Reassessment of treatment results in membranoproliferative glomerulonephritis, with emphasis on life-table analysis.

Treatment of idiopathic membranoproliferative glomerulonephritis remains an unsettled issue. The results of five randomized clinical trials have not provided convincing evidence for the effectiveness of any treatment. Follow-up periods in these trials were relatively short-term, ranging from 1 to 4 years. In three recently published long-term clinical studies of patients with membranoproliferative glomerulonephritis, 10-year cumulative survival free of renal failure improved by 20% to 35% over that described in studies published 5 to 10 years earlier. In each study, survival was estimated using life-table analysis. The survival curve in the treated group was then compared with that of a historical control group using the date of clinical onset as time zero. The survival curve in the treatment group was spuriously shifted to the right. By definition, those in the treatment groups had to survive from clinical onset to initiation of treatment, but the historical control group did not have such a constraint. The problem in this comparison is that treatments were not started in a large number of patients for years after clinical onset, resulting in a biased comparison in favor of the treatment groups. (Also, the conclusions drawn from the survival data were that the improvement related directly to various treatments that were used.) Survival was similarly improved in patients treated with dipyridamole and aspirin when survival was plotted against time after clinical onset. However, when the data were replotted and the platelet-inhibitor-treated group was compared with a contemporary randomized control group, no difference in either patient survival or survival free of renal disease was demonstrated.(ABSTRACT TRUNCATED AT 250 WORDS)