J Autoimmun. 2026 Sep 6;164:103616. doi: 10.1016/j.jaut.2026.103616. Online ahead of print.
ABSTRACT
BACKGROUND: The growing use of immune checkpoint inhibitors (ICIs) in the oncology field along with the potential development of the so-called immune-related adverse events (irAEs) has introduced new diagnostic and therapeutic challenges for rheumatologists. Rheumatic irAEs (Rh-irAEs), particularly those involving the musculoskeletal system, are relatively common and sometimes underdiagnosed, with an estimated prevalence of up to 7% of ICIs treated patients. In this context, we conducted a systematic review and pooled analyses of cohort studies to better characterize ICIs-induced-arthritis (ICIs-IA), defining demographic and clinical features of involved patients, the temporal relationship with treatment, and therapeutic implications.
METHODS: This systematic review was conducted in accordance with PRISMA framework and registered with the International Register of Prospective Reviews (PROSPERO) (registration CRD420251090954). A systematic search was performed in two electronic databases - PubMed and Scopus - including studies published from January 1st 2017, to May 31st 2024. Eligible studies reported cases of arthritis as an adverse event induced by immune checkpoint inhibitors (ICIs), including anti-PD1, anti-PDL1, anti-CTLA4, or their combinations. The Qumseya scale was used to assess the methodological quality of the studies included in the review. The meta-analysis of age and irAEs time to onset was performed using the metamean and metamedian functions and relative packages in R, which computes pooled means and confidence intervals while accounting for between-study heterogeneity.
RESULTS: We evaluated 37 studies, encompassing 882 patients with ICIs-IA. The most common underlying malignancy was melanoma (41.3%), followed by lung cancer (27.7%). Moving on treatment, 74.2% of patients received anti-PD1 drugs, 6.9% anti-PDL1, 16% anti-CTLA4; 20% of patients received a combination therapy (anti-CTLA4 plus anti-PD1). ICIs-IA patients were more frequently male (57.8% versus 42.2%), with a median age of 63.6 years (95% CI 61.9-65.2). The calculated time to onset of ICIs-IA from treatment initiation was 17.5 weeks (95% CI 13.7-21.2), ranging from 4.3 to 39.0 weeks. From a serological perspective, 7.5% of patients were positive for rheumatoid factor, 5.5% for anti-citrullinated protein antibodies and 23.4% for anti-nuclear antibodies. In terms of the clinical phenotype, 63.9% of patients experienced polyarthritis, 29.9% oligoarthritis, and 6.2% monoarthritis. Only 2.7% fulfilled classification criteria for inflammatory arthropathies or connective tissue diseases. As therapy, 67.9% of patients received glucocorticoids, 33.2% synthetic conventional DMARDs, and 11.0% biological DMARDs. Notably, ICIs-IA occurred predominantly in patients who also experienced other irAEs (64.5%).
CONCLUSIONS: In the absence of standardized diagnostic criteria, ICIs-induced arthritis remains a heterogeneous and incompletely characterized condition, as confirmed by our systematic review. Nevertheless, our analysis provides a more detailed characterization of this adverse event, which may facilitate earlier recognition and improved management.
PMID:42702183 | DOI:10.1016/j.jaut.2026.103616