BACKGROUND: Higher prostate-specific antigen (PSA) and alkaline phosphatase (ALK-P) levels predicted worse survival in men with metastatic hormone-refractory prostate cancer (HRPC). In the current study, the authors evaluated the combined effects of PSA and ALK-P on survival. METHODS: Two hundred twenty-four men who had HRPC with bone metastases and who were receiving chemotherapy were identified, and 143 of those men had data available on both ALK-P and PSA levels at chemotherapy initiation. The primary endpoint of the study was overall survival (OS) after chemotherapy. The men were dichotomized into normal and abnormal ALK-P groups according to levels based on institutional normal ranges. The effect of PSA was evaluated as both a categorical value and a continuous value using Cox regression. RESULTS: Eighty-nine of 143 patients (62%) had elevated ALK-P levels. The median PSA was 147 ng/mL (93 ng/mL in patients with normal ALK-P, 171 ng/mL in patients with elevated ALK-P). At a median follow-up of 30 months after chemotherapy initiation, 93 patients had died. The median OS after chemotherapy was 15.8 months (95% confidence interval, 12.8-18.4 months) and was significantly longer if ALK-P was in the normal range (21.3 months vs 14 months; P = .005). For the group with normal ALK-P levels, the median OS was 12.5 months, 24.5 months, and 36.9 months for patients with low, medium, and high PSA levels, respectively. In contrast, the effect of PSA on survival was not as evident in the group with elevated ALK-P levels (16.5 months vs 11.9 months vs 12.1 months, respectively; P = .14 for interaction). Age-adjusted multivariate analysis demonstrated statistically significant interactions of PSA and ALK-P with OS (P = .02). CONCLUSIONS: ALK-P significantly predicted OS in men with HRPC who had bone metastases. In patients with normal ALK-P levels, higher PSA levels were associated with improved survival. 2007 American Cancer Society
BACKGROUND: Higher prostate-specific antigen (PSA) and alkaline phosphatase (ALK-P) levels predicted worse survival in men with metastatic hormone-refractory prostate cancer (HRPC). In the current study, the authors evaluated the combined effects of PSA and ALK-P on survival. METHODS: Two hundred twenty-four men who had HRPC with bone metastases and who were receiving chemotherapy were identified, and 143 of those men had data available on both ALK-P and PSA levels at chemotherapy initiation. The primary endpoint of the study was overall survival (OS) after chemotherapy. The men were dichotomized into normal and abnormal ALK-P groups according to levels based on institutional normal ranges. The effect of PSA was evaluated as both a categorical value and a continuous value using Cox regression. RESULTS: Eighty-nine of 143 patients (62%) had elevated ALK-P levels. The median PSA was 147 ng/mL (93 ng/mL in patients with normal ALK-P, 171 ng/mL in patients with elevated ALK-P). At a median follow-up of 30 months after chemotherapy initiation, 93 patients had died. The median OS after chemotherapy was 15.8 months (95% confidence interval, 12.8-18.4 months) and was significantly longer if ALK-P was in the normal range (21.3 months vs 14 months; P = .005). For the group with normal ALK-P levels, the median OS was 12.5 months, 24.5 months, and 36.9 months for patients with low, medium, and high PSA levels, respectively. In contrast, the effect of PSA on survival was not as evident in the group with elevated ALK-P levels (16.5 months vs 11.9 months vs 12.1 months, respectively; P = .14 for interaction). Age-adjusted multivariate analysis demonstrated statistically significant interactions of PSA and ALK-P with OS (P = .02). CONCLUSIONS: ALK-P significantly predicted OS in men with HRPC who had bone metastases. In patients with normal ALK-P levels, higher PSA levels were associated with improved survival. 2007 American Cancer Society