Meta-Analysis of Diagnostic Performance of Next Generation Sequencing of Plasma Microbial Cell-Free DNA for Infections in Hematopoietic Stem Cell Transplant Patients

Abstract

Introduction Hematopoietic stem cell transplant (HSCT) recipients experience periods of profound deficiency in both innate and adaptive immunity putting them at risk for a wide spectrum of infections, including organisms which are not normally pathogenic in immunocompetent hosts. Next generation sequencing (NGS) of plasma microbial cell-free DNA (mcfDNA) allows non-invasive untargeted diagnosis of human pathogens, making this modality appealing for this patient population (1). Here we perform a meta-analysis to evaluate the diagnostic value of NGS of plasma mcfDNA for infections in HSCT recipients. Methods We searched for relevant articles in BASE, PubMed, and ClinicalTrials.gov from inception to May 2022. Studies were eligible for inclusion if they assessed the diagnostic performance of NGS of plasma mcfDNA, included HSCT recipients with individual level data, and provided sufficient data to construct a two-by-two table, and were excluded if they were case reports or studies which only focused on the diagnosis of only one set of microorganisms. Two clinical adjudicators independently assigned the plasma mcfDNA test result as true positive, true negative, false positive, or false negative. The mcfDNA results were analyzed as a whole rather than by each individual organism if there was more than one organism reported. These data were then synthesized using random effect models and statistical analysis provided by the metadta module within the STATA statistical software package. Results A total of 6 studies were included for the analysis, which included a total of 71 patients. The pooled sensitivity was 0.925 (97.5% CI 0.50-0.99) and the pooled specificity was 0.74 (0.47-0.90). The positive likelihood ratio, negative likelihood ratio, and diagnostic odds ratio were 3.59 (1.42-9.09), 0.10 (0.01-1.10), and 35.6 (1.9-677.1). Discussion NGS of plasma mcfDNA was effective in diagnosing infections in HSCT recipients. The high pooled diagnostic odds ratio found in this study suggests that NGS of plasma mcfDNA may have a unique role in this specific patient population, especially when the pre-test probability is high. The diagnostic accuracy of the test varies depending on host factors, preceding antimicrobial treatment, infection site, and the organism being evaluated. With a non-targeted test, it is possible that organisms are detected that may not be clinically significant, which would affect the test specificity, especially in populations with disrupted mucosal barriers, and therefore test results need to be carefully interpreted. The negative likelihood ratio of 0.10 suggests that there may be a role for negative tests as well, although this will need to be further explored. Overall, NGS plasma mcfDNA is a promising test for the diagnosis of infections in HSCT recipients.