Evaluation of Tacrolimus Levels and Risk of Graft-Versus-Host Disease, Relapse and Non-Relapse Mortality after Haploidentical Transplantation with Post-Transplant Cyclophosphamide

Abstract

Following haploidentical hematopoietic cell transplantation (haplo-HCT) with post-transplant cyclophosphamide (PTCy), the range of target tacrolimus blood levels is wide (5-15 ng/ml). This fact, combined with pharmacogenomic variation and effects of drug interactions, leads to significant variability in tacrolimus serum concentrations. In the setting of matched related or unrelated donor transplantation using conventional graft-versus-host disease (GVHD) prophylaxis, several published studies have correlated higher post-transplant tacrolimus levels with a lower risk of GVHD. Therefore, we hypothesized that early post-transplant tacrolimus levels following haplo-HCT might impact the incidence of GVHD and other post-transplant endpoints and that more precise tacrolimus level targeting may improve post-transplant outcomes. To this end, we analyzed the relationship between mean weekly tacrolimus levels and acute and chronic GVHD, relapse, non-relapse mortality (NRM) and survival in 158 consecutive patients with hematologic malignancies receiving haplo-HCT using PTCy, tacrolimus and mycophenolate for GVHD prophylaxis between 2013-2018. Transplanted patients had a median age of 51 years (range, 19-75) and comorbidity index was ≥3 in 97 (61%). Peripheral blood was the graft source in 142 (90%), and regimen intensity was myeloablative in 90 (57%). Mean weekly tacrolimus levels were calculated each week for the first 4 weeks starting on day +7 and ending day +35 post-transplant, as well as aggregate mean tacrolimus level for 2- and 4-week intervals. Table 1 depicts a summary of the univariate analysis of weekly mean tacrolimus levels and various clinical endpoints (hazard ratios are depicted for each 1 ng/ml increment in mean tacrolimus levels). In this analysis, there were no significant associations between higher mean weekly tacrolimus levels and acute GVHD, chronic GVHD, relapse, NRM, or survival. Using tertiles, mean tacrolimus levels during weeks 1-4 were divided into <10.9 ng/ml, 10.9-13 ng/ml, and ≥13 ng/ml subgroups. In univariate analysis, there were no significant differences between mean weekly tacrolimus level subgroups and cumulative incidence of acute GVHD, chronic GVHD, relapse or NRM (Figure 1). In conclusion, following haplo-HCT with PTCy, there were no significant associations found between serum tacrolimus concentrations and the incidence of GVHD or other post-transplant outcomes. Given the lack of an obvious detrimental effect of lower tacrolimus levels, combined with the known toxicities associated with tacrolimus use (i.e. renal, neurologic, cardiovascular, diabetes mellitus), we speculate that lower target tacrolimus serum concentrations may be preferable following PTCy-based haplo-HCT.