1273P - The prognostic value evaluation of a tissue Comprehensive Genomic Profiling (CGP)-informed personalized MRD detection assay in NSCLC

Background

ctDNA-based molecular residual disease (MRD) detection of postoperative plasma enables refining risk stratification to guide treatment precisely. The clinical efficacy of MNavigator V1 utilizing fixed panel detection scheme has been confirmed and published. Here, we reported its V2 improved version performance.

Methods

MNavigator V2 strategy combines the advantages of two classic technologies, WES-based personalized and non-customized off-the-shelf assay. In this assay, tissue somatic mutations were identified by CGP-sequencing with a 2.3 Mb 1021-gene panel, and then the top-ranked 20 mutations were selected. Tissue-specific customized panel combined with a lung universal core panel to monitor MRD by sequencing depth > 100,000x. The core panel can detect de novo mutations in ctDNA, especially targeted treatment-resistant mutations. Analytical validation was performed on a commercial MRD reference standard set, with VAF down to 0.005%. The retrospective cohort included 33 NSCLC patients (pts) treated with definitive surgery. Head-to-head compared MRD detection of this assay with MNavigator V1 that uses a 338-gene panel with a sequencing depth of 30,000x.

Results

This assay demonstrated the sample-level LOD of 0.005% by monitoring 20 variants, with a specificity of 100%. 10,625 NSCLC 1021panel sequencing data indicates 99% of pts available monitor 2 mutations, achieving 97.3% analytical sensitivity at 0.02% VAF. In the cohort, MRD was detected in 14/15 relapsed pts by both assays, apart from a brain-only metastasis pt. Recurrence-free pts with 2 years of follow-up were MRD-negative, yielding a specificity of 100% (18/18). MRD-detectable population was associated with a worse prognosis (HR=35.4). MNavigator V2 predicted relapse with a lead time of 143 days, 15 days ahead of V1. This may be related to more traced mutations (average 10 vs 5) and deeper sequencing depth, improving the capability to detect minimal tumor DNA residuals.

Conclusions

The CGP-based personalized MRD assay showed high analytical performance, and clinical assessment indicated superior sensitivity than fixed panel assays, providing greater potential for clinical benefit.

Clinical trial identification

Editorial acknowledgement

Legal entity responsible for the study

W. Gao.

Funding

Has not received any funding.

Disclosure

All authors have declared no conflicts of interest.