Alkaline phosphatase velocity predicts overall survival and bone metastasis in patients with castration-resistant prostate cancer

Introduction and objectives Identifying patients with prostate cancer (CaP) who will ultimately develop bone metastasis (BM) or die of disease is essential. Alkaline phosphatase velocity (APV) has been shown to predict overall survival (OS) and bone metastasis–free survival (BMFS) in an earlier study of an equal access military patient cohort of patients with castrate-resistant prostate cancer (CRPC). To confirm these findings, we examined a cohort of patients from a high-volume cancer center to validate a previous observation that faster alkaline phosphatase (AP) kinetics are predictive of OS and BMFS in this second cohort of patients. Materials and methods A retrospective cohort study was conducted of patients with CRPC treated at Memorial Sloan Kettering Cancer Center between 1989 and 2010. All patients who received androgen deprivation therapy (ADT) as primary treatment in response to a rising PSA after definitive surgery for CaP were eligible. For those who received primary ADT or surgery followed by ADT, CRPC was defined as one rising PSA value after a PSA nadir ≤4 ng/ml, and confirmed by a second rising PSA value, with concurrently documented testosterone levels <50 ng/dl. APV was computed as the slope of the linear regression line of all AP values (>2 values per patient) plotted against time. Study outcome included BMFS and OS. Univariable Kaplan-Meier analysis was used to examine time-to-event outcomes. Multivariable Cox proportional hazards regression analysis was used to model time to BMFS and OS. Results Of 89 patients with CRPC with evaluable data and CRPC, 17 (19%) experienced BM and 26 (29%) died. APV was dichotomized at the uppermost quartile split of all observed APV values:≥5.42 U/l/y vs. the lower 3 quartiles combined,<5.42 U/l/y. Patients with faster APV had significantly worse outcomes, including faster progression to BM and poorer OS when compared with those with slower APV (P = 0.0451 and P = 0.0109, respectively). There was strong correlation between PSA doubling time (PSADT) (<10,≥10 mo) and APV (≥5.42 U/l/y vs.<5.42 U/l/y) (P = 0.0289), preventing simultaneous evaluation of both factors in multivariable analysis. Kaplan-Meier analysis showed that PSADT was also predictive of BM and OS (log-rank P<0.0001). Separate multivariable Cox proportional hazards regression models were used to examine PSADT and APV, as predictors of each study outcome (BMFS and OS). Both PSADT and APV were strongly predictive of BMFS and OS (respectively). Conclusions APV and PSADT were predictors of BM and OS in patients with CRPC, respectively. These data are additional evidence of the potential value of AP kinetics in patients with advanced CaP. Prospective studies will be required to clarify these associations. However, given the restrictions on the current patient population in excluding metastatic disease within 12 months of ADT and a PSA nadir >4 ng/ml, the findings are not inappropriately generalized to other men.