Advertisement

Occurrence of menses or pregnancy after cessation of a continuous oral contraceptive

      Objective

      To evaluate the time to return to spontaneous menses in women after 1 year of daily continuous levonorgestrel (LNG) 90 μg/ethinyl E2 (EE) 20 μg.

      Design

      Observational study.

      Setting

      Gynecologic and primary care practices.

      Patient(s)

      Women aged 18–49 years with a history of regular menstrual cycles. After participation in an open-label, continuous oral contraceptive (OC) trial for at least 6 months, participants agreed to enroll in a separate study of the return to menses or pregnancy.

      Intervention(s)

      None.

      Main Outcome Measure(s)

      Time to return to spontaneous menses or pregnancy.

      Result(s)

      The 198 subjects had a mean age of 30.4 ± 6.6 years with 72% white, 13% Hispanic, and 7% African American. The mean duration of continuous LNG/EE treatment before enrollment was 349 ± 41 days. Of the 187 (94%) subjects who completed this study, 181 returned to spontaneous menses and 4 became pregnant within 90 days after the last dose of LNG 90 μg/EE 20 μg. The median time to return to menses in the completer population was 32 days, and the incidence of spontaneous menses or pregnancy at day ≤90 was 98.9%. The duration of amenorrhea during continuous LNG/ EE use before stopping treatment was unrelated to the time to the return to menses.

      Conclusions

      Spontaneous menses or pregnancy occurred in 98.9% of women after cessation of continuous LNG/EE.

      Key Words

      Until several years ago, available oral contraceptive (OC) regimens were limited to the traditional 28-day cycle with a monthly withdrawal bleed. This cyclic regimen was designed to resemble the menstrual cycle and thereby increase acceptability. However, the monthly bleeding with traditional cyclic OCs is an induced withdrawal bleed and is not a “natural” menstrual cycle. Recently, several investigators challenged the need for a monthly withdrawal bleed (
      • Coutinho E.M.
      Is menstruation normal? Suppression of the menstrual cycle in clinical practice.
      ,
      • Kaunitz A.
      Menstruation: choosing whether … and when.
      ,
      • Edelman A.
      Menstrual nirvana: amenorrhea through the use of continuous contraceptives.
      ), and surveys demonstrate that some women would prefer a reduced frequency of menses (
      • Sulak P.J.
      • Kuehl T.J.
      • Ortiz M.
      • Shull B.L.
      Acceptance of altering the standard 21-day/7-day oral contraceptive regimen to delay menses and reduce hormone withdrawal symptoms.
      ,
      • Andrist L.C.
      • Arias R.D.
      • Nucatola D.
      • Kaunitz A.M.
      • Musselman B.L.
      • Reiter S.
      • et al.
      Women's and providers' attitudes toward menstrual suppression with extended use of oral contraceptives.
      ). As yet, few large well-designed studies have evaluated the safety and efficacy of extended and continuous use OC regimens.
      Recently, two large, multicenter, phase 3 studies investigated the safety and contraceptive efficacy of continuous levonorgestrel (LNG) 90 μg and ethinyl E2 (EE) 20 μg (continuous LNG/EE, Lybrel). In these trials, Lybrel (Wyeth Pharmaceuticals, Collegeville, PA) was taken daily, without a pill-free or placebo interval for up to 1 year (
      • Archer D.F.
      • Jensen J.T.
      • Johnson J.V.
      • Borisute H.
      • Grubb G.S.
      • Constantine G.D.
      Evaluation of a continuous regimen of levonorgestrel/ethinyl estrodiol: phase 3 study results.
      ,
      • Teichmann A.T.
      • Kluft C.
      • Grubb G.
      • Constantine G.
      • Spielmann D.
      Comparative trial of continuous-use and 21-day cyclic levonorgestrel and ethinyl estradiol oral contraceptive.
      ) with good contraceptive efficacy and reassuring safety results. Menses inhibition with the elimination of regular menstrual bleeding, a decrease in the number of total bleeding or spotting days, and the development of amenorrhea was observed with the regimen in these trials provided benefits to women (
      • Archer D.F.
      • Jensen J.T.
      • Johnson J.V.
      • Borisute H.
      • Grubb G.S.
      • Constantine G.D.
      Evaluation of a continuous regimen of levonorgestrel/ethinyl estrodiol: phase 3 study results.
      ).
      We conducted a follow-up study to assess the incidence of and time to return to spontaneous menses or pregnancy in women who had taken continuous LNG/EE 90 μg/20 μg for 6 months to 1 year. Our objective was to address the effect, if any, of prolonged menstrual inhibition on the spontaneous menstrual cycle and associated return to fertility as spontaneous menses suggests a functioning hypothalamic-pituitary-ovarian axis with ovulation (
      • Malcolm C.E.
      • Cumming D.C.
      Does anovulation exist in eumenorrheic women?.
      ).

      Materials and methods

      This phase 3, multicenter, observational study began immediately after the completion of participation in the phase 3 efficacy and safety North American trial of daily continuous LNG 90 μg/EE 20 μg conducted in healthy women aged 18–49 years. All of the 2,134 subjects in the phase 3 efficacy and safety trial reported regular menstrual cycles of 21–35 days before enrollment. Those completing at least 168 days (6 months) of treatment were eligible for enrollment in this follow-up study.
      Women using any nonhormonal form of contraception (usually male condoms) or seeking pregnancy were eligible. Women initiating intrauterine or hormonal contraception and women older than 35 years who smoked >15 cigarettes per day were excluded. We did not record detailed data on the fertility intentions of participants at enrollment. This study was designed and performed according to the guidelines for Good Clinical Practice and conducted in accordance with the Declaration of Helsinki. An institutional review board (IRB) approved the protocol at each site.
      The primary objectives were to evaluate the incidence and timing of return to spontaneous menses or pregnancy after cessation of LNG/EE continuous OC. The 50 centers in the study contributed a median of 3.5 subjects per site (range 1–18). There was no intervention in this observational study. We also collected data on safety by using adverse event reporting.
      At the prestudy visit, which was the final visit of the preceding trial, eligible subjects provided informed consent. Participants were followed until they reported the return to menses or pregnancy, which occurred in all but two subjects within 90 days. Subjects performed a home urine pregnancy test once during the last 10 days of each 30-day interval if no spontaneous menses occurred and recorded the results on diary cards. Spontaneous menses was defined as at least 2 consecutive days of bleeding requiring sanitary protection that started no sooner than day 13 after stopping treatment in the previous trial. Safety assessments were based on a review of adverse events during each telephone contact and at the poststudy visit. Subjects recorded adverse events, bleeding or spotting, and any concomitant treatments on daily diary cards. Because there was no active treatment, adverse events that were not present before entry into the study or that worsened during the study were considered to be study-emergent, rather than treatment-emergent. Subjects who reported a return to menses or pregnancy were asked to return completed diary cards to the investigator by tracked mail (e.g., Express Mail, Federal Express).
      A poststudy visit was scheduled for subjects who did not report menses or pregnancy by day 90. At this visit, diary cards were collected, a serum pregnancy test (β-hCG) was performed, and adverse events were assessed.
      This analysis includes an intent-to-treat population comprising all subjects enrolled in the study, and a completer population comprising those subjects who spontaneously returned to menses, became pregnant, or completed diary records for 90 days after their last study medication intake in the preceding trial. The time to return to spontaneous menses was estimated using the Kaplan-Meier method (
      • Lee E.T.
      Statistical methods for survival data analysis.
      ).
      To investigate the relationship between the length of amenorrhea before the end of the treatment and the time to return to menses (or time to pregnancy, if there was no menses), we constructed regression models by using the covariates of body mass index (BMI), age, length of amenorrhea before the end of treatment, and a quadratic term, the square of the length of amenorrhea before the end of treatment (
      • Neter J.
      • Kutner M.H.
      • Wasserman W.
      • Nachtsheim C.J.
      Applied linear statistical models.
      ).

      Results

       Subject Disposition

      This study was conducted from January 2004 to January 2005 (trial registered with ClinicalTrials.gov, identifier: NCT00245921). There were 1,393 subjects who completed at least 6 months of the preceding study. Of these, 198 (14.2%) subjects enrolled in this study. Most of these subjects (162; 81.8%) had completed 12 months of treatment and 36 (18.2%) had 6 to 12 months of treatment. The mean duration of LNG/EE treatment before enrollment was 349 ± 41 days.
      There were 187 (94.4%) subjects who completed the study. Of those who did not complete the study, 7 (3.5%) were lost to follow-up and 4 (2.0%) discontinued because of protocol violations. A summary of demographic and baseline characteristics of the study population (intent-to-treat) is presented in Table 1.
      Table 1Summary of demographic and baseline characteristics for the intent-to-treat population.
      Treatment group, No study medication
      Characteristicn = 198
      Age (y)
       Mean ± SD30.4 ± 6.6
      Race n, (%)
       Asian7 (3.5)
       African American14 (7.1)
       Hispanic26 (13.1)
       Other8 (4.0)
       White143 (72.2)
      Duration on study (days)
       Mean ± SD57.5 ± 28.0
      Study completed n, (%)
       No11 (6)
       Yes187 (94)

       Return to Menses or Pregnancy

      Of the 187 subjects who completed the study, 185 (98.9%) returned to spontaneous menses (n = 181) or became pregnant (n = 4) within 90 days after stopping continuous LNG/EE. The number and percentage at each time interval are provided for the completer population in Table 2. Similar rates of return to spontaneous menses or pregnancy were observed in the intent-to-treat population. Figure 1 presents the estimated probability of returning to spontaneous menses or becoming pregnant as a function of time (completers). For the two subjects who did not experience menses during the study interval, one reported menses on day 124 and the other at approximately 2 months after the study ended. Results in the intent-to-treat population were similar to those observed in the completer population.
      Table 2Unadjusted return to spontaneous menses or pregnancy rates in the completer population.
      Menses or pregnancy occurrence (days after treatment)Incidence/total population (%)95% Exact CICumulative incidence/total population (%)95% Exact CI
      1–3075/187 (40.1)(33.0, 47.5)75/187 (40.1)(33.0, 47.5)
      31–60102/187 (54.5)(47.1, 61.8)177/187 (94.7)(90.4, 97.4)
      61–908/187 (4.3)(1.9, 8.3)185/187 (98.9)(96.2, 99.9)
      >90
      One subject returned to spontaneous menses on day 124 and one subject returned to menses approximately 2 months after the study after starting a cyclic oral contraceptive (OC) (Alesse).
      2/187 (1.0)(0.1, 3.8)187/187 (100.0)(98.1, 100.0)
      Note: CI = confidence interval.
      a One subject returned to spontaneous menses on day 124 and one subject returned to menses approximately 2 months after the study after starting a cyclic oral contraceptive (OC) (Alesse).
      Figure thumbnail gr1
      Figure 1Time to return to spontaneous menses or pregnancy: Kaplan-Meier survival function estimates in the completer population.

       Time to Return to Spontaneous Menses

      Table 3 shows the estimated 25th, 50th, and 75th percentiles of time (in days) to return to spontaneous menses for the completer population. The median time to return to menses was 32 days (95% confidence interval, 31–33 days) after the last study medication intake in the preceding trial. Similar results were observed in the intent-to-treat population.
      Table 3Days to return to spontaneous menses: completer population.
      PercentileEstimate (days)95% CI
      7536(35, 38)
      5032(31, 33)
      2529(27, 29)
      Note: CI = confidence interval. Subjects who became pregnant before onset of menses were included.
      The length of amenorrhea before the end of treatment was unrelated to the time to return of menses. The regression model for the length of last amenorrhea and the time to return to menses (Fig. 2) found that the full model (including both linear and quadratic terms, and demographic variables) yielded an R2 term (the proportion of variation explained by the model) of R2 = 0.1336 (i.e., 13.4% of variation explained). A model with only a quadratic term representing the length of the last amenorrhea interval (squared) accounted for 11.7% of the variance in the time to return to menses.
      Figure thumbnail gr2
      Figure 2Relationship between length of last amenorrhea interval (in preceding study) and time to return to menses (in the current study) using a quadratic regression model.

       Safety

      No treatment was administered in the current study, but 74 (37.4%) subjects reported study-emergent adverse events. However, there were no deaths, no serious adverse events, and no safety-related discontinuations or other clinically important adverse events. Of the four pregnancies that occurred before the return to menses, one subject was lost to follow-up, three subjects had full-term, uncomplicated pregnancies and delivered healthy infants.

      Discussion

      To our knowledge, this is the first study to assess the return of a spontaneous menses or natural menstrual function after a continuous combined OC regimen. The return to menses or pregnancy occurred in all but 1 of 187 subjects, and usually occurred within about 1 month after OC discontinuation with a median time to return to menses in the completer population of 32 days after stopping continuous LNG/EE. These results support earlier phase 2 trial observations of a rapid return to ovulation after 3 months of use of continuous LNG/EE. In that phase 2 trial, ovulation was completely suppressed during OC use in all 37 subjects evaluable for efficacy, yet an indication of ovulation occurred within 31 days after stopping treatment (
      • Archer D.F.
      • Kovalevsky G.
      • Ballagh S.
      • Grubb G.S.
      Effect on ovarian activity of a continuous-use regimen of oral levonorgestrel/ethinyl estradiol.
      ).
      In a substudy of the phase 3 efficacy and safety trial of continuous LNG/EE, endometrial biopsies were taken pretreatment and at pill pack 13 or at termination of treatment if it occurred earlier (
      • Johnson J.V.
      • Grubb G.S.
      • Constantine G.D.
      Endometrial histology following 1 year of a continuous daily regimen of levonorgestrel 90 mcg/ethinyl estradiol 20 mcg.
      ). Histology results in subjects who had completed at least 168 days of treatment showed generally inactive endometrium. Endometrial biopsy results in women on this continuous regimen for 1 year have shown the majority to have an endometrium categorized as inactive or benign, but <10% of these had atrophy of the glandular epithelium (
      • Johnson J.V.
      • Grubb G.S.
      • Constantine G.D.
      Endometrial histology following 1 year of a continuous daily regimen of levonorgestrel 90 mcg/ethinyl estradiol 20 mcg.
      ). It is likely that the estrogen (EE) dose of 20 μg is sufficient to prevent endometrial atrophy. These biopsy findings may help to explain why the time to return to menses in the current study is relatively rapid. A relatively rapid return to menses after long-term continuous use of other OCs was also described in a case report of three patients (
      • Craig L.B.
      • Miller L.
      • Criniti A.
      Rapid resumption of ovulation and menstruation following continuous use of the combination oral contraceptive.
      ).
      Our study showed no association between the length of amenorrhea before the end of continuous LNG/EE use (after 6–12 months of use) and the time to return to menses. Amenorrhea induced by continuous LNG/EE appears to be just as readily reversible after many months of amenorrhea as it is after a short period of amenorrhea.
      Protracted amenorrhea after discontinuing prolonged use of cyclic OCs has been reported (
      • Weisberg E.
      Fertility after discontinuation of oral contraceptives.
      ); however, the causative relationship is unproven. The amenorrhea observed after any OC use may be due to secondary causes, such as pre-existing oligomenorrhea or low body weight (
      • Weisberg E.
      Fertility after discontinuation of oral contraceptives.
      ,
      • Archer D.F.
      • Thomas R.L.
      The fallacy of the postpill amenorrhea syndrome.
      ). Our study did not include a control group to determine bleeding after cyclic OCs nor did our literature search identify studies of the return to menses after prolonged use of cyclic OCs among women with regular menstrual cycles before starting OCs.
      A limitation of the study is the short duration of follow-up. We are unable to determine what kind of bleeding pattern, if any, was established after the first menses. In addition, we cannot determine whether the first bleeding episode we observed was postovulatory and indicative of a return to fertility. However, the timing of the bleeding suggests that ovulation occurred. These results may not be generalizable to women who had irregular cycles before taking OCs, because only women with regular cycle lengths of 21–35 days before the phase 3 trial were enrolled in the study.
      Continuous LNG/EE is the only continuous OC regimen studied in large trials lasting more than 1 year (
      • Archer D.F.
      • Jensen J.T.
      • Johnson J.V.
      • Borisute H.
      • Grubb G.S.
      • Constantine G.D.
      Evaluation of a continuous regimen of levonorgestrel/ethinyl estrodiol: phase 3 study results.
      ,
      • Teichmann A.T.
      • Kluft C.
      • Grubb G.
      • Constantine G.
      • Spielmann D.
      Comparative trial of continuous-use and 21-day cyclic levonorgestrel and ethinyl estradiol oral contraceptive.
      ). In these studies, the regimen demonstrated comparable efficacy and safety versus standard cyclic regimens. The daily and cumulative doses of hormones in continuous LNG/EE are lower than those with some currently available cyclic OC regimens that have a placebo or pill-free interval. A recent published review of randomized studies of continuous and extended cycle cyclic OCs concluded that the data in those studies were of good quality and that continuous use OCs presented a reasonable option for contraception (
      • Edelman A.B.
      • Gallo M.F.
      • Jensen J.T.
      • Nichols M.D.
      • Schulz K.F.
      • Grimes D.A.
      Continuous or extended cycle vs. cyclic use of combined oral contraceptives for contraception.
      ).
      Our results indicated that inhibition of menses that occurred in women on continuous LNG/EE was readily and quickly reversible.

      Acknowledgments

      The authors wish to acknowledge the contribution of the principal investigators and staff at the clinical sites that conducted the study protocol, and would like to thank Lynne Smith, M.B.A., and Vincent Haudiquet, Ph.D., of Wyeth Research for the statistical analysis, and Iona Coutinho, PhD, Wyeth Research for assistance in the preparation of this manuscript.

      References

        • Coutinho E.M.
        Is menstruation normal? Suppression of the menstrual cycle in clinical practice.
        Reprod Med Rev. 2001; 9: 214-257
        • Kaunitz A.
        Menstruation: choosing whether … and when.
        Contraception. 2000; 62: 277-284
        • Edelman A.
        Menstrual nirvana: amenorrhea through the use of continuous contraceptives.
        Curr Womens Health Rep. 2002; 2: 434-438
        • Sulak P.J.
        • Kuehl T.J.
        • Ortiz M.
        • Shull B.L.
        Acceptance of altering the standard 21-day/7-day oral contraceptive regimen to delay menses and reduce hormone withdrawal symptoms.
        Am J Obstet Gynecol. 2002; 186: 1142-1149
        • Andrist L.C.
        • Arias R.D.
        • Nucatola D.
        • Kaunitz A.M.
        • Musselman B.L.
        • Reiter S.
        • et al.
        Women's and providers' attitudes toward menstrual suppression with extended use of oral contraceptives.
        Contraception. 2004; 70: 359-363
        • Archer D.F.
        • Jensen J.T.
        • Johnson J.V.
        • Borisute H.
        • Grubb G.S.
        • Constantine G.D.
        Evaluation of a continuous regimen of levonorgestrel/ethinyl estrodiol: phase 3 study results.
        Contraception. 2006; 74: 439-445
        • Teichmann A.T.
        • Kluft C.
        • Grubb G.
        • Constantine G.
        • Spielmann D.
        Comparative trial of continuous-use and 21-day cyclic levonorgestrel and ethinyl estradiol oral contraceptive.
        Obstet Gynecol. 2006; 107 ([abstract]): 12S
        • Malcolm C.E.
        • Cumming D.C.
        Does anovulation exist in eumenorrheic women?.
        Obstet Gynecol. 2003; 102: 317-318
        • Lee E.T.
        Statistical methods for survival data analysis.
        2nd ed. Wiley-Interscience, New York1992
        • Neter J.
        • Kutner M.H.
        • Wasserman W.
        • Nachtsheim C.J.
        Applied linear statistical models.
        5th ed. McGraw-Hill/Irwin, New York2004
        • Archer D.F.
        • Kovalevsky G.
        • Ballagh S.
        • Grubb G.S.
        Effect on ovarian activity of a continuous-use regimen of oral levonorgestrel/ethinyl estradiol.
        Fertil Steril. 2005; 84 ([abstract]): S24
        • Johnson J.V.
        • Grubb G.S.
        • Constantine G.D.
        Endometrial histology following 1 year of a continuous daily regimen of levonorgestrel 90 mcg/ethinyl estradiol 20 mcg.
        Contraception. 2007; 75 (23–6)
        • Craig L.B.
        • Miller L.
        • Criniti A.
        Rapid resumption of ovulation and menstruation following continuous use of the combination oral contraceptive.
        Fertil Steril. 2005; 83 ([abstract]): S16
        • Weisberg E.
        Fertility after discontinuation of oral contraceptives.
        Clin Reprod Fertil. 1982; 1: 261-272
        • Archer D.F.
        • Thomas R.L.
        The fallacy of the postpill amenorrhea syndrome.
        Clin Obstet Gynecol. 1981; 24: 943-950
        • Edelman A.B.
        • Gallo M.F.
        • Jensen J.T.
        • Nichols M.D.
        • Schulz K.F.
        • Grimes D.A.
        Continuous or extended cycle vs. cyclic use of combined oral contraceptives for contraception.
        Cochrane Database Syst Rev. 2005; 3 ([review]) (CD004695)