Oral potentially malignant lesions: first-level micro-histological diagnosis from tissue fragments sampled in liquid-based diagnostic cytology

Background: Scalpel biopsy may under-diagnose oral dysplasia/carcinoma in potentially malignant lesions (PMLs) because samples represent only one or a few sites. It is possible that small tissue specimens obtained from over the whole area of PMLs, by scraping with a dermatological curette, could be treated histologically and used as ‘micro’ biopsies. This study values the accuracy of micro-biopsies in the detection of dysplasia/carcinoma in oral PMLs. Methods: A prospective study was carried out on 164 patients with PMLs, with both scalpel and micro-biopsies, for the presence of dysplasia/carcinoma. The most severe diagnosis (obtained by either method) was used as the reference standard. The presence/absence of the basement membrane zone (BMZ) in the micro-biopsy specimens correlated with the site, the clinical features of the PMLs and the operator. Results: Micro-biopsy gave six of 164 (3.66%) inadequate specimens. Of 158 of 164 adequate samples, dysplasia/carcinoma was diagnosed in 85 of 158 cases; micro-biopsy diagnosis was in agreement with scalpel biopsy in 144 of 158 (91.14%) cases and showed a better sensitivity than did scalpel biopsy (97.65% vs. 85.88%), corresponding to two of 158 false-negative cases by micro-biopsy vs. 12 of 158 by scalpel biopsy. The BMZ was observed in 110 of 158 (69.62%) of all micro-biopsies and had no relationship with the sampling site, the clinical features of the PMLs or the operator. Conclusions: The negative predictive value (97.33%) suggests that micro-biopsy may well be an effective first-level diagnostic procedure for PMLs (especially in follow-ups and multiple lesions); moreover, in carcinoma (17% of cases) a definitive diagnosis could be made without further investigation.