Metatranscriptomic Evaluation of Pulmonary Complications after Pediatric Hematopoietic Cell Transplantation Reveals Pathogenic Microbes Linked to Dysregulated Human Immunologic Gene Expression.

Introduction Pulmonary complications (PCs) after pediatric hematopoietic cell transplant (HCT) are heterogeneous in nature and are strongly associated with transplant-related mortality. Microbial pathogens play a role in the evolution of many post-HCT PCs and may contribute to or develop in the setting of an altered inflammatory microenvironment in the lungs. Objectives To improve characterization of the pulmonary microbiome and transcriptome in pediatric HCT patients with PCs. Methods Bronchoalveolar lavage (BAL) was performed at the University Medical Center in Utrecht between April 2004 and November 2016 for all children <18 years of age prior to undergoing HCT and again in select patients after HCT. In this study, n=257 BALs (n=210 obtained pre-HCT and n=47 obtained post-HCT) underwent metatranscriptomic sequencing to a median depth of 41.5 million read-pairs per sample followed by alignment to human and microbial genomes. Pathogenic bacteria and fungi were identified if RNA was quantified ≥2 standard deviations above the pre-HCT cohort mean (Z-score ≥2). Pathogenic and non-pathogenic viruses were noted when quantified ≥10 and ≥1 reads per million total reads, respectively. Diversity of the bacterial microbiome was considered abnormal if the Simpson's Diversity Index was ≤0.6. Differential gene expression was calculated using DEseq2 with FDR-adjusted p-values <0.05, followed by Gene Ontology analysis. Results Relative to pre-HCT patients, post-HCT patients were more likely to have abnormal quantities of RNA aligning to pathogenic microbes (59.6% vs. 40.5%, p=0.017), including RNA aligning to pathogenic bacteria (23.4% vs. 11.0%, p=0.023) and pathogenic viruses (48.9% vs 29.0%, p=0.009), but not pathogenic fungi (6.4% vs 7.1%, p=0.854). In addition, post-HCT patients were also more likely to have depressed bacterial diversity (23.4% vs. 10.0%, p=0.012) and to harbor non-pathogenic viruses (27.7% vs. 7.1%, p<0.001). Relative to pre-HCT patients without an identified pathogen, post-HCT patients with a pathogen demonstrated at least 396 differentially expressed genes in the biological processes of immunity (leukocyte transendothelial migration; neutrophil degranulation; T-cell receptor activation, IL-1, IL-10 and TNF/NF-κβ mediated signaling; regulation of type 1 IFN production) and cellular response to stress (regulation of HIF-1, response to hypoxia, and regulation of apoptosis and programmed cell death). Conclusions As detection of pathogenic pulmonary microbes is exceedingly common pre-HCT and increases after HCT, incorporation of host gene expression profiles along with standard diagnostic criteria for pulmonary complications may improve classification accuracy and facilitate patient-targeted therapies. Further investigation into the microbiologic and human transcriptional heterogeneity within these high-risk patients is required. [ABSTRACT FROM AUTHOR]