Are there diagnostic alternatives to histopathology in detecting oral cancer?

Abstract

Background Oral potentially malignant disorders (OPMDs) include erythroplakia, erosive lichen planus, non-homogenous leukoplakia and many others. These lesions may revert to healthy tissues, remain unchanged, or undergo malignant change to oral squamous cell carcinoma (OSCC). Early detection allows early management, which subsequently improves survival rates for OSCC. Current diagnostic practice involves a tissue biopsy with histology. This can be uncomfortable for patients and delays diagnosis. Other, less invasive diagnostic tests are available which can provide immediate results.

Objectives Primary objective: to evaluate the diagnostic accuracy of index tests in detecting OSCC and OPMDs. Secondary objective: to determine the relative accuracy of alternative diagnostic tests.

Search methods Four online databases including Medline (OVID), Embase (OVID), the US National Institute of Health Ongoing Trials Register (clinicaltrials.gov) and the World Health Organisation International Clinical Trials Registry Platform were searched for ongoing trials to 20 October 2020. Citation searches were conducted, with no language restrictions, and reference lists were reviewed for any additional references.

Selection criteria Studies reporting the diagnostic test accuracy of the following index tests were included if used in conjunction with oral examination in determining OPMDs or OSCC: oral spectroscopy, light-based detection, vital staining, saliva and/or blood analysis and oral cytology.

Data collection and analysis Titles and abstracts were screened by two designated review authors to determine relevance. A minimum of two authors, both independently and together, assessed the eligibility of the papers and undertook the quality assessment and data extraction. The methodological quality was determined using the Quality Assessment of Diagnostic Accuracy Studies 2 (QUADAS-2). Results of each index test in various studies were amalgamated using meta-analysis following the bivariate approach to determine sensitivity and specificity.

Results Seventy-nine datasets were included, spanning from 1980 to 2020. A total of 7,942 lesions were examined. These studies assessed diagnostic sensitivity and specificity of tissue biopsy and histology in comparison to oral cytology (24 datasets), oral spectroscopy/light-based detection (24 datasets) and vital staining (22 datasets). Nine datasets evaluated two combined index tests. No eligible diagnostic accuracy studies assessing salivary or blood sample analysis were identified. Two studies were deemed low risk of bias in all categories (patient selection, index tests, reference standards and flow and timing) with 33 studies being low concern for applicability across three sectors. Patient selection, demographics, the index test and reference standard used were appropriate to a secondary care setting. Meta-analysis of eligible index tests showed the following results. Cytological sensitivity was 0.90 (CI 0.82-0.94) and specificity 0.94 (CI 0.88-0.97) from 24 datasets. For oral spectroscopy, sensitivity was 0.87 (CI 0.78-0.93) and specificity 0.50 (CI 0.32-0.68) from 24 datasets. For vital staining, sensitivity was 0.86 (CI 0.79-0.90) and specificity 0.68 (CI 0.58-0.77) from 22 datasets. For combined tests, sensitivity was 0.78 (CI 0.45-0.94) and specificity 0.71 (CI 0.53-0.84) from nine datasets.

Conclusion None of the adjunctive tests can presently be recommended in lieu of tissue biopsy and histological assessment. Of the modalities, oral cytology showed the greatest potential based on the high summary estimate values for sensitivity and specificity. Further research is needed into the utility of these alternative diagnostic tests.