# Metagenomic next-generation sequencing for diagnosis of immune checkpoint inhibitor-associated pneumonitis: a retrospective comparative clinical performance study

**Authors:** Ya-Lin Jiang, Shi-Zhen Dong, Yuan-Bo Xu, Jun-Li Fan, Yong-Mei Zhang, Shen-Shen Huang

PMC · DOI: 10.3389/fcimb.2026.1730022 · Frontiers in Cellular and Infection Microbiology · 2026-02-06

## TL;DR

This study shows that metagenomic next-generation sequencing is more accurate and faster than traditional methods in diagnosing immune-related lung inflammation versus infection in cancer patients receiving immunotherapy.

## Contribution

The study demonstrates mNGS's superior diagnostic performance and faster turnaround time for immune checkpoint inhibitor-associated pneumonitis compared to conventional microbiological testing.

## Key findings

- mNGS detected pathogens in 94% of infectious pneumonia cases versus 33% with conventional testing.
- mNGS had 88% sensitivity and 94% specificity for diagnosing CIP, outperforming conventional methods.
- mNGS results were available in 24 hours, compared to 121.5 hours for conventional testing.

## Abstract

To evaluate the diagnostic performance and clinical utility of metagenomic next-generation sequencing (mNGS) in distinguishing immune checkpoint inhibitor–related pneumonitis (CIP) from infectious pneumonia in cancer patients undergoing immunotherapy.

A retrospective tertiary hospital cohort included 34 cancer patients (Feb 2022–Jan 2024) with prior ICI exposure, new/worsening respiratory symptoms, imaging infiltrates, and both mNGS and conventional microbiological testing (CMT). Final diagnoses were adjudicated by a multidisciplinary panel. We compared pathogen detection rates, sensitivity, specificity, and turnaround times (TAT) between mNGS and CMT.

In the infectious pneumonia group, mNGS detected pathogens in 17/18 cases (94%), whereas CMT detected only 6/18 (33%). In the CIP group, mNGS was negative in 14/16 cases (88%), compared with 11/16 negatives by CMT (69%). Using the adjudicated diagnosis as the reference, mNGS showed sensitivity 88%, and specificity 94%. In contrast, CMT’s sensitivity was 69%, and specificity 33%. The median TAT for mNGS was 24 hours (IQR 22–31 h), versus 121.5 hours (IQR 80.5–156 h) for CMT (P < 0.001).

mNGS outperforms CMT in both diagnostic accuracy and timeliness for distinguishing CIP from infectious pneumonia among immunotherapy recipients. Incorporation of mNGS into the diagnostic workflow for suspected CIP may improve etiological discrimination and enable timely, individualized treatment. Further large-scale prospective studies are required to confirm these findings.

## Linked entities

- **Diseases:** pneumonitis (MONDO:0043905), cancer (MONDO:0004992)

## Full-text entities

- **Genes:** PDCD1 (programmed cell death 1) [NCBI Gene 5133] {aka ADMIO4, AIMTBS, CD279, PD-1, PD1, SLEB2}, CRP (C-reactive protein) [NCBI Gene 1401] {aka PTX1}, CTLA4 (cytotoxic T-lymphocyte associated protein 4) [NCBI Gene 1493] {aka ALPS5, CD, CD152, CELIAC3, CTLA-4, GRD4}, UROD (uroporphyrinogen decarboxylase) [NCBI Gene 7389] {aka PCT, UPD}, CD274 (CD274 molecule) [NCBI Gene 29126] {aka ADMIO5, B7-H, B7H1, PD-L1, PDCD1L1, PDCD1LG1}
- **Diseases:** interstitial abnormalities (MESH:D065167), infection (MESH:D007239), lung cancer (MESH:D008175), dyspnea (MESH:D004417), diabetes mellitus (MESH:D003920), radiation pneumonitis (MESH:D017564), lung disease (MESH:D008171), cancer (MESH:D009369), opportunistic infections (MESH:D009894), cough (MESH:D003371), cardiogenic pulmonary edema (MESH:D011654), mycobacterial (MESH:C564468), DM (MESH:D009223), lung injury (MESH:D055370), LRTIs (MESH:D012141), inflammation (MESH:D007249), respiratory symptoms (MESH:D012818), critically ill (MESH:D016638), Pneumocystis jirovecii (MESH:D011020), hypertension (MESH:D006973), Pseudomonas aeruginosa pneumonia (MESH:D011552), mNGS (MESH:D010855), fungal (MESH:D009181), Aspergillus infection (MESH:D001228), CAP (OMIM:115650), fever (MESH:D005334), pulmonary embolism (MESH:D011655), esophageal cancer (MESH:D004938), colonization (MESH:D003108), hypoxemia (MESH:D000860), Infectious (MESH:D003141), respiratory deterioration (MESH:D012131), lymphomas (MESH:D008223), cytomegalovirus (MESH:D003586), infiltrates (MESH:D017254), CMT (MESH:D013736), COPD (MESH:D029424), chest pain (MESH:D002637), interstitial lung disease (MESH:D017563), coronary artery disease (MESH:D003324), CIP (MESH:D011014), noninfectious (MESH:D000073296)
- **Chemicals:** CMT (-), water (MESH:D014867), steroid (MESH:D013256)
- **Species:** Staphylococcus aureus (species) [taxon 1280], Mycoplasmoides pneumoniae (Filterable agent of primary atypical pneumonia, species) [taxon 2104], Bacteria Latreille et al. 1825 (Bacteria stick insect, genus) [taxon 629395], Pseudomonas aeruginosa (species) [taxon 287], Haemophilus influenzae (species) [taxon 727], Chlamydia pneumoniae (species) [taxon 83558], Homo sapiens (human, species) [taxon 9606], Respiratory syncytial virus (no rank) [taxon 12814], Legionella pneumophila (species) [taxon 446], Streptococcus pneumoniae (species) [taxon 1313], Aspergillus (genus) [taxon 5052], Fungi (kingdom) [taxon 4751], Severe acute respiratory syndrome coronavirus 2 (no rank) [taxon 2697049], Pneumocystis jirovecii (species) [taxon 42068]

## Full text

_Full body text omitted from this summary view._ Fetch the complete paper as Markdown: https://tomesphere.com/paper/PMC12920478/full.md

## Figures

2 figures with captions in the complete paper: https://tomesphere.com/paper/PMC12920478/full.md

## References

27 references — full list in the complete paper: https://tomesphere.com/paper/PMC12920478/full.md

---
Source: https://tomesphere.com/paper/PMC12920478