Background:,Cutaneous dermatomyositis (DM) poses diagnostic and monitoring challenges due to its heterogeneous skin manifestations. The Cutaneous Dermatomyositis Disease Area and Severity Index (CDASI) is the standard for assessing disease activity, yet its complexity hinders routine use. Large language models (LLMs), such as Claude v3.5 Sonnet, offer potential for automated scoring through image interpretation and structured reasoning, but their utility in dermatologic assessment remains underexplored.,Materials:,We retrospectively analyzed 30 published DM cases with standardized clinical images suitable for CDASI scoring. Two expert rheumatologists independently scored each case using CDASI criteria. Claude v3.5 Sonnet was provided with identical images and prompted using a chain-of-thought strategy incorporating structured CDASI definitions. Intraclass correlation coefficients (ICCs) were calculated to assess concordance between Claude and human raters.,Results:,Claude demonstrated strong agreement with expert assessors in global CDASI scoring (ICC vs Expert 1: 0.92; vs Expert 2: 0.87), comparable to inter-expert reliability (ICC: 0.87). Domain-specific agreement was moderate for erythema, scaling, and ulceration (ICCs: 0.57–0.61), but lower for poikiloderma (ICC: 0.47) and chronic damage (ICC: 0.37). Hand assessments showed excellent reliability, including perfect agreement in periungual changes detection (ICC: 1.0). The LLM scored cases 92% faster than experts, averaging 42 seconds per case.,Conclusions:,Claude v3.5 Sonnet achieved expert-level reliability in scoring cutaneous DM, particularly in visually distinct domains. Its efficiency and scalability suggest potential utility in clinical trial screening and decision support, though limitations remain in recognizing subtle or chronic changes. Integration into hybrid human-AI workflows may enhance dermatologic assessment.

Large language models (LLMs) with multimodal capabilities may support automated assessment of cutaneous disease activity in dermatomyositis (DM). We evaluated the performance of Claude v3.5 Sonnet in scoring the Cutaneous Dermatomyositis Disease Area and Severity Index (CDASI) compared with expert rheumatologists. Thirty published DM cases with available clinical images were retrospectively analyzed. Two expert rheumatologists independently scored CDASI activity and damage domains. Claude v3.5 Sonnet assessed the same images using structured prompting based on CDASI definitions. Agreement was evaluated using intraclass correlation coefficients (ICCs) with a two-way random-effects model for absolute agreement. The LLM demonstrated better agreement for activity than for damage assessment. Global CDASI activity showed good agreement with expert raters (ICC 0.71, 95

Juvenile dermatomyositis (JDM), the most common pediatric inflammatory myopathy, is associated with significant morbidity despite therapeutic advances. Distinct clinical phenotypes have emerged, which can correlate with myositis-specific antibodies. Because translational data solidify the role of type I IFNs in JDM disease pathogenesis, integration of clinical and molecular phenotyping may impact the choice of targeted therapy. This paper reviews clinical and molecular phenotyping in JDM and translational insights into immune pathogenesis that have created emerging options for targeted therapy.

Recurrent exposures to a pathogenic antigen remodel the CD8(+) T cell compartment and generate a functional memory repertoire that is polyclonal and complex. At the clonotype level, the response to the conserved influenza antigen, M1(58-66) has been well characterized in healthy individuals, but not in patients receiving immunosuppressive therapy or with aberrant immunity, such as those with juvenile idiopathic arthritis (JIA). Here we show that patients with JIA have a reduced number of M1(58-66) specific RS/RA clonotypes, indicating decreased clonal richness and, as a result, have lower repertoire diversity. By using a rank-frequency approach to analyze the distribution of the repertoire, we found several characteristics of the JIA T cell repertoire to be akin to repertoires seen in healthy adults, including an amplified RS/RA-specific antigen response, representing greater clonal unevenness. Unlike mature repertoires, however, there is more fluctuation in clonotype distribution, less clonotype stability, and more variable IFNy response of the M1(58-66) specific RS/RA clonotypes in JIA. This indicates that functional clonal expansion is altered in patients with JIA on immunosuppressive therapies. We propose that the response to the influenza M1(58-66) epitope described here is a general phenomenon for JIA patients receiving immunosuppressive therapy, and that the changes in clonal richness and unevenness indicate a retarded and uneven generation of a mature immune response.

OBJECTIVE:The objective was to develop consensus treatment plans (CTPs) for patients with refractory moderately severe juvenile dermatomyositis (JDM) treated with biologic disease-modifying antirheumatic drugs (bDMARDs). METHODS:The Biologics Workgroup of the Childhood Arthritis and Rheumatology Research Alliance JDM Research Committee used case-based surveys, consensus framework, and nominal group technique to produce bDMARD CTPs for patients with refractory moderately severe JDM. RESULTS:Four bDMARD CTPs were proposed: tumor necrosis factor α (TNFα) inhibitor (adalimumab or infliximab), abatacept, rituximab, and tocilizumab. Each CTP has different options for dosing and/or route. Among 76 respondents, consensus was achieved for the proposed CTPs (93% [67 of 72]) as well as for patient characteristics, assessments, outcome measures, and follow-up. By weighted average, respondents indicated that they would most likely administer rituximab, followed by abatacept, TNFα inhibitor, and tocilizumab. CONCLUSION:CTPs for the administration of bDMARDs in refractory moderately severe JDM were developed using consensus methodology. The implementation of the bDMARD CTPs will lay the groundwork for registry-based prospective comparative effectiveness studies.

Journal Article Accepted manuscript Integrating Large Language Models in Medicine: A Study of Claude 2's Performance in MDAAT Scoring for idiopathic inflammatory myopathies Get access Vincenzo Venerito, Vincenzo Venerito Department of Emergency and Organ Transplantation, Rheumatology Unit, Bari, Italy https://orcid.org/0000-0002-2573-5930 Search for other works by this author on: Oxford Academic PubMed Google Scholar Marco Fornaro, Marco Fornaro Department of Emergency and Organ Transplantation, Rheumatology Unit, Bari, Italy https://orcid.org/0000-0003-1716-7432 Search for other works by this author on: Oxford Academic PubMed Google Scholar Sara Sabbagh, Sara Sabbagh Division of Rheumatology, Department of Pediatrics, Medical College of Wisconsin, Milwaukee, WI, USA https://orcid.org/0000-0003-3176-9958 Search for other works by this author on: Oxford Academic PubMed Google Scholar Shounak Ghosh, Shounak Ghosh Department of Rheumatology, Calcutta Medical Research Institute, Kolkata, India https://orcid.org/0000-0002-3546-4695 Search for other works by this author on: Oxford Academic PubMed Google Scholar Hector Chinoy, Hector Chinoy Division of Musculoskeletal and Dermatological Sciences, Faculty of Biology, Medicine and Health, The University of Manchester, Manchester, UKDepartment of Rheumatology, Salford Royal Hospital, Northern Care Alliance NHS Foundation Trust, Manchester Academic Health Science Centre, Salford, UK https://orcid.org/0000-0001-6492-1288 Search for other works by this author on: Oxford Academic PubMed Google Scholar Latika Gupta, Latika Gupta Department of Rheumatology, Salford Royal Hospital, Northern Care Alliance NHS Foundation Trust, Manchester Academic Health Science Centre, Salford, UKDepartment of Rheumatology, Royal Wolverhampton Hospitals NHS Trust, Wolverhampton, UK Corresponding author details: Latika Gupta, Department of Rheumatology, Royal Wolverhampton Hospitals NHS Trust, Wolverhampton WV10 0QP, UK. drlatikagupta@gmail.com https://orcid.org/0000-0003-2753-2990 Search for other works by this author on: Oxford Academic PubMed Google Scholar MyoLLM investigators MyoLLM investigators Search for other works by this author on: Oxford Academic PubMed Google Scholar Rheumatology, keae233, https://doi.org/10.1093/rheumatology/keae233 Published: 22 April 2024 Article history Received: 31 January 2024 Revision received: 12 April 2024 Accepted: 13 April 2024 Published: 22 April 2024

Objectives AECAs are detected in multiple forms of vasculitis or vasculopathy, including JDM. High levels of tropomyosin alpha-4 chain (TPM4) gene expression in cutaneous lesions and TPM4 protein expression in some endothelial cells (ECs) have been proven. Furthermore, the presence of autoantibodies to tropomyosin proteins have been discovered in DM. We therefore investigated whether anti-TPM4 autoantibodies are an AECA in JDM and are correlated with clinical features of JDM. Methods The expression of TPM4 protein in cultured normal human dermal microvascular ECs was investigated by Western blotting. Plasma samples from 63 children with JDM, 50 children with polyarticular JIA (pJIA) and 40 healthy children (HC) were tested for the presence of anti-TPM4 autoantibodies using an ELISA. Clinical features were compared between JDM patients with and without anti-TPM4 autoantibodies. Results Autoantibodies to TPM4 were detected in the plasma of 30% of JDM, 2% of pJIA (P < 0.0001) and 0% of HC (P < 0.0001). In JDM, anti-TPM4 autoantibodies were associated with the presence of cutaneous ulcers (53%; P = 0.02), shawl sign rash (47%; P = 0.03), mucous membrane lesions (84%; P = 0.04) and subcutaneous edema (42%; P < 0.05). Anti-TPM4 autoantibodies significantly correlated with the use of intravenous steroids and IVIG therapy in JDM (both P = 0.01). The total number of medications received was higher in patients with anti-TPM4 autoantibodies (P = 0.02). Conclusion Anti-TPM4 autoantibodies are detected frequently in children with JDM and are novel myositis-associated autoantibodies. Their presence correlates with vasculopathic and other cutaneous manifestations of JDM that may be indicative of more refractory disease.

OBJECTIVES Four-and-a-half LIM domains 1 (FHL1) is a muscle-specific protein. Autoantibodies against FHL1 were recently discovered in adults with idiopathic inflammatory myopathies (IIM) and were found to be associated with clinical features and outcomes indicative of increased disease severity. Anti-FHL1 autoantibodies have not been described in children. Here, the prevalence and clinical features associated with anti-FHL1 autoantibodies were examined in a large North American cohort of juvenile patients with IIM. METHODS Sera from 338 juvenile IIM patients and 91 juvenile healthy controls were screened for anti-FHL1 autoantibodies by ELISA. Clinical characteristics and HLA alleles of those with and without anti-FHL1 autoantibodies were compared among those with juvenile IIM. RESULTS Anti-FHL1 autoantibodies were present in 10.9% of juvenile IIM patients and 1.1% of controls. The frequency of anti-FHL1 autoantibodies among clinical and serologic subgroups did not differ. A higher percentage of Asian patients had anti-FHL1 autoantibodies (11% vs 0.7%, p= 0.002). Myositis associated autoantibodies (MAAs) (OR 2.09; CI 1.03, 4.32), anti-Ro52 autoantibodies specifically (OR 4.17; CI 1.83, 9.37), and V-sign rash (OR 2.59; CI 1.22, 5.40) were associated with anti-FHL1 autoantibodies. There were no differences in other features or markers of disease severity. No HLA associations with anti-FHL1 autoantibodies in Caucasian myositis patients were identified. CONCLUSION Anti-FHL1 autoantibodies are present in ∼11% of juvenile IIM patients and commonly co-occur with MAAs, including anti-Ro52 autoantibodies. In contrast to adult IIM, anti-FHL1 autoantibodies in juvenile myositis are associated with V-sign rash but not with other distinctive clinical features or worse outcomes.

Objectives JDM is an inflammatory myopathy characterized by prominent vasculopathy. AECAs are frequently detected in inflammatory and autoimmune diseases. We sought to determine whether AECAs correlate with clinical features of JDM, and thus serve as biomarkers to guide therapy or predict outcome. Methods Plasma samples from 63 patients with JDM, 49 patients with polyarticular JIA and 40 juvenile healthy controls were used to detect anti-heat shock cognate 71 kDa protein (HSC70) autoantibodies, a newly identified AECA, in ELISA assays. Clinical features were compared between JDM patients with and without anti-HSC70 autoantibodies. Results Anti-HSC70 autoantibodies were detected in 35% of patients with JDM, in 0% of patients with JIA (P < 0.0001) and in 0% of healthy donors (P < 0.0001). Both the presence of cutaneous ulcers (59% vs 17%, P < 0.002) and the use of wheelchairs and/or assistive devices (64% vs 27%, P < 0.007) were strongly associated with anti-HSC70 autoantibodies in JDM. High scores on the severity of myositis damage measures at the time of measurement of anti-HSC70 autoantibodies and an increased number of hospitalizations were also associated with anti-HSC70 autoantibodies. Intravenous immunoglobulin therapy was used more often in anti-HSC70 autoantibody-positive patients. Conclusion Anti-HCS70 autoantibodies are detected frequently in children with JDM and are novel myositis-associated autoantibodies correlating with disease severity.

ObjectiveTo define the prevalence and clinical phenotype of anti‐cortactin autoantibodies in adult and juvenile myositis.MethodsIn this longitudinal cohort study, anti‐cortactin autoantibody titers were assessed by enzyme‐linked immunosorbent assay in 670 adult myositis patients and 343 juvenile myositis patients as well as in 202 adult healthy controls and 90 juvenile healthy controls. The prevalence of anti‐cortactin autoantibodies was compared among groups. Clinical features of patients with and those without anti‐cortactin autoantibodies were also compared.ResultsAnti‐cortactin autoantibodies were more common in adult dermatomyositis (DM) patients (15%; P = 0.005), particularly those with coexisting anti–Mi‐2 autoantibodies (24%; P = 0.03) or anti–NXP‐2 autoantibodies (23%; P = 0.04). In adult myositis, anti‐cortactin was associated with DM skin involvement (62% of patients with anti‐cortactin versus 38% of patients without anti‐cortactin; P = 0.03), dysphagia (36% versus 17%; P = 0.02) and coexisting anti–Ro 52 autoantibodies (47% versus 26%; P = 0.001) or anti‐NT5c1a autoantibodies (59% versus 33%; P = 0.001). Moreover, the titers of anti‐cortactin antibodies were higher in patients with interstitial lung disease (0.15 versus 0.12 arbitrary units; P = 0.03). The prevalence of anti‐cortactin autoantibodies was not different in juvenile myositis patients (2%) or in any juvenile myositis subgroup compared to juvenile healthy controls (4%). Nonetheless, juvenile myositis patients with these autoantibodies had a higher prevalence of “mechanic's hands” (25% versus 7%; P = 0.03), a higher number of hospitalizations (2.9 versus 1.3; P = 0.04), and lower peak creatine kinase values (368 versus 818 IU/liter; P = 0.02) than those without anti‐cortactin.ConclusionThe prevalence of anti‐cortactin autoantibodies is increased in adult DM patients with coexisting anti–Mi‐2 or anti–NXP‐2 autoantibodies. In adults, anti‐cortactin autoantibodies are associated with dysphagia and interstitial lung disease.

OBJECTIVES Pneumocystis jirovecii pneumonia (PJP) is associated with significant morbidity and mortality in adult myositis patients; however, there are few studies examining PJP in juvenile myositis [juvenile idiopathic inflammatory myopathy (JIIM)]. The purpose of this study was to determine the risk factors and clinical phenotypes associated with PJP in JIIM. METHODS An research electronic data capture (REDCap) questionnaire regarding myositis features, disease course, medications and PJP infection characteristics was completed by treating physicians for 13 JIIM patients who developed PJP (PJP+) from the USA and Canada. Myositis features and medications were compared with 147 JIIM patients without PJP (PJP-) from similar geographic regions who enrolled in National Institutes of Health natural history studies. RESULTS PJP+ patients were more often of Asian ancestry than PJP- patients [odds ratio (OR) 8.7; 95% CI 1.3, 57.9]. Anti- melanoma differentiation associated protein 5 (MDA5) autoantibodies (OR 12.5; 95% CI 3.0, 52.4), digital infarcts (OR 43.8; 95% CI 4.2, 460.2), skin ulcerations (OR 12.0; 95% CI 3.5, 41.2) and interstitial lung disease (OR 10.6; 95% CI 2.1, 53.9) were more frequent in PJP+ patients. Before PJP diagnosis, patients more frequently received pulse steroids, rituximab and more immunosuppressive therapy compared with PJP- patients. Seven PJP+ patients were admitted to the intensive care unit and four patients died due to PJP or its complications. CONCLUSIONS PJP is a severe infection in JIIM that can be associated with mortality. Having PJP was associated with more immunosuppressive therapy, anti-MDA5 autoantibodies, Asian race and certain clinical features, including digital infarcts, cutaneous ulcerations and interstitial lung disease. Prophylaxis for PJP should be considered in juvenile myositis patients with these features.