Impact of age at diagnosis on survival of hormone-refractory prostate cancer (HRPC) patients
e16050 Background: Conflicting data exist for age as a determinant of overall survival (OS) in pts with HRPC. We hypothesize that young (<55) HRPC pts represent a more aggressive biological phenotype and therefore have a decreased OS. Methods: A retrospective chart review was conducted on 334 consective HRPC pts treated between 1995–2005. Summary statistics for demographic and clinical factors were generated, and Kaplan-Meier (KM) OS curves were created. Bivariate Cox Proportional-Hazards regression was used to test the association of age at diagnosis while adjusting for a covariate, with significant covariates entered into multivariate models. Results: Overall median survivals in the age stratified categories (<55, ≥55–65, ≥65–75, ≥75) were 5.5, 6.9, 7.9, and 4.3 yrs, with 5 yr survivals 51.9%, 67.4%, 67.0%, and 34.9%, respectively. KM curves showed divergence with an overall significant log-rank test (p < 0.0001). Compared to pts ≥65–75, the hazard ratios (HR) for HRPC pts <55 and ≥75 were 1.40 (95% CI 0.90–2.60) and 2.52 (95% CI 1.67–3.82), respectively. However, following multivariate analysis HRs for HRPC pts <55 and ≥75 were 1.60 (95% CI 0.98–2.62) and 1.25 (95% CI 0.71–2.20). Pts <55 and ≥75 presented with advanced stage at diagnosis and progressed to bone metastasis earlier. Pts ≥75 had decreased performance status, more comorbidities, higher PSA at diagnosis, shorter duration of hormone sensitive disease, and were less likely to receive chemotherapy than pts <75. The percentage of rapid PSA doubling times was highest in the <55 cohort. In multivariate analysis with age as a categorical variate, ECOG 3–4 (HR 2.65), time from diagnosis to both HRPC (HR 0.78) and bone metastasis (HR 0.80), and duration of response to androgen ablation (HR 0.86) remained highly predictive. Conclusions: Age at diagnosis influences OS in HRPC with a bimodal survival curve. Pts <55 and ≥75 present with more aggressive disease, translating into reduced median and 5 yr survivals. Other covariates, especially ECOG status, likely account for the decreased OS in the ≥75 cohort. Conversely, pts <55 had an adjusted HR of 1.60 (p = 0.06). Our study supports a growing body evidence that suggests a poor prognosis in younger men; correlating these differences at the molecular level could lead to better targeted therapies. No significant financial relationships to disclose.