To the Editor:
The reproductive risk to girls with pediatric systemic lupus erythematosus (pSLE) of varying disease severity and medication exposure is not well established. Because 15%–20% of patients with SLE are diagnosed before age 19 years1, many young women with pSLE will be affected by disease complications throughout their reproductive years. Although infertility in SLE has been attributed mainly to use of cyclophosphamide (CYC)2,3, further studies are needed to clarify the relationship between disease severity, medication use, and ovarian dysfunction. Because the incidence of premature primary ovarian insufficiency (POI) after CYC exposure in patients with pSLE < 21 years of age is estimated at 0–11%3, biomarkers for morbidities, such as infertility, are important because they may change current treatment strategies.
Anti-Müllerian hormone (AMH) is produced by granulosa cells of preantral and small antral follicles and has been identified as a sensitive biomarker of ovarian reserve. AMH concentrations are relatively stable throughout the menstrual cycle, are unaffected by hormonal contraceptives, and decline with advancing age, as does ovarian reserve/function4,5. Low to undetectable levels of AMH are found in women ages 25–46 years within 5 years of their final menstrual period, in cancer survivors exposed to chemotherapy and/or radiation therapy-induced follicle depletion, and in women with POI4,6,7.
To identify risk factors that may be associated with poor ovarian reserve and a risk of future infertility among patients with pSLE, we compared AMH levels among postmenarcheal young women with pSLE and a history of CYC exposure (pSLE+CYC; n = 6; median age 19.2 yrs), women with pSLE and no history of CYC exposure (pSLE–CYC; n = 17; median age 19.5 yrs), and healthy age-matched controls (n = 23; median age 20.3 yrs). All subjects were Tanner stage V for breast and pubic hair development.
The clinical features of our patients with pSLE are summarized in Table 1. SLE was classified according to the revised 1997 American College of Rheumatology (ACR) criteria at < 19 years of age. Disease activity and damage were assessed by the Systemic Lupus Erythematosus Disease Activity Index SELENA modification (SELENA-SLEDAI) and the Systemic Lupus International Collaborating Clinics (SLICC)/ACR Damage Index (SDI), respectively. Many patients were receiving prednisone (59% pSLE–CYC and 100% pSLE+CYC) and adjunctive immunosuppressive therapy (88% pSLE–CYC and 100% pSLE+CYC), such as azathioprine and mycophenolate mofetil, at time of AMH measurement. Patients with pSLE and +CYC had received CYC 0.4–8.6 years prior to measuring AMH. No patient received gonadotrophin-releasing hormone (GnRH) agonist cotherapy during CYC treatment. Assays for AMH were performed at Columbia University’s Center for Women’s Reproductive Care. Serum AMH levels (AMH Gen II ELISA; Beckman-Coulter) were determined in duplicate from blood samples taken irrespective of phase of menstrual cycle or use of oral contraceptives. The standard curve of this assay ranges from 0.05 to 10 ng/ml. Intra- and inter-assay coefficients of variation are 6.2% and 9.1%, respectively.
Exposure to CYC in patients with pSLE is associated with a significant reduction in AMH (Figure 1A). The median AMH level in pSLE+CYC patients was lower than in pSLE–CYC patients and controls, 0.79 ng/ml versus 2.01 ng/ml or 1.92 ng/ml, respectively (p = 0.02). Median SLEDAI scores were similar (Figure 1B), but SDI scores were significantly higher in pSLE+CYC patients (p = 0.02; Figure 1C). Spearman rank correlation indicated no association between AMH and SLEDAI (r = −0.16, p = 0.47), AMH and SDI (r = −0.30, p = 0.16), and AMH and disease duration (r = −0.27, p = 0.22).
Reduced AMH levels in pSLE+CYC patients suggests impairment in ovarian reserve despite menstrual regularity, and raises concern regarding prognosis for their future fertility. Consistent with these findings, Aikawa, et al recently reported increased follicle-stimulating hormone and lower AMH in a similar cohort of pSLE patients treated with CYC8. It is encouraging that menstrual cycles and estradiol levels returned to normal in 75% of women < 35 years of age with undetectable AMH levels after gonadotoxic chemotherapy7 and that low AMH in young healthy women ages 19–35 years did not predict reduced fecundability9. Oocyte quality of young women may allow for pregnancy despite reduced ovarian reserve (reflected by low AMH values). However, the significance of diminished AMH in women 15–25 years of age remains to be fully determined.
The effectiveness of GnRH agonist therapy for preservation of fertility in patients with pSLE who are undergoing chemotherapy has yet to be firmly established. However, Marder, et al10 recently showed AMH levels were higher in adult SLE patients who had received GnRH agonist cotreatment during CYC therapy. Such measures should be considered, because they may reduce risks of ovarian dysfunction in young women with severe disease manifestations. Although AMH levels increased by 0.5–0.8 ng/ml in a few patients who were 3 years post-CYC exposure5, the likelihood of AMH rebounding and the clinical significance of such increases in AMH post-CYC exposure remain unknown.
Our preliminary study is limited by the small number of patients from a single center and we lack ultrasound-derived antral follicle counts. Thus, our observations warrant confirmation with larger studies. Yet CYC is the standard of care treatment for severe manifestations of SLE, and therefore our observations should raise concern and encourage consideration of alternative therapies to CYC, such as mycophenolate mofetil. We continue to follow our pSLE patients prospectively to observe trends in their AMH levels and to determine the relationship between AMH, SLE disease activity, and treatment regimens, as well as the future risk of infertility when AMH is reduced in this young cohort.