Multimodal Regression Based Assessment of Longitudinal Evolution Trends in Inflammatory Myositis (IIM)
Prashanth Poulose1, Nalini Atchayaram1, Seena Vengalil5, Deepak Menon6, Saraswati Nashi7, Ravi Girikematha Shankar2, Karthik Kulanthaivelu3, Anita Mahadevan4
1Neurology, 2Department of Biostatistics, 3Neuroradiology, 4Department of Neuropathology, NIMHANS Bangalore, 5Neurology, Department of Neurology, NIMHANS, 6Neurology, National Institute of Mental Health and Neurosciences, 7Neurology, NIMHANS
Objective:
Study longitudinal evolution trends and see regression based treatment response prediction in a large Indian cohort
Background:
Literature gap on longitudinal evolution trends of IIM from subcontinent
Design/Methods:
92 M:F-38:54 with clinical & immunological diagnosis of IIM according to ACR/EULAR  criterion 2017, evaluated between 2014 -2020 were included in this retrospective analytic study 
Results:

Mean age of presentation and follow up are 38 yrs and 15 months respectively. MMT-8 at baseline(54) ,MMT-8 6m (69) ,final follow up (74).Dermatomyositis(DM) and overlap myositis (OM) were 37% each , necrotising myositis (NAM)19% and anti-synthetase syndrome (ASS)7% . Positive ANA (39%), myositis specific antibody 66% and myositis associated antibody 58% ,paraneoplastic antibody 6.5%.Pelvis and adductors muscle MRI showed fatty replacement in 83.6% (41/49) , edema 85%(42/49) .Anterior thigh FR-81.6%  ,edema 85.7% ,Posterior thigh FR-89.7% and edema 81.6% .Ant distal lower limb FR-63.2%  and edema -69.3% .In PDLL 67.3% FR and edema 63.2% . ILD was observed in 13%.FDG uptake also showed good correlation with histopathological features & muscle oedema in MRI.SUV mean ,SUV max ratio were higher among non responders. Muscle biopsy 36% cases showed muscle necrosis(28%)  perimysial infiltrates(20%) ,endomysial infiltrates(22%) ,perifasicular atrophy(11%) &fibrosis(23%).Outcomes showed best treatment responders(36%) , satisfactory responders(39%) ,unsatisfactory responders(13%) and non responders (11%).Non responders had higher ICU admissions=30.4%(7)(p<0.01),endotracheal intubations=13%(p<0.01), infections=39.1%(18), multiple immunomodulatory agents, severe weakness and rapidly progressive weakness=87%(n=20, p<0.01), chronic continuous course 82.6%(n=19, p=0.01) and higher number of relapses(p< 0.05).Multivariate logistic regression showed that lower score of  MMT-8 at baseline (p<0.01) and higher number of relapses (p<0.05) were independent predictors of non-response. These parameters could predict non-response with a sensitivity of 30.4% and specificity of 91.2% with area under the curve (ROC-0.775)

Conclusions:
Identifying distinctive trajectories paves path for creating predictive models and future precision based treatments. Highly specific model was another major highlight which aligns with precision diagnostic frameworks in tertiary care settings
10.1212/WNL.0000000000211694
Disclaimer: Abstracts were not reviewed by Neurology® and do not reflect the views of Neurology® editors or staff.