UvA-DARE (Digital Academic Repository)
Metformin in polycystic ovary syndrome
Moll, E.
Publication date
2013
Document Version
Final published version
Link to publication
Citation for published version (APA):
Moll, E. (2013). Metformin in polycystic ovary syndrome.
General rights
It is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), other than for strictly personal, individual use, unless the work is under an open content license (like Creative Commons).
Disclaimer/Complaints regulations
If you believe that digital publication of certain material infringes any of your rights or (privacy) interests, please let the Library know, stating your reasons. In case of a legitimate complaint, the Library will make the material inaccessible and/or remove it from the website. Please Ask the Library: https://uba.uva.nl/en/contact, or a letter
Etelka Moll
Metformin in
Polycystic Ovary Syndrome
© 2013 Etelka Moll, Amsterdam
Metformin in Polycystic Ovary Syndrome
PhD Thesis, University of Amsterdam – with references – with summary in Dutch
Keywords: polycystic ovary syndrome, metformin, ovulation induction, metabolic syndrome, subfertility
Cover: Sjoerd Repping & Etelka Moll
Printed by: Ridderprint
Publication of this thesis was supported by:
Metformin in Polycystic Ovary Syndrome
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
donderdag 25 april 2013, te 10.00 uur
door
Etelka Moll
Promotiecommissie
Promotores: prof.dr. F. van der Veen
prof.dr. P.M.M. Bossuyt
Co-promotores: dr. J.C. Korevaar
dr. M. van Wely
Overige leden: prof.dr. B.W.J. Mol
dr. M.J.M. Serlie
prof.dr. F.J.M. Broekmans
prof.dr. T. d’Hooghe
prof.dr. J.S.E. Laven
Faculteit Geneeskunde
This thesis was prepared at the Center for Reproductive Medicine, Department of Obstetrics and Gynecology, Academic Medical Center, University of Amsterdam, Amsterdam, the Netherlands.
Contents Page
Chapter 1
Introduction
7
Chapter 2
Effect of clomifene citrate plus metformin and clomifene citrate plus placebo on induction of ovulation in women with newly diagnosed polycystic ovary syndrome: randomised double blind clinical trial
Etelka Moll, Patrick M M Bossuyt, Johanna C Korevaar, Cornelis B Lambalk, Fulco van der Veen BMJ 2006; 332:1485-1489
19
Chapter 3
The role of metformin in polycystic ovary syndrome. A systematic review. Etelka Moll, Fulco van der Veen, Madelon van Wely
Human Reproduction Update 2007; 13:527-537
35
Chapter 4
Does adding metformin to clomifene citrate lead to higher pregnancy rates in a subset of women with polycystic ovary syndrome?
Etelka Moll, Joke C. Korevaar, Patrick M.M. Bossuyt, Fulco van der Veen Human Reproduction 2008; 23:1830-1834
71
Chapter 5
Does metformin treatment in women with polycystic ovary syndrome alter biomarkers associated with metabolic syndrome?
Etelka Moll, Madelon van Wely, Joost C.M. Meijers, Patrick M.M. Bossuyt, Cornelis B. Lambalk, Fulco van der Veen
Submitted
87
Chapter 6
Health-related quality of life in women with newly diagnosed polycystic ovary syndrome randomized between clomifene citrate plus metformin or clomifene citrate plus placebo.
Etelka Moll, Madelon van Wely, Cornelis B. Lambalk, Patrick M.M. Bossuyt, Fulco van der Veen Human Reproduction 2012; 27: 3273-3278
107
Chapter 7
Summary and implications for future research
125
Dankwoord 139
Portfolio 145
Chapter 1
Introduction
Introduction
In 1921 the first article of a severe case of - what we would now call- PCOS was published.1 The polycystic ovary syndrome (PCOS) affects 5% to 10% of women of reproductive age.2 PCOS is characterised by oligo-anovulation, clinical or biochemical hyperandrogenism and/or polycystic ovaries.3-5 In 1935 the first treatment, involving ovarian wedge resection in women with amenorrhea and enlarged polycystic ovaries, was published: “The mechanical theory: The overproduction of cystic follicles which crowd the ovarian cortex but which do not rupture on the surface of the ovary, together with the presence of a thickened tunic, prevents the immature follicles from ripening and reaching the surface. We observed that by removing the cystic cortex which formed the barrier, physiologic function was restored.” Since then the syndrome was named the Stein Leventhal syndrome, after the two authors, doctors Stein and Leventhal.6
Sixty years later, a connection was found between hyperandrogenism and hyperinsulinemia in PCOS patients,7 irrespective of weight.8 When these patients were treated with diazoxide, an insulin release inhibitor, insulin levels decreased and androgen parameters also decreased, whereas diazoxide in lean women without PCOS did not alter androgen parameters while insulin decreased.9-11 It was concluded that hyperandrogenism is the consequence of hyperinsulinemia and not vice versa. Insulin resistance (IR) accompanied by compensatory hyperinsulinemia constitutes a major biochemical feature of PCOS, which leads to early luteinizing hormone-sensitivity of the follicle and to stimulation of both ovarian and adrenal androgen production.12-16 Because of the link between IR and PCOS, metformin, an insulin sensitizer was put forward as a drug to improve the metabolic and endocrinological disturbances in women with PCOS.17
In 1994, more than 70 years after the first description of a patient with insulin resistance and hyperandrogenism, the first study on metformin in women with PCOS was published.18 Originally, this trial was meant to study metabolic and endocrinological parameters, but the authors noticed that some of the women (12%) conceived naturally. Following this initial study, many small trials were set up to test insulin sensitizers (mainly metformin) for ovulation induction in women with PCOS.19;20 It was at this time that we initiated the studies described in this thesis.
Results from a systemic review showed that metformin alone gave higher ovulation rates in women with PCOS, but not higher live birth rates than placebo.20 Metformin in combination with clomifene also resulted in higher ovulation rates and higher pregnancy rates, but not in higher live birth rates than clomifene alone. Critical analysis of the data presented in this review however, showed the following: Almost all included patients were obese. Obese women in general have less spontaneous and less clomifene induced ovulations.21;22 Therefore, the difference in ovulation and therefore also in live birth rate between placebo and clomifene treated women, may not be representative for the general PCOS population (obese and non-obese). Almost all patients receiving clomifene were previously diagnosed as clomifene resistant. This biases the results, because these patients will most likely benefit more from combination therapy than a group of patients that were treatment naive. Furthermore, the primary endpoint was ovulation and not live birth. Nevertheless, risk ratios for live birth rate were calculated using data not powered for this endpoint and the pregnancy results from the individual trials should therefore be interpreted with some caution.
Besides impaired reproductive function, PCOS can also result in several metabolic abnormalities including impaired glucose tolerance,7 dyslipidemia and hypertension.23 As a result, women with PCOS seem to be at increased risk of having metabolic syndrome24 and thereby at risk for cardiovascular disease and diabetes.25-27 According to the ESHRE/ASRM PCOS consensus workshop group a woman with PCOS has the metabolic syndrome when she suffers from at least three of the five following criteria: abdominal obesity (waist circumference), high triglycerides, low HDL, high blood pressure and insulin resistance.28 The prevalence of metabolic syndrome in women with PCOS varies largely among various populations and has been reported to range between 16% and 46%.29-34 Since insulin resistance is one of the criteria for the metabolic syndrome, metformin has been suggested to reduce risk for metabolic syndrome in women with PCOS.17;31;35 In studies where metformin was used to treat patients with diabetes, insulin values as well as lipid profile and coagulation and fibrinolytic factors were reduced.36;37 In
There are retrospective studies that claim to have data concerning cardiovascular disease in women with PCOS.38-41 None of these studies include patients with known or proven PCOS, let alone PCOS as defined following the Rotterdam criteria. Long term studies of women with well-defined PCOS are lacking. Therefore, it is not proven yet that women with PCOS suffer more from cardiovascular morbidity and mortality than women without PCOS.42 To improve the metabolic syndrome parameters seems worthwhile, but it is not clear if it will improve morbidity and mortality in the long term.
As PCOS is a heterogeneous condition,43 several authors have suggested that metformin would be most beneficial in specific subgroups of women with this condition. Increasing age and high waist hip ratio (WHR) are known risk factors in developing IR and, as such, these factors may also affect clinical response to metformin.44-47 One study showed that women with PCOS and with a high waist hip ratio (WHR) more often have insulin resistance and diabetes mellitus.45 In another study women with PCOS and a high WHR showed less insulin resistance and more weight loss as a reaction to metformin.47 At the time of our research it was not clear which specific subgroups of women with PCOS would benefit most from treatment with metformin.
In deciding to treat a patient with metformin or not, the burden of the treatment is important as well. Information on health related quality of life (HRQoL) of women with PCOS treated by different modalities for ovulation induction is limited. One study randomised between life style adjustments combined with metformin or placebo.48 Women were advised to use barrier contraception and to avoid pregnancy. This study had a significant amount of drop-outs (60 vs 73%). No difference was found in HRQoL between the two groups during the study. A second study randomised between oral contraceptives (OC) combined with metformin or placebo in adolescents (12-18years).49 No difference in HRQoL was found between the two groups. In women with fertility problems lower psychological stress scores are found once an ongoing pregnancy is reached,50 and differences between treatment groups disappear. Women will choose the more effective treatment, even if it will give more discomfort.51
Background of the research of this thesis
As clomifene is an effective treatment for ovulation induction in PCOS,22 the question arises what the added benefits of metformin would be. At the start of the studies described in this thesis, there was no randomised controlled trial comparing metformin combined with clomifene to placebo combined with clomifene for ovulation induction in women with PCOS.
At the time we started our studies the evidence on a beneficial effect of metformin in specific subgroups of patients was scarce. There was only one small study involving 32 patients with PCOS that had investigated the effect of metformin in specific subgroups. This study showed in a multivariable analysis that higher insulin, lower androstenedione and less severe cycle abnormalities appeared to be independent significant parameters for better response to metformin.52 We wanted to investigate if a treatment difference existed in the several subgroups of our population.
Data on prevalence and incidence of type 2 diabetes mellitus and cardiovascular disease in women with PCOS are poor. Studies that demonstrate a direct effect of life style adjustments or metformin on the incidence of these diseases are lacking. Thus, the clinical value of treatment of the surrogate endpoints –the metabolic syndrome- is still unknown. Nevertheless, some authors suggest that treatment of the metabolic syndrome in women with PCOS is beneficial for preventing disease later in life.17;38 We wanted to investigate whether metformin has a positive effect on several parameters of the metabolic syndrome.
Although it is generally assumed that ovulation induction with metformin is more burdensome than ovulation induction with clomifene due to the higher amount of tablets which ought to be ingested (three or four tablets a day until pregnancy is achieved in contrast to 5 to 15 tablets a month) and the high incidence of side-effects, there are no data in adult women who receive ovulation induction with the purpose of conceiving
Outline of this thesis
In Chapter 2 we describe a randomized trial designed to evaluate whether a strategy of adding metformin to the standard treatment with clomifene, results in a higher ovulation rate, higher pregnancy rate and less clomifene resistance. The study was a randomised clinical trial among newly diagnosed women with PCOS.
Chapter 3 offers an introduction to the whole spectrum of subfertility treatment in
PCOS with metformin and gives a critical appraisal of all existing studies, which will be helpful to guide clinical practice.
Chapter 4 reports whether co-treatment with metformin improves pregnancy rates
compared to the standard treatment of clomifene alone in subgroups of women based on clinical and biochemical variables. For this purpose we reanalysed the data from the randomised trial described in chapter 2.
Chapter 5 outlines whether metformin has a positive effect on a complete set of
biomarkers that is associated with the metabolic syndrome.
Chapter 6 examines the HRQoL of women treated with clomifene combined with
References
(1) Achard C, Thiers J. Le virilisme pilaire et son association a l'insuffisance glycolytique (diabete des femmes a barbe). Bull Acad Natl Med 1921; 86(29):51-66.
(2) Asuncion M, Calvo RM, San Millan JL, Sancho J, Avila S, Escobar-Morreale HF. A prospective study of the prevalence of the polycystic ovary syndrome in unselected Caucasian women from Spain. J Clin Endocrinol Metab 2000; 85(7):2434-2438.
(3) ESHRE/ASRM. Revised 2003 consensus on diagnostic criteria and long-term health risks related to polycystic ovary syndrome (PCOS). Hum Reprod 2004; 19(1):41-47.
(4) Franks S. Polycystic ovary syndrome. N Engl J Med 1995; 333(13):853-861.
(5) Knochenhauer ES, Key TJ, Kahsar-Miller M, Waggoner W, Boots LR, Azziz R. Prevalence of the polycystic ovary syndrome in unselected black and white women of the southeastern United States: a prospective study. J Clin
Endocrinol Metab 1998; 83(9):3078-3082.
(6) Stein IF, Leventhal ML. Amenorrhoea associated with bilateral polycystic ovaries. Am J Obstet Gynecol 1935; 29:181.
(7) Burghen GA, Givens JR, Kitabchi AE. Correlation of hyperandrogenism with hyperinsulinism in polycystic ovarian disease. J Clin Endocrinol Metab 1980; 50(1):113-116.
(8) Chang RJ, Nakamura RM, Judd HL, Kaplan SA. Insulin resistance in nonobese patients with polycystic ovarian disease. non-obese 1983; 57(2):356-359.
(10) Geffner ME, Kaplan SA, Bersch N, Golde DW, Landaw EM, Chang RJ. Persistence of insulin resistance in polycystic ovarian disease after inhibition of ovarian steroid secretion. Fertil Steril 1986; 45(3):327-333.
(11) Nestler JE, Barlascini CO, Matt DW, Steingold KA, Plymate SR, Clore JN et al. Suppression of serum insulin by diazoxide reduces serum testosterone levels in obese women with polycystic ovary syndrome. J Clin Endocrinol
Metab 1989; 68(6):1027-1032.
(12) Barbieri RL. Hyperandrogenism, insulin resistance and acanthosis nigricans. 10 years of progress. J Reprod Med 1994; 39(5):327-336.
(13) Dunaif A, Segal KR, Futterweit W, Dobrjansky A. Profound peripheral insulin resistance, independent of obesity, in polycystic ovary syndrome. Diabetes 1989; 38(9):1165-1174.
(14) Dunaif A. Insulin resistance and the polycystic ovary syndrome: mechanism and implications for pathogenesis. Endocr Rev 1997; 18(6):774-800.
(15) Nestler JE. Insulin regulation of human ovarian androgens. Hum Reprod 1997; 12 Suppl 1:53-62.
(16) O'Meara NM, Blackman JD, Ehrmann DA, Barnes RB, Jaspan JB, Rosenfield RL et al. Defects in beta-cell function in functional ovarian hyperandrogenism.
J Clin Endocrinol Metab 1993; 76(5):1241-1247.
(17) Nestler JE. Should patients with polycystic ovarian syndrome be treated with metformin?: an enthusiastic endorsement. Hum Reprod 2002; 17(8):1950-1953.
(18) Velazquez EM, Mendoza S, Hamer T, Sosa F, Glueck CJ. Metformin therapy in polycystic ovary syndrome reduces hyperinsulinemia, insulin resistance, hyperandrogenemia, and systolic blood pressure, while facilitating normal menses and pregnancy. Metabolism 1994; 43(5):647-654.
(19) Lord JM, Flight IH, Norman RJ. Metformin in polycystic ovary syndrome: systematic review and meta-analysis. BMJ 2003; 327(7421):951-953.
(20) Lord JM, Flight IH, Norman RJ. Insulin-sensitising drugs (metformin, troglitazone, rosiglitazone, pioglitazone, D-chiro-inositol) for polycystic ovary syndrome. Cochrane Database Syst Rev 2003;(3):CD003053.
(21) Norman RJ, Noakes M, Wu R, Davies MJ, Moran L, Wang JX. Improving reproductive performance in overweight/obese women with effective weight management. Hum Reprod Update 2004; 10(3):267-280.
(22) Imani B, Eijkemans MJ, te Velde ER, Habbema JD, Fauser BC. A nomogram to predict the probability of live birth after clomiphene citrate induction of ovulation in normogonadotropic oligoamenorrheic infertility. Fertil Steril 2002; 77(1):91-97.
(23) Ehrmann DA. Insulin resistance and polycystic ovary syndrome. Curr Diab
Rep 2002; 2(1):71-76.
(24) Reaven GM. Banting lecture 1988. Role of insulin resistance in human disease. Diabetes 1988; 37(12):1595-1607.
(25) Coviello AD, Legro RS, Dunaif A. Adolescent girls with polycystic ovary syndrome have an increased risk of the metabolic syndrome associated with increasing androgen levels independent of obesity and insulin resistance. J
Clin Endocrinol Metab 2006; 91(2):492-497.
(26) Legro RS, Kunselman AR, Dunaif A. Prevalence and predictors of dyslipidemia in women with polycystic ovary syndrome. Am J Med 2001; 111(8):607-613.
(27) Moran L, Teede H. Metabolic features of the reproductive phenotypes of polycystic ovary syndrome. Hum Reprod Update 2009; 15(4):477-488.
(28) ESHRE/ASRM. Revised 2003 consensus on diagnostic criteria and long-term health risks related to polycystic ovary syndrome (PCOS). Hum Reprod 2004;
(29) Apridonidze T, Essah PA, Iuorno MJ, Nestler JE. Prevalence and characteristics of the metabolic syndrome in women with polycystic ovary syndrome. J Clin Endocrinol Metab 2005; 90(4):1929-1935.
(30) Carmina E, Napoli N, Longo RA, Rini GB, Lobo RA. Metabolic syndrome in polycystic ovary syndrome (PCOS): lower prevalence in southern Italy than in the USA and the influence of criteria for the diagnosis of PCOS. Eur J
Endocrinol 2006; 154(1):141-145.
(31) Glueck CJ, Papanna R, Wang P, Goldenberg N, Sieve-Smith L. Incidence and treatment of metabolic syndrome in newly referred women with confirmed polycystic ovarian syndrome. Metabolism 2003; 52(7):908-915.
(32) Hahn S, Tan S, Sack S, Kimmig R, Quadbeck B, Mann K et al. Prevalence of the metabolic syndrome in German women with polycystic ovary syndrome.
Exp Clin Endocrinol Diabetes 2007; 115(2):130-135.
(33) Soares EM, Azevedo GD, Gadelha RG, Lemos TM, Maranhao TM.
Prevalence of the metabolic syndrome and its components in Brazilian women with polycystic ovary syndrome. Fertil Steril 2008; 89(3):649-655.
(34) Weerakiet S, Bunnag P, Phakdeekitcharoen B, Wansumrith S,
Chanprasertyothin S, Jultanmas R et al. Prevalence of the metabolic syndrome in Asian women with polycystic ovary syndrome: using the International Diabetes Federation criteria. Gynecol Endocrinol 2007; 23(3):153-160.
(35) Use of insulin sensitizing agents in the treatment of polycystic ovary syndrome. Fertil Steril 2004; 82 Suppl 1:S181-S183.
(36) Effect of intensive blood-glucose control with metformin on complications in overweight patients with type 2 diabetes (UKPDS 34). UK Prospective Diabetes Study (UKPDS) Group. Lancet 1998; 352(9131):854-865.
(37) Inzucchi SE. Oral antihyperglycemic therapy for type 2 diabetes: scientific review. JAMA 2002; 287(3):360-372.
(38) Wild S, Pierpoint T, Jacobs H, McKeigue P. Long-term consequences of polycystic ovary syndrome: results of a 31 year follow-up study. Hum Fertil
(Camb ) 2000; 3(2):101-105.
(39) Wild S, Pierpoint T, McKeigue P, Jacobs H. Cardiovascular disease in women with polycystic ovary syndrome at long-term follow-up: a retrospective cohort study. Clin Endocrinol (Oxf) 2000; 52(5):595-600.
(40) Pierpoint T, McKeigue PM, Isaacs AJ, Wild SH, Jacobs HS. Mortality of women with polycystic ovary syndrome at long-term follow-up. J Clin
Epidemiol 1998; 51(7):581-586.
(41) Cibula D, Cifkova R, Fanta M, Poledne R, Zivny J, Skibova J. Increased risk of non-insulin dependent diabetes mellitus, arterial hypertension and coronary artery disease in perimenopausal women with a history of the polycystic ovary syndrome. Hum Reprod 2000; 15(4):785-789.
(42) Legro RS. Polycystic ovary syndrome and cardiovascular disease: a premature association? Endocr Rev 2003; 24(3):302-312.
(43) Broekmans FJ, Fauser BC. Diagnostic criteria for polycystic ovarian syndrome. Endocrine 2006; 30(1):3-11.
(44) Despres JP. Abdominal obesity as important component of insulin-resistance syndrome. Nutrition 1993; 9(5):452-459.
(45) Legro RS, Kunselman AR, Dodson WC, Dunaif A. Prevalence and predictors of risk for type 2 diabetes mellitus and impaired glucose tolerance in polycystic ovary syndrome: a prospective, controlled study in 254 affected women. J Clin Endocrinol Metab 1999; 84(1):165-169.
(46) NCEP. Executive Summary of The Third Report of The National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, And
(47) Pasquali R, Gambineri A, Biscotti D, Vicennati V, Gagliardi L, Colitta D et al. Effect of long-term treatment with metformin added to hypocaloric diet on body composition, fat distribution, and androgen and insulin levels in abdominally obese women with and without the polycystic ovary syndrome. J
Clin Endocrinol Metab 2000; 85(8):2767-2774.
(48) Ladson G, Dodson WC, Sweet SD, Archibong AE, Kunselman AR, Demers LM et al. The effects of metformin with lifestyle therapy in polycystic ovary syndrome: a randomized double-blind study. Fertil Steril 2011; 95(3):1059-1066.
(49) Harris-Glocker M, Davidson K, Kochman L, Guzick D, Hoeger K.
Improvement in quality-of-life questionnaire measures in obese adolescent females with polycystic ovary syndrome treated with lifestyle changes and oral contraceptives, with or without metformin. Fertil Steril 2010; 93(3):1016-1019.
(50) van Wely M, Bayram N, Bossuyt PM, van der Veen F. Laparoscopic electrocautery of the ovaries versus recombinant FSH in clomiphene citrate-resistant polycystic ovary syndrome. Impact on women's health-related quality of life. Hum Reprod 2004; 19(10):2244-2250.
(51) Bayram N, van Wely M, van der Veen F, Bossuyt PM, Nieuwkerk P. Treatment preferences and trade-offs for ovulation induction in clomiphene citrate-resistant patients with polycystic ovary syndrome. Fertil Steril 2005; 84(2):420-425.
(52) Moghetti P, Castello R, Negri C, Tosi F, Perrone F, Caputo M et al. Metformin effects on clinical features, endocrine and metabolic profiles, and insulin sensitivity in polycystic ovary syndrome: a randomized, double-blind, placebo-controlled 6-month trial, followed by open, long-term clinical evaluation. J Clin
Chapter 2
Effect of clomifene citrate plus metformin and clomifene citrate plus placebo on induction of ovulation in women with newly diagnosed polycystic ovary
syndrome: randomised double blind clinical trial
Etelka Moll, Patrick M M Bossuyt, Johanna C Korevaar, Cornelis B Lambalk, Fulco van der Veen
Abstract
Objective To compare the effectiveness of clomifene citrate plus metformin and
clomifene citrate plus placebo in women with newly diagnosed polycystic ovary syndrome.
Design Randomised clinical trial.
Setting Multicentre trial in 20 Dutch hospitals.
Participants 228 women with polycystic ovary syndrome.
Interventions Clomifene citrate plus metformin or clomifene citrate plus placebo.
Main outcome measure The primary outcome measure was ovulation. Secondary
outcome measures were ongoing pregnancy, spontaneous abortion, and clomifene resistance.
Results 111 women were allocated to clomifene citrate plus metformin (metformin
group) and 114 women were allocated to clomifene citrate plus placebo (placebo group). The ovulation rate in the metformin group was 64% compared with 72% in the placebo group, a non-significant difference (risk difference −8%, 95% confidence interval −20% to 4%). There were no significant differences in either rate of ongoing pregnancy (40% v 46%; −6%, −20% to 7%) or rate of spontaneous abortion (12% v 11%; 1%, −7% to 10%). A significantly larger proportion of women in the metformin group discontinued treatment because of side effects (16% v 5%; 11%, 5% to 16%).
Conclusion Metformin is not an effective addition to clomifene citrate as the primary
Introduction
Polycystic ovary syndrome is characterised by any of oligoanovulation, clinical or biochemical hyperandrogenism, and polycystic ovaries.1;2. The syndrome affects approximately 4%-9% of women of reproductive age.3 Women most commonly seek counselling or treatment because of infertility due to chronic anovulation. Insulin resistance accompanied by compensatory hyperinsulinaemia constitutes another major biochemical feature of polycystic ovary syndrome, which leads to early luteinising hormone sensitivity of the follicle and to stimulation of both ovarian and adrenal androgen production.4-8
The first choice drug in women with newly diagnosed polycystic ovary syndrome is the antioestrogen clomifene citrate.9 Clomifene citrate enhances release of pituitary gonadotrophins, resulting in follicular recruitment. Three quarters of women with polycystic ovary syndrome will ovulate with clomifene citrate.10 Complications of treatment are rare and usually mild. Patients who do not ovulate on the maximum dose of 150 mg are considered to be clomifene citrate resistant.
Recently, the addition of metformin, an insulin sensitiser, to clomifene citrate has been proposed as an alternative treatment option for women with polycystic ovary syndrome. Insulin sensitisers improve hyperinsulinaemia and hyperandrogenism in these women.11;12 One study also showed that metformin regulated menstrual cycles and pregnancies.12;13
Metformin is now the most widely used insulin sensitiser for induction of ovulation in women with polycystic ovary syndrome and may improve ovulation rates when combined with clomifene citrate in clomifene citrate resistant women.14 This has led to the recommendation to use metformin alone or in combination with clomifene citrate as first line treatment in infertile women with polycystic ovary syndrome.
As clomifene citrate is an effective treatment,10 however, what are the added benefits of metformin? Two randomised controlled trials have examined this question. One small study found a significant increase in ovulation rates,14 while a larger study failed to find a significant difference.15 Sample sizes of both studies were small, performance was not double blinded, and previous treatment of the
less clomifene citrate resistance in a randomised clinical trial among women with newly diagnosed polycystic ovary syndrome.
Methods
Participants and experimental protocol
From June 2001 till May 2004 we invited relevant women from 20 Dutch hospitals to participate in the trial. All women had chronic anovulation (menstrual cycle 35 days, WHO type II, normogonadotropic, normo-oestrogenic, oligoanovulation or anovulation) and polycystic ovaries diagnosed by transvaginal ultrasonography and wanted to conceive.
We defined polycystic ovary syndrome according to current guidelines1. We excluded women with other causes of anovulation, age >40 years, and liver, kidney, or heart disease or failure (that is, abnormal results on liver function tests or serum creatinine concentration >95 mol/l or a history of heart disease or failure) and also those whose partner’s sperm quality indicated male factor subfertility (total motile count < 10x106). Tubal patency was not tested before induction of ovulation.
Women who gave informed consent were randomly allocated to clomifene citrate plus metformin (metformin group) or clomifene citrate plus placebo (placebo group). Randomisation was done in the coordinating centre (AMC, Amsterdam) by using computer generated blocks of four. Merck Santé, France, prepared the containers with the study medication. They determined the final allocation sequence and kept this list until inclusion was finished. The randomisation was stratified per centre, and the centres received blinded, numbered containers with medication. Each participant received the container with the next number in her own hospital.
We tested liver and kidney function before women started taking the study medication. The dose was increased from one to four tablets a day (that is, up to 2000 mg) over a period of seven days16. We used this “step up” regimen to limit side effects. Patients continued to take the study medication until they had a positive pregnancy test or six ovulatory cycles or developed clomifene citrate resistance, whichever came first.
Women took metformin or placebo for one month, to allow metformin enough time to have a sufficient insulin sensitising effect 14. If no spontaneous menstruation occurred and the pregnancy test was negative one month after the study medication
was started, we induced menstruation with dydrogesterone 10 mg three times a day for ten days. From the third or fifth day until the seventh or ninth day after (spontaneous or induced) menstruation, women took 50 mg clomifene citrate a day. If ovulation did not occur with this dose, it was increased with steps of 50 mg to a maximum of 150 mg a day in the next cycles. Ovulation was detected either with a biphasic basal temperature curve, a follicle with a diameter 16 mm on transvaginal ultrasonography, or progesterone 14 nmol/l in the second half of a menstrual cycle, or pregnancy. If a woman ovulated, she continued taking the same dose of clomifene citrate until an end point was reached.
If women had an ovulatory cycle with a certain dose of clomifene citrate followed by a cycle without ovulation at the same dose, they were classified as temporarily ovulatory. These women entered the next cycle with a higher dose of clomifene citrate.
Statistical analysis
The primary outcome measure was ovulation. Secondary outcome measures were ongoing pregnancy, spontaneous abortion, and clomifene resistance. We compared the cumulative rates of ovulation as well as other rates and proportions between groups using relative rates, rates differences, and χ2 test statistics using SPSS 11.5.1.
With an expected rate of ovulation of 75% in the placebo group,10;17 we needed 200 women to show an absolute increase of 15% in ovulation rate, with a power of at least 80% using a two sided χ2 test with a 5% significance level.
Results
We screened 228 women for eligibility. Three had to be excluded: two had type 2 diabetes and one had raised liver enzyme activity. Baseline characteristics were similar in the two groups apart from the mean total motile sperm count in partners (table 1).
In the 114 women allocated to the placebo group, four became pregnant and nine dropped out before they started taking the study medication. Ninety two women received 50 mg clomifene citrate, of whom 54 went on to take 100 mg clomifene citrate, and 23 eventually took 150 mg. Thirteen women developed resistance. In both groups none of the cycles was cancelled and no one developed ovarian hyperstimulation syndrome.
Cumulative rates of ovulation were slightly lower in the metformin group, but this difference was not significant, nor were differences in cumulative pregnancy and spontaneous abortion rates (table 2). There were no significant differences between the two groups when we analysed the data by clomifene citrate dose (table 3).
In the metformin group, 28 women stopped treatment before reaching an end point. Eighteen stopped because of side effects, 10 because of other or unspecified reasons.
In the placebo group, 21 women stopped treatment before reaching an end point. Six stopped because of side effects, 10 because of other or unspecified reasons, and five had a concomitant disease that prevented continuation of the study.
There was a significant difference between the metformin and placebo group in the discontinuation because of side effects (16% v 5%; risk difference 11%, 95% confidence interval 5% to 16%).
As 63 women discontinued study medication, 78 and 84 remained for the per protocol analyses. The results in those who did not withdraw showed no benefit with metformin (table 4).
By the end of follow-up there were 21 live births without complications in the metformin group, 13 spontaneous abortions and 14 ongoing singleton pregnancies. There were three premature deliveries (gestational age 36+0, 33+1, and 17+5 weeks). The last pregnancy concerned triplets; none of whom survived. One woman developed gestational diabetes, but she delivered without complications. Four patients developed hypertension; three delivered without complications, one had a child with Kartagener’s syndrome and hypospadias. One woman developed pre-eclampsia and delivered prematurely (32+6 weeks). The child had anal atresia. In the placebo group there were 30 live births without complications, 1 live twin birth without complications, 12 spontaneous abortions, 9 ongoing singleton pregnancies, and two ongoing twin pregnancies. There was one case of intra uterine growth retardation and two premature deliveries (34+1 and 18+0 weeks). The last delivery
concerned an anencephalic child that did not survive. Two women developed gestational diabetes and two developed hypertension, all four delivered without further complications. Three patients developed preeclampsia; two delivered without complications, one delivered prematurely (36+0 weeks).
There were no significant differences in complications of pregnancy, gestational age, complications of delivery, birth weights, or congenital malformations between the two groups. The power for these comparisons, however, was low.
Discussion
In the treatment of women with polycystic ovary syndrome who want to get pregnant we could not find any benefit of adding metformin to the standard treatment with clomifene citrate. We found no significant differences in outcome and can exclude any substantial improvement in rates of ovulation and ongoing pregnancy. Significantly more women in the metformin group discontinued treatment because of side effects.
Patients—We included women with polycystic ovary syndrome, defined according
to the 2003 consensus,1 who had never used clomifene citrate before and were seeking treatment for their fertility problems for the first time. We evaluated eligibility irrespective of their body mass index. By using these criteria our study group reflects the largest group of women with polycystic ovary syndrome a fertility clinic will see and treat.
Ovulation— Metformin monotherapy induces ovulation through its insulin
sensitizing effect. In our study, however, we failed to find an increase in ovulation rate of combined therapy compared with clomifene citrate alone. The effects of metformin on ovulation may not be sufficiently strong to improve on the already high ovulation rates with clomifene citrate in these women. This theory is strengthened by the fact that insulin resistance did not improve substantially. It is possible that those women who ovulate on metformin monotherapy would also ovulate on clomifene citrate monotherapy, explaining the absence of an added effect. Our participants were less obese than women in previous studies. Our group represents a normal
Dropouts—The number of dropouts in the metformin group was relatively high in
the first month of use. Previous studies have shown that it is not uncommon for patients to experience side effects but that these wear off after a certain time.18 There were no differences in baseline characteristics between the women who dropped out and those who completed the study (data not shown). Previous studies have not reported any dropouts.14;15
Comparison—Our findings confirm those of El-Biely and Habba15 but conflicts with those of Nestler et al.14 The mean body mass index was lower in our participants (28
v 32) and more were hyperandrogenic (free androgen index 12 v 3) than in the
Nestler study. Our participants were more representative of women with polycystic ovary syndrome in Europe.
Schedule—Previous studies have evaluated one menstrual cycle only14 or included patients for six months,15 whereas our study evaluated six ovulations or lasted until clomifene citrate resistance developed. We think this is a more realistic strategy from a clinical point of view. The downside of this strategy may have been the unexpectedly high rate of women dropping out. As we did not account for this in the power calculation, the power to detect smaller differences has been attenuated. Nevertheless, ovulation rates were lower in the metformin group, and the confidence intervals exclude substantial gains in ovulation and pregnancy rates from adding metformin.
Spontaneous abortion—Some authors have reported that metformin can decrease
the rate of spontaneous abortion in women with polycystic ovary syndrome.19-21 In our study the rate was similar in both groups and comparable with the rate in the general population.22-24 There was no evidence of any protective effect against spontaneous abortion in the metformin group, though this study was not powered for this question. Women had to discontinue their medication as soon as they had a positive pregnancy test as the safety and benefit of using metformin during pregnancy have not yet been proved. This might partially explain why the proportion of miscarriages was equal between the groups.
Conclusion
Based on the results of this trial, we cannot exclude the possibility that addition of metformin may lead to an increase in the ovulation rate of up to 5%, though whether such a small difference is clinically relevant is doubtful. Though metformin seems to be a relatively safe medication, it is associated with a high incidence of side effects.18
We conclude that metformin should not be added to clomifene citrate as primary method for induction of ovulation in women with polycystic ovary syndrome.
Acknowledgements
We thank the following participants for including and treating patients in this study: R E Bernardus (Ziekenhuis Gooi Noord, Blaricum); H E Bobeck (Rode Kruis Ziekenhuis, Beverwijk); D D M Braat, W N P Willemsen (Universitair Medisch Centrum, Sint Radboud, Nijmegen); J W Briët (Deventer Ziekenhuis, Deventer); J Dawson, I van der Laar (Sint Lucas Ziekenhuis, Amsterdam); H J H M van Dessel, T J G Griffioen (Twee Steden Ziekenhuis, Tilburg); J P R Doornbos (Zaans Medisch Centrum, De Heel, Zaandam); J J Duvekot, H Versendaal (Medisch Centrum Rijnmond Zuid, Rotterdam); M H Emanuel (Spaarne Ziekenhuis, Haarlem); P A Flierman, H Verhoeve (Onze Lieve Vrouwe Gasthuis, Amsterdam); D A Gietelink, A M Bongers (Amphia Ziekenhuis, Breda); M H A van Hooff (Sint Franciscus Gasthuis, Rotterdam); P W H Houben (Gemini Ziekenhuis, Den Helder); Y M van Kasteren, R Weselius (Medisch Centrum Alkmaar, Alkmaar); C B Lambalk, J J M L Dekker (Vrije Universiteit Medisch Centrum, Amsterdam); S van der Meer, C van Bolhuis (Medisch Centrum Haaglanden, Den Haag); A W J Omtzigt, P van Berlo (Flevoziekenhuis, Almere); CN M Renckens (Westfries Gasthuis, Hoorn); E Scheenjes (Ziekenhuis Gelderse Vallei, Ede); M A H M Wiegerinck, P Maas-van Son, B W Mol (Maxima Medisch Centrum, Veldhoven). We thank Nathalie Cugnardey, Merck Santé, France, and Louis LM van de Ven, Merck, Netherlands.
References
(1) ESHRE/ASRM. Revised 2003 consensus on diagnostic criteria and long-term health risks related to polycystic ovary syndrome (PCOS). Hum Reprod 2004; 19(1):41-47.
(2) Franks S. Polycystic ovary syndrome. N Engl J Med 1995; 333(13):853-861.
(3) Homburg R. What is polycystic ovarian syndrome? A proposal for a consensus on the definition and diagnosis of polycystic ovarian syndrome.
Hum Reprod 2002; 17(10):2495-2499.
(4) Barbieri RL. Hyperandrogenism, insulin resistance and acanthosis nigricans. 10 years of progress. J Reprod Med 1994; 39(5):327-336.
(5) Dunaif A, Segal KR, Futterweit W, Dobrjansky A. Profound peripheral insulin resistance, independent of obesity, in polycystic ovary syndrome. Diabetes 1989; 38(9):1165-1174.
(6) Dunaif A. Insulin resistance and the polycystic ovary syndrome: mechanism and implications for pathogenesis. Endocr Rev 1997; 18(6):774-800.
(7) Nestler JE. Insulin regulation of human ovarian androgens. Hum Reprod 1997; 12 Suppl 1:53-62.
(8) O'Meara NM, Blackman JD, Ehrmann DA, Barnes RB, Jaspan JB, Rosenfield RL et al. Defects in beta-cell function in functional ovarian hyperandrogenism.
J Clin Endocrinol Metab 1993; 76(5):1241-1247.
(9) ASRM. Use of clomiphene citrate in women. Fertil Steril 2003; 80(5):1302-1308.
(10) Imani B, Eijkemans MJ, te Velde ER, Habbema JD, Fauser BC. A nomogram to predict the probability of live birth after clomiphene citrate induction of ovulation in normogonadotropic oligoamenorrheic infertility. Fertil Steril 2002; 77(1):91-97.
(11) Nestler JE, Barlascini CO, Matt DW, Steingold KA, Plymate SR, Clore JN et al. Suppression of serum insulin by diazoxide reduces serum testosterone levels in obese women with polycystic ovary syndrome. J Clin Endocrinol Metab 1989; 68(6):1027-1032.
(12) Velazquez EM, Mendoza S, Hamer T, Sosa F, Glueck CJ. Metformin therapy in polycystic ovary syndrome reduces hyperinsulinemia, insulin resistance, hyperandrogenemia, and systolic blood pressure, while facilitating normal menses and pregnancy. Metabolism 1994; 43(5):647-654.
(13) Nestler JE, Jakubowicz DJ. Decreases in ovarian cytochrome P450c17 alpha activity and serum free testosterone after reduction of insulin secretion in polycystic ovary syndrome. N Engl J Med 1996; 335(9):617-623.
(14) Nestler JE, Jakubowicz DJ, Evans WS, Pasquali R. Effects of metformin on spontaneous and clomiphene-induced ovulation in the polycystic ovary syndrome. N Engl J Med 1998; 338(26):1876-1880.
(15) El-Biely MM., Habba M. The use of metformin to augment the induction of ovulation in obese infertile patients with polycystic ovary syndrome. Middle East Fertility SocietyJournal 6, 43-9. 2001.
(16) Garber AJ, Duncan TG, Goodman AM, Mills DJ, Rohlf JL. Efficacy of metformin in type II diabetes: results of a double-blind, placebo-controlled, dose-response trial. Am J Med 1997; 103(6):491-497.
(17) Eijkemans MJ, Imani B, Mulders AG, Habbema JD, Fauser BC. High singleton live birth rate following classical ovulation induction in normogonadotrophic anovulatory infertility (WHO 2). Hum Reprod 2003; 18(11):2357-2362.
(18) Lord JM, Flight IH, Norman RJ. Insulin-sensitising drugs (metformin, troglitazone, rosiglitazone, pioglitazone, D-chiro-inositol) for polycystic ovary syndrome. Cochrane Database Syst Rev 2003;(3):CD003053.
(20) Jakubowicz DJ, Iuorno MJ, Jakubowicz S, Roberts KA, Nestler JE. Effects of metformin on early pregnancy loss in the polycystic ovary syndrome. J Clin
Endocrinol Metab 2002; 87(2):524-529.
(21) Glueck CJ, Phillips H, Cameron D, Sieve-Smith L, Wang P. Continuing metformin throughout pregnancy in women with polycystic ovary syndrome appears to safely reduce first-trimester spontaneous abortion: a pilot study.
Fertil Steril 2001; 75(1):46-52.
(22) Alberman E. Spontaneous abortion: epidemiology. In: Spontaneous abortion: diagnosis and treatment., Stabile S, Grudzinkas G, Chard T, editors. London: Springer-Verlag; 1992. 9-20.
(23) Regan L, Braude PR, Trembath PL. Influence of past reproductive performance on risk of spontaneous abortion. BMJ 1989; 299(6698):541-545.
(24) Wilcox AJ, Weinberg CR, O'Connor JF, Baird DD, Schlatterer JP, Canfield RE et al. Incidence of early loss of pregnancy. N Engl J Med 1988; 319(4):189-194.
Figure 1. Overview of the clinical trial. Women shown as receiving 50 or 100mg
clomifene citrate received one to six cycles at the noted dose. Ovulatory=women with six ovulatory cycles in total, without pregnancy; temporarily ovulatory=women who ovulated on a certain dose of clomifene citrate at one point in the study and were anovulatory on the same dose at another point in the study. They were treated with a higher dose in the next cycle.
228 eligible patients
2 diabetes mellitus type II 1 elevated liver enzymes
7 pregnant
13 discontinued 100 received metformin 101 received placebo
7 pregnant 2 discontinued
6 ovulatory 27 anovulatory
17 temporarily ovulatory 2 ovulatory 41 anovulatory
13 temporarily ovulatory
2 ovulatory 12 anovulatory
5 temporarily ovulatory 3 ovulatory 15 anovulatory
8 temporarily ovulatory 6 pregnant
5 drop-out 111 allocated to metformin 114 allocated to placebo
4 pregnant 9 drop-out 80 received CC 50mg 92 received CC 50mg 24 pregnant 6 discontinued 29 pregnant 7 discontinued 44 received CC 100mg 54 received CC 100mg 16 pregnant 9 discontinued 18 pregnant 10 discontinued 17 received CC 150mg 23 received CC 150mg 4 pregnant 0 discontinued 6 pregnant 2 discontinued
Table 1. Baseline characteristics of women in study according to allocation to
clomifene citrate plus metformin or clomifene citrate plus placebo. Figures are numbers (percentages) of women unless stated otherwise.
Characteristics clomifene citrate and metformin n=111 clomifene citrate and placebo n=114 Mean (SD) age (years) 27.9 (3.7) 28.4 (4.7)
Parity:
Nulliparous 89 (80%) 95 (84%)
Multiparous 22 (20%) 18 (16%)
Mean (SD) duration of trying to conceive (years)
1.6 (1.2) 1.3 (1.1)
Mean(SD) BMI (kg/m2) 28.5 (7.1) 27.8 (6.7)
BMI 25 59/104 (57%) 60/109 (55%)
Mean (SD) waist:hip ratio 0.82 (0.1) 0.83 (0.1)
Mean (SD) LH:FSH ratio 1.92 (1.27) 2.13 (1.37)
Mean (SD) testosterone (nmol/l) 3.49 (3.68) 3.55 (3.54)
Testosterone concentration 4 nmol/l 37 (33%) 35 (31%)
Mean (SD) free androgen index 12.90 (17.95) 11.74 (11.89)
Free androgen index 8 33/70 (47%) 35/69 (51%)
Mean (SD) volume of ovaries (ml) 9.1 (6.1) 10.1 (4.9)
Ovary volume 10ml 19/70 (27%) 26/65 (40%)
Mean (SD) total motile sperm count (x106) 134 (223) 222 (370)
Homeostasis Model Assessment (HOMA) 4.6 (8.3) 3.8 (4.6)
BMI=body mass index; LH:FSH ratio=luteinising hormone:follicle stimulating hormone.
Table 2. Rates of ovulation, pregnancy and spontaneous abortion rates. Figures are
numbers (percentages) of women in each group.
Clomifene citrate + metformin (n=111) Clomifene citrate + placebo (n=114) Risk difference % (95% CI) Relative risk (95% CI) Ovulation 71 (64) 82 (72) -8 (-20 to 4) 0.89 (0.7 to 1.1) Ongoing pregnancy 44 (40) 52 (46) -6 (-20 to 7) 0.87 (0.6 to 1.2) Spontaneous abortion 13 (12) 12 (11) 1 (-7 to 10) 1.11 (0.5 to 2.3)
Table 3. Ovulation per dose of clomifene citrate. Figures are numbers (percentages)
of women who ovulated out of the total number of women.
Dose Clomifene citrate
+ metformin Clomifene citrate + placebo P value 50mg 49/80 (61) 50/92 (54) 0.36 100mg 27/44 (61) 35/53 (66) 0.63 150mg 8/17(47) 13/23 (57) 0.55
Table 4. Rates of ovulation, pregnancy and spontaneous abortion rates (per protocol
analysis). Figures are numbers (percentages) of women
Clomifene citrate + metformin (n=78 ) Clomifene citrate + placebo (n=84) Risk difference % (95% CI) Relative risk (95% CI) Ovulation 61 (78) 68 (81) -3 (-15 to 10) 0.97 (0.8 to 1.1) Ongoing 0.93
Chapter 3
The role of metformin in polycystic ovary syndrome. A systematic review.
Etelka Moll, Fulco van der Veen, Madelon van Wely
Abstract
This meta-analysis evaluated the effectiveness of metformin in subfertile women with polycystic ovary syndrome (PCOS). Only randomised trials investigating the effectiveness of metformin and PCOS definition consistent with the Rotterdam consensus criteria were eligible. Primary outcome was live birth rate. A literature search identified 27 trials. In therapy naive women, we found no evidence of a difference in live birth rate when comparing metformin with clomifene citrate (CC) [relative risk (RR) 0.73; 95% confidence interval (CI) 0.51 - 1.1] or comparing metformin plus CC with CC (RR 1.0; 95% CI 0.82 - 1.3). In CC-resistant women, metformin plus CC led to higher live birth rates than CC alone (RR 6.4; 95% CI 1.2 - 35); metformin also led to higher live birth rates than laparoscopic ovarian drilling (LOD) (RR 1.6; 95% CI 1.1 - 2.5). We found no evidence for a positive effect of metformin on live birth when added to LOD (RR 1.3; 95% CI 0.39 - 4.0) or FSH (RR 1.6; 95% CI 0.95 - 2.9), or when co-administered in IVF (RR 1.5; 95% CI 0.92 - 2.5). In IVF, metformin led to fewer cases of ovarian hyperstimulation syndrome (OHSS) (RR: 0.33; 95% CI: 0.13 - 0.80). This meta-analysis demonstrates that CC is still first choice therapy for women with therapy naïve PCOS. In CC-resistant women, the combination of CC plus metformin is the preferred treatment option before starting with LOD or FSH. At present, there is no evidence of an improvement in live birth when adding metformin to LOD or FSH. In IVF, metformin leads to a reduced risk of OHSS.
Introduction
The polycystic ovary syndrome (PCOS) affects 5% to 10% of women of reproductive age.1 PCOS is characterised by oligo-anovulation, clinical or biochemical hyperandrogenism and / or polycystic ovaries.2-4 Insulin resistance accompanied by compensatory hyperinsulinemia constitutes another major biochemical feature of PCOS.
In 1994, more than 70 years after the first description of a patient with insulin resistance and hyperandrogenism, the first study on the insulin sensitizer metformin in women with PCOS was published.5 Originally, this trial was meant to study metabolic and endocrinological parameters, but the authors noticed that some (12%) of the women conceived spontaneously.
From that moment on many trials were set up to test insulin sensitizers (mainly metformin) for ovulation induction in women with PCOS. These studies have been summarized in several reviews and meta-analyses.6-11 These meta-analyses were based on trials all consisting of a very small number of patients. In the analyses no consistent distinction between therapy naïve and clomifene citrate (CC)-resistant women was made. The reviews separately did not overview the total spectrum of treatment possibilities.
In addition, two large trials were recently published.12;13 The total number of patients in each separate trial exceeded the total number of patients in the existing reviews. Both trials found - in contrast to the previously published trials - that metformin does not lead to higher pregnancy rates when combined with CC and the same was true for metformin alone when compared with CC.
In view of this, we felt that updating our knowledge on metformin in subfertility and a critical appraisal of all existing studies might be helpful to guide clinical practice. In this review, we will therefore concentrate on the effect of metformin on live birth rate in women with PCOS for all comparisons studied so far.
Materials and Methods
Search Strategy
We searched the Cochrane Menstrual Disorders & Subfertility Group trials register, the Cochrane Central Register of Controlled Trials (both searched February 2007), MEDLINE (January 1966 to February 2007), the website for registration of controlled trials (controlled-trials.com) and several personal contacts with experts in this field (Balen, Nestler, Palomba). All electronic databases were searched using the following keywords: assisted reproduction, clomifene citrate, gonadotrophins, IVF, IUI, metformin, ovulation induction, PCOS, pregnancy. We handsearched the reference lists of selected trials and of recent reviews concerning this subject. No restrictions were held concerning publication year or language. All retrieved articles were of English language and published from 1996 till February 2007.
Study selection and data extraction
Studies were selected if the target population were women with PCOS. The definition of PCOS had to follow the standards of the ESHRE/ASRM 2003 consensus, or the criteria used in the article had to be, in retrospect, in consensus with the definition.4 If included patients did not meet the definition of ESHRE/ASRM, the study was not included in this review. Furthermore, the studies had to be of randomised design comparing the effect of metformin with placebo or no treatment, metformin with another ovulation induction agent or method or comparing the effect of metformin as co-treatment in IVF with no co-treatment.
The primary outcome of interest was live birth rate per randomised woman. Secondary outcomes were clinical pregnancy, multiple pregnancy and ovarian hyperstimulation syndrome (OHSS). It appeared that if live birth rate was not given in a manuscript, data on ongoing pregnancy were also not presented. Therefore, clinical pregnancy rate was chosen as a secondary outcome.
The review was undertaken by two reviewers (E.M., M.v.W.). The search strategy was employed to obtain titles and, where possible, abstracts of studies that were potentially relevant to the review. Both reviewers independently assessed whether the studies met the inclusion criteria, with disagreements resolved by discussion and final arbitration by F.V.
For each included trial, information was collected regarding the location of the study, methods of the study (as per quality assessment checklist), the participants
(age range, eligibility criteria), the nature of the interventions and data relating to the outcomes specified above. When possible, missing data were sought from the authors. The trial specific characteristics are expressed in Table 1.
We distinguished three indications: metformin as first-line treatment in therapy naive women; metformin as second-line treatment in CC-resistant women and metformin as co-treatment in women undergoing IVF. We subdivided first-line and second-line treatment into metformin monotherapy or co-treatment in combination with CC, FSH or laparoscopic ovarian drilling (LOD).
Statistical analysis
Relative risks (RR) with 95% confidence intervals (95% CI) were calculated for every study. Pooled RR were calculated using fixed effects models.14 If there was statistical heterogeneity, we performed a sensitivity analysis by pooling using a ‘random effects’-method.15
Statistical heterogeneity was assessed using forest plots, the I2 statistic and chi-square test. Clinical heterogeneity was assessed by reviewing differences across trials in characteristics of randomized patients.
Data from cross-over trials were only used from the first phase (i.e. before crossover). Any such trials that did not provide results at this point were excluded from the analysis. Review Manager-software (RevMan 4.2.7, Cochrane Collaboration, Oxford, UK) was used for the statistical analysis. We analysed the data on intention to treat basis.
Results of search
Our search selected 443 articles. After reading the titles, 296 articles did not answer our question. From the remaining 147 articles, 94 articles were discarded while they were non-original papers (reviews, letters). The 53 remaining articles were read. Thirty articles did not meet our inclusion criteria for not having a proper control group,5;16-24 for not using randomisation,25-27 for not providing clear information on live birth or clinical pregnancy rates,28-40 for using a cross-over design without clear rates
By handsearching reference lists, we came across four articles we did not find in the initial search.46-49 In total 27 studies were included in the analysis (Fig. 1).
Three studies compared metformin with placebo,46;50;51 two compared metformin with CC,13;52 12 compared metformin plus CC with CC,12;13;41;47;49;53-59 one compared metformin with LOD,60 one compared metformin plus LOD with LOD,61 one compared metformin plus CC with HMG,62 four compared metformin plus FSH with FSH48;63-65 and four compared metformin added in IVF versus IVF without metformin.34;66-68
Results
The quality and the main characteristics of the 27 trials included in this review are presented in Table 1. Most trials were of poor quality. Seventeen of 27 trials used an appropriate method of randomization, with 17 out of 27 having adequate concealment of allocation. A power calculation was only reported in eight trials. Trial size varied from 17 to 626 women.
Eight studies excluded women over the age of 35 years. Most studies did not have restrictions considering BMI. One study included women with BMI <25 kg/m2. Two studies included women with BMI <30 and <35 kg/ m2, respectively. Two studies included women with BMI >29 and 30 kg/m2, respectively.
Metformin in CC naïve women
Metformin monotherapy
We retrieved two randomised controlled trials in which metformin was compared with placebo for as first line treatment in 185 therapy naïve infertile women with PCOS (Table 1).46;51 None used HCG for triggering ovulation. Both trials reported clinical pregnancy rate. The pooled RR was 3.3 (95% CI: 0.92-11) (Fig. 2a). Visual examination of the forest plot and the I2 statistic (0%) suggested no heterogeneity in treatment effect across trials.
Live birth rate was not reported, nor multiple pregnancy rates. One study gave life style modification before starting drug therapy.51 The median weight loss was 2.8 versus 1.5%.
We found two double-blinded randomised controlled trials in which metformin was directly compared with CC as first line treatment in 509 infertile women with
PCOS (Table 1).13;52 HCG was not used for triggering ovulation. The pooled clinical pregnancy rate after six months of treatment was significantly lower after metformin (RR: 0.72; 95% CI: 0.54-0.97) (Fig. 2a). The pooled RR for live birth was 0.73 (95% CI: 0.51-1.1) (Fig. 2b). However, for both pregnancy outcomes there was significant heterogeneity in treatment effect across the two trials. When the data were pooled using a random effects model the RR was 0.88 (95% CI: 0.19 – 4.1) and 0.96 (95% CI: 0.11-8.2) for clinical pregnancy rate and live birth rate respectively. Furthermore, there was no evidence of a difference in multiple pregnancy rate between the two groups (RR: 0.38; 95% CI: 0.02-7.1).
Metformin as co-treatment in combination with CC
Seven randomised controlled trials compared CC plus metformin with CC in 985 infertile women with PCOS (Table 1).12;13;47;49;53-55 Two studies used HCG to trigger ovulation.47;55 After combining the data, there was a significantly higher clinical pregnancy rate in the metformin plus CC group (RR 1.5; 95% CI 1.2 - 1.8) (Fig. 2a). However, there was significant heterogeneity in treatment effect across the trials. When the data were pooled using a random effects model the difference in clinical pregnancy was still significant (RR 1.9; 95% CI: 1.2 – 3.3). The pooled RR for live birth was 1.0 (95% CI 0.82 - 1.3; three trials with 664 women) (Fig. 2b). For live birth, there was no indication for heterogeneity in treatment effect across trials.
Two studies reported multiple pregnancy rates.12;13 After combining these data no significant difference was seen (RR 0.38; 95% CI 0.09 - 1.5; 193 women).
Metformin in CC-resistant women Metformin monotherapy
We retrieved one randomised clinical trial in which metformin was compared with placebo in 18 infertile women with CC-resistant PCOS (Table 1).50 In this small number of women there was no evidence of a difference in clinical pregnancy or live birth rate (both RR: 0.50; 95% CI: 0.05-4.6) (Fig. 3a and b). No data on multiple pregnancy rates were given.
Metformin as co-treatment in combination with CC
We retrieved five randomised controlled trials in which CC plus metformin was compared with CC alone in 210 infertile women with CC-resistant PCOS (Table 1).41;56-59 Two trials used HCG to trigger ovulation.56;57
Combining the results showed that metformin plus CC led to a significantly higher clinical pregnancy rate than CC alone (RR: 5.6; 95% CI: 2.3-13) (Fig. 3a). Live birth rate was also in favour of metformin plus CC compared with the CC group (RR: 6.4; 95% CI: 1.2-34; 2 trials with 107 women) (Fig. 3b). For both pregnancy outcomes, visual examination of the forest plot and the I2 statistic (0%) suggested no heterogeneity in treatment effect across trials.
In the only trial that reported on multiple pregnancy no multiple pregnancies were observed in both groups.59
Metformin as opposed to LOD
Only one randomised trial was retrieved in which metformin treatment was compared with LOD (Table 1).60 There was no evidence of a difference in clinical pregnancy rate (RR: 1.3; 95% CI: 0.96-1.7) (Fig 3a). Live birth rate however was higher in the metformin group (RR: 1.6; 95% CI: 1.1-2.5) (Fig 3b). Multiple pregnancies were not observed.
Metformin as co-treatment in combination with LOD
One trial randomised 42 PCOS patients between LOD followed by metformin or LOD alone (Table 1).61 There were no significant differences in clinical pregnancy rate (RR: 2.3; 95% CI: 0.82-6.2) or live birth rate (RR: 1.3; 95% CI: 0.39-4.0) (Fig. 3a and 3b).
Metformin plus CC compared with gonadotrophins
In one randomised clinical trial, metformin plus CC was compared with gonadotrophins in 60 CC- resistant women (Table 1).62 Both groups were triggered for ovulation with HCG. There was no evidence of a difference in clinical pregnancy rate (RR: 0.71; 95% CI: 0.26-2.0) (Fig. 3a). There were no data on live birth, multiple pregnancy or OHSS.
Metformin plus FSH compared with FSH alone
In four randomised controlled trials FSH plus metformin was compared with FSH alone in 154 infertile women with PCOS (Table 1).48;63-65 All studies used HCG to trigger ovulation. The pooled clinical pregnancy rate was significantly higher in the FSH plus metformin group compared with FSH only group (RR: 1.7; 95% CI: 1.1-2.8) (Fig. 3a). A difference in live birth rate could however not be proven (RR: 1.6; 95% CI: 1.0-2.9) (Fig. 3b). For both pregnancy outcomes, visual examination of the forest plot and the I2 statistic (0%) suggested no heterogeneity in treatment effect across trials. Metformin led to less multiple pregnancies (RR: 0.26; 95% CI: 0.07-0.96). There was no evidence of a difference in OHSS (RR: 0.59; 95% CI: 0.17-2.1).
Metformin as additional treatment in controlled ovarian hyperstimulation in IVF
Four trials studied the effect of metformin during ovarian hyperstimulation in
IVF/ICSI in 283 women with PCOS (Table 1)34;66-68 (T.Tang, personal
communication).
In the first study, reasons for IVF treatment were not specified.34 In the second study, women received IVF or ICSI because of other fertility problems like tubal pathology, endometriosis or male subfertility.66 In the third study women with PCOS in whom conventional therapy had not lead to pregnancy, were included.67 In the fourth study, reasons for IVF were failure of conventional therapy and other fertility problems.68
All studies presented data in clinical pregnancy rate. Combining the results did not show a significant difference between the women treated with metformin or placebo (RR: 1.2; 95% CI: 0.85-1.6) (Fig. 4). Visual examination of the forest plot and the I2 statistic (0%) suggested no heterogeneity in treatment effect across trials. Live birth rate was reported in two studies.66;68 There was no evidence of a significant difference between the two groups (RR: 1.5; 95% CI: 0.92-2.5) (Fig. 4). Pooling the data of the two trials that reported multiple pregnancy gave no evidence of a significant difference between the two groups on multiple pregnancy rate66;68 (RR: 0.93; 95% CI: 0.42-2.1). OHSS was reported in all studies. When combining the
Discussion
In this review, we evaluated whether metformin leads to a more effective fertility treatment for women with PCOS.11;69-73 From the placebo controlled trials performed in infertile women with therapy naïve PCOS it is clear that metformin can induce ovulation and can lead to pregnancies. The important clinical question however is not whether metformin “works”, but whether it is better than CC in terms of live birth in CC naïve women or whether it has additional benefit in terms of live birth when used as co-treatment in therapy naïve or CC-resistant women.
At present, there is no evidence of a difference between metformin and CC in therapy naïve women in favour of metformin. The two trials that studied this comparison had conflicting results. In the study by Palomba et a 52 the live birth rate was three times higher in the metformin group. In contrast, in the study by Legro et
al13 with quadruple the number of patients, the live birth rate was three times lower in the metformin group. Of interest is that in the Palomba study, live birth rate in the CC group was unusually low due to a high miscarriage rate. Legro included patients previously treated with CC or metformin. We presumed these patients not to be CC-resistant. Through personal communication we were informed that it is not clear how many of these patients were CC-resistant. (R. Legro, personal communication). Still, this particular mixture of patients can explain the low live birth rate in this study.
Meta-analysis of the studies that compared co-treatment of metformin with CC versus CC alone did not show any benefit of metformin for live birth rate. These data, taken together, make it highly unlikely that metformin – as monotherapy or as co-treatment in combination with CC - is beneficial over CC in CC naïve women.
The clinical pregnancy rate in the comparison metformin plus CC versus CC in therapy naïve women was significantly higher in the metformin group. However, there was significant heterogeneity between studies as the small studies all favoured metformin plus CC above CC alone while this difference was not found in the larger studies. This difference between the larger and smaller studies may be a result of publication bias or low study quality bias.74;75 The sensitivity analysis using pooling with a random effects method was not helpful here as a random effects meta-analysis will award relatively more weight to smaller studies.
In CC-resistant women, two studies showed a clear benefit of adding metformin to CC over CC alone in terms of live birth. One should interpret these
results with some caution as one study was not blinded and the total number of patients in these two studies was only 107.
Metformin appears to be superior to LOD considering live birth rate in CC-resistant women, but these data are also based on a small number of women and from one monocenter study. This being so, metformin is quite a different treatment strategy than LOD and avoids the considerable risks of laparoscopic surgery, especially in obese patients.
No differences in live birth were detected when metformin was added to LOD compared with LOD alone and when metformin was added to FSH compared with FSH alone, but again few studies, including few patients, have been carried out so far.
Up till now, there is no evidence for better results on live birth rates when metformin is added during ovarian hyperstimulation in IVF. This is based on two studies with a limited number of patients and with probably totally different populations of women, as in one study a mix of women after failed ovulation induction and with other indications was included, while in the other studies only women with other fertility problems were included. Metformin may however, reduce the risk of OHSS.
In general, duration of metformin therapy differed substantially over the studies and we can only speculate which effects this will have on outcome parameters.
In summary, this meta-analysis demonstrates that CC is still first choice therapy for women with therapy naïve PCOS. In CC-resistant women, the combination of CC plus metformin is the preferred treatment option before starting with LOD or FSH. At present, there is no evidence of an improvement in live birth rates when adding metformin to LOD or FSH.
