Steroid Maintenance Therapy after Transplantation—Quo Vadis?

Since the seminal studies in the pioneer era (1), steroid treatment was regarded as an essential element of immunosuppression in the postoperative as well as in the maintenance phase after renal allograft transplantation. Very early on there had been concerns about the numerous long-term side effects, including the potential contribution to excess cardiovascular complications (2), the high rate of infections (3), and, more recently, the increasingly high rate of diabetes mellitus (4,5) as well as osteoporosis, osteonecrosis, cataract, weight gain, obesity, and dyslipidemia. Nevertheless, despite the pioneer efforts of some courageous nephrologists trying to administer less steroids (6), and in the maintenance phase frequently even no steroids, most nephrologists felt that the risk of rejection after withdrawal of steroids outweighs any benefit from lowering the rate of side effects. After the introduction of more potent immunosuppressants, Ponticelli et al. had repeatedly argued that steroids can be discontinued in the maintenance therapy of many graft recipients (7). The acceptance of this view was very mixed: the retrospective data of the Collaborative Transplant Study (CTS) documented that this approach was tried in no more than 10% of patients 1 yr after transplantation. Skepticism found further support in the results of the meta-analysis by Kasiske et al. in mostly short-term trials. He published in this journal (8) that after steroid withdrawal the risk of rejection was higher by 14 to 40%. In the study of Hricik et al. (9,10), acute allograft rejection episodes were indeed more frequent when steroids were avoided from the time of transplantation or withheld at some time after transplantation, but with the important qualification that patient or allograft survival was not adversely affected. In contrast, a Canadian study by Sinclair (11) with a prospective multicenter design had shown that patient survival was significantly lower in patients in whom steroids had been withdrawn. In this study, however, steroids were withdrawn remarkably early at 3 mo after transplantation and the difference between patients continuing or withdrawing steroids did not hold up when it was corrected for confounding factors. These results led to a cautious and skeptical attitude among most nephrologists. Against these results, however, stood the retrospective data of the CTS. In 12,000 renal graft recipients on cyclosporine, alone or combined with azathioprine, 5-yr patient and graft survival rates were higher, and, in contrast to the above studies, not lower, compared with patients on triple drug therapy including steroids (12). It was possible to argue, however, that retrospective data are not immune to bias. Against this background, and in view of the grave consequences of erring on either side, a controlled study was clearly indicated. The optimal solution undoubtedly would have been a randomized controlled trial, but a consensus conference convened by Opelz opted against this solution, not in the least because of the adverse results of the preceding Canadian study (11). The second best solution was adopted, i.e., a prospective follow-up of low-risk patients with renal allografts (and a relative small cohort of patients with cardiac allografts) in whom steroids were gradually withdrawn no earlier than 6 mo after transplantation. The comparator group was deliberately composed of graft recipients with lower baseline S-creatinine (which, if anything, would make it more difficult for patients in the steroid withdrawal group to end up with better outcomes). These control patients were matched for a number of relevant parameters such as age, gender, race, donor age, HLA mismatch, etc. The long-term results in 1110 cadaver renal graft and 450 cardiac graft patients from 43 centers are indeed impressive. The 7-yr renal graft survival was 6.6% better after steroid withdrawal (81.9 ± 1.% versus 75.3 ± 1.2%; P = 0.0001 in matched controls) and patient survival was also 4.5% better (88.8 ± 1.5% versus 84.3 ± 1.0%; P = 0.0016). The better graft survival was not an artifact resulting from better patient survival, because death-censored graft survival was better as well (91.8 ± 1.3% versus 87.9 ± 1.0%; P = 0.0091). A similar benefit on graft survival was seen in heart recipients (76.2 ± 2.4% versus 66.9 ± 1.7%; P = 0.0008). In contrast to the above studies (8,9,11), the rates of acute rejection episodes and chronic renal graft dysfunction did not differ between patients receiving steroid-free and steroid-containing maintenance therapy. The protocol allowed for administration of steroids when they were thought to be required according to local established protocols, but 58.6% of the renal and 44.3% of the cardiac allograft recipients never required steroids. As for the steroid side effects, somewhat surprisingly the number of patients requiring de novo antihypertensive therapy did not differ between the groups, presumably indicating that in the days of cyclosporine therapy the contribution of steroids to BP elevation has become minor. However, significantly fewer patients without steroids developed total cholesterol concentrations >300 mg/dl. New cases of osteoporosis and cataracts were not significantly different, but this may have been confounded by differences of prevalence at baseline and of cumulative steroid doses. Indeed, in patients in whom withdrawal of steroids was begun earlier, fewer cataracts were observed. In good agreement with previous, uncontrolled, single-center experiences (7,13), the main finding is that in a large proportion of kidney and heart graft recipients, steroids can be withdrawn when certain provisos are respected in suitable, low-risk patients: when the start of withdrawal is delayed beyond 6 mo, when the rate of withdrawal is gradual, and when patients are well-supervised to allow resumption of steroids for acute deterioration of renal function should a rejection episode supervene. Of particular importance is the demonstration that this procedure is safe in the long-term as documented by no differences in 7-yr renal function despite previous observations to the contrary (14). Therefore, delayed slow steroid withdrawal should not be tried in every graft recipient, but in a large proportion of patients it is feasible and safe with respect to graft function and beneficial with respect to patient survival, presumably because of a more favorable cardiovascular risk profile.

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.