The Epidemiologic Burden of Tacrolimus Variability among Kidney Transplant Recipients in the United States

Shah, Pratik B.
Ennis, Jennifer L.
Cunningham, Patrick N.
Josephson, Michelle A.
McGill, Rita L.

^aSection of Nephrology, University of Chicago, Chicago, Illinois, USA
^bLaboratory Corporation of America, Chicago, Illinois, USA

^*Pratik B. Shah, MD, Section of Nephrology, University of Chicago, 5814 South Maryland Avenue, MC 5100, Chicago, IL 60637 (USA), E-Mail [email protected]

Received June 20, 2019; Accepted September 03, 2019; previously published online September 25, 2019

American Journal of Nephrology 50(5):p 370-374, November 2019. | DOI: 10.1159/000503167

Abstract

Background:

Within-patient tacrolimus level variability >30% has been shown to be a risk factor for de novo donor-specific antibody formation and death-censored graft failure among kidney transplant recipients. The burden of tacrolimus variability and the correlation between variability and subtherapeutic tacrolimus levels were examined in a large national data set.

Methods:

All tacrolimus levels drawn at LabCorp® facilities in the United States with a diagnosis code for kidney transplant between November 2011 and September 2017 were examined, excluding values that could represent new allografts. Tacrolimus variability was calculated if at least 3 levels were available. The percentage of subtherapeutic (<4.0 ng/dL) tacrolimus levels (%subT) was also calculated. Interdependence between %subT and tacrolimus variability was assessed with correlation analysis and linear regression.

Results:

There were 410,257 tacrolimus levels among 27,375 patients, who had 11 (interquartile range [IQR] 6-20) tacrolimus levels over a median follow-up of 26.5 (IQR 12.8-46.1) months. Median tacrolimus variability was 30.6%, and 51.6% of patients exceeded 30% variability. Median %subT was 11.1% (IQR 0-30.8%), and 34.3% of patients had no subtherapeutic levels. The correlation coefficient between tacrolimus variability and %subT was 0.253 (p< 0.001). In linear regression, tacrolimus variability increased 1.86% for each 10% increase in %subT (p < 0.001), but R-squared for this model was only 0.06.

Conclusion:

More than half of established kidney transplant patients from a large national sample exhibited levels of tacrolimus variability that have been associated with inferior transplant outcomes. Tacrolimus variability has a weak association with subtherapeutic levels, but represents a more complicated constellation of clinical factors.