RESUMO
OBJECTIVE: To identify patient factors associated with a clinically significant improvement in semen parameters among infertile men treated with the aromatase inhibitor anastrozole. DESIGN: Multi-institutional retrospective cohort study. SETTING: Two Tertiary Academic Medical Centers. PATIENTS: A total of 90 infertile men treated at 2 tertiary academic medical centers who met inclusion criteria and obtained pretreatment and posttreatment semen analyses. INTERVENTION: Prescription of anastrozole (median 3 mg/wk). MAIN OUTCOME MEASURES: Upgrade in the World Health Organization sperm concentration category (WHO-SCC). Univariate logistic regression, multivariable logistic regression, and partitioning analyses were performed to identify statistically significant patient factors capable of predicting treatment response. RESULTS: With anastrozole treatment, 46% (n = 41/90) of men responded favorably with a WHO-SCC upgrade, and 12% (n = 11/90) experienced a downgrade. Responders exhibited lower pretreatment levels of luteinizing hormone (LH, 4.7 vs. 8.3 IU/L) and follicle-stimulating hormone (4.7 vs. 6.7 IU/mL), higher pretreatment levels of testosterone (T, 356 vs. 265 ng/dL), and similar baseline level of estradiol (E2, 73% vs. 70% with detectible level). Baseline semen parameters differed, with anastrozole responders demonstrating higher baseline semen concentration (3.6 vs. 0.3 M/mL) and higher total motile sperm counts (3.7 vs. 0.1 M). Anastrozole therapy converted 29% (n = 26/90) of the cohort to normozoospermia and enabled intrauterine insemination access in 31% (n = 20/64) of previously ineligible patients. Interestingly, neither body mass index nor the baseline E2 level or E2-T ratio was associated with WHO-SCC upgrade. Multivariable logistic regression revealed the T-LH ratio (odds ratio: 1.02, 95% confidence interval: 1.00-1.03) and baseline nonazoospermia (odds ratio: 9.4, 95% confidence interval: 1.1-78.9) to be statistically significant predictors of WHO-SCC upgrade (area under receiver operating characteristic curve: 0.77). The final user-friendly partitioning model consisting of the T-LH ratio ≥100 and baseline non-azoospermia was 98% sensitive and 33% specific for WHO-SCC upgrades (area under the curve: 0.77). CONCLUSION: Anastrozole therapy decreases serum E2 levels, increases serum gonadotropins, and clinically improves semen parameters in half of men with idiopathic infertility. Nonazoospermic infertile men with T-LH ratios ≥100 are likely to benefit from anastrozole treatment irrespective of baseline E2 level or E2-T ratio. Men with azoospermia rarely respond to anastrozole and should be counseled on alternative treatments.