Back

Evaluation of Prognostic Utility of Tumor Mutation Burden for Non Small Cell Lung Cancer (NSCLC) response to Immune Checkpoint Blockade Therapy: A Single Institute Study

Clemenceau, J. R.; Lee, S. H.; Peter, B.; Milinovich, A.; Jin, J.; Pennell, N.; Sohal, D.; Hwang, T. H.

2019-12-04 oncology
10.1101/19013433 medRxiv
Show abstract

IMPORTANCEThe role of Tumor Mutation Burden (TMB) as a prognostic and/or predictive biomarker for Immune Checkpoint Blockade (ICB) therapy in a real-world clinical setting is still unclear. OBJECTIVETo assess whether TMB status provided by a clinically and commercially available tumor genomic profiling (TGP) assay is associated with overall survival of Non-Small Cell Lung Cancer (NSCLC) patients treated with ICB from a single institute. DESIGN, SETTING, AND PARTICIPANTSOutcomes and genetic testing data were collected for 188 NSCLC patients treated within the Cleveland Clinic system between August 2012 and July 2017. MAIN OUTCOMES AND MEASURESOverall survival (OS) from time receiving ICB therapy. RESULTSAmong 188 patients with NSCLC (median age, 62 years; 49.5% female), 86 (45.7%) received ICB therapy. Patients were grouped into three categories based on the status of TMB (in mutations/Mb): high (>= 20/Mb), intermediate (>=5 to <= 20/Mb), and low (<5/Mb). In patients treated with ICB, TMB high status was not significantly associated with improved OS from therapy initiation (HR: 0.90 [95% CI, 0.52-2.49, P>0.8], median OS difference: 9.7 months). CONCLUSIONS AND RELEVANCEAmong patients with NSCLC from a single institute in a longitudinal database of clinical data including TGP results, exploratory analyses do not show statistical significance for the prognostic utility of TMB. These findings indicate that there is a need for more prospective data on the use of TMB status as a guide for ICB therapy in a routine care setting.

Matching journals

The top 12 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.