Abstract
Context:
In overweight/obese women with polycystic ovary syndrome (PCOS), the relative benefit of delaying infertility treatment to lose weight vs seeking immediate treatment is unknown.
Objective:
We compared the results of two, multicenter, concurrent clinical trials treating infertility in women with PCOS.
Design, Setting, and Participants:
This was a secondary analysis of two randomized trials conducted at academic health centers studying women 18–40 years of age who were overweight/obese and infertile with PCOS.
Intervention:
We compared immediate treatment with clomiphene from the Pregnancy in Polycystic Ovary Syndrome II (PPCOS II) trial (N = 187) to delayed treatment with clomiphene after preconception treatment with continuous oral contraceptives, lifestyle modification (Lifestyle: including caloric restriction, antiobesity medication, behavioral modification, and exercise) or the combination of both (combined) from the Treatment of Hyperandrogenism Versus Insulin Resistance in Infertile Polycystic Ovary Syndrome (OWL PCOS) trial (N = 142).
Main Outcome Measures:
Live birth, pregnancy loss, and ovulation were measured.
Results:
In PPCOS II, after four cycles of clomiphene, the cumulative per-cycle ovulation rate was 44.7% (277/619) and the cumulative live birth rate was 10.2% (19/187), nearly identical to that after oral contraceptive pretreatment in the OWL PCOS trial (ovulation 45% [67/149] and live birth: 8.5% [4/47]). In comparison, deferred clomiphene treatment preceded by lifestyle and combined treatment in OWL PCOS offered a significantly better cumulative ovulation rate compared to immediate treatment with clomiphene. (Lifestyle: 62.0% [80/129]; risk ratio compared to PPCOS II = 1.4; 95% confidence interval [CI], 1.1–1.7; P = .003; combined: 64.3% [83/129]; risk ratio compared to PPCOS II = 1.4; 95% CI, 1.2–1.8; P < .001 and a significantly better live birth rate lifestyle: 25.0% [12/48]; risk ratio compared to PPCOS II = 2.5; 95% CI, 1.3–4.7; P = .01 and combined: 25.5% [12/47]; risk ratio compared to PPCOS II = 2.5; 95% CI, 1.3–4.8; P = .01).
Conclusions:
These data show the benefit of improved ovulation and live birth with delayed infertility treatment with clomiphene citrate when preceded by lifestyle modification with weight loss compared with immediate treatment. Pretreatment with oral contraceptives likely has little effect on the ovulation and live birth rate compared with immediate treatment.
Comparing data from two randomized trials shows improved ovulation and live birth with delayed fertility therapy preceded by lifestyle modification compared with immediate therapy with clomiphene.
Couples face a dilemma when confronted with both infertility and a modifiable morbidity (such as obesity) that may limit their chance for conception and/or a healthy pregnancy: Should they start infertility therapy and ignore the morbidity or delay treatment to improve that morbidity? In general, there may be inadequate treatments to correct the morbidity or, as often may be the case, limited information exists supporting that modification improves outcomes. Finally, although there may be accepted treatments for the morbidity, the time required to achieve a benefit may significantly shrink the remaining fertility window of the woman, exhausting both patience and precious fecundability because advancing female age remains the single most important factor predicting infertility treatment failure (1). Delay in treatment as perceived by patients is like time's (and fertility's) winged chariot (2) speeding away without them.
Obesity contributes to significant periconceptional and perinatal morbidity in females and is associated with prolonged time to conception (3), increased pregnancy loss (4), and higher rates of adverse pregnancy outcomes such as preeclampsia and preterm labor, in turn, leading to fetal morbidity and mortality (5, 6). When combined with other medical conditions such as diabetes or polycystic ovary syndrome (PCOS), there are synergistic deteriorations in these outcomes (7–9). Expert opinion has uniformly recommended that obese women with PCOS delay infertility therapy and pursue lifestyle modification (10–12), though there is a paucity of high-quality evidence to document the efficacy (13, 14), let alone the optimum weight loss or duration of preconception intervention. Further, there are difficulties in designing such a trial of immediate fertility treatment vs delayed treatment after morbidity modification, especially if life table analyses are used for the primary outcome (ie, the delayed fertility treatment group by definition will start later than the immediate treatment group) (15, 16). The delayed treatment group will have a longer period of observation for pregnancy (eg, both preconception and infertility treatment vs immediate infertility treatment), which may serve to blunt the adverse effects of preconception intervention or augment it if successful. Further, the Kaplan-Meier curve would likely initially be skewed in favor of immediate treatment with its higher per-cycle pregnancy rate (compared to the low, if absent, pregnancy rate during preconception intervention). Thus an overall proportion comparing the cumulative live birth rates may be the preferred summary statistic for trials comparing varying interventions of different lengths.
Recently, we demonstrated a significant benefit of lifestyle modification on ovulation in overweight and obese women. We showed a trend towards improved live birth (The Treatment of Hyperandrogenism vs. Insulin Resistance in Infertile PCOS Women [OWL PCOS] study), an outcome the study was underpowered to address (17). Concurrently, we led a multicenter trial of immediate ovulation induction in infertile women with PCOS with clomiphene or letrozole without a preconception intervention (The Pregnancy in Polycystic Ovary Syndrome II [PPCOS II] study) (18). The intentional similar design in inclusion and exclusion criteria for PCOS, the administration of the same ovulatory agent, clomiphene citrate, and the tracking of pregnancy outcomes allows for a post hoc analysis to examine the effects of immediate treatment with clomiphene citrate in overweight/obese women with PCOS vs preconception treatment with either lifestyle modification, oral contraceptives, or the combination of the two.
Materials and Methods
Design of the trials
Both the OWL PCOS and the PPCOS II studies were multicenter trials supported by the National Institutes of Health with the same lead investigator (R.S.L.) and significant overlap in the investigative teams (R.S.L., W.C.D., C.L.G., S.J.E., A.R.K., C.C., K.T.B., A.D.) and trial design (Table 1). We have previously published the baseline (19) and main outcome data for PPCOS II (18) and the protocol is online at: http://www.nejm.org/doi/full/10.1056/NEJMoa1313517. We have also published the main outcomes for the OWL PCOS Study (17) and the protocol and case report forms can be accessed at http://ctsi.psu.edu/owl-pcos/. Both trials were approved by the investigational review boards at all sites and female participants and their male partners gave written informed consent. The trials were registered at Clinicaltrials.gov (OWL PCOS: NCT00704912 and PPCOS II: NCT00719186).
Table 1.
Comparison Between the OWL PCOS and the PPCOS II Trials
| OWL PCOS | PPCOS II | |
|---|---|---|
| Enrollment period | October 2008–December 2012 | February 2009–January 2012 |
| Primary outcome | Live birth | Same |
| Number of subjects | 149 (142 met criteria for this study) | 750 (376 in clomiphene arm and 187 met criteria for this study) |
| Inclusion criteria | ||
| PCOS | Modified Rotterdam criteria (2, 18): all women had ovulatory dysfunction with either hyperandrogenism (by hirsutism (19) or an elevated testosterone level (20)) or a polycystic ovary on transvaginal ultrasound (19) | Same |
| Age | 18–40 y | Same |
| BMI | 27–42 kg/m2 | No BMI restrictions |
| Exclusion criteria | ||
| Other comorbidities | Type 2 diabetes and major medical comorbidities | Same |
| Confounding medications | Sex steroids, insulin sensitizers, and other infertility drugs | Same |
| Other infertility factors | Tubal/uterine: at least one patent Fallopian tube and a normal uterine cavity | Same |
| Male: a sperm concentration of >14 million per milliliter with documented motility in at least one ejaculate within the past year and the commitment to have vaginal intercourse during the study with the intent of pregnancy | ||
| Treatment groups | ||
| Preconception intervention | 16-week preconception treatment with either continuous oral contraception, lifestyle modification, or a combination of both followed by four cycles of ovulation induction with clomiphene citrate | None |
| Infertility treatment | Four cycles of treatment with clomiphene citrate and timed intercourse | Five cycles of ovulation induction with letrozole or clomiphene citrate and timed intercourse |
| Blinding | Open-label preconception treatments and clomiphene | Double blinding of study drugs |
| Central laboratory | University of Virginia Core Ligand Laboratory | Same |
| Pregnancy outcomes | Maternal and infant pregnancy and hospital records reviewed | Same |
Abbreviations: BMI, body mass index; OWL PCOS, Treatment of Hyperandrogenism Versus Insulin Resistance in Infertile Polycystic Ovary Syndrome; PPCOS, Pregnancy in Polycystic Ovary Syndrome.
In brief, OWL PCOS was a randomized open-label two-site study of overweight/obese infertile women with PCOS with equal allocation to three treatment groups: 1) oral contraceptives (OCP) given continuously; 2) lifestyle modification (lifestyle), consisting of caloric restriction with meal replacements, antiobesity medication (sibutramine or orlistat) brief behavioral modification and increased physical activity; and 3) combined OCP and lifestyle modification (combined). Orlistat was substituted for sibutramine after the Food and Drug Administration issued an advisory about excess cardiovascular events associated with the drug and the drug was eventually removed from the market. In OWL PCOS, up to four cycles of induction of ovulation study medication were used (14). In brief, PPCOS II was a randomized, double-blind, multisite study of two ovulation induction agents: letrozole or clomiphene citrate in infertile women with PCOS. In PPCOS II, up to five cycles of induction of ovulation study medication were used (18), though we only examined the first four cycles in this study.
In both OWL PCOS and PPCOS II, oral ovulation induction agents were given in a standardized regimen. Baseline monitoring determined if patients were anovulatory or ovulatory. If anovulatory in PPCOS II, menses was induced with progestin and study drug started with subsequent menses. In OWL PCOS, investigators had the option to begin clomiphene immediately or induce menses with progestin. If ovulatory, women in both studies waited for a subsequent menses and began ovulation induction medication in the early follicular phase of their next menstrual cycle. Subjects in OWL PCOS were advised to maintain their weight and level of activity from the end of the preconception phase (ie, the end of the lifestyle modification and/or oral contraceptive phase), whereas meal replacements were discontinued. In PPCOS II, there were no instructions given regarding weight and activity during treatment. In both studies, a visit 3 weeks following clomiphene treatment ovulation was confirmed based on ultrasound monitoring and serum progesterone level. Patients with documented ovulation (progesterone level ≥3 ng/dl) had a serum pregnancy test 2 weeks later if no menses had occurred; with menses, they began clomiphene at the same dose at which they had ovulated. No other ultrasound monitoring was performed in either study.
In women without evidence of ovulation, the dose of clomiphene was increased to the next higher dose (100 mg for 5 days, up to a maximum of 150 mg for 5 days). In both studies, women who conceived were followed with serial serum human chorionic gonadotropin levels until an ultrasound could document fetal viability; subsequently, they were referred to local providers for prenatal care. Medical records from pregnancy (mother and infant) were obtained and reviewed to ascertain birth outcomes.
Assays
All reported laboratory values from both studies (except glucose and lipid levels) were determined in a central laboratory (Ligand Core Laboratory, University of Virginia) (19), using the same assays as previously reported in our PPCOS II study (18, 20). All assays had intra- and interassay coefficients of variation of less than 10% (18, 20). Glucose levels were measured by the glucose oxidase method at both the Penn State College of Medicine (OWL PCOS) and the Core Ligand Lab (PPCOS II). Lipid levels (total cholesterol, high-density lipoprotein cholesterol and triglyceride levels) were measured through commercial laboratories at both the Penn State College of Medicine (OWL PCOS) and the Core Ligand Lab (PPCOS II). The low-density lipoprotein cholesterol level was calculated using the Friedewald equation (21).
Statistical analysis
The primary outcome in both trials was live birth. We extracted from the PPCOS II database all patients from the clomiphene treatment group who met the same body mass index inclusion criteria of the OWL PCOS study and only used the first four ovulation induction cycles of the PPCOS II study to match the OWL-PCOS treatment protocol. Of the initial 200 women selected from the clomiphene arm of the PPCOS II database that met the OWL PCOS inclusion criteria, 20 had participated sequentially in both studies. For those 20 patients, our analysis used only the first study in which they participated. This yielded a final sample of 187 women from the PPCOS II study and 142 women from the OWL PCOS study. Predetermined secondary outcomes included ovulation rates, conception, and pregnancy loss rates. For binary outcomes (eg, live birth, pregnancy, ovulation), a log-binomial regression model was used to assess differences between the OWL PCOS treatment groups and the PPCOS II group. Because this was a prospective study, the log-binomial model allows us to estimate the risk ratio rather than the odds ratio, which is provided when using ordinary logistic regression. Linear regression was used to assess differences between the OWL PCOS treatment groups and the PPCOS II group for continuous birth outcomes (eg, infant birth weight). All hypothesis tests were two-sided and all analyses were performed using SAS software, version 9.4 (SAS Institute Inc.).
Results
Baseline characteristics of key demographic, reproductive, and metabolic parameters were evenly matched between treatment groups from the two studies (Table 2). Compared to immediate treatment with clomiphene in the PPCOS II trial, patients in the lifestyle group of the OWL PCOS study as well as patients in the combined group of the OWL PCOS study had a 2.5-fold (P = .01) increase in live births (Table 3). Patients in the Lifestyle and combined groups also had a 1.4-fold increase in cumulative ovulation rate (P = .003 and P < .001, respectively) compared to treatment with clomiphene in PPCOS II. In contrast, women in the OWL PCOS group who received OCP preconception treatment had no significant difference in live birth or ovulation rates compared to immediate treatment with clomiphene in PPCOS II. There were no differences in risk for pregnancy loss in any of the OWL PCOS treatment groups compared to treatment with clomiphene in PPCOS II.
Table 2.
Baseline Characteristics From the PPCOS II Clomiphene and OWL PCOS Selected Cohorts
| PPCOS II (n = 187) Mean (sd) | OWL PCOS: OCP (n = 47) Mean (sd) | OWL PCOS: Lifestyle (n = 48) Mean (sd) | OWL PCOS: Combined (n = 47) Mean (sd) | |
|---|---|---|---|---|
| Demographics | ||||
| Age (y) | 28.9 (4.1) | 29.8 (3.8) | 28.6 (3.4) | 28.5 (4.0) |
| Hispanic: N (%) | 41 (21.9) | 6 (12.8) | 5 (10.4) | 4 (8.5) |
| Caucasian: N (%) | 147 (78.6) | 38 (80.9) | 35 (72.9) | 29 (61.7) |
| Black/African-American: N (%) | 29 (15.5) | 7 (14.9) | 7 (14.6) | 13 (27.7) |
| Other/multiracial: N (%) | 11 (5.9) | 2 (4.3) | 6 (12.5) | 5 (10.6) |
| Duration of time attempting to conceive (mo) | 44.4 (38.3) | 43.3 (33.1) | 32.5 (27.5) | 34.5 (28.7) |
| Nulliparous: N (%) | 147 (78.6) | 37 (78.7) | 39 (81.3) | 42 (89.4) |
| Biometric | ||||
| Weight (kg), preweight reduction intervention | NA | 94.2 (14.3) | 96.0 (15.6) | 95.5 (14.8) |
| Weight (kg), at start of ovulation induction | 93.2 (14.2) | 93.5 (14.8) | 90.6 (13.9) | 88.4 (14.4) |
| % Weight Loss after weight reduction intervention | NA | 1.3 (3.1) | 6.4 (3.9) | 6.7 (4.2) |
| BMI (kg/m2), preweight reduction intervention | NA | 35.0 (4.2) | 35.0 (4.6) | 35.4 (4.5) |
| BMI (kg/m2), at start of ovulation induction | 34.8 (4.3) | 34.8 (4.4) | 33.0 (4.3) | 32.8 (4.2) |
| Waist (cm) | 107.1 (12.2) | 105.7 (11.0) | 107.2 (13.5) | 106.6 (12.1) |
| Systolic BP (mm Hg) | 121.0 (12.1) | 116.0 (10.2) | 114.7 (13.8) | 118.9 (11.8) |
| Diastolic BP (mm Hg) | 77.5 (9.1) | 73.6 (7.8) | 71.7 (8.2) | 74.6 (9.6) |
| Ferriman-Gallwey hirsutism score | 17.4 (8.2) | 16.6 (7.8) | 19.2 (8.8) | 18.0 (9.0) |
| Ultrasound parameters | ||||
| Antral follicle count (both ovaries) | 46.2 (27.6) | 63.6 (36.0) | 54.5 (34.4) | 60.1 (34.5) |
| Total ovarian volume (cm3)a | 22.1 (16.7–28.8) | 22.0 (17.3–29.1) | 21.3 (14.0–26.9) | 19.0 (13.7–27.2) |
| Endometrial thickness (mm) | 6.6 (3.0) | 6.5 (2.2) | 7.2 (2.5) | 7.0 (2.6) |
| Polycystic ovaries according to modified Rotterdam criteria: N (%) | 179 (95.7) | 43 (97.7) | 46 (97.9) | 43 (95.6) |
| Serum results | ||||
| AMH (ng/ml) | 8.1 (6.4) | 9.1 (5.2) | 8.8 (6.0) | 8.9 (5.5) |
| Testosterone (ng/dl)a | 51.4 (35.4–72.0) | 48.3 (36.8–72.9) | 52.7 (35.3–70.8) | 53.5 (38.4–79.0) |
| SHBG (nmol/liter)a | 22.3 (17.6–31.5) | 26.1 (20.2–34.4) | 28.9 (21.5–38.1) | 25.7 (18.6–40.4) |
| Free androgen indexa | 7.7 (5.0–12.0) | 6.0 (3.8–12.2) | 6.4 (4.0–8.9) | 6.7 (4.9–10.3) |
| Estradiol (pg/ml) | 52.6 (39.7) | 55.1 (41.4) | 54.2 (39.6) | 59.0 (47.7) |
| Progesterone (ng/ml) | 1.3 (2.4) | 1.5 (2.5) | 1.4 (2.4) | 1.5 (3.3) |
| Cholesterol (mg/dl) | 181.4 (39.5) | 187.9 (31.4) | 184.9 (34.8) | 184.9 (33.2) |
| HDL (mg/dl)a | 37.0 (30.0–44.0) | 44.0 (39.0–52.0) | 42.0 (36.0–48.0) | 43.0 (39.0–49.0) |
| LDL (mg/dl) | 123.1 (33.2) | 116.2 (26.7) | 114.1 (31.3) | 113.1 (31.1) |
| Triglycerides (mg/dl)a | 116.0 (83.0–167.0) | 117.0 (90.0–161.0) | 131.5 (99.5–174.5) | 120.0 (86.0–165.0) |
| Fasting glucose (mg/dl) | 87.0 (15.1) | 87.1 (9.6) | 86.8 (8.9) | 90.1 (15.0) |
Abbreviations: BP, blood pressure; NA, not applicable to the PPCOS II trial as there was no preconception intervention before ovulation induction; OWL PCOS, Treatment of Hyperandrogenism Versus Insulin Resistance in Infertile Polycystic Ovary Syndrome; PPCOS, Pregnancy in Polycystic Ovary Syndrome.
Median (25th–75th percentile) is reported.
Table 3.
Primary Outcome of Live Birth and Secondary Outcomes From the PPCOS II Clomiphene and OWL PCOS Selected Cohorts
| Outcome | PPCOS II No./Total No. (%) | OWL PCOS: OCP |
OWL PCOS: Lifestyle |
OWL PCOS: Combined |
|||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| No./Total No. (%) | RR (95% CI) Compared to PPCOSII | P Valuee | No./Total No. (%) | RR (95% CI) Compared to PPCOSII | P Valuee | No./Total No. (%) | RR (95% CI) Compared to PPCOSII | P Valuee | |||
| Ovulation | |||||||||||
| Ovulation (total number of ovulations/total treatment cycles)a | 277/619 (44.7) | 67/149 (45.0) | 1.0 (0.8–1.3) | .96 | 80/129 (62.0) | 1.4 (1.1–1.7) | 0.003 | 83/129 (64.3) | 1.4 (1.2–1.8) | <.001 | |
| Pregnancy | |||||||||||
| Conception (serum hCG level >10 mIU/ml) | 38/187 (20.3) | 7/47 (14.9) | 0.7 (0.3–1.5) | .41 | 15/48 (31.3) | 1.5 (0.9–2.6) | 0.10 | 14/47 (29.8) | 1.5 (0.9–2.5) | .15 | |
| Clinical pregnancy (fetal heart motion visualized on ultrasound) | 25/187 (13.4) | 6/47 (12.8) | 1.0 (0.4–2.2) | .91 | 12/48 (25.0) | 1.9 (1.0–3.4) | 0.04 | 13/47 (27.7) | 2.1 (1.1–3.7) | .02 | |
| Pregnancy loss | |||||||||||
| Pregnancy loss among subjects who conceivedb | 18/38 (47.4) | 3/7 (42.9) | 0.9 (0.4–2.3) | .83 | 3/15 (20.0) | 0.4 (0.1–1.2) | 0.11 | 2/14 (14.3) | 0.3 (0.1–1.1) | .08 | |
| Birth outcomes | |||||||||||
| Live birth | 19/187 (10.2) | 4/47 (8.5) | 0.8 (0.3–2.3) | .74 | 12/48 (25.0) | 2.5 (1.3–4.7) | .01 | 12/47 (25.5) | 2.5 (1.3–4.8) | .01 | |
| Infant birth weight (grams)c,d | 3198 (916) [19] | 3487 (284) [4] | 290 (−519 to 1098) | .47 | 3090 (629) [10] | −107 (−682 to 467) | .71 | 3409 (505) [12] | 212 (−330 to 753) | .44 | |
| Duration of pregnancy (weeks)c,d | 37.4 (4.5) [19] | 38.8 (0.5) [4] | 1.3 (−2.3 to 5.0) | .47 | 38.2 (2.7) [10] | 0.8 (−1.8 to 3.4) | .55 | 39.3 (1.3) [12] | 1.8 (−0.6 to 4.3) | .14 | |
| Fecundity per ovulated patient | |||||||||||
| Live birth | 19/135 (14.1) | 4/34 (11.8) | 0.8 (0.3–2.3) | .73 | 12/34 (35.3) | 2.5 (1.4–4.6) | .004 | 12/36 (33.3) | 2.4 (1.3–4.4) | .01 | |
Abbreviations: CI, confidence interval; hCG, human chorionic gonadotropin; OCP, oral contraceptives; OWL PCOS, Treatment of Hyperandrogenism Versus Insulin Resistance in Infertile Polycystic Ovary Syndrome; RR, risk ratio.
Cycles for which ovulation status could not be determined (no progesterone value/no pregnancy) are excluded.
Pregnancy loss does not include twin pregnancies that resulted in a singleton live birth.
Data are presented as mean (sd) [n] and difference in means (95% CI).
Calculations based on singleton deliveries that resulted in live birth, ranging in gestational age of 22–41 weeks; two twin deliveries from OWL PCOS are omitted.
Log-binomial model for categorical data and linear regression model for continuous data.
In regards to per-cycle results, only the lifestyle group in OWL PCOS had a significantly improved chance of pregnancy and live birth in the first ovulation induction cycle compared to the first ovulation induction cycle in PPCOS II (Table 4). The ovulation rate was significantly improved in the first, second, and fourth cycles of ovulation induction in the combined group from OWL PCOS compared to treatment with clomiphene in PPCOS II. Further, there was a significantly improved chance of ovulation in the first and fourth ovulation induction cycles with lifestyle alone in the OWL PCOS study compared to treatment with clomiphene in PPCOS II. No per-cycle differences were noted between the OCP group in OWL PCOS and treatment with clomiphene in PPCOS II.
Table 4.
Per-Cycle Results of Ovulation Induction in PPCOS II Clomiphene and OWL PCOS Selected Cohorts
| Outcome | PPCOS II No./Total No. (%) | OWL PCOS: OCP |
OWL PCOS: Lifestyle |
OWL PCOS: Combined |
||||||
|---|---|---|---|---|---|---|---|---|---|---|
| No./Total No. (%) | RR (95% CI) | P Valuea | No./Total No. (%) | RR (95% CI) | P Valuea | No./Total No. (%) | RR (95% CI) | P Valuea | ||
| Primary outcome: live birth per treatment cycle | ||||||||||
| Cycle 1 | 3/187 (1.6) | 1/47 (2.1) | 1.3 (0.1–12.5) | .80 | 5/48 (10.4) | 6.5 (1.6–26.2) | .01 | 3/47 (6.4) | 4.0 (0.8–19.1) | .08 |
| Cycle 2 | 8/184 (4.3) | 1/46 (2.2) | 0.5 (0.1–3.9) | .51 | 4/43 (9.3) | 2.1 (0.7–6.8) | .20 | 5/44 (11.4) | 2.6 (0.9–7.6) | .08 |
| Cycle 3 | 3/176 (1.7) | 1/45 (2.2) | 1.3 (0.1–12.2) | .82 | 2/39 (5.1) | 3.0 (0.5–17.4) | .22 | 1/39 (2.6) | 1.5 (0.2–14.1) | .72 |
| Cycle 4 | 5/173 (2.9) | 1/44 (2.3) | 0.8 (0.1–6.6) | .82 | 1/37 (2.7) | 0.9 (0.1–7.8) | .95 | 3/38 (7.9) | 2.7 (0.7–10.9) | .16 |
| Secondary outcome: clinical pregnancy (fetal heart motion on ultrasound) per treatment cycle | ||||||||||
| Cycle 1 | 5/187 (2.7) | 2/47 (4.3) | 1.6 (0.3–7.9) | .57 | 5/48 (10.4) | 3.9 (1.2–12.9) | .03 | 3/47 (6.4) | 2.4 (0.6–9.6) | .22 |
| Cycle 2 | 9/182 (4.9) | 2/45 (4.4) | 0.9 (0.2–4.0) | .89 | 4/43 (9.3) | 1.9 (0.6–5.8) | .27 | 6/44 (13.6) | 2.8 (1.0–7.3) | .04 |
| Cycle 3 | 5/173 (2.9) | 1/43 (2.3) | 0.8 (0.1–6.7) | .84 | 2/39 (5.1) | 1.8 (0.4–8.8) | .48 | 1/38 (2.6) | 0.9 (0.1–7.6) | .93 |
| Cycle 4 | 6/168 (3.6) | 1/42 (2.4) | 0.7 (0.1–5.4) | .70 | 1/37 (2.7) | 0.8 (0.1–6.1) | .79 | 3/37 (8.1) | 2.3 (0.6–8.7) | .23 |
| Secondary outcome: conception (serum hCG level >10 mIU/ml) per treatment cycle | ||||||||||
| Cycle 1 | 10/187 (5.3) | 3/47 (6.4) | 1.2 (0.3–4.2) | .78 | 5/48 (10.4) | 1.9 (0.7–5.4) | .20 | 4/47 (8.5) | 1.6 (0.5–4.9) | .41 |
| Cycle 2 | 11/177 (6.2) | 2/44 (4.5) | 0.7 (0.2–3.2) | .68 | 6/43 (14.0) | 2.2 (0.9–5.7) | .09 | 6/43 (14.0) | 2.2 (0.9–5.7) | .09 |
| Cycle 3 | 8/166 (4.8) | 1/42 (2.4) | 0.5 (0.1–3.8) | .50 | 3/37 (8.1) | 1.7 (0.5–6.0) | .43 | 1/37 (2.7) | 0.6 (0.1–4.3) | .58 |
| Cycle 4 | 9/158 (5.7) | 1/41 (2.4) | 0.4 (0.1–3.3) | .41 | 1/34 (2.9) | 0.5 (0.1–3.9) | .52 | 3/36 (8.3) | 1.5 (0.4–5.1) | .55 |
| Secondary outcome: ovulation per treatment cycle | ||||||||||
| Cycle 1 | 68/181 (37.6) | 15/41 (36.6) | 1.0 (0.6–1.5) | .91 | 27/42 (64.3) | 1.7 (1.3–2.3) | <.001 | 23/40 (57.5) | 1.5 (1.1–2.1) | .01 |
| Cycle 2 | 78/164 (47.6) | 21/39 (53.8) | 1.1 (0.8–1.6) | .46 | 19/35 (54.3) | 1.1 (0.8–1.6) | .45 | 25/35 (71.4) | 1.5 (1.2–2.0) | .003 |
| Cycle 3 | 74/146 (50.7) | 14/36 (38.9) | 0.8 (0.5–1.2) | .24 | 17/29 (58.6) | 1.2 (0.8–1.6) | .41 | 19/29 (65.5) | 1.3 (0.9–1.8) | .10 |
| Cycle 4 | 57/128 (44.5) | 17/33 (51.5) | 1.2 (0.8–1.7) | .46 | 17/23 (73.9) | 1.7 (1.2–2.3) | .001 | 16/25 (64.0) | 1.4 (1.0–2.0) | .04 |
Abbreviations: CI, confidence interval; hCG, human chorionic gonadotropin; OCP, oral contraceptives; OWL PCOS, Treatment of Hyperandrogenism Versus Insulin Resistance in Infertile Polycystic Ovary Syndrome; RR, risk ratio.
Log-binomial model.
Discussion
In this post hoc comparison of two randomized, concurrently performed clinical trials to treat infertility in women with PCOS, we found that pretreatment lifestyle modification for weight loss, with or without concurrent oral contraceptive therapy, was associated with a significant improvement in the rate of ovulation and an even greater increase in live birth rate than immediate fertility treatment with clomiphene. Further, ovulation and live birth rates were nearly identical between pretreatment with oral contraceptives vs immediate treatment with clomiphene, suggesting that there is little fertility benefit to pretreatment with hormonal suppression, alone or in combination with lifestyle modification. The weight maintenance in the oral contraceptive group further supports the conclusion that it was some aspect(s) of the weight loss intervention that led to the improved outcomes.
These data are relevant because we are unaware of any published randomized trials of lifestyle intervention promoting weight loss vs immediate treatment in women with PCOS (14). Patients and practitioners require a justification to delay immediate fertility treatment in the current practice environment with a greater tendency to immediate treatment and acceleration to expensive and invasive infertility therapies such as in vitro fertilization. Although there are data from at least one prospective randomized trial documenting a benefit to OCP pretreatment before clomiphene, (22), this was a single-center study done in women who were initially clomiphene resistant. This was not replicated in our study, which did not require clomiphene resistance as an entry criterion.
The strengths of the present study include the similarity of protocols, with a near uniform definition of PCOS and exclusion of other infertility disorders, the common means of administering and monitoring clomiphene, and the tracking of all pregnancy outcomes. Other strengths included the multicenter design, the concurrent and overlapping study sites, and that the similarity of metabolic and reproductive phenotypes of the population in both trials. There were, however, slight differences between the protocols, most prominently the focus on overweight/body mass index women with PCOS in the OWL PCOS, which we compensated for in the analysis by excluding women from PPCOS II who fell outside of these weight brackets. We allowed five cycles of ovulation induction in PPCOS II and only four in OWL PCOS, which we again compensated for by only including the first four cycles of both studies. Another difference was the option not to induce menses at the start of infertility treatment in OWL PCOS, which may have contributed to the markedly elevated chance of live birth in the first cycle of ovulation induction with clomiphene in the Lifestyle group (and trend in the combined group) vs immediate treatment with clomiphene in PPCOS II. We have previously shown that progestin exposure before ovulation is associated with lower pregnancy rates (23). Clomiphene was given open label in OWL PCOS and blinded in PPCOS II. Finally, the greatest discrepancy is the difference in sample size between the two study cohorts in this post hoc analysis where type I and type II errors are more likely. This may have contributed to the variable per-cycle ovulatory benefits of preconception weight loss compared to immediate treatment where some cycles lacked a statistically improved benefit. We note that with the larger sample sizes of our PPCOS I and II trials, we showed no time-related increase or decrease in the ovulation rate in response to clomiphene up to six cycles (17, 24).
One of the striking findings of this study is the marked improvement in live birth rates, out of proportion to the more modest, but still significant improvement, in ovulation rates, suggesting yet again that improving quality of ovulation is as important as improving the frequency of ovulation (24). Pregnancy loss rates were similar between the treatment groups in this study, so the differences in live birth rates were not completely explained by the trends favoring reduced pregnancy loss in the lifestyle modification groups. Weight loss is the simplest explanation for the improved ovulation and live birth rates, though there are concurrent improvements in a variety of metabolic parameters and to a lesser extent reproductive parameters with weight loss (17, 25, 26) that have also been implicated in poor reproductive outcomes. Thus, the mechanism(s) for improved fecundity per ovulation remain speculative, but the findings from the OCP arm in the OWL PCOS study support weight loss as a target goal of preconception treatment. To modify our opening analogy (with apologies to Andrew Marvell) (2), lightening time's winged chariot's load may allow a longer flight and greater chance of reaching a final destination of a healthy baby.
Our research holds significant implications for current practice and supports the concept of delaying fertility treatment to pursue lifestyle modification in overweight/obese women with PCOS. It provides momentum to test this concept more completely and prospectively in properly designed and adequately powered multicenter studies to generate level I evidence for the practice. Future studies may also want to use other ovulation induction agents in the infertility treatment phase such as low-dose gonadotropin (27) or letrozole, (28) which tend to have greater success rates combined with comparable rates of multiple pregnancy and congenital anomalies as clomiphene.
Acknowledgments
We acknowledge the outstanding leadership of the Treatment of Hyperandrogenism Versus Insulin Resistance in Infertile Polycystic Ovary Syndrome (OWL PCOS) study coordinators Patsy Rawa (Penn State) and Karen Lecks (University of Pennsylvania), the collaboration with the Core Ligand Lab at the University of Virginia under the leadership of Dan Haisenleider, PhD (University of Virginia), for their expertise in running our assays, and the Reproductive Medicine Network for providing us with the data from the Pregnancy in Polycystic Ovary Syndrome II study to perform our subanalysis of ovulation and live birth rates in women with overweight/obesity.
This project was supported by the Eunice Kennedy Shriver National Institutes of Child Health and Human Development (NICHD), National Center for Research Resources, and the National Center for Advancing Translational Sciences at the National Institutes of Health, through Grants R01 HD056510 (to R.S.L.), U10 HD27049 (to C.C.), U10 HD38992 (to R.S.L.), U10HD055925 (to H.Z.), U10 HD39005 (to M.P.D.), U10 HD38998 (to W.D.S.), U10 HD055936 (to G.M.C.), and U10 HD055944 (to P.R.C.), UL1 TR000127 (Penn State Clinical and Translational Institute), and U54 HD29834 (UVA Core Ligand Assay Core of the Specialized Cooperative Centers Program in Reproduction of the NICHD). The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.
For the Pregnancy in Polycystic Ovary Syndrome II study: In addition to the authors and personnel cite previously, other members of the National Institutes of Child Health and Human Development Reproductive Medicine Network were as follows. Penn State College of Medicine, Hershey: C. Bartlebaugh, S. Eyer, J. Ober, RN, B. Scheetz, J. Moessner, R. Ladda; University of Pennsylvania: P. Snyder, MD, K. Lecks, L. Martino, RN, R. Marunich, J. Spitzer, MS; SUNY Upstate Medical Center, Syracuse: J.C. Trussell; University of Texas Health Science Center at San Antonio: R. Bryzski, C. Easton, A. Hernandez, M. Leija, D. Pierce, R. Robinson; Wayne State University: A. Awonuga, L. Cedo, A. Cline, K. Collins, E. Pucscheck, M. Singh, M. Yoscovits; University of Colorado: A. Comfort, M. Crow; University of Vermont: A. Hohmann, S. Mallette; University of Michigan: D. Ohl, M. Ringbloom, J. Tang; University of Alabama Birmingham: S. Mason; Carolinas Medical Center: N. DiMaria; Virginia Commonwealth University: M. Rhea; Stanford University Medical Center: K. Turner; Yale University: D. DelBasso, Y. Li, R. Makuch, P. Patrizio, L. Sakai, L. Scahill, H. Taylor, T. Thomas, S. Tsang, M. Zhang; Eunice Kennedy Shriver National Institute of Child Health and Human Development; C. Lamar, L. DePaolo.
Trial Registration: clinicaltrials.gov: Treatment of Hyperandrogenism Versus Insulin Resistance in Infertile Polycystic Ovary Syndrome (PCOS) Women: NCT00704912 and PPCOS II: NCT00719186. Disclosure Summary: R.S.L. reports consulting fees from Euro screen, Astra Zeneca, Clarus Therapeutics, Takeda, and Kindex, and research funding from Ferring and Astra Zeneca. S.J.E. and W.M. reports research funding from AbbVie. A.D. reports consulting fees from JDS Therapeutics. A.R.K. reports ownership of Merck stock. D.B.S. reports consulting fees from BAROnova, EnteroMedics, and Ethicon. The other investigators report no disclosures.
Footnotes
- OWL PCOS
- Treatment of Hyperandrogenism Versus Insulin Resistance in Infertile Polycystic Ovary Syndrome
- PPCOS
- Pregnancy in Polycystic Ovary Syndrome.
References
- 1. Nelson SM, Telfer EE, Anderson RA. The ageing ovary and uterus: new biological insights. Human Reprod Update. 2013;19:67–83. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Marvell A. To His Coy Mistress. Unknown, early 1650's; http://www.poetryfoundation.org/poems-and-poets/poems/detail/44688 Accessed April 19, 2016.
- 3. Bolumar F, Olsen J, Rebagliato M, Saez-Lloret I, Bisanti L. Body mass index and delayed conception: a European multicenter study on infertility and subfecundity. Am J Epidemiol. 2000;151:1072–10791. [DOI] [PubMed] [Google Scholar]
- 4. Metwally M, Ong KJ, Ledger WL, Li TC. Does high body mass index increase the risk of miscarriage after spontaneous and assisted conception? A meta-analysis of the evidence. Fertil Steril. 2008;90:714–726. [DOI] [PubMed] [Google Scholar]
- 5. Galtier-Dereure F, Boegner C, Bringer J. Obesity and pregnancy: complications and cost. Am J Clin Nutr. 2000;71:1242S–1248S. [DOI] [PubMed] [Google Scholar]
- 6. Balen AH, Anderson RA, Policy, Practice Committee of the BFS. Impact of obesity on female reproductive health: British Fertility Society, Policy and Practice Guidelines. Hum Fertil (Camb). 2007;10:195–206. [DOI] [PubMed] [Google Scholar]
- 7. Rosenberg TJ, Garbers S, Lipkind H, Chiasson MA. Maternal obesity and diabetes as risk factors for adverse pregnancy outcomes: differences among 4 racial/ethnic groups. Am J Public Health. 2005;95:1545–1551. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Baghdadi LR, Abu Hashim H, Amer SA, et al. Impact of obesity on reproductive outcomes after ovarian ablative therapy in PCOS: a collaborative meta-analysis. Reprod Biomed Online. 2012;25:227–241. [DOI] [PubMed] [Google Scholar]
- 9. Joham AE, Ranasinha S, Zoungas S, Moran L, Teede HJ. Gestational diabetes and type 2 diabetes in reproductive-aged women with polycystic ovary syndrome. J Clin Endocrinol Metab. 2014;99:E447–E452. [DOI] [PubMed] [Google Scholar]
- 10. Consensus on infertility treatment related to polycystic ovary syndrome. Fertil Steril. 2008;89:505–522. [DOI] [PubMed] [Google Scholar]
- 11. Consensus on infertility treatment related to polycystic ovary syndrome. Hum Reprod. 2008;23:462–477. [DOI] [PubMed] [Google Scholar]
- 12. Legro RS, Arslanian SA, Ehrmann DA, et al. Diagnosis and treatment of polycystic ovary syndrome: an Endocrine Society clinical practice guideline. J Clin Endocrinol Metab. 2013;98:4565–4592. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Anderson K, Norman RJ, Middleton P. Preconception lifestyle advice for people with subfertility. Cochrane Database Syst Rev. 2010:CD008189. [DOI] [PubMed] [Google Scholar]
- 14. Sim KA, Partridge SR, Sainsbury A. Does weight loss in overweight or obese women improve fertility treatment outcomes? A systematic review. Obes Rev. 2014;15:839–850. [DOI] [PubMed] [Google Scholar]
- 15. Harbin Consensus Conference Workshop Group, Conference Chairs, Legro RS, Wu X, Scientific C, Barnhart KT, Farquhar C, Fauser BC, Mol B. Improving the reporting of clinical trials of infertility treatments (IMPRINT): modifying the CONSORT. Hum Reprod. 2014;29:2075–2082. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Harbin Consensus Conference Workshop Group. Improving the Reporting of Clinical Trials of Infertility Treatments (IMPRINT): modifying the CONSORT statement. Fertil Steril. 2014;102:952–959. [DOI] [PubMed] [Google Scholar]
- 17. Legro RS, Dodson WC, Kris-Etherton PM, et al. Randomized controlled trial of preconception interventions in infertile women with polycystic ovary syndrome. J Clin Endocrinol Metab. 2015;100:4048–4058. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Legro RS, Brzyski RG, Diamond MP, et al. Letrozole versus clomiphene for infertility in the polycystic ovary syndrome. N Engl J Med. 2014;371:119–129. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Legro RS, Brzyski RG, Diamond MP, et al. , National Institute of Child Health and Human Development Reproductive Medicine N. The Pregnancy in Polycystic Ovary Syndrome II study: baseline characteristics and effects of obesity from a multicenter randomized clinical trial. Fertil Steril. 2014;101:258–269 e258. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Legro RS, Myers ER, Barnhart HX, et al. The Pregnancy in Polycystic Ovary Syndrome study: baseline characteristics of the randomized cohort including racial effects. Fertil Steril. 2006;86:914–933. [DOI] [PubMed] [Google Scholar]
- 21. Friedewald WT, Levy RI, Fredrickson DS. Estimation of the concentration of low-density lipoprotein cholesterol in plasma, without use of the preparative ultracentrifuge. Clin Chem. 1972;18:499–502. [PubMed] [Google Scholar]
- 22. Branigan EF, Estes MA. A randomized clinical trial of treatment of clomiphene citrate-resistant anovulation with the use of oral contraceptive pill suppression and repeat clomiphene citrate treatment. Am J Obstet Gynecol. 2003;188:1424–1428; discussion 1429–1430. [DOI] [PubMed] [Google Scholar]
- 23. Diamond MP, Kruger M, Santoro N, et al. Endometrial shedding effect on conception and live birth in women with polycystic ovary syndrome. Obstet Gynecol. 2012;119:902–908. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Legro RS, Barnhart HX, Schlaff WD, et al. Clomiphene, metformin, or both for infertility in the polycystic ovary syndrome. N Engl J Med. 2007;356:551–566. [DOI] [PubMed] [Google Scholar]
- 25. Rochester D, Jain A, Polotsky AJ, et al. Partial recovery of luteal function after bariatric surgery in obese women. Fertil Steril. 2009;92:1410–1415. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Legro RS, Dodson WC, Gnatuk CL, et al. Effects of gastric bypass surgery on female reproductive function. J Clin Endocrinol Metab. 2012;97:4540–4548. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. Homburg R, Hendriks ML, Konig TE, et al. Clomifene citrate or low-dose FSH for the first-line treatment of infertile women with anovulation associated with polycystic ovary syndrome: a prospective randomized multinational study. Hum Reprod. 2012;27:468–473. [DOI] [PubMed] [Google Scholar]
- 28. Franik S, Kremer JA, Nelen WL, Farquhar C. Aromatase inhibitors for subfertile women with polycystic ovary syndrome. Cochrane Database Syst Rev. 2014;2:CD010287. [DOI] [PubMed] [Google Scholar]