The calcineurin inhibitor tacrolimus (Tac), when combined with mycophenolate mofetil and glucocorticoids, constitutes the standard immunosuppressive regimen for kidney transplant recipients [1, 2]. Tac is characterised by a narrow therapeutic index: low exposure is associated with higher risk of rejection [3] whereas overexposure leads to acute toxicity [4, 5] as well as to chronic graft injury [6]. In addition, Tac metabolism is subject to substantial intra- and inter-individual variability [7]. In practice, this translates into the necessity to carefully and regularly monitor Tac exposure in transplanted patients. Tac trough level (C0), being well correlated with 24 h-Tac Area Under Curve (AUC), is usually considered as a valid and clinically useful proxy for Tac exposure [8] even though direct evaluation of Tac AUC is recommended in specific situations [9, 10].

With the aim to better capture the clinical impact of Tac metabolization, a novel metric has recently emerged, the ratio between Tac C0 and Tac daily dose, also known as C0/D ratio [11]. Low C0/D ratio (usually below 1.05) identified so- called “fast metabolizer” patients who require higher Tac dose to achieve the desired C0 [11–13]. Retrospectively, the fast metabolizer status has been associated with lower graft function at 2 and 5 years post-transplantation [8], lower graft survival [12], increased risk of rejection [14], higher incidence of BK nephropathy [15] and histological lesions of interstitial fibrosis/tubular atrophy [16].

While a low C0/D ratio is linked to poorer outcome in kidney transplantation, the precise determinants driving this effect remain largely elusive. In this regard, a detrimental role of Tac maximal concentration (Cmax) could be discussed. Fast metabolizers receive higher doses of tacrolimus and might be exposed to higher Cmax levels (for the same C0) than non-fast metabolizers. This was actually reported in patients with the CYP3A5*1 allele, an allele that confers accelerated Tac metabolization [17, 18]. In the same line, Thölking and colleagues have observed higher C2 Tac blood concentrations in fast metabolizer patients, along with higher renal tubular cells toxicity [19].

While those data suggest a detrimental role of Tac Cmax, no study has specifically assessed the link between exposure to elevated Tac Cmax and graft outcome. We thus sought to retrospectively investigate the association between Tac Cmax, C0/D ratio and graft outcome in a cohort of kidney transplant patients who had Tac abbreviated pharmacokinetics as part of their routine follow-up.

In this study, we took advantage that Tac PK monitoring is routinely realized in our centre, to evaluate a possible detrimental effect of high Cmax exposure on transplant long term outcome. The main result of our study is that high Cmax exposure is significantly associated to decreased graft survival and that this detrimental effect is at least partially independent of Tac C0 and even of Tac AUC.

In the present study, we have first observed - consistently with previous reports- that fast metabolizers transplant patients exhibit significantly higher Cmax compared to slow/normal metabolizers [19]. It was thus tempting to speculate that elevated Cmax was responsible, at least partially, for the poor transplant outcome observed in recipients with low C0/D ratio, keeping in mind that several experimental and clinical data support a model of Tac Cmax induced nephrotoxicity [19]. However, as others previously reported, we were not able to identify a direct association between Cmax and transplant outcome. We believe that this apparent discrepancy can be explained by the fact that high Cmax actually reflects two different clinical situations. First, a scenario in which Tac overexposure is concomitant to high C0. In this scenario, exposition to high Cmax is obvious and likely to be rapidly corrected by decreasing Tac dosing. Second, a scenario in which elevated Cmax coexists with an apparently “normal” Tac C0 (i.e., within the expected range and thus falsely reassuring). In this scenario, elevated Cmax is bona fide not detectable and is likely to remain uncorrected. In consequence, testing the hypothesis of a direct toxicity of Tac Cmax may require to preferentially select the subpopulation of patients with a “high Cmax–normal C0” profile.

By re-expressing Cmax as Cmax/C0 ratio, we intended to better decipher the clinical impact of elevated Cmax, the hypothesis being that higher Cmax/C0 ratio confers higher risk. We did observe a statistically significant and independent association between Cmax/C0 and graft survival. In our cohort, an elevated Cmax/C0 clearly appears to be a marker of poor outcome suggesting that Tac Cmax overexposure could be deleterious (with a positive trend irrespective of the C0/D ratio, Supplementary Figure S3). While patients with high Cmax/C0 did not have more histological lesions related to Tac nephrotoxicity, they were significantly more likely to develop BK-associated nephropathy. The exact mechanisms underlying Cmax toxicity will however need to be thoroughly explored in larger and more appropriately designed studies.

Tac PK varies according to Tac formulation [8, 18, 20] and may impact Cmax/C0 ratio. In our study, we did not find any difference in Cmax alone or Cmax/C0 between different Tac formulations but the large majority of our patients (almost 75%) received conventional extended released-Tac. It is known that patients under the LCP formulation have significantly lower Cmax (−31%) especially when they expressed CYP3A5 fast metabolizer alleles [16]. In a retrospective proof of concept study [21], when patients were switched from intermediate release- to LCP formulation early after transplantation, fast metabolizers patients showed a noticeable and persistent increase of eGFR (+4.15 (mL/min/1,73 m²) whereas eGFR in slow metabolizers remained stable. While this was not directly evaluated in the present study, our data suggest that a lower Cmax exposure following the switch to LCP Tac might have played a role in renal graft function improvement. This speculation will need to be formally tested in a prospective and controlled manner, especially given that LCP-formulation has so far not demonstrated any consistent benefit over conventional extended-released formulation of Tac.

Interestingly, Tac Cmax/C0 and Tac AUC were poorly correlated (Supplementary Figure S1). Additionally, we did not observed a significant association between high Tac AUC and graft survival (Supplementary Figure S2). Together, those observations suggest that the detrimental effect of high Cmax/C0 is at least partially independent of Tac AUC.

Beyond the impact of Cmax, accumulation of Tac metabolites are also potential factors that could negatively influence graft outcome, especially in fast metabolizers [22, 23]. The increased Tac dosing which is required in fast metabolizers could result in intra renal metabolites accumulation responsible for cellular damages. This hypothesis is currently being explored in the TIPS study (ClinicalTrials.gov Identifier: NCT04526431).

Cmax value was defined in our study as the higher Tac concentration analysed between 2 predefined time points, namely, 90 and 180 min. Strictly speaking, this does not correspond to the real Cmax which would require a full PK analysis, a procedure hardly implementable in clinical practice. However, considering the PK profile of Tac [8, 24], we can reasonably assume that Cmax values determined from our abbreviated PK were close to the real Cmax. Of note, the pharmacokinetic status of each patient was determined from the median Tac concentrations measured at month 3 and month 12. We acknowledge that this relies on the assumption that the pharmacokinetic status is stable over this time-period, which was actually verified for a majority (81%) but not for all our patients.

Our study has other limitations, the first being its retrospective, monocentric nature. The absence of a systematic protocol biopsies policy in our centre may have also accounted for the low rate of histologically documented calcineurin inhibitor toxicity and our inability to demonstrate the histological impact of Tac overexposure that is usually expected [16]. Additionally, physicians had access to Cmax values which could have influenced subsequent prescriptions in adjusting Tac dosing. Finally, Tac PK results did not include AUC measurements. Tac AUC is considered the most reliable parameter to assess Tac exposure [9].

To the best of our knowledge, our study is the first one to have analysed the relationship between metabolization status and Tac peak concentration in regard with transplant outcome. Of note, all Tac measurements were done in the same laboratory at university hospital of Saint Etienne using a HPLC-MS-MS method, limiting biases due to the dosage technique. Another strength of our study lies in its relatively long median follow-up time of 8 years.

In conclusion, our data suggest that high Tac Cmax exposure is independently associated with poorer graft survival and support the hypothesis that fast metabolizers exhibit higher Tac Cmax even in situations where Tac C0 is not elevated.

Tac formulations prone to mitigate pharmacokinetic peak could thus be particularly beneficial to fast metabolizers kidney transplant patients and should be thoroughly evaluated in this indication.