Center-driven and clinically driven variation in US liver transplant maintenance immunosuppression therapy: A national practice patterns analysis

Background. Variation in the use of immunosuppression regimens after liver transplant has not been well described. Methods. Immunosuppression regimens used after liver transplant were identified in a novel database integrating national transplant registry and pharmacy fill records for 24 238 recipients (2006-2014). Bilevel hierarchical models were developed to quantify the effects of transplant program, recipient, and donor characteristics on regimen choice. Results. In the first 6 months after transplant, triple immunosuppression (tacrolimus, antimetabolite, corticosteroids) was the most common regimen (42.9%). By months 7 to 12, immunosuppression regimens were more commonly antimetabolite sparing (33.7%) or steroid sparing (26.9%), followed by triple (14.4%), mammalian target of rapamycin inhibitor (mTORi)-based (12.1%), or cyclosporine-based (9.2%). Based on intraclass correlation analysis, clinical characteristics explained less than 10% of the variation in immunosuppression choice, whereas program preference/practice explained 23% of steroid sparing, 26% of antimetabolite sparing, 28% of mTORi, and 21% of cyclosporine-based regimen use. Although case factors were not dominant practice drivers, triple immunosuppression in months 7 to 12 was more common among retransplant recipients and those with prior acute rejection. Hepatocellular carcinoma as cause of liver failure (adjusted odds ratio [aOR], 2.15; P<0.001), cancer within 6 months (aOR, 6.07; P<0.001), and 6-month estimated glomerular filtration rate less than 30 mL/min per 1.3 m ² (aOR, 1.98; P<0.001) were associated with mTORi use compared with triple immunosuppression in months 7 to 12, whereas acute rejection predicted lower use (aOR, 0.72; P=0.003). Conclusions. Liver transplant immunosuppression is dominantly driven by program preference, but case factors also affect regimen choice. This variation frames a natural experiment for future evaluations of comparative efficacy.