Back

Transcriptional phenocopies of deleterious KEAP1 mutations dictate survival outcomes in lung cancer treated with immunotherapy.

Scalera, S.; Ricciuti, B.; Marinelli, D.; Mazzotta, M.; Cipriani, L.; Bon, G.; Schiavoni, G.; Terrenato, I.; Di Federico, A.; Alessi, J.; Fanciulli, M.; Ciuffreda, L.; De Nicola, F.; Goeman, F.; Caravagna, G.; Santini, D.; De Maria, R.; Cappuzzo, F.; Ciliberto, G.; Jamal-Hanjani, M.; Awad, M.; McGranahan, N.; Maugeri-Sacca, M.

2023-12-14 oncology
10.1101/2023.10.30.23297743 medRxiv
Show abstract

Mutational models denoting KEAP1-NRF2 pathway activation have emerged as determinants of survival outcomes in non-small cell lung cancer (NSCLC). Hypothesizing that genetically distinct tumors recapitulate the transcriptional footprint of KEAP1 mutations (KEAPness), we identified a KEAP1-NRF2-related gene set shared by tumors with and without pathway mutations. KEAPness-dominant tumors were associated with poor survival outcomes and immune exclusion in two independent cohorts of immunotherapy-treated NSCLC (SU2C and OAK/POPLAR). Moreover, patients with KEAPness tumors had survival outcomes comparable to their KEAP1-mutant counterparts. In the TRACERx421, KEAPness exhibited limited transcriptional intratumoral heterogeneity and an immune-excluded microenvironment, as highlighted by orthogonal methods for T cell estimation. This phenotypic state widely occurred across genetically divergent tumors, exhibiting shared and private cancer genes under positive selection when compared to KEAP1-mutant tumors. Collectively, we discovered the pervasive nature of the KEAPness phenotypic driver across evolutionary divergent tumors. This model outperforms mutation-based classifiers in predicting survival outcomes.

Matching journals

The top 3 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.