P123 MRI based criteria to differentiate dysferlinopathies from other genetic muscle diseases

• Published in: Neuromuscular disorders : NMD Vol. 33; p. S78
• Main Authors: Diaz, C Bolaño, Verdu-Diaz, J., Gonzalez-Chamorro, A., Straub, V., Manera, J Diaz
• Format: Journal Article
• Language: English
• Published: Elsevier B.V 01-10-2023
• ISSN: 0960-8966; 1873-2364
• DOI: 10.1016/j.nmd.2023.07.058

Muscle MRI is a useful tool for the diagnosis of neuromuscular diseases as it identifies selective patterns of pathology that are characteristic of a specific diagnosis. In the case of dysferlinopathy (DYSF), we have described the muscle MRI features of 182 patients included in the Clinical Outcome Study for Dysferlinopathies (COS1). Using this data, we were able to identify eight characteristic MRI features labelled “pattern rules” based on the semi-quantitative Mercuri score that were present in nearly 90% of patients. Our aim was to apply this set of rules to patients diagnosed with other genetic muscle diseases (GMD) to determine whether they were a useful diagnostic tool and create a decision tree making algorithm. We used MRI data from 182 patients with DYSF from the COS1 study and 1000 MRI scans of patients with other GMD (ANO5 68, CAPN3 82, DUX4 269, DYS 46, LMNA 80, PABPN1 168, GAA 98, SCGC 76, FKRP 39, VCP 74). We calculated sensitivity (S), specificity (E), positive and negative predictive values (PPV/NPV), accuracy (Ac) and diagnostic odds ratio for each rule (significant p value <0.05). From the eight described rules, five were more frequently seen in the DYSF cohort. Patterns of muscle involvement of FKRP related limb girdle muscular dystrophy (LGMD), followed by ANO5 related LGMD and calpain3 related LGMD, were the three most similar to DYSF, whereas OPMD, laminopathies and dystrophinopathies were the three most different diseases. We applied in a sequential order the 5 rules associated with DYSF from the one with the highest sensitivity to the one with the lowest and, when applying the last rule to the remaining cohort of patients, obtained a S 95.9%, E 46.1%, Ac 66.8%, PPV 56% and NPV 94%. Our findings support the utility of MRI in the diagnostic process of patients with DYSF and their differentiation from other GMD. We have created a decision tree algorithm based on MRI findings that could be very useful in the diagnosis of patient.