Incorporating the epidermal growth factor receptor expression and histopathological parameters in a mathematical model to predict the prognosis of oral squamous cell carcinoma: a prospective observational cohort study.

Background: Oral squamous cell carcinomas (OSCC) are the most common type of variant causing oral cancers, contributing for around 90%. The overall survival of these patients is below 50%. The postoperative overall survival (OS) has not improved over years much despite of advanced surgical techniques and invention of various anticancer drugs. There was always a requirement for a non-invasive molecular marker to predict the prognosis of these patients. The epidermal growth factor and their receptors are not only thought to play a critical but also an influential role in growth of the cell and differentiation in normal/ healthy tissues. They also play an important role in malignant progression of disease and tumorigenesis. A better and sound understanding of mechanisms at molecular level and identification of potential oncogenes in OSCC may provide innovative therapeutic decisions such as targeted therapy in management of these cancer patients.

Aim & Objectives: The aim of this study is to check whether epidermal growth factor expression is a prognosticator in oral squamous cell carcinoma and also to propose a mathematical model to find the prognosis of the patients which have not been done so far in the literature.

Materials & Methods: The study was a prospective cohort study with 25 patients who had biopsy proven OSCC who reported to our hospital from July 2017 to June 2019. The data collected from their histopathological report for this prospective study and model were: surgical margins (superior, inferior, anterior and posterior), depth of tumor, lymph nodal metastasis, lymphovascular invasion and scoring of epidermal growth factor receptor (EGFR) expression done by immunohistochemistry (IHC) on wax blocks.

Results: EGFR expression on surgical margins was found to have a -value of 0.023 which was significant statistically. EGFR expression showed a statistically significant value 0.002 as independent marker in prognosis with sensitivity of 97.7% and specificity of 61.2%. The tumor depth of infiltration showed an insignificant correlation with pathological Tumor, Node, Metastasis (TNM) staging with a p value of 0.860. A mathematical model linear regression equation was proposed which predicted a cutoff value above 16, the prognosis of the patient being bad (Stages III and IV), and below 16, the prognosis of the patient being good (Stages I and II).

Conclusion: This study put forward a proposed mathematical model by incorporating all important parameters to predict the prognosis of the patients. EGFR expression is one such important parameter to be considered to develop anti-EGFR agents to improve the OS of the patients.

Supplementary Information: The online version contains supplementary material available at 10.1007/s12663-022-01797-0.