Predicting Outcomes and Prognosis in Adult Myelopathies in Adult Non-traumatic Myelopathies
Maria Garcia-Dominguez1, Paula Barreras2, Carlos Pardo-Villamizar3
1UMass Memorial Medical Center, 2Johns Hopkins, 3Johns Hopkins U, Med Dept of Neurology
Objective:
 to develop a predictive score for long-term functional disability in myelopathypatients.
Background:
Inflammatory myelopathies (IM) and vascular myelopathies (VM) represent the two most common etiologies of non-traumatic myelopathies in adults. Tools predicting long term outcomes are lacking.
Design/Methods:
In a retrospective observational study that included 211 subjects with IM and VMfollowed at the Johns Hopkins Myelitis Myelopathy Center, between 2010 and 2020, clinical features, neuroimaging, and CSF profiles at the initial presentation were used to identify predictors of severe disability (defined as modified Rankin scale >3) using multivariable logistic regression models. Risk assessment for significant variables and predictive scores were calculated on long-term follow-up (32 months from onset).
Results:
 Factors associated with the worst functional prognosis (mRS>3) in non-traumatic myelopathies included rapid progression of initial symptoms, <48 hours to nadir (coefficient0.53, p=0.017), longitudinally extensive lesions, >3 spinal levels) on initial MRI (coefficient 0.6,p=0.005), and pleocytosis with WBC > 50 in CSF (coefficient 0.71, p=0.024). The presentation consistent with vascular myelopathy (coefficient, p value 0.008). Assigning one point to each of those factors, an outcome predictive score was developed (Total score of 4). Having a score of four was associated with a risk of 72% for mRS >3 at long-term follow-up. A score of one was associated with 27%, a score of 2 was associated with 43% risk, and a score of 3 was associated with 58 % of risk. Patients with a score of 0 have an 18% risk of mRs>3 at 32 months of follow-up.
Conclusions:

A score assessment that considers the clinical, neuroimaging, and CSF profiles at initial presentation in IM and VM may provide a useful tool to estimate the risk of adverse functional outcomes (mRs>3).

 

10.1212/WNL.0000000000205946