Risk factors for high level cytomegalovirus viremia in liver transplant recipients and associated outcomes

To evaluate epidemiology, risk-factors, and outcomes of high-level (HL) cytomegalovirus (CMV) viremia in liver transplant recipients.

Adult patients receiving a liver transplant between 1/1/2017 and 9/30/2020 were evaluated. Viral loads at University of Wisconsin Health Clinical Laboratories were required to allow for numerical comparison. Primary objective was incidence and outcomes of HL CMV viremia (viral-load > 100 000 IU/ml). Secondary objective was to elucidate risk factors to allow targeted interventions.

Two hundred nine patients met inclusion criteria; 175 kept their graft for at least 240 days. Of these nine patients developed HL CMV, 28 developed low-level (LL CMV, viral-load 250-100 000 IU/ml), and 138 did not develop CMV viremia. When comparing these three groups via classic statistical methods time from transplant to viremia was similar (HL 158 ± 77 days, LL 150 ± 76 days). Clinical factors were also similar with the exception of donor seropositivity (HL 87.5%, LL 70.4%, No CMV 49.6%, p = 0.025). HL CMV was significantly associated with graft loss (p < 0.0001) on Kaplan-Meier analysis; graft loss in the LL CMV group did not differ from the no CMV group (p = 0.96). To allow valid assessment of risk factors in the total study population (n = 209), models of time-varying covariates were used, and Cox proportional hazards ratios were calculated. In this analysis, HL CMV was associated with a significantly increased risk of graft loss (HR 5.6, p = 0.0016). When investigating risk factors associated with HL CMV, donor seropositivity significantly increased risk (HR 8.85, 95% CI 1.13-71.43, p = 0.038). Pretransplant total bilirubin (HR 1.04, 95% CI 0.998-1.07, p = 0.06) trended toward significance. Recipient seronegativity, liver disease, clinical and allocation model for end-stage liver disease (MELD), transplant surgery duration, age, sex, induction immunosuppression, and maintenance immunosuppression were not significantly associated with development of HL CMV.

HL CMV after liver transplant is uncommon but is associated with a significantly increased risk of graft loss that is not present in those patients who develop LL CMV or do not develop CMV viremia. Given these negative graft effects, CMV stewardship interventions targeting recipients of CMV seropositive allografts are warranted. Future larger scale studies evaluating the potential role of other factors in risk stratification are needed.