Recurrent Miscarriage: Screening for Polycystic Ovaries and Subsequent Pregnancy Outcome

EDITORIAL COMMENT: One of the reviewer's comments and the letter of response from the authors are published below, with permission of the writers, since the cut and thrust will interest readers and add to their understanding of polycystic ovaries and subsequent pregnancy outcome. The editorial subcommittee often considers that it is a pity that readers never see a reviewer's report on a published paper and the author's response that may clinch the decision to publish. REVIEWER'S COMMENTS: This study by Liddell and colleagues compares the outcome amongst 2 groups of women who are habitual aborters. Their study group are women who have polycystic ovaries on ultrasound examination and their control group are habitual aborters who do not have this ultrasound appearance. My feeling is that they have slightly missed the point. The hypothesis that exists is that women with Polycystic Ovary Syndrome (PCOS) (i.e. with clinical manifestations due to excess androgen secretion) have poorer obstetric outcome. This is highlighted in their reference 4 from Regan et al (Hypersecretion of LH, infertility and miscarriage). I do not believe that it has been hypothesized that the appearance of polycystic ovaries on its own is what causes early pregnancy loss, but rather that it is the hypersecretion of LH which does so. One of the first reports of this phenomena came from Howells and colleagues (Lancet 1986; 3: 521–522) who reported on the high frequency of abortion in pregnancy after in vitro fertilization in women with elevated LH. The other early report by Stanger and Yovich (Br J Obstet Gynaecol 1985; 92: 385–393) also correlated poor fertilization rate in women with high LH levels. Why I believe that the authors have missed the boat here is that they have purely gone on ultrasonic ovarian appearance. As can be seen from table 3 the women did not appear to have PCOS as their hormonal profile is entirely normal without the PCOS associated elevation of LH and androgens. Therefore, one can conclude from their study that women who do not have hormonal disruptions of PCOS do not have an increased miscarriage rate. which does not really add much to the scientific literature. I believe that a similar study which compared women with true PCOS to those that did not have it would be interesting. AUTHOR'S RESPONSE: In answer to your reviewer's comments, we would disagree with the statement that the ‘hypothesis that exists is that women with Polycystic Ovary Syndrome, i.e. with clinical manifestations due to excess androgen secretion, have poor obstetric outcome.’ The original description of the association between polycystic ovaries and recurrent miscarriage was the paper published by Sagle et al in the BMJ in 1988, who found that 82% of women with recurrent miscarriages had polycystic ovaries diagnosed by an ultrasound of the pelvis. They also pointed out that most of these women had normal hormonal indices. The paper by Tulppala (reference 10) again found gonadotropin and androgen levels similar between women found to have polycystic ovaries on ultrasound to control women. This study, published in the Br J Obstet Gynaecol showed a 50% miscarriage rate in the group who had PCO on ultrasound. We feel that the association between polycystic ovaries and recurrent miscarriage has been overstated in recent years. It has been our hypothesis that the appearance on ultrasound of polycystic ovaries does not have great significance, and we believe that this is the first paper to have shown it. It also emphasizes that the majority of women with a polycystic ovary appearance on ultrasound had normal hormonal indices and we think that the paper shows very conclusively that it is not appropriate to overemphasize the diagnosis of polycystic ovaries in women with recurrent miscarriage as they have a very good outcome with purely supportive care. This is in keeping with recent studies that show that even with women who are shown to have higher levels of LH, complex endocrinological treatment regimens to reduce the LH do not improve pregnancy outcome. We think that there is still much confusion over how to define the group with hypersecretion of LH and there seems no agreed practicable method of clinically screening patients for hypersecretion ofLH. The St Mary's group has been looking at using daily urinary samples but is still modifying this method, therefore we think it is quite valid that we have used a method used by a number of other groups who have again shown a very high miscarriage rate with a single blood test in the follicular phase showing an elevation of LH. We were not able to confirm that we could define a group of women with recurrent miscarriage where this occurred or that they had a deleterious outcome. We have worked in the field of recurrent miscarriage for many years and firmly believe that new, much vaunted explanations for the causes of recurrent miscarriage, especially if they are accompanied by potentially hazardous and expensive treatment programmes designed to correct the so‐called abnormality, must be very carefully evaluated before they are adopted into clinical practice. We have learnt much from the use of immunotherapy in this group and its subsequent disproof of being a valid method of treatment. We think that it is vitally important to publish negative research, especially in the field of recurrent miscarriage, to emphasize over and over again how this group of patients will often have a good outcome with the minimum of treatment. The numbers in our study are comparable to other clinics who have published on the subject of polycystic ovaries and recurrent miscarriage, and it emphasizes the fact that even in the very large recurrent miscarriage clinic that we run in Auckland, we were not able to identify a group of women with elevated LH levels, polycystic ovaries, and a poor reproductive outcome and this emphasizes the low level of importance that this condition contributes to the causation of recurrent miscarriage. Summary: A population of women with a history of recurrent miscarriage were screened for polycystic ovaries (PCO) by an ultrasound, LH, FSH, free testosterone in the follicular phase, then luteal phase progesterone and body mass index (BMI). Twenty six of the 73 women screened (36%) had an ultrasound demonstrating PCO; of these 21 (81%) became pregnant and 17 were given supportive and observational care only. The miscarriage rate was 18% with 14 (82%) having livebirths. Twenty seven of the 47 women with normal ovaries (74%) became pregnant; 31 had supportive care only and 6 (19%) miscarried with 25 (81%) having a livebirth. We conclude that the ultrasound diagnosis of PCO in women with a history of recurrent miscarriage does not necessarily predict a poor outcome in subsequent pregnancy.