Background Rheumatoid arthritis (RA) is a chronic inflammatory disease characterized by bone, cartilage and muscle loss. While bone and cartilage damage have been extensively studied in the past, the effects of RA on volume and composition of hand muscles have not yet been studied. This situation is surprising, since visible hand muscle atrophy is a hallmark of RA and quantification of muscle composition is of growing interest.
Objectives Quantitative assessment of hand muscle volume and fat using MRI.
Methods A random forest based method was used to segment hand muscle in T1 weighted MR scans of 76 RA patients (37 males, 26–87 years, mean 61 years). The segmentation procedure is fully automated but allows for manual corrections of wrongly segmented areas. Multimodal registration of the muscle segmentation masks to MR Dixon Fat Fraction images was used for fat quantification.
Outcome parameters were absolute hand and muscle volume (VabsH resp. VabsM), relative muscle volume VrelM = VabsM/VabsH and absolute and relative fat content (VabsF resp. VrelF).
Student's t-tests were performed for gender discrimination. Linear regression was used to model age dependency. Further a multivariate regression was used to model dependence on predictors BMI, RA disease duration, DMARD treatment duration, HAQ, DAS28, RF, ESR and CRP after age adjustment. Dixon sequences were not available for all patients, therefore fat analysis could only be done in 17 females.
Results VabsM in males was significantly higher than in females (p<0.001, means: 63 cm3 vs. 29 cm3). Similar results were observed for VabsH (p<0.001, means: 243 cm3 vs. 163 cm3) and VrelM (p=0.02, means: 0.256 vs. 0.239).
The table shows the results of the linear regression analysis for significant predictors of males and females, respectively. The figure shows age dependence of VabsM and VabsH.
VabsF and VrelF showed no dependence on age or disease duration. VabsF but not VrelF was highly correlated with VabsM and VabsH (both: p<0.001, R2=0.75). VrelF was positively correlated with BMI (p=0.03, R2=0.32), a weaker but not significant effect was observed for VabsF.
Conclusions Results for muscle volume were highly sex specific. Males showed higher VabsM, VabsH and VrelM. Nevertheless VabsM and VabsH in males showed significant decrease with age while the opposite but weaker effects were observed in females. Relative fat volume depends on BMI but is independent of muscle volume.
Disclosure of Interest None declared