Randomised controlled trial of steroid withdrawal in renal transplant recipients receiving triple immunosuppression

Background The combination of cyclosporin, azathioprine, and prednisolone (triple immunosuppression) is the most commonly used immunosuppressive regimen early after renal transplantation, but the risks and benefits of maintaining the steroid component of this regimen in the long term are uncertain. Methods A randomised controlled trial of steroid withdrawal was conducted among renal transplant patients receiving triple immunosuppression. Between one and six years after transplantation, 100 such patients were randomised either to reduce prednisolone treatment to zero over about four months or to maintain their triple immunosuppression unchanged. Outcome was analysed according to ''intention-to-treat''. Findings In 42 (86%) of 49 patients allocated to steroid withdrawal, complete steroid withdrawal was achieved. Although these patients did not experience defined acute rejection episodes, insidious increases in plasma creatinine were observed more frequently in this group than in the controls. In 97 patients surviving one year after trial entry, plasma creatinine exceeded the baseline by more than 25% at some time in the first year in 25 (53%) of 47 in the steroid withdrawal group compared with 9 (18%) of 50 in the control group (p<0.001, chi-square test). In some cases these increases were transient. However, when corrected for the baseline (entry) value significant differences between groups were apparent in both mean plasma creatinine and mean creatinine clearance; mean (SD) plasma creatinine values at entry, immediately after withdrawal, and at one year were 138 (27), 151 (36), and 150 (36) mu mol/L in the steroid withdrawal group versus 138 (34), 140 (51), and 139 (47) mu mol/L in the control group (p=0.017, analysis of covariance). Steroid withdrawal patients showed a further rise in mean plasma creatinine to 160 (44) and 161 (65) mu mol/L at two and three years from trial entry. Changes in several clinical and metabolic indices were also observed in association with steroid withdrawal. Blood pressure declined but the reduction was incompletely sustained, being more evident immediately after steroid withdrawal than at one year. Total cholesterol declined about 1 mmol/L in the steroid withdrawal group. Other changes associated with steroid withdrawal were reductions in white cell count and haemoglobin and increases in plasma phosphate and alkaline phosphatase. Interpretation Late steroid withdrawal is feasible in most patients with stable graft function on triple immunosuppression and has potentially beneficial metabolic effects. However, a substantial proportion of patients show a reduction in graft function, indicating a need for caution in considering the longterm outcome.