Randomized trial of cyclosporine microemulsion (neoral) versus conventional cyclosporine in liver transplantation: MILTON study

Background. The new microemulsion formulation of cyclosporine (Neoral) has been developed in an effort to improve the reliability of drug absorption. The objectives of this study were to assess the efficacy, safety, and toler ability of Neoral compared to the original formulation (Sandimmun) in liver transplant recipients. Methods. In a double-blind, parallel group study conducted in 28 centers ac ross Europe and the United States, patients receiving primary orthotopic li ver allografts were randomized within 24 hr of transplantation, 198 to Neor al and 192 to Sandimmun, Patients with and without T-tube biliary drainage were included. Postoperatively, all patients also received intravenous (i.v .) cyclosporine, together with prednisolone and azathioprine, Antibody indu ction was excluded, Efficacy measures were rejections, graft failure, patie nt survival, and the efficacy of the study medication in achieving the desi red cyclosporine blood levels. Safety was assessed by reported adverse even ts, blood pressure, serum creatinine, and other routine laboratory measurem ents. Results. Kaplan-Meier analyses showed that the Neoral group performed bette r than the Sandimmun group, with the estimates for patients free of treated rejection and histologically confirmed rejection either showing or approac hing statistical significance at the 5% level. By 52 weeks, 5.8% (95% confi dence limits: -4.4-15.9%) fewer patients required treatment of acute reject ion in the Neoral group. The proportion of: patients experiencing at least one treated rejection episode by 2 weeks was 29.8% for Neoral and 43.2% for Sandimmun. For histologically confirmed rejection, these proportions were 32.8% and 44.3%, respectively, The proportion of patients experiencing at l ease one steroid-resistant rejection was 2.0% for Neoral and 6.3% for Sandi mmun at week 2, and 3.0% and 9.9%, respectively, at week 3. All these diffe rences were significant at P<0.05. By 52 weeks, graft failure was 6.3% on N eoral and 11.4% on Sandimmun, with respective patient survival figures of 8 5.4% and 85.8%. The median duration of the initial episode of i.v. cyclospo rine was 4.0 days for Neoral, compared to 6.5 days for Sandimmun (P<0.001). Within the first 2 weeks, a larger percentage of patients in the Neoral gr oup reached the lower target level of cyclosporine (P less than or equal to 0.01). The; weight-adjusted daily doses of study medication were lower in the Neoral group (median dose: 4.86 vs. 5.42 mg/kg/day, P=0.001), but the b lood levels of cyclosporine showed no difference. For those with a T-tube, more of the patients on Neoral remained free of treated rejection throughou t the study period (P=0.042, Wilcoxon), By week 2, 44.9% of these patients in the Sandimmun group required treatment for rejection compared to 30.2% i n the Neoral group (P=0.007). There was no significant difference between t he groups for serum creatinine, blood pressure, other biochemical and hemat ological variables, or reported adverse events. Conclusions. In liver transplantation in the normal clinical setting, the p harmacokinetic advantages of Neoral translate into clinical superiority ove r Sandimmun without a negative impact on safety. Recent data indicate that it is not optimal to use i.v. cyclosporine initially in this type of study, but the benefit was seen despite this. In keeping with the previous pharma cokinetic studies, patients managed by T-tube biliary drainage, and hence w ith no or limited bile available in the gastrointestinal tract, did particu larly well with Neoral.