Actionable Genomic Alterations and Outcomes in Lung Squamous and Adenosquamous Cell Carcinoma

Background
Actionable genomic alterations (AGA) are uncommon in lung squamous/adenosquamous carcinoma (LUSC/LUASC) and may contribute to poorer survival than lung adenocarcinoma (LUAD). Though a subset of LUSC patients harbor AGA and may benefit from targeted therapy, outcomes are poorly characterized. We hypothesized that AGA- LUSC patients treated with first-line (1L) targeted therapy have superior survival to those receiving chemotherapy and that AGA-LUSC has a distinct immune tumor microenvironment (TME) compared to non-AGA-LUSC.

Methods
We identified and analyzed LUSC (6,970) and LUASC (227) patients from the Tempus Lens database with DNA (xT, 648-gene) and whole-transcriptome RNA (xR) NGS using the Lens Platform. AGA were defined as ALK, ROS1, RET, NTRK1/2/3 fusions, EGFR, KRAS, BRAF p.V600E, MET exon 14 skipping, or ERBB2 alterations. KRAS, BRAF p.V600E, and MET exon 14 skipping were classified as immune-associated AGA. RNA data were quantified as transcripts per million (TPM) and reported as log2(TPM+1). We assessed tumor mutational burden (TMB), TTF1 and TP63 gene expression, and TME (via quanTIseq). OS and PFS were measured from 1L treatment initiation to death, progression/death or last follow up, respectively.

Results
AGA incidence was 7.5% in LUSC and 45% in LUASC. Non-immune AGA had significantly lower median TMB than immune-AGA and non-AGA in both LUSC (5.3 vs 7.4 vs 7.9) and LUASC (3.7 vs 5.5 vs 8.4) (both p < 0.001) and had numerically lower CD8 T-cell infiltration in LUSC. TTF1 expression did not differ among AGA and non-AGA groups; but TP63 expression was significantly higher in LUSC than LUASC (8.70 vs 6.23, p < 0.001) and in non- AGA compared to AGA in both LUSC (p < 0.001) and LUASC (p = 0.027). In AGA-LUSC, 1L targeted therapy (n = 15) was associated with numerically longer PFS (27.7 mo vs 8.1 mo, p = 0.17) and OS (19.2 mo vs 15.7 mo, p = 0.78) compared with chemoimmunotherapy/chemotherapy. No difference in outcomes was observed between non- AGA and AGA LUSC patients treated with chemoimmunotherapy/chemotherapy.

Conclusions
Given the observed rates of AGA in both LUSC and LUASC and numerically improved outcomes in AGA-LUSC patients receiving 1L targeted therapy, routine NGS testing is warranted. TME differs between non-immune and immune-AGA, and TP63 expression is inversely associated with AGA in both tumor types.