Oral biopsy in mucous membrane pemphigoid and pemphigus vulgaris with gingival expression: the optimal site. A systematic review and meta-analysis.

ABSTRACT

PURPOSE: In order to diagnose mucous membrane pemphigoid (MMP) and pemphigus vulgaris (PV) with gingival expression, clinical data must be compared with immunohistochemical data obtained using direct immunofluorescence (DIF). It is therefore essential to carry out a good quality mucosal biopsy for this vital additional test. To date, no study has been able to effectively guide clinicians in their choice of oral site for biopsy to guarantee the efficient contribution of DIF to diagnosis. We propose a systematic review of the literature and a meta-analysis to clarify this issue. MATERIALS AND METHODS: Electronic databases and bibliographies of articles were searched in April 2023. The primary outcome was the rate of DIF + contribution to diagnosis according to the location of the oral site biopsied. RESULTS: 16 studies were included. Gingival biopsies showed a rate of DIF + 100% [97%-100%] p = 0.998 I2 = 0.0% with no heterogeneity for PV, and 90.2% [66.5%-100%] p < 0.001 I2 = 89.6% with high heterogeneity for MMP. For the other oral sites, this rate was 95.7% [87.4%- 100%] p = 0.011 I2 = 73.0% with moderate heterogeneity for PV, and 87.4% [70.1%- 98.7%] p < 0.001 I2 = 92.6% with high heterogeneity for MMP. In addition, meta-regression confirmed the significant association between the appearance of the biopsied mucosa and the rate of DIF + in MMP (p < 0.001), with no influence on residual heterogeneity. CONCLUSION: The nature of the oral mucosa biopsied does not influence the rate of DIF + to diagnosis. The choice of biopsy site should only take into account the characteristics of the clinical picture and the benefit/risk balance of the surgical protocol. The sample must be taken in healthy aeras as close as possible of active lesions on the gingiva if the MMP and PV are strictly gingival, on the alveolar mucosa if the whole gingiva is altered and on any healthy mucosa if a large number of oral sites are affected. CLINICAL TRIALS CRD42023392345.