Postmenopausal bleeding: Studies on the diagnostic work-up Anne Timmermans. Anne Timmermans

Transcription

1 Postmenopausal bleeding: Studies on the diagnostic work-up Anne Timmermans Postmenopausal bleeding: Studies on the diagnostic work-up Anne Timmermans

2 Postmenopausal bleeding: studies on the diagnostic work-up

3 Financial support for printing of this thesis was provided by: Vrouwenkliniek AMC Wetenschapsfonds TweeSteden ziekenhuis Tilburg Vakgroep gynaecologie TweeSteden ziekenhuis Tilburg Johnson & Johnson Medical BV Sigma Medical BV Erbe Benelux BV Olympus Nederland BV Cook Medical SWOAHS Medical Dynamics Postmenopausal bleeding: studies on the diagnostic work-up Timmermans, Anne Thesis University of Amsterdam. With a summary in Dutch ISBN Cover design: Angela Gasseling Lay-out & Print: Gildeprint Drukkerijen B.V. Enschede, The Netherlands A. Timmermans, Utrecht, 2009 All rights reserved. No part of this thesis may be reproduced or transmitted in any form by any means, without permission of the copyright owner.

4 Postmenopausal bleeding: studies on the diagnostic work-up academisch proefschrift ter verkrijging van de graad van doctor aan de Universiteit van Amsterdam, op gezag van de Rector Magnificus, prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Agnietenkapel op vrijdag 13 februari 2009, te 10:00 uur door Anne Timmermans geboren te Nijmegen

5 Promotiecommissie Promotor: Prof. dr. B.W.J. Mol Co-promotores: Dr. B.C. Opmeer Dr. F.W. Jansen Overige leden: Prof. dr. M.J. Heineman Prof. dr. M.P.M. Burger Prof. dr. E. Schadé Prof. dr. H.A.M. Brölmann Prof. dr. M.E. Vierhout Dr. J.B. Reitsma Dr. T.J. Clark Faculteit der Geneeskunde

6 Voor mijn ouders, pa mi madrina en oma Limburg

7

8 Contents Chapter 1 11 General introduction Chapter 2 21 Patients preferences in the evaluation of postmenopausal bleeding (BJOG 2007; 114: ) Chapter 3 31 Meta-analysis of endometrial thickness measurement for detecting endometrial cancer in women with postmenopausal bleeding: a different approach using individual patient data (submitted) Chapter 4 47 Office hysteroscopy in women with postmenopausal bleeding: see and treat of endometrial polyps using a Duckbill polypsnare (Gyn Surg 2004; 1: ) Chapter 5 53 The diagnostic accuracy of endometrial thickness to exclude polyps in women with postmenopausal bleeding (J Clin Ultrasound 2008; 36: ) Chapter 6 63 Hysteroscopy and removal of endometrial polyps: a Dutch survey (Eur J Obstet Gynecol Reprod Biol 2008; 138: 76-9) Chapter 7 73 What is the recurrence rate of postmenopausal bleeding in women who have a thin endometrium during a first episode of postmenopausal bleeding? (Acta Obstet Gynecol Scand 2008; 87: 89-93)

9 Chapter 8 85 Follow-up of women after a first episode of postmenopausal bleeding and endometrial thickness greater than 4 millimeters (Obstet Gynecol 2008; 111: ) Chapter 9 99 Should endometrial polyps be removed in women with postmenopausal bleeding? An assessment of study designs and report of a failed trial (submitted) Chapter Summary, general discussion and recommendations Nederlandse samenvatting 129 List of co-authors and their affiliations 139 Dankwoord 143 About the author 149 List of publications 153

10

11

12 1 General introduction

13 Abnormal postmenopausal bleeding (PMB) can be defined as uterine bleeding occurring at least one year after menopause. PMB is a common clinical problem in both general and hospital settings. 1,2 The incidence of spontaneously occurring PMB in the general population can be as high as 10% immediately after menopause. 3 Moreover, in 10-15% of the women with PMB endometrial carcinoma is diagnosed. 4,5 Endometrial cancer presents itself with PMB in more than 95% of women. 6 The incidence of PMB decreases with age, while the incidence of endometrial carcinoma increases. 7 Furthermore, the probability of endometrial carcinoma in women with PMB rises from less than 1% in women younger than 50 years to 23.8% in women older than 80 years. 7 Although patients with PMB have an increased risk for endometrial carcinoma, the majority of the patients will have non-malignant endometrium, atrophy or benign pathology. A frequent finding in women with PMB are benign endometrial polyps, with a prevalence ranging between 20 to 40%. 4,5,8 Chapter 1 12 It is due to this risk of endometrial carcinoma that diagnostic assessment of women with PMB focuses on the exclusion of (pre)malignant disease. Since two decades transvaginal ultrasonography (TVU) has become in use widely to evaluate the endometrium in women with PMB. Prior to the introduction of TVU, women with PMB were scheduled for dilatation and curettage (D&C). The goal of TVU assessment of the endometrium is to exclude endometrial pathology. Different guidelines in various countries advise measurement of endometrial thickness by TVU as a first step in the evaluation of women with PMB These guidelines are based on various studies summarized in the meta-analysis by Smith-Bindman et al. 14 The meta-analysis by Smith-Bindman et al. 14 was the first one published on this subject. Several years after that two other meta-analyses were published. 15,16 Although three meta-analyses on this subject were published, guidelines have mostly implemented the meta-analysis of Smith-Bindman et al. 14 Furthermore, the meta-analysis by Smith-Bindman et al. 14 is the most cited publication (total citation 169 by Web of Science) compared to the other two meta-analyses (both total citation 37). 15,16 The meta-analysis of Smith-Bindman et al. 14 used traditional methods of combining published data from different studies. Using the reported data, 2x2 tables per included study were constructed that compared endometrial thickness measured at TVU to presence or absence of endometrial carcinoma. Results across studies were combined in a summary Receiver Operator Characteristics (ROC) Curve. At a 5 mm cut-off the sensitivity for detecting endometrial cancer was 96% for a 39% false-positive rate. Such a combination of sensitivity and specificity

14 would reduce a pre-test probability of 10% for endometrial cancer to a post-test probability of 1%. Based on this post-test probability, expectant management is at present recommended to these women. In contrast, the two other lesser quoted meta-analyses came to different conclusions. 15,16 Gupta et al. 15 conducted a comprehensive systematic review in which they focused on the study quality assessment of each study. Only four studies were identified as best-quality studies For each article a 2x2 table was constructed and likelihood ratios (LR) were calculated. Pooling of the results of these four studies for endometrial thickness 5 mm resulted in a LR of a negative test of Such a LR would implicate that in a patient with a negative test result a pre-test probability of 10% would change to a post-test probability of 2.5%. Tabor et al. 16 included only studies from which they were able to get the original data from the authors. For each study they calculated median endometrial thickness per center and used multiples of the median for endometrial thickness to pool data. They reported a sensitivity of 96% for a specificity of 50% and concluded that such a sensitivity with a 4% false-negative rate was too high. Therefore, in their opinion endometrial thickness measurement does not reduce the need for invasive diagnostic testing. Based on the meta-analysis by Smith-Bindman et al. 14 the Dutch guideline on PMB advocates expectant management once malignancy has been excluded by TVU (i.e. endometrial thickness 4 mm). 9 Adherence to this guideline has been shown to be fairly good: 2/3 of women presenting with PMB are managed conform this guideline. 21 This adherence study demonstrates that in the Netherlands expectant management in women with PMB after exclusion of malignancy by TVU is at present generally accepted. 21 General introduction 13 Although guidelines leave room for individual choices, the preference of the patients has never been systematically taken into account when these guidelines were made. The question is to what extent women are willing to accept a higher probability of cancer (> 1%) associated with expectant management in case of thin endometrium, in order to avoid further invasive procedures. Although current guidelines presume that women consider this small risk as acceptable, empirical data on this presumption are lacking. The usefulness of endometrial thickness measurement in women with PMB might thus be overestimated. Not only is the preference of the patient unclear, but it is also questionable what the diagnostic accuracy of endometrial thickness measurement is. Although the diagnostic assessment of women with PMB focuses on exclusion of (pre)malignant disease, further diagnostic assessment may centre on benign intrauterine pathology

15 (i.e. endometrial polyps). This diagnostic assessment can be undertaken at the first episode of PMB, after exclusion of malignancy or after expectant management in case the bleeding persists or recurs Since endometrial polyps in women with PMB have a prevalence ranging between 20% and 40%, diagnostic assessment of benign disease aims to detect this pathology. Saline Infusion Sonography (SIS) as well as hysteroscopy are highly accurate in diagnosing focal endometrial pathology with comparable diagnostic accuracy with SIS. 22,23 Hysteroscopy has the advantage of allowing simultaneous removal of intracavitary pathology at the time of diagnosis. The Dutch guideline on PMB first became available in ,24 The revised version differed from the first version only with respect to the recommendation of performing a SIS or an office hysteroscopy in women with endometrial thickness > 4 mm. 9 In hospitals without access to these ambulatory procedures further uterine cavity evaluation is not advised. Chapter 1 14 Traditionally hysteroscopy was performed under general anaesthesia. With the development of smaller diameter hysteroscopic systems and the introduction of a vaginoscopic approach to hysteroscopy, patient compliance has improved considerably and hysteroscopy nowadays can be performed in an outpatient setting without the use of anaesthesia. 25 New developments in hysteroscopic armentarium and the smaller diameter hysteroscopic systems ( mm diameter) have made outpatient hysteroscopic surgery possible. 26 Large series of outpatient hysteroscopic polypectomy have been described. 26,27 Furthermore outpatient hysteroscopic polyp removal has been reported to be better than its traditional inpatient counterpart with respect to cost-effectiveness, and patient s satisfaction rates In contrast, outpatient hysteroscopic polyp removal does not seem to be widely available. 30 The recommendation to perform SIS or office hysteroscopy to detect benign intrauterine pathology implicitly suggests that treatment of such pathology will be beneficiary to the patient. A benefit could be that patients experience less recurrent PMB if benign intrauterine pathology is diagnosed and treated at first work-up. Another possible benefit could be that hysteroscopy detects more patients with malignancy than office endometrial sampling. At present, hysteroscopy or SIS can be incorporated in the diagnostic work-up at first episode of bleeding or in case the bleeding persists or recurs Few studies report on the incidence of recurrent PMB, with incidences varying between 6% to 40%. 12,32-35 None of these studies related initial diagnostic work-up to the incidence of recurrent PMB. This issue raises several questions. First, it can be questioned what the recurrence rate of bleeding after a first episode of PMB is. If the recurrence rate would be low, then a policy with TVU and office endometrial sampling would be sufficient. Second, it is questionable if the recurrence rate of PMB depends on the performance and outcome of the initial diagnostic work-up, i.e. whether women with a hysteroscopy at the initial work-up experience less recurrent bleeding than patients without hysteroscopy. Third, is malignancy present in women with recurrent bleeding? If, the recurrence rate would be high, or if malignancy would be diagnosed at follow-up, one could advocate that hyster-

16 oscopy should be applied immediately. Furthermore, if it would be possible to identify those patients at increased risk of recurrent bleeding, one could apply a policy based on individual patient characteristics. One should keep in mind how diagnostic tests affect patient outcome. Diagnostic hysteroscopy itself does not influence patient outcome. A woman suffering from PMB, in whom hysteroscopy does not show any abnormalities, does not have a decreased probability of recurrent bleeding after hysteroscopy. In contrast, the woman in whom a polyp is diagnosed at hysteroscopy might benefit from the hysteroscopy, but only if removal of the polyp reduces the probability of recurrent bleeding. The majority of gynaecologists advocates removal of endometrial polyps. 36 It is considered that the removal of endometrial polyps reduces the probability of recurrent bleeding with reported success rates varying between 40% to 100%. However, there still is a lack of highquality evidence regarding the efficacy of intrauterine polypectomy with respect to abnormal uterine bleeding. 37 Endometrial polyps are highly prevalent in women with PMB 8 and, if causative, may be responsible for significant morbidity. Alternatively we may be subjecting women to unnecessary interventions, risks and wasting valuable health care resources. No studies exist that included a control group when reporting on the efficacy of polypectomy. From this point of view it can be questioned if the diagnosis of endometrial polyps is important and randomized controlled trials on this subject are advocated. 37 Therefore, the recommendation to perform SIS or office hysteroscopy to diagnose benign intrauterine pathology (i.e. endometrial polyps) is not supported by current literature. The question that remains to be answered is whether endometrial polyps in patients with PMB should be removed when malignancy has been excluded, in view of recurrent bleeding symptoms. On the one hand, this will depend on the prevalence of benign endometrial polyps in patients with PMB, and whether such polyps can be diagnosed accurately, for example with TVU or with SIS. On the other hand, this depends on whether removal of a benign endometrial polyp improves outcome. General introduction 15

17 Aims of the thesis The aim of this thesis is to evaluate the current diagnostic work-up in women with PMB. Outline of the thesis Chapter 2 studies the preference of patients for diagnostic management of PMB. Women were asked to make a trade-off between expectant management after transvaginal ultrasound or complete diagnostic work-up including invasive diagnostic procedures, i.e. hysteroscopy. This study was performed among women with PMB who had had a hysteroscopy in the diagnostic work-up. Chapter 3 determines the diagnostic accuracy of endometrial thickness measurement in the detection of endometrial cancer among women with PMB with individual patient data using different meta-analytic strategies. Chapter 4 describes the prevalence of endometrial polyps at hysteroscopy in women with PMB. In addition, it describes the possibility of diagnosis and treatment of these polyps in the same outpatient hysteroscopy session. Chapter 1 16 Chapter 5 determines the diagnostic accuracy of endometrial thickness measurement with transvaginal ultrasound for detection of endometrial polyps in women with PMB. Data on endometrial thickness measurement with transvaginal ultrasound and findings at hysteroscopy were combined to perform a receiver operator characteristics (ROC) analysis to assess the discriminative capacity of endometrial thickness for the presence of endometrial polyps. Chapter 6 describes current practice of Dutch gynaecologists in polyp removal. This was surveyed by a mailed self-administered questionnaire regarding polyp removal. Gynaecologists were asked about their individual performance of polypectomy: setting, form of anaesthesia, method and instrument used. Chapter 7 determines the incidence and significance of recurrent PMB among women diagnosed with an endometrial thickness 4 mm after a first episode of PMB. Patients who presented with PMB were registered prospectively in a multi-center study between January 2001 and June Diagnostic assessment of the patients followed the Dutch guideline on this subject. For women with endometrial thickness 4 mm we determined whether recurrent bleeding had occurred.

18 Chapter 8 estimates the incidence and significance of recurrent PMB among women diagnosed with an endometrial thickness > 4 mm after a first episode of PMB. Patients who presented with PMB were registered prospectively in a multi-center study between January 2001 and June Diagnostic assessment of the patients followed the Dutch guideline on this subject. For women with endometrial thickness > 4 mm we determined time to recurrent bleeding and diagnosis at recurrent bleeding. Furthermore, we evaluated if incorporation of hysteroscopy or polyp removal at initial work-up had any influence on recurrent bleeding. Chapter 9 evaluates the design of a randomized controlled trial in women with PMB and an endometrial polyp. We set up a prospective cohort study with an embedded randomized clinical trial comparing polypectomy and expectant management. We report this study design and its problems and propose a possible alternative study design. Chapter 10 summarizes the results of the studies presented in this thesis and gives clinical implications and implications for future research. General introduction 17

19 References Chapter Clark TJ, Barton PM, Coomarasamy A, Gupta JK, Khan KS. Investigating postmenopausal bleeding for endometrial cancer: cost-effectiveness of initial diagnostic strategies. BJOG 2006;113: Dijkhuizen FP. Diagnosis of Uterine Cavity Abnormalities. Studies on test performance and clinical value. Thesis University Utrecht 2000;ISBN Astrup K, Olivarius Nde F. Frequency of spontaneously occurring postmenopausal bleeding in the general population. Acta Obstet Gynecol Scand 2004;83: Emanuel MH, Verdel MJ, Wamsteker K, Lammes FB. An audit of true prevalence of intrauterine pathology: the hysteroscopic findings, controlled for patient selection in 1,202 patients with abnormal uterine bleeding. Gynaecol Endosc 1995;4: Dijkhuizen FP, Brolmann HA, Potters AE, Bongers MY, Heinz AP. The accuracy of transvaginal ultrasonography in the diagnosis of endometrial abnormalities. Obstet Gynecol 1996;87: Gredmark T, Kvint S, Havel G, Mattsson LA. Histopathological findings in women with postmenopausal bleeding. Br J Obstet Gynaecol 1995;102: Epstein E, Ramirez A, Skoog L, Valentin L. Dilatation and curettage fails to detect most focal lesions in the uterine cavity in women with postmenopausal bleeding. Acta Obstet Gynecol Scand 2001;80: NVOG (Dutch Society of Obstetrics and Gynaecology). NVOG-Richtlijn 4. Diagnostiek bij abnormaal vaginaal bloedverlies in de postmenopauze [in Dutch] Epstein E. Management of postmenopausal bleeding in Sweden: a need for increased use of hydrosonography and hysteroscopy. Acta Obstet Gynecol Scand 2004;83: Scottish Intercollegiate Guidelines Network. Investigation of postmenopausal bleeding. 1st edn edinburgh, UK: Scottisch Intercollegiate Guidelines Network, Royal College of Physicians 2002;( 12. Gull B, Karlsson B, Milsom I, Granberg S. Can ultrasound replace dilation and curettage? A longitudinal evaluation of postmenopausal bleeding and transvaginal sonographic measurement of the endometrium as predictors of endometrial cancer. Am J Obstet Gynecol 2003;188: Goldstein RB, Bree RL, Benson CB, et al. Evaluation of the woman with postmenopausal bleeding: Society of Radiologists in Ultrasound-Sponsored Consensus Conference statement. J Ultrasound Med 2001;20: Smith-Bindman R, Kerlikowske K, Feldstein VA, et al. Endovaginal ultrasound to exclude endometrial cancer and other endometrial abnormalities. JAMA 1998;280: Gupta JK, Chien PF, Voit D, Clark TJ, Khan KS. Ultrasonographic endometrial thickness for diagnosing endometrial pathology in women with postmenopausal bleeding: a meta-analysis. Acta Obstet Gynecol Scand 2002;81: Tabor A, Watt HC, Wald NJ. Endometrial thickness as a test for endometrial cancer in women with postmenopausal vaginal bleeding. Obstet Gynecol 2002;99:

20 17. Grigoriou O, Kalovidouros A, Papadias C, Antoniou G, Antonaki V, Giannikos L. Transvaginal sonography of the endometrium in women with postmenopausal bleeding. Maturitas 1996;23: Nasri MN, Coast GJ. Correlation of ultrasound findings and endometrial histopathology in postmenopausal women. Br J Obstet Gynaecol 1989;96: De Silva BY, Stewart K, Steven JD, Sathanandan M. Transvaginal ultrasound measurement of endometrial thickness and endometrial pipelle sampling as an alternative diagnostic procedure to hysteroscopy and dilatation and curettage in the management of post-menopausal bleeding. J Obstet Gynaecol 1997;17(4): Gupta JK, Wilson S, Desai P, Hau C. How should we investigate women with postmenopausal bleeding? Acta Obstet Gynecol Scand 1996;75: Werkgroep Dutch Study in Postmenopausal B. [Gynaecological diagnosis of postmenopausal women with abnormal vaginal bleeding: a comparison with the guideline] Diagnostiek door gynaecologen bij vrouwen met abnormaal vaginaal bloedverlies in de postmenopauze; vergelijking met de richtlijn [in Dutch]. Ned Tijdschr Geneeskd 2005;149: Van Dongen H, Kolkman W, Jansen FW. Implementation of hysteroscopic surgery in The Netherlands. Eur J Obstet Gynecol Reprod Biol 2007;132: de Kroon CD, de Bock GH, Dieben SW, Jansen FW. Saline contrast hysterosonography in abnormal uterine bleeding: a systematic review and meta-analysis. BJOG 2003;110: NVOG (Dutch Society of Obstetrics and Gynaecology). NVOG-Richtlijn 4. Abnormaal vaginaal bloedverlies in de postmenopauze [in Dutch] Bettocchi S, Selvaggi L. A vaginoscopic approach to reduce the pain of office hysteroscopy. J Am Assoc Gynecol Laparosc 1997;4: Bettocchi S, Ceci O, Di Venere R, et al. Advanced operative office hysteroscopy without anaesthesia: analysis of 501 cases treated with a 5 Fr. bipolar electrode. Hum Reprod 2002;17: Bettocchi S, Ceci O, Nappi L, et al. Operative office hysteroscopy without anesthesia: analysis of 4863 cases performed with mechanical instruments. J Am Assoc Gynecol Laparosc 2004;11: Kremer C, Duffy S, Moroney M. Patient satisfaction with outpatient hysteroscopy versus day case hysteroscopy: randomised controlled trial. BMJ 2000;320: Marsh FA, Rogerson LJ, Duffy SR. A randomised controlled trial comparing outpatient versus daycase endometrial polypectomy. BJOG 2006;113: Clark TJ, Godwin J, Khan KS, Gupta.J.K. Ambulatory endoscopic treatment fo symptomatic benign endometrial polyps: a feasibility study. Gynaecol Endosc 2002;11: Marsh F, Kremer C, Duffy S. Delivering an effective outpatient service in gynaecology. A randomised controlled trial analysing the cost of outpatient versus daycase hysteroscopy. BJOG 2004;111: Feldman S, Shapter A, Welch WR, Berkowitz RS. Two-year follow-up of 263 patients with post/perimenopausal vaginal bleeding and negative initial biopsy. Gynecol Oncol 1994;55:56-9. General introduction 19

21 33. Epstein E, Valentin L. Rebleeding and endometrial growth in women with postmenopausal bleeding and endometrial thickness < 5 mm managed by dilatation and curettage or ultrasound follow-up: a randomized controlled study. Ultrasound Obstet Gynecol 2001;18: Epstein E, Jamei B, Lindqvist PG. High risk of cervical pathology among women with postmenopausal bleeding and endometrium <or=4.4 mm: long-term follow-up results. Acta Obstet Gynecol Scand 2006;85: Gull B, Carlsson S, Karlsson B, Ylostalo P, Milsom I, Granberg S. Transvaginal ultrasonography of the endometrium in women with postmenopausal bleeding: is it always necessary to perform an endometrial biopsy? Am J Obstet Gynecol 2000;182: Clark TJ, Khan KS, Gupta JK. Current practice for the treatment of benign intrauterine polyps: a national questionnaire survey of consultant gynaecologists in UK. Eur J Obstet Gynecol Reprod Biol 2002;103: Nathani F, Clark TJ. Uterine polypectomy in the management of abnormal uterine bleeding: A systematic review. J Minim Invasive Gynecol 2006;13: Chapter 1 20

22 2 Patients preferences in the evaluation of postmenopausal bleeding A. Timmermans, B.C. Opmeer, S. Veersema, B.W.J. Mol (BJOG 2007; 114: )

23 Abstract Objective: To assess patients preferences for diagnostic management of postmenopausal bleeding (PMB). Methods: A structured interview was taken from 39 women who had had an office hysteroscopy in the diagnostic work-up for PMB. Women were informed about the probability of endometrial carcinoma versus benign disease and about advantages and disadvantages of different diagnostic strategies, i.e. expectant management after ultrasound or complete diagnostic work-up including invasive procedures. Women were asked to make a trade-off between different options. Results: Most women wanted to be 100% certain that carcinoma could be ruled out. Only 5% of the women were willing to accept more than 5% risk of false reassurance. If the risk of recurrent bleeding due to benign disease exceeded 25% the majority of women would prefer immediate diagnosis and treatment of benign lesions. Conclusion: Women with PMB are prepared to undergo hysteroscopy to rule out any risk on cancer. This finding implicates that the measurement of endometrial thickness with transvaginal ultrasound as a first line test in the assessment of PMB should be reconsidered. Chapter 2 22

24 Introduction Postmenopausal bleeding (PMB) is generally considered as a first symptom of endometrial cancer. Different authors have suggested starting the diagnostic work-up of women with PMB with transvaginal ultrasound (TVU) for the measurement of endometrial thickness. 1-3 Guidelines in various countries recommend a cut-off value of 4 or 5 mm double layer of endometrial thickness. 1,2,4 At this level, measurement of endometrial thickness has a sensitivity of 96% to diagnose endometrial carcinoma for a specificity of 50%. At a prevalence of 10% this implicates that a woman with an endometrial thickness 4 mm still has almost a 1% chance of having an endometrial carcinoma. Based on this post-test probability, expectant management is at present recommended to these women. In case the endometrial thickness is more than 4 mm, endometrial sampling is recommended. Other diagnostic procedures such as diagnostic hysteroscopy can also be used to further rule out endometrial cancer, but these procedures are substantially more invasive and incur additional health risks, increased discomforts and costs. Diagnostic hysteroscopy is at present not offered to women with endometrial thickness 4 mm. Although guidelines leave room for individual choices, the preference of the women has never been systematically taken into account when these guidelines were made. 1,2,4 The question is to what extent women are willing to accept a higher probability of cancer (> 1%) associated with expectant management in case of thin endometrium, in order to avoid further invasive procedures. Although current guidelines presume that women consider this small risk as acceptable, empirical data on this presumption are lacking. 1,2,4 The present study therefore addresses trade-offs and preferences for diagnostic management of women presenting with PMB at risk for endometrial carcinoma. Patients preferences and PMB 23 Materials and Methods Study sample The study was conducted in the St. Antonius Hospital, Nieuwegein, The Netherlands. The St. Antonius Hospital is a teaching hospital with an outpatient hysteroscopy department, where 500 hysteroscopies are performed annually. In this hospital, all women with PMB and an endometrial thickness > 4 mm undergo an outpatient hysteroscopy. Easily contactable women were selected on a sequential basis from an electronic database in which all women that visit the outpatient gynaecology department for PMB are recorded. Inclusion criteria for the study were: PMB, a completed diagnostic work-up including an outpatient hysteroscopy, benign diagnosis, ability to read and sufficient command of the Dutch language. The interviewer invited women by telephone to participate in a 45-minute interview at the hospital after explaining the study objective and procedure. The interviews were taken from January

25 2005 until March Women who had visited the outpatient clinic in 2004 were eligible for this study. Structured interview A structured interview was designed for systematic assessment of patient s preferences in the diagnostic work-up of PMB. The interview was reviewed by the department of communication of the St. Antonius Hospital with respect to patients acceptability to the interview (readability, comprehensibility). All interviews were performed in an identical setting by the first author (A.T.). During the interview women were offered the questions on paper. The interviewer clarified problems, and discussed the questions with the women where necessary. The women provided their answers on paper. Chapter 2 24 The first part of the interview focused on endometrial carcinoma and consisted of questions regarding a trade-off between expectant management and performing a complete diagnostic work-up including endometrial sampling, providing reassurance with a false negative rate close to 0%. Women were informed about this trade-off and the different diagnostic tests involved (transvaginal ultrasound (TVU), endometrial sampling [Pipelle], saline infusion sonography [SIS], and diagnostic hysteroscopy). Women were presented with two scenarios: (A) TVU as a first step and in case of unsuspicious TVU result, expectant management thereby accepting a small risk of missing an endometrial carcinoma and avoiding further diagnostic invasive measures and (B) immediate invasive diagnostic procedure (office hysteroscopy with endometrial sampling) for all patients with almost complete reassurance. Women were asked whether they were willing to accept expectant management (i.e. no further procedures after ultrasound unless bleeding recurred), and what they considered an acceptable risk for missing a carcinoma (Table 1). Table 1. Questions regarding the diagnostic management Questions If I had to choose between expectant management and thereby avoiding further invasive diagnostic procedures (A) or immediate reassurance with a complete diagnostic work-up (B), I would choose A, surely A, probably No preference B, probably B, surely Would you accept a small chance of missing of tumour to avoid unnecessary, invasive diagnostic procedures? Yes, surely Yes, probably Maybe No, probably No, surely How many percent do you need to be sure of absence of a tumour to avoid invasive diagnostic procedures?

26 The second part concerned management of polyps. Women were informed that in case malignancy was excluded with endometrial biopsy (Pipelle), there still remained a chance of presence of a benign endometrial polyp. Thereby, they were presented with two scenarios: (A) expectant management and only in case of recurrent bleeding hysteroscopy to evaluate presence or absence of endometrial polyp, and treatment in case of a polyp or (B) immediate hysteroscopy to diagnose endometrial polyp and in case of presence of a polyp, treatment in the same session. In scenario A, the chance of recurrent bleeding was systematically increased from 0 to 100% until women indicated indifference or preference for immediate diagnostic assessment. Costs associated with diagnostic procedures were not considered in the scenarios. Results In 2004, 70 women had a hysteroscopy because of PMB, of which 7 women had endometrial carcinoma. Of the remaining 63 women, 49 women were randomly selected from the database. Three women did not meet the inclusion criteria (one woman was deaf, one woman was mentally disabled and one woman could not read or speak Dutch). Three women refused to participate, six women were not mobile enough to come to the hospital and three women did not show up on their interview appointment and for these women three other women were selected in the same manner by which the other women had been selected. Therefore, 40 women were interviewed. At the interview one woman had turned out to have an endometrial carcinoma diagnosed; she was not included in the analysis. With respect to the diagnostic management to rule out endometrial carcinoma, 35 women (90%; 95% CI: 76 to 96%) indicated that they definitely wanted to rule out presence of carcinoma. Four (10%; 95% CI: 4 to 24%) would opt for expectant management. Subsequently women were asked to give a percentage on the absence of endometrial carcinoma in order to accept expectant management. Twenty-three women (59%; 95% CI: 43 to 73%) wanted to be 100% certain that carcinoma could be ruled out, 14 women (36%; 95% CI: 23 to 52%) wanted to be sure for % that carcinoma could be ruled out, while two women (5%; 95% CI: 1.4 to 17%) were willing to accept a risk > 5% (Figure 1). Patients preferences and PMB 25

27 Figure 1. Acceptance of expectant management % accepting expectant management % 0.1% 1% 10% 100% risk of false reassurance (1-sensivity) Chapter 2 26 In the trade-off between immediate diagnosis and subsequent treatment of polyps on one hand or expectant management on the other hand, women were more likely to accept an expectant approach. The lower the risk of recurrent bleeding was made, the more women would opt for expectant management. Only if the risk of recurrent bleeding within one year exceeded 25%, the majority of women would prefer immediate hysteroscopy for polyp detection (Figure 2). Figure 2. Acceptance of expectant management depending on risk of recurrent bleeding after malignancy has been excluded % preferring expectant management % 10% 25% 50% 75% 95% risk of recurrent bleeding within 1 year

28 Discussion To our knowledge, this study is the first to elicit preferences from women with PMB with respect to the diagnostic management. Most women with PMB expressed a strong preference for being informed about the diagnostic management of their complaints and the different options available to them, and they preferred to participate in this decision. The vast majority of women wanted cancer to be ruled out with almost 100% certainty. In this study women were given a choice between an invasive test with almost 100% certainty and expectant management with less but still high certainty. The women in this study were aware of the fact that in the context of diagnostic testing 100% certainty is not available. With respect to the diagnostic work-up of benign pathology (endometrial polyps) women were more likely to accept expectant management, but only if the risk of recurrent bleeding did not exceed 25%. A potential limitation of our study is that women were interviewed after establishment of a diagnosis and that all but one woman had benign disease. However, we speculate that women with a diagnosis of endometrial carcinoma are more willing to rule out uncertainty. As women were almost unanimous in their choice to rule out cancer with almost 100% certainty, it is not likely that including women with endometrial carcinoma in the sample would alter our main findings. Moreover, our sample was relatively small and of 49 women meeting the inclusion criteria, 40 women participated. As nonparticipation (17%) was merely due to logistic problems, it is not likely that these women would have deviating preferences, although this cannot be completely ruled out. However, the results were homogenous and show strong pointing into an almost 100% certainty. Another source of bias could result from the fact that all women had undergone a hysteroscopy at the time of the interview. Due to a phenomenon called cognitive dissonance reduction, people change their perceptions in order to make their situation seem better. This might have influenced the results in a more positive attitude towards hysteroscopy. However, the actual experience of the hysteroscopy has provided women with realistic information of this procedure. Most women deemed office hysteroscopy a rather easy and minimal invasive procedure. Patients preferences and PMB 27 What are the implications of our findings for clinical practice? The definition of evidencebased medicine is to integrate the best available research evidence with clinical expertise and patients values. 5 At present, guidelines for the diagnostic work-up of women with PMB recommend to start with TVU and advise expectant management in case endometrial thickness < 5 mm. 1-4 Our study shows that this is not in line with patients preferences as thin endometrium does not rule out cancer and the probability of false reassurance is 1%. One should keep in mind that the prevalence of endometrial carcinoma can vary among different populations.

29 In the Netherlands the prevalence of endometrial carcinoma lies around 10%. 4 In the UK, the prevalence has been reported to be 5%, this changes the probability of false reassurance after TVU to 0.4%. 6 With respect to endometrial carcinoma, 60% wanted to be sure for 100%. The majority of the patients would not even accept a risk of 0.4% of missing endometrial carcinoma, as is the case in a population with a prevalence of 5%. This implicates that the position of TVU as a first test in women with PMB should be reconsidered. Women expressed a preference for immediate diagnosis and treatment of endometrial polyps in case the probability of recurrence of bleeding was estimated to be 25% or more. This contradicts with current practice where expectant management usually is advised once cancer has been ruled out by endometrial sampling and evaluation of the uterine cavity for the presence of polyps is only undertaken in case the bleeding recurs. However, the risk of recurrent bleeding in case an endometrial polyp left in situ is not known, therefore it is difficult to formulate an evidence-based policy for benign lesions. 7 In summary we found that women with PMB show a strong risk aversion. As a consequence, they are prepared to undergo invasive procedures to rule out endometrial cancer. In our opinion the choice of TVU as a first step in the work up for PMB should be reconsidered. Acknowledgements Chapter 2 28 Supported by VIDI Grant, Amstelveen, The Netherlands. Grant Price winning paper at the Annual congress of the European Society of Gynaecological Endoscopy, October 2005, Athens, Greece.

30 References 1. Epstein E. Management of postmenopausal bleeding in Sweden: a need for increased use of hydrosonography and hysteroscopy. Acta Obstet Gynecol Scand 2004;83: Gupta JK, Chien PF, Voit D, Clark TJ, Khan KS. Ultrasonographic endometrial thickness for diagnosing endometrial pathology in women with postmenopausal bleeding: a meta-analysis. Acta Obstet Gynecol Scand 2002;81: Smith-Bindman R, Kerlikowske K, Feldstein VA, et al. Endovaginal ultrasound to exclude endometrial cancer and other endometrial abnormalities. JAMA 8;280: NVOG (Dutch Society of Obstetrics and Gynaecology). NVOG-Richtlijn Abnormaal vaginaal bloedverlies in de menopauze [in Dutch]. (NVOG Guideline: Abnormal vaginal bleeding during menopause) Sackett DL, Rosenberg WM, Gray JA, Haynes RB, Richardson WS. Evidence based medicine: what it is and what it isn t. BMJ 1996;312: Bachmann LM, ter Riet G, Clark TJ, Gupta JK, Khan KS. Probability analysis for diagnosis of endometrial hyperplasia and cancer in postmenopausal bleeding: an approach for a rational diagnostic workup. Acta Obstet Gynecol Scand 2003;82: Nathani F, Clark TJ. Uterine polypectomy in the management of abnormal uterine bleeding: A systematic review. J Minim Invasive Gynecol 2006;13: Patients preferences and PMB 29

31

32 3 Meta-analysis of endometrial thickness measurement for detecting endometrial cancer in women with postmenopausal bleeding: a different approach using individual patient data A. Timmermans, B.C. Opmeer, K.S. Khan, L.M. Bachmann, E. Epstein, T.J. Clark, J.K. Gupta, S.H. Bakour, T. van den Bosch, H.C. van Doorn, S.T. Cameron, M.G. Giusa, S. Dessole, F.P.H.L.J. Dijkhuizen, G. ter Riet, B.W.J. Mol

33 Abstract Chapter 3 Objective: To determine the diagnostic accuracy of endometrial thickness measurement in the detection of endometrial cancer among women with postmenopausal bleeding (PMB) with individual patient data using different meta-analytic strategies. Methods: We identified studies reporting on endometrial thickness measurements and endometrial carcinoma in women with PMB. Several approaches were used in the analyses of the acquired data. First, we performed Receiver Operator Characteristics (ROC) analysis per study, resulting in a summary Area Under the ROC Curve (AUC) calculated as a weighted mean of AUCs from original studies. Second, individual patient data were pooled and analyzed with ROC analyses irrespective of study, with standardization of distributional differences across studies using multiples of the median and by random effects logistic regression. Finally, we also used a two stage procedure, calculating sensitivities and specificities for each study, and using the bivariate random effects model to estimate summary estimates for diagnostic accuracy. Results: We contacted 79 primary investigators to obtain the individual patient data of their reported studies, of which 13 could provide data. Data on 2,896 patients, of which 259 had carcinoma, were included. All different ROC analyses resulted in rather comparable ROC curves with AUCs of 0.82 and 0.84 respectively, and summary estimates for sensitivity and specificity located along these curves. These curves were indicated a lower AUC than previously reported meta-analyses using conventional techniques. Conclusion: Previous meta-analyses on endometrial thickness measurement have overestimated its diagnostic accuracy in the detection of endometrial carcinoma. 32

34 Introduction Since two decades transvaginal ultrasonography (TVU) has become widely used in the evaluation of women with postmenopausal bleeding (PMB). Before TVU was introduced in the early 1990s, women with PMB were scheduled for dilatation and curettage (D&C). The goal of TVU assessment of the endometrium is to exclude endometrial carcinoma. In case of a thin endometrium, the chance of endometrial carcinoma is considered to be low and expectant management can be recommended, thus preventing endometrial sampling with D&C or office endometrial biopsy in these patients. In case the endometrial thickness is thickened, additional endometrial sampling is warranted. 1-5 Different guidelines in various countries advise measurement of endometrial thickness by TVU as a first step in the evaluation of women with PMB. 1-5 These guidelines mostly refer to the meta-analysis of Smith-Bindman et al. 6 Two other meta-analyses on this subject are rarely referred to in guidelines. 7,8 Similarly, the meta-analysis by Smith-Bindman et al. 6 is the most cited publication (total citation 169 by Web of Science) compared to the other two meta-analyses (citations of 37 and 37 respectively by Web of Science). 7,8 The meta-analysis of Smith-Bindman et al. 6 used traditional methods of combining published data from different studies. Using the reported data from each study 2x2 tables were constructed that compared endometrial thickness measured at TVU versus presence or absence of endometrial carcinoma. Results across studies were combined in a summary Receiver Operator Characteristics (ROC) Curve. At a 5 mm cut-off the sensitivity for detecting endometrial cancer was 96% for a 39% falsepositive rate. Such a combination of sensitivity and specificity would reduce a pre-test probability of 10% for endometrial cancer to a post-test probability of 1%. IPD-meta-analysis of endometrial thickness With respect to meta-analysis of randomized controlled trials, individual patient data (IPD) meta-analysis is considered to be superior over meta-analysis of the literature. 9 Meta-analyses using individual patient data instead of published summary data have come to less optimistic, but more accurate conclusions. Individual patient data meta-analysis is even reported to be the gold standard for meta-analyzing randomized controlled trials In diagnostic research the same might apply. Instead of combining reported accuracy measurements, individual patient data can be used to come to a more precise conclusion on diagnostic performance. With respect to TVU measurement of endometrial thickness, the exact endometrial thickness of each patient in each study can be used, instead of classification in a two-by-two table in which the endometrial thickness is classified above or below a particular cut-off level.

35 In this study we report an individual patient data meta-analysis on the diagnostic accuracy of endometrial thickness measurement in the detection of endometrial cancer in women with PMB. As there is little evidence regarding the preferred analytical strategy in analyzing individual patient data on diagnostic test accuracy, we used different statistical approaches to combine and summarize the individual patient data to explore whether different approaches also yield different results. Methods Chapter 3 34 Literature search, study selection We started our meta-analysis with three previous meta-analyses on this subject. 6-8 These meta-analyses were checked for included studies and after crosschecking duplicates were excluded. 6-8 In their meta-analysis, Tabor et al. 7 had already contacted individual authors with the request to supply data on mean and median endometrial thickness and 10th and 90th centiles. Those authors that had been unable to provide data for the meta-analysis Tabor et al. 7 were not contacted for the present IPD-meta-analysis, since we considered it highly unlikely that those authors that could not provide the required data to Tabor et al. 7, would be able to provide us with their original data for the present meta-analysis. Furthermore, a MEDLINE search was performed identifying those studies reported since the published meta-analyses ( ). Studies were included if they reported on women with PMB in whom endometrial thickness was measured by TVU. The outcome endometrial carcinoma had to be confirmed histologically. Histological specimens were collected using different methods. In more recent studies verification of diagnosis through histology sampling could have been omitted, since in these studies the established cut-off level for endometrial thickness may have been accepted and verification deemed unnecessary. If description of follow-up to verify outcome was sufficiently sought for all women recruited in the study, these studies were eligible for the meta-analysis. Data gathering Contact information of the first and last authors was sought through MEDLINE (recent publications) or internet (e.g. google). The authors were then contacted through , fax, conventional mail or telephone. In case contact information was not available or in case the first author did not respond, the last author was contacted. In a general invitation letter authors were explained about the concept of IPD meta-analysis. Initially, they were asked if they were interested to participate. In case they responded positively, they were asked to provide the original data for the analysis. The authors were asked to provide data on the absolute value of endometrial thickness, data on presence of endometrial cancer for each patient, and additional patient characteristics (age, time since menopause, HRT use, presence of diabetes,

36 hypertension, anticoagulants use and BMI). Authors could supply their data in any convenient format. If authors had new, original data available that were not included in the original publications, they were asked whether they were willing to share these data, on the condition that the data fulfilled the inclusion criteria of our study. Statistical analysis Distribution of endometrial thickness was calculated per study for patients with and patients without endometrial carcinoma separately. Subsequently, we applied different methods to determine diagnostic accuracy of endometrial thickness. First, for each study a Receiver Operator Characteristics (ROC) analysis on endometrial thickness for endometrial carcinoma was performed and for each study an Area Under the Curve (AUC) was calculated. A mean AUC across studies was estimated, weighted for the sample size of each study. Second, data from all studies were pooled and directly analyzed for diagnostic accuracy: (1) ROC-analysis based on logistic regression irrespective of the original study; (2) ROC-analysis, in which we adjusted for differences in the distribution of endometrial thickness between studies in the pooled data by expressing endometrial thickness as multiples of the median (MoM) within each study; and (3) ROC-analysis based on a random effects logistic regression model, where the model allowed for different odds-ratios per study. The three ROC curves based on direct analyses of the individual patient data were then compared by plotting them in one graph, with the ROC curve based on the earlier reported conventional meta-analysis as a comparator. 6 Finally, a two stage approach was used, in which we introduced different cut-off levels for endometrial thickness per study of 1, 2, 3, 4, 5, 6, and 7 mm respectively. This meant that for a cut-off value of 4 mm, endometrial thickness > 4.0 mm was considered abnormal and endometrial thickness 4.0 mm was considered normal. For each cut-off level and for each study 2x2 tables were constructed, for which sensitivity and specificity were calculated. These sensitivities and specificities were analyzed simultaneously by using a non-linear random effects model to calculate summary estimates of sensitivity and specificity for a range of endometrial thickness cut-off levels. These summary estimates of sensitivity and 1-specificity were plotted along the other ROC curves. IPD-meta-analysis of endometrial thickness 35

37 Results A total of 90 publications were identified. Of 11 authors contact information was either not available or not correct. All remaining 79 authors were contacted, of which 33 authors replied and 13 authors provided their data (Figure 1) Figure 1. Flow chart of included studies Eligible studies Smith-Bindman N = 35 Gupta N = 57 Tabor N = 48 Exclusion based on Tabor N = 24 After exclusion of duplicates N = 74 Total studies N = 49 MEDLINE search N = 41 Chapter 3 Identified studies N = Authors contacted N = 79 Author contact details unavailable N = 11 Response N = 33 authors Data available N = 19 authors Data obtained N = 13 authors Exclusion dichotomized data Total data obtained N = 3,435 Total data included N = 2,896

38 One author (Epstein) supplied data on several publications ; the data for these publications were gathered in one database containing consecutive patients during one study period; these data were considered to come from one study Another author (Van den Bosch) provided data from two publications 13,28 ; the data for these publications were gathered in different hospitals in different time periods, therefore these data were not taken together, but considered to come from two studies. 13,28 Data on 3,435 patients, of which 296 had endometrial carcinoma, were available. Of two studies original endometrial thickness measurements had been dichotomized to endometrial thickness above or below 5 mm. 21,27 These studies were excluded in the present analysis, leaving data on 2,896 patients of which 259 had endometrial carcinoma. All other authors provided data on original endometrial thickness, age and diagnosis of endometrial carcinoma ,22-26,28 Table 1 presents information on different study characteristics, number of patients per study, prevalence of endometrial carcinoma, whether verification of diagnosis was complete or partial (through follow-up), study period and AUC of ROC-analysis. Figure 2 shows the range and median endometrial thickness in women with and women without endometrial carcinoma. The median endometrial thickness in patients without endometrial carcinoma varied from 2 to 7 mm between studies. The median endometrial thickness in patients with endometrial carcinoma varied more between the different studies (from 6 to 21 mm) (Figure 2). Table 1. Information on different included studies Author No. of pts No of EC (%) Verification Period AUC (95% CI) Vd Bosch (5.8) Complete ( ) Epstein (10.7) Complete ( ) IPD-meta-analysis of endometrial thickness Timmermans (6.8) Partial ( ) 37 V Doorn (11.9) Partial ( ) Bakour 53 5 (9.4) Complete ( ) Cameron 40 1 (2.5) Complete ( ) Giusa-Chiferi (23) Complete ( ) Bachmann (4.9) Partial ( ) Dessole (4.5) Complete ( ) Dijkhuizen (14.3) Complete ( ) Vd Bosch 99 6 (6.1) Complete ( ) Total (8.9) 0.84 EC, endometrial carcinoma; AUC, Area Under the Curve (95% Confidence Intervals)

39 Figure 2. Distribution of endometrial thickness in patients with and patients without endometrial carcinoma per study Endometrial thickness (mm) patients without endometrial carcinoma patients with endometrial carcinoma Chapter 3 38 vdbosch 2 Dijkhuizen Dessole Bachmann Cameron Giusa Bakour v Doorn Timmermans Epstein vdbosch ROC analysis on endometrial thickness for endometrial carcinoma per study yielded AUCs between 0.68 and 0.93 (Table 1 and Figure 3). The mean AUC across studies weighted by the sample size of each study was When data from all studies were pooled and analyzed irrespective of the original study, the ROC curve had an AUC of 0.82 (95% CI ). When endometrial thickness was expressed as MoM per study, the AUC was 0.83 (95% CI ), and the AUC based on the random effects logistic regression model was 0.84 (95% CI ).

40 Figure 3. ROC-curves for each individual study (thickness of the lines represent sample size of each study) sensitivity ROCs for individual studies v Doorn Epstein Timmermans 0.6 vd Bosch 0.4 Bachmann Dessole 0.2 vd Bosch (2) Giusa 0.0 Dijkhuizen Bakour 1-specificity Cameron Finally, for cut-off levels between 1 and 7 mm endometrial thickness, summary estimates based on the bivariate model decreased from 99.8 (95% CI 99.0 to 99.9) to 83.9 (95% CI 72.7 to 91.0) for sensitivity and increased from 5.4 (95% CI 3.1 to 9.1) to 68.7 (95% CI 62.4 to 74.3) for specificity. These estimates of sensitivity and 1-specificity were plotted as an ROC curve (Figure 4), and are located near the other ROC curves. IPD-meta-analysis of endometrial thickness 39 When these ROC curves were compared with the ROC of Smith-Bindman et al. 6 the overall diagnostic accuracy was lower (Figure 4). We looked at the commonly used different cut-off levels (4 mm and 5 mm). 1,5,29 In the present meta-analysis these were found to have a sensitivity of 94.8% (95% CI %) and a specificity of 46.7% (95% CI %) for 4 mm, and 90.3% (95% CI %) and 54.0% (95% CI %) for 5 mm respectively. A cut-off value of 3 mm was found to have a sensitivity of 97.9% (95% CI %) for a specificity of 35.4% (95% CI %).

41 Figure 4. Receiver Operator Characteristics Curve of endometrial thickness. SROC S-B MoM Logistic RE logistic Bivariate RE 2 mm 3 mm 4 mm 5 mm specificity 6 mm 7 mm 8 mm S-B: ROC curve of Smith-Bindman; MoM: ROC curve based on Multiples of the Median; Logistic: ROC curve based on logistic regression; RE logistic: ROC curve based on logistic regression with random-effects model; Bivariate RE: random-effects model combining different cut-off values (1-7 mm) Chapter 3 Discussion 40 The present meta-analysis used individual patient data to estimate the diagnostic accuracy of endometrial thickness measurement by TVU for presence of endometrial carcinoma among women with PMB. Different methods of ROC analysis were used, which all resulted in comparable ROC curves. In all these analyses, we found that the diagnostic accuracy was lower than that of Smith-Bindman et al. 6 The commonly used cut-off value of 4 mm and 5 mm were found to have a sensitivity of 95% en 90% respectively. Only a cut-off values of 3 mm, yielded a high sensitivity of 98%. The ROC curve is an informative way to present accuracy for each possible cut-off value of endometrial thickness measurement. For clinical practice, usually one cut-off value is used. The optimal cut-off value is determined by the consequences associated with false positive and false negative results. In the case of endometrial carcinoma, the consequences of false negative results are far stronger than the consequences of false positive results, i.e. one aims to exclude endometrial carcinoma with a high amount of certainty, and accepts therefore many

42 false positive diagnoses resulting in unnecessary endometrial biopsies. Therefore, clinicians aim for virtually 100% sensitivity. A data-driven selection of optimal cut-off values is prone to bias. It potentially leads to overestimation of the sensitivity and specificity of the test under study. A data-driven approach to select the optimal cut-off value usually leads to a point above the true underlying ROC curve. 30 Consequently, combining these different reported cut-off values of smaller studies in a metaanalysis could lead to overestimation of diagnostic accuracy. The use of the original individual patient data can correct for this bias, as it ignores the reported optimal cut-off values, but incorporates the whole range of data. Moreover estimates are more precise because of increasing sample size. Unfortunately, not all studies available on diagnostic accuracy of endometrial thickness could be included in the present meta-analysis. This was mainly due to the fact that authors did not have the original data available anymore. Since the mid 90 s, in which the majority of the papers was published, many authors have moved and changed affiliation. Furthermore, information technology has rapidly changed leading to different computer programs and different ways of storing data. Although this lack of inclusion is a potential source of bias, we were able to report on 2,896 women across different studies. It seems likely that the lack of inclusion did not affect our results, since we were able to report on a high number of women. However, for future IPD research, solutions to solve this problem have to be explored. A possible solution could be that it becomes mandatory to supply journals with original databases once an article is published and that these databases are stored. Apart from our meta-analysis there are three meta-analyses on this subject which have used different methods and have come to different conclusions. 6-8 Smith-Bindman et al. 6,9 performed a conventional meta-analysis using reported cut-off values, combining these in a summary ROC. Tabor et al. 7 used original data on endometrial thickness and for each study they calculated median endometrial thickness per center and used multiples of the median for endometrial thickness to pool data. They reported a sensitivity of 96% for a specificity of 50%, implicating that a 10% pre-test probability would result in a post-test probability of 0.8%. Gupta et al. 8 conducted a comprehensive systematic review in which they focused on the study quality assessment of each study. Only four studies were identified as best-quality studies. Pooling of the results of these four studies for endometrial thickness 5 mm resulted in a Likelihood Ratio (LR) of a negative test of Such a LR would implicate that in a patient with a negative test result a pre-test probability of 10% would change to a post-test probability of 2.5%. Although Gupta et al. 8 and Tabor et al. 7 have come to less optimistic conclusions on the diagnostic accuracy of endometrial thickness measurement, the meta-analysis by Smith-Bindman et al. 6 has had the most impact on clinical practice. This meta-analysis is IPD-meta-analysis of endometrial thickness 41

43 mostly quoted in different guidelines and has the highest average citations per year in web of science (15.36 for Smith-Bindman et al. 6 compared to 5.29 for both Tabor et al. 7 and Gupta et al. 8 ). The present IPD meta-analysis came to an overall lower diagnostic accuracy than the meta-analysis by Smith-Bindman et al. 6 When we looked at different cut-off values, a cut-off value of 3 mm (i.e. endometrial thickness > 3 mm warrants endometrial sampling) was found to have a high sensitivity of 98%. Cut-off values of 4 and 5 mm had a lower sensitivity than the previously reported 95% and 90% compared to 96% reported by Smith-Bindman et al. 6 This would argue for use of a 3 mm cut-off value. To guide clinical decision making there is a need for good quality primary accuracy studies, of which the original data are retained so that they can eventually be included in subsequent IPD meta-analyses. Systematic reviews of test accuracy test have come available to summarize primary accuracy studies. However, there is ongoing debate about the most appropriate methodology for summarizing the available evidence in literature. Chapter 3 In conclusion, using individual patient data instead of reported data and optimal cut-off values of endometrial thickness measurement for endometrial carcinoma in women with PMB, leads to a less optimistic estimation of this diagnostic accuracy. Nonetheless, the a cut-off value of 3 mm has a sensitivity of 98%. Therefore, TVU measurement of endometrial thickness in women with postmenopausal bleeding using a cut-off value of 3 mm is still clinically useful and using such a cut-off value can reliably exclude endometrial carcinoma in women with PMB. 42

44 References 1. Epstein E. Management of postmenopausal bleeding in Sweden: a need for increased use of hydrosonography and hysteroscopy. Acta Obstet Gynecol Scand 2004;83: Scottish Intercollegiate Guidelines Network. Investigation of postmenopausal bleeding. 1st edn edinburgh, UK: Scottisch Intercollegiate Guidelines Network, Royal College of Physicians 2002; ( 3. Gull B, Karlsson B, Milsom I, Granberg S. Can ultrasound replace dilation and curettage? A longitudinal evaluation of postmenopausal bleeding and transvaginal sonographic measurement of the endometrium as predictors of endometrial cancer. Am J Obstet Gynecol 2003;188: Goldstein RB, Bree RL, Benson CB, et al. Evaluation of the woman with postmenopausal bleeding: Society of Radiologists in Ultrasound-Sponsored Consensus Conference statement. J Ultrasound Med 2001;20: NVOG (Dutch Society of Obstetrics and Gynaecology). NVOG-Richtlijn Abnormaal vaginaal bloedverlies in de menopauze [in Dutch]. (NVOG Guideline: Abnormal vaginal bleeding during menopause) Smith-Bindman R, Kerlikowske K, Feldstein VA, et al. Endovaginal ultrasound to exclude endometrial cancer and other endometrial abnormalities. JAMA 1998;280: Tabor A, Watt HC, Wald NJ. Endometrial thickness as a test for endometrial cancer in women with postmenopausal vaginal bleeding. Obstet Gynecol 2002;99: Gupta JK, Chien PF, Voit D, Clark TJ, Khan KS. Ultrasonographic endometrial thickness for diagnosing endometrial pathology in women with postmenopausal bleeding: a meta-analysis. Acta Obstet Gynecol Scand 2002;81: Stewart LA, Parmar MK. Meta-analysis of the literature or of individual patient data: is there a difference? Lancet 1993;341: Simmonds MC, Higgins JP, Stewart LA, Tierney JF, Clarke MJ, Thompson SG. Meta-analysis of individual patient data from randomized trials: a review of methods used in practice. Clin Trials 2005;2: Stewart LA, Tierney JF. To IPD or not to IPD? Advantages and disadvantages of systematic reviews using individual patient data. Eval Health Prof 2002;25(1): Horton R. The information wars. Lancet 1999;353: Van den Bosch T, Vandendael A, Van Schoubroeck D, Wranz PA, Lombard CJ. Combining vaginal ultrasonography and office endometrial sampling in the diagnosis of endometrial disease in postmenopausal women. Obstet Gynecol 1995;85: Epstein E, Ramirez A, Skoog L, Valentin L. Transvaginal sonography, saline contrast sonohysterography and hysteroscopy for the investigation of women with postmenopausal bleeding and endometrium > 5 mm. Ultrasound Obstet Gynecol 2001;18: IPD-meta-analysis of endometrial thickness 43

45 Chapter Epstein E, Ramirez A, Skoog L, Valentin L. Dilatation and curettage fails to detect most focal lesions in the uterine cavity in women with postmenopausal bleeding. Acta Obstet Gynecol Scand 2001;80: Epstein E, Valentin L. Rebleeding and endometrial growth in women with postmenopausal bleeding and endometrial thickness < 5 mm managed by dilatation and curettage or ultrasound follow-up: a randomized controlled study. Ultrasound Obstet Gynecol 2001;18: Epstein E, Valentin L. Gray-scale ultrasound morphology in the presence or absence of intrauterine fluid and vascularity as assessed by color Doppler for discrimination between benign and malignant endometrium in women with postmenopausal bleeding. Ultrasound Obstet Gynecol 2006;28: Bakour SH, Dwarakanath LS, Khan KS, Newton JR, Gupta JK. The diagnostic accuracy of ultrasound scan in predicting endometrial hyperplasia and cancer in postmenopausal bleeding. Acta Obstet Gynecol Scand 1999;78: Giusa-Chiferi MG, Goncalves WJ, Baracat EC, de Albuquerque Neto LC, Bortoletto CC, de Lima GR. Transvaginal ultrasound, uterine biopsy and hysteroscopy for postmenopausal bleeding. Int J Gynaecol Obstet 1996;55: Cameron ST, Walker J, Chambers S, Critchley H. Comparison of transvaginal ultrasound, saline infusion sonography and hysteroscopy to investigate postmenopausal bleeding and unscheduled bleeding on HRT. Aust N Z J Obstet Gynaecol 2001;41: Clark TJ, Bakour SH, Gupta JK, Khan KS. Evaluation of outpatient hysteroscopy and ultrasonography in the diagnosis of endometrial disease. Obstet Gynecol 2002;99: Bachmann LM, ter Riet G, Clark TJ, Gupta JK, Khan KS. Probability analysis for diagnosis of endometrial hyperplasia and cancer in postmenopausal bleeding: an approach for a rational diagnostic workup. Acta Obstet Gynecol Scand 2003;82: Litta P, Merlin F, Saccardi C, et al. Role of hysteroscopy with endometrial biopsy to rule out endometrial cancer in postmenopausal women with abnormal uterine bleeding. Maturitas 2005;50: Dijkhuizen FP, Brolmann HA, Potters AE, Bongers MY, Heinz AP. The accuracy of transvaginal ultrasonography in the diagnosis of endometrial abnormalities. Obstet Gynecol 1996;87: Timmermans A, Gerritse MB, Opmeer BC, Jansen FW, Mol BW, Veersema S. Diagnostic accuracy of endometrial thickness to exclude polyps in women with postmenopausal bleeding. J Clin Ultrasound 2008;36: van Doorn LC, Dijkhuizen FP, Kruitwagen RF, et al. Accuracy of transvaginal ultrasonography in diabetic or obese women with postmenopausal bleeding. Obstet Gynecol 2004;104: Gupta JK, Wilson S, Desai P, Hau C. How should we investigate women with postmenopausal bleeding? Acta Obstet Gynecol Scand 1996;75: Van den Bosch T, Van Schoubroeck D, Ameye L, Van Huffel S, Timmerman D. Ultrasound examination of the endometrium before and after Pipelle endometrial sampling. Ultrasound Obstet Gynecol 2005;26:283-6.

46 29. Clark TJ, Barton PM, Coomarasamy A, Gupta JK, Khan KS. Investigating postmenopausal bleeding for endometrial cancer: cost-effectiveness of initial diagnostic strategies. BJOG 2006;113: Leeflang MM, Moons KG, Reitsma JB, Zwinderman AH. Bias in sensitivity and specificity caused by data-driven selection of optimal cutoff values: mechanisms, magnitude, and solutions. Clin Chem 2008;54: IPD-meta-analysis of endometrial thickness 45

47

48 4 Office hysteroscopy in women with postmenopausal bleeding: see and treat of endometrial polyps using a Duckbill polypsnare A. Timmermans, S. Veersema (Gyn Surg 2004; 1: )

49 Abstract Objective: To describe the use of office hysteroscopy in the diagnosis and treatment in women with postmenopausal bleeding (PMB). Methods: In 83 women with PMB in whom an endometrial thickness of more than 4 mm was found, hysteroscopy was performed. Hysteroscopy was performed in an office setting, using a vaginoscopic approach. If a polyp was visualised this polyp was removed in the same session (see and treat) using the Duckbill polyp snare. Results: In 10.8% of the women endometrial carcinoma was diagnosed. In 41% of the women a polyp was diagnosed and treated. Conclusion: Office hysteroscopy in women with PMB and an endometrial thickness of more than 4 mm offers the possibility of diagnosis as well as treatment in the same session using a Duckbill polyp snare. Chapter 4 48

50 Introduction Between 10 and 15% of women with postmenopausal bleeding (PMB) have endometrial carcinoma or a premalignant disorder of the endometrium. 1,2 Transvaginal ultrasound (TVU) with measurement of endometrial thickness can be used to discriminate between normal and pathological endometrium. When a cut-off value of 5 mm is used, a prevalence for endometrial polyps of up to 43% can be found. 2 Women with PMB undergo endometrial biopsy to exclude endometrial cancer. Women who are biopsied and are found to have either benign diagnosis or insufficient tissue for diagnosis may still have endometrial polyps. Endometrial cancer is then excluded but benign polyps are not diagnosed or treated. Hysteroscopy with endometrial resection has been proven to be superior to dilatation and curettage especially in the presence of focal lesions. 3 Nowadays, the diagnostic and therapeutic utility of hysteroscopy in intracavitary pathologies is widely recognised. Improvements in endoscopic technology and the introduction of office hysteroscopy have made minimally invasive options available for both the diagnosis and treatment of endometrial polyps. 4 We describe the use of office hysteroscopy in the diagnosis and treatment in women with PMB. Materials and Methods Patients From January 2002 until December 2002 women with PMB presenting at our clinic underwent TVU examination. When on TVU an endometrial thickness of more than 4 mm was found, these women were scheduled for office hysteroscopy. In 2002, 83 women underwent office hysteroscopy because of PMB and endometrial thickness of more than 4 mm. Office hysteroscopy and PMB 49 Office hysteroscopy Office hysteroscopy was performed using a 5.5 mm rigid continuous-flow hysteroscopic system with a 5 Fr. working channel (Olympus). A vaginoscopic approach to hysteroscopy was used, i.e. without speculum, tenaculum, dilatation or anaesthesia. 5 The uterine cavity was distended using Sorbitol. If during hysteroscopy a lesion suspect for malignancy was seen, this lesion was biopsied. If no lesion was seen, a random biopsy from the endometrium was taken. If a focally growing (benign) lesion was identified (a polyp), if possible this lesion was removed in the same session: see and treat. Small (< 5 mm) polyps were removed using mechanical instruments such as endoscopic forceps or scissors; polyps larger than 5 mm were removed using the Duckbill Polyp Snare.

51 Operative technique: the Duckbill Polyp Snare (Figure 1) The snare (5 Fr.) was introduced through the working channel of the scope until the tip was visualised and advanced to the area of the polyp. The loop was opened to hang the polyp. By moving the tip of the sheath and advancing the snare, the loop was placed to the base of the polyp. Then after closing the loop tightly, the wire was charged with appropriate cutting current (70 Watt) thereby cutting the polyp at its base. The loop was then opened again to catch the floating polyp, and after closing the loop firmly around the free polyp, the scope was removed with the polyp simultaneously ending the procedure. In case of larger polyps in which the base of the polyp could not be visualised or reached, the procedure was repeated until the polyp was removed completely. All tissue was sent for histological examination. The final diagnosis of each woman was made on the basis of the histological diagnosis. In case of (pre)malignancy the patient was scheduled for further treatment. Figure 1. The Duckbill Polyp Snare Chapter 4 50 Results In 2002, 83 patients underwent office hysteroscopy because of PMB and an endometrial thickness of more than 4 mm on TVU. The mean age of the women was 60 years (range 46-88). Of these 83 women, 10.8% of the patients had an endometrial carcinoma, whereas 41% had an endometrial polyp (Table 1). In 35 cases a polyp was identified. Of the 35 polyps, 20 polyps were removed with the Duckbill Polyp Snare, this accounted for 60% of the polypectomies; nine polyps were removed with the use of mechanical instruments, and 5 polyps were biopsied. Of these 35 polyps, in 34 cases a benign endometrial polyp was confirmed by histology. In one case, however, a carcinoma was found inside the polyp. This accounts for 2.8% of malignancy inside polyps. In 46% of the women a polyp (endocervical or endometrial) was identified and treated in the same session. In 13% of the women a (pre)malignant lesion was seen and hysteroscopy served

52 for diagnosis and tissue sampling. After hysteroscopy and histology results, the women could be scheduled for further treatment. In 41% women intracavitary pathology could be excluded by hysteroscopy. Table 1. Hysteroscopy results N % Atrophy Active benign Endometrial polyp Endocervical polyp Submucous myoma Hyperplasia Atypical hyperplasia Endometrial carcinoma Cervical carcinoma Total Conclusion Office hysteroscopy in women with PMB and an endometrium thickness of more than 4 mm offers the possibility of diagnosis as well as treatment in the same session. A decision-analysis regarding the diagnosis and treatment in women with PMB is necessary. Office hysteroscopy and PMB 51

53 References 1. Dijkhuizen FP, Brolmann HA, Potters AE, Bongers MY, Heinz AP. The accuracy of transvaginal ultrasonography in the diagnosis of endometrial abnormalities. Obstet Gynecol 1996;87: Emanuel MH, Verdel MJ, Wamsteker K, Lammes FB. An audit of true prevalence of intrauterine pathology: the hyesteroscopic findings, controlled for patient selection in 1,202 patients with abnormal uterine bleeding. Gynaecol Endosc 1995;4: Epstein E, Ramirez A, Skoog L, Valentin L. Dilatation and curettage fails to detect most focal lesions in the uterine cavity in women with postmenopausal bleeding. Acta Obstet Gynecol Scand 2001;80: Bettocchi S, Ceci O, Di Venere R, et al. Advanced operative office hysteroscopy without anaesthesia: analysis of 501 cases treated with a 5 Fr. bipolar electrode. Hum Reprod 2002;17: Bettocchi S, Selvaggi L. A vaginoscopic approach to reduce the pain of office hysteroscopy. J Am Assoc Gynecol Laparosc 1997;4: Chapter 4 52

54 5 The diagnostic accuracy of endometrial thickness to exclude polyps in women with postmenopausal bleeding A. Timmermans, M.B.E. Gerritse, B.C. Opmeer, F.W. Jansen, B.W.J. Mol, S. Veersema (J Clin Ultrasound 2008; 36: )

55 Abstract Objective: To determine the accuracy of endometrial thickness measurement with transvaginal ultrasound (TVU) to diagnose endometrial polyps in women with postmenopausal bleeding (PMB) in whom a carcinoma has been ruled out. Methods: In women with PMB endometrial thickness was measured with TVU. If endometrial thickness was > 4 mm, office hysteroscopy was performed. At hysteroscopy, the uterine cavity was assessed for the presence of polyps. Women with malignancy were excluded. We used Receiver Operating Characteristics (ROC) analysis to assess the capacity of TVU endometrial thickness measurement to diagnose endometrial polyps. Findings at hysteroscopy were considered to be the reference standard. Results: We included 178 women with PMB and endometrial thickness > 4 mm. Hysteroscopy showed an endometrial polyp in 90 women (50%). The ROC analysis revealed that endometrial thickness had an area under the curve of 0.64 in the diagnosis of endometrial polyps. Conclusion: In women with PMB in whom carcinoma has been ruled out, measurement of endometrial thickness with TVU is not useful in the diagnosis of endometrial polyps. Chapter 5 54

56 Introduction Postmenopausal bleeding (PMB) is often caused by abnormalities of the endometrium, whether they are benign or malignant. At present, transvaginal ultrasound (TVU) is used as a first step in the evaluation of women with PMB. 1-3 The probability of malignant pathology is strongly reduced in the presence of a distinct endometrial ultrasound with an endometrial thickness 4 mm. 4 Endometrial sampling is not recommended below this cut-off value, but it is warranted to rule out malignancy above this cut-off value. 5 The diagnostic approach is less clear with respect to benign disease. The prevalence of endometrial polyps in women with PMB and endometrial thickness > 4 mm is estimated to be approximately 40%. 6 The majority of gynaecologists advocate removal of endometrial polyps. Removal of such polyps might reduce the probability of recurrent bleeding and allow histological assessment of the removed polyp. 7 From that point of view, it would be important for TVU to diagnose endometrial polyps so that a selective hysteroscopy could be undertaken with the intention of removing the polyps. The role for measuring endometrial thickness in order to exclude endometrial carcinoma is clearly established. However, it is unclear whether measurement of endometrial thickness is clinically useful for predicting the presence or absence of endometrial polyps. In pre-menopausal women, a thin endometrium has been shown to reduce the probability of intrauterine abnormalities such as polyps, but did not exclude them. 8 No such data are available for symptomatic postmenopausal patients. Therefore the aim of the present study was to determine the accuracy of TVU endometrial thickness measurement in the diagnosis of endometrial polyps in women with PMB. Endometrial thickness and polyps 55 Material and Methods The study was performed between January 1 st 2002 and July 1 st 2005 in the St. Antonius Hospital, a university-affiliated teaching hospital in Nieuwegein, The Netherlands. The St. Antonius Hospital has a fixed local protocol for the diagnostic work-up of women with PMB. This protocol includes TVU measurement of endometrial thickness and office hysteroscopy in women with endometrial thickness > 4 mm. Consecutive women undergoing office hysteroscopy are documented in a database. Women who underwent hysteroscopy in the study period because of PMB were identified from this database. The medical files of these women were reviewed to obtain information about endometrial thickness, hysteroscopic findings and histology results. This study was limited to women with endometrial thickness > 4 mm or endometrial thickness not measurable.

57 The work-up was started with TVU which was performed by a gynaecologist or gynaecological resident to measure endometrial thickness. TVU was performed with 5 MHz vaginal probe transducers (UST 984-5) connected to an Aloka SSD 900 scanner (Aloka, Tokyo, Japan). The thickness of the endometrium was measured from a longitudinal sonogram through the thickest area of the endometrium. The endometrial stripe was scanned longitudinally from right to left until the thickest point was found, and this thickness was measured. The measurements of endometrial thickness included both layers. When the endometrial layers were separated by intracavitary fluid, each layer was measured and the sum was recorded. In case the endometrium was too ill-defined to allow a reliable measurement result, this was recorded as endometrial thickness not measurable. Furthermore, intracavitary pathology could be suspected on TVU, by a focal or circumscribed isoechogenic lesion of the endometrium. In case of endometrial thickness > 4 mm or in case of endometrial thickness not measurable, patients were scheduled for a hysteroscopy. Chapter 5 56 Hysteroscopy was performed in an office setting using a 5.5 mm continuous flow hysteroscope (3 mm telescope and 5.5 mm operative with 5 French channel sheats; (Olympus, Olympus America Inc., Melville, NY) by a gynaecologist with expertise in hysteroscopy or by a resident under direct supervision of this gynaecologist. The hysteroscopist was aware of the results of the TVU. Hysteroscopy was performed using a vaginoscopic approach, without speculum, tenaculum or local anaesthetics. During hysteroscopy the uterine cavity was described in a standard way, which included a description of polyps, submucous myomas, suspicion of malignancy, normal endometrium or atrophy. A polyp was defined as a benign, localized overgrowth of endometrial tissue covered by epithelium. Polyps could be sessile or pedunculated and they could be single (i.e. one polyp) or multiple (i.e. at least two polyps). In case of abnormalities in the uterine cavity, these abnormalities were resected or a biopsy was taken and the material was sent for pathological examination. Hysteroscopic findings and histology results were combined for final diagnosis. Patients with final diagnosis of malignancy were excluded. The STARD (Standards for Reporting of Diagnostic Accuracy) initiative checklist was used to report the results of the study. 9 Statistical analysis In the analysis, endometrial thickness was related to the final disease status (i.e. the presence of a benign polyp or not). We categorized the results of endometrial thickness and tabulated them against the presence or absence of polyps. Endometrial thickness that could not be measured was considered as a separate category, as was suspicion of intracavitary pathology, based on the TVU image. Likelihood ratios (LRs) with 95% confidence intervals (95% CIs) were calculated for different ranges of endometrial thickness, as well as for the diagnostic categories suspicion of intracavitary pathology and endometrial thickness not measurable. We performed Receiver Operating Characteristics (ROC) analysis to assess the discriminative

58 capacity of endometrial thickness for the presence of benign polyps. The area under the ROC curve (AUC) reflects the diagnostic accuracy of a test, incorporating sensitivity and specificity for all possible thresholds, thus allowing detection of an optimal cut-off point for further clinical management. An AUC of 0.5 indicates no discriminative capacity whereas an AUC of 1 indicates perfect discriminative capacity. Statistical analysis was performed with SAS version 9.1 (SAS Institute Inc., Cary, NC, USA). Results Two hundred forty-five women underwent office hysteroscopy for PMB. At hysteroscopy, 29 of the 245 women were diagnosed with endometrial carcinoma, leaving 216 women with a benign diagnosis. Among these 216 women, there were 100 women in whom hysteroscopy showed an endometrial polyp. In 38 of these 216 women, endometrial thickness was unknown, thus these women were excluded from further analysis. Endometrial thickness was recorded in 153 women. Endometrial thickness was not measurable in five of the 153 women and intracavitary pathology was suspected on TVU in 20 women (Figure 1). Figure 1. Flow Chart Malignancy (excluded) (N = 29) Women with ET > 4 mm undergoing hysteroscopy No malignancy (N = 216) Endometrial thickness and polyps 57 Endometrial thickness unknown (excluded) (N = 38) Endometrial thickness measurement (N = 153) Endometrial thickness not measurable (N = 5) Intracavitary pathology suspected by TVU (N = 20) Endometrial polyp (N = 10) Endometrial polyp (N = 76) Endometrial polyp (N = 3) Endometrial polyp (N = 11) Endometrial polyps were found in three women in whom endometrial thickness was not measurable (n=5), in 11 patients in whom intracavitary pathology was suspected (n=20) and in 76 women with recorded endometrial thickness (n=153). ROC analysis was performed us-

59 ing the 153 women with recorded endometrial thickness. Women with endometrial thickness not measurable, and women with suspected intracavitary pathology on TVU were analyzed as separate categories. Table 1 shows the likelihood for different TVU results. The LRs varied between 0.6 and 1.2, indicating poor diagnostic performance of TVU. Endometrial thickness not measurable showed a sensitivity of 3% (95% CI %) and specificity of 98% ( %). For TVU suspicion of intracavitary pathology the sensitivity was 12% (95% CI %) and specificity 90% (95% CI %). Table 1. Diagnostic value of endometrial thickness measured via TVU in the detection of endometrial polyps as established at hysteroscopy Endometrial thickness (mm) No. of patients No. of patients (%) with endometrial polyp at hysteroscopy LR (95% CI) for endometrial polyp (34%) 0.68 ( ) (64%) 1.28 ( ) > (61%) 1.20 ( ) Not measurable 5 3 (60%) 1.19 ( ) Polyp suspected on TVU image (55%) 1.09 ( ) CI, confidence interval; LR, Likelihood Ratio Chapter 5 58 Figure 2 shows the ROC curve for endometrial polyps. The AUC was 0.64, confirming the poor diagnostic performance that was derived from the LRs. Figure 2. ROC analysis of diagnostic value of endometrial thickness for detection of endometrial polyps sensitivity specificity

60 Discussion This study shows that endometrial thickness measured with TVU is not helpful in the diagnosis of endometrial polyps in women with PMB. Whereas in women with PMB a cut-off value of the endometrial thickness of 4 mm identifies women at low risk for endometrial carcinoma, no such cut-off level of endometrial thickness could be identified for endometrial polyps. Both the ROC analysis and likelihood ratios indicated poor discriminative capacity. A limitation of this study is that data were collected retrospectively. Although the protocol in our clinic is to measure endometrial thickness in case of PMB, and all women with endometrial thickness > 4 mm undergo office hysteroscopy, it is possible that not all women with an endometrial thickness > 4 mm had had an office hysteroscopy. Alternatively, they could either have had an outpatient endometrium sampling (Pipelle) without hysteroscopy or they could have been scheduled for an inpatient hysteroscopy under general anaesthesia. In 2004, we recorded all women that visited the outpatient department of our clinic because of PMB. Based on unpublished data from this database, 5% of the women with endometrial thickness> 4 mm did not have an office hysteroscopy. Because this 5% is rather low, and because the choice for not having an office hysteroscopy is rather random, we feel that it is unlikely that these dropouts have affected our results. Another limitation of this study is the fact that the hysteroscopist was aware of the sonographic findings. Women with thicker endometrium or women with suspicion of intracavitary pathology at TVU might therefore be evaluated more thoroughly at hysteroscopy than women with thinner endometrium. However, the definition of a polyp was well described in this study, and it seems unlikely that polyps might have been missed in case of thinner endometrium. Furthermore, the pathologist who was unaware of the TVU findings made the final diagnosis of a polyp. Finally, if such a bias were to be present, it would lead to an overestimation of accuracy. Thus, the poor discriminative capacity that we found might even be an overestimation of the accuracy. Endometrial thickness and polyps 59 According to current guidelines the work-up of women with PMB focuses on exclusion of endometrial carcinoma, and the work-up starts with measurement of endometrial thickness via TVU. In cases of endometrial thickness > 4 mm, further evaluation of the endometrium is required to exclude malignancy using either hysteroscopy with endometrial biopsy or by blind office endometrial sampling. 1-3 In cases where malignant pathology has been excluded by office endometrial sampling (e.g. Pipelle), significant benign pathology (i.e. endometrial polyps) may have been overlooked. Imaging of the distended uterine cavity is required to most accurately diagnose focal lesions such as endometrial polyps, and the techniques most often employed are hysteroscopy or Saline Infusion Sonography (SIS). Hysteroscopy or SIS

61 with the aim of detecting endometrial polyps can be incorporated in the diagnostic work-up at first episode of bleeding or in case the bleeding persists or recurs. 1,3,10 Hysteroscopy has the advantage of allowing simultaneous removal of polyps at the time of diagnosis, although data regarding the efficacy of polypectomy in treating PMB are scarce. 11 Therefore it is still questionable if the removal of polyps reduces the probability of recurrent bleeding and from this point of view it can be questioned if the diagnosis of endometrial polyps is important. Although data regarding efficacy of polypectomy are scarce, polyps are a frequent finding in the work-up of women with PMB. The prevalence of polyps in women with PMB and an endometrial thickness of more than 4 mm has been reported to be as high as 40%, a finding that was confirmed in this study. 6,12 Conventional TVU is available to most gynaecologists but SIS or outpatient hysteroscopy may not be. In current practice hysteroscopy or SIS are often performed in case of PMB and endometrial thickness > 8 mm or suspicion of intracavitary pathology on TVU. 1 Chapter 5 This study addressed the question of whether endometrial thickness measurement via TVU, with its established role in the diagnostic work up of PMB to exclude endometrial carcinoma, could also provide useful information about the presence or absence of the most likely benign intrauterine pathology, endometrial polyps. If so, subsequent testing via hysteroscopy or SIS could be employed to confirm or refute the diagnosis of an endometrial polyp in appropriate women. However, our study demonstrates that TVU endometrial thickness measurement has poor discriminative ability for detecting or excluding endometrial polyps and thus is not clinically useful in the diagnosis of such benign focal pathology. 60

62 References 1. Epstein E. Management of postmenopausal bleeding in Sweden: a need for increased use of hydrosonography and hysteroscopy. Acta Obstet Gynecol Scand 2004;83: Gupta JK, Chien PF, Voit D, Clark TJ, Khan KS. Ultrasonographic endometrial thickness for diagnosing endometrial pathology in women with postmenopausal bleeding: a meta-analysis. Acta Obstet Gynecol Scand 2002;81: NVOG (Dutch Society of Obstetrics and Gynaecology). NVOG-Richtlijn Abnormaal vaginaal bloedverlies in de menopauze [in Dutch]. (NVOG Guideline: Abnormal vaginal bleeding during menopause) Smith-Bindman R, Kerlikowske K, Feldstein VA, et al. Endovaginal ultrasound to exclude endometrial cancer and other endometrial abnormalities. JAMA 1998;280: Dijkhuizen FP, Mol BW, Brolmann HA, Heintz AP. The accuracy of endometrial sampling in the diagnosis of patients with endometrial carcinoma and hyperplasia: a meta-analysis. Cancer 2000;89: Epstein E, Ramirez A, Skoog L, Valentin L. Dilatation and curettage fails to detect most focal lesions in the uterine cavity in women with postmenopausal bleeding. Acta Obstet Gynecol Scand 2001;80: Clark TJ, Khan KS, Gupta JK. Current practice for the treatment of benign intrauterine polyps: a national questionnaire survey of consultant gynaecologists in UK. Eur J Obstet Gynecol Reprod Biol 2002;103: Dueholm M, Jensen ML, Laursen H, Kracht P. Can the endometrial thickness as measured by transvaginal sonography be used to exclude polyps or hyperplasia in pre-menopausal patients with abnormal uterine bleeding? Acta Obstet Gynecol Scand 2001;80: Bossuyt PM, Reitsma JB, Bruns DE, et al. Towards complete and accurate reporting of studies of diagnostic accuracy: the STARD initiative. BMJ 2003;326: Scottish Intercollegiate Guidelines Network. Investigation of postmenopausal bleeding. 1st edn edinburgh, UK: Scottisch Intercollegiate Guidelines Network, Royal College of Physicians 2002; ( 11. Nathani F, Clark TJ. Uterine polypectomy in the management of abnormal uterine bleeding: A systematic review. J Minim Invasive Gynecol 2006;13: Timmermans A, Veersema S. Office hysteroscopy in women with postmenopausal bleeding: see and treat of endometrial polyps using a Duckbill polyp snare. Gynecol Surg 2004;1: Endometrial thickness and polyps 61

63

64 6 Hysteroscopy and removal of endometrial polyps: a Dutch survey A. Timmermans, H. van Dongen, B.W.J. Mol, S. Veersema, F.W. Jansen (Eur J Obstet Gynecol Reprod Biol 2008; 138: 76-9)

65 Abstract Objective: To evaluate current practice of Dutch gynaecologists in polyp removal. Methods: All practicing gynaecologists in The Netherlands in 2003 were surveyed by a mailed self-administered questionnaire about polypectomy. Gynaecologists were asked about their individual performance of polypectomy: setting, form of anaesthesia, method and instrument used. Results: The response rate was 74% (553 of 752 gynaecologists). Among the respondents, 455 (83%) stated that they removed polyps themselves. Polyps were most commonly removed in an inpatient setting (71%), under general or regional anaesthesia (77%), and under direct hysteroscopic visualization (69%). Gynaecologists working in a teaching hospital removed polyps more often in an outpatient setting compared to gynaecologists working in a non-teaching hospital(93 (39%) versus 36 (19%) p<0.001). Conclusion: In the Netherlands, outpatient polyp removal is not practiced on a large scale. However teaching hospitals are more often performing polypectomy in an outpatient setting. Therefore, we expect that there must be a tendency towards outpatient hysteroscopic removal of polyps for the future. Further research is required to assess the efficacy of polyp removal. Chapter 6 64

66 Introduction Endometrial polyps are associated with abnormal uterine bleeding in both pre- and postmenopausal women. Its prevalence ranges from 20% in symptomatic premenopausal women up to 20 to 40% in women with postmenopausal bleeding (PMB). 1,2 Traditionally, endometrial polyps are removed by dilatation and curettage (D&C) under general anaesthesia. However, blind curettage has been shown to miss endometrial polyps in 50 to 85% of cases. 2-4 A retrospective analysis showed that D&C for removal of endometrial polyps decreased from 45% to 14% in the period 1990 to In this same period hysteroscopic resection increased significantly from 12.7% to 50.5%. 5 In the Netherlands, polypectomy is one of the most commonly performed surgical hysteroscopic procedures (31% of all surgical hysteroscopic procedures) and has the lowest complication rate (0.4%) compared to other surgical hysteroscopic procedures. 6 New developments in hysteroscopic armentarium (e.g. 5 Fr bipolar electrodes [Versapoint]) and smaller diameter hysteroscopic systems allow this type of surgery to be performed in an outpatient setting. 7 In literature large series of outpatient hysteroscopic polypectomy have been described and it can be considered as an effective procedure. 8,9 Furthermore, four studies have reported efficacy, cost-effectiveness and patient s satisfaction rates of outpatient hysteroscopy and outpatient hysteroscopic polypectomy to be better than in a day-case setting When outpatient hysteroscopy was compared to day-case hysteroscopy it offered faster recovery, less time away from work and home and cost savings when women were randomized to either outpatient hysteroscopy or day-case hysteroscopy. 11,12 For polyp removal it was demonstrated in a feasibility study, in 58 women with symptomatic endometrial polyps allocated to either inpatient or outpatient polypectomy, that ambulatory hysteroscopic polypectomy was efficacious and cost-effective. 10 Furthermore, a randomized controlled trial demonstrated that, in a group of 40 women diagnosed with an endometrial polyp, outpatient polypectomy showed minimal intra-operative discomfort, faster recovery and was preferred by women when compared with day-case polypectomy. 13 Survey on hysteroscopic polyp removal 65 There is evidence that outpatient polyp removal is better than its traditional inpatient counterpart. However, there still is a lack of high-quality evidence regarding the efficacy of intrauterine polypectomy with respect to abnormal uterine bleeding, although the majority of gynaecologists advocate removal of endometrial polyps. 14 Furthermore the utility of polyp removal is further confused by the variety of approaches now available. 14 In contrast to all evidence in literature regarding outpatient hysteroscopic polyp removal, Clark et al. 15 showed that still 90% of gynaecologists in the UK remove polyps in an inpatient

67 setting under general anaesthesia. The method most commonly performed, was dilatation and curettage (D&C). When we look at the situation in the UK and the evidence in literature, it seems that implementation of outpatient hysteroscopic polyp removal is progressing slowly. We therefore conducted a national questionnaire survey to assess the current practice regarding the removal of endometrial polyps in the Netherlands. Material and Methods For this study, all 752 practicing gynaecologists holding membership of the Dutch Society of Obstetrics and Gynaecology (NVOG) in 2003 were identified from the national database and were sent a questionnaire. The questionnaire was brief, explicit, had a structured format consisting of a series of closed questions. After pilot testing on seven members of the Dutch Working Committee of Gynaecological Endoscopy, the questionnaire was sent to all gynaecologists. Gynaecologists in training were not included. The survey included a covering letter and a prepaid response envelop. After eight weeks a single mailed reminder was sent to the non-responders. Chapter 6 66 The gynaecologists were asked whether they performed polypectomy for endometrial polyps themselves. Those that did were then requested to report about setting (inpatient or outpatient), form of anaesthesia (general, regional, local or none), method of removal (removal by D&C, removal by D&C following hysteroscopy or removal under direct hysteroscopic visualization) and type of hysteroscopic instrument used (mechanical instruments, 5 Fr electrosurgical instruments, resectoscope). A formal operating theatre with an anaesthesiologist present for general or regional anaesthesia was considered an inpatient setting. An outpatient setting was considered a setting where a patient could come in, have the procedure done and walk out again right after the procedure. Since general or regional anaesthesia requires an anaesthesiologist these two categories were considered as one entity. In parallel, the same was done for local or no anaesthesia. Respondents were asked to report whether they used the different modalities routinely, occasionally or never at all. The options that were chosen as a routine were used for further analysis. It was possible to leave questions open or give more than one answer to a question (e.g. general and regional anaesthesia as a routine). Statistical analysis The received information was collected in the statistical SPSS program (SPSS, version 12, SPSS Inc., Chicago, IL). For this study all gynaecologists were stratified to type of clinic in which they were working: teaching hospitals with a residency program for gynaecology (n=41) or nonteaching hospitals (n=59). Per answering option proportions of gynaecologists working in TH were compared with proportions of gynaecologists working in non-th. Comparisons were

68 tested for statistical significance using the Chi-square test, a p-value < 0.05 was supposed to indicate statistical significance. The same was done for method of removal compared to form of anaesthesia, using Chi-square test to compare and test for statistical significance. Results There were 458 gynaecologists (61%) who responded the first time, whereas another 95 questionnaires were returned after the single mail reminder. In total, 553 gynaecologists responded, corresponding to a response rate of 74%. Of these responders, 455 (83%) gynaecologists performed polypectomy themselves in the treatment of endometrial polyps. Answers about setting, anaesthesia, method and instrument were calculated as a percentage of the 455 gynaecologists that removed polyps themselves. There were 321 gynaecologists (71%) that removed polyps in an inpatient setting, 326 (77%) gynaecologists used general or regional anaesthesia and 290 (69%) gynaecologists choose removal under direct hysteroscopic vision as the routinely used method. Removal under direct hysteroscopic visualization was practiced routinely with mechanical instruments and the resectoscope by 197 and 159 gynaecologists respectively (48% and 38% respectively). An outpatient polypectomy setting was used by 129 (29%) of the gynaecologists and 98 (23%) gynaecologists used no or local anaesthesia. New 5 Fr electrosurgical instruments (e.g. bipolar electrode [Versapoint ] and monopolar snare) were used by 14% (56 gynaecologists) on a routine basis (Table 1). Survey on hysteroscopic polyp removal 67

69 Table 1. Current practice concerning setting, anaesthesia, method and instrument for removal of endometrial polyps, number of performing gynaecologists (%) Routine Total Teaching (n=241) Setting Non-teaching (n=194) p-value Inpatient 321 (71) 149 (62) 152 (78) p<0.001 Outpatient 129 (29) 93 (39) 36 (19) p<0.001 Anaesthesia General/ Regional 326 (77) 170 (70) 156 (80) ns Local/no anaesthesia 98 (23) 68 (28) 30 (15) p=0.001 Method D&C 17 (4) 4 (1.6) 13 (6.7) p=0.007 D&C after hysteroscopic localisation Removal under direct hysteroscopic vision 115 (27) 58 (24) 57 (29) ns 290 (69) 173 (72) 117 (59) p=0.008 Instrument Mechanical 197 (48) 108 (44) 89 (46) ns 5 Fr electrosurgical 56 (14) 37 (15) 19 (10) ns Chapter 6 Resectoscope 159 (38) 94 (39) 65 (34) ns teaching: academic and non-academic teaching hospitals; ns: not significant note: the denominators do not add up to 455 since it was possible to leave a question open 68 In teaching hospitals gynaecologists removed polyps significantly more in an outpatient setting as compared to gynaecologists in non-teaching hospitals (non-th): 93 (39%) versus 36 (19%) (p<0.001). Local or no anaesthesia was used more in teaching hospitals compared to non-th: 68 (28%) gynaecologists versus 30 (15%) gynaecologists (p=0.001). In non-th significantly more gynaecologists (6.7%) were removing polyps by D&C ( blindly ) compared to the teaching hospitals (1.6%)(p=0.007), whereas gynaecologists in teaching hospitals made more use of the hysteroscope for direct removal (72% vs. 59%, p=0.008). In case general or regional anaesthesia was used for polyp removal significantly more gynaecologists performed D&C following hysteroscopic localization when compared to removal under direct hysteroscopic vision (31% vs. 16%, p=0.001). If local or no anaesthesia was used the polyp was more often removed under direct hysteroscopic visualization (81% vs. 65%, p=0.007) (Table 2).

70 Table 2. Method of removal against form of anaesthesia: number of gynaecologists (%) General/regional anaesthesia Local/no anaesthesia Chi-square Blind 16 (4) 3 (3) ns Blind following hysteroscopy 115 (31) 16 (16) Under direct hysteroscopic vision 244 (65) 83 (81) Total ns: not significant Discussion According to our study, the current practice for polypectomy in the Netherlands is removal in an inpatient setting, using general or regional anaesthesia. Gynaecologists remove polyps under direct hysteroscopic visualization using mechanical instruments or the resectoscope. The findings reported should be considered valid and generalisable in view of the response rate of 74%. This is because it is above the desirable response rate of 70%, above which the impact of non-response biases is negligible. 16 In 2002 in the Netherlands 15,354 hysteroscopic procedures were performed: 11,428 diagnostic (74%) and 3,926 surgical (26%).The most commonly performed hysteroscopic surgical procedure was polypectomy. 17 Hysteroscopic polypectomy is therefore an extensive used procedure, although it s efficacy in the treatment of abnormal uterine bleeding is still matter of debate. There is high-quality evidence in favour of outpatient hysteroscopy and polyp removal. Four studies have reported efficacy, cost-effectiveness and patients satisfaction rates to be better in an outpatient setting compared to a day-case setting Survey on hysteroscopic polyp removal 69 In the Netherlands the outpatient setting for polyp removal was favoured by 29% of the gynaecologists. We related method of anaesthesia to method of removal. We found that the form of anaesthesia influenced method of removal. Under general or regional anaesthesia more gynaecologists removed polyps by D&C after hysteroscopic localization than with local or no anaesthesia (31% vs. 16%). When local or no anaesthesia was applied more gynaecologists removed polyps under direct hysteroscopic visualization compared to general or regional anaesthesia (81% vs. 65%). In the UK still 91% of the gynaecologists favoured an inpatient setting and the most favoured method for polyp removal there was D&C after hysteroscopic localization. 15 Although we observed a difference regarding outpatient hysteroscopy between the Dutch situation and current practice in the UK, in both countries the current situation is still far away from the situation as it is described in literature. The differences between the UK and the Netherlands might be explained by time, since the situation in the Netherlands was

71 evaluated at a later time than in the UK. Furthermore the differences might be due to different health care systems. When we look at the differences between teaching hospitals and non-th, the outpatient setting and local or no anaesthesia were used more by gynaecologists in teaching hospitals. Hysteroscopic surgery for polyp removal was performed more by gynaecologists in teaching hospitals than in non-th. Therefore, we stipulate that there must be a tendency towards outpatient hysteroscopic removal of polyps for the future, especially because the new generation gynaecologists are being taught to use the hysteroscope in an outpatient setting. This speculation is supported by the fact that another survey showed that in the UK 80% of the consultants without access to outpatient hysteroscopic facilities would like to have access to such facilities. 18 In conclusion, we can say that although hysteroscopic polypectomy can successfully be performed without anaesthesia in an outpatient setting, outpatient polyp removal is not practiced on a large scale. 10,13 As teaching hospitals are more often using an outpatient setting, we expect that there must be a tendency towards outpatient hysteroscopic removal of polyps for the future. Although further research is required to further assess the efficacy of polyp removal especially with respect to abnormal uterine bleeding, implementation of outpatient hysteroscopy and polyp removal has to be kept under surveillance. Chapter 6 70

72 References 1. Emanuel MH, Verdel MJ, Wamsteker K, Lammes FB. An audit of true prevalence of intrauterine pathology: the hyesteroscopic findings, controlled for patient selection in 1,202 patients with abnormal uterine bleeding. Gynaecol Endosc 1995;4: Epstein E, Ramirez A, Skoog L, Valentin L. Dilatation and curettage fails to detect most focal lesions in the uterine cavity in women with postmenopausal bleeding. Acta Obstet Gynecol Scand 2001;80: Emanuel MH, Wamsteker K, Lammes FB. Is dilatation and curettage obsolete for diagnosing intrauterine disorders in premenopausal patients with persistent abnormal uterine bleeding? Acta Obstet Gynecol Scand 1997;76: Maia H, Jr., Barbosa IC, Farias JP, Ladipo OA, Coutinho EM. Evaluation of the endometrial cavity during menopause. Int J Gynaecol Obstet 1996;52: Chavez NF, Garner EO, Khan W, et al. Does the introduction of new technology change population demographics? Minimally invasive technologies and endometrial polyps. Gynecol Obstet Invest 2002;54: Jansen FW, Vredevoogd CB, van Ulzen K, Hermans J, Trimbos JB, Trimbos-Kemper TC. Complications of hysteroscopy: a prospective, multicenter study. Obstet Gynecol 2000;96: Bettocchi S, Selvaggi L. A vaginoscopic approach to reduce the pain of office hysteroscopy. J Am Assoc Gynecol Laparosc 1997;4: Bettocchi S, Ceci O, Di Venere R, et al. Advanced operative office hysteroscopy without anaesthesia: analysis of 501 cases treated with a 5 Fr. bipolar electrode. Hum Reprod 2002;17: Bettocchi S, Ceci O, Nappi L, et al. Operative office hysteroscopy without anesthesia: analysis of 4863 cases performed with mechanical instruments. J Am Assoc Gynecol Laparosc 2004;11: Clark TJ, Godwin J, Khan KS, Gupta.J.K. Ambulatory endoscopic treatment fo symptomatic benign endometrial polyps: a feasibility study. Gynaecol Endosc 2002;11: Kremer C, Duffy S, Moroney M. Patient satisfaction with outpatient hysteroscopy versus day case hysteroscopy: randomised controlled trial. BMJ 2000;320: Marsh F, Kremer C, Duffy S. Delivering an effective outpatient service in gynaecology. A randomised controlled trial analysing the cost of outpatient versus daycase hysteroscopy. BJOG 2004;111: Marsh FA, Rogerson LJ, Duffy SR. A randomised controlled trial comparing outpatient versus daycase endometrial polypectomy. BJOG 2006;113: Nathani F, Clark TJ. Uterine polypectomy in the management of abnormal uterine bleeding: A systematic review. J Minim Invasive Gynecol 2006;13: Clark TJ, Khan KS, Gupta JK. Current practice for the treatment of benign intrauterine polyps: a national questionnaire survey of consultant gynaecologists in UK. Eur J Obstet Gynecol Reprod Biol 2002;103: Lydeard S. Commentary: avoid surveys masquerading as research. BMJ 1996;313: Survey on hysteroscopic polyp removal 71

73 17. van Dongen H, Kolkman W, Jansen FW. Implementation of hysteroscopic surgery in The Netherlands. Eur J Obstet Gynecol Reprod Biol 2007;132: Rogerson LJ, Duffy S. A national survey of outpatient hysteroscopy. Gynaecol Endosc 2001;10: Chapter 6 72

74 7 What is the recurrence rate of postmenopausal bleeding in women who have a thin endometrium during a first episode of postmenopausal bleeding? H.C. van Doorn, A. Timmermans, B.C. Opmeer, R.F.M.P. Kruitwagen, F.P.H.L.J. Dijkhuizen, G.S. Kooi, P.H.M. van de Weijer, B.W.J. Mol (Acta Obstet Gynecol Scand 2008; 87: 89-93)

75 Abstract Chapter 7 Objective: To determine the incidence and significance of recurrent postmenopausal bleeding (PMB) among women who were diagnosed with an endometrial thickness 4 mm after a first episode of PMB. Methods: Consecutive patients not using hormone replacement therapy (HRT) presenting with a first episode of PMB and an endometrial thickness 4 mm at transvaginal ultrasound (TVU) were managed expectantly. In case of recurrent bleeding the patient was evaluated according to hospital s local policy with TVU, office endometrial sampling, hysteroscopy or dilatation and curettage (D&C) or a combination of these tests. We evaluated the incidence of recurrent bleeding, potential risk factors for recurrent bleeding and the diagnosis that was made after recurrent bleeding. Results: We registered 607 patients with a first episode of PMB, of whom 249 had an endometrial thickness 4 mm. Follow-up took place with a median of 174 weeks (range 4 to 250 weeks). During follow-up 25 of the 249 (10%; 95% CI 6.6 to 14%) patients had recurrent bleeding. Median time until recurrence of bleeding was 49 weeks (range 9 to 186 weeks). Two patients with recurrent bleeding turned out to have an endometrial carcinoma (8%; 95% CI 2.2% to 25%) and one patient had a malignant melanoma. Time since menopause, age, body mass index, hypertension, diabetes and anticoagulants were not predictive for recurrent bleeding. Conclusion: The recurrence rate after a first episode of PMB managed expectantly is low and can not be predicted by patient s characteristics. Patients with recurrent bleeding should be re-evaluated, as they bear a considerable risk of carcinoma. 74

76 Introduction Postmenopausal bleeding (PMB) is frequently caused by abnormalities of the endometrium, either benign or malignant. At present, transvaginal ultrasound (TVU) is used as a first step in the evaluation of women with PMB. 1-4 The probability of malignancy is reduced to < 1% in cases of an endometrial thickness 4 mm. 2,5,6 Below this cut-off level most guidelines do not recommended endometrial sampling, whereas above this cut-off level endometrial sampling is warranted to rule out malignancy. 1,3,4,7,8 Cost-effectiveness analyses show a diagnostic strategy starting with TVU followed by endometrial sampling in case of a thickened endometrium (> 4 mm) was the most cost-effective strategy. 9,10 An important assumption in such a strategy is that the recurrence rate of PMB in women without malignancy is low, whereas women with a malignancy who are initially missed (false negatives) experience recurrent or persistent bleeding. Few studies address the long-term outcome of conservative management of women with a first episode of PMB and endometrial thickness 4 mm. A randomized trial in which women with an endometrial thickness < 5 mm were either allocated to dilatation and curettage (D&C) or expectant management, showed that the recurrence rate of bleeding was 21% after D&C versus 33% after expectant management during a 12-month follow-up period (p = 0.17). 11 In another study, the recurrence rate of bleeding after a first episode endometrial thickness 4.4 mm was reported to be 13% after a median time of 13 months. 12 Gull et al. 13 reported a recurrence rate of 6% after one year follow-up and 17% when these patients were followed 10 years. 8 None of these studies diagnosed endometrial carcinoma during follow-up, although in the second study by Gull et al. 8, two cases of atypical hyperplasia were diagnosed. None of these studies examined risk factors for recurrent bleeding. 8,13 Recurrent PMB after thin endometrium 75 With respect to the incidence and significance of PMB in women with an initial endometrial thickness 4 mm, several issues are at stake. If the recurrence rate was low, then a policy with TVU would be sufficient. If, however, the recurrence rate was high or if malignancy was diagnosed at follow-up, one could advocate that endometrial sampling should be applied immediately. Furthermore, if it was possible to know what patients would experience recurrent bleeding, one could apply a policy based on individual patient characteristics. We hypothesized that the recurrence rate among women with normal findings at the initial work-up would be low. To answer these questions, we performed a prospective cohort study among women with a first episode of PMB and endometrial thickness 4 mm.

77 Materials and Methods The study was performed in a university hospital and seven university-affiliated teaching hospitals in The Netherlands (listed at the end of this article). Between January 2002 and June 2003 consecutive patients who presented with PMB were registered prospectively. During this study period all women with PMB were registered prospectively. The study was limited to patients with a first episode of PMB, not using hormone replacement therapy (HRT) and who had an endometrial thickness 4 mm as measured with TVU. We recorded data regarding endometrial thickness, body mass index (BMI), relevant medical history, anticoagulant therapy and co-morbidity. Patients were evaluated according to the guideline of the Dutch Society for Obstetrics and Gynaecology, which recommends starting the work-up with TVU measurement of endometrial thickness. 3 The endometrial thickness was measured from a longitudinal sonogram through the thickest area of the endometrium, and from the outermost borders of the endometrium. All measurements were done with callipers on a frozen ultrasound image. In case it was not possible to measure the endometrial thickness in a reliable way, this was recorded. When the endometrial layers were separated by intracavitary fluid, each layer was measured and the sum recorded. Chapter 7 76 In case of endometrial thickness > 4 mm or endometrial thickness not measurable, endometrial sampling was performed; these groups were not included in this study. In case of an endometrial thickness 4 mm expectant management was recommended. Some gynaecologists performed endometrial sampling either with an office endometrial sampling device or with hysteroscopy and/or D&C, despite an endometrial thickness 4 mm. Follow-up started on completion of the initial work-up. If endometrial sampling was performed, only patients with benign histology were included for follow-up. Patient were instructed to contact their gynaecologist again in case the bleeding recurred. The primary endpoints were incidence of recurrent bleeding and time to recurrent bleeding. In case of recurrent bleeding, the patient was evaluated according to the protocol of the local hospital. This work-up contained different combinations of test (TVU alone, office endometrial sampling, hysteroscopy, D&C). We registered both the type of test(s) used at recurrent bleeding as well as the final diagnosis. From November 2005 until February 2006 hospital records and patients charts were systematically reviewed to assess if the patient had returned to the hospital with recurrent bleeding. Since patients were instructed to contact the hospital in case of recurrent bleeding, we assumed that if the patients had not contacted the hospital, they had not experienced recurrent bleeding. Time to recurrence of bleeding was censored by

78 this date of chart review if the patient had not contacted the hospital. If the patient had undergone a hysterectomy during the follow-up period for other indications (prolapse surgery) or if the patient had deceased, this date was taken for censoring. Statistical analysis We used Kaplan-Meier analysis to assess time to recurrent bleeding. The prognostic value of potential indicators (age, BMI, anticoagulants, co-morbidity and time since menopause) was evaluated using log-rank statistics. A p-value of 0.05 was considered to indicate statistical significance. If log-rank statistics indicated statistical significance and the risk appeared to be proportionally constant over time, Cox regression analysis was performed and Hazard Rate Ratio s (HRR) were calculated. The diagnosis at recurrent bleeding was also evaluated. Patients in whom an office endometrial sampling, hysteroscopy and/or D&C had been performed at initial work-up were compared with patients who had undergone only TVU at initial work-up. Calculations were performed with SPSS 12.0 (SPSS Inc., Chicago, IL., USA). Table 1. Patients characteristics All patients TVU alone (N=181) TVU and OES (N=47) TVU and Hyst/ D&C (N=21) Mean age (years) (SD) 61.8 (9.6) 61.7 (9.7) 61.4 (9.8) 62.2 (8.8) Time since menopause (years) (SD) Mean endometrial thickness (mm) (SD) 11.4 (9.9) 11.6 (10.5) 10.7 (8.8) 11.7 (7.8) 2.4 (0.96) 2.2 (0.9) 2.8 (1.0) 2.8 (1.1) Mean BMI (kg/m 2 ) (SD) 26.5 (5.0) 26.3 (5.1) 27.5 (5.0) 27.6 (5.0) Recurrent PMB after thin endometrium 77 Diabetes Mellitus (%) Hypertension (%) Anticoagulants users (%) TVU, transvaginal ultrasound; OES, office endometrial sampling; Hyst/D&C, hysteroscopy and/or dilatation and curettage

79 Results We registered 607 patients with a first episode of PMB, of whom 249 patients had an endometrial thickness 4 mm at TVU. The characteristics of the patients are presented in Table 1. In 181 (73%) patients only TVU was performed. At the initial work-up 52 (21%) patients had an office endometrial sampling. Five of these patients also underwent a hysteroscopy and/or D&C, whereas 16 other patients had a hysteroscopy and/or D&C but not an office endometrial sampling performed at their initial work-up. There were no malignancies diagnosed in the patients with endometrial sampling and/or hysteroscopy/d&c. Figure 1. Kaplan-Meier curve showing time until recurrent bleeding women with recurrent bleeding (%) Chapter time to recurrent bleeding (weeks) no. of patients at risk remaining

80 Median duration of follow-up was 174 weeks (range: 4 to 250 weeks, interquartile range: 157 to 188 weeks). During follow-up, 25 (10%; 95% CI 6.6 to 14%) patients contacted the clinic with recurrent bleeding. Median time until recurrence of bleeding was 49 weeks (interquartile range: 26 to 112 weeks) (Figure 1). None of the patient characteristics was associated with incidence of recurrent bleeding or with time to recurrent bleeding (Table 2). There was no statistically significant difference with respect to incidence of recurrent bleeding or time to recurrence between patients with TVU alone at initial work-up (15/181) and patients in whom an office endometrial sampling had been performed at the initial work-up (5/47) (log rank statistics 0.19, p=0.66). However, there was a statistically significant difference in recurrent bleeding between patients with hysteroscopy and/or D&C at initial work-up (5/21) and patients without hysteroscopy and/or D&C (20/208) (p=0.04). Therefore, Cox regression analysis was performed and a HRR was calculated as 2.8 (95% CI 1.05 to 7.5). Table 2. Log rank statistics for time to recurrent bleeding for potential indicators of recurrent bleeding Patient s characteristic Log rank p-value BMI > Age > Hypertension Diabetes Anticoagulants use TMP < 3 years Recurrent PMB after thin endometrium TMP > 7 years TMP > 15 years* TMP, time since menopause Astrup 14 : higher risk of bleeding first 3 years after menopause Van Doorn 15 : higher risk of cancer more than 7 years after menopause * Bruchim 16 : higher risk of cancer more than 15 years after menopause Table 3 shows the findings at the time of recurrent bleeding. In two patients (8%; 95% CI 2.2 to 25%) with recurrent bleeding an endometrial carcinoma was diagnosed, and in a third patient a malignant melanoma of the vagina was found. One patient with subsequent endometrial cancer had only undergone TVU at the initial work-up, whereas in the other patient a representative office endometrial sample was obtained as well, showing benign histology. Time to recurrent bleeding in these two patients with endometrial carcinoma was 44 weeks and 16 weeks respectively.

81 Table 3. Findings at recurrent bleeding in 25 women with initially an endometrial thickness 4 mm Findings No. (%) Endometrial thickness 4 mm 7 (29) Benign endometrium 10 (38) Endometrial carcinoma 2 (8.3) Malignant melanoma 1 (4.1) Other 5 (21) Total 25 Discussion This study shows that recurrent bleeding in patients with PMB and endometrial thickness 4 mm occurs in only 10% of the patients. The median time to recurrent bleeding was approximately one year. We could not identify patient factors associated with recurrent bleeding, nor did we observe a difference in occurrence of recurrent bleeding for patients with TVU alone at initial work-up and patients with TVU and subsequent office endometrial sampling. However, patients with hysteroscopy and/or D&C at initial work-up were at higher risk for recurrent bleeding than patients without. Chapter 7 80 A potential limitation of this study is that patients were not systematically contacted for follow-up. Although patients were clearly instructed to contact the hospital in case of recurrent bleeding, it might be possible that patients experienced recurrent bleeding but did not contact the hospital or were evaluated in another hospital. Therefore the true incidence of recurrent bleeding might be underestimated. However, we feel that our follow-up was reliable, as patients were clearly instructed. The incidence of recurrent bleeding in our study is in accordance with the incidence of 6% to 33% previously reported. 8,11-13 Follow-up in these studies also took place by reviewing medical records, hospital registries and regional registries. Differences in study design, duration of follow-up, mean age, and frequency of HRT use might explain the variations in recurrent bleeding between these studies. In our study patients using HRT were excluded, whereas in the other studies HRT use was used by 35% to 68% of the patients. 8,11,13 Furthermore, follow-up in the two largest studies took place for a minimum of 2.9 years to over 10 years, compared to minimum follow-up of 4 weeks in our study. 8,12

82 We found no difference in recurrence rates between patients who had undergone TVU alone and patients who had both TVU and office endometrial sampling at the initial work-up. Epstein et al. 11 reported a lower recurrent bleeding rate in a group initially managed with D&C when compared to a group managed with TVU alone. However, the difference observed was not statistically significant, possibly explained by a lack of power of that study. In contrast, in our study the group of patients with hysteroscopy and/or D&C at initial work-up had more frequent recurrent bleeding than the group of patients without such tests. This would argue for restrictive diagnostic management of women presenting with PMB and thin endometrium. In summary, the recurrence rate of PMB after an initial reassuring TVU was low and not related to patients characteristics. In our opinion, this is supportive for the use of TVU as a first test in the work-up for PMB. However, when recurrent bleeding occurred, a substantial number of patients was found to have malignancy. Therefore, patients managed expectantly should be instructed to contact their gynaecologist at recurrent bleeding and repeated gynaecological examination and subsequent endometrial sampling is warranted at that time. Acknowledgements Participating hospitals A.P.M. Heintz MD, PhD, University Medical Centre Utrecht, Utrecht, The Netherlands R.F.M.P. Kruitwagen MD, PhD, TweeSteden Hospital, Tilburg, The Netherlands F.P.H.L.J. Dijkhuizen MD, PhD, Rijnstate Hospital, Arnhem, The Netherlands G.S. Kooi MD, PhD, Albert Schweitzer Hospital, Dordrecht, The Netherlands M.V.A.M. Kroeks MD, PhD, Diakonessenhuis, Utrecht, The Netherlands P.H.M. van de Weijer MD, PhD, Gelre Hospital, Apeldoorn, The Netherlands M.J. Duk MD, PhD, Meander Medical Centre, Amersfoort, The Netherlands A.M. Bouwmeester MD, Mesos Medical Centre, Utrecht, The Netherlands Recurrent PMB after thin endometrium 81

83 References Chapter Epstein E. Management of postmenopausal bleeding in Sweden: a need for increased use of hydrosonography and hysteroscopy. Acta Obstet Gynecol Scand 2004;83: Gupta JK, Chien PF, Voit D, Clark TJ, Khan KS. Ultrasonographic endometrial thickness for diagnosing endometrial pathology in women with postmenopausal bleeding: a meta-analysis. Acta Obstet Gynecol Scand 2002;81: NVOG (Dutch Society of Obstetrics and Gynaecology). NVOG-Richtlijn Abnormaal vaginaal bloedverlies in de menopauze [in Dutch]. (NVOG Guideline: Abnormal vaginal bleeding during menopause) Scottish Intercollegiate Guidelines Network. Investigation of postmenopausal bleeding. 1st edn edinburgh, UK: Scottisch Intercollegiate Guidelines Network, Royal College of Physicians 2002; (www. sign.ac.uk). 5. Smith-Bindman R, Kerlikowske K, Feldstein VA, et al. Endovaginal ultrasound to exclude endometrial cancer and other endometrial abnormalities. JAMA 1998;280: Tabor A, Watt HC, Wald NJ. Endometrial thickness as a test for endometrial cancer in women with postmenopausal vaginal bleeding. Obstet Gynecol 2002;99: Goldstein RB, Bree RL, Benson CB, et al. Evaluation of the woman with postmenopausal bleeding: Society of Radiologists in Ultrasound-Sponsored Consensus Conference statement. J Ultrasound Med 2001;20: Gull B, Karlsson B, Milsom I, Granberg S. Can ultrasound replace dilation and curettage? A longitudinal evaluation of postmenopausal bleeding and transvaginal sonographic measurement of the endometrium as predictors of endometrial cancer. Am J Obstet Gynecol 2003;188: Clark TJ, Barton PM, Coomarasamy A, Gupta JK, Khan KS. Investigating postmenopausal bleeding for endometrial cancer: cost-effectiveness of initial diagnostic strategies. BJOG 2006;113: Dijkhuizen FP, Mol BW, Brolmann HA, Heintz AP. Cost-effectiveness of the use of transvaginal sonography in the evaluation of postmenopausal bleeding. Maturitas 2003;45: Epstein E, Valentin L. Rebleeding and endometrial growth in women with postmenopausal bleeding and endometrial thickness < 5 mm managed by dilatation and curettage or ultrasound follow-up: a randomized controlled study. Ultrasound Obstet Gynecol 2001;18: Epstein E, Jamei B, Lindqvist PG. High risk of cervical pathology among women with postmenopausal bleeding and endometrium <or=4.4 mm: long-term follow-up results. Acta Obstet Gynecol Scand 2006;85: Gull B, Carlsson S, Karlsson B, Ylostalo P, Milsom I, Granberg S. Transvaginal ultrasonography of the endometrium in women with postmenopausal bleeding: is it always necessary to perform an endometrial biopsy? Am J Obstet Gynecol 2000;182: Astrup K, Olivarius Nde F. Frequency of spontaneously occurring postmenopausal bleeding in the general population. Acta Obstet Gynecol Scand 2004;83:203-7.

84 15. van Doorn HC, Opmeer BC, Burger CW, et al. Inadequate office endometrial sample requires further evaluation in women with postmenopausal bleeding and abnormal ultrasound results. Int J Gynaecol Obstet 2007;99: Bruchim I, Biron-Shental T, Altaras MM, et al. Combination of endometrial thickness and time since menopause in predicting endometrial cancer in women with postmenopausal bleeding. J Clin Ultrasound 2004;32: Recurrent PMB after thin endometrium 83

85

86 8 Follow-up of women after first episode of postmenopausal bleeding and endometrial thickness > 4 mm A. Timmermans, H.C. van Doorn, B.C. Opmeer, M.V.A.M. Kroeks, M.J. Duk, A.M. Bouwmeester, R.F.M.P. Kruitwagen, F.P.H.L.J. Dijkhuizen, B.W.J. Mol (Obstet Gynecol 2008; 111: )

87 Abstract Objective: To estimate the incidence of recurrent postmenopausal bleeding (PMB) among women who were diagnosed with an endometrial thickness > 4 mm. Methods: We designed a prospective cohort study and included consecutive women not using hormone replacement therapy, presenting with a first episode of PMB. We evaluated patients who had an endometrial thickness > 4 mm at transvaginal ultrasound (TVU) and benign endometrial sampling; presence of carcinoma was ruled out by office endometrial sampling, hysteroscopy and/or dilatation and curettage (D&C). Time until recurrent bleeding was measured and diagnosis at recurrent bleeding was recorded. Results: Among 318 patients who had an endometrial thickness > 4 mm, 222 patients had benign histology results and were available for follow-up. During follow-up 47 (21%) patients had recurrent bleeding with a median time to recurrent bleeding of 49 weeks (interquartile range 18 to 86 weeks). There was no difference with respect to recurrence rate between patients with polyp removal, patients with a normal hysteroscopy, and patients with office endometrial sampling alone at the initial work-up. Two patients were diagnosed with atypical endometrial hyperplasia upon recurrent bleeding. Conclusion: The recurrence rate of PMB in women with endometrial thickness > 4 mm is 21%. This recurrence is not related to incorporation of hysteroscopy or polyp removal at the initial work-up. Chapter 8 86

88 Introduction Postmenopausal bleeding (PMB) is often caused by abnormalities of the endometrium, being either benign or malignant. Dilatation and curettage (D&C) have been used for the diagnosis of endometrial carcinoma in women with PMB. At present, transvaginal ultrasound (TVU) is used as a first step in the evaluation of women with PMB. 1-5 The probability of malignant pathology is strongly reduced in cases wherein the endometrial thickness is less than 5 mm. 2,6,7 Above this cut-off level, endometrial sampling is warranted to rule out malignancy. 2,6 Office endometrial sampling is a minimally invasive alternative to D&C. It has been shown to be highly accurate in the detection of endometrial carcinoma in women with PMB with a sensitivity of 99.6%. 8 After malignancy has been ruled out, expectant management can be recommended. 3,4 Alternatively, hysteroscopy or Saline Infusion Sonography (SIS) can be used to assess the uterine cavity for the presence of benign pathology, mostly endometrial polyps. 1,3,4,9,10 Hysteroscopy or SIS can be incorporated in the diagnostic work-up at first episode of bleeding or in case the bleeding persists or recurs. 1,3,4 Little is known about the recurrence of bleeding in women with PMB and endometrial thickness > 4 mm after initial benign sampling. Feldman et al. 11 reported that 33% of the patients with PMB and initial benign sampling underwent another sampling during a follow-up of 2 years, but endometrial thickness was not measured in that study. Gull et al. 12 reported a recurrence of bleeding of 40% during 10 years follow-up in patients with endometrial thickness > 4 mm and initial benign sampling after a first episode of PMB. This issue raises several questions. First, what is the recurrence rate of bleeding after a first episode of PMB, in patients who have an endometrial thickness > 4 mm and benign histology? Second, does the recurrence rate depend on the performance and outcome of the initial diagnostic work-up; that is, do women with a hysteroscopy and/or D&C at the initial work-up of PMB experience less recurrent bleeding than patients with office endometrial sampling only? Third, is malignancy present in women with recurrent bleeding? If the recurrence rate would be low, then a policy with endometrial sampling would be sufficient. If, however, the recurrence rate would be high, or if malignancy would be diagnosed at follow-up, one could advocate that hysteroscopy and/ or D&C should be applied immediately. To answer these questions, a prospective cohort study was performed among women not using hormone replacement therapy (HRT) with a first episode of PMB, endometrial thickness > 4 mm and benign histology. We hypothesized that the recurrence rate among women with normal findings at the initial work-up would be low. Recurrent PMB after thickened endometrium 87

89 Materials and Methods The study was performed in one university hospital and seven university-affiliated teaching hospitals in the Netherlands. Between January 2002 and June 2003 consecutive patients who presented with a first episode of PMB at the outpatient department were registered prospectively. The study was limited to women not using HRT who had an endometrial thickness > 4 mm as measured with TVU. Data on body mass index (BMI), anticoagulant therapy, comorbidity, endometrial thickness and histology sampling, were recorded. Patients were evaluated according to the guideline of the Dutch Society of Obstetrics and Gynaecology, which starts the work-up with measurement of endometrial thickness by TVU. 3 The thickness of the endometrium was measured from a longitudinal sonogram through the thickest area of the endometrium, measuring the outermost borders of the endometrium. All measurements were done with callipers on a frozen ultrasound image. Measuring the endometrium in postmenopausal women can be difficult due to an upright position of the uterus, vessel calcifications, and a diffuse endometrial-myometrial border. Cases where it was not possible to measure the endometrial thickness in a reliable way were recorded. When the endometrial layers were separated by intracavitary fluid, both layers were measured and the sum recorded. Chapter 8 88 In patients with an endometrial thickness > 4 mm and in patients in whom endometrial thickness was not measurable, endometrial sampling was performed either with an office endometrial sampling device, hysteroscopy with guided biopsies, D&C or a combination of these tests. Although the national Dutch guideline states that office endometrial sampling alone is sufficient, individual doctors could decide for subsequent testing by hysteroscopy alone or D&C or hysteroscopy combined with D&C. 3 Hysteroscopy and/or D&C were also performed in case office endometrial sampling showed insufficient material for diagnosis. Hysteroscopy and/or D&C were performed as is customary in the residential hospital, which could be in an inpatient setting under general anaesthesia or in an outpatient setting with local anaesthesia. Patients whose histology results showed malignancy or atypical hyperplasia, were scheduled for further treatment and not included in the follow-up. In cases of benign histology results expectant management was recommended. The primary endpoint was recurrent bleeding. Each patient was instructed to contact her gynaecologist in case the bleeding recurred. In case of recurrent bleeding, the patient was evaluated according to the protocol of the local hospital. From November 2005 until February 2006 hospital records and patients charts were systematically reviewed to assess recurrence of bleeding. Since patients were instructed to contact the hospital in case of recurrent bleeding, it was assumed that if the patients had not contacted the hospital, they had not experienced recurrent bleeding. Time to recurrence of

90 bleeding was censored by this date of chart review if the patient had not contacted the hospital. If the patient had undergone a hysterectomy during the follow-up period for other indications (prolapse surgery) or if the patient had deceased, this date was taken for censoring. Statistical analysis Time to recurrent bleeding was assessed using Kaplan-Meier analysis. Subsequently, we evaluated whether the performance of a hysteroscopy and/or D&C, and the performance of polypectomy at the initial work-up were associated with recurrence of bleeding. The log-rank statistic was used to test for statistical significance. A p-value of < 0.05 was considered to indicate statistical significance. If differences were found to be proportional over time, Cox regression analysis was performed and a Hazard Rate Ratio (HRR) was calculated. Calculations were performed with SPSS 12.0 (SPSS Inc., Chicago, IL., USA). Results We registered 607 patients with a first episode of PMB, of whom 270 patients had an endometrial thickness > 4 mm and 48 patients had an endometrial thickness that could not be measured. Histology results showed endometrial carcinoma in 60 patients and atypical endometrial hyperplasia in eight patients. Five patients were diagnosed with other malignancies (three breast cancer metastases, one bladder carcinoma and one cervical cancer). Four patients were treated for squamous intraepithelial lesions of the cervix and one patient died shortly after initial work-up from a cause not related to her PMB. Seven patients underwent a hysterectomy for prolapse surgery shortly after their first visit. As a consequence, 233 patients were included in the present study. Of these 233 patients, 11 patients were lost to follow-up, as hospital records were not available or lost; these patients could not be included in the analysis. Patients characteristics are summarized in Table 1. From the 222 patients in whom followup was known, 107 patients had had office sampling during the initial work-up and 115 patients had had a hysteroscopy with or without a D&C. In 49 of the patients undergoing hysteroscopy endometrial polyp(s) had been diagnosed and removed (Figure 1). Recurrent PMB after thickened endometrium 89

91 Table 1. Patients characteristics of patients with first episode of postmenopausal bleeding, not on HRT, endometrial thickness > 4 mm or not measurable and benign histology All patients (N=222) OES (N=107) Hyst/D&C (N=115) p-value Mean age (years) 60.9 (±9.7) 59.4 (±8.6) 63.1 (±9.7) ns Mean time since menopause (years) 10.8 (±10.6) 9.2 (±9.0) 13.0 (±11.2) 0.01 Mean endometrial thickness (mm) 9.8 (±5.2) 9.6 (±4.6) 11.1 (±6.2) Mean BMI (kg/m 2 ) 29.7 (±6.7) 29.5 (±6.3) 30.4 (±7.1) ns Diabetes Mellitus Hypertension ns Anticoagulants users OES, office endometrial sampling; Hyst/D&C, hysteroscopy and/or dilatation and curettage; BMI, body mass index; ns: not significant Data are mean (±standard deviation) or % Figure 1. Flow chart of women with postmenopausal bleeding and endometrial thickness > 4 mm Chapter 8 First episode of PMB with endometrial thickness > 4 mm N = Benign diagnosis after initial work-up N = 222 Hospital records lost N = 11 Not included for follow-up (malignancy after initial work-up or hysterectomy) N = 85 Office endometrial sampling at initial work-up N = 107 Recurrent bleeding N = 24 Hysteroscopy and/or dilatation and curettage at initial work-up N = 115 Recurrent bleeding N = 23 Endometrial polyp at initial work-up N = 49

92 Median duration of follow-up was 176 weeks (interquartile range 156 to 194 weeks, range 50 to 260 weeks). During follow-up 47 of the 222 (21%; 95% CI: 16-27%) patients had recurrent bleeding (Figure 1). Median time until recurrent bleeding was 43 weeks (interquartile range 18 to 86 weeks) (Figure 2). Figure 2. Kaplan-Meier analysis for recurrent postmenopausal bleeding risk of recurrent bleeding time to recurrent bleeding (weeks) Recurrent PMB after thickened endometrium 91 Solid line, survival function; cross, censored. There were no statistically significant differences in patient characteristics in the recurrent bleeding group when compared with the group of patients without recurrent bleeding. In the group of patients with hysteroscopy and/or D&C 23 of 115 (20%; 95% CI: 14-28%) patients experienced recurrent bleeding and this was 24 of 107 (22%; 95% CI: 16-30%) in the group of patients undergoing office sampling. There was no statistically significant difference with respect to incidence of recurrent bleeding or time to recurrent bleeding between patients with hysteroscopy and/or D&C at initial work-up or patients without (Figure 3), nor was there a difference between patients in whom an endometrial polyp had been identified and removed at initial

93 Figure 3. Kaplan-Meier analysis for recurrent postmenopausal bleeding risk of recurrent bleeding Chapter time to recurrent bleeding (weeks) Hysteroscopy or dilatation and curettage at work-up: Solid line, no; dashed line, yes; cross, censored. work-up (Figure 4) and patients in whom such an endometrial polyp was not diagnosed. There were no statistically significant differences in patient characteristics between patients with hysteroscopy and/or D&C at initial work-up and recurrent bleeding and patients with office endometrial sampling and recurrent bleeding, nor was there such a difference for patients with polyp diagnosis and removal. The subsequent diagnostic process and findings were evaluated for patients with recurrent bleeding. Patients with a hysteroscopy performed at recurrent bleeding had a shorter time to recurrent bleeding (median 32 weeks interquartile range 18 to 66 weeks) than patients in whom hysteroscopy at recurrence was not performed (median 64 weeks, interquartile range 49 to 180 weeks) (Log rank statistics 8.13; p=0.004). Cox regression analysis showed a Hazard Rate Ratio of 2.8 (95% CI ).

94 Figure 4. Kaplan-Meier analysis for recurrent postmenopausal bleeding risk of recurrent bleeding time to recurrent bleeding (weeks) Endometrial polyp at initial work-up: Solid line, no; dashed line, yes; cross, censored. Recurrent PMB after thickened endometrium 93 At recurrent bleeding two of 47 (4.3%; 95% CI: %) patients were diagnosed with atypical endometrial hyperplasia. The first patient was a 53-year old multipara with diet-regulated diabetes, BMI of 43, and 1.6 years postmenopausal. Office endometrial sampling at initial bleeding episode had shown benign histology; hysteroscopy was also performed at initial work-up, during which an endometrial polyp had been removed, also with benign histology. Time to recurrent bleeding was 49 weeks, and atypical hyperplasia was diagnosed with office endometrial sampling. The second patient was a 69-year-old multipara with no history of diabetes or hypertension and BMI of 32, and she was 16 years postmenopausal. Hysteroscopy was performed at initial work-up, during which an endometrial polyp was removed, with benign histology. Time to recurrent bleeding was 3.5 years in this patient, and atypical hyperplasia was diagnosed with office endometrial sampling.

95 In 13 patients an endometrial polyp was diagnosed at recurrent bleeding. Seven of these patients had not had a hysteroscopy at initial work-up, in two patients the hysteroscopy at the initial work-up revealed a normal uterine cavity, and the other four patients had had an endometrial polyp removed at the primary work-up. The 32 other patients with recurrent bleeding had benign office endometrial sampling or no intracavitary pathology at hysteroscopy. Discussion This study shows that recurrent bleeding in patients with PMB and endometrial thickness > 4 mm occurred in 21% of the patients. The median time to recurrent bleeding was approximately one year. There was no difference in recurrent bleeding for patients with hysteroscopy and/or D&C at initial work-up and patients with office endometrial sampling, nor was there a difference between patients with an endometrial polyp diagnosed and treated at initial workup and patients without a polyp. During follow-up, two (4.3%) endometrial pre-malignancies were found among patients with recurrent bleeding. Chapter 8 94 A potential limitation of this study is the fact that patients were not systematically contacted for follow-up. Although patients were instructed to contact the hospital in case of recurrent bleeding, it might be possible that patients experienced recurrent bleeding and did not contact the hospital or were evaluated in another hospital. Therefore, the true incidence of recurrent bleeding might be underestimated. However, it seems unlikely that patients would not contact the hospital in case of recurrent bleeding, especially since it is known that patients with PMB, would like as much certainty as possible in the diagnostic process. 13 Furthermore, the Dutch guideline for general practitioners advises to refer patients with PMB to a gynaecologist. 14 Therefore, even if the patient would contact her general practitioner instead of her gynaecologist with recurrent bleeding, it would have been highly unlikely if the patient had not been referred to the gynaecologist. Another potential limitation of this study is the fact that initial endometrial sampling procedure was based on clinical choice of the physician. Patients with higher risk profile or more suspicious ultrasound findings (Table 1) may have tended to be treated initially with hysteroscopy rather than office endometrial sampling. However, at recurrent bleeding, patient characteristics were not statistically significantly different between the hysteroscopy group and the office endometrial sampling group; therefore, we think that this did not influence our results. The found absence of significant differences in the recurrent bleeding could be explained by the limited power of the current study. Post-hoc power analyses indicated, however, that we would have needed a sample sizes of approximately 1,400 women to detect a reduction in the 5-year recurrence rate from 90% to 85% with an α error of 5% and a β error of 20%.

96 The incidence of recurrent bleeding in our study (21%) is lower than the incidence found in the study of Gull et al. 12 (40%).This difference might be explained by the fact that women on HRT were not included in our study whereas they were included in the study by Gull et al. 12 where 35% of the women used HRT. An additional explanation for this difference might be the fact that the follow-up in our study was shorter than in the study by Gull et al. 12 as in that study patients were followed for more than 10 years. In both our study as well as in the study by Gull et al. 12 follow-up took place by reviewing patients medical records and hospital registries. In cases where malignant pathology has been excluded by office endometrial sampling without performing hysteroscopy, significant benign pathology, that is endometrial polyps, may have been overlooked. In patients with PMB and an endometrial thickness > 4 mm the prevalence of endometrial polyps has been reported to be 40%. 15,16 Both D&C and office endometrial sampling are known to miss such focal disease. 16,17 Imaging of the distended uterine cavity is required to most accurately diagnose focal lesions such as endometrial polyps and the techniques most often employed are hysteroscopy or SIS. 9 Hysteroscopy has the advantage over SIS of allowing simultaneous removal of polyps at the time of diagnosis, although data regarding the efficacy of polypectomy in treating PMB are scarce. 18 In our study, there was no difference in recurrent bleeding between patients with a hysteroscopy incorporated in the initial work-up and patients without a hysteroscopy, nor was there a difference between patients with polypectomy at initial work-up and patients without a polypectomy. Therefore, with respect to recurrent bleeding symptoms, it is not beneficial for the patient to include a hysteroscopy in the initial work-up. Patients preference analysis showed that if the risk of recurrent bleeding due to benign pathology exceeded 25%, the majority of the patients would prefer a hysteroscopy to diagnose and treat such benign pathology. 13 As the risk of recurrent bleeding in this study was approximately 21%, it therefore seems justified to refrain from uterine cavity evaluation at the initial work-up of PMB. However, a randomized controlled trial comparing polypectomy with expectant management in case of PMB and cost-effectiveness analysis would be needed to fully answer this question. 18 Recurrent PMB after thickened endometrium 95 The recurrence rate after a first episode of PMB and endometrial thickness > 4 mm was estimated to be 21% and not related to incorporation of hysteroscopy or endometrial polyp removal at initial work-up. Therefore, the performance of hysteroscopy at the initial work-up can be questioned. Endometrial sampling has to be undertaken if patients experience recurrent bleeding, because two patients were diagnosed with atypical endometrial hyperplasia at time of recurrent bleeding. Further research should focus on the diagnostic work-up of women with recurrent PMB.

97 Acknowledgements This work was supported by grant from the Healthcare Insurance Board, Amstelveen, The Netherlands, and grant in the VIDI-program of ZonMW, The Hague, The Netherlands. Participating hospitals A.P.M. Heintz MD, PhD, University Medical Centre Utrecht, Utrecht, The Netherlands R.F.M.P. Kruitwagen MD, PhD, TweeSteden Hospital, Tilburg, The Netherlands F.P.H.L.J. Dijkhuizen MD, PhD, Rijnstate Hospital, Arnhem, The Netherlands G.S. Kooi MD, PhD, Albert Schweitzer Hospital, Dordrecht, The Netherlands M.V.A.M. Kroeks MD, PhD, Diakonessenhuis, Utrecht, The Netherlands P.H.M. van de Weijer MD, PhD, Gelre Hospital, Apeldoorn, The Netherlands M.J. Duk MD, PhD, Meander Medical Centre, Amersfoort, The Netherlands A.M. Bouwmeester MD, Mesos Medical Centre, Utrecht, The Netherlands Chapter 8 96

98 References 1. Epstein E. Management of postmenopausal bleeding in Sweden: a need for increased use of hydrosonography and hysteroscopy. Acta Obstet Gynecol Scand 2004;83: Gupta JK, Chien PF, Voit D, Clark TJ, Khan KS. Ultrasonographic endometrial thickness for diagnosing endometrial pathology in women with postmenopausal bleeding: a meta-analysis. Acta Obstet Gynecol Scand 2002;81: NVOG (Dutch Society of Obstetrics and Gynaecology). NVOG-Richtlijn Abnormaal vaginaal bloedverlies in de menopauze [in Dutch]. (NVOG Guideline: Abnormal vaginal bleeding during menopause) Scottish Intercollegiate Guidelines Network. Investigation of postmenopausal bleeding. 1st edn edinburgh, UK: Scottisch Intercollegiate Guidelines Network, Royal College of Physicians 2002;(www. sign.ac.uk). 5. Goldstein RB, Bree RL, Benson CB, et al. Evaluation of the woman with postmenopausal bleeding: Society of Radiologists in Ultrasound-Sponsored Consensus Conference statement. J Ultrasound Med 2001;20: Smith-Bindman R, Kerlikowske K, Feldstein VA, et al. Endovaginal ultrasound to exclude endometrial cancer and other endometrial abnormalities. JAMA 1998;280: Tabor A, Watt HC, Wald NJ. Endometrial thickness as a test for endometrial cancer in women with postmenopausal vaginal bleeding. Obstet Gynecol 2002;99: Dijkhuizen FP, Mol BW, Brolmann HA, Heintz AP. The accuracy of endometrial sampling in the diagnosis of patients with endometrial carcinoma and hyperplasia: a meta-analysis. Cancer 2000;89: Clark TJ. Outpatient hysteroscopy and ultrasonography in the management of endometrial disease. Curr Opin Obstet Gynecol 2004;16: Epstein E, Ramirez A, Skoog L, Valentin L. Transvaginal sonography, saline contrast sonohysterography and hysteroscopy for the investigation of women with postmenopausal bleeding and endometrium > 5 mm. Ultrasound Obstet Gynecol 2001;18: Feldman S, Shapter A, Welch WR, Berkowitz RS. Two-year follow-up of 263 patients with post/perimenopausal vaginal bleeding and negative initial biopsy. Gynecol Oncol 1994;55: Gull B, Karlsson B, Milsom I, Granberg S. Can ultrasound replace dilation and curettage? A longitudinal evaluation of postmenopausal bleeding and transvaginal sonographic measurement of the endometrium as predictors of endometrial cancer. Am J Obstet Gynecol 2003;188: Epstein E, Ramirez A, Skoog L, Valentin L. Dilatation and curettage fails to detect most focal lesions in the uterine cavity in women with postmenopausal bleeding. Acta Obstet Gynecol Scand 2001;80: NHG (Dutch College of General Practitioners). NHG-standaard vaginaal bloedverlies [in Dutch]. (NHG-Practice Guideline: Vaginal bleeding). M28. Recurrent PMB after thickened endometrium 97

99 15. Emanuel MH, Wamsteker K, Lammes FB. Is dilatation and curettage obsolete for diagnosing intrauterine disorders in premenopausal patients with persistent abnormal uterine bleeding? Acta Obstet Gynecol Scand 1997;76: Timmermans A, Veersema S. Office hysteroscopy in women with postmenopausal bleeding: see and treat of endometrial polyps using a Duckbill polyp snare. Gynecol Surg 2004;1: Nathani F, Clark TJ. Uterine polypectomy in the management of abnormal uterine bleeding: A systematic review. J Minim Invasive Gynecol 2006;13: Timmermans A, Opmeer BC, Veersema S, Mol BW. Patients preferences in the evaluation of postmenopausal bleeding. BJOG 2007;114: Chapter 8 98

100 9 Should endometrial polyps be removed in women with postmenopausal bleeding? An assessment of study designs and report of a failed trial A. Timmermans, S. Veersema, T.C. van Kerkvoorde, L.F. van der Voet, B.C. Opmeer, M.Y. Bongers, B.W.J. Mol

101 Abstract Objective: To assess whether diagnosis and treatment of endometrial polyps in patients with a first episode of postmenopausal bleeding (PMB) is useful. Methods: A single-blinded randomized controlled trial (ISRCTN ) was designed. Patients with PMB were scheduled for TVU to measure endometrial thickness. In case the endometrial thickness was > 4 mm, hysteroscopy was scheduled. If during hysteroscopy a benign endometrial polyp was diagnosed, the patient was asked to participate into this trial. After informed consent patients were randomly allocated either to resection of this polyp in the same session or expectant management. Patients were not informed about the result of the randomization (single-blinding). Recurrence of bleeding was determined during a follow-up of six months. We assumed a reduction in recurrent bleeding of 40% and after power calculation showed that we needed 60 patients, 30 patients per arm. Results: The study suffered from lack of recruitment. After 26 months, only four patients had agreed to participate in the trial. Since both patients and clinicians appeared to be reluctant towards randomization when the polyp was already visualized at hysteroscopy, we decided to stop the trial. Conclusion: In our setting a trial, in which patients with PMB and an endometrial polyp at hysteroscopy were randomized to polypectomy or expectant management, turned out not to be feasible. Alternative trial designs may be more successful to answer the question whether removal of benign endometrial polyps in patients with PMB is effective. Chapter 9 100

102 Introduction Postmenopausal bleeding (PMB) is often caused by abnormalities of the endometrium, being either benign or malignant. At present, transvaginal ultrasound (TVU) measurement of endometrial thickness is used as a first step in the evaluation of patients with PMB. 1-5 Below a cut-off level of 4 mm endometrial sampling is not recommended whereas above this cut-off level endometrium sampling is warranted to rule out malignancy. 3,6,7 This strategy is cost-effective over strategies involving initial evaluation with test combinations or hysteroscopy alone. 8,9 After malignancy has been ruled out, expectant management can be recommended. 4,5 Alternatively, hysteroscopy or Saline Infusion Sonography (SIS) can be used to assess the uterine cavity for the presence of benign pathology, mostly endometrial polyps. 1,4,5,10,11 The Dutch guideline on PMB advocates expectant management once malignancy has been excluded by TVU (i.e. endometrial thickness 4 mm) or office endometrial sampling (e.g. Pipelle). 4 Adherence to this guideline has been shown to be fairly good: 2/3 of women presenting with PMB are managed conform this guideline. 12 Only in 10% of women presenting with PMB hysteroscopy was performed whereas office endometrial sampling would have been sufficient. 12 This adherence study demonstrates that in the Netherlands expectant management in women with PMB after exclusion of malignancy is at present generally accepted. Premalignant and malignant disorders of the endometrium are found in about 12% of the patients with PMB. 13,14 In contrast, the prevalence of endometrial polyps in patients with PMB and endometrial thickness of more than 4 mm is estimated to be around 40%. 15,16 Such benign lesions can be diagnosed with hysteroscopy or SIS. Hysteroscopy has the advantage of allowing simultaneous removal of polyps at the time of diagnosis. The question is whether it is effective to advise expectant management after exclusion of endometrial malignancy and withhold uterine cavity evaluation until the bleeding recurs or persists. Trial design for efficacy of polyp removal in PMB 101 One could hypothesize that the removal of endometrial polyps reduces the probability of recurrent bleeding. 17 From that point of view, it would be important to diagnose endometrial polyps, which can then be removed hysteroscopically. At present, high-quality evidence on this subject is lacking, as no studies prospectively compared polypectomy to expectant management. Simple polypectomy is known to lead to subjective improvement in symptoms of bleeding and high satisfaction rates The question that remains to be answered is whether endometrial polyps in patients with PMB should be removed when malignancy has been excluded, in view of recurrent bleeding symptoms. On the one hand, this will depend on the prevalence of benign endometrial polyps in patients with PMB, and whether such polyps can be diagnosed accurately, for example with TVU or with SIS. On the other hand, this depends on whether removal of a benign endometrial polyp improves outcome. To answer this question, we set up a prospective cohort study with an embedded randomized clinical trial comparing

103 polypectomy and expectant management. The trial was stopped due to lack of recruitment after 26 months. With respect to efficacy of polypectomy in women with abnormal uterine bleeding, randomized clinical trials are advocated. 17 Our trial on this subject failed by lack of recruitment and to avoid other investigators who might encounter the same problems, we report the study design and inclusion until discontinuation, discuss explanatory factors for this failure and propose a possible alternative study design. Methods The study was performed between July 2005 and September All patients visiting the outpatient clinic with postmenopausal vaginal bleeding were included in a cohort study. A history was taken to determine the nature of the bleeding (i.e. vaginal bleeding and not rectal bleeding) and to establish whether there had been a single bleeding episode, whether the bleeding persisted or whether the bleeding had occurred more than once. Subsequently, a gynaecological examination was performed including a cervical smear and a TVU. Chapter The thickness of the endometrium was measured from a longitudinal sonogram through the thickest area of the endometrium, and from the outermost border of the endometrium on one side to that on the other side. All measurements were done with callipers on a frozen image. The measurements of endometrial thickness included both layers and any expansive process or fluid in the endometrial cavity. The endometrial measurement was classified in full millimetres. In case the endometrial thickness was 4 mm the patient was reassured and discharged in accordance with the Dutch guideline. 4 In case the endometrial thickness was > 4 mm or if the endometrial thickness was not clearly measurable, an office endometrial sampling (e.g. Pipelle biopsy) was taken to ensure histology, and subsequently in all patients a hysteroscopy was scheduled (Figure 1). Office endometrial sampling has been shown to have a sensitivity of 99.6% in excluding endometrial cancer. 20 In those patients in whom office endometrial sampling is unsuccessful or histology shows insufficient material for diagnosis, hysteroscopy is at present advocated. Therefore the trial design reflects current practice with respect to diagnosis of endometrial cancer or hyperplasia.

104 Figure 1. Flow chart of study design Postmenopausal vaginal bleeding Vaginal examination incl. PAP smear, TVU ET 4 mm ET > 4 mm Pipelle Hysteroscopy No polyp Polyp Randomization Resection Expectant management Trial design for efficacy of polyp removal in PMB 103 Follow-up Follow-up Patients were informed that at present the Dutch guideline advocates exclusion of malignancy by office endometrial sampling. 4 They were informed that the diagnostic work-up of women with PMB does not in general include a hysteroscopy. Furthermore, they were informed that for this trial a hysteroscopy was performed, which was additional to office endometrial sampling in their diagnostic work-up. Patients were then scheduled for hysteroscopy and received information regarding the possible presence of an endometrial polyp. They received written information about the ongoing trial and were asked to consider participation in case an endometrial polyp was detected at hysteroscopy. Before the hysteroscopy patients were asked whether they would agree to participate in the trial if during hysteroscopy an endometrial

105 polyp would be diagnosed, a principle for consent was asked prior to the hysteroscopy. At hysteroscopy, the uterine cavity was systematically assessed. A polyp was defined as a benign, localized overgrowth of endometrial tissue covered by epithelium. Patients with a lesion suspect for malignancy were excluded, as were patients using Tamoxifen (Nolvadex ). If, during hysteroscopy a polyp was diagnosed, the patient was again asked to confirm consent for randomization. Patients who confirmed their consent were randomized to immediate resection of the polyp or expectant management (Figure 1). Assisting operating nurse performed the randomization which took place by a computer through block-randomizations, stratified for one or multiple polyps. Patients were not informed about the result of the randomization (single-blinding). Resection of the endometrial polyp was performed in the same session in which the polyp was diagnosed. Resection was performed following standard procedures: smaller polyps (< 5 mm) were resected using endoscopic forceps and/or scissors; larger polyps (> 5 mm) were resected using a monopolar polyp snare. 21 In patients that were randomized for expectant management a sham resection procedure was performed for five minutes. Chapter The primary outcome measure was recurrence of PMB. Patients were asked to record bleeding on a bleeding chart. We planned to assess time to recurrence of PMB using Kaplan-Meier analysis and to test for significance with the log-rank statistic. Beforehand we assumed the recurrence of PMB in case of an endometrial polyp to be 60% after 6 months. In case a polyp was resected, we expected the probability of recurrence of bleeding to be 20%. In view of these assumptions, and using a two-sided test with conventional characteristics (alpha-error 5%, beta-error 20%), we needed 60 patients to be randomized to the two arms of the study (i.e. 30 patients per arm). The trial was registered in the clinical trial register (ISRCTN ) and approved by the institutional review board of the St. Antonius Hospital in Nieuwegein, The Netherlands (registration number TME/C.04.08, date of approval 17 th December 2004). All gynaecologists working in the outpatient department expressed their willingness to participate in this trial. During several meetings they were informed about the trial and instructed on how to counsel eligible patients at the outpatient clinic. All hysteroscopies were performed by three gynaecologists who were all members of the scientific committee of this trial.

106 Results The trial started on July 15 th On September 15 th 2007 the trial was stopped due to lack of recruitment. From the start onwards, the study suffered from lack of recruitment. In the study period, 361 patients visited the outpatient clinic with PMB (Figure 2). We performed 255 outpatient hysteroscopies in patients with PMB. Two patients met the exclusion criteria and seven patients were not found suitable by their doctor. Of the remaining 246 patients, 105 patients were informed about the trial before hysteroscopy. Hysteroscopy showed an endometrial polyp in 94 patients. Of these, 43 (46%) patients were informed about the trial and 4 out of these 43 (9%) patients gave informed consent (Table 1). These four patients were all randomized to resection. Of the 94 patients with endometrial polyp(s), 67 patients underwent immediate resection of the polyp in the same session, in the remaining 27 the endometrial polyp was not resected in that session. Table 1. Findings at hysteroscopy related to informed consent for randomization Findings at hysteroscopy No. Asked Informed consent Randomization No intracavitary pathology (53%) 13 (32%) NA Atrophy 18 6 (33%) 1 (17%) NA Hypertrophy 4 0 (0%) 0 (0%) NA Endometrial polyp (46%) 4 (9%) 4 Myoma 12 2 (17%) 2 (100%) NA Endometrial carcinoma 14 2 (14%) 1 (50%) NA Synecchia 1 1 (100%) 0 (0%) NA Trial design for efficacy of polyp removal in PMB 105 Procedure stopped (29%) 2 (20%) NA NA: not applicable

107 Figure 2. Flow chart of patients eligible for RCT Patients with postmenopausal bleeding N = 361 Office hysteroscopy N = 255 Exclusion N = 9 Informed about trial N = 105 Not informed about trial N = 141 Endometrial polyp N = 43 Endometrial polyp N = 51 Informed consent N = 4 Randomization to resection N = 4 Chapter Discussion In this paper, we reported on a randomized trial that was discontinued due to a lack of recruitment. In our study design, we performed a complete diagnostic work-up of patients presenting with PMB including a hysteroscopy. Since a large majority of patients did not give informed consent once the polyp was diagnosed, we could only include four patients in 26 months. Based on this low recruitment we decided to stop the trial after 26 months. With respect to efficacy of polypectomy in women with abnormal uterine bleeding, randomized clinical trials are advocated. 17 Our trial on this subject failed by lack of recruitment and to advise other investigators who might encounter the same problems, we report this trial and explore possible explanatory factors for this. The low recruitment in this trial was probably related both to the doctors at the outpatient clinic as well as the patients. Several factors can affect the granting of consent to participate

108 in a clinical trial. Patients can expect some therapeutic benefit from participation in a trial. 22 In our trial, participation led to the possibility of expectant management after randomization, with a possible disadvantage of recurrent bleeding and the burden of an extra hysteroscopy at the end of the follow-up period. Furthermore, placebo-controlled studies are known to induce patient s unwillingness to participate, especially if placebo means no treatment. 22 Therefore the design of the trial itself incurs less willingness to participate. Extra attention to communication and reduction of uncertainty in these trials is known to lead to improvement of participation. 22 The relationship between the person asking for informed consent and the patient is also a known factor to influence trial participation. Some patients are more willing to participate if their own doctor advises them to participate or if they know to please their doctor. 22 If the informed-consent seeker has moral or ethical objections against the trial or has insufficient knowledge about the trial, this could obstruct recruitment. 22 In our trial there was not one single informed-consent seeker. Although all doctors at the outpatient clinic were informed about the details of the trial so that they could inform patients about the trial, trial specific training was not given. Lack of this training and lack of time at the outpatient department might have led to insufficient communication about the trial and therefore lack of recruitment. Furthermore, the ethics of clinical research requires equipoise. Equipoise is defined as a state of genuine uncertainty on the part of the clinical investigator regarding the comparative therapeutic merits of each arm in a trial, which entails that the investigator has no treatment preference. 23 It is possible that individual doctors did have a treatment preference and that their believes have led to less recruitment of patients. An alternative concept of equipoise would be clinical equipoise in which controversy exists in the clinical community over the preferred treatment. 23 It is known that patients do not fully understand or accept the theoretical concept of equipoise, even when given explicit information about the fact that one treatment is not better than the other. 24 Even if they have some understanding of the concept of equipoise the majority of the patients would ask their doctor about his opinion about the best treatment option. 24 Trial design for efficacy of polyp removal in PMB 107 Finally, a patients preference study was performed alongside this trial. 25 With respect to the diagnostic work-up of endometrial polyps women were more likely to accept expectant management, but only if the risk of recurrent bleeding did not exceed 25%. This suggests that patients are willing to accept expectant management, however the risk of recurrent bleeding was not known during the trial, since this was subject of the trial. Recent literature has shown that the risk of recurrent bleeding probably lies around this 25%. 26 This implicates that patients might be less willing to participate in this trial with expectant management. Furthermore, the preference study demonstrated that most women deemed office hysteroscopy a rather easy and minimal invasive procedure and are willing to undergo invasive procedures.

109 At present no universal algorithm exists in the diagnostic management of women with PMB. Guidelines in different countries advocate evaluation of the uterine cavity and removal of benign endometrial polyps. 1,4,5,10,11 However, high-quality evidence on this subject is lacking. 17 Endometrial polyps are highly prevalent in women with PMB and as such may be responsible for significant morbidity and high resource use. 15,16 Alternatively, we may be subjecting women to unnecessary interventions, risks and wasting valuable health care resources. No studies exist that included a control group when reporting on the efficacy of polypectomy. A systematic review advocated RCT s on this subject. 17 It is therefore important to report this trial, since it s failure due to lack of recruitment adds valuable information to the discussion on efficacy of polypectomy. Although an RCT should be undertaken to fully answer the question if polypectomy is beneficial to patients, the present design does not seem to be feasible. It is important for clinicians to realize this. Chapter The design of RCTs that evaluate diagnostic tests, has been subject to debate. 27 Although trials are often undertaken for issues in therapy and prevention, there is no a-priori reason why they should not be used to resolve difficulties in diagnosis and monitoring. Yet, one should keep in mind how diagnostic tests affect patient outcome. Hysteroscopy itself does not influence patient outcome. A woman suffering from PMB, in whom hysteroscopy does not show any abnormalities, does not have a decreased probability of recurrent bleeding after hysteroscopy. In contrast, the woman in whom a polyp is diagnosed at hysteroscopy might benefit from the hysteroscopy, but only if removal of the polyp reduces the probability of recurrent bleeding. The latter question was addressed in our unsuccessful trial. We can distinguish two types of trials that evaluate diagnostic tests. We applied a design in which all patients were scheduled for hysteroscopy after which only those patients in whom the hysteroscopy showed a polyp were eligible for the study (Figure 2). In this so-called discordance design, patients with a polyp at hysteroscopy were randomized to either resection or expectant management. The effectiveness of hysteroscopy as a test can then be evaluated by integrating the effectiveness of removing the polyp and data on the prevalence of polyps. An alternative design is a trial in which patients are allocated to undergo the test or not to undergo the test. In our case this would imply that patients are randomized either to hysteroscopy or to expectant management without hysteroscopy (Figure 3).

110 Figure 3. Alternative study design Postmenopausal vaginal bleeding Vaginal examination incl. PAP smear, TVU ET 4 mm ET > 4 mm No polyp Hysteroscopy Pipelle Benign Polyp Resection Expectant management Trial design for efficacy of polyp removal in PMB 109 Follow-up Follow-up Follow-up Both designs have their advantages and disadvantages. In the discordance design, the impact of random error on the outcomes in the study stays limited, as patients without abnormalities can cause random error due to coincidental PMB. Second, the number of patients that is asked informed consent for randomization, which is in itself a time consuming and therefore costly procedure, will be limited. This advantage could not be applied in our study as we tried to inform patients prior to the hysteroscopy. Finally, when only the patients with a polyp are randomized, the follow up can be limited to those patients, whereas patients without an abnormality do not need follow-up, also increasing the efficiency of the trial.

111 An alternative design (Figure 2) that addresses the effectiveness of hysteroscopy and subsequent polypectomy in patients with PMB would be to randomize patients with benign histology results at office endometrial sampling to office hysteroscopy or to expectant management. Patients in whom office endometrial sampling is not successful, or shows insufficient material for diagnosis or patients with malignancy are excluded. In patients scheduled for office hysteroscopy in whom a polyp is visualized, the polyp can be removed during this procedure. The primary outcome measure would be similar to our trial, i.e. the recurrence of PMB. An obvious disadvantage of this study design is that patients can not be blinded for the treatment arm. The analysis would be performed in a similar way, i.e. with the construction of Kaplan-Meier curves demonstrating time to recurrent bleeding. We expect the probability of recurrence of PMB without hysteroscopy to be 36%. A strategy with hysteroscopy is thought to reduce this percentage to 20%. 26 We need 270 patients randomized to two groups of 135 patients to show such a difference. Chapter 9 In summary, the limited evidence that is available suggests that hysteroscopic polypectomy is a technical successful procedure that might improve abnormal uterine bleeding symptoms. However, there is lack of high clinical evidence to reliably inform clinical practice regarding the efficacy of endometrial polyp removal in patients with abnormal bleeding in pre- and postmenopausal patients. 17 We conducted a RCT in which patients with PMB in whom hysteroscopy showed an endometrial polyp, but our effort failed as a result of insufficient recruitment. A RCT on the subject is still warranted and an alternative trial design might be more feasible to fill this gap in evidence. 110

112 References 1. Epstein E. Management of postmenopausal bleeding in Sweden: a need for increased use of hydrosonography and hysteroscopy. Acta Obstet Gynecol Scand 2004;83: Goldstein RB, Bree RL, Benson CB, et al. Evaluation of the woman with postmenopausal bleeding: Society of Radiologists in Ultrasound-Sponsored Consensus Conference statement. J Ultrasound Med 2001;20: Gupta JK, Chien PF, Voit D, Clark TJ, Khan KS. Ultrasonographic endometrial thickness for diagnosing endometrial pathology in women with postmenopausal bleeding: a meta-analysis. Acta Obstet Gynecol Scand 2002;81: NVOG (Dutch Society of Obstetrics and Gynaecology). NVOG-Richtlijn Abnormaal vaginaal bloedverlies in de menopauze [in Dutch]. (NVOG Guideline: Abnormal vaginal bleeding during menopause) Scottish Intercollegiate Guidelines Network. Investigation of postmenopausal bleeding. 1st edn edinburgh, UK: Scottisch Intercollegiate Guidelines Network, Royal College of Physicians 2002; (www. sign.ac.uk). 6. Smith-Bindman R, Kerlikowske K, Feldstein VA, et al. Endovaginal ultrasound to exclude endometrial cancer and other endometrial abnormalities. JAMA 1998;280: Tabor A, Watt HC, Wald NJ. Endometrial thickness as a test for endometrial cancer in women with postmenopausal vaginal bleeding. Obstet Gynecol 2002;99: Clark TJ, Barton PM, Coomarasamy A, Gupta JK, Khan KS. Investigating postmenopausal bleeding for endometrial cancer: cost-effectiveness of initial diagnostic strategies. BJOG 2006;113: Dijkhuizen FP, Mol BW, Brolmann HA, Heintz AP. Cost-effectiveness of the use of transvaginal sonography in the evaluation of postmenopausal bleeding. Maturitas 2003;45: Clark TJ. Outpatient hysteroscopy and ultrasonography in the management of endometrial disease. Curr Opin Obstet Gynecol 2004;16: Epstein E, Ramirez A, Skoog L, Valentin L. Transvaginal sonography, saline contrast sonohysterography and hysteroscopy for the investigation of women with postmenopausal bleeding and endometrium > 5 mm. Ultrasound Obstet Gynecol 2001;18: Werkgroep Dutch Study in Postmenopausal B. [Gynaecological diagnosis of postmenopausal women with abnormal vaginal bleeding: a comparison with the guideline] Diagnostiek door gynaecologen bij vrouwen met abnormaal vaginaal bloedverlies in de postmenopauze; vergelijking met de richtlijn [in Dutch]. Ned Tijdschr Geneeskd 2005;149: Dijkhuizen FP, Brolmann HA, Potters AE, Bongers MY, Heinz AP. The accuracy of transvaginal ultrasonography in the diagnosis of endometrial abnormalities. Obstet Gynecol 1996;87: Emanuel MH, Verdel MJ, Wamsteker K, Lammes FB. An audit of true prevalence of intrauterine pathology: the hyesteroscopic findings, controlled for patient selection in 1,202 patients with abnormal uterine bleeding. Gynaecol Endosc 1995;4: Trial design for efficacy of polyp removal in PMB 111

113 Chapter Epstein E, Ramirez A, Skoog L, Valentin L. Dilatation and curettage fails to detect most focal lesions in the uterine cavity in women with postmenopausal bleeding. Acta Obstet Gynecol Scand 2001;80: Timmermans A, Veersema S. Office hysteroscopy in women with postmenopausal bleeding: see and treat of endometrial polyps using a Duckbill polyp snare. Gynecol Surg 2004;1: Nathani F, Clark TJ. Uterine polypectomy in the management of abnormal uterine bleeding: A systematic review. J Minim Invasive Gynecol 2006;13: Cravello L, D Ercole C, Azoulay P, Boubli L, Blanc B. [Hysteroscopic treatment of uterine fibromas] Le traitement hysteroscopique des fibromes uterins. J Gynecol Obstet Biol Reprod (Paris) 1995;24: Tjarks M, Van Voorhis BJ. Treatment of endometrial polyps. Obstet Gynecol 2000;96: Dijkhuizen FP, Mol BW, Brolmann HA, Heintz AP. The accuracy of endometrial sampling in the diagnosis of patients with endometrial carcinoma and hyperplasia: a meta-analysis. Cancer 2000;89: Timmermans A, Veersema S. Ambulatory transcervical resection of polyps with the Duckbill polyp snare: a modality for treatment of endometrial polyps. J Minim Invasive Gynecol 2005;12: Nievaard MA, de Vos R, de Haes JC, Levi M. [Reasons why patients do or do not participate in clinical trials; a systemic review of the literature] Redenen voor patienten om (niet) te participeren in klinische trials; een systematisch literatuuroverzicht. Ned Tijdschr Geneeskd 2004;148: Freedman B. Equipoise and the ethics of clinical research. N Engl J Med 1987;317: Robinson EJ, Kerr CE, Stevens AJ, et al. Lay public s understanding of equipoise and randomisation in randomised controlled trials. Health Technol Assess 2005;9:1-192, iii-iv. 25. Timmermans A, Opmeer BC, Veersema S, Mol BW. Patients preferences in the evaluation of postmenopausal bleeding. BJOG 2007;114: Timmermans A, van Doorn LC, Opmeer BC, et al. Follow-up of women after a first episode of postmenopausal bleeding and endometrial thickness greater than 4 millimeters. Obstet Gynecol 2008;111: Bossuyt PM, Lijmer JG, Mol BW. Randomised comparisons of medical tests: sometimes invalid, not always efficient. Lancet 2000;356:

114 10 Summary, general discussion and recommendations

115 Summary This thesis aims to evaluate the diagnostic work-up for postmenopausal bleeding (PMB) from several viewpoints. We focused on patients preferences, thereby determining what risk of false reassurance women would consider acceptable to what cost; we investigated whether the diagnosis of benign disease should be incorporated in the initial diagnostic work-up; and we assessed whether the risk of recurrent bleeding is related to the results of initial work-up. This chapter summarize and discusses the findings from our research, furthermore clinical and future research implications are put forward. The introduction (chapter 1) addresses the clinical problem of postmenopausal bleeding. It gives an outline of the diagnostic work-up in women with PMB and presents the aim of this thesis. The problem of PMB has two components: (1) the risk of endometrial carcinoma; (2) diagnosis of benign intracavitary pathology once malignancy has been ruled out. The role of transvaginal ultrasound (TVU) in the diagnostic work-up of women with PMB to exclude endometrial carcinoma is presented in this chapter. The diagnostic accuracy of TVU was established in three previous meta-analysis and as such is now implemented in guidelines and clinical practice. However, these meta-analyses have come to different conclusions on the diagnostic accuracy of TVU. Furthermore the preference of the patients regarding the diagnostic work-up in case of PMB was never systematically taken into account. Chapter Although guidelines on diagnostic work-up of PMB focus on exclusion of endometrial carcinoma, further diagnostic assessment can be undertaken to diagnose benign intracavitary pathology, mainly endometrial polyps. TVU, Saline Infusion Sonography and hysteroscopy are available to evaluate the uterine cavity for benign endometrial polyps. During SIS or hysteroscopy in women with PMB endometrial polyps are frequently found, but whether or not these polyps are causative of PMB, is not clear. Women suffering from PMB might benefit from uterine cavity evaluation and diagnosis of endometrial polyps, if consecutive removal of endometrial polyps will lead to less recurrent bleeding. However, there is a lack of high-quality evidence regarding the efficacy of polyp removal in women with PMB. In chapter 2 we systematically assess patients preferences for the diagnostic management of PMB. For this purpose we designed a structured interview, which was taken from 39 women who had had an office hysteroscopy in the diagnostic work-up of PMB. Women were informed about the probability of endometrial carcinoma versus benign disease and about advantages and disadvantages of different diagnostic strategies, i.e. expectant management after TVU or complete diagnostic work-up including invasive procedures. Most women wanted to be 100% certain that carcinoma could be ruled out. Only 5% of the women were willing to accept a

116 more than 5% risk of false reassurance. If the risk of benign disease exceeded 25%, the majority of women would prefer immediate diagnosis and treatment of benign pathology. From this preference study we concluded that women with PMB are prepared to undergo hysteroscopy to rule out any risk on cancer and that women aim for a high accuracy of the diagnostic workup. This implicates that measurement of endometrial thickness with TVU as a first-line test in the assessment of PMB should be reconsidered. In chapter 3 we determined the diagnostic accuracy of endometrial thickness measurement in the detection of endometrial carcinoma among women with PMB. We performed a meta-analysis using individual patient data (IPD) from individual studies. We included data on 2,896 patients of which 259 patients had endometrial carcinoma. Different Receiver Operator Characteristics (ROC) analyses were performed. All these ROC analyses resulted in similar ROC curves, with Area Under the Curve s (AUC) varying between 0.82 to These ROC curves indicated a lower AUC than the previously reported meta-analysis using conventional techniques. The commonly used cut-off values of 4 mm and 5 mm were found to have a sensitivity of 95% (95% CI %) and 90% (95% CI %) respectively. Only a cut-off value of 3 mm, yielded a high sensitivity of 98% (95% CI %). We concluded that previous meta-analyses on endometrial thickness measurement have overestimated its diagnostic accuracy in the detection of endometrial carcinoma. However, TVU measurement of endometrial thickness in women with postmenopausal bleeding using a cut-off value of 3 mm is still clinically useful and using such a cut-off value can reliably exclude endometrial carcinoma in women with postmenopausal bleeding. In chapter 4 the use of office hysteroscopy in the diagnosis and treatment in women with PMB is described. Endometrial polyps have a prevalence 41% in women with PMB and endometrial thickness > 4 mm. Office hysteroscopy offers the possibility of diagnosis as well as treatment of these polyps. This treatment can be performed in the same session in which endometrial polyps are diagnosed. It was concluded that a decision analysis regarding the diagnosis and treatment in women with PMB is necessary. Summary, general discussion and recommendations 115 In chapter 5 we determine the accuracy of endometrial thickness measurement with TVU to diagnose endometrial polyps in women with PMB in whom a carcinoma has been ruled out. In women with PMB endometrial thickness was measured with TVU. If endometrial thickness was > 4 mm, office hysteroscopy was performed. At hysteroscopy the uterine cavity was assessed for the presence of endometrial polyps. ROC analysis was performed to assess the capacity of TVU endometrial thickness measurement to diagnose endometrial polyps. We included 178 patients with endometrial thickness > 4 mm of which 90 (50%) had an endometrial polyp. ROC analysis showed an AUC of Therefore, we concluded that in women with PMB

117 in whom carcinoma has been ruled out, measurement of endometrial thickness with TVU is not useful in the diagnosis of endometrial polyps. In chapter 6 we evaluate current practice of Dutch gynaecologists in polyp removal. All practicing gynaecologists in the Netherlands in 2003 were surveyed by a mailed self-administered questionnaire about polyp removal. Polyps were commonly removed in an inpatient setting (71%), under general or regional anaesthesia (77%), and under direct hysteroscopic visualization (69%). Gynaecologists working in a teaching hospital removed polyps more often in an outpatient setting compared to gynaecologists working in a non-teaching hospital (39% versus 19%, p<0.001). Thus, in the Netherlands outpatient polyp removal is not practiced on a large scale. However, teaching hospitals are more often performing polypectomy in an outpatient setting. Therefore, we expect that there must be tendency towards outpatient hysteroscopic removal of polyps for the future. Chapter 10 In chapter 7 we determined the incidence and significance of recurrent PMB among women diagnosed with an endometrial thickness 4 mm after a first episode of PMB. Women with a first episode of PMB and an endometrial thickness 4 mm were managed expectantly. Of 249 women with endometrial thickness 4 mm, 25 (10%) women had recurrent PMB with a median time until recurrent bleeding of 49 weeks. Two women with recurrent bleeding turned out to have an endometrial carcinoma (8%) and one woman had malignant melanoma. Time since menopause, age, body mass index, hypertension, diabetes and anticoagulants were not predictive for recurrent bleeding. We concluded that the recurrence after a first episode of PMB managed expectantly is low and cannot be predicted by patient characteristics. Women with recurrent bleeding should be re-evaluated as they bear a considerable risk of carcinoma. 116 In chapter 8 we estimated the incidence of recurrent PMB among women who were diagnosed with endometrial thickness > 4 mm after a first episode of PMB. Women with a first episode of PMB and endometrial thickness > 4 mm underwent endometrial sampling. Women with benign endometrial sampling were included for follow-up. After diagnostic work-up 222 women were included for follow-up. During follow-up 47 (21%) women had recurrent PMB, with a median time to recurrent bleeding of 49 weeks. There was no difference with respect to recurrence rate between patients with polyp removal, patients with normal hysteroscopy, and patients with office endometrial sampling alone at the initial work-up. Two patients were diagnosed with atypical endometrial hyperplasia upon recurrent bleeding. From this study, we concluded that the recurrence rate of PMB in women with endometrial thickness > 4 mm is 21%. This recurrence rate is not related to incorporation of hysteroscopy or polyp removal at the initial work-up.

118 In chapter 9 we describe the design of a randomized controlled trial to evaluate the efficacy of polyp removal in women with PMB. We designed a trial in which patients with PMB and endometrial thickness > 4 mm undergo hysteroscopy. If during hysteroscopy a polyp was diagnosed, patients were asked to participate in this trial and after informed consent allocated to immediate removal of the polyp or expectant management. This trial suffered from lack of recruitment related both to doctors seeking for informed consent as well as patients unwillingness to participate in this trial. However, a randomized controlled trial on this subject is still necessary to evaluate the efficacy of uterine cavity evaluation in the diagnostic work-up of women with PMB, focussing on benign pathology. Therefore, we propose an alternative design which might be more feasible. In this alternative design patients undergo TVU and office endometrial sampling and in case of benign histology, patients are then allocated to either uterine cavity evaluation (including SIS and hysteroscopy) with removal of endometrial polyps or expectant management. Such a trial seems more feasible, since it is more in line with current clinical practice. A randomized clinical trial on the subject is still warranted to fill this gap in evidence. Summary, general discussion and recommendations 117

119 General discussion At present the diagnostic work-up of women with postmenopausal bleeding (PMB) focuses on how (pre)malignancy or benign intrauterine pathology can be diagnosed. This diagnostic work-up is subject of a NVOG guideline 1 which includes a diagnostic algorithm and gives three key recommendations: 1. Expectant management is justified in women with an endometrial thickness 4 mm. 2. In case of an endometrial thickness > 4 mm office endometrial sampling can reliably diagnose endometrial carcinoma. After office endometrial sampling, Saline Infusion Sonography (SIS) can be performed to diagnose benign intrauterine pathology. 3. In case of persistent or recurrent PMB a diagnostic hysteroscopy with endometrial sampling should be performed. Chapter Expectant management in women with endometrial thickness 4 mm Based on the Dutch guideline, transvaginal ultrasound (TVU) is used as a first step in women with PMB to stratify them in having a high versus low probability of endometrial carcinoma. 1 A previous study showed that Dutch gynaecologists adhere in most cases (73%) to the recommendation of expectant management in case of endometrial thickness 4 mm. 2 The first recommendation of the guideline seems widely implemented. However, the guideline seems to be formulated without a systematic consideration of patients preferences. 2 It is of clinical importance that guidelines fulfil the expectations of the patient. In chapter 2, we systematically assessed patients preferences regarding the trade-off between diagnostic accuracy and invasiveness in the diagnostic work-up in case of PMB. Women with PMB appeared to be prepared to undergo a rather invasive and painful diagnostic procedure such as hysteroscopy to rule out any risk on cancer. This finding would imply that, from a patient s point of view, the measurement of endometrial thickness with TVU as a first line test in the assessment of PMB should be reconsidered. However, guidelines represent a broader perspective than the individual patient and from a societal perspective also other factors (including costs) are to be considered. The Scotland National Health Services emphasizes the woman s input in further investigation if she is thought to be at low risk after TVU: If the clinician, the patient or both are not satisfied with this level of reassurance, further investigation is justified. 3 This should include an endometrial biopsy to obtain a histological assessment. Before subjecting all women to office endometrial biopsy (Pipelle), one should keep in mind that the failure rate of office endometrial sampling has been reported to be as high as 25% and that in women in whom office endometrial sampling failed, hysteroscopy is advised. 4 A strategy, in which all women undergo office endometrial sampling unconditional upon TVU results, can therefore result in more hysteroscopies in women with PMB. In this way, we may be subjecting women to unnecessary

120 interventions, risks and wasting valuable health care resources. Cost-effectiveness analyses have shown that a strategy starting with TVU was most cost-effective. 5,6 Only in case of a high disease prevalence (> 10-15%) strategies with office endometrial sampling or hysteroscopy became more cost-effective. 6 Another aspect of expectant management in women with endometrial thickness 4 mm is the diagnostic accuracy of TVU. This diagnostic accuracy of TVU in women with PMB has previously been estimated based on individual studies with varying positivity cut-off values as well as meta-analyses of the reported results With respect to meta-analysis of randomized controlled trials, individual patient data (IPD) meta-analysis is considered to be superior over meta-analysis of the literature Meta-analyses using individual patient data instead of published summary data have come to less optimistic, but more accurate, conclusions. Limitations of conventional meta-analysis based on published data, might be overcome by meta-analysis of individual patient data. In chapter 3, we used a meta-analytic approach in which individual patient data from a series of original studies are combined. We showed that in previous studies and meta-analyses the diagnostic accuracy of TVU has been overestimated. We demonstrated a lower diagnostic accuracy for TVU than was reported previously, which resulted in a sensitivity of 95% for a specificity of 47% at a cut-off level of 4 mm. At a cut-off level of 3 mm for endometrial thickness, we found a sensitivity of 98% for specificity of 35%. Therefore, based on the present meta-analysis, the use TVU measurement of endometrial thickness remains justified but we recommend the use of a cut-off level of 3 mm. Such a cut-off value reduces a 10% pre-test probability to a 0.6% post-test probability. Although patients aim for a post-test probability of 0%, such a post-test probability seems virtually impossible. One should keep in mind that the risk of endometrial carcinoma in a population of asymptomatic postmenopausal bleeding is reported to be 0.2%. 17 Although, the use of a cut-off level of 3 mm results in a higher sensitivity which is more in-line with patients preferences, other diagnostic strategies have to be evaluated. Cost-effectiveness analyses have to be repeated to evaluate whether a strategy starting with TVU and using a cut-off value of 3 mm remains cost-effective as a strategy with TVU using a cut-off value of 4 mm. Alternatively a strategy starting with office endometrial sampling might become more cost-effective. Summary, general discussion and recommendations 119 Another aspect of expectant management in women with endometrial thickness 4 mm is the chance of recurrent bleeding. In chapter 7 we studied this chance, and showed that recurrent bleeding in patients with PMB occurs in only 10% with a median time to recurrent bleeding of one year. We could not identify patient factors associated with recurrent bleeding. This low recurrent bleeding rate is supportive for the use of TVU as a first test in the work-up for PMB.

121 Diagnosis of benign pathology The second recommendation of the NVOG guideline focuses on the diagnosis of benign intrauterine pathology. 1 Endometrial polyps are prevalent (40%) in women with PMB and endometrial thickness > 4 mm (chapter 4 & 9). They are easily diagnosed and removed at office hysteroscopy (chapter 4). Whether or not SIS or an office hysteroscopy (i.e. uterine cavity evaluation) should subsequently be performed in these women is left to the discretion of the individual gynaecologists. A guideline adherence studied showed that in 13 to 17% of the cases, gynaecologists proceeded with hysteroscopy where office endometrial sampling would have been sufficient to exclude malignancy. 2 Uterine cavity evaluation (SIS or hysteroscopy) is often performed in case of PMB and endometrial thickness > 8 mm or suspicion of intracavitary pathology on TVU. 18 However, our study (chapter 5) demonstrates that TVU endometrial thickness measurement has poor discriminative ability for detecting or excluding endometrial polyps and therefore is not clinically useful in the diagnosis of such benign focal pathology. TVU can not reliably detect those patients that have a high probability of focal endometrial pathology and as such can not be used in the work-up for benign pathology. Chapter It can be questioned if the diagnosis of endometrial polyps is of clinical importance. To answer this question, four aspects need to be taken into consideration. First, one can assume that focal carcinomas in an endometrial polyp are only detected if this polyp is removed. The prevalence of serious endometrial disease within polyps is low. 19,20 If endometrial polyps with a focal carcinoma inside the polyp are not removed, it is to be expected that these cancers will become manifest with clinical signs of recurrent PMB and the patient will then undergo another endometrial sampling. Whether this delay in detection of tumors significantly alters the prognosis remains unclear. 21,22 The second argument for diagnosis and removal of endometrial polyps is that women who underwent polypectomy are thought to experience less recurrent bleeding. However it has been shown that (smaller) endometrial polyps may regress spontaneously and as such may not be responsible for PMB. 19,23 Furthermore, endometrial polyps are also found in women without PMB with a prevalence around 10%. 24,25 Alternatively, endometrial polyps may be causative of PMB and removal will therefore lead to less recurrent bleeding. In chapter 8, we showed that the recurrence rate after a first episode of PMB and endometrial thickness > 4 mm is 21%. There was no difference in recurrent bleeding for patients with hysteroscopy and/or dilatation and curettage (D&C) at initial work-up and patients with office endometrial sampling, nor was there a difference between patients with an endometrial polyp diagnosed and treated at initial work-up and patients without a polyp. Unfortunately, D&C was used in this study and D&C has been shown to miss endometrial polyps in 50% to 85% of the cases and as such D&C should not be performed for diagnosis or removal of endometrial polyps. 26,27 Hysteroscopy is generally considered to be superior for endometrial polyp removal. Given the fact that hysteroscopic removal of endometrial polyps was performed in at least part of this group and therefore part of this group was treated adequately, one would still

122 expect the recurrence bleeding rate to be lower compared to the group without hysteroscopy or D&C. Since this lower recurrence rate was not found, the removal of polyps does not seem to lead to less recurrent PMB and from this point of view diagnosis of polyps whether with SIS or hysteroscopy is not clinically useful. A third argument can be found in patients preferences regarding diagnosis and treatment of benign pathology. In chapter 2 we found that patients would prefer immediate diagnosis and treatment of endometrial polyps in case the probability of recurrence of bleeding was estimated to be more than 25%. Although the recurrence bleeding rate of 21% lies close to this point, the exact recurrence bleeding rate caused by an endometrial polyp left in situ is still not known. Furthermore, given the present knowledge, this recurrence bleeding rate is not lowered by uterine cavity evaluation (including polyp removal) at first work-up. Uterine cavity evaluation and subsequent polyp removal can be performed at first work-up, however this does not seem to lead to less recurrent bleeding and that is what patients expect if they prefer immediate uterine cavity evaluation. Although SIS and office hysteroscopy are easily performed, the effectiveness of uterine cavity evaluation (SIS or hysteroscopy) at the initial work-up of PMB should be subject of future studies. Therefore, uterine cavity evaluation for the sole purpose of detecting endometrial polyps in case of PMB has only to be undertaken after counseling the patient about the above mentioned arguments or within study settings. At present hysteroscopic polyp removal in an outpatient setting is not common practice in the Netherlands (chapter 6). Furthermore, 30% of the gynaecologists will remove endometrial polyps by D&C (chapter 6). D&C has been shown to miss endometrial polyps in 50% to 85% and should therefore not be performed for diagnosis or removal of endometrial polyps. 26,27 Since hysteroscopy under general anaesthesia carries a higher risk of complications than office hysteroscopy, it is questionable if in this clinical setting hysteroscopy has to be incorporated in the work-up of women with PMB. 28 Therefore it is of importance that those gynaecologists that proceed with uterine cavity evaluation (SIS or hysteroscopy) as is stated in the second recommendation of the guideline, do so in an office setting (SIS or office hysteroscopy). Before office hysteroscopic polyp removal becomes widely available, evidence is needed to support the efficacy of polyp removal in case of PMB. Summary, general discussion and recommendations 121 A randomized controlled trial (RCT) is needed, to unequivocally answer the question of efficacy of polypectomy in case of PMB. A trial in which patients with PMB and an endometrial polyp at hysteroscopy were randomized to polypectomy or expectant management, turned out not to be feasible, at least in our setting (chapter 9). The low recruitment in this trial was probably related both to the doctors at the outpatient clinic as well as the patients. Expectant management in case of endometrial thickness 4 mm or in case of benign histology after endometrial thickness > 4 mm, might be generally accepted. 2 In contrast, expectant management in case

123 of a hysteroscopically diagnosed polyp is not so easy to accept, neither for patients nor for doctors. The question whether removal of benign endometrial polyps in patients with PMB is effective, remains therefore to be answered. An alternative design of a RCT, as proposed in this chapter, might be more feasible. Until such distinct evidence becomes available, gynaecologists should be hesitant to incorporate hysteroscopy or SIS in the diagnostic work-up of women with PMB. Chapter Diagnostic work-up of women with recurrent PMB The third recommendation of the NVOG guideline focuses on the work-up of women with recurrent PMB. The chance of recurrent PMB is 10% in women with endometrial thickness 4 mm at initial work-up and 21% in women with endometrial thickness > 4 mm. At recurrent bleeding endometrial carcinoma was diagnosed in 8% of the women with initial endometrial thickness 4 mm and in 4.3% of the women with initial endometrial thickness > 4 mm. Therefore, all patients with recurrent PMB should undergo histology sampling to exclude malignancy, irrespective of the initial assessment with TVU. Whether this should be done by hysteroscopy and targeted endometrial biopsy or by office endometrial sampling, is not yet clearly established. It can be argumented that in women in whom only TVU was performed at initial work-up ( 4 mm), office endometrial sampling should be performed at recurrent bleeding; whereas in those women in whom already office endometrial sampling was performed, immediate office hysteroscopy with targeted biopsy could be performed. Women with recurrent bleeding are at risk of endometrial carcinoma (4% to 8%) and these women are possibly falsely reassured at initial work-up (false-negatives of initial diagnostic test). Therefore the work-up at recurrent bleeding should consist of a different diagnostic test with a higher diagnostic accuracy. However, precisely how the diagnostic algorithm of women with recurrent bleeding should be, remains to be determined in future research.

124 Future research Expectant management in women with endometrial thickness 4 mm Future research should focus on the consequences of a cut-off level of 3 mm. A sensitivity of 98% for a specificity of 35% implies that we find 2% more endometrial carcinomas for 15% more invasive diagnostic procedures. What effect this has on cost-effectiveness has to be determined in future research. In this light other diagnostic strategies have to explored, such as a diagnostic strategy starting with office endometrial sampling. Patients and doctors aim for a higher diagnostic accuracy than is presently offered with TVU measurement of endometrial thickness. Such a higher accuracy might be achieved by incorporation of patient s characteristics (e.g. age, presence of diabetes, Body Mass Index (BMI), presence of hypertension) in the diagnostic work-up. The incorporation of TVU with patient s characteristics in a diagnostic strategy has been studied and resulted in higher diagnostic accuracy. 4,29-31 Statistical methods can be used to develop and further improve such models and incorporating patient s characteristics with diagnostic tests. 31,32 Furthermore, by combining and analysing individual patient data from different studies using (IPD meta analyses), larger databases can be obtained, in which previously described models can be externally validated. Such models could be incorporated in clinical prediction rules, where the individual probability for endometrial cancer is obtained for each individual woman, and a diagnostic algorithm is developed to maximize the diagnostic accuracy at an acceptable patient burden and health care costs. Such prediction rules are currently also available in reproductive medicine or similar to a risk of malignancy index. After developing such clinical prediction rules, diagnostic accuracy and clinical applicability should be tested in clinical practice in a prospective multicenter study. If indeed, such a model would lead to higher diagnostic accuracy than TVU alone, office endometrial sampling could then for example be offered only to those women with a high probability of endometrial cancer. Summary, general discussion and recommendations 123 Diagnosis of benign pathology With respect to benign pathology, future research should focus on the diagnosis and treatment of endometrial polyps in women with PMB. Diagnosis is only useful if treatment is beneficiary. This question still remains to be answered. An RCT in which women were randomized after the diagnosis of endometrial polyps was not feasible. An alternative design in which women are randomized after exclusion of malignancy to either SIS and hysteroscopy or expectant management, seems more feasible, since this is more in line with current practice. Preparations for such a trial are undertaken (POMPOEN-trial). In this trial women with PMB and endometrial thickness > 4 mm undergo office endometrial sampling. In case office endometrial sampling shows benign histology, women are randomized to expectant management or further work-up. If randomized to further work-up, all women undergo SIS and office hysteroscopy; in case endometrial polyps are diagnosed these are then removed. If in the end,

125 removal of endometrial polyps in women with PMB is shown to be effective (less recurrent PMB), then a diagnostic strategy focussing on exclusion of malignancy and diagnosis and treatment of benign intrauterine pathology has to be developed. The optimal strategy should be based on the diagnostic accuracy, costs and patient burden of different test combinations. Furthermore, the preference of the patient should be taken into account, as the trade-off between accuracy, costs and test burden depends on how patients value each aspect. Although SIS and hysteroscopy have comparable diagnostic accuracy, SIS is reported to be less painful than hysteroscopy. 33,34 However, patient s might be willing to accept more pain and discomfort in exchange for diagnosis and treatment options in one session, such as is offered with hysteroscopy. Further studies should focus on different diagnostic strategies and patients preferences regarding such strategies. Ideally, the diagnostic strategy for exclusion of malignancy should be combined with the diagnostic strategy for benign pathology, resulting in a diagnostic algorithm in which exclusion of malignancy and treatment of benign pathology (if this has been proven to be beneficiary) are combined. Clinical implications Chapter This thesis supports the use of TVU as a first step in women with PMB. However, it advocates the use of a cut-off value of 3 mm instead of the currently used 4 or 5 mm. With a sensitivity of 98% for a specificity of 35%, a cut-off value of 3 mm will reduce a 10% post-test probability of endometrial carcinoma, to a post-test probability of 0.6%. With respect to diagnosis of benign intracavitary pathology (endometrial polyps), this thesis emphasizes the lack of high-quality evidence regarding the diagnosis and removal of endometrial polyps with respect to recurrent PMB. Considering the results of our cohort study in women with PMB, endometrial thickness > 4 mm and benign histology, uterine cavity evaluation (i.e. SIS or hysteroscopy) does not seem to be beneficiary to patients with PMB with respect to recurrent bleeding. Clinicians should therefore be hesitant to routinely incorporate uterine cavity evaluation in the diagnostic work-up of women with PMB with the sole purpose of diagnosing and treating benign pathology. Uterine cavity evaluation (SIS and hysteroscopy) should therefore be performed in trial settings.

126 References 1. NVOG (Dutch Society of Obstetrics and Gynaecology). NVOG-Richtlijn Abnormaal vaginaal bloedverlies in de menopauze [in Dutch]. (NVOG Guideline: Abnormal vaginal bleeding during menopause) Werkgroep Dutch Study in Postmenopausal Bleeding. [Gynaecological diagnosis of postmenopausal women with abnormal vaginal bleeding: a comparison with the guideline] Diagnostiek door gynaecologen bij vrouwen met abnormaal vaginaal bloedverlies in de postmenopauze; vergelijking met de richtlijn [in Dutch]. Ned Tijdschr Geneeskd 2005;149: Scottish Intercollegiate Guidelines Network. Investigation of postmenopausal bleeding. 1st edn edinburgh, UK: Scottisch Intercollegiate Guidelines Network, Royal College of Physicians 2002;( 4. van Doorn HC, Opmeer BC, Burger CW, et al. Inadequate office endometrial sample requires further evaluation in women with postmenopausal bleeding and abnormal ultrasound results. Int J Gynaecol Obstet 2007;99: Dijkhuizen FP, Mol BW, Brolmann HA, Heintz AP. Cost-effectiveness of the use of transvaginal sonography in the evaluation of postmenopausal bleeding. Maturitas 2003;45: Clark TJ, Barton PM, Coomarasamy A, Gupta JK, Khan KS. Investigating postmenopausal bleeding for endometrial cancer: cost-effectiveness of initial diagnostic strategies. BJOG 2006;113: Nasri MN, Coast GJ. Correlation of ultrasound findings and endometrial histopathology in postmenopausal women. Br J Obstet Gynaecol 1989;96: Tabor A, Watt HC, Wald NJ. Endometrial thickness as a test for endometrial cancer in women with postmenopausal vaginal bleeding. Obstet Gynecol 2002;99: Gupta JK, Chien PF, Voit D, Clark TJ, Khan KS. Ultrasonographic endometrial thickness for diagnosing endometrial pathology in women with postmenopausal bleeding: a meta-analysis. Acta Obstet Gynecol Scand 2002;81: Smith-Bindman R, Kerlikowske K, Feldstein VA, et al. Endovaginal ultrasound to exclude endometrial cancer and other endometrial abnormalities. JAMA 1998;280: Karlsson B, Granberg S, Wikland M, et al. Transvaginal ultrasonography of the endometrium in women with postmenopausal bleeding a Nordic multicenter study. Am J Obstet Gynecol 1995;172: Granberg S, Ylostalo P, Wikland M, Karlsson B. Endometrial sonographic and histologic findings in women with and without hormonal replacement therapy suffering from postmenopausal bleeding. Maturitas 1997;27(1): Granberg S, Ylostalo P, Wikland M, Karlsson B. Endometrial sonographic and histologic findings in women with and without hormonal replacement therapy suffering from postmenopausal bleeding. Maturitas 1997;27: Stewart LA, Parmar MK. Meta-analysis of the literature or of individual patient data: is there a difference? Lancet 1993;341: Summary, general discussion and recommendations 125

127 Chapter Simmonds MC, Higgins JP, Stewart LA, Tierney JF, Clarke MJ, Thompson SG. Meta-analysis of individual patient data from randomized trials: a review of methods used in practice. Clin Trials 2005;2: Stewart LA, Tierney JF. To IPD or not to IPD? Advantages and disadvantages of systematic reviews using individual patient data. Eval Health Prof 2002;25(1): Gull B, Karlsson B, Milsom I, Wikland M, Granberg S. Transvaginal sonography of the endometrium in a representative sample of postmenopausal women. Ultrasound Obstet Gynecol 1996;7: Epstein E. Management of postmenopausal bleeding in Sweden: a need for increased use of hydrosonography and hysteroscopy. Acta Obstet Gynecol Scand 2004;83: Savelli L, De Iaco P, Santini D, et al. Histopathologic features and risk factors for benignity, hyperplasia, and cancer in endometrial polyps. Am J Obstet Gynecol 2003;188: Anastasiadis PG, Koutlaki NG, Skaphida PG, Galazios GC, Tsikouras PN, Liberis VA. Endometrial polyps: prevalence, detection, and malignant potential in women with abnormal uterine bleeding. Eur J Gynaecol Oncol 2000;21: Martin-Ondarza C, Gil-Moreno A, Torres-Cuesta L, et al. Endometrial cancer in polyps: a clinical study of 27 cases. Eur J Gynaecol Oncol 2005;26: Gerber B, Krause A, Muller H, et al. Ultrasonographic detection of asymptomatic endometrial cancer in postmenopausal patients offers no prognostic advantage over symptomatic disease discovered by uterine bleeding. Eur J Cancer 2001;37: DeWaay DJ, Syrop CH, Nygaard IE, Davis WA, Van Voorhis BJ. Natural history of uterine polyps and leiomyomata. Obstet Gynecol 2002;100: Lieng M, Qvigstad E, Sandvik L, Jorgensen H, Langebrekke A, Istre O. Hysteroscopic resection of symptomatic and asymptomatic endometrial polyps. J Minim Invasive Gynecol 2007;14: Antunes A, Jr., Costa-Paiva L, Arthuso M, Costa JV, Pinto-Neto AM. Endometrial polyps in pre- and postmenopausal women: factors associated with malignancy. Maturitas 2007;57: Epstein E, Ramirez A, Skoog L, Valentin L. Dilatation and curettage fails to detect most focal lesions in the uterine cavity in women with postmenopausal bleeding. Acta Obstet Gynecol Scand 2001;80: Emanuel MH, Wamsteker K, Lammes FB. Is dilatation and curettage obsolete for diagnosing intrauterine disorders in premenopausal patients with persistent abnormal uterine bleeding? Acta Obstet Gynecol Scand 1997;76: Jansen FW, Vredevoogd CB, van Ulzen K, Hermans J, Trimbos JB, Trimbos-Kemper TC. Complications of hysteroscopy: a prospective, multicenter study. Obstet Gynecol 2000;96: Opmeer BC, van Doorn HC, Heintz AP, Burger CW, Bossuyt PM, Mol BW. Improving the existing diagnostic strategy by accounting for characteristics of the women in the diagnostic work up for postmenopausal bleeding. BJOG 2007;114: Bruchim I, Biron-Shental T, Altaras MM, et al. Combination of endometrial thickness and time since menopause in predicting endometrial cancer in women with postmenopausal bleeding. J Clin Ultrasound 2004;32:

128 31. Bachmann LM, ter Riet G, Clark TJ, Gupta JK, Khan KS. Probability analysis for diagnosis of endometrial hyperplasia and cancer in postmenopausal bleeding: an approach for a rational diagnostic workup. Acta Obstet Gynecol Scand 2003;82: Khan KS, Bachmann LM, ter Riet G. Systematic reviews with individual patient data meta-analysis to evaluate diagnostic tests. Eur J Obstet Gynecol Reprod Biol 2003;108: van Dongen H, de Kroon CD, van den Tillaart SA, Louwe LA, Trimbos-Kemper GC, Jansen FW. A randomised comparison of vaginoscopic office hysteroscopy and saline infusion sonography: a patient compliance study. BJOG 2008;115: Van den Bosch T, Verguts J, Daemen A, et al. Pain experienced during transvaginal ultrasound, saline contrast sonohysterography, hysteroscopy and office sampling: a comparative study. Ultrasound Obstet Gynecol 2008;31: Summary, general discussion and recommendations 127

129

130 Nederlandse samenvatting

131 Dit proefschrift handelt over de diagnostiek bij vrouwen met postmenopauzaal bloedverlies (PMB). Postmenopauzaal bloedverlies is gedefinieerd als bloedverlies na de overgang. Vrouwen die zich presenteren met PMB hebben de kans dat een endometriumcarcinoom (kwaadaardige afwijking van het baarmoederslijmvlies) de onderliggende oorzaak is van dit bloedverlies. Deze kans bedraagt ongeveer 10%. Vanwege dit risico moet er bij vrouwen met PMB diagnostiek worden verricht om een endometriumcarcinoom uit te sluiten. Nederlandse samenvatting 130 Traditioneel werd bij vrouwen met PMB onder narcose een curettage van de baarmoederholte verricht om weefsel te verkrijgen voor het opsporen of uitsluiten van een endometriumcarcinoom. Sinds het begin van de jaren 90 is het echter mogelijk om met transvaginale echoscopie (TVE) bij een deel van de patiënten endometriumcarcinoom uit te sluiten. De Nederlandse Vereniging voor Obstetrie en Gynaecologie (NVOG) geeft in een richtlijn (Diagnostiek naar abnormaal bloedverlies in de postmenopauze 2003) aan hoe de gynaecoloog bij voorkeur dient te handelen in geval van PMB. Deze richtlijn adviseert het verrichten van een uitstrijkje van de baarmoedermond en het verrichten van TVE waarbij de dubbele dikte van het endometrium (baarmoederslijmvlies) wordt gemeten. Indien een dubbele endometriumdikte van 4 mm wordt gemeten, kan bij normale uitslag van het uitstrijkje een afwachtend beleid worden voorgesteld. Indien een endometriumdikte van meer dan 4 mm wordt gevonden, dient aanvullend weefsel-onderzoek van het endometrium plaats te vinden. Dit weefsel kan eenvoudig poliklinisch verkregen worden door met een slangetje endometrium uit het cavum uteri (baarmoederholte) op te zuigen (ook wel Pipelle genoemd). Door histologisch onderzoek van dit weefsel wordt dan een endometriumcarcinoom uitgesloten. Nadat dit weefselonderzoek heeft plaatsgevonden, kan nog diagnostiek worden ingezet naar de aanwezigheid van benigne (goedaardige) afwijkingen in de baarmoederholte, veelal gaat het dan om endometriumpoliepen (poliepen van het baarmoederslijmvlies). Bij vrouwen met een endometriumdikte > 4 mm worden poliepen in ongeveer 40% van de vrouwen gevonden. Diagnostiek naar benigne endometriumpoliepen kan worden verricht door middel van een watercontrastecho (SIS) of door middel van een hysteroscopie (kijkbuisonderzoek). Beide onderzoeken kunnen poliklinisch worden verricht, maar in tegenstelling tot een SIS kunnen tijdens een hysteroscopie eventuele afwijkingen (in dezelfde sessie) worden weggehaald. De hysteroscopie werd echter van oudsher op de operatiekamers verricht, waarbij algehele of regionale anesthesie (narcose of ruggenprik) werd toegepast. De poliklinische hysteroscopie, zowel voor diagnostiek als voor therapie, is nog geen gemeengoed. De NVOG richtlijn adviseert om diagnostiek naar goedaardige behandelbare afwijkingen alleen in te zetten als SIS of poliklinische hysteroscopie beschikbaar zijn. De richtlijn adviseert dat diagnostiek naar benigne afwijkingen kan worden verricht tijdens een eerste episode van PMB of bij herhaald PMB. Bij herhaald PMB moet in ieder geval een hysteroscopie worden verricht.

132 In dit proefschrift is gekeken naar de voorkeuren van de patiënt met PMB voor het diagnostisch traject. Daarbij werd onderzocht welke afwegingen worden gemaakt in het diagnostisch traject en welke risico s, zoals bijvoorbeeld de kans op complicaties of de kans op gemist carcinoom, door vrouwen met PMB acceptabel werden geacht. Voorts werd geëvalueerd of goedaardige afwijkingen hier een rol speelden, en of diagnostiek hiernaar moet worden verricht in het diagnostisch traject. Hierbij keken we naar het risico op herhaald bloedverlies en onderzochten we of de kans op herhaald bloedverlies afhankelijk is van het initieel gevolgde diagnostische traject. Uiteindelijk werden de bevindingen tegen de achtergrond geplaatst van de huidige NVOG richtlijn. Hoofstuk 1: Inleiding In dit hoofdstuk wordt het klinisch probleem van postmenopauzaal bloedverlies uiteengezet (PMB). Er worden twee aspecten nader belicht, te weten (1) het risico op endometriumcarcinoom en (2) diagnostiek naar benigne intracavitaire pathologie nadat een maligniteit is uitgesloten. Transvaginale echoscopie (TVE) wordt in de NVOG richtlijn genoemd als de eerste stap in het diagnostisch traject bij vrouwen met PMB. De diagnostische accuratesse van TVE en meting van het endometrium werd eerder vastgesteld in drie meta-analyses, maar de auteurs van deze meta-analyses kwamen niet tot dezelfde conclusies. Daarnaast werd nooit systematisch gekeken naar de voorkeuren van de patiënt in het diagnostisch traject. De mening van de patiënt kan een rol spelen om de mate van invasiviteit van het onderzoek (door dokters voorgesteld) te bepalen. De richtlijn richt zich in de eerste plaats op het uitsluiten van endometriumcarcinoom, hetgeen kan gebeuren met TVE en indien nodig een Pipelle. Echter, nadat een endometrium carcinoom is uitgesloten kan nog verdere diagnostiek naar benigne behandelbare intracavitaire afwijkingen worden verricht. Het gaat dan voornamelijk om benigne endometriumpoliepen. Hoewel endometriumpoliepen frequent worden gevonden bij vrouwen met PMB, is het niet duidelijk of er een causaal verband bestaat tussen endometriumpoliepen en PMB. Het is tot op heden onduidelijk of verwijdering van deze poliepen leidt tot minder herhaald PMB. Nederlandse samenvatting 131 Hoofdstuk 2: Onderzoek naar patiëntenvoorkeuren bij diagnostiek van postmenopauzaal bloedverlies In dit hoofdstuk hebben we de voorkeur van vrouwen met PMB ten aanzien van het te volgen beleid systematisch in kaart gebracht. Hiertoe werd een gestructureerd interview ontworpen, waarin de verschillende diagnostische mogelijkheden werden uitgelegd. Dit interview werd afgenomen bij vrouwen met PMB, die het hele diagnostische traject hadden doorlopen inclusief een poliklinische hysteroscopie. Deze vrouwen werden geïnformeerd over de kans

133 op endometriumcarcinoom en de kans op benigne pathologie. Verder kregen ze informatie over de verschillende diagnostische trajecten, bijvoorbeeld alleen echo of een compleet diagnostisch traject inclusief meer invasieve diagnostiek. De meerderheid van de vrouwen wilden een endometriumcarcinoom met 100% zekerheid uitgesloten hebben. Slechts 5% van de vrouwen was bereid een risico van 5% of meer te accepteren op het missen van een endometriumcarcinoom. Wat betreft de diagnostiek naar benigne afwijkingen werd gekeken hoe groot de herhalingskans op recidief bloedverlies ten gevolge van die afwijking moest zijn, waarbij zij verdergaande diagnostiek wilden ondergaan. De meerderheid van de vrouwen gaf aan, dat indien het risico op recidief bloedverlies meer dan 25% zou bedragen, ze direct diagnostiek en therapie naar benigne pathologie wilden, d.w.z. bij de eerste episode PMB en niet wachten tot een recidief episode. Geconcludeerd werd dat vrouwen met PMB bereid zijn om invasieve diagnostiek te ondergaan om ieder risico op kanker uit te sluiten. Vrouwen verwachten dus diagnostiek met een zeer hoge diagnostische accuratesse en weinig kans op fout-negatieve test uitslag. Dit impliceert dat de TVE en de meting van endometriumdikte als start van het diagnostisch proces bij vrouwen met PMB onder druk komt te staan. Nederlandse samenvatting 132 Hoofdstuk 3: Meta-analyse op individuele patiëntendata van endometriumdiktemeting bij vrouwen met postmenopauzaal bloedverlies voor het uitsluiten van endometriumcarcinoom In dit hoofstuk wordt de diagnostische accuratesse van echografische endometriumdiktemeting voor het uitsluiten van endometriumcarcinoom bij vrouwen met PMB onderzocht. Conventionele meta-analyses hebben mogelijk deze diagnostische accuratesse overschat. Met behulp van de individuele patiëntendata van verschillende individuele studies werd een nieuwe meta-analyse verricht. Zo n meta-analyse van individuele patiëntendata (IPD) wordt tegen woordig superieur beschouwd t.o.v. conventionele meta-analyses die gebruik maken van gepubliceerde data. Bij deze IPD meta-analyse konden data worden geïncludeerd van patiënten, waarvan 259 patiënten met een endometriumcarcinoom. Verschillende Receiver Operator Characteristics (ROC) analyses werden verricht om te kijken naar de diagnostische accuratesse van endometriumdiktemeting. De verschillende ROC analyses resulteerden allemaal in een zelfde ROC curve. Deze curve werd vergeleken met de eerder gepubliceerde ROC curve, die was verkregen in een conventionele meta-analyse. Hierbij bleek dat in de eerdere meta-analyse een overschatting van de diagnostische accuratesse voor endometriumdiktemeting was gevonden. De afkapwaardes voor endometriumdikte van 4 en 5 mm hadden respectievelijk een sensitiviteit van 95% en 90%. Alleen een afkapwaarde van 3 mm had een acceptabele sensitiviteit van 98%. Geconcludeerd werd dat de diagnostische accuratesse van TVE en endometriumdiktemeting in eerdere meta-analyse te hoog was ingeschat. Echter, een 3 mm afkapwaarde voor endometriumdikte kan endometriumcarcinoom betrouwbaar uitsluiten en derhalve is de transvaginale echo met deze aangepaste afkapwaarde nog steeds bruikbaar in de dagelijkse praktijk.

134 Hoofdstuk 4: Poliklinische hysteroscopie biedt de mogelijheid van diagnose en therapie in dezelfde sessie bij vrouwen met postmenopauzaal bloedverlies In dit hoofdstuk wordt beschreven hoe poliklinische hysteroscopie bij vrouwen met PMB de mogelijkheid biedt van diagnostiek en therapie in dezelfde sessie. Endometriumpoliepen worden bij 40% van de vrouwen met PMB en een endometriumdikte > 4 mm gevonden. Door middel van poliklinische hysteroscopie kunnen deze poliepen worden gediagnosticeerd en in dezelfde sessie eveneens worden verwijderd. Echter, op grond van de huidige aanwezige literatuur is het nog niet duidelijk wat de klinische waarde is van het verwijderen van deze poliepen. Aan de hand van onze studie wordt dan ook geconcludeerd dat een beslisboom nodig is naar diagnostiek en therapie bij vrouwen met PMB waarbij ook gekeken wordt naar benigne afwijkingen. Hoofdstuk 5: Diagnostische accuratesse van endometriumdikte meting ter uitsluiting van endometriumpoliepen bij vrouwen met postmenopauzaal bloedverlies In dit hoofdstuk wordt de waarde van echografische endometriumdiktemeting geëvalueerd bij het opsporen of uitsluiten van endometriumpoliepen bij vrouwen met PMB, waarbij een maligniteit is uitgesloten. Bij vrouwen met PMB werd een transvaginale echo (TVE) verricht. Bij die vrouwen die een endometriumdikte van > 4 mm hadden, werd vervolgens een poliklinische hysteroscopie verricht. Tijdens de hysteroscopie werd het cavum uteri (de baarmoederholte) bekeken op de aanwezigheid van endometriumpoliepen. Er werd een ROC analyse verricht waarbij gekeken werd naar de capaciteit van endometriumdiktemeting voor de diagnose van endometriumpoliepen. Er werden 178 patienten met een endometriumdikte > 4 mm geïncludeerd. Van deze 178 patienten hadden 90 patienten (50%) een endometriumpoliep tijdens de hysteroscopie. ROC analyse liet een diagnostische accuratesse van 0,64 zien, uitgedrukt als oppervlakte onder de ROC curve (AUC). Geconcludeerd werd dat bij vrouwen met PMB waarbij een maligniteit is uitgesloten, endometriumdiktemeting met TVE niet bruikbaar is om die vrouwen met een endometriumpoliep te kunnen onderscheiden van de vrouwen zonder een endometriumpoliep. Derhalve is endometriumdiktemeting niet behulpzaam in het opsporen van endometriumpoliepen bij vrouwen met PMB. Nederlandse samenvatting 133 Hoofdstuk 6: Enquête naar poliepverwijdering in Nederland Dit hoofdstuk evalueert de huidige Nederlandse praktijk voor het verwijderen van endometriumpoliepen. In 2003 werd aan alle gynaecologen in Nederland gevraagd om een enquête over het verwijderen van poliepen in te vullen. De meerderheid van de gynaecologen (71%) gaf aan poliepen in een algemene operatiekamer te verwijderen. Hierbij werd veelal (door 77% van de gynaecologen) gebruik gemaakt van algehele of regionale anesthesie (narcose of ruggenprik). De meerderheid van de gynaecologen (69%) gaf aan endometriumpoliepen met behulp van hysteroscopie te verwijderen. Daarentegen werden poliepen vaker verwijderd in een poliklinische setting door gynaecologen in opleidingsziekenhuizen dan door gynaecolo-

135 gen werkzaam in een niet-opleidingsziekenhuis: 39% versus 19% (p<0,001). Geconcludeerd werd dat poliklinische hysteroscopische verwijdering van poliepen nog niet op grote schaal plaatsvindt. In opleidingsziekenhuizen gebeurt dit vaker in een poliklinische setting, derhalve wordt verwacht dat in de toekomst meer gynaecologen poliepen in een poliklinische setting zullen gaan verwijderen zeker met de huidige ontwikkeling van kleinere hysteroscopen en meer mogelijkheden tot poliklinische operatiekamers. Nederlandse samenvatting Hoofdstuk 7: Herhalingskans op bloedverlies bij vrouwen met een eerste episode postmenopauzaal bloedverlies en een endometriumdikte 4 mm In dit hoofdstuk werd gekeken naar de herhalingskans en de impact van herhaald bloedverlies na een eerste episode PMB en een aanvankelijke endometriumdikte 4 mm. Vrouwen met een eerste episode PMB ondergingen TVE. Bij die vrouwen waarbij een endometriumdikte 4 mm werd gevonden, werd een afwachtend beleid gevoerd. Vrouwen werden geïnstrueerd, zich opnieuw te melden bij herhaald of aanhoudend bloedverlies. Er werden 249 vrouwen met endometriumdikte 4 mm geïncludeerd. Van deze 249 vrouwen hadden 25 vrouwen (10%) herhaald of aanhoudend bloedverlies. De mediane tijd tot herhaald bloedverlies bedroeg 49 weken. Twee vrouwen met herhaald bloedverlies bleken een endometriumcarcinoom te hebben (8%) en één vrouw had een maligne melanoom in de vagina. Periode na de overgang, leeftijd, Quetelet-index (body mass index), hypertensie, diabetes en antistolling konden de kans op recidief bloedverlies niet voorspellen. Geconcludeerd werd dat de herhalingskans op bloedverlies na een eerste episode PMB en endometriumdikte 4 mm laag is. Deze herhalingskans kan niet worden voorspeld aan de hand van patiënten-kenmerken. Vrouwen met herhaald PMB hebben een risico op endometriumcarcinoom en moeten daarom opnieuw worden geëvalueerd. 134 Hoofdstuk 8: Herhalingskans op bloedverlies bij vrouwen met een eerste episode postmenopauzaal bloedverlies en endometriumdikte > 4 mm In dit hoofdstuk werd de herhalingskans op PMB geschat van vrouwen na een eerste episode met PMB en endometriumdikte > 4 mm. Bij vrouwen met een eerste episode PMB en een endometriumdikte > 4 mm, werd weefsel voor histologisch onderzoek uit het cavum uteri (baarmoederholte) verkregen. Dit weefsel kon verkregen zijn met behulp van een Pipelle of door middel van hysteroscopie al dan niet gecombineerd met curettage. Vrouwen met benigne histologie (goedaardig weefsel) werden gevolgd. Vrouwen werd geïnstrueerd zich opnieuw te melden bij herhaald of aanhoudend PMB. Er werden 222 vrouwen geïncludeerd, hiervan hadden 47 vrouwen (21%) herhaald bloedverlies. De mediane tijd tot herhaald bloedverlies bedroeg 49 weken. Er werd geen verschil in herhaald bloedverlies gevonden tussen vrouwen waarbij tijdens het eerste diagnostische traject een hysteroscopie werd verricht en vrouwen waarbij alleen een Pipelle was verricht. Voorts werd geen verschil in herhaald bloedverlies gevonden tussen vrouwen bij wie een poliep was verwijderd en vrouwen met een normale

136 hysteroscopie. Twee vrouwen met herhaald bloedverlies bleken atypische hyperplasie van het endometrium (premaligne afwijking van het baarmoederslijmvlies) te hebben. Geconcludeerd werd dat de kans op herhaald bloedverlies bij vrouwen na een eerste episode PMB en een endometriumdikte > 4 mm 21% bedraagt. Deze herhalingskans is niet gerelateerd aan een hysteroscopie, al dan niet met curettage of poliepverwijdering, tijdens het eerste diagnostische traject. Hoofdstuk 9: Ontwerp van een gerandomiseerd geblindeerd onderzoek naar de verwijdering van endometriumpoliepen bij vrouwen met postmenopauzaal bloedverlies In dit hoofdstuk wordt het ontwerp van een gerandomiseerd onderzoek beschreven naar het effect van verwijderen van endometriumpoliepen bij vrouwen met PMB. In dit gerandomiseerd onderzoek ondergingen patiënten met PMB en endometriumdikte > 4 mm een diagnostische hysteroscopie. Vrouwen met een maligniteit (zoals vastgesteld d.m.v. Pipelle) werden uitgesloten voor dit onderzoek. Tijdens de hysteroscopie kon een poliep worden vastgesteld. Indien een poliep werd gediagnosticeerd werd vrouwen gevraagd mee te doen aan het onderzoek. Als vervolgens toestemming was verkregen, werd geloot tussen het wel of niet weghalen van deze poliep. Vrouwen werden vervolgens vervolgd voor de herhalingskans op PMB. Er konden te weinig vrouwen worden geïncludeerd voor deze studie: in twee jaar tijd werden vier patiënten geïncludeerd i.p.v de berekende 60 patiënten die nodig waren. Deze ondermaatse inclusie werd waarschijnlijk veroorzaakt door factoren die hadden te maken met de dokters, die de patiënten moesten informeren over het onderzoek als ook door factoren bij de patiënten zelf. Accepteren van een afwachtend (expectatief) beleid bij een reeds vastgestelde poliep bleek moeilijk te zijn. Ondanks de tegenvallende inclusie is gerandomiseerd onderzoek op dit gebied nog steeds nodig om te evalueren of diagnostiek naar benigne afwijkingen bij PMB zinvol is. Een alternatief ontwerp voor een gerandomiseerd onderzoek wordt hier dan ook voorgesteld. In dit alternatief ontwerp ondergaan alle vrouwen met PMB een TVE. In geval van een endometriumdikte > 4 mm wordt bij alle vrouwen weefselonderzoek door middel van Pipelle verricht. Indien dit benigne histologie (goedaardig weefselonderzoek) laat zien, wordt geloot tussen een afwachtend beleid of verder evaluatie van het cavum uteri (baarmoederholte), hetgeen dan vervolgens met SIS (watercontrastecho) en hysteroscopie wordt verricht. Indien dan een poliep wordt vastgesteld, wordt deze poliep tijdens hysteroscopie verwijderd. Ingeschat wordt dat het voorgestelde gerandomiseerd onderzoek beter haalbaar is, daar het sterk overeenkomt met de huidige dagelijkse praktijk. Nederlandse samenvatting 135 Hoofdstuk 10: Discussie en aanbevelingen Dit proefschrift toont aan dat de diagnostische waarde van transvaginale echo (TVE) en endometriumdiktemeting bij vrouwen met PMB in eerdere studies en meta-analyses is overschat. De diagnostische accuratesse werd opnieuw geëvalueerd door middel van meta-analyse van individuele patiëntendata (IPD), de huidige gouden standaard voor meta-analyses. Bij een

137 afkapwaarde voor endometriumdikte van 4 mm wordt een sensitiviteit van 95% gevonden, hetgeen resulteert in een kans op endometriumcarcinoom van 1,2% bij vrouwen met endometriumdikte 4 mm. Patiënten geven te kennen een dergelijke achterafkans te hoog te vinden. Een afkapwaarde van 3 mm resulteerde in een sensitiviteit van 98%, wat bij een prevalentie van 10% voor endometriumcarcinoom, resulteert in een achterafkans op endometriumcarcinoom van 0,6% bij een endometriumdikte 3 mm. Hoewel patiënten het liefst een achterafkans van 0,0% zien, lijkt zo n achterafkans klinisch niet haalbaar, maar is een achterafkans van 0,6% als zeer acceptabel te duiden. Nederlandse samenvatting 136 Er is verder onderzoek nodig voordat deze afkapwaarde van 3 mm ook in de richtlijnen en in de dagelijkse praktijk moet worden geïmplementeerd. Dit onderzoek moet zich ook richten op de kosteneffectiviteit van verschillende diagnostische trajecten. Hoewel momenteel een diagnostisch traject startend met TVE als meest kosteneffectief wordt beschouwd, kan dit veranderen indien de afkapwaarde van 3 mm wordt gehanteerd. Deze afkapwaarde resulteert immers in meer invasieve diagnostiek. Een alternatief zou kunnen zijn om bij alle vrouwen met PMB direct een Pipelle te verrichten onafhankelijk van gemeten endometriumdikte. Zo n traject zou meer kosteneffectief kunnen blijken te zijn dan een traject startend met TVE waarbij een afkapwaarde van 3 mm wordt gehanteerd. Enerzijds heeft een Pipelle een hoge sensitiviteit van 99.6% en komt daarmee tegemoet aan de hoge zekerheid die patiënten wensen, anderzijds resulteert het mislukken van de Pipelle in meer hysteroscopiën en dus meer invasieve diagnostiek. Het verhogen van de diagnostische accuratesse van de echo door het incorporeren van patiëntenkenmerken (zoals leeftijd, diabetes, hypertensie en body mass index), is een ander alternatief wat verder onderzocht dient te worden. Wat betreft diagnostiek en behandeling van benigne afwijkingen werd geconcludeerd dat de waarde hiervan tot op heden onvoldoende is aangetoond. Derhalve is er nog onvoldoende bewijs om het klinisch geheel te implementeren. Zo blijkt verwijdering van endometriumpoliepen niet te leiden tot een vermindering van het herhaald optreden van postmenopauzaal bloedverlies. Kanttekening bij deze studie is wel dat er niet alleen hysteroscopiën maar ook curettages werden verricht. Bij een curettage worden endometriumpoliepen in 50 to 85% van de gevallen gemist. Een curettage is dan ook niet geschikt voor de diagnostiek of behandeling van endometriumpoliepen. In onze studie werd niet bij alle patiënten een curettage verricht, maar een deel van de groep patiënten onderging alleen een hysteroscopie met poliep verwijdering. Geconcludeerd kan worden dat in een deel van de groep diagnostiek en behandeling volgens de huidige standaard plaatvond. Dit resulteerde niet in een lagere incidentie van bloedverlies in de groep met hysteroscopie en/of curettage. Een gerandomiseerd onderzoek naar de waarde van poliepverwijdering bij vrouwen met PMB mislukte. De aanbeveling in de NVOG richtlijn dat diagnostiek van het cavum uteri kan worden verricht bij vrouwen met een endometriumdikte > 4 mm, is gericht op het opsporen van endometriumpoliepen. Hiertoe

138 wordt SIS of poliklinische hysteroscopie geadviseerd. Er is echter onvoldoende bewijs om te stellen dat dit resulteert in gezondheidswinst, zoals reductie van de kans op bloedverlies of afname van andere morbiditeit. Ons voorstel is dan ook om de waarde van de SIS en de poliklinische hysteroscopie bij vrouwen met PMB verder te evalueren in een multicenter onderzoeksverband. In zo n gerandomiseerd onderzoek in de vorm van de POstMenoPauzaal bloedverlies ONderzoek (POMPOEN)-studie zullen patiënten met een endometriumdikte > 4 mm gerandomiseerd worden tussen een expectatief beleid na een benigne Pipelle versus SIS en poliklinische hysteroscopie en zo nodig poliepverwijdering. Deze studie doet recht aan de huidige richtlijn en probeert uiteindelijk een antwoord te vinden op de vraag of cavum diagnostiek bij vrouwen met PMB een zinvolle interventie is. Nederlandse samenvatting 137

139

140 List of co-authors and their affiliations

141 List of co-authors and their affiliations 140 Lucas M. Bachmann PhD Horten Centre, University of Zurich, Switzerland Shagaf H. Bakour MD PhD Department of Obstetrics and Gynaecology, Birmingham Women s Health Care NHS Trust, Birmingham, UK Marlies Y. Bongers MD PhD Department of Obstetrics and Gynaecology, Maxima Medical Centre, Veldhoven, the Netherlands Thierry van den Bosch MD PhD Department of Obstetrics and Gynaecology, University Hospital Leuven, Leuven, Belgium Annette M. Bouwmeester MD Department of Obstretrics and Gynaecology, MESOS Medical Centre, Utrecht, the Netherlands Sharon M. Cameron MD PhD Department of Reproductive and Developmental Sciences, University of Edinburgh, UK T. Justin Clark MD PhD Department of Obstetrics and Gynaecology, Birmingham Women s Health Care NHS Trust, Birmingham, UK Salvatore Dessole MD PhD Department of Gynaecology, University of Sassari, Sassari, Italy Heleen van Dongen MD Department of Obstretrics and Gynaecology, Leiden University Medical Centre, Leiden, the Netherlands Lena C. van Doorn MD PhD Department of Gynaecological Oncology, Erasmus Medical Centre, Rotterdam, the Netherlands M. Jitze Duk MD PhD Department of Obstetrics and Gynaecology, Meander Medical Centre, Amersfoort, the Netherlands F. Paul H.L.J. Dijkhuizen MD PhD Department of Obstetrics and Gynaecology, Hospital Rijnstate, Arnhem, the Netherlands Elisabeth Epstein MD PhD Department of Obstetrics and Gynaecology, Lund University Hospital, Lund, Sweden Maaike B.E. Gerritse MD Department of Perinatology and Gynaecology, University Medical Centre, Utrecht, the Netherlands M. Gabriella Giusa MD PhD Department of Gynaecology, Escola Paulista de Medicina, Federal University of Sao Paolo, Brazil Janesh K. Gupta MD PhD Department of Obstetrics and Gynaecology, Birmingham Women s Health Care NHS Trust, Birmingham, UK Frank Willem Jansen MD PhD Department of Obstetrics and Gynaecology, Leiden University Medical Centre, Leiden, the Netherlands Tjitkse C. van Kerkvoorde Department of Obstetrics and Gynaecology, St. Antonius Hospital, Nieuwgein, the Netherlands Khalid S. Khan MD PhD Department of Obstetrics and Gynaecology, Birmingham Women s Health Care NHS Trust, Birmingham, UK G. Sjarlot Kooi MD PhD Department of Obstretrics and Gynaecology, Schweitzer Hospital, Dordrecht, the Netherlands

142 Maurice V.A.M. Kroeks MD PhD Roy F.M.P. Kruitwagen MD PhD Ben W.J. Mol MD PhD Brent C. Opmeer PhD Gerben Ter Riet Md PhD Sebastiaan Veersema MD Lucet F. van der Voet MD Peter H.M. van de Weijer MD PhD Department of Obstetrics and Gynaecology, Diakonessen Hospital, Utrecht, the Netherlands Department of Obsterics and Gynaecology, TweeSteden Hospital, Tilburg, the Netherlands Department of Obstetrics and Gynaecology, University Medical Centre, Maastricht, the Netherlands Department of Obstetrics and Gynaecology, Academic Medical Centre, Amsterdam, the Netherlands Department of Obstetrics and Gynaecology, Maxima Medical Centre, Veldhoven, the Netherlands Department of Clinical Epidemiology, Biostatistics and Bio-informatics. Academic Medical Centre, Amsterdam, the Netherlands Department of General Practice, Academic Medical Centre, Amsterdam, the Netherlands Department of Obstetrics and Gynaecology, St. Antonius Hospital, Nieuwegein, the Netherlands Department of Obstetrics and Gynaecology, St. Antonius Hospital, Nieuwegein, the Netherlands Department of Obstretrics and Gynaecology, Deventer Hospitals, Deventer, the Netherlands Department of Obstetrics and Gynaecology, Gelre Hospitals, Apeldoorn, the Netherlands List of co-authors and their affiliations 141

143

144 Dankwoord

145 Onderzoek doe je niet alleen! In dit dankwoord wil ik iedereen bedanken dankzij wie (direct of indirect betrokken) dit proefschrift tot stand is gekomen, ik hoop daarbij niemand te vergeten. Allereerst patiënten die in de verschillende databases zijn verdwenen : zonder jullie had dit onderzoek nooit plaats kunnen vinden. Zulke databases zijn erg waardevol waarin de gegevens van verschillende patiënten anoniem worden verwerkt. Een speciaal woord van dank aan alle patiënten die zich hebben laten interviewen voor de preferentiestudie. Het vraagstuk wat jullie voorgelegd kregen was niet makkelijk, maar jullie kijk op het vraagstuk, is erg verhelderend geweest. Collega s van verschillende ziekenhuizen die zeer trouw gegevens hebben verzameld van patiënten met postmenopauzaal bloedverlies, ook jullie wil ik bedanken voor het verzamelen van al die gegevens en later nog terug gaan naar de statussen als dit nodig was! Het onderzoek De onderzoeksgroep met Prof. dr. B.W. J. Mol, drs. S. Veersema, dr. B.C. Opmeer, dr. F.W. Jansen: Bas, als beginnende AIOS begon ik samen met jou aan een klein onderzoekje wat uiteindelijk hoofdstuk 4 in dit proefschrift is geworden. Jij bracht me met Ben-Willem in contact en uiteindelijk was dat kleine onderzoekje naar de poliepsnaar en de hysteroscopie het begin van een mooi onderzoeksproject. Zonder jou, was dit boekje niet begonnen en geëindigd! Dankwoord 144 Ben-Willem, tussen het door een design stoel zakken in Nieuwegein en groene thee drinken in Almere, zaten een aantal onderzoeksjaren waarin dit boekje tot stand is gekomen. Je was er altijd, zelfs om 5 uur s nachts als ik tijdens een nachtdienst even was gaan liggen en jij naar de VK belde voor overleg. Ik ben je dankbaar voor al het vertrouwen wat je in me hebt en al het geduld dat je met me hebt (gehad). Brent, met de preferentie studie raakte jij betrokken bij dit proefschrift en bleef je betrokken, ook waar je misschien niet rechtstreeks bij het onderzoek betrokken was geweest. Die paar maanden die ik voltijds op jouw afdeling door heb gebracht, heb ik als zeer waardevol ervaren. Je deur stond altijd open, als ik weer eens de juiste knop op SPSS niet kon vinden. Ik zal je meer malen tot wanhoop hebben gedreven met deze vragen, maar meer nog met mijn verzoekjes om even en liefst ZSM een bepaald plaatje te helpen construeren of een analyse te herhalen. Ik ben blij dat je zo betrokken bent geweest bij dit proefschrift. Frank-Willem, vanaf mijn eerste praatje op een endoscopie congres (ESGE Luxemburg 2003) toonde jij je betrokkenheid. Deze betrokkenheid resulteerde uiteindelijk tot een aantal onderzoeken en je werd mijn co-promotor. Jij zorgde altijd voor een endoscopische en klinisch blik, naast alle wetenschappelijke evidentie.

146 De leden van de commissie Prof. dr. M.J. Heineman, Prof. dr. M.P.M. Burger, Prof. dr. E. Schadé, Prof. dr. H.A.M. Brölmann, Prof. dr. M.E. Vierhout, Dr. J.B. Reitsma, Dr. T.J. Clark wil ik danken voor hun bereidheid om het manuscript te beoordelen en plaats te willen nemen in de commissie. Dr. Clark, Dear Justin, I feel honoured that you are one of my opponents. Alle artsen van de deelnemende klinieken: dank voor jullie inzet om patiënten te includeren in de verschillende cohortstudies en in de trial. Marlies Bongers dank voor je betrokkenheid, ook al liep onze trial uiteindelijk niet zoals verwacht, dank voor je enthousiasme, hopelijk gaan we het met de POMPOEN beter doen. Lena van Doorn, dank voor de prettige samenwerking en je immer kritische blik. Jouw DUPOMEB heeft toch nog mooi twee extra artikelen opgeleverd. Secretariaat: St. Antonius ziekenhuis, Andrea, Coby, Kitty en Ineke, jullie hielden trouw de lijsten met PMB patienten bij, maakten zelfs een begin van mijn database en waren altijd bereid om ZSM statussen bij elkaar te zoeken! Bertina, ook al had je met mijn hele onderzoek niets te maken, je hebt me met regelmaat met kleine en grotere klussen geholpen. Secretariaat gynaecologie AMC, Ingrid: het manuscript moest geprint, en dan het liefst niet 1 letter per pagina! Ik was niet bereikbaar en jullie hebben het toch mooi gechefd! Jullie allemaal zijn super en ik kon ook nog altijd voor een praatje bij jullie terecht! Medeonderzoekers AMC en mede-uitvinders UMC Utrecht; als vreemde eend in de bijt (UMC AIOS die in het AMC promoveert): ik wil jullie bedanken dat ik op maandag mee mocht lunchen als dat zo uitkwam (UMC Utrecht) en dat jullie ondanks mijn afwezigheid in het AMC toch altijd betrokken waren. Vooral Marsha van Leeuwen en Sjors Coppus met wie het altijd goed toeven was op de KEBB. Dankwoord 145 Tijdelijke collega s op de KEBB (AMC): gedurende 5 maanden verbleef ik op jullie afdeling, waar ik me welkom voelde en veel geleerd heb, waarvoor veel dank! De Opleiding Gynaecologen St. Antonius ziekenhuis Nieuwegein: Ton Hameeteman, Jules Schagen van Leeuwen, Sien The, Lucie Ribbert, Peppino Graziosi, Eric van Beek, Jessica van der Leij en Bas Veersema. Bij jullie leerde ik de eerste kneepjes van het vak als AGNIO en vervolgde ik mijn weg als AIOS. Ik wil jullie niet alleen danken voor de prettige samenwerking en mijn vorming tot gynaecoloog, maar ook voor de database met patiënten met postmenopauzaal bloed verlies, die door jullie toedoen gevuld kon worden. Het was vertrouwd om bij jullie de opleiding en dit onderzoek te beginnen en na al die tijd nog steeds terug te kunnen komen.

147 Prof. dr. A.P.M. Heintz, Prof. dr. G.H.A. Visser en Prof. dr. H.W. Bruinse: onder jullie leiding werd ik als gynaecoloog academisch verder gevormd. Dank voor jullie interesse in dit onderzoek ook al vond dat niet direct onder jullie hoede of de hoede van het UMC plaats. Tenslotte gynaecologen TweeSteden ziekenhuis Tilburg: Thierry van Dessel, Roy Kruitwagen, Tanja Roosen, Hanny Pijnenborg, Bas Keijser, Marjan Stegeman, Annette ter Haar, Bart Broekman en Addy Drogtrop. Bij jullie kreeg ik de kans mijn opleiding en ook dit proefschrift te voltooien. Ik heb veel bij jullie en van jullie geleerd, al zullen sommige zeggen dat ik hardleers ben... Ik ben ook erg blij dat ik nog wat langer in jullie vakgroep en onder jullie vleugels kan verblijven en de endoscopie me nog meer eigen kan en mag maken. Het was een feest om mijn opleiding bij jullie te eindigen. Collega s: collega s in het St. Antonius ziekenhuis in Nieuwegein, in het UMC Utrecht en in het TweeSteden Ziekenhuis te Tilburg, bedank ik voor de fijne sfeer waarin ik altijd opleiding en onderzoek heb kunnen combineren. Het is dankzij jullie dat ik af en toe een onderzoekdag kon organiseren, een afspraak met promotor kon plannen of een cursus epidemiologie kon volgen. Een aantal van jullie wil ik met name noemen, door jullie werd een wetenschappelijke stage mogelijk: Diana Helmink, Laura Zuurendonk, Ayten Elvan, Mark van der Laan, Yvo Laeven, Pieter van Runnard Heimel, Ralph Niewenweg, Ilse van Rooij en Fleur Koenders. Verder Merel Breijer, ik ben blij dat jij dit onderzoek voort gaat zetten en ik ben er trots op dat we dat nog een tijdje samen gaan doen. Dankwoord 146 Vrienden en Vriendinnen Lieve Vrienden en Vriendinnen! Van middelbare school via oud-huisgenoten, oud-collega s tot de kroeg, jullie waren er voor de gezelligheid of voor de afleiding. Balletchica s (Melissa, Liza, en Angela) na al die jaren nog steeds door dik en dun, wat ben ik blij met jullie, trouwe vriendinnen! Lieve Angela: Wat is de voorkant mooi geworden! Lieve Claire en Melissa: wat hebben jullie fijn meegedacht en wat kon ik fijn gebruik maken van jullie SIS-ontwerp expertise (en dan heb ik het niet over de watercontrastecho). Geitjes (Maaike, Maartje, Maria, Marieke, Hester en Hester) de 5 e en laatste promovenda van dit stel: zie de stellingen. Anouk en Martine, allebei oud-huisgenoten (maar niet hetzelfde huis), in korte of langere tijd zijn jullie me zeer dierbaar geworden, jullie zijn er gewoon als het nodig is. Robbert, de start van dit proefschrift maakte je van dichtbij mee, ik ben blij dat je nu ook het einde kunt bewonderen. Maaike en Job! Na jullie Merel op de wereld te hebben gezet, kunnen jullie nu de geboorte van mijn kindje meemaken. In Mallorca kreeg ik alle tijd, rust en ruimte om de laatste weeën op te vangen.

148 Paranimfen Lieve Heleen, dat jij mijn paranimf zou worden was me al snel duidelijk. Maar kort samengewerkt in Nieuwegein, maar lang bij elkaars onderzoek betrokken en een trouwe vriendin geworden! Ik ben er ook trots op dat ik op 26 februari naast jou mag staan. Lieve Sarah, klein zusje! Naast mijn zusje ben je een lief vriendinnetje en straks word je ook nog collega. Fijn dat je me zo geholpen hebt en naast me wilt staan. Familie Lieve Joris en Lucas, lieve broertjes! Joris, als enige niet medicus is het soms vechten tegen de bierkaai, terwijl jij toch echt de enige echte wetenschapper bij ons bent. Lucas, je zult altijd ons kleine broertje blijven, al ben je ook hard op weg om een echte dokter te worden en ben je met je 1.93 niet meer klein te noemen. Fijn om zulke trouwe broertjes en zusje te hebben. Lieve Papa en Mama! Drie dokters (onder weg) en drie doctors (onder weg). Ik weet hoe trots jullie op ons zijn. Papa zoals je zelf al zei, met drie medici en één fysicus heb je de strijd verloren. Papa en Mama jullie hebben me altijd in alles gesteund, ook al waren mijn keuzes soms lastig. Het doorzettingsvermogen wat nodig was voor dit boekje heb ik van jullie gekregen. Het is nog steeds heerlijk om thuis te komen in het warme nest waarin jullie ons hebben opgevoed! Dankwoord 147

149

150 About the author

151

152 Anne Timmermans was born on September 12, 1976 in Nijmegen, the Netherlands. In 1994 she graduated from secondary school at the Stedelijk Gymnasium Nijmegen. She attended medical school at the University of Maastricht from During her medical studies, for a period of six months, she was involved in a research project The role of VEGF and sflt-1 in predicting pre-eclampsia at the Department of Obstetrics and Gynaecology of the New York University School of Medicine, United States of America (Dr. E.F. Funai), with Dr. J.P.M. Offermans (Department of Obstetrics and Gynaecology of the Maastricht University Hospital) as her Dutch Supervisor. After obtaining her medical degree in May 2001 she worked as a resident at the Department of Obstetrics and Gynaecology of the St. Antonius Hospital in Nieuwegein for a period of one year. In 2002 she started her training in Obstetrics and Gynaecology at the same St. Antonius Hospital Nieuwegein (Dr. H.S. The). During this period she also started the first study on postmenopausal bleeding under supervision of Drs. S. Veersema (St. Antonius Hospital Nieuwegein), and continued with the studies as described in this thesis (promoters: Prof. Dr. B.W.J. Mol [Academic Medical Centre University of Amsterdam], Dr. F.W. Jansen [Leiden University Medical Centre] and Dr. B.C. Opmeer [Academic Medical Centre University of Amsterdam]). She received the Raoul Palmer Award for her study on Patients preferences in the evaluation of postmenopausal bleeding at the Annual congress of the European Society of Gynaecological Endoscopy in 2005, Athens, Greece. She continued her official residency in Obstetrics and Gynaecology at the University Medical Centre Utrecht (Prof. Dr. A.P.M. Heintz and Prof. Dr. G.H.A. Visser) and will finish her residency training on February 12, 2009, at the Tweesteden Ziekenhuis Tilburg (Dr. H.J.H.M. van Dessel). Sarah Timmermans (paranimf) About the author 151

153

154 List of publications