Whole-body MRI in patients with arthralgia or inflammatory arthritis after exposure to immune checkpoint inhibitors: a single-centre prospective imaging study.

Background: Musculoskeletal adverse events due to immune checkpoint inhibitors are common and can present clinically as inflammatory arthritis, polymyalgia rheumatica, or arthralgia. The pathoanatomy of musculoskeletal adverse events related to immune checkpoint inhibitors remains undefined, with a paucity of available imaging data. We aimed to investigate the whole-body imaging phenotype of arthralgia and inflammatory arthritis following exposure to immune checkpoint inhibitors, to fully characterise the pattern of inflammation in these patients and subsequently inform clinical management.

Methods: In this prospective imaging study, patients aged 18 years or older with new musculoskeletal symptoms that started during or up to 6 months after receiving an immune checkpoint inhibitor and healthy controls aged 18 years or older, with no personal history of rheumatological autoimmune disease, no active cancer, and no self-reported joint pains in the 4 weeks before their MRI scan date, were recruited at the Leeds Rheumatology department of Chapel Allerton Hospital, Leeds, UK, and underwent gadolinium contrast-enhanced whole-body MRI. Joint, tendon, bursal, entheseal, and whole spinal imaging lesions were graded by two independent masked assessors and consensus reported. Inflammatory whole-body MRI patterns were analysed and patients were followed up for 6 months. People with lived experience of inflammatory arthritis and musculoskeletal toxicity related to immune checkpoint inhibitors highlighted the importance of knowing and understanding imaging findings to help inform risk versus benefit decisions about immunosuppressive treatments.

Findings: Between Oct 20, 2021, and May 22, 2024, 60 patients (35 [58%] with arthralgia and 25 [42%] with inflammatory arthritis) and 20 healthy controls were recruited. The mean age of patients was 65 years (SD 11) and that of healthy controls was 62 years (7); 34 (57%) patients were men and 26 (43%) were women, and 12 (60%) healthy controls were men and eight (40%) were women. All patients and healthy controls were White. Median total joint synovitis, joint erosions, enthesitis, and tenosynovitis scores were significantly higher in patients with arthralgia or inflammatory arthritis induced by immune checkpoint inhibitors compared with healthy controls, without significant differences between the inflammatory arthritis and arthralgia subgroups. Acromioclavicular (46 [77%] of 60), glenohumeral (45 [75%] of 60), wrist (43 [73%] of 59), and metacarpophalangeal (35 [59%] of 59) joints were the most frequently affected by synovitis in all patients. There were three distinct global inflammatory patterns: peripheral inflammatory arthritis in 22 (37%) of 60 patients; polymyalgia rheumatica in seven (12%), and an overlapping phenotype of polymyalgia rheumatica and peripheral inflammatory arthritis in 12 (20%). Axial inflammation was only identified in one patient. Four of the five patients requiring disease-modifying antirheumatic drug therapy were in the peripheral inflammatory arthritis group, which also had the highest initial and ongoing glucocorticoid requirement.

Interpretation: MRI inflammation and erosions are as prevalent in patients with arthralgia exposed to an immune checkpoint inhibitor as in those with inflammatory arthritis. This finding suggests that the overall burden of musculoskeletal toxicity associated with immune checkpoint inhibitors is currently under-recognised. Patients who develop inflammatory arthritis or arthralgia after exposure to immune checkpoint inhibitors have three main imaging patterns: polymyalgia rheumatica, peripheral inflammatory arthritis, and an overlap of polymyalgia rheumatica and inflammatory arthritis. Patients with peripheral inflammatory arthritis were most likely to require disease-modifying antirheumatic drugs.

Funding: The National Institute for Health and Care Research (NIHR) Leeds Biomedical Research Centre.