Why Do Patients Develop Proteinuria With Sirolimus? Do We Have the Answer?

Sirolimus is a potent suppressor of T-cell proliferation and expansion by inhibition of the Target of Rapamycin Complex 1 (TORC1) protein kinase. Sirolimus was initially developed as a potential solution to the problem of chronic calcineurin inhibitor (CNI) nephrotoxicity.1, 2 Both experimental and clinical studies revealed that sirolimus had immunosuppressive efficacy associated with little of the nephrotoxicity at standard immunosuppressant doses that was associated with cyclosporine.3 It was approved by the US Food and Drug Administration in 1999 as an immunosuppressant to prevent allograft rejection, and has become a widely established agent used in kidney, islet cell, liver, and heart transplants.4 While the precise role of sirolimus in kidney transplants continues to evolve, there is increasing recognition that substantial numbers of patients develop proteinuria when converted to sirolimus.5, 6 The underlying mechanisms for this effect have been elusive, in part because the context of most studies has been comparison with calcineurin inhibitors, which have known antiproteinuric effects.7 In this issue of the American Journal of Kidney Diseases, Franz and colleagues provide further information about the mechanisms of sirolimus-associated proteinuria, by comparing the effect of sirolimus to cyclosporin on glomerular proteinuria and proximal tubular function during the first 3 months following a kidney transplant.8

Head-to-head comparison of sirolimus with either tacrolimus or cyclosporin in de novo use for low-risk patients revealed a higher incidence of acute rejection and worse outcomes,9 and clinical studies designed to supplant cyclosporin with sirolimus as maintenance therapy in patients exhibiting chronic dysfunction with histological lesions of what was then described as “chronic allograft nephropathy” failed to demonstrate benefit associated with sirolimus, especially in those with worse function at conversion.10, 11 The Sirolimus Renal Conversion Trial (CONVERT) also revealed that underlying proteinuria was a risk factor for poor outcomes in patients switched to sirolimus therapy.10 Overall, these and other data indicate that sirolimus might best serve as a maintenance immunosuppressant following early CNI withdrawal in patients with low immunological risk,11, 12 and/or in the prevention or management of post-transplant malignancy.10, 12, 13, 14, 15

What do we know about the mechanisms of proteinuria? In vitro studies suggest that glomerular proteinuria may reflect direct toxicity of sirolimus to the glomerular podocyte-endothelial axis through inhibition of vascular endothelial growth factor (VEGF),16, 17 although strong clinical evidence to support this hypothesis is yet to be furnished (Fig 1). In contrast, the tubular effects of sirolimus appear to be more consistently established. In vitro and in vivo studies show that clinically relevant concentrations of sirolimus have antiproliferative and proapoptotic effects on tubular cells.18, 19 The latter is particularly accentuated when there is a relative deficiency of epidermal growth factor, as in acute tubular necrosis.18, 19 Indeed, clinical studies in transplantation show that delayed graft function is increased with the use of sirolimus and may be associated with intratubular cast nephropathy, related to reduction in tubular epithelial cell survival leading in turn to intranephron obstruction and acute kidney failure,20 as seen in animal models of renal ischemia.19 Impairment in tubular uptake of protein has also been postulated to be a mechanism for proteinuria in sirolimus-associated delayed graft function.21

View Large Image Download to PowerPoint

Figure 1. Potential hypothetical effects of sirolimus in glomerular and tubulointerstitial disease states, as summarized from experimental and clinical studies reported in the literature. The figure shows a schematic diagram of the nephron, highlighting the podocyte-endothelial-VEGF (vascular endothelial growth factor) axis and the tubular epithelial EGF axis of the nephron. Sirolimus may have detrimental effects when proliferation plays an important regenerative role and there is relative underexpression of EGF, as in acute kidney injury.

In the study by Franz and colleagues, 200 patients undergoing cadaveric or living donor transplants between 2001 and 2004 were screened for inclusion into a single-center, prospective, open-label randomized trial (Calfree Study). Of these, 127 patients were randomized before transplant to receive either (1) sirolimus, mycophenolate mofetil, and steroids, or (2) cyclosporin, mycophenolate mofetil, and steroids. The investigators found that while kidney function was similar in both groups as defined by serum creatinine and estimated glomerular filtration rate using the 4-variable Modification of Diet in Renal Disease Study equation, markers of glomerular proteinuria (albumin and transferrin) and tubular proteinuria (alpha 1-microglobulin and transferrin) as well as tubular glycosuria were increased approximately 2-fold in the sirolimus-treated group at days 90 and 180.

The strengths of this study are that it is the largest clinical trial that has compared markers of glomerular and tubular toxicity and that the urinary measurements are backed up by histology of the transplanted kidneys undertaken by protocol biopsy. The study has demonstrated substantial differences in several different markers of tubular function and the histology has revealed clear differences in tubular damage. The case for tubular toxicity of sirolimus after kidney transplant thus appears to be strong.

The weaknesses of this study in regards to the attempt to prove definitively the tubulotoxic properties of sirolimus in kidney transplant are, however, worthy of mention. While this is a randomized, controlled, head-to-head study of sirolimus and cyclosporin, the different pharmacokinetic effects mean that concomitant mycophenolic acid exposure would have been lower in the cyclosporin group. Acute rejection rates were not different, but the protocol histology revealed higher subclinical rejection in the sirolimus group, questioning the role of rejection in the observed differences in tubular dysfunction. Attempts to correct for this involved exclusion of data from patients with subclinical rejection, but one must maintain the view that rejection plays a part in tubular damage after kidney transplant and that this could have confounded the study. The fact that the histologist was not blinded and the samples were scored with knowledge of each patient's therapy leads to concern that a bias could have been introduced into the data, leading perhaps to slightly less weighting of the pathological findings. Additionally, the study involved preemptive transplant in 20% of the participants, and 30% had glomerulonephritis as a cause of chronic kidney disease, suggesting that the urine protein would have been contributed by both native and transplanted kidneys in substantial numbers of patients. A part of the observed effect of sirolimus could thus be on the residual function of the damaged native kidneys.

Does this study help us determine the role of sirolimus and its class partner, everolimus, following kidney transplant? The consensus of clinical opinion formed to date on the de novo use of sirolimus has been heavily influenced by 2 issues in the early posttransplant period: (1) the higher incidence of early adverse events such as slow wound healing, lymphoceles, and prolonged acute tubular necrosis, and (2) the lower immunosuppressive potency compared to calcineurin inhibitors, as demonstrated in a number of trials. In this context, it is unlikely that these data will alter that view. At the other extreme, clinical opinion strongly favors long-term use of sirolimus in the context of patients with actual or high risks of malignancy. Importantly, the jury remains undecided on the role of conversion from a calcineurin inhibitor to a TOR inhibitor at 6, 12, 26, or 52 weeks, or at some other defined time after transplant, to avoid long-term nephrotoxicity. New data are awaited from a large trial of everolimus used in this context and the field is ripe for a new meta-analysis of the studies available in the literature. In the context of these clinical decision–forming studies, does this new demonstration of tubular toxicity of sirolimus influence us? The suggestion of the authors is that the tubulotoxic effect of sirolimus is reversible and thus detection may lead to clinical actions which might potentially avert long-term damage. Perhaps the most useful role for this data is in helping us to differentiate the patients who, after conversion to sirolimus, develop low levels of proteinuria from those that become frankly nephrotic, as the explanation may lie in tubular toxicity rather than glomerular injury.