Predictors of clomiphene citrate response in the treatment of men with testosterone deficiency

BACKGROUND: Clomiphene citrate (CC) is an established treatment for men with low testosterone, but predictors of treatment response remain poorly defined. AIM: To identify factors associated with clinically meaningful increases in serum testosterone levels during CC therapy. METHODS: This retrospective study analyzed men diagnosed with low testosterone and treated with CC. Inclusion criteria were (1) a diagnosis of low testosterone (total testosterone (TT) ≤300 ng/dL with symptoms) or borderline low testosterone (TT 300-400 ng/dL with objective signs of low testosterone (low bone density or elevated HbA1c)), (2) laboratory follow-up within 12-weeks of initiation, and (4) no prior testosterone therapy. Initial CC dosing was 25 mg every other day (QOD), with escalation to 50 mg QOD if TT remained <400 ng/dL. Labs were redrawn every 4-weeks following dose changes until TT was at goal or until CC discontinuation. Discontinuation within 12-weeks without documented response constituted treatment failure. TT was assessed using liquid chromatography-mass spectrometry. Multivariable models were used to identify predictors of treatment response. OUTCOMES: The primary outcome was achievement of treatment response, defined as TT ≥400 ng/dL on treatment plus an increase in TT ≥200 ng/dL from baseline. RESULTS: The study included 292 men with median age of 60 (IQR 50, 66) years, median baseline TT (219, 314) 264 ng/dL, and median baseline luteinizing hormone (LH) of 3.5 (2.6, 5.1) mIU/mL. Comorbidities included diabetes (18%), hyperlipidemia (46%), hypertension (44%), prior radical prostatectomy (41%), prostate radiotherapy (12%), and androgen deprivation therapy (ADT) (4.5%). Treatment response was achieved in 136 of 292 (47%) patients; 156 (53%) failed to meet response criteria within 12-weeks. Multivariable analysis identified baseline LH (increase per mIU/mL) (OR 0.82, CI = 0.71-0.95, P = .008) as a significant negative predictor for achieving treatment response. Likewise, prior ADT was predictive for poor response (OR 0.11, CI = 0.01-0.6, P = .039). Baseline TT and age at start of CC treatment were not predictive. The study included 292 men with median age of 60 (IQR 50, 66) years, median baseline TT (219, 314) 264 ng/dL, and median baseline LH of 3.5 (2.6, 5.1) mIU/mL. Comorbidities included diabetes (18%), hyperlipidemia (46%), hypertension (44%), prior radical prostatectomy (41%), prostate radiotherapy (12%), and ADT (4.5%). Treatment response was achieved in 136 of 292 (47%) patients; 156 (53%) failed to meet response criteria within 12-weeks. Multivariable analysis identified baseline LH (increase per mIU/mL) (OR 0.82, CI = 0.71-0.95, P = .008) as a significant negative predictor for achieving treatment response. Likewise, prior ADT was predictive for poor response (OR 0.11, CI = 0.01-0.6, P = .039). Baseline TT and age at start of CC treatment were not predictive. CLINICAL IMPLICATIONS: Identification of predictors for CC treatment response enables individualized counseling, leading to better informed treatment decisions and avoidance of treatment failure. STRENGTHS AND LIMITATIONS: Strengths include cohort size, utilization of the gold-standard lab assessment for TT (LCMS), and standardized reproducible clinical pathways. Limitations include having a study population that is skewed older and less healthy than average, lack of long-term follow-up, and lack of quantifiable data on symptoms. CONCLUSION: In men with low testosterone, this study found that higher baseline LH and history of prior ADT predicted worse response to CC therapy.