Those studies, wherein CNI regimen were changed after occurrence of renal dysfunction, raised the question whether CNI delay after transplantation would be beneficial for preserving renal function. MMF de novo was the immunosuppression of choice in most of these studies, often combined with induction therapy using an anti IL2 receptor antibody. The first prospective randomized trial with delayed TAC by Yoshida et al. , revealed a significantly better GFR in patients at 6 months after transplantation as compared with those who received TAC from day 1 after LTx. This is somehow surprising, as the TAC delay was only for 4–6 days after LTx. But also studies by Soliman et al.  and Bajjoka et al.  in which CNI delay lasted only 3 days (in combination with induction therapy) showed a significant increase in kidney function compared with CNIs from day 1.
These data point out that kidney function is at particular risk very early after LTx with a relatively long-lasting effect, especially if kidney function was already impaired. This is also confirmed by the ReSpeCT Study (A. D. Mayer, unpublished data). Concerning rejection, no significant increase was found with delayed CNI therapy. This therapeutic regimen – with delayed CNIs, MMF and induction along with corticosteroids – corresponds with the regimen we installed at our center at the University of Regensburg/Germany after implementation of the MELD allocation system in the Eurotransplant area. Moreover, patients needing renal replacement therapy did not receive CNIs before recovery of renal function. Although the number of patients needing renal replacement therapy before and after transplantation was the same in both groups (which is in accordance with the study by Machicao et al. ), the duration of renal replacement therapy after transplantation was shorter with delayed CNI therapy. A prospective study (PATRON trial) will follow (A. A. Schnitzbauer et al., unpublished data).
CNI reduction and withdrawal strategies
The initial studies with CNI minimization were performed in patients that had already developed CNI nephropathy [36,41,50]. The follow up was up to 1 year in smaller patient numbers. These studies showed for the first time that introduction of MMF combined with the reduction of at least 50% of CNI dose allowed the renal function of liver transplant recipients to significantly improve at 1 year. Now longer term results are coming up, however still retrospective. In the abstract by J. A. Thompson (unpublished data) CNIs were taken off after 3 years. In 58 patients during 2 years of observation, one biopsy-proven rejection occurred, which led to the death of the patient at re-transplantation. In 46% of the patients, creatinine improved significantly. These data support that CNI withdrawal is mostly safe and improves renal function in a large fraction of the patients. Nevertheless, rejections on MMF monotherapy remained a problem for some patients. They can occur surprisingly late after LTx, as shown in these studies.
Another important point has to be considered in these studies. Only half of the patients showed a significant improvement of renal function after withdrawal of CNIs. This means that either they were taken off too late – when CNI nephropathy was already irreversible – or CNIs did not contribute to nephropathy in these patients. Thus, it should be distinguished between CNI-nephropathy and nephropathy caused by other reasons like diabetes, vasculitis or glomerulonephritis. Renal biopsy is a reliable tool to diagnose calcineurin-inhibitor-induced nephrotoxicity. In chronic toxicity, the medial hyaline deposits beaded in afferent arterioles are found and the interstitium displays striped fibrosis and tubular atrophy [51–54]. Therefore, it should be considered imperative to perform a renal biopsy before CNI withdrawal or if CNI withdrawal did not improve renal function.
A further step to avoid CNI nephropathy is replacement of CNIs by mTOR inhibitors like sirolimus. However, the evolving literature on mTOR inhibitors shows mixed results concerning conversion from CNIs to mTOR inhibitors because of renal preservation . Watson et al.  found a significant, if modest, improvement in the change of GFR at 1 year from the time of conversion, but no significant difference in absolute GFR. Shenoy et al. showed in a late conversion trial (mean time of conversion 4.4 years) an improvement in CrCl at 3 months but not at 12 months. Fairbanks et al.  found an increase in eGFR after switching from CNI to sirolimus in 25% of the patients. Nair et al.  found no improvement in renal insufficiency in 43% of the patients after switching to SRL. The recent study by DuBay et al.  showed a stabilization of CrCl but no improvement after switching to SRL. These results are in contrast to the above-mentioned studies. Differences in study methodology may partly explain these conflicting results. DuBay et al. used a well-matched CNI reduction arm, which was not done in other studies. Ultimately, a multicenter, randomized controlled trial is necessary to clarify the role of sirolimus in ameliorating CNI- induced renal toxicity after LTx. Furthermore, in the study by DuBay et al. , especially patients with severe renal dysfunction had more side-effects because of SRL. Regarding the lack of effect of a switch to sirolimus, in particular for patients with already severe renal dysfunction, it has to be noted that CNI nephropathy is probably irreversible if detected too late. Thus, a switch too late only offers probably more side-effects than benefits for the patients. Furthermore, it has to be noted that a switch to sirolimus can induce proteinuria in patients after heart or kidney transplantation [12,55,56].
In the very recent ‘Spare the Nephron liver study’ (A. Sebastian et al., unpublished data) (preliminary analysis), immunosuppression was started with MMF + CNIs. CNIs were replaced by sirolimus in the study group 54 days (mean) post-transplant. After 12 months, renal function improved significantly compared with the CNI-containing regimen; however, severe rejections were more frequent in the sirolimus-containing group. Furthermore, 31% of the patients did not tolerate MMF + sirolimus because of severe side-effects, compared with 15% in the MMF + CNI group. Similar side-effects of MMF and sirolimus, like bone marrow suppression, anemia and hyperlipidemia, are a disadvantage of this combination. Also, reduced compliance because of gastrointestinal side-effects of MMF may add to higher rejection risks. This is supported by a recent abstract (H. Sollinger, unpublished data) where compliance and rejection were investigated retrospectively in 1700 kidney transplant recipients receiving either CellCept or Myfortic. Drug discontinuation and dose reduction occurred more often in the CellCept group; this was paralleled by a higher risk of rejection.
Taken together, total avoidance of CNIs is not without risk. This regime – or at least CNI delay – should be considered for patients with impaired renal function at the time of transplantation. However, only patients with CNI nephropathy benefit from CNI withdrawal concerning renal function. A total avoidance of CNIs from the beginning of immunosuppression (not only a delay) would be desirable concerning nephropathy. However, an immunosuppressive regimen based only on MMF and steroids for all patients appears too risky. This supports the idea of an individualized, tailor-made immunosuppressive regime, where renal function has to play a very important role in the decision making. As shown by the studies by Yoshida et al. , Soliman et al.  and the recent, however preliminary results from the ReSpECT study (A. D. Mayer, unpublished data) CNI delay seems to be a reasonable approach – in combination with induction therapy – to prevent patients at risk for renal failure from the need for renal replacement therapy.
On the other hand, induction therapy with long-lasting effects like alemtuzumab is not without risk [57–60]. Thus, it should be considered to use only moderate induction therapy in severely sick patients with higher MELD scores.
In the future, new diagnostic tools (e.g. metabolic profiling) should be developed to identify patients at risk of renal failure earlier, before probably irreversible structural damages occur. Different studies showed that immunosuppressant-induced changes of metabolite patterns in urine were associated with a combination of changes in glomerular filtration, changes in secretion/absorption by tubulus cells, and changes in kidney cell metabolism [61,62]. A combination of these biomarkers including urine metabolites could probably provide this information in the future.
Until then, especially patients with higher MELD scores have to be regarded at risk for renal dysfunction after LTx. For those patients, CNI delay is recommended. To quantify and detect renal dysfunction better, CrCl and quantitative measurement of proteinuria should be documented before LTx and during follow up. Patients developing a CNI nephropathy during maintenance immunosuppression thus can be detected early. Then a kidney biopsy should be performed and taken into account to prove CNI nephropathy.