G Bathla1, C W Freeman2, T Moritani3, J W Song2, S Srivastava4, N Soni5, C Derdeyn5, S Mohan2. 1. Department of Radiology, University of Iowa Hospitals and Clinics, Iowa City, IA, USA. Electronic address: girish-bathla@uiowa.edu. 2. Department of Radiology, University of Pennsylvania Health System, Philadelphia, PA, USA. 3. Department of Radiology, University of Michigan, Ann Arbor, MI, USA. 4. Department of Statistics and Actuarial Science, University of Iowa, Iowa, USA. 5. Department of Radiology, University of Iowa Hospitals and Clinics, Iowa City, IA, USA.
Abstract
AIM: To assess the prevalence of various imaging manifestations in neurosarcoidosis (NS) patients at presentation and to explore if specific imaging findings may cluster in different sub-groups. MATERIALS AND METHODS: A retrospective, dual-institution, systematic imaging review was undertaken of the magnetic resonance imaging (MRI) findings in 100 consecutive NS patients who presented over a 15-year period. Clustering analysis (k-mode) was performed to evaluate co-occurrence of imaging findings. RESULTS: Non-enhancing white matter (NEWM) lesions were the most common imaging abnormality (56%), followed by leptomeningeal (47%) and pachymeningeal (32%) involvement. Other common manifestations included cranial nerve involvement (30%), parenchymal granulomas (27%), hypothalamic-pituitary-adrenal axis involvement (26%), and hydrocephalus (14%). Additionally, a higher prevalence of perivascular enhancement (23%), cerebrovascular events (including ischaemic and haemorrhagic events; 17%), and ependymal involvement (20%) were noted than recognised previously. Additional k-mode analysis was performed to explore underlying disease sub-clusters. This was evaluated for clusters varying between two though five (k=2-5). For k=4, the analysis revealed that the imaging findings may possibly be divided into disease sub-sets of four groups, each with varying distribution of imaging manifestations and clinical manifestations. CONCLUSION: Overall, NEWM lesions and meningeal involvement are the most common imaging manifestations of NS. The prevalence of perivascular enhancement, cerebrovascular events, and ependymal involvement is likely higher than reported previously. Additionally, different imaging findings in NS may cluster together and imaging subtypes in NS possibly exist.
AIM: To assess the prevalence of various imaging manifestations in neurosarcoidosis (NS) patients at presentation and to explore if specific imaging findings may cluster in different sub-groups. MATERIALS AND METHODS: A retrospective, dual-institution, systematic imaging review was undertaken of the magnetic resonance imaging (MRI) findings in 100 consecutive NSpatients who presented over a 15-year period. Clustering analysis (k-mode) was performed to evaluate co-occurrence of imaging findings. RESULTS: Non-enhancing white matter (NEWM) lesions were the most common imaging abnormality (56%), followed by leptomeningeal (47%) and pachymeningeal (32%) involvement. Other common manifestations included cranial nerve involvement (30%), parenchymal granulomas (27%), hypothalamic-pituitary-adrenal axis involvement (26%), and hydrocephalus (14%). Additionally, a higher prevalence of perivascular enhancement (23%), cerebrovascular events (including ischaemic and haemorrhagic events; 17%), and ependymal involvement (20%) were noted than recognised previously. Additional k-mode analysis was performed to explore underlying disease sub-clusters. This was evaluated for clusters varying between two though five (k=2-5). For k=4, the analysis revealed that the imaging findings may possibly be divided into disease sub-sets of four groups, each with varying distribution of imaging manifestations and clinical manifestations. CONCLUSION: Overall, NEWM lesions and meningeal involvement are the most common imaging manifestations of NS. The prevalence of perivascular enhancement, cerebrovascular events, and ependymal involvement is likely higher than reported previously. Additionally, different imaging findings in NS may cluster together and imaging subtypes in NS possibly exist.
Authors: Michael A Paley; Brandi J Baker; S Richard Dunham; Nicole Linskey; Claudia Cantoni; Kenneth Lee; Lynn M Hassman; Jennifer Laurent; Elisha D O Roberson; David B Clifford; Wayne M Yokoyama Journal: J Neuroimmunol Date: 2022-04-01 Impact factor: 3.221
Authors: Girish Bathla; Lama Abdel-Wahed; Amit Agarwal; Tracey A Cho; Sarika Gupta; Karra A Jones; Sarv Priya; Neetu Soni; Bruce A Wasserman Journal: Neurol Neuroimmunol Neuroinflamm Date: 2021-08-04