Correlation of antral follicle count and antimullerian hormone in women with sickle cell disease Free

Background: Ovarian reserve is compromised in adults with sickle cell disease (SCD), but whether SCD treatment exposures drive this finding is unclear as is the relationship between compromised ovarian reserve and future fertility. Previous studies show that AMH, a serum marker of ovarian reserve is lower in adults with SCD compared to unaffected age-matched adults. The extent to which AMH, produced by granulosa cells, correlates with antral follicle count (AFC), the number of secondary follicles visualized on ovarian ultrasound, is unknown in SCD. This information is essential to determine if AMH is a reasonable proxy measure of ovarian reserve and to draw meaningful conclusions about oocyte quantity, particularly in the setting of high stakes treatment decisions.

Methods: This single center cross-sectional, IRB approved study enrolled subjects at the Johns Hopkins Sickle Cell Center for Adults between January 2018 and June 2025. We recruited subjects with ovaries and any SCD genotype, aged 18 and older. We excluded those with transformative therapy exposure or known endocrine disorders. We measured AMH, AFC, follicle stimulating hormone (FSH) and estradiol on menstrual cycle days 3 to 5, or any time in subjects on hormonal contraception.

In this analysis, we included subjects < 35 years old. Continuous variables are summarized as median and interquartile range (IQR) and categorical variables as counts. We compared ovarian reserve measures in subjects with hemoglobin SS/HbSβ⁰-thalassemia and those with hemoglobin SC/Sβ⁺-thalassemia, performing univariate analysis with the χ2 test for categorical covariates and Mann-Whitney U test for continuous variables. In the larger HbSS/HbSβ⁰ cohort, we used log linear regression to examine the association of AMH and AFC with current hydroxyurea therapy and serologic markers from the lab draw closest to the study visit including hemoglobin F percent, hemoglobin S percent, ferritin, mean corpuscular volume (MCV), and absolute neutrophil count (ANC).

Results:Among 56 participants, AMH was available for 56 and ovarian ultrasound performed in 38 with 76 ovaries imaged. The cohort median (IQR) age was 25 (23.5, 28.5) years, body mass index (BMI) was 24.7 kg/m² (21.8-29.1). Twenty-one used hormone-containing contraception; 51 /56 had HbSS/HbSβ⁰ and among them, 27 took hydroxyurea and 16 received chronic transfusion therapy (CTT). Among five subjects with HbSC/Sβ⁺, none took hydroxyurea and two received CTT.

We compared the HbSC/Sβ⁺ to SS/HbSβ⁰ groups. There was no difference in median (IQR) age (23(22,24) vs 25(23,29), p=0.27), BMI (32.1 (31.6, 35.5) vs 24.7 (21.3, 27.5), p=0.28), hormonal contraception use (1/5 vs 20/51, p=0.40), AMH (4.5 (2.55, 5.89) vs 1.67 (0.98, 3.58), p=0.06), FSH (4.6 (4.2, 6.5) vs 7.55 (5.8, 10), p=0.09), estradiol (22.05 (20, 24.1) vs 46.25 (31.10, 86.2), p=0.07) or AFC (31(28,34) vs 17 (12, 26.5), p=0.19). No subject had premature ovarian insufficiency.

In the whole cohort, log-linear regression showed AMH and AFC were associated (β=0.03 (95%CI 0.003, 0.06), p = 0.03).

There was no hydroxyurea exposure in the small HbSC/HbSβ⁺. Thus, we further analyzed AMH in the HbSS/HbSβ⁰subjects. AMH was negatively associated with MCV (β=-.03 (95%CI -.06, -.01), p=0.02), and HbF percent (β=-.04 (95%CI -.08, -.01), p=0.02); and positively associated with ferritin (β=.0001 (95%CI .000025, .00018), p=0.01). AMH was not associated with HbS percent (β=-.01 (95%CI -.02, 0), p=0.07), current hydroxyurea use (β=-0.58 (95%CI -1.21, 0.04), p=0.07), or ANC (β=.06 (95%CI -.03, .16), p=0.2).

Conclusions: In adults, AMH predicts fertility preservation outcomes and is used as an indirect measure of ovarian reserve because of demonstrated correlation with AFC. Here AMH and AFC are correlated. A recent study suggested no effect of hydroxyurea on ovarian follicle density in a cohort of mostly premenarchal girls with SCD. Our data suggests that in adults, hydroxyurea's effects may be different. Here, markers of hydroxyurea treatment adherence, MCV and %HbF were associated with lower AMH. Findings in this larger cohort are consistent with our prior work that raised the possibility that hydroxyurea negatively affects ovarian reserve. These findings need further study in a large prospective multi-center study that will define the relationship of AMH and AFC in SCD and determine their relationship with SCD treatment, pregnancy and fertility preservation outcomes.