ESMO Congress 2023 | OncologyPRO

Abstract 907P

Background

Adenoid cystic carcinoma (ACC) is characterized by multiple relapses and a poor prognosis. It is commonly considered a radioresistant tumor, indeed particle radiotherapy is especially indicated in ACC. Recently, ACC was stratified in two clusters with a different prognosis, ACC type I and II, according to distinct molecular features i.e. NOTCH activating mutation and TP63 upregulation, respectively. We aimed to disclose molecular profiles predictive of radiosensitivity or radioresistance features.

Methods

The study cohort comprised 54 patients with ACC of salivary gland origin, treated with surgery plus photon beam radiotherapy (R) or R alone. A survey in the literature yielded a list of 15 signatures including pan-cancer or different tumor-specific models (e.g. head and neck; breast cancer; glioblastoma; prostate cancer) with a total of 141 genes. The expression levels of these genes were retrieved and signatures scores were imputed following the methods described in the original papers. Log2-trasformed transcripts per million mapped reads values were retrieved from Ferrarotto et al. [PMC7854509]. Clinical and follow-up data were updated at 31st December 2022; overall survival (OS) as endpoint was estimated with the Kaplan-Meier method.

Results

Only the signatures developed on breast cancer gene-expression datasets among those investigated were able to stratify ACC patients; in particular, we focused on an immune-radiosensitivity signature [PMC8984882] able to reach HR=7.59, 95%CI 3.19-18.09, p=1e-07. The predicted radiosensitivity of this signature was significantly associated with ACC type II and lack of NOTCH mutations.

Conclusions

Molecular profile of ACC type II is significantly associated with an intrinsic radiosensitivity. The association of ACC radiosensitivity to breast cancer signatures warrants further research on the biology behind ACC and its molecular similarities.