CHARACTERIZING KI67 ANTIBODY STAINING FOR NON-HODGKIN'S LYMPHOMA USING VISIBLE AND NEAR-INFRARED HYPERSPECTRAL IMAGING TECHNIQUES

Objective To assess the model of quantitation of the Ki67 index by employing visible and infrared (Vis-NIR) hyperspectral imaging (HSI) in Non-Hodgkin Lymphomas (NHL) as time and cost-efficient tissue-sparing alternative to standard diagnostic procedures. Material and Methods Ki67 quantitation was done on 19 formalin-fixed paraffin embedded specimens (6 follicular lymphomas (FL); 12 diffuse large B-cell lymphomas (DLBCL) and one Burkitt-lymphoma (BL) by applying classical immunohistochemistry using an automated platform. Subsequent visual quantification of the Ki67 index was comparatively assessed visually and to HSI-based Ki67 index quantification. Results HSI was successfully implemented for analyzing histopathological slides using the IMEC SNAPSCAN camera. There was no significant difference between visual assessment (VA) and HSI (p = 0.257) in FL, while in DLBCL VA resulted in significantly higher Ki67 values than HSI (p = 0.006). Cohens Kappa showed a fair correlation for both FL and DLBCL (Kappa: 0.333 and 0.316, respectively). The case of BL added as a”super-positive” control was consistently measured by both techniques. Conclusion HSI can be used to quantify marker expression using the Ki67 index model. HSI offers additional information on the biochemical composition of tissues, while preserving limited samples. HSI is easy to install enabling optimization of workflows by restricting analysis times at low costs. Future application by automated imaging analyses systems may allow quantitative measurement of more significant areas of tissue samples, diminishing selection bias in commonl visual pathology and improving information on the tissue heterogeneity of a given marker. In summary, HSI is a cost-effective and quick alternative imaging platform when compared to other digital image analysis systems that are significantly more costly and time-consuming.