Evaluation of Stimulated Raman Histology as a Novel Method for Pathologic Examination of Urothelial Carcinoma With Unprocessed, Fresh Bladder Cancer Tissue
Philipp Maisch, Michael Mühlberger, Annika Beck, Wilhelm Schreen, Miles P. Mannas, Cornelia Horsch, Christian BolenzIntroduction
Stimulated Raman histology (SRH) is a recent microscopic technique allowing real‐time, label free, high‐resolution microscopic images of unprocessed, unsectioned tissue. Tissue samples are imaged in the operating room using a mobile SRH microscope, visualizing histology within minutes.
Methods
Aim of this prospective first in human feasibility study was to evaluate the use of SRH to identify bladder cancer (BCa) and to test whether a histopathological grading between low- and high-grade tumors can be determined. Therefore, patients with suspicion of BCa, who underwent transurethral resection of bladder tumor were included. Suspicious papillary tissue was resected and subsequently scanned with an SRH microscope (NIO, Invenio Imaging Inc.). After a 10-sample training set, pathologists were tested on a 30-sample test set. Images which were initially classified wrong were re-evaluated within a consensus conference.
Results
The scanning time was less than 180 seconds for all tissues samples. Individual pathologists’ overall combined accuracy (CA) for interpretation of low and high grade SRH images was 76.5 % [95% confidence interval (CI): 65.8-85.3]. After the consensus conference pathologists’ overall CA for interpretation of low and high grade SRH images was 96.3 % [95% CI: 89.6-99.2], respectively. The sensitivity for the identification of high-grade tumors was between 66.0 % and 66.7 % before the consensus conference, increasing to 100 % after pathologic consensus. The specificity of the three pathologists was between 83.3 % and 100 %. After the consensus conference, the specificity was on average 91.7 %.
Conclusion
After further investigations, including the testing against benign tissue, SRH might be a reliable point of care-technique to produce high-resolution images of unprocessed bladder tumors, allowing histopathological identification of BCa. Low- and high-grade tumors can be accurately discriminated.