Pathogen detection by targeted next-generation sequencing test in adult hematological malignancies patients with suspected infections

Background Infections in patients with hematological malignancies (HM) are a significant cause of morbidity and mortality. Timely and effective empirical anti-infective treatment can reduce the infection-related mortality rate. Targeted next-generation sequencing (tNGS) offers a rapid diagnostic approach for identifying diverse pathogens in these patients. However, relevant research is still limited to adult patients with HM. Methods We conducted a retrospective analysis of adult HM patients admitted to our hospital from March 2023 to September 2023, focusing on their clinical characteristics and the results of both tNGS and conventional microbiological tests (CMTs). We evaluated the performance of tNGS and CMTs in pathogenic diagnosis and described the distribution characteristics of pathogens in adult HM patients with infections. Results The study included 209 samples collected from 137 patients. Results showed that the overall pathogen detection rate differed significantly between tNGS and CMTs (60.3% vs. 24.4%, p < 0.001). The sensitivity (69.7% vs. 35.9%), negative predictive value (NPV) (48.2% vs. 42.4%), and accuracy (66.5% vs. 56.5%) of pathogen detection were notably superior with tNGS compared to CMTs. Among the 142 samples with clinically diagnosed infections, tNGS combined with CMTs identified a definite or probable microbial etiology in 114 samples (80.3%). Of the 36 samples with concordant positive results from both tNGS and CMTs, 72.2% (26/36) exhibited full or partial agreement. Our study further showed the highest detection rate for viral infections (57.0%), predominantly for Epstein–Barr virus (DNA-V, 18.3%). Followed by bacterial infections (46.5%), the detection rate of Gram-negative bacteria (G + , 35.9%) was higher than that of Gram-positive bacteria (G − , 21.8%) in this study. Klebsiella pneumoniae (G − , 12.7%) had the highest detection rate among these emerging bacteria, followed by Pseudomonas aeruginosa (G − , 10.6%) and Enterococcus faecium (G + , 7.7%). Bacterial-viral coinfections were the most common type of mixed infection (35.5%). Conclusion In conclusion, tNGS outperforms CMTs in both sensitivity and pathogen spectrum. Therefore, it can serve as an adjunct to CMTs to facilitate the precise adjustment of anti-infective regimens for adult HM patients. Our findings establish a basis for formulating empirical anti-infective therapy strategies tailored to the pathogen distribution in this patient population.