Non-invasive Methods for Oral Cancer Screening

Oral cancer is a common malignancy worldwide, with approximately 250,000 new cases diagnosed annually and a high mortality rate. Its primary risk factors are smoking, alcohol consumption, and betel nut chewing. Current screening methods heavily rely on visual inspection by specialist physicians and invasive biopsy of lesions, but they are limited by insufficient sensitivity, biopsy sampling errors, and low patient compliance.

To address these issues, this study proposes an innovative cross-sectional randomized trial to evaluate two non-invasive techniques-high-resolution microendoscopy (HRME) and liquid-based cytology (LBC) of the oral mucosa-with the aim of optimizing screening for oral precancerous lesions.

The study will recruit 400 participants with visually identified oral potentially malignant disorders (e.g., leukoplakia or erythroplakia) and randomly assign them to two groups:

a clinical biopsy group (biopsy site selected based on conventional visual inspection), and an HRME-guided biopsy group (biopsy site directed by high-resolution microendoscopy).

Both groups will also undergo liquid-based cytology (LBC) through collection and analysis of oral mucosal cells.

The study objectives include:

Comparing the detection rates of precancerous lesions (mild, moderate, severe dysplasia, or carcinoma in situ, or invasive cancer) between the two groups; Analyzing the intra-lesional consistency of biopsy sites in both groups; Validating the diagnostic performance of LBC alone or in combination with HRME.

Inclusion criteria

We proposed to include individuals aged 30-80 years with prevalent oral precancer with or without a prior history of head and neck cancer (oral cavity, oropharynx, hypopharynx, larynx). The rationale for including those with history of cancer is that they are still at risk of developing oral cavity cancer and therefore will benefit from a better screening algorithm. Also, in our experience lesions in cancer survivors are phenotypically similar to those without a prior history of head and neck cancer. Only lesions with at least 1 cm in greatest diameter will be considered eligible. Additionally, because there could be issues in HRME imaging due to keratinization at some anatomic sites, we will divide lesions into approximate quadrants and exclude any lesions that do not have good nuclear resolution on at least one representative image per quadrant.

-

Exclusion criteria

We exclude those who: 1) are undergoing current cancer treatment or had a cancer within the last two years to avoid any effects of the cancer treatments; 2) are unwilling or unable to provide informed consent to participate; and/or 3) based on clinical judgement, should not be included in the study e.g., has other medical conditions that would preclude them from participating.

-