Introduction: Tacrolimus is commonly used for prophylaxis of graft versus host disease (GVHD) after allogeneic hematopoietic cell transplant (HCT). Dosing strategies are weight-based, and trough levels are monitored for toxicity and efficacy. Tacrolimus metabolism is regulated by the CYP3A5 sub-family of the cytochrome P450 protein. Occurrence of the main single nucleotide polymorphism related to the functional variation of cytochrome P450 results in the absence of a functional CYP3A5 protein in patients who are homozygous carriers (CYP3A5*3/*3) versus patients with functional CYP3A5 (CYP3A5*1/*1 or CYP3A5*1/*3) [Niioka T et al, 2015]. Patients (pts) with the functional CYP3A5 protein metabolize tacrolimus rapidly; they require a higher drug dose to achieve the same therapeutic blood concentration compared to pts with the CYP3A5*1/*1 or *1/*3 genotypes [Rojas L et al, 2015; Terrazzino S et al, 2012]. We hypothesized that pts with *1/*1 or *1/*3 genotype (hyporesponders) will take a longer time to achieve therapeutic blood levels using weight-based dosing algorithms and potentially have an increased incidence of acute GVHD (aGVHD).

Methods: The current study is a single-institution, retrospective cohort analysis evaluating the effect of tacrolimus metabolism, as studied by CYP3A5 genotype, on incidence of aGVHD. All HCT recipients undergo Pharmacogenomic Resource for Enhanced Decisions in Care and Treatment (PREDICT) testing, which evaluates 184 SNPs within 34 genes relating to the metabolism of five drugs including tacrolimus (via CYP3A5 analysis). Adult pts 18 years and older were included if they received tacrolimus for GVHD prophylaxis after allogeneic HCT and had no evidence of moderate to severe hepatic dysfunction at baseline. Tacrolimus levels were monitored in a uniform manner and doses were adjusted based on levels. Data was collected by manual review of the pts' electronic medical record.

Results: Forty-six out of 73 pts screened met the inclusion criteria and were analyzed for the current study. Pts were grouped as either standard responders (39 patients) or hyporesponders (7 patients). Both groups were similar with respect to baseline demographics, disease and transplant-related factors, with the exception of a statistically significant difference in racial composition. There were no African-American pts in the standard responder group and 29% in the hyporesponder group (p=0.003), similar to prior studies [Birdwell K et al, 2012].

All pts received an initial tacrolimus dose of 0.045 mg/kg ideal body weight/dose by mouth (PO) twice daily, starting 3 days prior to HCT. Though not statistically significant, pts in the standard responder group achieved a therapeutic tacrolimus dose faster than those in the hyporesponder group (median 14 versus 16 days, p=0.382). The median total daily dose, in PO equivalents, ultimately required for pts who were hyporesponders was 0.125 mg/kg IBW (range 0.063-0.204), which is significantly higher than the median total daily dose required in the standard responder group, 0.070 mg/kg IBW (range 0.014-0.134), (p=0.001).

Acute GVHD, only chronic GVHD and no GVHD was seen in 21 (46%), 4 (8%) and 21 (46%) patients, respectively. Thirty-eight of 39 (97%) standard responders and 5 of 7 (71%) hyporesponders survived and maintained their graft to Day +100. Amongst these pts, aGVHD was seen in 18 (47%) pts (grades [gr.] 2-4, 13 patients; gr. 3-4, 2 patients) in the standard group and in 3 (60%) pts (gr. 2-4, 2 patients; gr. 3-4, 2 patients) in the hyporesponder group. Acute GVHD occurred earlier in hyporesponders versus standard responders (median 24 and 27.5 days, respectively) and was more severe. One hundred percent of pts with aGVHD in the standard responder group required systemic steroid therapy, versus 72% of standard responders (p=0.549).

Conclusion: Hyporesponders to tacrolimus require increasing doses to achieve similar therapeutic levels compared to standard responders, predisposing them to aGVHD with increased severity. There appears to be a racial disparity in the metabolism of tacrolimus. Additionally, standard responders ultimately required less than our standard initial tacrolimus regimen to achieve a therapeutic level, and may be at excess risk for toxicity when dosed by a standardized regimen. Validation of these findings in a larger, more racially diverse cohort is warranted; results may lead to genotype-driven tacrolimus dosing.


No relevant conflicts of interest to declare.

Author notes


Asterisk with author names denotes non-ASH members.