Introduction: High variability of tacrolimus trough levels in the early post transplantation period was shown to be associated with kidney graft loss.  The long term effects of exposure to cumulative drug level variability as well as interactions between high variability and exposure to values below and above the therapeutic range weren’t adequately evaluated, and the cutoff value of variability exposure that is associated with graft loss wasn’t determined
Patients and Methods: All tacrolimus drug values of patients that were transplanted in our center between the years 2006-2015 were collected and exposure to variability was calculated by time-weighted coefficient of variability (TWCV) defined as time-weighted standard deviation divided by the mean drug level. Time dependent univariate and multivariate Cox proportional hazard model was used with the primary outcome of patients and graft survival.  In order to find a cutoff value,the effect of different values of TWCV on survival (below20%, 20%-25%, 25%-30%, 30%-35% and above 35%) was evaluated.
Results: Between 1/1/2006 and 31/12/2015 1129 kidney transplantations were performed, of them 878 (77.8%) were included in the study with median follow of 1237 (181- 3649) days. While TWCV between 20%-25% was not associated with reduced survival, all ranges of TWCV above 25% were associated with significantly reduced graft survival . Patients with cumulative TWCV greater than 25% had increased rate of graft loss (Hazard ratio (HR) 3.66, 95% confidence interval (CI) 2.3 - 5.8, p<0.001)  the association remained highly significant after extensive multivariate adjustment. The interaction of high TWCV and exposure to low drug levels < 5 ng/ml was associated with reduced graft survival (HR 5.8, 95% CI 3.36-10.08 ) in comparison to patients with  TWCV below 25% and no exposure to drug levels< 5 ng/ml.
Conclusion: TWCV greater than 25% is a cutoff value predicting reduced graft survival. The combination TWCV greater than 25% and exposure to low drug levels < 5 is highly associated with an increased rate of graft loss .  Monitoring TWCV and exposure to sub-therapeutic drug levels can help detect the high risk patients. Further studies are needed to evaluate the effect of interventions intended to reduce variability on long term graft survival.