Computer Vision Analysis of Rheumatoid Arthritis Synovium Reveals Lymphocytic Inflammation Is Associated With Immunoglobulin Skewing in Blood.

OBJECTIVE: We sought to develop computer vision methods to quantify aggregates of cells in synovial tissue and compare these with clinical and gene expression parameters. METHODS: We assembled a computer vision pipeline to quantify five features encompassing synovial cell density and aggregates and compared these with pathologist scores, disease classification, autoantibody status, and RNA expression in a cohort of 156 patients with rheumatoid arthritis (RA) and 149 patients with osteoarthritis (OA). RESULTS: All five features were associated with pathologist scores of synovial lymphocytic inflammation (P < 0.0001). Three features that related to the cells per unit of tissue were significantly increased in patients with both seronegative and seropositive RA compared with those with OA; on the other hand, aggregate features (number and diameter) were significantly increased in seropositive, but not seronegative, RA compared with OA. Aggregate diameter was associated with the gene expression of immunoglobulin heavy-chain genes in the synovial tissue. Compared with blood, synovial immunoglobulin isotypes were skewed from IGHM and IGHD to IGHG3 and IGHG1. Further, patients with RA with high levels of lymphocytic infiltrates in the synovium demonstrated parallel skewing in their blood with a relative decrease in IGHGM (P < 0.002) and IGHD (P < 0.03) and an increase in class-switched immunoglobulin genes IGHG3 (P < 0.03) and IGHG1 (P < 0.002). CONCLUSION: High-resolution automated identification and quantification of synovial immune cell aggregates uncovered skewing in the synovium from naïve IGHD and IGHM to memory IGHG3 and IGHG1 and revealed that this process is reflected in the blood of patients with high inflammatory synovium.

Machine learning identification of thresholds to discriminate osteoarthritis and rheumatoid arthritis synovial inflammation.

BACKGROUND: We sought to identify features that distinguish osteoarthritis (OA) and rheumatoid arthritis (RA) hematoxylin and eosin (H&E)-stained synovial tissue samples. METHODS: We compared fourteen pathologist-scored histology features and computer vision-quantified cell density (147 OA and 60 RA patients) in H&E-stained synovial tissue samples from total knee replacement (TKR) explants. A random forest model was trained using disease state (OA vs RA) as a classifier and histology features and/or computer vision-quantified cell density as inputs. RESULTS: Synovium from OA patients had increased mast cells and fibrosis (p < 0.001), while synovium from RA patients exhibited increased lymphocytic inflammation, lining hyperplasia, neutrophils, detritus, plasma cells, binucleate plasma cells, sub-lining giant cells, fibrin (all p < 0.001), Russell bodies (p = 0.019), and synovial lining giant cells (p = 0.003). Fourteen pathologist-scored features allowed for discrimination between OA and RA, producing a micro-averaged area under the receiver operating curve (micro-AUC) of 0.85±0.06. This discriminatory ability was comparable to that of computer vision cell density alone (micro-AUC = 0.87±0.04). Combining the pathologist scores with the cell density metric improved the discriminatory power of the model (micro-AUC = 0.92±0.06). The optimal cell density threshold to distinguish OA from RA synovium was 3400 cells/mm2, which yielded a sensitivity of 0.82 and specificity of 0.82. CONCLUSIONS: H&E-stained images of TKR explant synovium can be correctly classified as OA or RA in 82% of samples. Cell density greater than 3400 cells/mm2 and the presence of mast cells and fibrosis are the most important features for making this distinction.

Alanine Aminotransferase as a Monitoring Biomarker in Children with Nonalcoholic Fatty Liver Disease: A Secondary Analysis Using TONIC Trial Data.

BACKGROUND: Validated noninvasive biomarkers to assess treatment response in pediatric nonalcoholic fatty liver disease (NAFLD) are lacking. We aimed to validate alanine aminotransferase (ALT), a monitoring biomarker for change in liver histology. METHODS: A retrospective analysis using data from the TONIC trial. NAFLD histologic assessments were defined by: Fibrosis score, NAFLD activity score (NAS), nonalcoholic steatohepatitis (NASH), and a combination of NASH resolution and fibrosis (NASH + fibrosis). Analysis was performed using classification and regression trees (CART) as well as logistic regression. RESULTS: Mean ALT for the child over 96 weeks and percent change of ALT from baseline to 96 weeks were significant predictors of progression of NAFLD for each histologic assessment (p < 0.001 for fibrosis score, NASH, and NASH + fibrosis and p < 0.05 for NAS). Mean ALT adjusted for age, sex and ethnicity was a better predictor for change in NASH (81.8 (11.0) ROC (receiver operating characteristic curve) mean (SD (Standard derivation))) and NASH + fibrosis (77.8 (11.2)), compared to change in NAS (63 (17.7)) and fibrosis (58.6 (11.1)). CONCLUSION: Mean ALT over 96 weeks is a reasonable proxy of histologic improvement of NASH and NASH + fibrosis. These findings support ALT as a valid monitoring biomarker of histologic change over time in children with NASH and fibrosis.