# NGS coverage accurately predicts MET and HER2 (ERBB2) gene amplifications in a real-world non-small cell lung cancer cohort

## TL;DR

This study shows that next-generation sequencing (NGS) can accurately predict gene amplifications in non-small cell lung cancer, potentially replacing the traditional FISH method.

## Contribution

The study demonstrates that NGS coverage can reliably detect MET and HER2 gene amplifications in real-world lung cancer samples.

## Key findings

- NGS fold changes strongly correlate with FISH Gene/CEN ratios and gene copy number per cell.
- A fold change cutoff of 2.0 effectively distinguishes amplified from non-amplified cases.
- NGS achieves high predictive reliability for gene amplification detection across multiple genes.

## Abstract

Fluorescence in situ hybridization (FISH) is the current standard for detecting gene amplifications, yet its low throughput and practical constraints call for alternative methods. This study evaluates next-generation sequencing (NGS) as a potential tool for accurately predicting gene amplifications.

We analyzed 66 primary non-small cell lung cancer (NSCLC) samples, tested by both NGS and FISH. FISH was conducted to detect gene amplifications in MET in 26 samples, in HER2 (ERBB2) in 21 samples, in PIK3CA in 9 samples, and KRAS in 9 samples, with one tumor tested for both MET and ERBB2. NGS fold changes, reflected by gene coverage, were calculated as the ratio of the highest gene-specific coverage to the mean coverage across all genes.

Amplification was detected in 46 (68.7%) samples. NGS fold changes correlated strongly with FISH Gene/CEN ratios (Spearman’s ρ = 0.720, p < 0.001) and gene copy number per cell (Spearman’s ρ = 0.847, p < 0.001). Among FISH-negative cases, NGS fold change ranged from 0.57 to 1.95, while in FISH-positive cases, it ranged from 2.11 to 25.08.

NGS fold changes demonstrate significant correlation with FISH metrics, supporting NGS as a promising marker for gene amplification. A fold change cutoff of 2.0 effectively distinguishes amplified from non-amplified cases, with NGS achieving a high degree of predictive reliability across the tested genes.