ObjectiveRadiographs are frequently obtained for patients with knee osteoarthritis (KOA), with magnetic resonance imaging (MRI) reserved for those with complex KOA. There are few data on how often subsequent MRI reveals clinically actionable but unanticipated findings. The purpose of this study is to estimate the prevalence of these findings on MRI for patients managed nonoperatively for suspected meniscal tears.MethodsThe Treatment of Meniscal Problems and Osteoarthritis (TeMPO) study enrolled patients aged 45 to 85 years with knee pain, osteoarthritis (Kellgren-Lawrence [KL] grades 0-3), and suspected meniscal tear. We reviewed baseline MRI and recorded notable findings, including subchondral insufficiency fractures of the knee (SIFKs), avascular necrosis (AVN), tumors, and nonsubchondral fractures. Other baseline data included demographic characteristics, Knee Injury and Osteoarthritis Outcome Score, duration of knee symptoms, and KL grade.ResultsStudy-ordered MRI was performed on 760 patients, with 61 concerning findings identified (8.03%, 95% confidence interval 6.09%-9.96%). A total of 25 participants had SIFKs, 10 had nonsubchondral fractures, 4 had AVN, 8 had benign tumors, and 14 had other clinically relevant findings.ConclusionWe estimated the prevalence of clinically relevant incidental findings on MRI to be 8.03% in middle-aged adults with mild to moderate KOA and suspected meniscal tear. These data may prompt clinicians to be more aware of the range of findings that can underlie knee symptoms, some of which could change management but may require different modalities of imaging to detect. Future research is needed to pinpoint factors associated with these concerning findings so that patients who are at risk can be identified and referred for advanced imaging.

Objective: To investigate immunomodulator use, risk factors and management for rheumatoid arthritis (RA) flares, and mortality for patients with pre-existing RA initiating immune checkpoint inhibitors (ICI) for cancer. Methods: We performed a retrospective cohort study of all patients with RA meeting 2010 ACR/EULAR criteria that initiated ICI for cancer at Mass General Brigham or Dana-Farber Cancer Institute in Boston, MA (2011-2022). We described immunomodulator use and changes at baseline of ICI initiation. We identified RA flares after baseline, categorized the severity, and described the management. Baseline factors were examined for RA flare risk using Fine and Gray competing risk models. We performed a landmark analysis to limit the potential for immortal time bias, where the analysis started 3 months after ICI initiation. Among those who survived at least 3 months, we examined whether RA flare within 3 months after ICI initiation was associated with mortality using Cox regression. Results: Among 11,901 patients who initiated ICI for cancer treatment, we analyzed 100 pre-existing RA patients (mean age 70.3 years, 63 % female, 89 % on PD-1 monotherapy, 50 % lung cancer). At ICI initiation, 71 % were seropositive, 82 % had remission/low RA disease activity, 24 % were on glucocorticoids, 35 % were on conventional synthetic disease-modifying antirheumatic drugs (DMARDs), and 10 % were on biologic or targeted synthetic DMARDs. None discontinued glucocorticoids and 3/35 (9 %) discontinued DMARDs in anticipation of starting ICI. RA flares occurred in 46 % (incidence rate 1.84 per 1000 person-months, 95 % CI 1.30, 2.37); 31/ 100 flared within 3 months of baseline. RA flares were grade 1 in 16/46 (35 %), grade 2 in 25/46 (54 %), and grade 3 in 5/46 (11 %); 2/46 (4 %) required hospitalization for RA flare. Concomitant immune-related adverse events occurred in 15/46 (33 %) that flared. A total of 72/100 died during follow-up; 21 died within 3 months of baseline. Seropositivity had an age-adjusted sdHR of 1.95 (95 % CI 1.02, 3.71) for RA flare compared to seronegativity, accounting for competing risk of death. Otherwise, no baseline factors were associated with RA flare, including cancer type, disease activity, RA duration, and deformities. 9/46 (20 %) patients had their ICI discontinued/paused due to RA flares. In the landmark analysis among 79 patients who survived at least 3 months, RA flare in the first 3 months was not associated with lower mortality (adjusted HR 1.24, 95 % CI 0.71, 2.16) compared to no RA flare. Conclusion: Among patients with pre-existing RA, few changed immunomodulator medications in anticipation of starting ICI, but RA flares occurred in nearly half. RA flares were mostly mild and treated with typical therapies. Seropositivity was associated with RA flare risk. A minority had severe RA flares requiring disruption of ICI, and RA flares were not associated with mortality

We report the case of an 80-year-old man with restrictive strabismus in lateral gaze following multiple oculoplastic procedures for idiopathic epiphora. Despite excellent initial response to nasal conjunctival recession with lysis of adhesions and a miminal recession of the medial rectus muscle, the patient suffered recurrence of diplopia associated with limitation of abduction due to aggressive, deep, subconjunctival scarring. Given the history of oral lichen planus (LP), the patient was diagnosed with ocular involvement of LP. He underwent a second conjunctival recession, this time accompanied by an intensive LP treatment regimen. Nine months after surgery, he remained diplopia free and orthophoric in primary gaze. Surgeons treating restrictive strabismus in patients with LP should consider implementing systemic and topical immunosuppressive treatment simultaneously with surgical management.

OBJECTIVE:Inflammation is a potential pain generator and treatment target in knee osteoarthritis (OA). Inflammation can be detected on magnetic resonance imaging (MRI) and by synovial fluid white blood cell count (WBC). However, the performance characteristics of synovial fluid WBC for the detection of synovitis have not been established. This study was undertaken to determine the sensitivity and specificity of synovial fluid WBC in identifying inflammation in knee OA using MRI effusion-synovitis as the gold standard.METHODS:We identified records of patients seen at an academic center with a diagnosis code for knee OA, a procedural code for knee aspiration, and a laboratory order for synovial fluid WBC in the same encounter, as well as an MRI within 12 months of the aspiration. MRIs were read for effusion-synovitis using the MRI OA Knee Score (MOAKS). We dichotomized effusion-synovitis as 1) none or small, or 2) medium or large. We calculated the sensitivity and specificity of synovial fluid WBC using MRI effusion-synovitis (medium/large) as the gold standard. We used the Youden index to identify the best cut point.RESULTS:We included 75 patients. Mean ± SD age was 63 ± 12 years, and 69% were female. The synovial fluid WBC was higher in the medium/large effusion-synovitis group (median 335 [interquartile range (IQR) 312]) than in the none/small group (median 194 [IQR 272]). The optimal cut point was 242, yielding a sensitivity of 71% (95% confidence interval [95% CI] 56-83%) and specificity of 63% (95% CI 41-81%).CONCLUSION:The sensitivity and specificity of synovial fluid WBC in identifying effusion-synovitis on MRI were limited. Further research is needed to better understand the association between MRI and effusion-synovitis measured by synovial fluid and to determine which measure more strongly relates to synovial histopathology and patient outcomes.

Objective To systematically review the published literature on genicular artery embolization (GAE) for osteoarthritis (OA) related knee pain. Materials and Methods Using three databases, a systematic review was performed following Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) guidelines. Outcome measures included the Visual Analog Scale (VAS) and the Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC). Results Three single-arm studies were included from an initial search yielding 305 results. One hundred and eighty-six knees in 133 patients with either mild-to-moderate (174/186, 94%) or severe (12/186, 6%) OA underwent embolization with either imipenem/cilastatin sodium (159/186, 85%) or embozene (27/186, 15%). Technical success was 100%. Average VAS improved from baseline at 1 day, 1 week, 1 month, 3 months, 4 months, 6 months, 1 year and 2 years (66.5 at baseline vs 33.5, 32.7, 33.8, 28.9, 29.0, 22.3, 14.8 and 14.0, respectively). Average WOMAC scores improved from baseline at 1, 3, 4, 6, 12 and 24 months (45.7 at baseline vs 24.0, 31.0, 14.8, 14.6, 8.2 and 6.2). Severe OA in 12 cases showed initially improved VAS, but was not sustained. Minor adverse events such as erythema in the region of embolization (21/186, 11%), puncture-site hematoma (18/186, 10%), paresthesia (2/186, 1%) and fever (1/186, 0.5%) were reported. Conclusion Limited single-arm studies report GAE is promising for treating OA-related pain. Most treatments performed for mild-to-moderate OA demonstrated durable clinical responses from 6 months to 4 years. Limited data for severe OA suggest a non-durable response. Future studies should be standardized to facilitate comparison and control for placebo effect.

Purpose: Meniscal tear in persons > 45 years old is often initially managed with exercise or physical therapy (PT). Patients who do not respond may be offered arthroscopic partial meniscectomy (APM). A concern of APM is the risk for progression of underlying osteoarthritic changes due to reduced load bearing from incremental partial meniscectomy. We examined radiographic changes over 5 years in a randomized trial cohort.

Objective Randomized controlled trials (RCTs) are a gold standard for estimating the benefits of clinical interventions, but their decision-making utility can be limited by relatively short follow-up time. Longer-term follow-up of RCT participants is essential to support treatment decisions. However, as time from randomization accrues, loss to follow-up and competing events can introduce biases and require covariate adjustment even for intention-to-treat effects. We describe a process for synthesizing expert knowledge and apply this to long-term follow-up of an RCT of treatments for meniscal tears in patients with knee osteoarthritis (OA). Methods We identified 2 post-randomization events likely to impact accurate assessment of pain outcomes beyond 5 years in trial participants: loss to follow-up and total knee replacement (TKR). We conducted literature searches for covariates related to pain and TKR in individuals with knee OA and combined these with expert input. We synthesized the evidence into graphical models. Results We identified 94 potential covariates potentially related to pain and/or TKR among individuals with knee OA. Of these, 46 were identified in the literature review and 48 by expert panelists. We determined that adjustment for 50 covariates may be required to estimate the long-term effects of knee OA treatments on pain. Conclusion We present a process for combining literature reviews with expert input to synthesize existing knowledge and improve covariate selection. We apply this process to the long-term follow-up of a randomized trial and show that expert input provides additional information not obtainable from literature reviews alone.

OBJECTIVE:Meniscal tear in persons aged ≥45 years is typically managed with physical therapy (PT), and arthroscopic partial meniscectomy (APM) is offered to those who do not respond. Prior studies suggest APM may be associated with greater progression of radiographic changes. METHODS:We assessed changes between baseline and 60 months in the Kellgren-Lawrence (KL) grade and OARSI radiographic score (including subscores for joint space narrowing and osteophytes) in subjects aged 45-85 years enrolled into a seven-center randomized trial comparing outcomes of APM with PT for meniscal tear, osteoarthritis changes, and knee pain. The primary analysis classified subjects according to treatment received. To balance APM and PT groups, we developed a propensity score and used inverse probability weighting (IPW). We imputed a 60-month change in the OARSI score for subjects who underwent total knee replacement (TKR). In a sensitivity analysis, we classified subjects by randomization group. RESULTS:We analyzed data from 142 subjects (100 APM, 42 PT). The mean ± SD weighted baseline OARSI radiographic score was 3.8 ± 3.5 in the APM group and 4.0 ± 4.9 in the PT group. OARSI scores increased by a mean of 4.1 (95% confidence interval [95% CI] 3.5-4.7) in the APM group and 2.4 (95% CI 1.7-3.2) in the PT group (P < 0.001) due to changes in the osteophyte component. We did not observe statistically significant differences in the KL grade. Sensitivity analyses yielded similar findings to the primary analysis. CONCLUSION:Subjects treated with APM had greater progression in the OARSI score because of osteophyte progression but not in the KL grade. The clinical implications of these findings require investigation.

Background: Patients with pre-existing rheumatoid arthritis (RA) initiating immune checkpoint inhibitors (ICI) for cancer may be at risk for increased mortality, RA flare, and other immune-related adverse events (irAE).Methods: We did a retrospective comparative cohort study investigating mortality and irAEs for patients at Mass General Brigham and Dana-Farber Cancer Institute (1/Apr/2011-21/Apr/2021) with pre-existing RA initiating ICI for cancer. Pre-existing RA cases met 2010 ACR/EULAR criteria. We matched each case to up to 3 comparators at index date of ICI initiation by age, sex, year, ICI target, and cancer type/stage. Deaths were identified by medical record/obituary review. Medical record review determined RA flare and irAE presence, type, and severity. Multivariable Cox regression estimated adjusted hazard ratios (aHRs) and 95% confidence intervals (CIs) for mortality, any irAE, severe (grade 3+) irAE, and irAE types for pre-existing RA cases vs comparators.Findings: We analyzed 84 pre-existing RA cases (mean age 69.9 years, 60% female) and 203 comparators (mean age 69.4 years, 62% female). The most common cancer types were lung cancer (49% vs 56%) and melanoma (24% vs 25%); PD-1 inhibitors were the most common ICI (92% vs 93%). There were 58 (69%) deaths among pre-existing RA cases and 122 (60%) deaths among comparators (adjusted [a] HR 1.16, 95%CI 0.87-1.57). Any irAE was experienced by 61% of pre-existing RA cases and 49% of comparators (aHR 1.72, 95%CI 1.20-2.47); severe irAE occurred in 14% vs 15% (p=0.83). RA flares occurred in 48% of pre-existing RA cases and inflammatory arthritis occurred in 7% of comparators (p<0.0001). Pre-existing RA cases were less likely to experience rash/dermatitis (6% vs 14%, p=0.048), endocrinopathy (2% vs 11%, p=0.0078), and hepatitis (4% vs 9%, p=0.043).Interpretation: Patients with pre-existing RA initiating ICI had similar risks for mortality and severe irAE as comparators. While pre-existing RA was associated with higher risk for any all-grade irAE than comparators, this was mostly due to RA flares that were typically mild. Some irAE types were less commonly experienced among those with RA. These results suggest that pre-existing RA should not be considered a contraindication for ICI for cancer treatment.Funding Information: Dr. Yoshida is supported by the National Institute of Arthritis and Musculoskeletal and Skin Diseases (grant numbers K23 AR076453 and R03 AR081309). Dr. MacFarlane is supported by the National Institute of Arthritis and Musculoskeletal and Skin Diseases (grant number K23 AR080206). Dr. Sparks is supported by the National Institute of Arthritis and Musculoskeletal and Skin Diseases (grant numbers R01 AR077607, P30 AR070253, and P30 AR072577), the R. Bruce and Joan M. Mickey Research Scholar Fund, and the Llura Gund Award for Rheumatoid Arthritis Research and CareDeclaration of Interests: Dr. Yoshida receives consulting fees from OM1, Inc. Dr. Shadick is supported by Mallinckrodt, Lilly, BMS, AMGEN, ABBVIE, Aqtual. Dr. LeBoeuf receives consulting ees from Bayer, Seattle Genetics, Sanofi, Silverback, and Synox Therapeutics outside the submitted work. Dr. Sparks has received research support from Bristol Myers Squibb and performed consultancy for AbbVie, Amgen, Boehringer Ingelheim, Bristol Myers Squibb, Gilead, Inova Diagnostics, Janssen, Optum, and Pfizer unrelated to this work. Dr. Gedmintas is on the advisory board for Merck, BMS, and Novartis. Dr. Gedmintas is supported by Genentech. Dr. Sparks has received research support from Bristol Myers Squibb and performed consultancy for AbbVie, Amgen, Boehringer Ingelheim, Bristol Myers Squibb, Gilead, Inova Diagnostics, Janssen, Optum, and Pfizer unrelated to this work. All other authors report no conflicts of interest.Ethics Approval Statement: This study was approved by the institutional review boards of MGB and DFCI.

There is increasing use of ultrasound to identify effusion and synovitis, which are proxies for inflammation, in knee osteoarthritis (OA). Ultrasound appears to be a valid modality for assessing effusion and synovitis when compared to MRI, but its reliability has been questioned. We aimed to assess the agreement of ultrasound-identified effusion and synovitis between two readers. We included 30 unique knees from a cohort with symptomatic and radiographic knee OA. A single ultrasonographer performed ultrasound of the suprapatellar recess. Two readers with different levels of experience read the images. We specified maximal effusion depth first as continuous (mm) and then as binary, dichotomized at ≥4mm (present) vs. <4mm (absent). We graded synovitis in the suprapatellar recess (0-3) based on the Outcome Measures in Rheumatology (OMERACT) guidelines. We used proportions to determine inter-rater agreement for continuous variables and Cohen's Kappa or weighted Cohen's Kappa for agreement of binary and ordinal variables. The mean age of the participants was 66 years. Seventeen (57%) were female. For maximal effusion depth the readers differed by ≤ 2mm 80% of the time. For the presence of effusion (binary), the Cohen's Kappa for the two readers was 0.69 (95% CI 0.40-0.97). Synovitis grade (0-3) for the two readers had a weighted Cohen's Kappa of 0.76 (95% CI 0.58-0.94). We found moderate to strong inter-reader agreement of ultrasound identified effusion-synovitis in knee OA between two readers with differing ultrasound experience. Ultrasound may be a viable bedside tool for identifying the subset of OA patients with effusion or synovitis.

Immune checkpoint inhibitor (ICI) therapies used to treat cancer, such as anti–PD-1 antibodies, can induce autoimmune conditions in some individuals. The T cell mechanisms mediating such iatrogenic autoimmunity and their overlap with spontaneous autoimmune diseases remain unclear. Here, we compared T cells from the joints of 20 patients with an inflammatory arthritis induced by ICI therapy (ICI-arthritis) with two archetypal autoimmune arthritides, rheumatoid arthritis (RA) and psoriatic arthritis (PsA). Single-cell transcriptomic and antigen receptor repertoire analyses highlighted clonal expansion of an activated effector CD8 T cell population in the joints and blood of patients with ICI-arthritis. These cells were identified as CD38 hi CD127 − CD8 T cells and were uniquely enriched in ICI-arthritis joints compared with RA and PsA and also displayed an elevated interferon signature. In vitro, type I interferon induced CD8 T cells to acquire the ICI-associated CD38 hi phenotype and enhanced cytotoxic function. In a cohort of patients with advanced melanoma, ICI therapy markedly expanded circulating CD38 hi CD127 − T cells, which were frequently bound by the therapeutic anti–PD-1 drug. In patients with ICI-arthritis, drug-bound CD8 T cells in circulation showed marked clonal overlap with drug-bound CD8 T cells from synovial fluid. These results suggest that ICI therapy directly targets CD8 T cells in patients who develop ICI-arthritis and induces an autoimmune pathology that is distinct from prototypical spontaneous autoimmune arthritides.

Objective: Inflammation, manifesting as effusion and synovitis, is thought to contribute to pain in knee osteoarthritis (OA). We conducted a pilot study to investigate recruitment feasibility and assess whether effusion on ultrasound of the knee was associated with greater reduction in knee pain after corticosteroid injection. Methods: A pilot study was conducted from 2020 to 2021 including patients >= 40 years with knee OA undergoing clinically indicated corticosteroid injections. At baseline, participants completed the Knee Injury and Osteoarthritis Outcome Score (KOOS) Pain subscale and had an ultrasound of the injected knee(s) to assess for effusion, defined as anechoic material in the suprapatellar recess. KOOS Pain was re-assessed two weeks following injection. We used mixed linear models to evaluate the change in KOOS Pain scores for knees with and without effusion to determine estimates of the magnitude of association. Results: We recruited 10 participants who contributed 16 knees from 4 clinical sessions. The mean age was 68 years (standard deviation [SD] 13) and 90% were female. Six knees had effusion. At baseline, knees without effusion had greater pain (mean KOOS Pain 44, SD 19) compared to those with effusion (mean KOOS Pain 51, SD 15). Knees without effusion had a 6 point (95% CI -16, 28) greater improvement in KOOS Pain 2-weeks post injection compared to those with effusion. Conclusion: This pilot study demonstrated clinic-based recruitment was feasible. We did not observe clinically important or statistically significant differences in pain relief post corticosteroid injection between knee OA patients with or without effusion.

Objective Middle‐aged subjects with meniscal tear treated with arthroscopic partial meniscectomy (APM) experience greater progression of damage to joint structures on imaging than subjects treated nonoperatively. It is unclear whether these changes are clinically relevant. The goal of this study was to assess whether worsening in magnetic resonance imaging (MRI)–assessed tissue damage over 18 months leads to subsequent worsening in knee pain over the subsequent 3.5 years. Methods We used data from the Meniscal Tear in Osteoarthritis Research (MeTeOR) trial of APM versus physical therapy for subjects ages ≥45 years with knee pain, cartilage damage, and meniscal tear. We assessed whether change in cartilage surface area damage score (and other structural measures) from baseline to 18 months, assessed on MRI with the MRI Osteoarthritis Knee Score (MOAKS) system, was associated with change in Knee Injury and Osteoarthritis Outcome Score (KOOS) pain score (range 0–100; 100 = worst) from 18 to 60 months. Results The primary analysis included 168 subjects with complete MRI data at baseline and 18 months and KOOS data at 18 and 60 months. We did not observe clinically important associations between change in cartilage surface area score between baseline and 18 months and change in pain scores from 18 to 60 months. Pain scores in the worst tertile for cartilage surface area damage score progression worsened by 0.45 points more than in the best tertile (95% confidence interval –4.45, 5.35). Similarly, we did not observe clinically important associations between changes in bone marrow lesions, osteophytes, or synovitis and subsequent pain. Conclusion We did not observe clinically important associations between early changes in cartilage damage and other structural measures and worsening in pain over the subsequent 3.5 years. Further follow‐up is required to assess this association over a longer follow‐up period.

OBJECTIVE:Several current and many emerging osteoarthritis (OA) treatments are intraarticular (IA) injections; however, little is known about physicians' perceptions and beliefs regarding IA injections or their considerations when deciding whether to recommend them to patients. We aimed to investigate physician-perceived benefits and drawbacks of offering IA injections.METHODS:We conducted individual interviews with orthopedic surgeons, rheumatologists, and physiatrists who treat patients with knee OA. We conducted a thematic analysis to identify factors that physicians weigh when making IA injection recommendations.RESULTS:We interviewed 18 physicians from academic and community practices. We identified the following 4 categories of themes that influenced providers' recommendations to their patients regarding injections: 1) the physician's knowledge, beliefs, and concerns, including their propensity to rely on guidelines versus clinical experience, and understanding of the efficacy and risks associated with injectables, such as possible cartilage damage; 2) the characteristics of the injectable product, such as ease or number of administrations needed; 3) the individual patient-specific factors, including OA severity, comorbidities, and patient preference for and expectations of specific IA injections; and 4) the financial and administrative factors, including insurance coverage and out-of-pocket costs.CONCLUSION:Physicians factor the uncertain efficacy of injectable treatments and the need to manage patient expectations into their decision to offer IA therapies. Some providers relied on evidence and guidelines while others were swayed more by clinical experience. High out-of-pocket costs were seen as a barrier to use. These findings may help in the delivery of IA injections for OA and in the development of injectable products.

OBJECTIVE:To identify predictors of rheumatic immune-related adverse events (irAEs) following immune checkpoint inhibitor (ICI) treatment for cancer. METHODS:We performed a case-control study to predict the occurrence of rheumatic irAEs in cancer patients who initiated ICI treatment at Mass General Brigham and the Dana-Farber Cancer Institute between 2011 and 2020. We screened for the presence of rheumatic irAEs by reviewing the medical records of patients evaluated by rheumatologists or those prescribed nonglucocorticoid immunomodulatory drugs after the time of ICI initiation (baseline). Review of medical records confirmed the presence of rheumatic irAEs and the indications necessitating immunomodulatory drug treatment. Controls were defined as patients who did not experience rheumatic irAEs, did not have preexisting rheumatic disease, did not have a clinical evaluation by a rheumatologist after ICI treatment, did not receive an immunomodulatory drug after ICI, did not receive systemic glucocorticoids after ICI, and survived at least 6 months after the initial ICI treatment. We used logistic regression to estimate the odds ratios (ORs) (with 95% confidence intervals [95% CIs]) for the risk of a rheumatic irAE in the presence of various baseline predictors. RESULTS:A total of 8,028 ICI recipients were identified (mean age 65.5 years, 43.1% female, 31.8% with lung cancer). After ICI initiation, 404 patients (5.0%) were evaluated by rheumatologists, and 475 patients (5.9%) received an immunomodulatory drug to treat any irAEs. There were 226 confirmed rheumatic irAE cases (2.8%) and 118 de novo inflammatory arthritis cases (1.5%). Rheumatic diseases (either preexisting rheumatic diseases or rheumatic irAEs) were a common indication for immunomodulatory drug use (27.9%). Baseline predictors of rheumatic irAEs included melanoma (multivariable OR 4.06 [95% CI 2.54-6.51]) and genitourinary (GU) cancer (OR 2.22 [95% CI 1.39-3.54]), both relative to patients with lung cancer; combination ICI treatment (OR 2.35 [95% CI 1.48-3.74]), relative to patients receiving programmed death 1 inhibitor monotherapy; autoimmune disease (OR 2.04 [95% CI 1.45-2.85]) and recent glucocorticoid use (OR 2.13 [95% CI 1.51-2.98]), relative to patients not receiving a glucocorticoid, compared to the 2,312 controls without rheumatic irAEs. Predictors of de novo inflammatory arthritis were similar to those of rheumatic irAEs. CONCLUSION:We identified novel predictors of rheumatic irAE development in cancer patients, including baseline presence of melanoma, baseline presence of GU tract cancer, preexisting autoimmune disease, receiving or having received combination ICI treatment, and receiving or having received glucocorticoids. The proportion of cancer patients experiencing rheumatic irAEs may be even higher than was reported in the present study, since we used stringent criteria to identify cases of rheumatic irAEs. Our findings could be used to identify cancer patients at risk of developing rheumatic irAEs and de novo inflammatory arthritis and may help further elucidate the pathogenesis of rheumatic irAEs in patients with cancer who are receiving ICI treatment.