Objectives: To evaluate the degree and causes of underdiagnosis of incisional biopsy in oral potentially malignant lesions (PML).
Design: Cross-sectional retrospective study.
Materials and methods: All pathology slides (incisional biopsy and resection specimen) from 46 patients with oral PML were reviewed and scored blind. The provisional diagnosis was considered in order to provide underdiagnosis or overdiagnosis according to whether this score was lower or higher than the final diagnosis.
Main outcome measures: Concordance between the scores for provisional and final diagnoses was compared for reviewed and unreviewed assessments.
Results: Upon review, underdiagnosis significantly decreased from 23.9% to 4.4% (P=0.016, chi-square with the Yates' correction). As the underdiagnosis ratio was not significantly influenced by intra- or inter-observer variability, it is likely to be related to pathologists' interpretation difficulties when examining incisional biopsy specimens.
Conclusions: The good quality of incisional biopsy samples was demonstrated, but clinicians should be conscious of the possible underdiagnosis by incisional biopsy when planning therapy in suspected early oral squamous cell carcinoma lesions.