CMV INFECTION

W. Bosch¹, M.G. Heckman², N.N. Diehl², J.A. Shalev³, W.C. Hellinger^1
1Infectious Diseases, Mayo Clinic Florida, Jacksonville/FL/UNITED STATES OF AMERICA, ²Biostatistics, Mayo Clinic Florida, Jacksonville/FL/UNITED STATES OF AMERICA, ³Transplant Department, Mayo Clinic Florida, Jacksonville/FL/UNITED STATES OF AMERICA

Body: Introduction During the era of effective antiviral prophylaxis, CMV disease has been inconsistently associated with increased mortality of liver transplant recipients (LTR). A study of the association of CMV infection or disease with death, or with the combined endpoint of graft loss or death within one year of first liver transplant (LT) in high risk CMV donor seropositive/recipient seronegative (D+/R-) recipients was undertaken. Methods We performed a retrospective cohort study of 240 CMV (D+/R-) LTR undergoing first LT between 2003 and 2008. All patients received prophylactic ganciclovir or valganciclovir through day 100 after LT. CMV infection and disease were managed using standardized treatment protocols. Potential risk factors for CMV infection, CMV disease, death and graft loss were collected from transplant center databases and medical records. CMV infection and disease were identified using consensus definitions and standardized surveillance. The Kaplan-Meier method was used to estimate the cumulative proportion of patients with CMV infection and CMV disease within the first year of LT. The associations of CMV infection or disease with death and with the combined endpoint of graft loss or death were evaluated using Cox proportion hazards models. Single variable models were utilized as well as multivariable models adjusting for potentially confounding variables. A Holm step-down adjustment for multiple testing was applied with p-values ≤ 0.025 considered to be statistically significant. Results CMV infection and disease occurred in 100 (48%) and 45 (21%) of the 240 CMV (D+/R) LTR. Forty-five (21%) patients died, and 56 (27%) patients reached the combined endpoint of graft loss or death, with a median follow-up of 40.5 months. At 3, 6, and 12 months following LT, the estimated cumulative proportion of patients who had developed CMV infection were 11.6%, 36.0%, and 44.1%, respectively. At 3, 6, and 12 months following LT, the estimated cumulative proportion of patients who had developed CMV disease were 0.9%, 15.3%, and 20.4%, respectively. In multivariable analysis, CMV infection was strongly associated with death (RR: 2.35, 95% CI: 1.27 – 4.33, P=0.0063) and graft loss or death (RR: 2.69, 95% CI: 1.52 – 4.77, P=0.0007). CMV disease was also associated with an increased risk of death (RR: 1.90, 95% CI: 0.91 – 3.97, P=0.087) and graft loss or death (RR: 2.21, 95% CI: 1.08 – 4.53, P=0.030) in multivariable analysis, however these results were not quite statistically significant. ConclusionsDespiteeffective antiviral prophylaxis through day 100, CMV infection and disease occur frequently in high risk CMV (D+/R-) LTR. Our results suggest that CMV disease within the first year after LT isassociated with an increased risk of death and with graft loss or death, however these associations remain uncertain and require further examination. CMV infection before the end of the first yearafter LT is associated with increased risk of death and with increased risk of graft loss or death. Investigation of additional measures to prevent CMV infection during the first year after LT in CMV(D+/R-) LTR is warranted.

Disclosure: All authors have declared no conflicts of interest.