Quantifying potential confounders of panel-based tumor mutational burden (TMB) measurement

Area of Science:

• Genomic medicine
• Cancer biomarker research
• Computational biology

Background:

• Tumor mutational burden (TMB) is gaining traction as a predictive biomarker for immune checkpoint blockade therapy.
• Accurate measurement of TMB via panel sequencing (psTMB) is critical, yet influenced by multiple factors.
• Existing literature lacks an integrated analysis quantifying these psTMB measurement confounders.

Purpose of the Study:

• To comprehensively quantify, compare, and combine all potential confounders affecting panel sequencing-based TMB (psTMB) measurements.
• To develop a statistical framework for analyzing complex genomic biomarker variability.
• To assess the impact of panel size, germline filtering, and biological/technical variance on psTMB accuracy.

Main Methods:

• Separated and modeled individual confounders of psTMB: panel size, germline mutation filtering, biological variance, and technical variance.
• Fitted published experimental psTMB data to developed error models to quantify each confounder's contribution.
• Summed individual variance contributions to determine total psTMB variance.

Main Results:

• Total error rates for a 1-1.5 Mbp panel were 57%, 42%, 34%, and 28% for psTMB of 5, 10, 20, and 40 muts/Mbp, respectively.
• Stochastic error from panel size was the largest contributor to psTMB variance, particularly for TMB ≤ 20 muts/Mbp.
• Other variability sources can be managed through stringent quality control, laboratory best practices, and optimized bioinformatics.

Conclusions:

• Developed a statistical framework to analyze complex genomic biomarker variability, applicable to psTMB.
• Highlights the significant impact of panel size on psTMB accuracy, emphasizing the need for careful consideration in biomarker development.
• The framework supports the clinical implementation of quantitative biomarkers like TMB.