INTERLEUKIN-6-ASSOCIATED ANEMIA - DETERMINATION OF THE UNDERLYING MECHANISM

Recombinant human interleukin-6 (rhIL-6) is a pluripotent cytokine wit h proinflammatory, antitumor, and growth factor effects. Clinical inve stigations of rhIL-6 either alone as immunotherapy or as a colony-stim ulating factor in conjunction with chemotherapy have shown a dose-depe ndent, rapid onset, and largely reversible decrease in venous hematocr it levels. In an effort to determine the mechanism for the rhIL-6-asso ciated anemia, we measured red blood cell volume serially in patients receiving rhIL-6 at either 30 mu g/ kg/day as a 120-hour continuous in travenous infusion (renal cell carcinoma) or 100 mu g/kg/d intravenous ly over 1 hour for 5 days (melanoma) as part of two separate phase II trials. Radioisotope dilution assays with Cr-51-labeled autologous red blood cells and hemolysis screens were performed on day 1 before the initiation of therapy and on day 5 shortly before the end of therapy. In the 6 patients studied, the mean decrease in hemoglobin concentrati on was 1.9 +/- 0.94 g/dL. The mean decrease in the hematocrit level wa s 6% +/- 2% and the mean increase in total blood volume was 731 +/- 33 7 mL. These changes were explained by a mean decrease in red blood mas s of 106 +/- 109 mL and a mean increase in plasma volume of 743 +/- 28 9 mL. The decrease in red blood cell mass was largely explained by phl ebotomy during the hospitalization, but was not statistically signific ant (paired t-test, P = .06). All other changes were statistically sig nificant (P < .05). Simple regression analysis indicated that the decr ease in hematocrit level and increase in plasma volume were related (y = -1.78 - .0066X; R = -.74). Measurements of lactate dehydrogenase, b ilirubin, haptoglobin, and reticulocyte counts and serial stool hemocc ults did not indicate hemolysis or blood loss. We conclude that the an emia caused by IL-6 is caused by an increase in plasma volume. (C) 199 5 by The American Society of Hematology.