Morphological and immunophenotypical differences between chronic periodontitis and peri-implantitis - a cross-sectional study.

Abstract

Purpose: To examine differences in morphology and in immunophenotype subsets between chronic periodontitis (CP) and peri-implantitis (P-I) lesions and to test the diagnostic agreement (CP vs P-I) among three independent observers, based on histopathological features.

Materials and methods: This observational cross-sectional study included 15 gingival biopsies of CP lesions and 15 biopsies of P-I lesions for double-blinded examination by three independent pathologists. Inflammatory infiltrate intensity was assessed semiquantitatively on a 4-point scale, determining the percentage of lymphocytes, plasma cells, monocytes/macrophages, and granulocytes and the presence/absence of bacterial colonies. Immunohistochemical analyses were performed to verify the inflammatory infiltrate subset data (CD45, CD38, CD68 and myeloperoxidase [MPO]-positive), and number of vessels. Kappa statistic was used to evaluate the degree of diagnostic concordance among examiners.

Results: Inflammatory infiltrate was significantly more severe in P-I cases (P = 0.01), which showed a significantly higher percentage of plasma cells (P = 0.004) than in CP cases. Immunohistochemically, the percentage of leukocyte subsets was generally lower in CP (CD38: 32.05%; CD68: 6.45% and MPO: 8.62%) than in P-I (CD38: 61.13%; CD68: 9.09% and MPO: 7.47%) (CD38 P = 0.001, P = 0.955 and P = 0.463, for remaining subsets, respectively; Mann-Whitney U-test). The inter-observer diagnostic agreement was poor or slight (kappa = -0.18 to 0.13).

Conclusions: Despite the significantly more severe general inflammatory infiltrate and plasma cells in P-I cases, it proved difficult to detect reliable differential morphological features based on histopathological images of these CP and P-I soft-tissue samples, obtaining low inter-observer and intra-observer diagnostic agreement. Conflict of interest statement: This investigation was partially supported by Research Groups #CTS-138 and #CTS-583 (Junta de Andalucía, Spain). No conflict of interest.