Hematocrit was not validated as a surrogate end point for survival among epoetin-treated hemodialysis patients

To evaluate the use of hematocrit as a surrogate end point for survival among end-stage renal disease (ESRD) patients treated with epoetin. Using United States Renal Data System (USRDS) data, we conducted an observational prospective study to analyze the relationships among epoetin dose, hematocrit, and survival for 31,301 facility-based hemodialysis patients incident to ESRD therapy in 1998. To address our objective, we used criteria developed by Prentice based on results from a Cox regression model. Results indicate that hematocrit is inversely associated with epoetin dose. For the same epoetin treatment-related achieved hematocrit levels, there were widely varying treatment-related survival outcomes, thereby challenging a central criterion required to empirically validate a surrogate end point. Our results support earlier clinical trial and epidemiological data suggesting that hematocrit may not be a valid surrogate for survival among the epoetin-treated renal failure population. We hypothesize that hematocrit may not be in the causal pathway or that epoetin may have important mechanisms of action apart from increasing hematocrit. Effective treatment for anemia may therefore not be simply a matter of increasing hematocrit. This study has potential implications for revising the existing treatment guidelines for anemia management and selecting an appropriate treatment regimen.