Alternative Viewpoint on Tacrolimus Concentration‐to‐Dose Ratios in Kidney Transplant Recipients and Relationship to Clinical Outcomes

Bartlett et al. reported in a large US-based single-center study that the tacrolimus metabolism rate, defined as the concentration-to-dose ratio (CDR), does not have an impact on biopsy-proven acute rejection as previously shown in kidney transplant recipients. In contrast to our results, they observed a higher mortality rate in nonrapid metabolizers (NRM) compared to rapid metabolizers (RM). Although estimated glomerular filtration rate values were lower at several time points in RMs compared to NRM, they vote against CDR-based tailoring of the tacrolimus therapy. First, Bartlett et al. defined RM and NRM differently compared to previous studies. Their CDR cut-off for RM/NRM was 2.04 (ng/ml)*1/ mg that is considerably higher compared to previous definitions [< 1.05 (ng/ml)*1/mg] 4 and < 0.67 (ng/ml)*1/mg. In accordance, they identified 74.3% RM compared to 39.1% and 25.2% RM. Hence, a substantial number of their RM would have been classified as NRM in the other studies. Notably, as kidney transplant recipients with a CDR between 1.05 and 1.54 (ng/ml)*1/ mg (intermediate group) showed comparable results to patients with a CDR ≥ 1.55 (ng/ml)*1/ mg we have even chosen to combine both groups in later analyses. Second, in contrast to European studies, Bartlett et al.’s study included 40% African Americans; ethnicity has a known impact. 7 Third, the low rejection rate (e.g., in contrast to Egeland et al.), might have masked differences as already stated by the authors. Different induction therapies and missing protocol biopsies, frequently detecting rejection in stable grafts, might be responsible. 8 Fourth, Bartlett et al. use a lower tacrolimus trough level compared with our data. 4 Lower troughs might mitigate tacrolimus adverse effects. 9 Fifth, the authors speculate that NRM have a higher exposure to tacrolimus over time, leading to increased toxicity, which is not supported by data. In fact, others showed comparable area under the receiver operating characteristic curve between RM and NRM. 11 A reanalysis of the study data using different CDR cut-offs would be interesting.