Reproductive Medicine and Surgery Unit, University of Sheffield, Sheffield Teaching Hospitals, Jessop Wing, Tree Root Walk, Sheffield S10 2SF, UK 2

Transcription

1 RBMOnline - Vol 19. No Reproductive BioMedicine Online; on web 19 August 2009 Article How common is polycystic ovary syndrome in recurrent miscarriage? Dr Karen Cocksedge began her research career in the very different field of astrophysics. She has a degree in Physics from Oxford University and a PhD in Astrophysics from Manchester University. She published over 60 papers in this field and the calibre of her work was recognised through the award a Royal Society Research Fellowship. In 2006, Dr Cocksedge became a medical student at Sheffield University. On the basis of her previous experience, she began research with Professor TC Li in reproductive medicine whilst continuing her studies. This study is her third publication in the field of recurrent miscarriage and PCOS. She has recently been awarded a Foulkes Foundation Fellowship on the basis of her research in this field. Dr Karen Cocksedge KA Cocksedge 1,2, SH Saravelos 1, M Metwally 1,TCLi 1 1 Reproductive Medicine and Surgery Unit, University of Sheffield, Sheffield Teaching Hospitals, Jessop Wing, Tree Root Walk, Sheffield S10 2SF, UK 2 Correspondence: mdc06kc@sheffield.ac.uk Abstract The association between polycystic ovary syndrome (PCOS) and recurrent miscarriage (RM) has been long established, but the relative importance of this condition as a cause of RM is far from clear. Previous studies on the prevalence of PCOS in RM have been hampered by a lack of objective and universally accepted criteria for the diagnosis of PCOS, resulting in considerable controversy. However, the Rotterdam criteria have since been accepted as the gold standard for diagnosis of PCOS, and therefore these criteria have been used to produce a much clearer and more objective assessment of the prevalence of PCOS in RM. Three hundred women with recurrent miscarriage were studied. A diagnosis of PCOS was established via measurement of cycle length and day 21 serum progesterone, determination of the free androgen index and pelvic ultrasonography. All ultrasound reports prior to publication of the Rotterdam criteria were reviewed, ensuring consistency in the diagnosis of a polycystic ovary. Ultrasound scans of 27 patients confirmed polycystic ovaries with a further 10 scans suggestive of polycystic ovaries, but with insufficient information for the Rotterdam criteria to be applied. Hence, (9.0 12%) patients presented with ultrasonographic polycystic ovaries. Using the Rotterdam criteria, (8.3 10%) patients had PCOS. It is concluded that the prevalence of PCOS in RM is considerably lower than has previously been accepted. Keywords: diagnosis, polycystic ovary syndrome, prevalence, recurrent miscarriage 572 Introduction Recurrent miscarriage (RM), defined as the loss of three or more consecutive pregnancies, is a heterogeneous condition affecting approximately 1% of women (Stirrat, 1990). A number of causes of this condition have been identified, including antiphospholipid antibody syndrome, uterine abnormalities, parental chromosomal abnormalities and polycystic ovary syndrome (PCOS) (Li et al., 2002). PCOS is a common endocrine disorder that typically presents with a wide spectrum of clinical features, including infertility, hirsutism, acne, obesity and alopecia (Norman et al., 2007). Further investigation typically demonstrates chronic anovulation, ultrasonographic evidence of polycystic ovaries, elevated serum androgens and insulin resistance (Norman et al., 2007). The association between PCOS and RM has long been established in the medical literature, with the majority of publications still referring to the study of Sagle et al. (1988), which demonstrated that a very high proportion (82%) of their RM population had identifiable polycystic ovaries. However, subsequent studies that investigated the prevalence of PCOS in RM reported a wide variation in prevalence of between 5% and 81% (Tulppala et al., 1993; Watson et al., 1993; Liddell et al., 1997; Okon et al., 1998; Bussen et al., 1999; Li et al., 2000; Rai et al., 2000; Diejomaoh et al., 2002; Yang et al., 2006). Hence, whilst the association between the two conditions is generally accepted, the relative importance of the contribution of PCOS to RM is far from clear (Cocksedge et al., 2008a). The wide variation in prevalence can be attributed to a number of factors. Most importantly, there is significant variation in the criteria used to define PCOS. The majority of Ó 2009 Published by Reproductive Healthcare Ltd, Duck End Farm, Dry Drayton, Cambridge CB23 8DB, UK

2 studies, including that of Sagle et al. (1988), used the presence of ultrasonographic polycystic ovaries alone to define PCOS when this does not necessarily indicate the complete syndrome of PCOS (Balen et al., 2003). Furthermore, different criteria have been used to determine polycystic ovaries in these studies and all are somewhat subjective, requiring an increased or abnormal stroma without precise definition (Stray-Pedersen and Stray-Pedersen, 1984; Adams et al., 1986; Ardaens et al., 1991). The remaining studies used different criteria to define PCOS: Okon et al. (1998) used either ultrasonagraphic polycystic ovaries or elevated LH and Yang et al. (1996) used clinical findings and either ultrasound or biochemical testing but did not describe which biochemical tests or clinical findings were used. To resolve the controversy on the prevalence of PCOS in RM, a more reliable and universally accepted definition of PCOS is required. In 2003, the diagnostic criteria for PCOS were agreed during a Consensus Workshop in Rotterdam [Rotterdam European Society for Human Reproduction and Embryology (ESHRE)/American Society for Reproductive Medicine (ASRM)-Sponsored PCOS Consensus Workshop Group, 2004]. These criteria state that a positive diagnosis should be made when at least two of the following features are present: oligo- or anovulation, clinical and/or biochemical signs of hyperandrogenism and ultrasonographic polycystic ovaries. Furthermore, these criteria include the first clear and objective definition of the polycystic ovary. Since none of the previous publications used these now widely accepted criteria to examine the prevalence of PCOS in women with RM, the prevalence has been reappraised using the Rotterdam criteria. Materials and methods This study was conducted at the Recurrent Miscarriage Clinic of the Jessop Wing of the Royal Hallamshire Hospital, Sheffield, UK. Between July 1992 and September 2007, investigations were completed on a total of 300 women who presented to the clinic with three or more consecutive miscarriages. These investigations followed an established protocol (Li, 1998), which included chromosome analysis of both partners, testing for anticardiolipin antibodies and lupus anticoagulant, coagulation studies, transvaginal ultrasonography, hysterosalpingogram and endocrinological investigations. Endocrinological investigations included serum measurements of day 2 5 FSH, LH, oestradiol, prolactin, testosterone, androstenedione, sex hormone-binding globulin, thyroid function test and day 21 progesterone. In order to make a diagnosis of PCOS using the Rotterdam criteria, a minimum of two of the following features are required: (i) Oligo/anovulation. The Rotterdam Workshop did not define what precisely constitutes oligo/anovulation, but instead noted that a range in ovulatory function is seen in PCOS, ranging from chronic oligo/amenorrhoea through to regular cycles. Although oligo/anovulation typically presents as oligo/amenorrhoea, it is also possible for anovulation to present with regular cycles. In view of this and in the absence of a clear definition provided by the Rotterdam Workshop, biochemical criteria have been used to define anovulation, based on the day 21 progesterone <30 nmol/ l, which is commonly used in the UK (Li et al., 2000). For oligoovulation, the definition of oligomenorrhoea was adopted, and therefore cycle length 36 days was used. Hence, in this study, oligo/anovulation was considered to be present if either the cycle length 36 days or the day 21 progesterone <30 nmol/l. (ii) Hyperandrogenism. The Rotterdam Workshop recommended that hyperandrogenaemia be assessed by either free testosterone or the free androgen index (FAI). In this study, an FAI >5 was used to define hyperandrogenaemia (Cocksedge et al., 2008b) and this factor alone was used to define hyperandrogenism rather than also including the somewhat subjective diagnosis of hirsutism or acne to define hyperandrogenism. (iii) Ultrasonographic polycystic ovaries. The Rotterdam Workshop defined a polycystic ovary as one containing either 12 or more follicles of 2 9 mm diameter or with an increased ovarian volume (>10 ml). Furthermore, only one ovary satisfying this condition was deemed sufficient to define polycystic ovaries. Since the definition of the polycystic ovary has varied considerably during the timescale of this study, all ultrasound reports prior to the publication of the Rotterdam criteria were reviewed, to ensure consistency. Androgen measurements From 1992 until 2001, the total serum testosterone (T) was measured using a coated tube radio-immunoassay (Coat-A- Count, Diagnostic Products Corporation, USA) and from 2001 onwards it was measured using an automated chemiluminescent immunoassay method (Advia Centaur, Bayer, Germany). From 1992 until 2000, the sex hormone-binding globulin (SHBG) was measured using a column saturation method published by Iqbal and Johnson (1977) and from 2000 onwards it was measured using an automated chemiluminescent immunoassay (Immulite analyser; Diagnostic Products Corporation). The intra- and inter-assay coefficients of variation for the measurements were: T ( ) 7.5%, 8.0%; T (2001 onwards) 6.2%, 4.4%; SHBG (2000 onwards) 6.8%, 9.4%. FAI was calculated using FAI = (T/SHBG) 100. Progesterone measurements Progesterone was measured using automated chemiluminescent immunoassay methods. From 1992 until 1999, the Immulite analyser (Diagnostic Products Corporation) was used and from 1999 onwards, the Advia Centaur (Bayer) was used. The intra- and inter-assay coefficients of variation for the measurements were: %, 12.4%; 1999 onwards 4.6%, 5.5%. 573

3 Results The mean (±SD) age of the women studied (n = 300) was 32.1 (±5.2) years. One hundred and seventy-six women (59%) had primary recurrent miscarriage and the remaining 124 (41%) had secondary recurrent miscarriage. The median (range) number of miscarriages prior to referral was three (3 14). Of the 300 patients studied, the ultrasound scan reports of 27 patients confirmed a diagnosis of polycystic ovaries using the Rotterdam criteria. A further 10 patients had scan reports suggestive of polycystic ovaries, but with insufficient information for the Rotterdam criteria to be applied. Hence, a total of patients (9.0 12%) presented with ultrasonographic polycystic ovaries. A total of 119 (40%) presented with oligo/anovulation (of which 91% were anovulatory) and 31 (10%) presented with hyperandrogenaemia (Table 1). The number of patients with the various combinations of positive diagnoses is best illustrated with a Venn diagram (Figure 1). Using the Rotterdam criteria, it was found that (8.3 10%) patients have PCOS. Of the patients with PCOS, 90 92% have oligo/anovulation, 63 72% have hyperandrogenaemia and 56 70% have ultrasonographic polycystic ovaries. Of the patients who presented with oligo/anovulation, 43 52% were purely anovulatory, % presented purely with oligoovulation and 41 48% presented with both anovulation and oligoovulation. Discussion This study is the first to have employed the Rotterdam criteria to investigate PCOS in women with RM. It has been found that the prevalence is %, which is significantly lower than the value initially quoted in the literature (82%; Sagle et al., 1988) and at the lower end of the range derived from other studies (5 81%) (Tulppala et al., 1993; Watson et al., 1993; Liddell et al., 1997; Okon et al., 1998; Bussen et al., 1999; Li et al., 2000; Rai et al., 2000; Diejomaoh et al., 2002; Yang et al., 2006). Prior to the publication of these criteria, the diagnosis of polycystic ovaries by ultrasound was often criticized as subjective and without consistency. The Rotterdam criteria have introduced objectivity into the ultrasound assessment and have therefore been widely welcomed. In view of the fact that the Rotterdam ultrasound criteria were introduced during the duration of the study, all ultrasound reports prior to the publication of these criteria were reviewed, to ensure consistency in the diagnosis of polycystic ovary. It was found that % of patients have ultrasonographic polycystic ovaries compared with the wide range in prevalence previously reported (5 82%) (Sagle et al., 1988; Tulppala et al., 1993; Watson et al., 1993; Liddell et al., 1997; Bussen et al., 1999; Li et al., 2000; Rai et al., 2000; Diejomaoh et al., 2002). Whilst the definition of the polycystic ovary from the Rotterdam criteria is both precise and objective, it is recognized that the remainder of the criteria defining PCOS, although substantially improved, are still not perfect. Some difficulty was found in applying these criteria, since Rotterdam do not precisely define either hyperandrogenism or oligo/anovulation. As detailed above, it was felt that the diagnoses of hirsutism and acne are both too subjective to be used precisely for research purposes, and therefore hyperandrogenaemia alone was used as the definition of hyperandrogenism. Hence it is accepted that this does not strictly adhere to the Rotterdam criteria, but it is the most practical application available. Furthermore, the Rotterdam criteria do not define oligo/anovulation. As detailed above, oligo/anovulation frequently presents as oligo/amenorrhoea but anovulation could also present as regular cycles. In view of this, defining oligo/anovulation as oligo/amenorrhoea would involve underestimation of the prevalence. Hence, it is therefore preferable to use biochemical criteria to define anovulation and in this respect, a limit of <30 nmol/l has been used, since this is the value traditionally used in the UK (Li et al., 2000). However, it is recognized that values quoted in the literature range from 9.6 to 38 nmol/l (Malcolm and Cumming, 2003). Since a surprisingly high proportion of patients with oligo/anovulation (40%, the majority of whom were anovulatory) were found, this could suggest that the progesterone concentration used was too high. If, instead, the suggested lower value of 9.6 nmol/l was used, then this reduces the prevalence of oligo/anovulation from 40% to 20%, and the prevalence of PCOS from 10 to 7.0% (or from 8.3% to 6.0%, using only the confirmed cases of ultrasonographic polycystic ovaries). Hence, the change to the prevalence calculation is relatively insignificant and certainly does not detract from the conclusion that the prevalence is considerably lower than previously considered. In fact, it is found that the prevalence of PCOS in this study is not too dissimilar to the prevalence of PCOS in the general population as determined by other recent studies using the Rotterdam criteria ( %; De Sutter et al., 2008; Mueller et al., 2008). It would therefore be of interest to Table 1. The prevalence of the various diagnostic factors for polycystic ovary syndrome (PCOS) (oligo- or anovulation, hyperandrogenaemia, polycystic ovaries on ultrasound) and the prevalence of PCOS according to the Rotterdam criteria for the patients with recurrent miscarriage (n = 300). Factor Prevalence n (%) 574 Oligo- or anovulation 119 (40) Hyperandrogenaemia 31 (10) Ultrasonographic polycystic ovaries (9.0 12) PCOS (8.3 10)

4 diagnosis of PCOS. It is believed that they most accurately reflect the broad spectrum of the disease, since they are the only criteria to incorporate ultrasonographic evidence, hyperandrogenism and oligo/anovulation. They are also the only criteria to objectively define a polycystic ovary. As a result of the application of these criteria, the present study has provided clarity to the prevalence of PCOS in RM. It is concluded that approximately 10% of the women with RM fulfil the criteria for diagnosis of PCOS, which is considerably lower than has previously been accepted. Although this result suggests that PCOS is not one of the main contributors to RM, this prevalence is by no means clinically insignificant. For those approximately 10% of women who are diagnosed with RM associated with PCOS, further work is needed to determine whether the diagnosis of PCOS is a prognostic factor for subsequent miscarriage. Acknowledgements The authors thank staff nurses Barbara Anstie and Kathryn Wood for their help in entry of clinical data, David Drew for his help with the database of laboratory measurements and Martin Loxley for his advice on assay measurements. Figure 1. Venn diagrams to illustrate the proportion of patients with recurrent miscarriage who have oligo- or anovulation, hyperandrogenaemia and/or polycystic ovaries on ultrasound, including (a) the minimum number of patients with ultrasonographic polycystic ovaries (n = 27) and (b) the maximum number of patients with ultrasonographic polycystic ovaries (n = 37). FAI = free androgen index. conduct a population study to directly compare the prevalence of PCOS in RM with the prevalence in the general population. It should be noted that within the literature, there is still debate by some groups as to whether alternative criteria, such as the previous 1990 National Institutes of Health (NIH) criteria, may be more appropriate (Norman et al., 2007). One of the strengths of this study is that the values shown in the Venn diagrams of Figure 1 can be used to determine the prevalence of PCOS using alternative criteria such as the NIH criteria, or in the future if the Rotterdam criteria are subsequently modified and improved. Despite the recognized problems with the Rotterdam criteria, they are generally regarded as the gold standard in the References Adams J, Polson DW, Franks S 1986 Prevalence of polycystic ovaries in women with anovulation and idiopathic hirsutism. British Medical Journal (Clinical Research) 293, Ardaens Y, Robert Y, Lemaitre L et al Polycystic ovarian disease: contribution of vaginal endosonography and reassessment of ultrasonic diagnosis. Fertility and Sterility 55, Balen AH, Laven JS, Tan SL 2003 Ultrasound assessment of the polycystic ovary: international consensus definitions. Human Reproduction Update 9, Bussen S, Sutterlin M, Steck T 1999 Endocrine abnormalities during the follicular phase in women with recurrent spontaneous abortion. Human Reproduction 14, Cocksedge KA, Li TC, Saravelos SH et al. 2008a A reappraisal of the role of polycystic ovary syndrome in recurrent miscarriage. Reproductive BioMedicine Online 17, Cocksedge KA, Saravelos SH, Wang Q et al. 2008b Does free androgen index predict subsequent pregnancy outcome in women with recurrent miscarriage? Human Reproduction 23, De Sutter P, Dutré T, Vanden Meerschaut F et al PCOS in lesbian and heterosexual women treated with artificial donor insemination. Reproductive BioMedicine Online 17, Diejomaoh MF, Al-Azemi M, Jirous J et al The aetiology and pattern of recurrent pregnancy loss. Journal of Obstetrics and Gynaecology 22, Iqbal MJ, Johnson MW 1977 Study of steroid protein binding by a novel two-tier column employing Cibacron Blue F3G-A- Sepharose 4B. I Sex hormone binding globulin. Journal of Steroid Biochemistry 8, Liddell HS, Sowden K, Farquhar CM 1997 Recurrent miscarriage: screening for polycystic ovaries and subsequent pregnancy outcome. Australian and New Zealand Journal of Obstetrics and Gynaecology 37, Li TC 1998 Recurrent miscarriage: principles of management. Human Reproduction 13, Li TC, Makris M, Tomsu M et al Recurrent miscarriage: aetiology, management and prognosis. Human Reproduction Update 8,

5 Li TC, Spuijbroek MD, Tuckerman E et al Endocrinological and endometrial factors in recurrent miscarriage. British Journal of Obstetrics and Gynaecology 107, Malcolm CE, Cumming DC 2003 Does anovulation exist in eumenorrheic women? Obstetrics and Gynecology 102, Mueller A, Gooren LJ, Naton-Schotz S et al Prevalence of polycystic ovary syndrome and hyperandrogenemia in femaleto-male transsexuals. Journal of Clinical Endocrinology and Metabolism 93, Norman RJ, Dewailly D, Legro RS et al Polycystic ovary syndrome. Lancet 370, Okon MA, Laird SM, Tuckerman EM et al Serum androgen levels in women who have recurrent miscarriages and their correlation with markers of endometrial function. Fertility and Sterility 69, Rai R, Backos M, Rushworth F et al Polycystic ovaries and recurrent miscarriage a reappraisal. Human Reproduction 15, Rotterdam ESHRE/ASRM-Sponsored PCOS Consensus Workshop Group 2004 Revised 2003 consensus on diagnostic criteria and long-term health risk related to polycystic ovary syndrome (PCOS). Human Reproduction 19, Sagle M, Bishop K, Ridley N et al Recurrent early miscarriage and polycystic ovaries. British Medical Journal 297, Stirrat GM 1990 Recurrent miscarriage I: definition and epidemiology. Lancet 336, Stray-Pedersen B, Stray-Pedersen S 1984 Etiologic factors and subsequent reproductive performance in 195 couples with a prior history of habitual abortion. American Journal of Obstetrics and Gynecology 148, Tulppala M, Stenman UH, Cacciatore B et al Polycystic ovaries and levels of gonadotrophins and androgens in recurrent miscarriage: prospective study in 50 women. British Journal of Obstetrics and Gynaecology 100, Watson H, Kiddy DS, Hamilton-Fairley D et al Hypersecretion of luteinizing hormone and ovarian steroids in women with recurrent early miscarriage. Human Reproduction 8, Yang CJ, Stone P, Stewart AW 2006 The epidemiology of recurrent miscarriage: a descriptive study of 1214 prepregnant women with recurrent miscarriage. Australian and New Zealand Journal of Obstetrics and Gynaecology 46, Declaration: The authors report no financial or commercial conflicts of interest. Received 4 December 2008; refereed 22 January 2009; accepted 1 June