Differences in the probability of discontinued immunosuppressive treatment in patients treated with prednisone alone (- - -) or with CSP plus prednisone (—) were not statistically significant.

Lower curves and left-side scale in each panel show the cumulative incidence of discontinued immunosuppressive treatment (IS) before death or recurrent malignancy (P = .32). Upper curves and right-side scale show the cumulative incidence of death or recurrent malignancy during continued immunosuppressive treatment as competing risks. The space between the 2 curves indicates the proportion of patients surviving without recurrent malignancy and continuing with immunosuppressive treatment. Numbers at the end of each curve indicate point estimates at the end of follow-up.