Post‐transplant monitoring of NK cell counts as a simple approach to predict the occurrence of opportunistic infection in liver transplant recipients

Background Monitoring of peripheral blood lymphocyte subpopulation ( PBLS ) counts might be useful for estimating the risk of infection after liver transplantation ( LT ). Methods We prospectively measured total lymphocyte and PBLS counts at baseline and post‐transplant months 1 and 6 in 92 LT recipients. PBLS were enumerated by single‐platform 6‐color flow cytometry technology. Areas under receiver operating characteristic ( ROC ) curves were used to evaluate the accuracy of different PBLS for predicting cytomegalovirus ( CMV ) disease and overall opportunistic infection ( OI ). Adjusted hazard ratios ( aHR s) for both outcomes were estimated by Cox regression. Results After a median follow‐up of 730.0 days, 29 patients (31.5%) developed 38 episodes of OI (including 22 episodes of CMV disease). The counts of CD 3 + , CD 4 + , and CD 8 + T cells, and CD 56 + CD 16 + natural killer ( NK ) cells at month 1 were significantly lower in patients subsequently developing OI . The NK cell count was the best predictive parameter (area under ROC curve for predicting CMV disease: 0.78; P ‐value = 0.001). Patients with an NK cell count <0.050 × 10 3 cells/μL had higher cumulative incidences of CMV disease ( P ‐value = 0.001) and overall OI ( P ‐value <0.001). In the multivariate models, an NK cell count <0.050 × 10 3 cells/μL at month 1 post transplantation remained as an independent risk factor for CMV disease ( aHR : 5.54; P ‐value = 0.003) and overall OI ( aHR : 7.56; P ‐value <0.001). Conclusion Post‐transplant kinetics of NK cell counts may be used as a simple and affordable proxy to the cell‐mediated immunity status in LT recipients and to their associated risk of OI .