Effects of human recombinant erythropoietin on inflammatory status in peritoneal dialysis patients.

Treatment with recombinant human erythropoietin (rHuEPO) in dialysis patients has been associated with improvement of nutritional and immune status through an increase of cytokine production [such as tumor necrosis factor alpha (TNF alpha)]. The high cytokine production can be a double-edged sword owing to the relationship of cytokines with the systemic inflammatory process, which has been associated with many complications of uremic status. Our aim was to analyze the medium-long term effects of rHuEPO treatment on uremic inflammatory markers. We studied 45 peritoneal dialysis (PD) patients divided in two groups: a rHuEPO group (40-70 subcutaneous units/kg weekly) and a control group (no rHuEPO). The treated group was analyzed in four periods. Period 1 (rHuEPO-1) included 24 patients who had been using rHuEPO at long-term. Period 2 (rHuEPO-2; n = 21) looked at the patients 2 months after rHuEPO withdrawal. Period 3 (rHuEPO-3; n = 19) looked at the patients after 2 months under rHuEPO therapy. Period 4 (rHuEPO-4; n = 17) looked at the patients after 4 months on rHuEPO treatment. With the reintroduction of rHuEPO, we observed a progressive, statistically significant (p < 0.05), and temporary increase in TNF alpha plasma levels, from 44 +/- 24 pg/mL (rHuEPO-2) to 76.8 +/- 25 pg/mL (rHuEPO-3), and then to 83 +/- 27 pg/mL (rHuEPO-4). But in the long term, TNF alpha decreased [33.5 +/- 10 pg/mL (rHuEPO-1)]. Similarly, albumin increased in the short term (3.73 +/- 0.5 g/dL to 4 +/- 0.5 g/dL, and then to 4 +/- 0.43 g/dL), and then decreased (3.8 +/- 0.44 g/dL). The normalized protein catabolic rate (nPCR) increased from 1 +/- 0.2 g/kg daily to 1.12 +/- 0.3 g/kg daily (rHuEPO-4). Long term, nPCR decreased to 1.06 +/- 0.3 g/kg daily. Leptin initially increased (60.1 +/- 48 ng/mL to 42.8 +/- 22 ng/mL, and then to 38 +/- 18.2 ng/mL); it also increased in the long term (62 +/- 50.9, p < 0.05). At baseline, we found a significant positive linear correlation (p < 0.05) between TNF alpha and leptin (r = 0.52), TNF alpha and C-reactive protein [(CRP) r = 0.4], CRP and leptin (r = 0.49), fibrinogen and CRP (r = 0.78, p < 0.01), fibrinogen and leptin (r = 0.37), and leptin and body mass index [(BMI) r = 0.67]. In conclusion, rHuEPO induces a temporary, non inflammatory immune hyperactivity mediated by TNF alpha, without the adverse effects associated with that cytokine. By decreasing leptin, rHuEPO could increase food intake and improve the nutritional status of PD patients. At baseline, we confirm the existence of a chronic inflammatory process in uremia.