364: The Relationship between Tacrolimus Serum Concentrations and Acute Graft-Versus-Host Disease in Allogeneic Stem Cell Recipients

Tacrolimus is commonly used as prophylaxis of GVHD following allogeneic HSCT. While there is an association between increasing serum tacrolimus levels and the incidence of renal toxicity, there is no established relationship between tacrolimus levels and the incidence of GVHD. We studied the outcome of 92 consecutive allogeneic blood HSCT recipients receiving tacrolimus-based GVHD prophylaxis. 55 had received conventional-intensity conditioning (usually busulfan-fludarabine; 34 unrelated donors and 21 sibling donors) where tacrolimus was combined with short-course methotrexate, and 36 had received reduced-intensity conditioning (melphalan ± cyclophosphamide; all unrelated donors) where tacrolimus was combined with mycophenolate mofetil. For each patient, a weekly average tacrolimus level was determined for each of the first 7 weeks post-transplant by averaging all the levels available for that week. The relationship between the average weekly tacrolimus levels and the occurrence of any grade of acute GVHD was examined. The ANOVA test was used to compare the weekly average tacrolimus levels between patients who developed GVHD and those who did not. The cumulative incidence of acute GVHD was compared by tacrolimus level categories each week. A total of 1385 tacrolimus levels were available (6–45 per patient; median 14). The cumulative incidence of acute GVHD at 60 days was 37% (95% CI: 29–49%). As the table below shows, the average tacrolimus levels amongst patients going on to develop acute GVHD tended to be lower than those not developing acute GVHD; significantly so for week 4 and showing a trend towards significnace for weeks 5 and 6. The cumulative incidence of acute GVHD amongst patients whose week 3 average tacrolimus was <10 was 51% (95% CI: 39–68%) compared with 24% (95% CI: 14–41%) for those with a level of ≥10 (P = 0.015). Similarly, the cumulative incidence of acute GVHD amongst patients whose week 4 average tacrolimus was <10 was 43% (95% CI: 32–59%) compared with 20% (95% CI: 10–41%) for those with a level of ≥10 (P = 0.046). The relationship between higher acute GVHD incidence and lower tacrolimus levels was observed for other weeks and tacrolimus cut-off levels, but the differences were not statistically significant. Our data suggest that there is a relationship between tacrolimus levels and acute GVHD. How this affects relapse rates, transplant-related mortality, and survival remains to be determined.Tabled 1Median average tacrolimus levels in patients with or without acute GVHDWeekAcute GVHDNo acute GVHDP113.613.10.97211.112.10.2138.910.90.2047.38.50.03158.610.20.05767.89.10.0677.58.10.25 Open table in a new tab