Lymphocyte Area Under the Curve as a Predictive Factor for Viral Infection after Allogenic Hematopoietic Stem Cell...

Viral infection is a serious complication that can greatly affect patient mortality and morbidity after allogenic hematopoietic stem cell transplantation (allo-HSCT). For the early identification of patients at high risk for viral infection, we evaluated the impact of lymphocyte area under the curve (AUC) value as a new predictive factor for early immune reconstitution after allo-HSCT against viral infection. This study included 286 patients who underwent their first allo-HSCT at Kyoto University Hospital between 2005 and 2017. Lymphocyte AUC from day 0 to day +15 was calculated in the analysis of human herpesvirus 6 (HHV-6), and lymphocyte AUC from day 0 to day +30 was calculated in the analysis of other viruses (cytomegalovirus [CMV], adenovirus, BK virus, JC virus, and varicella zoster virus). The risk factors for each viral reactivation/infection were assessed by multivariate analysis. The median age at transplantation was 51years (range, 17 to 68 years). The median lymphocyte AUC was 63/μL (range, 0 to 5620/μL) at day +15 and 3880 (range, 0 to 118,260/μL) at day +30. An increase in lymphocyte AUC was significantly associated with a high frequency of HHV-6 reactivation (P = .033) and a low frequency of CMV antigenemia (P = .014). No apparent association was found between lymphocyte AUC and reactivation/infection of other viruses. Aplastic anemia as a primary disease (hazard ratio [HR], 5.34; P < .001) and cord blood as a donor source (HR, 3.05; P = .006) were other risk factors for HHV-6 reactivation. Other risk factors for CMV antigenemia included the occurrence of acute graft-versus-host disease (HR 2.21; P < .001) and recipient age (HR 1.55; P = .017). Higher lymphocyte AUC at day +30 was significantly associated with low treatment-related mortality (HR, .47; P = .045). Lymphocyte AUC may be a good predictive factor for immune reconstitution against CMV reactivation. It also provides valuable information for predicting HHV-6 reactivation and treatment-related mortality.