FSH Requirements for Follicle Growth During Controlled Ovarian Stimulation

Abstract

Introduction: Ovarian follicle growth is a key step in the success of assisted reproductive treatment, but limited data exists to directly relate follicle growth to recombinant FSH (rFSH) dose. In this study, we aim to evaluate FSH requirements for follicular growth during controlled ovarian stimulation. Method: Single center retrospective cohort study of 1,034 IVF cycles conducted between January 2012-January 2016 at Hammersmith Hospital IVF unit, London, UK. Median follicle size after 5 days of stimulation with rFSH and the proportion of antral follicles recruited were analyzed in women treated with rFSH alone to induce follicular growth during IVF treatment. Results: Starting rFSH dose adjusted for body weight (iU/kg) predicted serum FSH level after 5 days of rFSH (r ² = 0.352, p < 0.0001), median follicle size after 5 days of rFSH, and the proportion of antral follicles recruited by the end of stimulation. Day 5 median follicle size predicted median follicle size on subsequent ultrasound scans (r ² = 0.58-0.62; p < 0.0001), and hence time to oocyte maturation trigger (r ² = 0.22, P < 0.0001). Insufficient rFSH starting dose that required >5% dose-increase was associated with increased variability in follicle size on the day of oocyte maturation trigger, and negatively impacted the number of mature oocytes retrieved. Conclusion: Weight-adjusted rFSH dose correlates with follicular growth during ovarian stimulation. Early recruitment of follicles using a sufficient dose of rFSH from the start of stimulation was associated with reduced variability in follicle size at time of oocyte maturation trigger and an increased number of mature oocytes retrieved.

Keywords: fertility; follicle growth; in vitro fertilization (IVF); ovarian response; recombinant FSH; reproduction.

Figure 1

Relationship between serum FSH level and follicle growth. Serum AMH and FSH levels were available in a subset of patients triggered with kisspeptin (serum AMH >10 pmol/L; n = 147). (A) Mean (±SD) of median follicle size on ultrasound by categories of steady state serum FSH levels (iU/L) at 5 days after starting recombinant FSH dose (rFSH) is presented (n = 144; A). Categories were compared by one-way ANOVA with post hoc Tukey's multiple comparisons test. (B) Dose of rFSH adjusted for body weight (iU/kg) predicted serum FSH level (n = 166) by simple linear regression: Day 4–5 serum FSH level = 2.79 x starting rFSH dose (iU/kg) + 1.8, r² = 0.352, p < 0.0001. (C) Mean (±SD) of median follicle size on ultrasound by categories of starting rFSH adjusted for weight (iU/kg) at 5 days after starting rFSH is presented (n = 147). Categories were compared by one-way ANOVA with post hoc Tukey's multiple comparisons test. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.

Figure 2

Effect of starting rFSH dose on follicle growth in predicted “good responders” (AFC>15 and age 18–34 years; n = 354). (A) Mean (±SD) starting Gonal F/kg is positively associated with median follicle size on ultrasound (mm) after 5 days of rFSH treatment. Categories were compared by one-way ANOVA with post hoc Tukey's multiple comparisons test (n = 241). (B) Median (±IQR) starting dose of Gonal F/kg is positively associated with the proportion of antral follicles recruited by the final ultrasound scan during controlled ovarian stimulation. Categories were compared by the Kruskal-Wallis test with post hoc Dunn's multiple comparisons test (n = 324). (C) Median (±IQR) starting dose of rFSH adjusted for weight (iU/kg) is associated with the proportion of antral follicles that result in a mature oocyte being retrieved. Categories were compared by the Kruskal-Wallis test with post hoc Dunn's multiple comparisons test (n = 324). (D) Mean (±SD) of the proportion of 12–19 mm follicles at the final scan is positively associated with categories of starting rFSH dose adjusted for body weight. Categories were compared by one-way ANOVA with post hoc Tukey's multiple comparisons test (n = 324). (E) Median (±IQR) of the proportion of 12–19 mm follicles at the final scan as a function of the antral follicle count is positively associated with categories of starting rFSH dose adjusted for body weight. Categories were compared by the Kruskal-Wallis test with post hoc Dunn's multiple comparisons test (n = 324). (F) Median (±IQR) of the proportion of 12–19 mm follicles at the final scan as a function of the antral follicle count is positively associated with categories of unadjusted starting rFSH dose. Categories were compared by the Kruskal-Wallis test with post hoc Dunn's multiple comparisons test (n = 324). *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.

Figure 3

Effect of starting rFSH dose on variability in follicle size on day of trigger. (A) Median (±IQR) of the number of mature oocytes is presented by proportion of follicles sized 12–19 mm at the final ultrasound scan prior to administration of oocyte maturation trigger. Categories were compared by the Kruskal-Wallis test with post hoc Dunn's multiple comparisons test (n = 1029). (B) Increasing an insufficient starting rFSH dose during the cycle resulted in a reduced proportion of follicles sized 12–19 mm (n = 1034). Mean (±SD) is presented. Categories were compared by one-way ANOVA with post hoc Tukey's multiple comparisons test. (C) Increasing an insufficient starting rFSH dose during the cycle resulted in fewer mature oocytes retrieved. Median (±IQR) number of mature oocytes is presented. Categories were compared by the Kruskal-Wallis test with post hoc Dunn's multiple comparisons test (n = 1034). **p < 0.01, ***p < 0.001, ****p < 0.0001.

Figure 4

Relationship between serum estradiol, follicle size and number of mature oocytes retrieved. (A) Serum estradiol (pmol/L) on the day of the final scan before administration of the trigger of oocyte maturation predicts the number of mature oocytes subsequently retrieved (n = 419; r² = 0.17, P <0.0001). (B) Serum estradiol (pmol/L) at the final scan before administration of the trigger of oocyte maturation also correlates with the cumulative sum of follicle sizes (mm) of all follicles by ultrasound scan (n = 419; r² = 0.30, P < 0.0001). Whilst the sum of follicle sizes on the final scan before administration of oocyte maturation trigger predicted the number of mature oocytes retrieved (C, n = 1034; r² =0.40, P < 0.0001), this was better predicted by the number of follicles sized 12–19 mm in diameter (D, n = 1031; r² = 0.44, P < 0.0001).