| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Department of Obstetrics and Gynecology (S.P., A.F., T.R., F.Z.), University "Magna Graecia" of Catanzaro, 88100 Catanzaro, Italy; and Departments of Endocrinology (F.O.) and Obstetrics and Gynecology (A.T.), University "Federico II" of Naples, 80131 Naples, Italy
Address all correspondence and requests for reprints to: Stefano Palomba, M.D., Department of Gynecology and Obstetrics, University "Magna Graecia" of Catanzaro, Via Pio X, 88100 Catanzaro, Italy. E-mail: stefanopalomba{at}tin.it.
 |
Abstract |
|---|
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Objective: The aim of the study was to compare in a clinical setting the efficacy of CC and metformin as first-line approaches for treating anovulation in infertile PCOS patients.
Design: This study was a multicenter, nonrandomized, prospective, controlled study.
Setting: The study was conducted at the Department of Obstetrics and Gynecology, Universities "Magna Graecia" of Catanzaro and "Federico II" of Naples, Italy.
Patients: Patients included 80 infertile anovulatory patients with PCOS allocated in two body mass index- and age-matched groups (experimental and control groups).
Interventions: Six months of 1700 mg/d metformin treatment (experimental group) or CC was administered using a traditional incremental-dose protocol (control group). In both groups, patients who ovulated under treatment continued the therapy for a total of 6 months.
Main Outcome Measures: Reproductive outcomes were measured.
Results: Experimental and control groups were studied for a total of 204 and 169 cycles, respectively. No difference between the two groups was observed in ovulation (55.4 vs. 59.8%, respectively; P = 0.396), pregnancy (10.8 vs. 11.2%, respectively; P = 0.888), and abortion (19.5 vs. 26.3%, respectively; P = 0.530) rates. The cumulative pregnancy rate was not different between groups (62.9 vs. 48.6%, respectively; P = 0.225).
Conclusion: A 6-month course of 1700 mg/d metformin treatment and CC administered in an escalation protocol are two effective first-line approaches for improving fertility in anovulatory PCOS women.
|
Introduction |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Several approaches have been proposed to induce ovulation in women with PCOS (4).
Clomiphene citrate (CC) was the first agent used in experiments for ovulation induction in oligomenorrheic women, and it represents the first therapeutic option for treating the anovulatory infertility from many years (5). CC, in fact, is effective and characterized by low costs, limited dose-dependent side effects, and simplicity of administration and management (5).
More recently metformin, an oral biguanide used for type 2 diabetes mellitus, has been shown to induce ovulatory cycles in anovulatory CC-resistant or nonresistant patients with PCOS, also improving the ovulation rate as an additional treatment in patients who received CC (6, 7).
An experimental head-to-head randomized, controlled trial demonstrated that, in PCOS patients with anovulatory infertility, 6 months of CC and metformin administration resulted similarly effective in terms of ovulations, even if a significant superiority of metformin was observed in terms of pregnancies and live births (8).
Moreover, these data (8) were obtained in a well-selected population of nonobese PCOS patients who underwent ovulation monitoring and timed intercourses. In addition, to simplify the study design, i.e. double-blind, double-dummy, placebo-controlled, and to minimize the number of CC-resistant patients, CC was administered at fixed dosages using a dosage of 150 mg daily that has shown to exert deleterious effects at both endometrial (9) and ovarian (10) levels.
Based on these considerations, the aim of the present study was to compare in a clinical setting the efficacy of 6 months of metformin vs. CC administrated at incremental doses as first-line approaches for treating the anovulatory infertility in PCOS patients.
|
Subjects and Methods |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Between February 2006 and April 2006, 80 primary infertile anovulatory women with PCOS were enrolled in the current multicenter, prospective, nonrandomized, controlled study protocol. They were consecutively screened in the Department of Obstetrics and Gynaecology of the Universities "Magna Graecia" of Catanzaro and "Federico II" of Naples and in the Department of Endocrinology of the University "Federico II" of Naples.
In particular, we studied 40 consecutive PCOS infertile patients who received experimental treatment (experimental group) and another 40 age- and body mass index (BMI)-matched PCOS infertile patients who received traditional treatment and were considered controls (control group). The matching procedure was one to one, and two patients were defined as age and BMI matched when the differences between them were less than 2 yr and 1 kg/m2 for age and BMI, respectively.
In all cases, the diagnosis of PCOS was based on the presence of clinical [Ferriman-Gallwey score 
8 (11)] or biochemical hyperandrogenism [serum testosterone levels > 2 SD above our reference mean values] and chronic anovulation [serum luteal progesterone (P) < 2 ng/ml)].
The exclusion criteria for all subjects included neoplastic, metabolic, hepatic, and cardiovascular disorders or other concurrent medical illnesses; hypothyroidism, hyperprolactinemia, Cushing’s syndrome, and nonclassical congenital adrenal hyperplasia as excluded by appropriate tests; or current or previous use (within the last 6 months) of oral contraceptives, glucocorticoids, antiandrogens, ovulation induction agents, antidiabetic and antiobesity drugs or other hormonal drugs. Other exclusion criteria were organic pelvic diseases, previous pelvic surgery, suspected peritoneal factor infertility, or tubal or male factor infertility or subfertility as excluded by hysterosalpingogram and semen analysis, respectively. We also excluded women who intended to start a diet or a specific program of physical activity.
All subjects had a normal physical activity and none drank alcoholic beverages.
After enrollment, all patients were treated and followed up in the two Departments of Obstetrics and Gynecology of Catanzaro and Naples. In particular, the patients selected and enrolled in the Department of Endocrinology of the University "Federico II" of Naples were treated and followed up in the Department of Obstetrics and Gynecology of the same university, whereas the patients enrolled in the Department of Obstetrics and Gynecology of the University "Magna Graecia" of Catanzaro were treated and followed up in the same department.
At study entry, all subjects underwent clinical evaluations and venous blood drawing to evaluate complete hormonal assays and fasting serum glucose and insulin levels.
Clinical evaluation consisted of anthropometric measurements [including height, weight, BMI, and waist to hip ratio (WHR)] and Ferriman-Gallwey score. Specifically, BMI was calculated as the ratio between the weight and the square of the height, and WHR as the ratio between the waist (considered to be the smallest circumference of torso between the 12th rib and the iliac crest) and the circumference of the hip (considered as the maximal extension of the buttocks). All measurements were performed when the patients were in a standing position with relaxed abdomen, arms at their sides, and joined feet (12). The Ferriman-Gallwey score was calculated by the standard method (11).
During the same visit, a semiquantitative questionnaire to evaluate the patients’ daily physical activity, job and daily activities (13), and a transvaginal ultrasonography were performed in each patient.
Blood samples were obtained in the morning between 0800 and 0900 h after 12-h overnight fasting and resting in bed during the early proliferative phase (second to third day) of the P-induced withdrawal uterine bleeding (100 mg natural P, im).
Complete hormonal assays consisting of FSH, LH, TSH, prolactin (PRL), 17-ß estradiol (E2), P, 17-hydroxyprogesterone (17-OHP), total testosterone (T), androstenedione (A), dehydroepiandrosterone sulfate (DHEAS), and SHBG were evaluated. As previously reported (14), all plasma hormone concentrations were measured by specific RIA, whereas SHBG levels using an immunoradiometric assay. Serum insulin was assayed by a solid-phase chemiluminescent enzyme immunoassay using commercially available kits, and serum glucose levels were determined by the glucose oxidase method (14).
Finally, the homeostasis model of assessment [fasting glucose (millimoles per liter) x fasting insulin (microunits per milliliter)/22.5], the fasting glucose to insulin ratio (milligrams per 10–4 U), and the free androgen index [T (nanomoles per liter)/SHBG x 100]) were calculated in each subject.
The experimental group was treated with metformin at a dosage of 850 mg twice daily. Metformin was given starting from the third day of a P-induced withdrawal bleeding (10 mg natural P, im) using a dose of 850 mg (one tablet daily) and increasing the dosage after 1 wk up to 1700 mg/d (two tablets daily). Metformin was taken before lunch during the first week and thereafter before lunch and dinner. All subjects who became pregnant throughout the study suspended the metformin administration. Patients who did not ovulate under metformin for 6 months also suspended the treatment, whereas those who started to ovulate with metformin administration continued the treatment for a total of six ovulatory cycles.
The control group received CC using a classical incremental regimen. In particular, CC was administered for 5 d beginning on cycle d 3 after a P-induced withdrawal bleeding (10 mg natural P, im) using a starting dose of 50 mg daily. If ovulation did not occur, the dose was increased by 50 mg in successive cycles until the ovulation was achieved or up to a maximal dose of 250 mg daily. Patients who did not ovulate under 250 mg/d CC were classically defined as CC resistant and suspended the treatment, whereas patients who ovulated under CC continued the treatment at the same dosage for a total of six ovulatory cycles.
No agent to induce ovulation, e.g. human chorionic gonadotropin (hCG), was administered throughout the study. All subjects, in the absence of spontaneous withdrawal bleeding after 35 d from last P-induced uterine bleeding and after exclusion of a pregnancy with a serum ß-hCG assay, received a further dose of 100 mg natural P, im.
Throughout the study, no change in diet and physical activity was implemented. On the contrary, the subjects were instructed to follow their usual diet and physical activity.
Each patient was motivated and instructed to have sexual intercourses regularly (at least one intercourse every 3 d for four times) starting 9 d after the P-induced uterine bleedings (the day after the CC suspension for the control group).
During the study, the ovulation, pregnancy, and abortion rates were evaluated in each woman.
No ovulation monitoring was performed throughout treatment administration, but the ovulations were defined by plasma P assay [>10 ng/ml (SI: 32 nmol/liter)] performed 21 d after the spontaneous or P-induced bleedings (7 d before the expected menses).
Ovulation rate was calculated as the percentage of ovulatory cycles per total observed cycles. The pregnancy rate was defined as the percentage of pregnancies per total observed cycles. The rate of pregnancies for each ovulatory cycle was also calculated. A regular rising ß-hCG and the sonographic evidence of intrauterine gestational sac were considered criteria to define a pregnancy. The abortion rate was defined as a percentage of miscarriage during the first 12 wk of gestation per total pregnancies. Cumulative pregnancy rate was defined as ratio between number of pregnant patients and total patients. The cumulative pregnancy rate was also calculated according to time to the first event as detailed in the statistical analysis section.
All subjects were instructed to report the characteristics of their menstrual cycles and the onset of any adverse events (AEs) in a personal daily diary. For each AE reported, severity, duration, and possible cause-effect relationship with drug administrations were noted. To evaluate the compliance with the treatment and protocol, the number of skipped tablets and changes in diet, physical activity, and weight as well as the characteristics (number and timing) of the sexual intercourses were also recorded in the same diary.
Statistical analysis
The primary end point of the current study was the pregnancy rate. One goal of the proposed study was to test the null hypothesis that the proportion positive was identical in both populations. The criterion for significance (alpha) was set at 0.05. The test was two tailed, which means that an effect in either direction will be interpreted.
Based on our previous data (8), we expected a cumulative pregnancy rate of 0.68 and 0.34 after metformin and CC, respectively, with a difference in proportion of 0.34. Thus, we needed to enroll 33 patients for each group to yield a statistically significant result with a power study of 80.3%. To allow for an unpredictable number of withdrawals, we decided to enroll a total of 80 patients in the expectation that at least 33 patients would be left in each group. The power analysis and the sample size calculation has been performed using SamplePower release 2.0 (SPSS Inc., Chicago, IL).
Data were analyzed using the intention-to-treat method on the basis of treatment assignment and not on treatment receipt.
For categorical variables, the Pearson 
2 test was performed; conversely, the Fisher’s exact test was required for the frequency tables when more than 20% of the expected values were less than 5.
The normal distribution of continuous variables data were evaluated with the use of the Kolmogrov-Smirnov test. Thus, our data were expressed as median and interquartile range (IQR) with minimum-maximum values, and the differences between groups were analyzed using the Mann-Whitney U test.
Cumulative event (pregnancy) rate was calculated by the Kaplan-Meier method, using the time to a first event as the outcome variable, and the differences between groups were tested with the use of log-rank test. In addition, Cox proportional-hazards model was used to calculate the hazard ratio and its 95% confidence interval for pregnancy in patients who received experimental or control treatment. The hazard ratio also represented the relative risk because it was calculated for a dichotomous variable in which there are two levels.
Data were also analyzed with the use of general linear model univariate procedure to evaluate the influence of the centers involved (Catanzaro and Naples) on the results.
Statistical significance was set at P < 0.05. The Statistics Package for Social Science (SPSS 14.0.1; SPSS Inc.) was used for all statistical analyses. The number needed to treat was calculated with StatsDirect, release 2.4.3 (StatsDirect, Cheshire, UK).
|
Results |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
No difference (P = 0.460) in proportion of subjects studied for each center was detected [14 (40.0%) and 18 (48.6%) for experimental and control group, respectively, in the Department of Obstetrics and Gynecology of Catanzaro vs. 21 (60.0%) and 19 (51.4%) for experimental and control group, respectively, in the Department of Obstetrics and Gynecology of Naples].
At baseline, no difference in any demographic data was observed between the two groups that were well matched for age and BMI (Table 1
). No difference between the experimental and control groups was detected in the proportion of lean (11.1 vs. 8.1%, respectively; P = 0.468), normal-weight (28.6 vs. 29.7%, respectively; P = 0.914), overweight (37.1 vs. 37.8%, respectively; P = 0.952), and obese (22.9 vs. 24.3%, respectively; P = 0.884) patients.
|
At the end of the study, the median of follow-up cycles was significantly longer in the experimental than in the control group [six (1 IQR; range 2–10) and five (4 IQR; range 1–9), respectively; P = 0.003].
The experimental and control groups were studied for a total of 204 and 169 cycles, respectively.
Table 2 summarizes the reproductive outcomes recorded throughout the study in both groups.
|
). Two groups did not differ for number of sexual intercourses for each cycle of observation (data not shown).
No significant difference between groups was observed in the rate of pregnancies for each ovulatory cycle during the study (Fig. 1).
|
The probability of first pregnancy by the Kaplan-Meier survival analysis was also similar between groups (P = 0.246) with an estimated median of 7 and 5 months for the experimental and control groups, respectively (Fig. 2). Using a Cox regression analysis, the benefit associated with the metformin administration was not significant (P = 0.269). In fact, infertile patients under experimental treatment, compared with those under control treatment, had a relative risk of pregnancy of 1.4 (95% confidence interval 0.755–2.747).
|
One case of multiple pregnancy (twin) was reported in the control group (150 mg daily CC) (Table 2).
During the study, the two treatments were well tolerated and the total incidence of AEs was not significantly different between groups. No serious AE or laboratory abnormalities were reported in either group during the study.
Using the general linear model, no influence of the investigational center involved on the results was detected at the end of the study.
|
Discussion |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
In the current study, we compared the efficacy of metformin vs. CC as first-line approaches for treating anovulatory infertile PCOS patients in a more realistic scenario. For this aim we studied a very heterogeneous population of PCOS patients (consecutive cases of infertile PCOS patients) divided into two treatment groups closely matched for age and BMI.
Our data showed that 6 months of metformin administrated at a dosage of 1700 mg daily is effective as well as CC administrated using a classical protocol at incremental doses in terms of ovulation, pregnancy, and abortion rates.
These findings disagree with not only two recent clinical studies (18, 19) but also our previous data (8).
Previously we showed that metformin had significant advantages in comparison with CC in terms of pregnancy and abortion rates as fist-line treatment for ovulatory infertility in PCOS (8). Although this clinical trial was formally well designed, i.e. randomized, controlled, double-blind, double-dummy fashion, the study population, composed of only nonobese patients and subjects screened for diabetes and/or glucose intolerance, was not really representative of the entire PCOS population (8). In fact, nonobese PCOS patients (20) seems to have a better response to metformin administration, and to date, it is not known whether metformin acts better or worst in patients more or less insulin resistant.
In addition, for simplifying the double-blind design and to limit the rate of CC-resistant patients, CC was administered at fixed dosages of 150 mg daily for six consecutive cycles (8). In this regard, our subanalyses demonstrated an effect of 150 mg CC on endometrial thickness, pattern and vascularity (9), and perifollicular blood flows (10). These last findings seem to suggest that the repetitive administration of CC at high and fixed dosages could have a deleterious effect on endometrial receptivity and/or oocyte/embryo quality, whereas the escalation-dose protocol is probably the safest and most effective protocol to administer CC in clinical practice (21, 22).
A recent retrospective study by Neveu et al. (18) demonstrated that metformin is better than CC as first-line approach to induce ovulation in anovulatory PCOS women, whereas the pregnancy and abortion rates resulted similar in both treatment groups, without further advantages obtained by the combination of metformin plus CC. Notwithstanding the several limits due to the lack of randomization of the study population and standardization of the procedures and treatments, the results of this last study (18) seem to be close to a real-life clinical setting.
Our findings differ also from a recent large multicenter randomized, controlled trial (19) that demonstrated that CC was more effective than metformin to treat anovulation in PCOS, without any significant benefit of the combination of metformin plus CC. Furthermore, this last study (19) presented several selection biases (23). First of all, more than 50% of patients received previous study drug treatments, whereas patients not exposed to study drug potentially used other hormonal treatments, inducing a bias due to the selection of patients and persistence of the effects of the previous treatments (24). In addition, the population studied was composed for a large quantity of severely and/or morbid obese patients who have a worse response to metformin (20) and/or are usually under insulin-sensitizing drugs, cholesterol-lowering drugs, dietary treatments, and physical exercise programs or undergo bariatric surgery (25, 26, 27). Finally, couples with potential subfertility factors (28, 29, 30, 31) were not excluded from the study by Legro et al. (19).
Even if our study was not powered to demonstrate a difference in the rate of pregnancies for ovulation cycles, it is possible to observe a trend in different clinical responses to metformin and CC throughout the study (Fig. 1). In fact, as already previously suggested (8, 18), the percentage of pregnancies relative to the ovulatory cycles was constant throughout metformin administration. On the contrary, in agreement with other authors (18, 21, 22), more than 70% of pregnancies occur within the first three cycles of CC administration, with a significant reduction over the time.
These crucial observations suggest that a different therapeutic strategy should be adopted to optimize metformin and CC treatment.
In particular, patients who remain anovulatory after 6 months of metformin administration seem to have no benefit to continue the treatment for a longer time, as already observed in our previous study (32). Indeed, subjects who ovulate under metformin should be treated for a period of at least 6 months because the probability of pregnancy remains unchanged over the time. In fact, metformin administration showed to improve fertility acting on several surrogate markers of endometrial receptivity and oocyte/embryo quality (10, 33).
Conversely, our findings demonstrated that conceptions mainly occur within the first months of CC therapy, and no significant advantages can be obtained using CC at high dosages. Thus, patients who do not ovulate after 3 months of CC at a daily dose of 150 mg should be switched to other second-line treatments. On the other hand, women who do not achieve a pregnancy during the first ovulatory cycles again have few advantages to continue the CC administration. In fact, as discussed before, CC administered for a longer time period could exert its adverse antiestrogenic effects on the reproductive apparatus (9, 10).
In this regard, in the study by Legro et al. (19), subjects were treated for six cycles, and these protocols of administration did not optimize the therapeutic effects of both metformin and CC. In fact, in several patients, who became ovulating at the last cycles, treatment was probably stopped much too early. In addition, keeping in mind that metformin exerts a clinical response later than CC, patients received metformin for a suboptimal treatment period (19).
Based on these considerations, this crucial limit could explain the very low pregnancy and live-birth rates observed in women treated with metformin (19).
In a clinical setting, it is not possible to compare two very different strategies during the same study period because different approaches need different times to obtain the maximal clinical effectiveness. In this regard, in the current study, a significant difference between metformin and CC protocols was observed in the follow-up period. In addition, the Kaplan-Meier survival analysis showed that the first pregnancy occurred later after metformin than CC with an estimated median of 7 and 5 months, respectively.
Although the probability of first pregnancy by the same analysis did not result in significant difference between groups, further studies on a larger study sample and with a long-term follow-up are needed to drawn definitive conclusions. This last observation gives rise to an important question regarding the issue of the infertility treatment: is it better to choose an approach that gives the higher clinical response or one that acts more quickly?
At the moment, two main controversies should be solved in the near future: first, to define the best protocol for metformin administration and the most effective dosage of metformin for different PCOS subpopulations, and second, to define predictive factors for a good response to metformin to optimize the treatment(s) of the anovulatory infertility related to PCOS.
In conclusion, our study demonstrates that both CC and metformin are two effective first-line options to treat anovulatory infertility in PCOS patients, even if the efficacy of these approaches has a different proceeding over time. Further studies on well-selected populations are needed to optimize the treatment protocols according to specific predictive factors.
|
Footnotes |
|---|
First Published Online June 26, 2007
Abbreviations: A, Androstenedione; AE, adverse event; BMI, body mass index; CC, clomiphene citrate; DHEAS, dehydroepiandrosterone sulfate; E2, 17-ß estradiol; hCG, human chorionic gonadotropin; IQR, interquartile range; 17-OHP, 17-hydroxyprogesterone; P, progesterone; PCOS, polycystic ovary syndrome; PRL, prolactin; T, testosterone; WHR, waist to hip ratio.
Received May 10, 2007.
Accepted June 14, 2007.
|
References |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  R. Shao, M. Nutu, B. Weijdegard, E. Egecioglu, J. Fernandez-Rodriguez, L. Karlsson-Lindahl, K. Gemzell-Danielsson, C. Bergh, and H. Billig Clomiphene Citrate Causes Aberrant Tubal Apoptosis and Estrogen Receptor Activation in Rat Fallopian Tube: Implications for Tubal Ectopic Pregnancy Biol Reprod, June 1, 2009; 80(6): 1262 - 1271. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Palomba, A. Falbo, F. Zullo, and F. Orio Jr. Evidence-Based and Potential Benefits of Metformin in the Polycystic Ovary Syndrome: A Comprehensive Review Endocr. Rev., February 1, 2009; 30(1): 1 - 50. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. C. Vantyghem, D. Vincent-Desplanques, F. Defrance-Faivre, J. Capeau, C. Fermon, A. S. Valat, O. Lascols, A. C. Hecart, P. Pigny, B. Delemer, et al. Fertility and Obstetrical Complications in Women with LMNA-Related Familial Partial Lipodystrophy J. Clin. Endocrinol. Metab., June 1, 2008; 93(6): 2223 - 2229. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
