Objectives: To test the hypothesis that chromogranin A (CgA) levels are prognostic in patients with metastatic hormone-refractory prostate cancer (HRPC). The extent of neuroendocrine differentiation in prostate cancer correlates with aggressive disease and with progression to HRPC. Plasma CgA levels in patients with prostate cancer may reflect the extent of the tumor neuroendocrine phenotype.
Methods: Pretreatment plasma was collected from 390 patients with metastatic HRPC enrolled in the Cancer and Leukemia Group B (CALGB) 9480 trial, a study of three different doses of suramin. Plasma CgA levels were determined in 321 samples in duplicate using a quantitative sandwich immunoassay. The proportional hazards model was used to assess the prognostic significance of CgA in predicting overall survival.
Results: The median plasma CgA level was 12 U/L (interquartile range 7.7 to 19.3). In univariate analysis, plasma CgA correlated inversely with survival times, with a survival time of 17 months for low CgA (less than 12 U/L, 95% CI 14 to 19) compared with 11 months for high CgA (95% CI 8 to 14, P = 0.014) and at all exploratory cutpoints, including CgA of 9.5 U/L or less versus greater than 9.5 U/L, with survival of 19 months compared with 12 months (P = 0.0015). In multivariate models (adjusting for performance status, prostate-specific antigen, and lactate dehydrogenase), the plasma CgA levels remained predictive of overall survival.
Conclusions: These results support the hypothesis that serum CgA levels correlate with outcome in patients with HRPC, although the clinical significance needs to be established in confirmatory studies before incorporation of CgA measurements in clinical practice.