Keywords: Luteinizing hormone, Estradiol, Ovulation, Progesterone, Letrozole, Aromatase inhibitor
Introduction
Classic teaching of the hormonal events which trigger female ovulation stresses the importance of the hypothalamic-pituitary-ovarian axis [1,2]. Estradiol produced from the dominant follicle rises rapidly by cycle day 7 and continue to increase through the mid-follicular phase [1,3]. The mid-follicular peak in estradiol exerts a positive feedback influence on gonadotropin releasing hormone (GnRH) which leads to a surge of LH release and subsequent ovulation [1,2].
Estradiol is synthesized from androgen by aromatase enzyme activity in granulosa cells [4,5]. Letrozole is a nonsteroidal third generation aromatase inhibitor (AI) which results in low serum estrogen levels during ovulation induction [6].
Since the common understanding of ovulation in a natural cycle suggests that a sustained, elevated estradiol level is required to trigger the LH surge, administration of letrozole throughout the cycle should lower estradiol levels and prevent the LH surge from occurring. We sought to determine if the LH surge, ovulation, and subsequent progesterone levels increase in spite of maintaining low estradiol levels by daily administration of letrozole in normal ovulatory volunteers in a prospective study.
Materials and Methods
After IRB approval and informed consent were obtained, ten willing volunteers who met inclusion criteria (no hormonal contraception within 3 months, regular menstrual cycles 26-30 days, normal thyroid function and normal prolactin, and no pregnancy currently or within 3 months) were monitored for one month without treatment for evaluation of normal ovulation.
Natural control cycle
The subjects used home urine LH tests (Clearblue® Easy, SPD Swiss Precision Diagnostics, Switzerland) on days 10-18 to monitor for the LH surge in both the initial natural cycle and the letrozole cycle. In the natural cycle and in the letrozole cycle, blood was drawn every other day starting on day 12 of the cycle through day 22 to measure estradiol and progesterone levels, the LH surge was monitored with home urine ovulation tests on days 10-18, and follicular development was monitored one time using transvaginal ultrasound on cycle day 12-14.
Letrozole cycle
In the next cycle, all subjects were administered oral letrozole 5 mg daily (Femara®, Novartis Pharmaceuticals Corporation, East Hanover, NJ ) starting on cycle day 1-3 and continuing through cycle day 22. Table 1 illustrates protocols for both the natural control cycle and the letrozole study cycle.
The primary outcome, assessment of ovulation in letrozole cycles, was determined by the presence or absence of progesterone elevation (>1.5 ng/mL) and the presence or absence of a positive urinary LH test. The bioequivalence evaluation of two cycles (before and after letrozole administration) was based on pharmacokinetic parameters such as area under the serum concentration-time curve (AUC), the peak serum concentration (Cmax) and the time of peak serum concentration (Tmax). Cmax and Tmax were determined by visual inspection from each volunteer's serum concentration-time curve for estradiol and progesterone. AUC was calculated by the linear trapezoidal method from day 12 through day 22 in both the initial natural cycle and the letrozole cycle.
Paired t-tests, or Wilcoxon Signed Rank tests if non-normally distributed, were used to evaluate the statistical significance of the mean values of the pharmacokinetic parameters. The McNemar test was used to assess the difference in LH surge and follicular development before and after letrozole administration. A standard of statistical significance (alpha) of 0.05 was used in all cases. The SAS System (SAS Institute, Cary, NC) was used for all analyses.
Results
As expected, the peak serum E2 levels were significantly lower in the letrozole cycle than in the natural control cycle (118 +/- 53 pg/mL vs. 232 +/- 95 pg/mL; p<0.05). Additionally, the E2AUC was significantly lower in the letrozole cycle than in the natural cycle (478 +/- 275 vs. 1410 +/- 347; p<0.05). One subject ovulated with a peak estradiol level less than 100 pg/mL.
Despite significantly lower E2 concentrations, all volunteers ovulated during the letrozole cycle, while 9/10 ovulated during the natural cycle.
In spite of continuing letrozole, the peak progesterone was significantly higher during the letrozole cycle compared to the normal cycle (42.4 +/- 27.6 ng/mL vs. 18.4 +/- 5.3 ng/mL; p<0.05). Additionally, progesterone AUC was significantly increased in the letrozole cycle compared to the natural cycle (207 +/- 38 vs. 93 +/- 39; p<0.05). Other results include a trend toward larger lead follicular diameter in the letrozole cycle (diameter 20.8 +/- 7.4 mm vs 17.2 +/- 5.8 mm) and slightly decreased endometrial thickness during the letrozole cycle (8.1 +/- 1.5 mm vs. 8.4 +/- 2.5 mm) compared to the control natural cycle.Results are depicted in Table 2 and Figure 1.
The demographic data are presented in Table 3.
Discussion
The results of this study found that women taking letrozole ovulated even though estradiol levels were significantly lower than in a natural ovulation cycle. This observation is contrary to the classical teaching that sustained elevated estradiol is the trigger for ovulation [1-6]. These results demonstrate that a high sustained estradiol level is not required to trigger ovulation as commonly believed. Perhaps estradiol has a permissive role in ovulation induction, or it may be supplemented by other factors, likely one or more substances produced by the mature preovulatory follicle.
The exact trigger for ovulation induction has been the subject of much investigation, but it remains unclear. One theory suggests that progesterone augments the effects of estradiol to induce the LH surge and subsequent ovulation [1,7]. Estradiol induces expression of progesterone receptors in the hypothalamus, and this increases the hypothalamic sensitivity to circulating progesterone [1,7]. Another theory involving progesterone as an adjuvant trigger is based on the discovery of de novo steroidogenesis in neural tissue [8]. Estradiol binds to estrogen receptor alpha (ERα) receptors on astrocycte membranes and induces de novo progesterone production, and this progesterone augments the peak estradiol to induce a GnRH pulse and the subsequent LH surge [9]. However, this theory is refuted by the observations in the current study, as ovulation occurred even though estradiol levels were low. Finally, kisspeptin is involved in generation of the luteinizing hormone surge and inducing ovulation, and kisspeptin-54 has been used to trigger egg maturation in women undergoing IVF [9]. However, since kisspeptin is at least partially regulated by estrogen, it is surprising that the LH surge and ovulation occurred even though estradiol levels were low when letrozole was administered.
Strengths include this being a randomized controlled study with patients serving as their own controls. The protocol design is simple; however there are limitations with this work. Only one dosage 5 mg of letrozole was studied. We also chose to use home urine LH tests rather than serum or urine LH levels or serial follicular ultrasounds to determine if ovulation occurred, which was done both for ease of our subjects and because home urine LH tests were thoughts to be accurate to determine ovulation. The results of this research and clinical observation suggest that estradiol is not the trigger to induce the LH surge in humans. Supraphysiologic estradiol levels do not induce the LH surge. Estradiol levels may be markedly elevated during ovarian stimulation for in vitro fertilization (IVF), but premature ovulation is rare when the leading follicle diameter is less than 14 mm [10]. Likewise, in our study, we showed that the LH surge occurs in spite of a subphysiologic estradiol levels. It is most likely that the true trigger for the LH surge comes from some other factor secreted by the dominant follicle in humans. The factor that triggers ovulation has yet to be elucidated.
Acknowledgments
None
Conflict of Interest
The authors have no conflicts of interest to report.