Together, our data indicate that Lupeol may orchestrate a bifurcated regulation of neoplastic growth and apoptosis in head and neck cancers and may serve as a promising agent for the management of tumors that have progressed on a platinum-based treatment regimen.

Epidermal growth factor receptor (EGFR) pathway is overexpressed in head and neck cancer (HNC). Lupeol, a natural triterpene (phytosterol found in fruits, vegetables, etc.), has been reported to be effective against multiple cancer indications. Here we investigate the antitumor effects of Lupeol and underlying mechanism in oral cancer. Lupeol-induced antitumor response was evaluated in two oral squamous cell carcinoma (OSCC) cell lines (UPCI:SCC131 and UPCI:SCC084) by viability (MTT), proliferation, and colony formation assays. Lupeol-mediated induction of apoptosis was examined by caspase 3/7 assay and flow cytometry. Effect of Lupeol on EGFR in the presence or absence of EGF was delineated by Western blot. The mRNA stability assay was performed to check the role of Lupeol on COX-2 mRNA regulation. Lupeol inhibited proliferation of OSCC cells in vitro by inducing apoptosis 48 h post treatment. Ligand-induced phosphorylation of EGFR and subsequent activation of its downstream molecules such as protein kinase B (PKB or AKT), I kappa B (IκB), and nuclear factor kappa B (NF-κB) was also found to be, in part, suppressed. Interestingly, Lupeol suppressed expression of COX-2 at mRNA and protein level in a time-dependent manner. Primary explants from oral squamous cell carcinoma tissues further confirmed significant inhibition of proliferation (Ki67) in Lupeol-treated explants as compared to untreated control at 48 h. Together these data suggest that Lupeol may act as a potent inhibitor of the EGFR signaling in OSCC and therefore imply its role in triggering antitumor efficacy.

We assessed the pan-cancer predictability of multi-omic biomarkers from haematoxylin and eosin (H&E)-stained whole slide image (WSI) using deep learning and standard evaluation measures throughout a systematic study. A total of 13,443 deep learning (DL) models predicting 4,481 multi-omic biomarkers across 32 cancer types were trained and validated. The investigated biomarkers included genetic mutations, transcriptomic (mRNA) and proteomic under- and over-expression status, metabolomic pathways, established markers relevant for prognosis, including gene expression signatures, molecular subtypes, clinical outcomes and response to treatment. Overall, we established the general feasibility of predicting multi-omic markers across solid cancer types, where 50% of the models could predict biomarkers with the area under the curve (AUC) of more than 0.633 (with 25% of the models having AUC larger than 0.711). Aggregating across the omic types, our deep learning models achieved the following performance: mean AUC of 0.634 ±0.117 in predicting driver SNV mutations; 0.637 ±0.108 for over-/under-expression of transcriptomic genes; 0.666 ±0.108 for over-/under-expression of proteomes; 0.564 ±0.081 for metabolomic pathways; 0.653 ±0.097 for gene signatures and molecular subtypes; 0.742 ±0.120 for standard of care biomarkers; and 0.671 ±0.120 for clinical outcomes and treatment responses. The biomarkers were shown to be detectable from routine histology images across all investigated cancer types, with aggregate mean AUC exceeding 0.62 in almost all cancers. In addition, we observed that predictability is reproducible within-marker and less dependent on sample size and positivity ratio, indicating a degree of true predictability inherent to the biomarker itself.