Assessment of ovarian function in adolescents and young adults after childhood cancer treatment-How accurate are young adult/parent proxy-reported outcomes?

Abstract

Providers often rely on self-reported ovarian function in adolescent and young adult (AYA)-aged childhood cancer survivors when making clinical decisions. This study described reported menstrual patterns and the agreement between respondent-reported and biochemical premature ovarian insufficiency (POI) in this population. This was a cross-sectional study of survivors (or their parent proxy) aged 13-21.9 years who received gonadotoxic therapy and were enrolled in a longitudinal health survey. Participants reported menstrual regularity, hormone-replacement therapy (HRT) use, and ovarian dysfunction. Respondent-reported POI was defined as the survivor taking HRT for ovarian failure or having been told she had ovarian failure. Biochemical POI was defined as follicle-stimulating hormone (FSH) level ≥40 mIU/mL. The agreement between respondent-reported and biochemical POI was determined using Cohen's kappa coefficient (κ) and analyzed by demographic and clinical factors. Among 182 AYA-aged survivors (72.5% non-Hispanic White, 46.7% leukemia survivors), 14.8% reported requiring HRT to have menses but 55.5% reported regular menses without HRT use. Among survivors with FSH measurements (n=130), 17.7% reported POI whereas 18.5% had FSH ≥40 mIU/mL (κ=0.66, sensitivity 70.8%, specificity 94.3%). The highest agreement between respondent-reported and biochemical POI was with young adult self-report (κ=0.78) and survivors with >5 survivor clinic (κ=0.83) and/or >5 endocrinologist (κ=1.00) visits. The majority of AYA-aged survivors reported having regular menses without HRT support. The accuracy of respondent-reported POI increased with repeated survivor clinic or endocrinologist visits, highlighting the importance of continued education. Survivors must be informed about their ovarian function to enable them to advocate for their reproductive health.