The purpose of this prospective pilot study was to determine whether a strategy of early steroid reduction, but not complete withdrawal, in association with low dose cyclosporine, MPA and Thymoglobulin, would maintain low rates of AR while minimizing the development of NODM. In 48 patients followed for 6 months, there was only one episode of AR, two new cases of NODM (both with impaired glucose tolerance pre-transplant) and two new cases of impaired glucose tolerance. These results compare favorably to those in a historical cohort who in particular had a higher incidence of NODM. We believe that this protocol was successful in minimizing both important metabolic and immunologic complications of transplantation over the short-term.
The benefits of early steroid withdrawal are controversial. In 2009, a large retrospective review of US data showed no reduction in 4-year graft and patient survival in patients discharged from hospital on a steroid-free regimen . Although, in fact graft and patient survival were significantly better in steroid-free patients, the authors comment that the difference was slight and cannot be presumed to be related to steroids because of the retrospective registry-based nature of the study. Another similar retrospective review found that steroids are initiated in a significant percentage of patients initially discharged without them and that this group has lower graft survival than those discharged on steroids . A recent study from the Minneapolis group showed reduced NODM (based on treatment alone), avascular necrosis, cataracts and CMV infection compared to a retrospective control group. However, it should be noted that much higher steroid doses were used in that group compared with practice today, that is 0.4 mg/kg at 1 month and down to 0.15 mg/kg/day at 1 year. Therefore, it is by no means clear that early steroid withdrawal would have these benefits versus the much lower steroid doses used today in steroid-containing protocols. Accordingly, the optimal long-term strategy remains unclear.
Other larger, randomized controlled trials have attempted to maintain freedom from both AR and NODM. The multicenter US study  was a randomized double-blind controlled trial over 5 years. In that study, tacrolimus was the calcineurin inhibitor used, steroids were withdrawn completely and some patients received Thymoglobulin, whereas others received anti-CD25 antibodies. Both biopsy-proven AR and chronic allograft nephropathy were more common in the steroid withdrawn group. However, there was no overall difference in NODM based on blood glucose testing although fewer subjects in the steroid withdrawal group needed insulin. There were also no major benefits shown in other cardiovascular risk factors. In the Freedom study, there were 3 groups: standard steroids, steroid withdrawal and a group receiving no steroids. Once again steroid withdrawal (and steroid-free) protocols were associated with significantly more AR. Reduced NODM was only seen in the steroid-free group, which also had the highest AR rate. Of note, two earlier studies did not demonstrate a higher AR rate after steroid withdrawal, but in both cases patients in the control group did not receive induction, potentially accounting for a higher incidence of rejection in those patients [11, 12].
The most common causes of late graft loss are patient death and combined interstitial fibrosis and tubular atrophy. Cardiovascular disease is the commonest cause of patient death . Several studies have shown that NODM is associated with both patient death and death-censored graft loss [2, 14]. Recent work has demonstrated that NODM occurs in up to 20% of patients in the first year post-transplant [15, 16]. It has been suggested, based on retrospective data, which likely underestimates NODM incidence, that NODM is associated with a 90% increase in mortality and 60% increase in graft loss within 3 years post-transplant . Accordingly reduction in NODM is an important goal with the potential to significantly improve long-term graft loss.
Previous studies have shown that tacrolimus is associated with a greater risk of NODM than cyclosporine A [14, 17], with similar AR rates when cyclosporine is monitored by C2 . Accordingly we chose cyclosporine as the calcineurin inhibitor in this protocol to minimize NODM risk and antibody induction, to reduce AR risk and permit reduced cyclosporine exposure [4, 18]. We chose Thymoglobulin induction based on US data showing a trend towards less AR versus CD25 antibody induction in patients in the early steroid withdrawal arm .
While steroid dose reduction would be expected to improve glycemic control, a previous study has shown no additional benefit in glucose control of complete steroid withdrawal below 5 mg/day of prednisone . Accordingly, in an effort to keep AR rates low, we chose to reduce steroids rapidly but to maintain prednisone at 5 mg/day. Our very low incidence of NODM seems superior to that seen by Vincenti et al. at 1 year , using higher cyclosporine targets, of 14.7% in patients on cyclosporine with standard steroids, or 12.2% in those in whom steroids were withdrawn in 1 week. However, that study included many Agfrican American Black patients, known to be at higher risk of NODM.
The issue of the long-term safety of steroid-sparing regimens is historically linked to the Canadian Multicentre Transplant Study Group  because of a higher risk of graft loss in those in whom steroids were discontinued. However, that study had significant methodological problems and the immunosuppression protocols were significantly less effective than those in current use. Despite equivalent graft and patient survival at 5 years in the US study, long-term safety concerns were renewed with findings from a secondary analysis involving a sample of patients with available biopsy data that showed a significantly higher rate of chronic allograft nephropathy in the corticosteroid withdrawal (CSWD) group  It was never clarified whether this was attributable to patients with AR in the CSWD group. A recent observational study from the University of Minnesota reported excellent 10-year outcomes with rapid steroid discontinuation and should address the concerns regarding the long-term safety of CSWD at least in Caucasian patients .
The major strength of our study is exclusion of diabetic patients from the pilot study and rigorous assessment of NODM after transplantation. The major limitation is a lack of prospective control group and relatively few patients at high risk for NODM based on demographic features. Comparisons with the historical control group reinforce our clinical impression that this regimen safely reduces NODM and AR, and we have continued to use this regimen in clinical practice in low immunological risk patients with high risk of NODM, based on historical features, while we plan a definitive study enriched for patients at greater risk for NODM than those included in this pilot study.