(102) Comparison of Image Analysis Techniques of Immunohistochemical Stains From Patients With Neuroproliferative Vestibulodynia

Abstract Introduction Neuroproliferative vestibulodynia (NPV) is a provoked genital pain disorder that is difficult to diagnose because the underlying pathology is not visible during physical examination and resides underneath the epithelium. Diagnosis is confirmed in vestibular tissue by excess mast cells and nerves/nociceptors identified by immunohistochemical (IHC) staining for CD117, consistent with mast cells, and PGP9.5, consistent with nerves. Objective Compare two methods of assessing immunostaining density for mast cells and nerves in vestibular tissue from patients with suspected NPV. Methods Slides from pathology specimens from the 1:00-11:00 vestibule and 12:00 vestibule were immunostained for CD117 and PGP9.5 to confirm NPV diagnosis. Slides were digitally photographed with a LCD35 CMOS color camera attached to an Olympus CX43 microscope at 100x magnification. For each stained slide, at least 2 photomicrographs were taken of representative regions of the epithelial basement membrane and adjacent subepithelium. Density of mast cells and nerves was automatically determined using Fiji version 1.53 after separating diaminobenzidine immunostain from hematoxylin counterstain with the color deconvolution application. Multiple measurements of percent cross-sectional area were performed for each tissue section. A comparison was made between this assessment technique and a previous technique that included a manual measurement of immunostaining density without color deconvolution. All manual measurements were performed by the same reader. Percent cross-sectional area of immunostaining was averaged for each patient, and the mean density of mast cells and nerves was calculated for the study cohort. Results All slides examined exceeded 8 mast cells per high-powered field, confirming the NPV diagnosis in this cohort. Using the automated method after color deconvolution, median cross-sectional area for CD117 immunostaining was 0.57% and 0.63% for the 1:00-11:00 and 12:00 regions, respectively, while the manual method without color deconvolution resulted in median CD117 immunostaining of 0.69% and 0.73% for the 1:00-11:00 and 12:00 regions, respectively. Using the automated method, median cross-sectional area for PGP9.5 immunostaining was 0.33% and 0.41% for the 1:00-11:00 and 12:00 regions, respectively, while the manual method without color deconvolution resulted in median PGP9.5 immunostaining of 0.47% and 0.31% for the 1:00-11:00 and 12:00 regions, respectively. There was no statistically significant difference between the two techniques for both CD117 and PGP9.5 immunostaining density (Fig. 1). Conclusions Both techniques, manual and automatic, provide data consistent with NPV diagnosis. The manual technique involves more time and is subjective, potentially increasing variability between measurements and among different readers. The automated technique enables a faster workflow and is less dependent on the experience of the reader, and is therefore preferable. Disclosure No.