Tissue-based genomic instability markers for predicting malignant transformation in oral leukoplakia and proliferative verrucous leukoplakia: a systematic review.

OBJECTIVES

Although several biomarkers have been described for predicting malignant transformation in oral leukoplakias (OLs) and proliferative verrucous leukoplakias (PVLs), no systematic review has comprehensively evaluated tissue-based genomic instability markers. This review aimed to evaluate the evidence for these markers and their potential role in biomarker panel development.

METHODS

A systematic review across PubMed, Embase and Cochrane Library was performed to identify studies evaluating the differences in tissue-based genomic markers between OL and PVL patients with and without malignant transformation.

RESULTS

34 observational studies comprising 3,237 patients were included, and genomic aberrations were categorised into DNA-level, chromosomal, and gene-specific alterations. For studies on OLs, DNA-level and chromosomal markers for which individual studies reported associations with malignant transformation included aneuploidy, impaired DNA repair capacity, loss of heterozygosity, chromosomal instability, and copy number alterations. Multiple gene-specific alterations also showed associations (e.g., TP53, MKI67, FGFR1), but findings varied across studies. The genomic markers of PVLs differed substantially, with fewer consistent predictors found. No meta-analysis was performed as all included studies were observational.

CONCLUSIONS

Genomic instability across multiple levels contributes to malignant transformation, and represents a promising biological framework for predicting malignant transformation for OLs. While no single marker reliably demonstrates sufficient predictive performance, the integration of complementary genomic alterations with clinical and histopathological risk factors may provide a basis for the development of robust multi-marker panels. Future prospective studies using standardised detection methods and multivariable prediction models are required before clinical implementation.

SYSTEMATIC REVIEW REGISTRATION: identifier CRD42024585830.