) as a predictor of a more than average reduction in the volume receiving 107% of the prescribed dose to the breast (V54.23) with IMRT-SIB when compared to 3D-CRT-SIB. Significantly different reductions in lung dose with IMRT-SIB were found for the various boost locations. In the case of cranial boost locations, IMRT-SIB showed larger mean reductions in lung dose (V20: 1.3%; mean lung dose: 0.9 Gy) than in the case of caudal boost locations (V20: 0.6%; mean lung dose: 0.1 Gy, both p <0.001). Similarly, in the case of medial boost locations, IMRT-SIB showed larger mean reductions in lung dose (V20: 1.0%; mean lung dose: 0.7 Gy) than in

82 Strategies for simultaneously integrated breast boost the case of lateral boost locations (V20: 0.7%, p = 0.02; mean lung dose: 0.4 Gy, p = 0.03). However, with regard to lung dose, NTCP values were almost similar with each of the various techniques. Therefore the relevance of using the boost location as a predictor for a benefit of IMRT-SIB is unclear. When the OHB factor and the size of the boost PTV would both be used to identify patients that might benefit from IMRT-SIB, 18 of the 30 patients examined in this study would be identified, 8 on the basis of the OHB factor alone and 14 on the basis of the size of the boost PTV alone. Discussion The application of SIB in breast conserving RT has proven feasible with both 3D-CRT and IMRT [6,9]. While reports on whole breast IMRT have demonstrated that dose delivered to OARs can be reduced as compared to whole breast 3D-CRT [7,8], the results of the present study demonstrated, that in the case of SIB, relevant reductions with IMRT are present in subgroups of patients only. The benefits of IMRT-SIB compared to 3D-CRT-SIB appeared to be a reduced V54.23 and a slightly lower heart V30 in subgroups of patients. In fact, when these reductions in dose were analyzed in relation to the size of the boost PTV and the OHB value, it appeared that the added value of using IMRT-SIB instead of 3D- CRT-SIB was minimal in patients with a boost PTV volume 125 cm 3 and an OHB 1.4 cm. Although we also found that with IMRT-SIB, reductions in lung dose can be obtained, this reduction did not translate into a reduced NTCP-value of radiation pneumonitis, i.e., the probability of this complication was similar, regardless of the SIB technique used. With regard to physical dose reductions in heart and breast tissue, it remains unclear if these might result in reduced NTCPvalues of cardiac morbidity and breast fibrosis, as reliable and validated NTCPmodels for these endpoints are still lacking. Future research should therefore be focused on the validation of existing NTCP models [14,15], and the collection of data providing insight in the clinical relevance of different dose distributions in various OARs

83 Chapter 5 Although specific subgroups of patients could benefit more from IMRT-SIB because of the lower heart V30 in some patients, it should be noted that the use of breath-hold techniques may also reduce heart dose values [16]. As breath-hold techniques can also be used in combination with 3D-CRT-SIB, this could further diminish the additional benefit of IMRT-SIB relative to 3D-CRT-SIB with respect to potential heart dose reduction. Some may argue that as many patients as possible should be treated with IMRT, regardless the magnitude of its potential benefit. However, it should be noted that this can only be the realised in institutes that have a long experience with IMRT, have access to multiple accelerators that are IMRT compatible, have invested in treatment planning software up to twice the cost of conventional software, have educated all personnel to ensure quality of each step in the IMRT process, have the disposal of additional physicists to perform the necessary quality assurance tests and can still treat all patients in the case IMRT would result in increased treatment delivery time per patient. In addition, the total annual radiotherapy costs would largely increase when all breast cancer patients are to receive IMRT instead of 3D-CRT. In a comprehensive report evaluating the cost effectiveness of IMRT, the Belgium government has calculated in 2007 that implementation of IMRT instead of 3D-CRT for each Belgium breast cancer patient would imply a total increase of the total annual radiotherapy costs of 18.7% [11]. Therefore, the authors feel that in the absence of clinical trials proving that IMRT- SIB results in improved local control or decreased side effects, alternative treatments that are easier to implement but have similar dose distributions, should be considered. Moreover, it is of particular interest to know whether particular patients may benefit from the use of IMRT-SIB or, conversely, that similar results can also be obtained with 3D-CRT-SIB. It was already demonstrated by Hurkmans and colleagues for whole breast IMRT that the maximum heart distance, defined as the maximum distance of the heart contour to the medial field edge as seen in a beam s-eye-view of the medial tangential field, correlated with dose levels delivered to the heart [7]. However, when applying (fully segmented) IMRT-SIB, no conformal beams are available for

84 Strategies for simultaneously integrated breast boost appropriate measurement of the maximum heart distance. Therefore we propose the use of the OHB factor as a criterion when identifying patients for IMRT-SIB. In the present study, we examined two IMRT-SIB strategy variables: 1) the method of IMRT optimisation, and 2) the method of IMRT segmentation. With regard to the optimisation method, we found that dosimetric results with CS plans, generated on the basis of a standard set of structures and objectives, were similar to that with OS plans thoroughly optimised to minimise dose to OARs. With regard to the segmentation method it appeared that with 100%IMRT-SIB, dose delivered to the heart could be slightly lower than with 25%IMRT-SIB. This might indicate the CS 100%IMRT-SIB as the preferred strategy when IMRT-SIB is implemented. In our previous study on 3D-CRT-SIB [6], target coverage was considered adequate when 95% of the prescribed dose was delivered to at least 99% of the PTV, while in the current study we accepted a coverage of 98%. This allowed for better sparing of the heart with both 3D-CRT-SIB and IMRT-SIB. As a result, OAR dose values with 3D-CRT-SIB in the current study are somewhat lower than those in our previous study. Conclusions In breast conserving RT, similar results can be obtained with forward planned 3D-CRT-SIB and inversely planned IMRT-SIB for a large proportion of patients. Although IMRT-SIB is generally capable to reduce absolute dose values to OARs, the actual benefit of IMRT-SIB was different for specific patient subgroups. Patient characteristics, such as the OHB factor and the size of the boost PTV, can be used to identify patients that are most likely to benefit from IMRT-SIB

85 Chapter 5 References 1. Early Breast Cancer Trialists' Collaborative Group. Effects of radiotherapy and surgery in early breast cancer. An overview of the randomized trials. Early Breast Cancer Trialists' Collaborative Group. N Engl J Med 1995;333: Fisher B, Anderson S, Bryant J, et al. Twenty-year follow-up of a randomized trial comparing total mastectomy, lumpectomy, and lumpectomy plus irradiation for the treatment of invasive breast cancer. N Engl J Med 2002;347: Bartelink H, Horiot JC, Poortmans PM, et al. Impact of a higher radiation dose on local control and survival in breast-conserving therapy of early breast cancer: 10-year results of the randomized boost versus no boost EORTC trial. J Clin Oncol 2007;25: Veronesi U, Luini A, Del Vecchio M, et al. Radiotherapy after breast-preserving surgery in women with localized cancer of the breast. N Engl J Med 1993;328: Guerrero M, Li XA, Earl MA, et al. Simultaneous integrated boost for breast cancer using IMRT: a radiobiological and treatment planning study. Int J Radiat Oncol Biol Phys 2004;59: van der Laan HP, Dolsma WV, Maduro JH, et al. Three-dimensional conformal simultaneously integrated boost technique for breast-conserving radiotherapy. Int J Radiat Oncol Biol Phys 2007;68: Hurkmans CW, Cho BC, Damen E, et al. Reduction of cardiac and lung complication probabilities after breast irradiation using conformal radiotherapy with or without intensity modulation. Radiother Oncol 2002;62: Ahunbay EE, Chen GP, Thatcher S, et al. Direct aperture optimization-based intensity-modulated radiotherapy for whole breast irradiation. Int J Radiat Oncol Biol Phys 2007;67: Hurkmans CW, Meijer GJ, van Vliet-Vroegindeweij C, et al. High-dose simultaneously integrated breast boost using intensity-modulated radiotherapy and inverse optimization. Int J Radiat Oncol Biol Phys 2006;66: Haffty BG, Buchholz TA, McCormick B. Should intensity-modulated radiation therapy be the standard of care in the conservatively managed breast cancer patient? J Clin Oncol 2008;26:

86 Strategies for simultaneously integrated breast boost 11. Van den Steen D, Hulstaert F, Camberlin C. Intensiteitsgemoduleerde Radiotherapie (IMRT), Health Technology Assessment (HTA), KCE reports 62A (D2007/10.273/32). Brussel: Federaal Kenniscentrum voor de Gezondheidszorg (KCE); Seppenwoolde Y, Lebesque JV, de Jaeger K, et al. Comparing different NTCP models that predict the incidence of radiation pneumonitis. Normal tissue complication probability. Int J Radiat Oncol Biol Phys 2003;55: Semenenko VA, Li XA. Lyman-Kutcher-Burman NTCP model parameters for radiation pneumonitis and xerostomia based on combined analysis of published clinical data. Phys Med Biol 2008;53: Gagliardi G, Lax I, Ottolenghi A, et al. Long-term cardiac mortality after radiotherapy of breast cancer--application of the relative seriality model. Br J Radiol 1996;69: Alexander MA, Brooks WA, Blake SW. Normal tissue complication probability modelling of tissue fibrosis following breast radiotherapy. Phys Med Biol 2007;52: Remouchamps VM, Vicini FA, Sharpe MB, et al. Significant reductions in heart and lung doses using deep inspiration breath hold with active breathing control and intensity-modulated radiation therapy for patients treated with locoregional breast irradiation. Int J Radiat Oncol Biol Phys 2003;55:

87 - 86 -

88 Chapter 6 Comparison of normal tissue dose with three-dimensional conformal techniques for breast cancer irradiation including the internal mammary nodes Hans Paul van der Laan, Wil V. Dolsma, Aart A. van t Veld, Henk P. Bijl, Johannes A. Langendijk International Journal of Radiation Oncology Biology Physics 2005; 63:

89 Chapter 6 Abstract Purpose: To compare the Para Mixed technique for irradiation of the internal mammary nodes (IMN) with three commonly used strategies, by analyzing the dose to the heart and other organs at risk. Materials and Methods: Four different three-dimensional conformal dose plans were created for 30 breast cancer patients. The IMN were enclosed with the Para Mixed technique by a widened medio-lateral tangential photon beam and an anterior electron beam, with the Patched technique by an anterior electron beam, with the Standard technique by an anterior photon and electron beam, and with the PWT technique by partially wide tangential beams. All techniques were optimised for conformity and produced equally adequate target coverage. Results: Heart dose was lowest with the Para Mixed and Patched technique for all patients and with the PWT technique for right-sided treatment only. Lung dose was highest with the PWT, lowest with the Patched, and intermediate with the Para Mixed and Standard techniques. Skin dose was highest with the Patched, lowest with the PWT, and intermediate with the Para Mixed and the Standard techniques. The Para Mixed technique resulted in a 13-Gy lower dose in an overlap area, and the PWT technique was the only technique that incorporated considerable volumes of the contralateral breast. Conclusion: The Para Mixed technique yielded the overall best results. No other technique resulted in a lower heart dose. Lung and skin were equally spared instead of one of them being compromised, and the contralateral breast was avoided

90 3D-CRT for loco-regional breast cancer irradiation Introduction There is no general consensus with regard to the indications for elective irradiation of the internal mammary nodes (IMN) in the treatment of breast cancer [1]. The higher probability of radiation-induced toxicity is of concern, and it remains unclear whether such treatment will provide better loco-regional control and prevent further dissemination. The IMN were routinely incorporated into the target volume in a number of breast cancer trials that proved the benefits of radiotherapy with respect to survival. However, no difference in outcome has been observed yet in randomized studies that specifically focused on the possible benefit of IMN irradiation [2,3]. Only a few studies showed a decreased incidence of locoregional recurrence in small subgroups of patients [4,5]. A number of studies pointed out that patients who received partial irradiation of the heart had an increased risk of dying from cardiac disease [6-9]. This thin line between advantages and side effects has caused actual practise to be culture driven rather than evidence-based [10,11]. Nevertheless, it should be realised that in many studies and trials that proved the benefits and harms of IMN irradiation, outdated treatment techniques were used, and in many cases relatively large volumes of the heart were irradiated. Currently, a number of tools and techniques are available that enable us to reduce the dose to organs at risk, such as the heart and lungs. While treatment planning moved toward the era of three-dimensional conformal radiotherapy (3D-CRT), Marks et al. [12] proposed a possible solution to reduce normal tissue dose with the so-called partially wide tangents (PWT) technique. With this technique, the heart and lungs could be spared more adequately than with conventional two-dimensional techniques. In 1998, Jansson et al. [13] drew attention to the drawbacks of the PWT technique, which included the relatively high lung dose, the high heart dose in left-sided treatment, and the irradiation of the contralateral breast with the PWT technique. The Patched technique was proposed. With this technique, the medial chest wall and the IMN were treated with electrons. Although the dose to the heart, lungs, and contralateral breast was further reduced compared with the PWT technique, this Patched

91 Chapter 6 technique had the drawback of a high skin dose and matchline problems. In 2000, Hurkmans et al. [14,15] described two techniques to treat the IMN with a mix of photons and electrons. With these techniques, both skin dose and lung dose were reduced. On the basis of these and other techniques, a number of investigators performed comparative planning studies. Although the PWT technique was advocated in three of these studies [16-18], questions could be raised regarding the way treatment planning was performed. Non-conformal photon electron techniques were compared with PWT techniques that were constructed to effectively shield normal tissue. In recent studies, similar normal tissue complication probabilities (NTCPs) for heart and lung were predicted for intensitymodulated radiotherapy (IMRT) and photon electron techniques [19,20]. In these studies, similar target volumes and margins were used with all techniques. However, the photon electron plans were not yet created with the optimal use of 3D-CRT. At the University Medical Center Groningen, we developed a new technique to irradiate the parasternal region with a mix of photons and electrons, the Para Mixed technique [21]. This technique is a further improvement of the techniques proposed by Hurkmans and others because all beams are constructed, shaped, and weighted according to individual patient characteristics. The beam configuration used enables a further reduction of normal tissue dose while adequate target coverage is maintained. In this study, we compared the Para Mixed technique with three other commonly used strategies that were all optimised with the use of 3D-CRT. Our purpose was to test the hypothesis that the Para Mixed technique is the most optimal technique in case of IMN irradiation in terms of heart dose, lung dose, skin dose, and dose in the contralateral breast. Materials and Methods Patients Thirty patients scheduled for radiotherapy after mastectomy or breastconserving surgery were included in this study. All patients were planned to receive irradiation of the IMN and supraclavicular nodes as part of the treatment. Thirteen

92 3D-CRT for loco-regional breast cancer irradiation patients were treated for right-sided breast cancer (breast = 6, chest wall = 7), and 17 patients were treated for left-sided breast cancer (breast = 8, chest wall = 9). A planning CT scan was made for each patient. Of the 35 patients who were initially included in this study, 5 with IMN exceeding a depth of 5 cm were excluded because the adequate use of electrons would be precluded in these patients. Patient positioning and acquisition of CT data Patients were positioned on a breast board with the sternum parallel to the table, and the ipsilateral arm was abducted above the head. Before the CT scan, skin marks were placed to enable patient repositioning during treatment. Radiopaque catheters and markers were placed to locate the palpable breast, scar, and skin marks on the CT images. Patients were scanned from the level of the larynx to the level of the upper abdomen, including both lungs, with a scan thickness and index of 5 mm. CT data for all patients were transferred to the Helax- TMS 3D treatment-planning system (Nucletron, Veenendaal, The Netherlands), version 6.1B. Target volumes and organs at risk After breast-conserving surgery, the clinical target volume (CTV) consisted of the ipsilateral breast, the IMN, and the supraclavicular nodes. The breast CTV was defined as the breast parenchyma visible on the CT images, which was generally consonant with the palpable breast marked by the radiopaque catheter. After mastectomy, the CTV consisted of the residual tissue of the chest wall, the IMN, and the supraclavicular nodes. The chest wall CTV extended within the confines of the markers placed on the skin surface before the CT scan. The most posterior aspect of the chest wall CTV was the anterior rib surface. The supraclavicular nodes CTV was defined from the level of the cricoid cartilage to the supraclavicular fossa. The IMN CTV was defined by an elliptical cylinder with lateral and anterior posterior diameters of 15 mm and 10 mm, respectively, placed adjacent to the edge of the sternum medially and to the pleura posteriorly. The IMN CTV extended from the inferior aspect of the ipsilateral clavicular head through the fourth intercostal

93 Chapter 6 space. The planning target volume (PTV) was obtained for each CTV by applying a margin of 5 mm in three dimensions. The heart was contoured to the level of the pulmonary trunk superiorly, excluding the major vessels. Both lungs were contoured with the automatic contouring tool of the Helax-TMS planning system, edited, and then verified. The contralateral breast was contoured as the breast parenchyma visible on the CT images. General conditions for treatment planning An identical matchline was used with all techniques. Superior to the matchline, the supraclavicular nodes PTV was treated. Seventy-five percent of the prescribed dose was delivered to the supraclavicular nodes with an anterior photon beam, and 25% with a posterior photon beam. The central axes and inferior borders of these beams were aligned with the matchline. All techniques produced the same dose distribution superior to the matchline. The relative volume of breast or chest wall PTV receiving at least 95% of the prescribed dose was never more than 2% different between the techniques. At least 98% of the IMN PTV received 85% of the prescribed dose with all techniques, which is consistent with the requirements of European Organization for Research and Treatment of Cancer (EORTC) protocol [2]. This protocol states that the target area of the IMN should be covered by at least the 85% isodose. When electron beams were used, they were shaped and directed in such a way that the IMN PTV received the required dose. Multileaf collimator shielding was used with all photon beams. A heavy metal mould was used to collimate the electron beams. All shielding was conformal to the PTV, with a margin of 5 mm outside the PTV to account for penumbra. Wedges were used in the tangential beams with all patients. In addition to this, a maximum of three small sub-beams was used to obtain a homogeneous dose distribution. The wedge fraction and the relative weights of any sub-beams were weighted manually (i.e., by forward planning). The treatment plans were normalized at the International Commission on Radiation Units and Measurements (ICRU) reference point of the breast or chest wall PTV [22]. The point was chosen to conform to the General

94 3D-CRT for loco-regional breast cancer irradiation Recommendations for Reporting Doses [22] at the center of the PTV, at or close to the isocenter, and always in a region where there is no steep dose gradient. The ICRU reference dose was 2 Gy per fraction, with a total treatment dose of 50 Gy. The dose delivered to the IMN region was prescribed to an IMN dose calculation point. The photon energy was 6 MV in all cases. The electron energy ranged from 8 to 14 MeV. The pencil beam algorithm of the Helax-TMS treatment-planning system was used to calculate full 3D dose distributions, and lung density corrections were used for all techniques. Para Mixed technique treatment planning Inferior to the matchline, three tangential photon beams and one anterior electron beam were matched. For the tangential beams, tabletop rotations and collimator rotations were used to create a divergence-free match to the inferior borders of the supraclavicular beams. Two standard tangential beams and one additional widened medio-lateral tangential beam were used to treat the breast or chest wall PTV. The rotation point of these beams was placed centrally in the breast or chest wall PTV. The widened medio-lateral tangential beam enclosed the IMN PTV and the medial part of breast or chest wall PTV that was excluded with the standard tangential beams. The dose delivered to these regions was then supplemented with an anterior electron beam (gantry angle 0, source-to-surface distance 100 cm). There was an overlap area between the electron beam and the two standard tangential beams. The ipsilateral border of the electron field was defined slice by slice to include the PTV region that was excluded with the standard tangential beams. The location of the overlap area was anticipated when the gantry angles of the standard tangential beams were defined. For chest wall irradiation, a more lateral ipsilateral location was chosen for the overlap area to reduce the dose to the heart and the ipsilateral lung. As a result of this, not only the IMN region, but also the medio-caudal part of the chest wall was irradiated with a mix of photons and electrons

95 Chapter 6 The electron energy was determined by the depth of the IMN. The IMN dosecalculation point was placed centrally in the electron beam at the depth of the dose maximum. An average of 62% of the prescribed dose was delivered to the IMN dose-calculation point with the electron beam. The widened medio-lateral photon beam was weighted to deliver the remaining percentage of the prescribed dose to the IMN dose-calculation point. In this way, the volume of lung receiving 20 Gy was limited, and skin dose in the parasternal region was limited to a dose of 36 Gy. The 36-Gy skin dose (defined at a depth of 1 mm) resulted from the photon electron mix of 38%/62% when 10-MeV or 12-MeV electrons were used. When 14-MeV electrons were used, in case of deeper IMN, electrons became less efficient, and skin dose increased. For this reason, a lower percentage of electrons (58%) was used in a mix containing 14-MeV electrons and a higher percentage of electrons (66%) was used in a mix containing 8-MeV electrons. The combined weight of the two medio-lateral tangential beams was identical to that of the single latero-medial tangential beam. The gantry angle for the widened medio-lateral tangential beam was chosen to minimise lung dose and heart dose, without including the contralateral breast. Shielding of the heart was further optimised by changing the multileaf collimator settings of the standard tangential beams, as long as the ipsilateral border of the electron beam could be shaped accordingly without compromising adequate target coverage. Patched, Standard, and PWT technique treatment planning The Para Mixed technique was compared with the Patched, the Standard, and the PWT techniques (Fig. 1). With the Patched technique, a conformal anterior electron beam was used to treat the IMN. The beam configuration was identical to that with the Para Mixed technique, except that the widened medio-lateral tangential beam was excluded. The two remaining tangential beams were of the same weight, and the weight of the electron beam was adjusted to deliver 100% of the prescribed dose to the IMN

96 3D-CRT for loco-regional breast cancer irradiation Figure 1. Overview of the techniques compared in this study The Para Mixed technique (A); The Patched technique (B); The Standard technique (C); and The Partially Wide Tangents technique (D). The dose distribution is represented by the 20%, 50% (dark blue), 85% (blue), 95% (green), 125% (orange), and 140% (red) isodose lines

97 Chapter 6 With the Standard technique, a mix of anterior photon and electron beams was used to deliver the prescribed dose to the IMN. The beam configuration with the Standard technique was identical to that with the Patched technique, except that a 6-MV photon beam was added with a field shape identical to that of the electron beam. Heavy metal shielding blocks were used, and the superior border was matched divergence-free to the matchline. The contribution of the electron beam was similar to that with the Para Mixed technique. With the PWT technique, wide tangential photon beams were matched divergence-free inferior to the matchline. With the use of beam s-eye-view projections, optimal gantry angles were determined to achieve maximum avoidance of the heart and the ipsilateral lung, without including more than 25% of the contralateral breast. Analyses of normal tissue dose In studies focusing on patients receiving radiotherapy for breast cancer and Hodgkin s disease, 30 Gy has been suggested as a threshold dose for ischemic heart disease [9,23-25]. Therefore, the relative volume of the heart receiving 30 Gy (V30) and the mean heart dose were obtained from the dose volume histograms (DVHs) of the different techniques. The mean lung dose and the relative volume of lung receiving 20 Gy (V20) are regarded to be predictors of radiation pneumonitis [26,27]. Therefore, the mean lung dose and the lung V20 (considering both lungs as one organ) were obtained from the DVHs. Irradiation of the contralateral breast gives reason for concern. It is demonstrated that the risk of breast cancer is closely associated with breast tissue dose [28]. A linear dose relationship is maintained at lower radiation doses, and there exists no low-dose threshold below which there is no excess risk. The contralateral breast mean dose and its relative volume receiving 3 Gy (V3) and 10 Gy (V10) were used to quantify the volume of contralateral breast irradiated with the different techniques. Bad cosmetic results and up to a 72% chance for late radiation-induced telangiectasia have been reported in patients receiving full electron irradiation for

98 3D-CRT for loco-regional breast cancer irradiation breast cancer [29,30]. Therefore, the dose delivered to the skin in the IMN region was evaluated for all techniques. Owing to limitations of the treatment-planning system, it was not possible to obtain accurate DVHs of the skin. The skin dose was therefore estimated according to data on beam characteristics of photons and electrons measured at our institution. When beams overlap below the surface of the skin, high doses can accumulate, and subcutaneous fibrosis is often seen as a late side effect. For all techniques in which an overlap was present, the maximum dose was calculated by the treatment-planning system. The average maximum equivalent dose in 2-Gy fractions was calculated with an α/β of 1.7 Gy, as defined for subcutaneous fibrosis by Bentzen and Overgaard [31]. Statistical analysis To compare DVH parameters of the different techniques, the mean values were analyzed with the Wilcoxon signed ranks test or the paired-samples t-test whenever appropriate. All tests were two-tailed, and differences were considered statistically significant when the p value was 0.05 or less. Table 1. Mean dose in Gy and percent volume of heart, lung, and contralateral breast irradiated Heart Lungs Contralateral breast Right-sided treatment Left-sided treatment Technique Mean dose V30 Mean dose V30 Mean dose V20 Mean dose V3 V10 Para Mixed 3.8 (1.3) 1.2 (1.8) 8.7 (3.5) 7.4 (5.8) 9.3 (1.8) 18.0 (4.3) 1.0 (0.3) 2.4 (2.1) 0.5 (0.9) Patched 4.3 (1.6) a 1.9 (2.2) a 7.5 (2.6) a 7.1 (5.3) 7.9 (1.5) a 14.4 (3.3) a 0.6 (0.2) a 0.3 (0.8) a 0.1 (0.3) a Standard 7.2 (1.7) a 3.8 (2.9) a 12.9 (3.8) a 9.6 (5.7) a 9.5 (2.0) 16.5 (4.2) a 0.7 (0.2) a 0.5 (0.9) a 0.1 (0.3) a PWT 3.4 (1.0) 1.9 (3.4) 11.1 (4.4) a 16.0 (9.2) a 12.2 (2.5) a 23.3 (5.3) a 2.8 (1.6) a 15.7 (7.8) a 5.2 (4.3) a Abbreviation: PWT = partially wide tangents. Dose data presented in Gy and volume data in %, with the standard deviation in parenthesis. a Significantly different from Para Mixed technique, p

99 Chapter 6 Results Heart With left-sided treatment, the heart received the least dose with the Patched and Para Mixed technique, considering both the mean heart dose and the heart V30 (Table 1). With right-sided treatment, the heart received the least dose with the Para Mixed and the PWT technique. The heart dose was significantly higher with the Standard technique for both left- and right-sided treatment together with the PWT technique for left-sided treatment. The only significantly better result compared with the Para Mixed technique was the mean heart dose with the Patched technique for left-sided treatment (7.5 Gy vs. 8.7 Gy). Lungs The mean lung dose and lung V20 with the Para Mixed technique were lower than with the PWT technique but higher than with the Patched technique (Table 1). The lung dose with the Para Mixed technique was comparable to that with the Standard technique. The PWT technique resulted in a relative volume of lung receiving 35 Gy that was almost two times greater than with the other techniques (Fig. 2). All techniques resulted in a significantly greater lung V20 and mean lung dose for the chest wall cases, whereas the absolute differences between the techniques were generally the same for the breast and chest wall cases. Contralateral breast The contralateral breast was excluded in most cases with the Para Mixed technique. Because the contralateral breast was often close to the widened tangential photon beam, a higher dose was delivered to the contralateral breast with the Para Mixed technique than with the Patched and Standard techniques (Table 1). The greatest volume of contralateral breast was irradiated with the PWT technique. The relative volume receiving 10 Gy with the PWT technique was 10 times greater than with the Para Mixed technique

100 3D-CRT for loco-regional breast cancer irradiation Figure 2. Cumulative average dose volume histograms Dose volume histograms represent the heart in left-sided treatment (n = 17), the heart in right-sided treatment (n = 13), and the lungs (n = 30). Abbreviation: PWT = partially wide tangents

101 Chapter 6 Table 2. Skin dose IMN region and maximum dose overlap area Skin Estimated dose (Gy) Average maximum dose (Gy) Overlap area Average maximum equivalent dose in 2 Gy fractions (Gy) a Para Mixed Patched Standard PWT Abbreviations: IMN = internal mammary nodes; PWT = partially wide tangents. a The average maximum equivalent dose in 2-Gy fractions was calculated with an α/β of 1.7 Gy, as defined for subcutaneous fibrosis by Bentzen and Overgaard [31]. Skin and overlap The estimated skin dose in the area covered by the electron beam was highest with the Patched technique (Table 2). Skin dose with the Para Mixed and Standard techniques was estimated to be 7 Gy lower than with the Patched technique. The PWT technique resulted in the lowest skin dose because no electron beams were used. The Para Mixed technique resulted in an average maximum dose in the overlap area that was 13 Gy lower compared with the Patched and Standard techniques. No overlap was present in the PWT technique. Discussion In the present study, the Para Mixed technique was compared with three commonly applied radiotherapy strategies. 3D-CRT treatment planning was performed for all techniques, and all techniques produced equally adequate target coverage of the PTV. The Para Mixed technique performed well compared with the other techniques. The Patched technique resulted in a lower lung dose but delivered the highest dose to the skin and the overlap area. The Standard technique resulted in a higher heart dose and a higher dose in the overlap area. The PWT technique was easier to construct because no electron beams were used. Although high doses to the skin and overlap area were avoided, the PWT resulted in the highest lung dose, a higher heart dose in left-sided treatment, and a considerable

102 3D-CRT for loco-regional breast cancer irradiation volume of the contralateral breast was irradiated. Unlike with the other techniques, partial irradiation of the contralateral breast could not be avoided with the PWT technique because this would have resulted in a significant increase of the dose delivered to the heart and lungs. The findings of the present study are not consistent with reports from three other studies that proposed the PWT technique to be the most favourable technique. In a study conducted by Severin et al. [16], the PWT technique was compared with a technique in which a mix of photons and electrons was used to treat the IMN region. Patients with deep IMN (>6 cm) were not excluded, resulting in high-energy photons and electrons with the photon electron technique, although an adequate coverage of the IMN could not be achieved. This is why patients with deep IMN were excluded in the present study and another study [14]. In the studies conducted by Severin et al. [16] and Arthur et al. [17], customshaped blocks were used with the PWT technique. The heart and ipsilateral lung were effectively shielded, allowing only limited volumes to be irradiated. The PWT technique was compared with a photon electron technique in which standard beams without shielding of the heart and the ipsilateral lung were used. In a study conducted by Pierce et al. [18], the PWT technique was modified to completely shield the heart and then treat the shielded area with electrons. Although the method used in constructing these electron beams was not specified, it seems that heart and lung dose were minimised because no margins for penumbra or position uncertainties were used to define the field edges around the IMN with the PWT technique, and relatively low-energy electrons were used (6 9 MeV). The PWT technique was compared with two photon electron techniques, in which standard IMN beams were used without effective shielding for the heart and the ipsilateral lung. Severin et al. [16] reported that 24.5% of the contralateral breast received 2.5 Gy with the PTW technique. Comparative planning studies in which similar margins and shielding were applied for all techniques consistently show similar or better results with photon electron techniques than with techniques in which wide tangential photon beams are used [13,19,20,32]. Two photon electron techniques were compared by

103 Chapter 6 Hurkmans et al. [14,15]: the Standard technique and the Improved technique. With both techniques, beam configurations with a fixed position and width for the IMN beams were defined, and a standard overlap size was used. In the present study, the Standard technique was modified and optimised to be compared with the Para Mixed technique. Although the Para Mixed technique was based on the ideas proposed by Hurkmans et al., some technical improvements were introduced. A new beam configuration was used to enable more effective shielding and optimisation. Moreover, the Para Mixed technique is a real 3D-CRT technique with an integrated electron beam, which is individually weighted and shaped for each patient to obtain adequate target coverage and limit the radiation dose to normal tissue. In the studies conducted by Hurkmans et al., the Improved technique was proposed to be the most favourable technique. It was later referred to as the Oblique electron technique in a study by Cho et al. [20]. They compared it with a PWT technique and an IMRT technique. Higher NTCPs for heart and lungs were found with the PWT technique compared with the Oblique electron technique and the IMRT technique. The heart and lung NTCPs were similar for the Oblique electron technique and the IMRT technique. In a study conducted by Johansson et al. [19], similar NTCPs were found for IMRT and the Patched technique. However, the photon electron plans discussed were not yet created with the optimal use of 3D-CRT, and they can be further improved. 3D-CRT solutions like the Para Mixed technique require more resources than conventional two-dimensional plans. However, the level of complexity involved with the construction and delivery of the Para Mixed technique is similar to that of other advanced multi-beam 3D-CRT techniques. Whereas treatment-planning is performed by a select group of dosimetrists, the delivery is performed by all radiation technologists. During this study, IMN irradiation was restricted to patients who had an internal mammary sentinel node that was identified by lymphoscintigraphy (either unexamined or proved positive by histopathologic evaluation). This resulted in a group of more than 30 patients who were treated with the Para Mixed technique

104 3D-CRT for loco-regional breast cancer irradiation Study outcomes are influenced by the way target volumes and margins are defined. For many years, tangential radiation fields were defined during simulation, according to X-ray imaging and exterior patient anatomy. Even at present, when CT data are available, many studies use standard field borders for clinical target definition. There seems to be a certain reluctance to use density information and margins when these might result in larger fields including larger volumes of normal tissue within the treated volume. The same inconsistency is observed regarding the definition of the IMN CTV. Although the internal mammary vessels were used as a reference in many studies, varying volumes were delineated to include the IMN. In this study, the IMN CTV extended through the fourth intercostal space, in accordance with EORTC protocol [2]. This protocol states that the IMN target volume should include at least the IMN chain in the first three intercostal spaces but, depending on tumour location, can be extended through the fifth intercostal space. We chose an intermediate range as a constant factor with all patients. The ipsilateral edge of the sternum and the pleura were regarded as the outer boundaries enclosing the IMN in lateral and dorsal directions, respectively. It is clear that questions regarding this issue remain and that study results are affected by the way the target volumes are defined. Recently published material might help provide solutions [33,34]. It also remains unclear which margins should be used around the CTV to take into account position uncertainties and penumbra. We believe that the margins used in this study were sufficient for breast cancer radiotherapy. Most importantly, the margins were kept the same for all techniques to make a reliable assessment of the normal tissue dose with the different techniques. In the case of loco-regional irradiation, the target volume often has a curved shape, with internal concave regions. One can use PWT to avoid the problems of a photon electron match, but this will include all normal tissue within the concavity of the target volume. Conversely, one can use an electron beam to treat the medial part of the target volume to reduce the size of the concavity that is included within the tangential photon beams. This will involve problems of a photon electron match and sometimes hot spots. With the Para Mixed technique, a compromise is

105 Chapter 6 reached with respect to both the problems of the photon-electron match and the dose delivered to the normal tissues internally. In many other planning studies, oblique electron beams were used. In the present study anterior electron beams were used, and an overlap with the tangential photon beams, resulting in a local hot spot, was accepted. To quantify the dimensions of the hot spot with the Para Mixed technique, the local volumes receiving 107%, 130%, and 140% of the prescribed dose were determined for 6 typical patients (breast = 3, chest wall = 3). The volumes were 43 cm 3, 15 cm 3, and 8 cm 3 for the breast cases and 46 cm 3, 8 cm 3, and 3 cm 3 for the chest wall cases, respectively. The average length of the overlap area was 12.0 cm for the breast cases and 15.3 cm for the chest wall cases. This illustrates that the maximum dose delivered to the overlap area (Table 2) is confined to a relatively small volume. Although side effects have not been specifically scored, a few cases of acute erythema, but no severe detrimental effects, have yet been observed in the group of patients treated with the Para Mixed technique in our institution since We investigated the behaviour of anterior and oblique electron beams at our institution. It seemed that with an anterior electron beam, the deepest points within the IMN PTV could be reached with the use of lower electron energies. This resulted in a lower heart dose, lung dose, and skin dose. The target coverage within the IMN PTV was more adequate with the anterior electron beam. Even in rather slim patients, oblique electron beams failed to yield adequate target coverage. This is consistent with reports from others [15,20,35]. When oblique electron beams are used, there is a greater risk of under-dosage to the breast or chest wall PTV at the place where the photon and electron beams are matched, due to position uncertainties and patient motion during treatment. To maintain adequate target coverage with all techniques and for all patients, we used anterior electron beams. With the Para Mixed technique, the electron beam is integrated in the treatment plan. In this way apertures conformal to the PTV can be created in beam s-eyeview. Photons and electrons can be mixed to achieve an optimal sparing of normal tissue while adequate target coverage is maintained. The electron beam model implemented in our treatment-planning system, in conjunction with the pencil

106 3D-CRT for loco-regional breast cancer irradiation beam algorithm used, has been compared with measurements in water at our institution and with dose distributions in patient CT, calculated by a state-of-theart algorithm [36]. It proved to be accurate with respect to the parameters reported in this study, beam output, depth of 90% electron isodose, and depth of dose profiles, as well as the maximum overall dose in the treatment plan. Its poor penumbra modelling had no influence on these parameters. Although electrons have valuable dose-limiting characteristics, their role is still limited in 3D-CRT. In this study, we demonstrated that the integration of electrons in 3D-CRT can further improve existing techniques and should therefore be the subject of future research. Application of better electron dose calculation algorithms provides a valuable improvement for this purpose [36]. Conclusions Normal tissue dose and especially heart dose can outweigh the benefits of IMN irradiation. The Para Mixed technique is an improved 3D-CRT technique in which the IMN are treated by a mix of photons and electrons. In this study, we compared the dose to the heart and other organs at risk with the Para Mixed technique with that with three other commonly applied radiotherapy strategies. All techniques were optimised and produced equally adequate target coverage. The Para Mixed technique yielded the overall best results. No other technique resulted in a lower heart dose. Lung and skin were equally spared instead of one of them being compromised, and the contralateral breast was avoided

107 Chapter 6 References 1. Recht A, Edge SB, Solin LJ, et al. Postmastectomy radiotherapy: Clinical practice guidelines of the American Society of Clinical Oncology. J Clin Oncol 2001;19: Van den Bogaert W, Struikmans H, Fourquet A, et al. Internal mammary and medial supraclavicular (IM-MS) lymph node chain irradiation in stage I-III breast cancer. A phase III randomised trial of the EORTC, protocol , May Brussels, Belgium: European Organization for Research and Treatment of Cancer; Kaija H, Maunu P. Tangential breast irradiation with or without internal mammary chain irradiation: Results of a randomized trial. Radiother Oncol 1995;36: Le MG, Arriagada R, de Vathaire F, et al. Can internal mammary chain treatment decrease the risk of death for patients with medial breast cancers and positive axillary lymph nodes? Cancer 1990;66: Yamashita T, Hurukawa M, Sekiguchi K, et al. Efficacy of loco-regional lymphnodes irradiation after mastectomy for breast cancer with biopsy proven parasternal lymphnodes metastases. A randomized study [Abstract]. Int J Radiat Oncol Biol Phys 1996;36: Cuzick J, Stewart H, Rutqvist L, et al. Cause-specific mortality in long-term survivors of breast cancer who participated in trials of radiotherapy. J Clin Oncol 1994;12: Rutqvist LE, Lax I, Fornander T, et al. Cardiovascular mortality in a randomized trial of adjuvant radiation therapy versus surgery alone in primary breast cancer. Int J Radiat Oncol Biol Phys 1992;22: Adams MJ, Hardenbergh PH, Constine LS, et al. Radiationassociated cardiovascular disease. Crit Rev Oncol Hematol 2003;45: Gagliardi G, Lax I, Ottolenghi A, et al. Long-term cardiac mortality after radiotherapy of breast cancer-application of the relative seriality model. Br J Radiol 1996;69: Lievens Y, Van den Bogaert W. Internal mammary and medial supraclavicular lymph node irradiation: The thin line between advantages and side effects. Radiother Oncol 2002;65: Taghian A, Jagsi R, Makris A, et al. Results of a survey regarding irradiation of internal mammary chain in patients with breast cancer: Practice is culture driven rather than evidence based. Int J Radiat Oncol Biol Phys 2004;60:

108 3D-CRT for loco-regional breast cancer irradiation 12. Marks LB, Hebert ME, Bentel G, et al. To treat or not to treat the internal mammary nodes: A possible compromise. Int J Radiat Oncol Biol Phys 1994;29: Jansson T, Lindman H, Nygard K, et al. Radiotherapy of breast cancer after breastconserving surgery: An improved technique using mixed electron-photon beams with a multileaf collimator. Radiother Oncol 1998;46: Hurkmans CW, Borger JH, Bos LJ, et al. Cardiac and lung complication probabilities after breast cancer irradiation. Radiother Oncol 2000;55: Hurkmans CW, Saarnak AE, Pieters BR, et al. An improved technique for breast cancer irradiation including the locoregional lymph nodes. Int J Radiat Oncol Biol Phys 2000;47: Severin D, Connors S, Thompson H, et al. Breast radiotherapy with inclusion of internal mammary nodes: A comparison of techniques with three-dimensional planning. Int J Radiat Oncol Biol Phys 2003;55: Arthur DW, Arnfield MR, Warwicke LA, et al. Internal mammary node coverage: An investigation of presently accepted techniques. Int J Radiat Oncol Biol Phys 2000;48: Pierce LJ, Butler JB, Martel MK, et al. Postmastectomy radiotherapy of the chest wall: Dosimetric comparison of common techniques. Int J Radiat Oncol Biol Phys 2002;52: Johansson J, Isacsson U, Lindman H, et al. Node-positive left-sided breast cancer patients after breast-conserving surgery: Potential outcomes of radiotherapy modalities and techniques. Radiother Oncol 2002;65: Cho BC, Hurkmans CW, Damen EM, et al. Intensity modulated versus non-intensity modulated radiotherapy in the treatment of the left breast and upper internal mammary lymph node chain: A comparative planning study. Radiother Oncol 2002;62: van der Laan HP, van t Veld AA, Bijl HP, et al. Treating the internal mammary nodes with the Para Mixed Technique limits dose to the heart and other organs at risk [Abstract]. Radiother Oncol 2004;73:S International Commission on Radiation Units and Measurements (ICRU). Prescribing, recording, and reporting photon beam therapy. ICRU Report 50. Bethesda, MD: ICRU Publications; Reinders JG, Heijmen BJ, Olofsen-van Acht MJ, et al. Ischemic heart disease after mantlefield irradiation for Hodgkin s disease in long-term follow-up. Radiother Oncol 1999;51:

109 Chapter Thilmann C, Zabel A, Milker-Zabel S, et al. Number and orientation of beams in inversely planned intensity-modulated radiotherapy of the female breast and the parasternal lymph nodes. Am J Clin Oncol 2003;26:e136-e Venables K, Miles EA, Deighton A, et al. Irradiation of the heart during tangential breast treatment: A study within the START trial. Br J Radiol 2004;77: Seppenwoolde Y, Lebesque JV. Partial irradiation of the lung. Semin Radiat Oncol 2001;11: Martel MK, Ten Haken RK, Hazuka MB, et al. Dose-volume histogram and 3-D treatment planning evaluation of patients with pneumonitis. Int J Radiat Oncol Biol Phys 1994;28: Carmichael A, Sami AS, Dixon JM. Breast cancer risk among the survivors of atomic bomb and patients exposed to therapeutic ionising radiation. Eur J Surg Oncol 2003;29: Johansen J, Overgaard J, Rose C, et al. Cosmetic outcome and breast morbidity in breast-conserving treatment-results from the Danish DBCG-82TM national randomized trial in breast cancer. Acta Oncol 2002;41: Huang EY, Chen HC, Wang CJ, et al. Predictive factors for skin telangiectasia following post-mastectomy electron beam irradiation. Br J Radiol 2002;75: Bentzen SM, Overgaard M. Relationship between early and late normal-tissue injury after postmastectomy radiotherapy. Radiother Oncol 1991;20: Kong FM, Klein EE, Bradley JD, et al. The impact of central lung distance, maximal heart distance, and radiation technique on the volumetric dose of the lung and heart for intact breast radiation. Int J Radiat Oncol Biol Phys 2002;54: Dijkema IM, Hofman P, Raaijmakers CP, et al. Loco-regional conformal radiotherapy of the breast: Delineation of the regional lymph node clinical target volumes in treatment position. Radiother Oncol 2004;71: Saarnak AE, Hurkmans CW, Pieters BR, et al. Accuracy of internal mammary lymph node localization using lymphoscintigraphy, sonography and CT. Radiother Oncol 2002;65: Woudstra E, van der Werf H. Obliquely incident electron beams for irradiation of the internal mammary lymph nodes. Radiother Oncol 1987;10: Cygler JE, Daskalov GM, Chan GH, et al. Evaluation of the first commercial Monte Carlo dose calculation engine for electron beam treatment planning. Med Phys 2004;31:

110 Chapter 7 Minimising contralateral breast dose in postmastectomy intensity-modulated radiotherapy by incorporating conformal electron irradiation Hans Paul van der Laan, Erik W. Korevaar, Wil V. Dolsma, John H. Maduro, Johannes A. Langendijk Accepted for publication in Radiotherapy & Oncology

111 Chapter 7 Abstract Purpose: To assess the potential benefit of incorporating conformal electron irradiation in intensity-modulated radiotherapy (IMRT) for loco-regional postmastectomy RT. Materials and Methods: Ten consecutive patients that underwent left-sided mastectomy were selected for this comparative planning study. Three-dimensional conformal radiotherapy (3D-CRT) photon-electron dose plans were compared to photon-only IMRT (IMRT p ) and photon IMRT with conformal electron irradiation (IMRT p/e ). The planning target volume (PTV) was prescribed 50 Gy and included the chest wall and the internal mammary and supra-clavicular lymph node regions. It was attempted to minimise dose delivered to heart, lungs and contralateral breast, while maintaining adequate PTV coverage. Results: All plans complied with objectives for PTV coverage. IMRT p/e eliminated volumes receiving 70 Gy (V70) that were present in 3D-CRT at the junction of photon and electron beams. Both IMRT strategies reduced heart V30 significantly below 3D-CRT levels. Mean heart dose with IMRT p/e was lowest and equal to that with 3D-CRT. Minimising heart dose with IMRT p resulted in irradiated contralateral breast volumes much larger than with 3D-CRT. With IMRT p/e, contralateral breast dose was only slightly increased when compared to 3D-CRT. Mean lung dose values were similar for IMRT and 3D-CRT. With IMRT, lung V20 was smaller, whereas V5 values for heart, lung and contralateral breast were higher than with 3D-CRT. Conclusion: Incorporation of conformal electron irradiation in postmastectomy IMRT p/e enables a heart dose reduction that in IMRT p can only be obtained by allowing large irradiated volumes in the contralateral breast

112 Electron irradiation in post-mastectomy IMRT Introduction The results of a number of randomised clinical trials revealed that post mastectomy radiotherapy (PMRT) significantly improves survival in high-risk breast cancer patients as compared to surgery alone [1-3]. PMRT often involves irradiation of regional lymph nodes, such as the internal mammary nodes (IMN) and supraclavicular nodes (SCN). As the involvement of these lymph node regions often results in larger irradiation fields and larger irradiated volumes, organs at risk (OAR) can receive a considerable radiation dose. Several authors demonstrated that higher cardiac dose in loco-regional PMRT resulted in an increased risk of late cardiac mortality [4-7]. The relatively high dose delivered to OARs might explain the relatively limited benefit of PMRT in terms of overall survival [8]. Ongoing research in the field of breast cancer radiotherapy resulted in the clinical introduction of new radiation delivery techniques that improved radiation dose coverage of the designated target volumes and, at the same time, enabled improved sparing of OARs [9,10]. These techniques include: three-dimensional conformal RT (3D-CRT) [11]; intensity modulated RT (IMRT) [12]; and the use of breath-hold techniques [13,14]. Although 3D-CRT and IMRT have shown to reduce the heart dose, both techniques also have some limitations. For instance, 3D-CRT with combined photon and electron irradiation may limit the dose to the heart and the contralateral breast, but it often results in hot-spots at the junction between photon and electron beams [11,15,16]. Similarly, 3D-CRT with wide tangential photon beams and multi-beam photon IMRT have shown to increase dose delivered to the contralateral breast [17,18]. The latter is of particular concern in the light of recent reports on the increased incidence of second primary malignancies in the contralateral breast following breast cancer radiotherapy, that may be dose dependent [19-21]. From this point of view, combining the benefits of electron irradiation and photon IMRT may further improve the therapeutic ratio. The use of conformal electron irradiation already proved beneficial in 3D-CRT with regard to sparing the heart and contralateral breast [11], and comparative dose planning studies have

113 Chapter 7 suggested a potential benefit of modulated and fixed electron beams incorporated in photon IMRT [22,23]. Since manufactures of treatment planning software provided the use of Monte Carlo electron dose calculation, the introduction of combining conformal electron irradiation with post-mastectomy IMRT has now become clinically feasible. This might be of particular benefit for patients that receive extensive loco-regional treatment, which often requires larger treatment fields and often results in higher doses in OARs. The main objective of this planning comparative study was to test the hypothesis that photon IMRT combined with conformal electron irradiation results in a significant reduction of the radiation dose to the contralateral breast as compared to photon IMRT alone. In this study, our current standard, i.e., photon 3D-CRT combined with electrons [11], was used as a reference for both new techniques. Materials and Methods Patients and CT acquisition The study population was composed of 10 consecutive patients who underwent left-sided mastectomy. A computed tomography (CT) scan was made in treatment position (supine) for the purpose of treatment planning with a slice thickness and index of 3 mm in all patients. CT-data were transferred to the Pinnacle 3 treatment planning system (TPS) incorporating Monte Carlo electron dose calculation (research version 8.1y, Philips Radiation Oncology Systems, Fitchburg WI, USA). Regions of interest The clinical target volume (CTV) consisted of the chest wall, the IMN and the SCN. The chest wall CTV extended from the ipsilateral edge of the sternum medially to the mid-axillary line laterally. It excluded the anterior rib surface and the pectoral muscle. The IMN CTV was defined from the inferior aspect of the clavicular head through the fourth inter-costal space by an elliptical cylinder with a

114 Electron irradiation in post-mastectomy IMRT lateral and anterior-posterior diameter of 15 and 10 mm, respectively. It was placed adjacent to the left edge of the sternum and ventral to the pleura. The SCN CTV included the supraclavicular lymph nodes regions and extended in the cranial direction until the level of the cricoid. The planning target volumes (PTVs) were obtained for each CTV by applying a 5 mm margin in three dimensions. The SCN and chest wall PTVs were restricted to 5 mm and 2 mm under the skin surface, respectively, to exclude the build up region from the PTVs (additional bolus was applied on the chest wall only). The heart was contoured until the level of the pulmonary trunk superiorly, excluding the major vessels, including the pericardium. Both lungs were contoured using the automatic contouring tool of the TPS and were edited manually when necessary. The contralateral breast was contoured excluding the skin and pectoral muscles. Regions of interest were defined in addition to the CTVs and OARs for the purpose of IMRT optimisation (Fig. 1). Photon-electron 3D-CRT treatment planning Treatment planning procedures for photon-electron 3D-CRT have been described previously in more detail [11]. Briefly, a total dose of 50 Gy was prescribed to each PTV in 25 daily fractions of 2 Gy. A matchline was determined at the level of inferior aspect of the clavicular head. Superior to the matchline, the SCN PTV was treated with conformal anterior-posterior and posterior-anterior photon beams, delivering roughly 75% and 25% of the prescribed dose, respectively. Inferior to the matchline, the major part of the chest wall PTV was treated with shallow tangential photon beams. These excluded the medial part of the chest wall PTV and the IMN PTV. A conformal medio-lateral beam, was used to deliver approximately 0.75 Gy per fraction to the entire chest wall PTV and the IMN PTV. All photon beams had a common isocenter that was placed on the matchline. An anterior-posterior electron beam was added to complete the remaining 1.25 Gy to the medial part of the chest wall PTV and the IMN PTV. The medial border of the electron beam was shaped by means of a lead cut-out conformal to the medial edge of the PTV, whereas the ipsilateral border of the

115 Chapter 7 electron beam slightly overlapped the shallow tangential photon beams. The ipsilateral border of the electron beam was shaped in such a way that the size of the overlap area was minimal and adequate PTV coverage was maintained. Figure 1. Structures created for intensity modulated radiotherapy optimisation Structures were created for direct aperture optimisation-based intensity modulated radiotherapy (IMRT): structure A) extending from the planning target volumes (PTVs) in the medio-dorsal direction with a margin of 3 cm; structure B) extending 4 mm inside the posterior part of the PTVs; and structure C) including the tissue ventral to the chest wall PTV and ~3 cm of air to provide skin flash in IMRT beams. A bolus of 5 mm (grey) was used in each photon beam inferior to the matchline. During optimisation, objective values and weights were optimised for the various structures, including heart (maroon), lungs (teal), contralateral breast (green), chest wall PTV (red), internal mammary nodes PTV (blue) and supraclavicular nodes PTV (not shown)

116 Electron irradiation in post-mastectomy IMRT Electron energy (8-14 MeV) was determined by the depth of the IMN, that did not exceed 5 cm in the patients included in this study. Wedges were used in the photon beams and a maximum total of three small manually shaped photon beams were added, when necessary, to obtain a homogeneous dose distribution. Wedge fractions and relative weights of photon beams were weighted manually, i.e., by forward planning. The plan was optimised in such a way that 98% of the chest wall PTV and the SCN PTV received 95% of the prescribed dose and 98% of the IMN PTV received 90% of the prescribed dose (a slightly lower dose was accepted in the deepest parts of the IMN PTV because with the use of anterior-posterior electron irradiation, a PTV margin of 5 mm for position inaccuracies in the dorsal direction is debatable and may unnecessarily increase the dose to the heart). It was attempted to minimise the dose delivered to heart, lungs and contralateral breast by effective use of multileaf collimator shielding and careful selection of gantry angles. A skin flash of ~3 cm was used in photon beams wherever appropriate and a bolus of 5 mm, covering the patient s skin inferior to the matchline, was applied in all (3D-CRT and IMRT) photon beams. IMRT treatment planning Two direct aperture optimisation (DAO)-based IMRT treatment plans were constructed for each patient: one photon-only IMRT plan (IMRT p ) and one with an electron beam identical to that used in the 3D-CRT plan (IMRT p/e ). When used, the contribution of the electron beam was fixed in the optimisation process at a dose of 1.25 Gy. In each IMRT plan, 9 different photon beam directions were used: 300, 330, 0, 30, 60, 90, 120, 150 and 180. The actual gantry angles of the 330 and 150 beams were slightly adjusted for each patient in such a way that these were similar to that of the shallow tangential beams of the 3D-CRT plan. The photon beam isocenter of the IMRT plans was always similar to that of the 3D-CRT plan. DAO-settings were the same for each IMRT plan: segment size 6 cm 2 ; minimum leaf separation 2 cm; number of leaf pairs 3; monitor units per segment 4; and a maximum total of 40 segments for each plan. The IMRT optimisation process was performed on the basis of a series of structures that were

117 Chapter 7 created in addition to the OARs and PTVs (Fig. 1). Objectives and weights were entered for each volume of interest. By a trial-and-error adaptive adjustment of the objectives, the doses to both lungs, heart and contralateral breast were reduced as much as possible, while attempting to preserve target coverage similar to that as achieved by the 3D-CRT plan. Although it was attempted to minimise the contralateral breast dose at all times, reducing the mean heart dose and heart V30 to levels similar or below that with 3D-CRT plan had priority. In the IMRT plans, no dose 60 Gy was accepted, with the exception of the area of overlap between electrons and photons in the IMRT p/e plan, where no dose 70 Gy was accepted. Target coverage, dose homogeneity and OAR dose Target coverage and target dose homogeneity were determined for all plans by evaluating the proportions of chest wall PTV and SCN PTV receiving 95% and 107% of the prescribed dose and the proportions of IMN PTV receiving 90% of the prescribed dose. In addition, absolute volumes receiving 60 Gy and 70 Gy, and the general maximum dose were determined for all plans. Heart dose was determined by calculating the proportion of the heart receiving 5 Gy (V5), 30 Gy (V30) and the mean heart dose. Lung dose (both lungs combined as one organ) was determined by calculating the proportion of lung receiving 5 Gy (V5), 20 Gy (V20), and the mean lung dose. Similarly, we determined the contralateral breast V0.05, V0.6, V1, V2 and V5 and the contralateral breast mean dose, as some of these parameters were recently proposed as maximum dose objectives in treatment planning to limit the risk of second breast cancer [19-21]. Statistics To compare the various dose-volume parameters, the mean values were analysed with the Wilcoxon signed ranks test or the paired-samples t-test on statistical significance whenever appropriate. All tests were two-tailed, and differences were considered statistically significant at p

118 Electron irradiation in post-mastectomy IMRT Table 1. Dose-volume results with 3D-CRTp/e, IMRTp and IMRTp/e 3D-CRT p/e vs. IMRT p 3D-CRT p/e vs. IMRT p/e IMRT p vs. IMRT p/e 3D-CRT p/e IMRT p IMRT p/e p-value p-value p-value Target Coverage PTV receiving 95% ( 47.5 Gy) CW PTV (%) 98.7 (0.3) 98.5 (0.2) 98.8 (0.3) SCN PTV (%) 98.8 (0.5) 99.3 (0.3) 99.2 (0.3) PTV receiving 90% ( 45 Gy) IMN PTV (%) 98.5 (0.4) 98.5 (0.4) 98.6 (0.5) Irradiated volumes PTV receiving 107% ( 53.5 Gy) CW PTV (%) 19.0 (6.3) 25.3 (8.7) 27.5 (10.0) SCN PTV (%) 18.3 (8.7) 4.3 (4.3) 3.3 (2.1) Volume receiving 60 Gy (cm 3 ) 20.3 (7.2) 0.2 (0.3) 5.5 (3.8) Volume receiving 70 Gy (cm 3 ) 1.8 (2.0) 0.0 (0.0) 0.0 (0.0) Integral mean dose (Gy) 9.9 (0.9) 11.1 (1.1) 10.7 (1.1) Integral max dose (Gy) 74.3 (3.6) 60.3 (1.6) 67.0 (2.4) Heart Mean dose (Gy) 8.9 (1.8) 10.1 (1.8) 8.9 (2.0) V30 (%) 8.2 (4.0) 5.3 (2.8) 5.2 (3.0) V5 (%) 36.5 (6.1) 77.4 (12.6) 54.2 (18.0) Lungs Mean dose (Gy) 10.3 (1.7) 10.2 (1.3) 9.9 (1.7) V20 (%) 21.3 (4.4) 15.9 (3.3) 17.0 (3.9) V5 (%) 33.5 (6.0) 45.4 (5.5) 42.5 (6.3) Contralateral breast Mean dose (Gy) 0.8 (0.2) 2.7 (1.1) 1.3 (0.3) V5 (%) 0.3 (0.5) 8.9 (8.0) 1.0 (1.0) V2 (%) 8.3 (4.2) 47.6 (18.0) 18.3 (7.8) V1 (%) 29.1 (8.5) 73.7 (19.6) 53.3 (13.2) V0.6 (%) 49.0 (10.7) 88.6 (14.9) 76.9 (13.9) V0.05 (%) 95.0 (1.7) 99.5 (1.0) 99.0 (1.6) Abbreviations: 3D-CRTp/e = photon-electron three-dimensional conformal radiotherapy; IMRTp/e = photon-electron intensity modulated RT; IMRTp = photon-only IMRT; IMN = internal mammary nodes; CW = chest wall; SCN = supraclavicular nodes; PTV = planning target volume; Vx = proportion of organ at risk receiving x Gy. Data presented as mean values with standard deviation in parenthesis. Results Target coverage and target dose homogeneity Each dose plan complied with the minimum objectives for target coverage (Table 1). Dose levels 70 Gy in a mean volume of 1.8 cm 3 were found in the 3D- CRT plans at the level of the overlap between photon and electron beams. With IMRT p/e, no such high dose values were found, and V60 was significantly reduced

119 Chapter 7 when compared to 3D-CRT. With both IMRT p and IMRT p/e, doses up to 60 Gy were accepted in the chest wall PTV in an attempt to minimise the dose delivered to OARs [24,25]. This may also explain why the proportion of chest wall PTV receiving 107% of the prescribed dose with IMRT was roughly comparable to that with 3D-CRT. While with 3D-CRT, the chest wall PTV volumes receiving 107% were located primarily in the vicinity of the beam junctions, with IMRT they were found scattered in the chest wall PTV (Fig. 2). The proportions of SCN PTV receiving 107% of the prescribed dose were largest with 3D-CRT because, in contrast to IMRT, only two 3D-CRT beams were used superior to the matchline. Figure 2. Axial representation of dose distributions with three techniques Axial representation of dose distributions with photon-electron three-dimensional conformal radiotherapy (3D-CRTp/e), photon-electron intensity modulated radiotherapy (IMRTp/e) and photononly IMRTp