Prediction of severe infections in chronic lymphocytic leukemia: a simple risk score to stratify patients at diagnosis

Abstract

Chronic Lymphocytic Leukemia (CLL) is well-known for increasing susceptibility to infections. Factors such as immune dysregulation, IGHV status, hypogammaglobulinemia, and patient comorbidity and treatment, contribute to higher infection rates and mortality. However, the impact of hypogammaglobulinemia on infection rates is controversial. We aimed to identify clinical and biological parameters linked to the risk of severe infectious events. Additionally, we set up a straightforward risk infection score to stratify CLL patients at diagnosis, thereby enabling the development of suitable infection prevention strategies. We retrospectively evaluated 210 unselected CLL patients diagnosed between 1988 and 2018. This evaluation encompassed demographics, Binet stage, immunoglobulin (Ig) levels, treatment history, comorbidities, and IGHV mutational status at diagnosis. The frequency and severity of infectious events were recorded. Analysis revealed that age, IGHV mutational status, Binet stage, and hypogammaglobulinemia were statistically associated with the Time to First Infection (TTFI) in univariate and multivariate analyses. Using hazard ratios from the multivariate analysis, we finally devised a risk scoring system that integrated age, IGHV mutational status, immunoglobulin levels, and Binet stage to stratify patients at diagnosis based on their specific infection risk. In our cohort, disease progression and infections were the leading cause of death. These findings pointed out the clinical need for a screening process strategic for defining infectious risk at the time of CLL diagnosis, with a significant enhancement in the clinical management of these patients.

Chronic Lymphocytic Leukemia: Management of Adverse Events in the Era of Targeted Agents

“…However, current guidelines do not provide clear indications on patient stratification, either at diagnosis or before treatment initiation [ 12 , 195 ]. We proposed an easy-to-use scoring system to identify diagnosis patients at high risk of developing severe infections, based on disease stage, age, IGHV (immunoglobulin heavy chain variable region) mutational status, and hypogammaglobulinemia [ 196 ]. The CLL-TIM is a machine-learning model designed to identify patients at risk of infection within two years of a CLL diagnosis [ 197 ].…”