Characterization of the genomic and transcriptomic landscape of invasive non-mucinous lung adenocarcinoma based on IASLC grading.

BACKGROUND

The IASLC grading system has prognostic utility and potential therapeutic implications in invasive non-mucinous lung adenocarcinoma (LUAD), but the molecular basis underlying the grading spectrum remains unclear.

METHODS

We performed whole-genome sequencing in 138 Chinese patients with invasive non-mucinous LUAD and RNA sequencing of 96 matched tumor-normal tissue pairs to systematically characterize the molecular features across grades, including coding driver events, mutational signatures, non-coding regulatory disruptions, and transcriptional programs.

RESULTS

Compared with Grade 1-2 tumors, Grade 3 LUADs exhibited heightened invasive potential, manifested by more advanced stage, more frequent spread through air spaces, and independently worse survival. Grade 3 tumors had elevated tumor mutational burden and were enriched for alterations in genome maintenance and cell-cycle genes, including TP53, as well as genes implicated in DNA damage response, including ZFHX4. APOBEC-associated mutagenesis was selectively enriched in Grade 3 tumors independent of smoking status, consistent with an instability-associated phenotype. Recurrent non-coding regulatory disruptions affected lung lineage-defining genes, particularly surfactant-associated genes, and were correlated with reduced expression. Transcriptomic profiling revealed epithelial dedifferentiation, loss of pulmonary homeostatic programs, and activation of proliferative and stress-related pathways. Notably, MUC16 emerged as a convergent event linking genomic and transcriptional dysregulation, with coding mutations associated with higher expression and increased expression in Grade 3 tumors correlating with the proportion of high-grade histologic patterns.

CONCLUSIONS

These findings provide a molecular framework for the IASLC grading spectrum and identify Grade 3 LUAD as a distinct instability-associated and dedifferentiated biological state.