Ab0246 Comparison in the Detection Ability of Myositis-Specific and Myositis-Associated Antibodies by Lineblot Method and Immunoprecipitation in Patients With Autoimmune Inflammatory Myopathies

Background: Idiopathic inflammatory myopathies (IIM) represent a group of systemic autoimmune diseases with different clinical phenotypes and associated antibodies: myositis-specific antibodies (MSA) and myositis-associated antibodies (MAA). The reference method for detection of MAA/MSA is considered immunoprecipitation (IP) [1], but in clinical practice it is replaced by semi-automated methods, such as lineblot (LB). Objectives: to evaluate the concordance rate between LB and IP in detecting MSA/MAA and the agreement with clinical diagnoses. Methods: Sixty-nine patients diagnosed with IIM [10 antisynthetase syndrome (ASS), 27 dermatomyositis (DM), 24 polymyositis (PM) and 8 overlap myositis] were included. Each serum was analyzed in LB (Myositis Antigens Profile 3 (IgG) EUROLINE, Euroimmun) and IP (both RNA and protein antigens). The degree of concordance was calculated by Cohen's coefficient. Results: MSA were detected in 37% of sera with LB and 55% with IP, while multiple MSAs were found (only with LB) in 2 sera (2.8%). The most frequent MSAs with LB and IP are anti-Jo1 (16% and 14% respectively), followed by anti-ASS not Jo-1 (10% and 7% respectively), anti-SRP, anti-Mi2, anti-NXP2. Among MAAs, the most represented are anti-Ro52, followed by anti-PM/Scl and anti-Ku (Table 1). A good concordance rate between methods was found for anti-Jo-1 (k=0.72), anti-Ku (k=0.73), anti-PM/Scl (k=0.72) and anti-Mi2 (k=0.65); a moderate concordance rate for anti-TIF1ɣ (k=0.48); anti-EJ, anti-SRP and anti-Ro52 have a low concordance rate (k=0.37, 0.24, 0.2, respectively). In particular, LB detects anti-Ro52 in 46.3% of sera while IP in 11.6%. Anti-NXP2 and anti-MDA5 are positive only in IP, in 9 and 5 sera respectively. Assessing concordance with clinical diagnosis, anti-Jo1 associates with ASS (72% in LB, 50% in IP) or with PM (27% with LB, 50% with IP). Non-Jo-1 anti-ASS associates with ASS only in 28% with both methods, the remaining with DM or PM. Patients with anti-SRP (5 with LB and 3 with IP) have a diagnosis of PM in 4 cases (2/5 with LB and 2/3 with IP), while the rest shows an incongruous diagnosis (DM, ASS and overlap myositis). Anti-MDA5, TIF1ɣ and NXP2 (detected predominantly with IP) are all associated with DM; anti-Mi2 is associated with DM in 2/3 cases (both in LB and IP), 1/3 with PM. Anti-PM/Scl positives have overlap diagnosis in 2/10 with LB and 2/4 in IP; for anti-Ku such diagnosis is present in 2/5 in LB and 2/3 in IP. Conclusion: The two methods show good concordance for Jo1, Mi2 and two MAAs (PM/Scl and Ku). These specificities also show good concordance with the related clinical diagnoses. IP shows a very good performance for the detection of DM-associated MSA, poorly detected with LB, in particular anti-NXP2 and anti-MDA5 were positive only with IP, and none were found with LB. LB appears more sensitive in detecting anti-Ro 52kD antibodies but also identified multiple MSAs not detected by IP REFERENCES: [1] Betteridge Z, et al. J Autoimmun 2019;101:48-55. Acknowledgements: NIL. Disclosure of Interests: None declared.