SMARCA4 Deficiency and Mutations Are Frequent in Large Cell Lung Carcinoma and Are Prognostically Significant.

Alvin Ho-kwan Cheung,Kit-Yee Wong,Shuk-Ling Chau,Fuda Xie,Zeta Mui,Gordon Yuan-Ho Li,Molly Siu Ching Li,Joanna Tong,Calvin Sze-Hang Ng,Tony S. Mok,Wei Kang,Ka-Fai To
DOI: https://doi.org/10.1016/j.pathol.2023.12.414
IF: 5.335
2024-01-01
Pathology
Abstract:SMARCA4 mutation has emerged as a marker of poor prognosis in lung cancer and has potential predictive value in cancer treatment, but recommendations for which patients require its investigation are lacking. We comprehensively studied SMARCA4 alterations and the clinicopathological significance in a large cohort of immunohistochemically-subtyped non-small cell lung cancer (NSCLC). A total of 1416 patients was studied for the presence of SMARCA4 deficiency by immunohistochemistry (IHC). Thereafter, comprehensive sequencing of tumours was performed for 397 of these patients to study the mutational spectrum of SWI/SNF and SMARCA4 aberrations. IHC evidence of SMARCA4 deficiency was found in 2.9% of NSCLC. Of the sequenced tumours, 38.3% showed aberration in SWI/SNF complex, and 9.3% had SMARCA4 mutations. Strikingly, SMARCA4 aberrations were much more prevalent in large cell carcinoma (LCC) than other histological tumour subtypes. SMARCA4-deficient and SMARCA4-mutated tumours accounted for 40.5% and 51.4% of all LCC, respectively. Multivariable analyses confirmed SMARCA4 mutation was an independent prognostic factor in lung cancer. The immunophenotype of a subset of these tumours frequently showed TTF1 negativity and HepPAR1 positivity. SMARCA4 mutation or its deficiency was associated with positive smoking history and poor prognosis. It also demonstrated mutual exclusion with EGFR mutation. Taken together, the high incidence of SMARCA4 aberrations in LCC may indicate its diagnostic and prognostic value. Our study established the necessity of SMARCA4 IHC in the identification of SMARCA4-aberrant tumours, and this may be of particular importance in LCC and tumours without known driver events.
What problem does this paper attempt to address?