876P - Gene expression analysis in oral potentially malignant disorders (OPMD) with dysplasia identifies patients at high risk of malignant transformation

Background

Despite surgical resection, Oral Potentially Malignant Disorders show a rate of malignant transformation varying from 3% to 17.5%. Clinical and histological factors outside dysplasia and previous oral cavity carcinoma failed to identify OPMDs at higher risk of malignant transformation (MF). We previously identified six clusters of transcriptomic data able to stratify Head and Neck Squamous Cell Carcinoma (HNSCC) that also applied to OPMD. The primary objective of the current study was to validate the stratification by the transcriptional signature in an independent series of dysplastic OPMD.

Methods

We collected a retrospective series (3/1996 to 11/2019) of completely excised OPMDs with any grade of dysplasia at the University of Brescia. Clinical/histological data were retrieved. Library preparation for gene expression profiling was conducted QuantSeq 3'mrna-seq (Lexogen) and pooled libraries were sequenced by NextSeq500 (Illumina). Finally, previously identified molecular clusters based on main biological characteristics and de-regulated signaling pathways were associated with the risk of MF.

Results

We identified 106 consecutive patients with OPMDs. Of these, 66 had available material for gene expression analysis. A slight majority were male (53.4%), with a median age of 65 years (33-95); 17% had a previous diagnosis of HNSCC. With a median follow-up of 53 months (IQR 91), 23% experienced a MF with a median time to transformation of 30 months (3—195). After stratification based on our previously reported clusters, OPMDs were classified as mesenchymal, hypoxia, defense response, classical, and immunoreactive clusters constituted 8%, 15%, 27%, 9%, and 41% of our cohort, respectively. Hypoxia cluster showed a higher risk of malignant transformation (p<0.0001).

Conclusions

OPMDs gene expression clustering allows stratifying patients according to different risk of MF. Among them, hypoxia cluster had the highest probability of cancerization. A more in-depth analysis of this cluster is required, both to better characterize the mechanisms of MF and to develop customized prevention strategies.

Clinical trial identification

Editorial acknowledgement

Legal entity responsible for the study

Medical Oncology Unit, Department of Medical & Surgical Specialties, Radiological Sciences & Public Health, University of Brescia, ASST Spedali Civili, 25123 Brescia, Italy.

Funding

Funded by AIRC (IG 21740 to PB).

Disclosure

All authors have declared no conflicts of interest.