ONETest PathoGenome: A Multi-Cohort Evaluation of an Optimized NGS Assay for Detection of Lower Respiratory Pathogens in Bronchoalveolar Lavage

The ONETest PathoGenome assay, an optimized hybrid-capture metagenomic NGS test, demonstrates high sensitivity and specificity for detecting lower respiratory pathogens in bronchoalveolar lavage samples, significantly improving diagnostic yield and turnaround time compared to conventional culture and whole-metagenome shotgun sequencing.

The Gist

Imagine you are a detective trying to solve a mystery in a very crowded, noisy room. The room is a patient's lungs (specifically, a sample called Bronchoalveolar Lavage, or BAL), and the mystery is: What is making this person sick?

Usually, detectives (doctors) use two main tools:

1. The "Grow-It" Method (Culture): They take a swab and try to grow the bacteria in a petri dish. It's reliable, but it's slow. Some bad guys are too shy to grow, or they've already been scared off by antibiotics.
2. The "Wanted Poster" Method (PCR): They check for specific names on a list. It's fast, but if the bad guy isn't on the list, they miss them.

The Problem: Sometimes, the "Grow-It" method takes too long, and the "Wanted Poster" list is too short. The patient gets sicker while waiting.

The New Solution: Enter ONETest PathoGenome. Think of this as a high-tech, super-smart search engine for the lungs. Instead of waiting for bacteria to grow or checking a short list, it reads the genetic "DNA fingerprints" of everything in the sample to find the bad guys.

The Challenge: The "Hostile Crowd"

The lungs are full of human cells (the "host"). If you try to read the DNA of the whole room, 99% of what you find is just human stuff. The bad bacteria are hiding in the crowd, like a few spies in a stadium. Traditional methods (called Whole-Metagenome Shotgun Sequencing) try to read everything, but it's like trying to find a needle in a haystack by reading every single piece of hay. It's expensive, slow, and the "needle" (the bacteria) often gets lost in the noise.

The Innovation: The "VIP Pass"

The ONETest assay is like giving the spies a VIP Pass.

• How it works: The test uses millions of tiny "bait" hooks (probes) designed to grab only the DNA of known respiratory germs (bacteria, fungi, viruses) and ignore the human DNA.
• The Result: It filters out the crowd. Instead of reading 100% human DNA, the test focuses 26 times more energy on the actual germs. It's like turning down the volume on the stadium crowd so you can clearly hear the spies whispering.

The Investigation: What Did They Find?

The researchers tested this new tool on three groups of patients (cohorts) to see how it compared to the old methods.

1. The "Sensitivity" Test (Can it find the bad guys?)
They spiked fake bacteria into clean lung fluid to see how little of a germ the test could find.

• The Verdict: It found bacteria even when there were very few of them (about 23,000 per milliliter). It's sensitive enough to catch the "shy" germs that culture methods often miss.

2. The "Head-to-Head" Test (vs. The Old Way)
They compared ONETest against the standard "Grow-It" culture method in 360 patients.

• The Match: When the old method found a germ, ONETest found it 87% of the time.
• The Bonus: In 21% of cases, ONETest found germs that the old method completely missed.
  • Analogy: Imagine the old method found a thief in a bank. ONETest found that thief plus two other thieves hiding in the back room that the security guard didn't see.
• The "False Alarm" Check: Sometimes, finding too much is bad. If you find a germ, is it causing the sickness, or is it just a harmless visitor (colonization)? The test uses smart math to filter out the "noise" (harmless background germs) so doctors don't treat patients for things that aren't actually making them sick.

3. The "Speed" Test

• The Verdict: While a culture test might take 3 to 5 days to grow a germ, ONETest can give results in less than 24 hours. This is a game-changer for doctors who need to know now which antibiotic to give.

The Bottom Line

Think of ONETest PathoGenome not as a replacement for the old methods, but as a super-charged sidekick.

• The Old Way (Culture): Reliable but slow; misses the shy or fastidious germs.
• The New Way (ONETest): Fast, broad, and sensitive. It finds the germs the old way misses and does it overnight.

Why does this matter?
For patients with severe pneumonia, especially those with weak immune systems (like transplant patients), knowing exactly what is infecting them quickly can mean the difference between life and death. This test helps doctors stop guessing and start treating with the right medicine much faster, while also avoiding unnecessary antibiotics for harmless germs.

In short: It's like upgrading from a magnifying glass to a high-powered, noise-canceling microscope that can spot the invisible enemies in your lungs before they take over.

Technical Summary

1. Problem Statement

Lower Respiratory Tract Infections (LRTIs) are a leading cause of mortality, yet diagnosing the causative pathogen remains challenging.

• Limitations of Current Standards: Conventional culture methods are slow, have low sensitivity (especially after antibiotic exposure), and fail to detect non-culturable or fastidious organisms. PCR panels offer speed but are limited by predefined targets, missing novel or unexpected pathogens.
• Limitations of Existing NGS: Whole-metagenome shotgun sequencing (WmGS) offers broad coverage but suffers from high host-background noise (human DNA dominates the reads), high cost, computational complexity, and difficulty in distinguishing true pathogens from commensals or contaminants in non-sterile samples like Bronchoalveolar Lavage (BAL).
• The Gap: There is a need for an automated, targeted metagenomic sequencing (TmGS) assay that improves signal-to-noise ratios, reduces sequencing burden, and provides actionable, contamination-aware results for LRTIs.

2. Methodology

The study evaluated ONETest™ PathoGenome (OT), an automated, hybrid-capture TmGS assay designed for BAL fluid.

• Assay Design:

  • Technology: Uses a 6.2 million probe "QuantumProbe" (QP) panel targeting core-genome loci across 50 microbial families (>250 respiratory pathogens).
  • Workflow: Fully automated DNA-to-results workflow (Fusion Genomics) with integrated bioinformatics (FusionCloud) that applies species-specific thresholds to distinguish signal from background/contamination.
  • Comparison: Benchmarked against WmGS, routine bacterial/acid-fast bacillus (AFB) culture, and an amplicon-based NGS comparator (MicroGenDx/MGDx).

• Study Cohorts:

  • Cohort 1 (Technical Performance, n=119): Compared OT vs. WmGS on BAL specimens to assess diversity preservation, host-read reduction, and on-target signal enrichment.
  • Cohort 2 (Clinical Accuracy, n=360): Retrospective clinical trial benchmarked against routine bacterial and AFB culture. Included orthogonal validation via BioFire Pneumonia Panel and MGDx.
  • Cohort 3 (LDT Validation, n=43): Laboratory-developed test validation benchmarked against culture to support regulatory implementation.
  • Analytical Performance: Limit of Detection (LoD) assessed using whole-cell spike-ins of 6 organisms (5 bacteria, 1 fungus) into culture-negative BAL.

• Statistical Analysis:

  • Metrics included sensitivity, specificity, positive/negative predictive values (PPV/NPV), and Cohen's $\kappa$ for agreement.
  • Diversity was measured using Shannon index; signal enrichment was measured by FPKM and genome coverage.

3. Key Contributions

• Development of a Hybrid-Capture Panel: Introduction of a 6M-probe panel specifically optimized for core-genome enrichment of respiratory pathogens, balancing broad coverage with high specificity.
• Context-Aware Reporting: Implementation of a proprietary statistical framework using species-specific background baselines (derived from negative controls and historical data) to mitigate false positives from commensals and reagent contaminants.
• Multi-Cohort Validation: A comprehensive evaluation spanning analytical sensitivity, technical comparison with WmGS, and clinical performance across three distinct cohorts, including a CAP-accredited LDT validation.
• Differentiation from Amplicon Sequencing: Direct comparison demonstrating the superiority of hybrid capture over amplicon-based methods (MGDx) for recovering culture-confirmed pathogens in complex BAL samples.

4. Key Results

Analytical Performance (LoD)

• Sensitivity: The assay achieved an operational LoD of 23,000 CFU/mL for all six spiked organisms (including Mycobacterium kansasii, Pseudomonas aeruginosa, and Aspergillus flavus) in BAL matrix.
• Reproducibility: 100% reproducibility across intra-assay and inter-assay runs.

Technical Performance (Cohort 1: OT vs. WmGS)

• Signal Enrichment: OT increased normalized on-target microbial abundance 26-fold compared to WmGS while preserving within-sample microbial diversity (Shannon index correlation $R=0.74$).
• Host Reduction: OT reduced the human-read fraction by 4.2% compared to WmGS at matched library sizes.
• Pathogen Detection: OT yielded significantly higher per-pathogen signal (e.g., 61-fold for P. aeruginosa, 56-fold for S. pneumoniae) compared to full-depth WmGS, enabling detection of organisms missed by WmGS (e.g., M. kansasii).

Clinical Performance (Cohort 2: n=360)

• Agreement with Culture:
  • OT detected at least one culture-confirmed organism in 87% (100/115) of culture-positive specimens.
  • Species-level Sensitivity: 80%; Specificity: 99%.
  • Sample-level Sensitivity: 83%; Specificity: 69% (lower specificity is expected in non-sterile BAL due to colonization vs. infection ambiguity).
• Additive Yield: OT identified 21% (76/360) additional specimens with ≥1 pathogen not found by culture.
  • Orthogonal Support: 7.5% (27/360) of specimens had additional findings supported by orthogonal testing (BioFire or MGDx).
• Comparison with MGDx: In a subset (n=66), OT detected 86% of culture-confirmed isolates vs. 54% for MGDx, demonstrating higher sensitivity for culture-confirmed targets.

LDT Validation (Cohort 3: n=43)

• OT identified ≥1 culture-confirmed organism in 85% (34/40) of culture-positive specimens.
• Overall accuracy: 84%; Sensitivity: 85%; Specificity: 67% (limited by small number of culture-negative controls).
• Performance was robust for common bacteria (P. aeruginosa, A. xylosoxidans) but showed variability for fungi and NTM (Non-Tuberculous Mycobacteria), largely attributed to low organism burden and sample volume constraints.

5. Significance and Conclusion

• Clinical Utility: ONETest PathoGenome serves as a powerful adjunct to conventional microbiology, particularly for immunocompromised patients, those with prior antibiotic exposure, or cases with polymicrobial/slow-growing infections.
• Efficiency: By using hybrid capture, the assay reduces sequencing burden and host background while maintaining broad pathogen coverage, making it more feasible for routine clinical implementation than WmGS.
• Interpretation Framework: The study emphasizes that NGS results in BAL must be interpreted with "contamination-aware" thresholds. The assay's ability to distinguish true pathogens from background noise is critical to preventing unnecessary antimicrobial treatment.
• Future Impact: The data supports the use of OT to expand microbiologic findings, detect co-infections, and identify slow-growing organisms that conventional culture misses. The authors call for prospective studies with adjudicated clinical endpoints to further refine the distinction between colonization and causative infection.

In summary, the paper validates ONETest PathoGenome as a high-performance, automated TmGS assay that improves pathogen detection in BAL fluid, offering a faster and more comprehensive alternative to culture and PCR, with a specific advantage in signal enrichment over WmGS.