Abstract
Purpose :
Quantifying the number of IgG4 plasma cells and percentage of IgG4/IgG plasma cells is crucial for diagnosis of patients with IgG4-RD. Digital image analysis (DIA) is able to count IHC stained cells using Positive Cell Detection. According to the criteria, IgG4-related disease can be diagnosed based on the absolute number of IgG4 positive cells/high power field and the IgG4/IgG ratio. The goal is to evaluate the accuracy and efficacy of digital pathology tools to diagnose IgG4-RD on tissue images.
Methods :
We have 14 tissue biopsies from patients with inflammatory orbital or ophthalmic diseases (4 male, 10 female), including 8 from lacrimal glands, 5 orbital tissue, and 1 conjunctiva. The age range were 29-70 years old (mean= 53.07). We collected images of IgG4, IgG or CD138 for each biopsies making a total of 28 images. There were 6 positive and 8 negative cases. To identify the plasma cells, IgG stain was used to identify the plasma cells in 3 cases and CD138 was used in 11 cases. The parameters for the images are in png format, scan factor of 40x, and with a regional size of 5,000 x 5,000 pixels. Each image was analyzed by DIA and manual counting, which we believe is a gold standard, for IgG4, IgG or CD138, and IgG4/IgG or IgG4/CD138 ratio. The statistical analysis used was Intraclass Correlation Coefficient (ICC) to describe the correlation between the measurements from DIA and manual counting groups.
Results :
The quantification of absolute IgG4 positive cells with DIA had excellent correlation with manual counting (ICC= 0.981(p<0.001). The quantification of IgG positive cells had poor correlation (ICC= 0.373(p=0.129). The quantification of CD138 positive cells had moderate correlation (ICC=0.750 (p=0.001). IgG4/IgG ratio showed poor correlation. (ICC=0.390(p=0.294). IgG4/CD138 showed good correlation (ICC=0.759(p=0.003).
Conclusions :
Using DIA for quantification of IgG4 positive cells appears to be the most reliable method. There was much variability in quantifying in IgG stains, CD138 stains, IgG4/IgG ratios with DIA. Additional studies with larger sample size are necessary to further evaluate the efficacy of DIA in diagnosing IgG4-related disease.
This abstract was presented at the 2023 ARVO Annual Meeting, held in New Orleans, LA, April 23-27, 2023.