Gender modulates responsiveness to recombinant erythropoietin.

Several investigators reported that individuals with diabetes and women on hemodialysis treated with recombinant erythropoietin (EPO) attained lower hematocrits than individuals without diabetes and men. It is unclear whether these observed differences in achieved hematocrits are caused by inherent biological differences in responsiveness to EPO or undetected differences in modifiable factors that affect response to EPO. Also potentially modulating response to EPO is diurnal variation in the bioavailability of serum iron. To address these issues, we studied 309 patients undergoing hemodialysis in two large facilities in New York City. Retrospective data collected monthly for 3 months included patients' hematocrit, dose of EPO, urea reduction ratio (URR), total amount of intravenous iron administered, serum albumin concentration, transferrin saturation, and time of day patient underwent dialysis. The 309 study subjects (165 women, 144 men) included 207 blacks (67%), 74 Hispanics (24%), 23 whites (7%), and 5 Asians (2%) with a mean age of 55.4 +/- 15.6 (SD) years. Despite a greater mean URR (74% +/- 6.4% versus 71% +/- 6%; P = 0.001) and a 39% greater dose of EPO (97 +/- 65 versus 59 +/- 53 U/kg; P = 0.001), women (36% +/- 3.5%) had hematocrits equivalent with men (36.5% +/- 3.7%; P = not significant [NS]). There was no difference in the amount of intravenous iron administered to men (375 +/- 389 mg) and women (377 +/- 413 mg; P = NS). Diabetes mellitus (P = 0.48) did not significantly affect the odds of attaining a hematocrit greater than 33% after adjustment for URR, EPO dose, and amount of intravenous iron administered. The time of day a patient underwent dialysis (P = 0.93) had no effect on their response to EPO. We conclude that gender, but not diabetes status or time of dialysis, modulates response to EPO in hemodialysis patients.

Adequacy of dialysis and differences in hematocrit among dialysis facilities.

Despite the clearly established relationship between adequacy of dialysis and response to erythropoietin, recent guidelines on anemia management in end-stage renal disease (ESRD) omit mention of dialysis adequacy while advocating the use of large amounts of intravenous iron. To determine the relative effects of adequacy of dialysis and intravenous iron on hematocrit, we studied 309 hemodialysis patients and analyzed data from 141 hemodialysis facilities in New York State (ESRD Network 2), as well as data from all 18 ESRD Networks in the United States, for the last quarter of 1997. Among the 309 subjects, mean hematocrit differed between quartiles of urea reduction ratio (URR; F statistic = 4; P: = 0.008). Patients with URRs greater than 70% were 2.6 times more likely to have hematocrits greater than 33% (odds ratio, 2.6; 95% confidence interval [CI], 1.3 to 5.3; P: = 0.009) after adjustment for other factors. Mean dialysis facility (n = 141) hematocrits correlated directly with mean URRs (r = 0.32; P: = 0.001). Facilities with a mean URR greater than 70% were three times more likely to have a mean hematocrit greater than 33% (odds ratio, 3; 95% CI, 1.2 to 7.5; P: = 0.02). The percentage of patients in each of the 18 ESRD Networks with hematocrits of 33% or greater correlated inversely with the percentage of patients administered intravenous iron (r = -0.53; P: = 0.03) after adjustment for dose of erythropoietin. We conclude that adequacy of dialysis predicts the response to erythropoietin at both patient and dialysis facility levels. Patients with low hematocrits primarily because of inadequate dialysis may inappropriately be administered excess intravenous iron intended as a corrective measure.