Quantitative Evaluation of Inflammatory Markers in Peri-Implantitis and Periodontitis Tissues: Digital vs. Manual Analysis—A Proof of Concept Study

Abstract

Background and Objectives: In dentistry, the assessment of the histomorphometric features of periodontal (PD) and peri-implant (PI) lesions is important to evaluate their underlying pathogenic mechanism. The present study aimed to compare manual and digital methods of analysis in the evaluation of the inflammatory biomarkers in PI and PD lesions. Materials and Methods: PD and PI inflamed soft tissues were excised and processed for histological and immunohistochemical analyses for CD3+, CD4+, CD8+, CD15+, CD20+, CD68+, and CD138+. The obtained slides were acquired using a digital scanner. For each marker, 4 pictures per sample were extracted and the area fraction of the stained tissue was computed both manually using a 594-point counting grid (MC) and digitally using a dedicated image analysis software (DC). To assess the concordance between MC and DC, two blinded observers analysed a total of 200 pictures either with good quality of staining or with non-specific background noise. The inter and intraobserver concordance was evaluated using the intraclass coefficient and the agreement between MC and DC was assessed using the Bland–Altman plot. The time spent analysing each picture using the two methodologies by both observers was recorded. Further, the amount of each marker was compared between PI and PD with both methodologies. Results: The inter-and intraobserver concordance was excellent, except for images with background noise analysed using DC. MC and DC showed a satisfying concordance. DC was performed in half the time compared to MC. The morphological analysis showed a larger inflammatory infiltrate in PI than PD lesions. The comparison between PI and PD showed differences for CD68+ and CD138+ expression. Conclusions: DC could be used as a reliable and time-saving procedure for the immunohistochemical analysis of PD and PI soft tissues. When non-specific background noise is present, the experience of the pathologist may be still required.