Abstract
BACKGROUND: Although patients with idiopathic inflammatory myopathy (IIM) have a higher cancer risk than the general population, this risk varies among IIM patients based on clinical factors such as IIM subtype, clinical features, and autoantibody profiles. This study aimed to establish a risk-stratification system to identify those with high-risk for cancer in patients with IIM. METHODS: This study included 481 patients with IIMs from four independent cohorts in South Korea. The primary outcome was myositis-associated cancer, defined as cancers occurring within 3 years before or after the diagnosis of IIM. Logistic regression with Least Absolute Shrinkage and Selection Operator (LASSO) regularization was used to select predictors from 24 prespecified factors for the multivariable model. A scoring system, entitled the SCRIM score, was established based on the β coefficient of each predictor in the final model. A conditional inference tree analysis was used to identify the optimal cut-off point for group classifications. The cancer risk of each group compared to that of the general population was assessed using standardized incidence ratio (SIR), which was calculated using the 2022 age-matched general population in South Korea as the reference. RESULTS: In the study population, 12.9% (n = 62) had myositis-associated cancer. The LASSO regression identified older age, male sex, current smoking status, dermatomyositis, and anti-TIF1γ as high-risk factors for cancer; however, anti-synthetase syndrome, overlap myositis, arthritis, and interstitial lung disease were associated with a lower cancer risk. The SCRIM score based on these factors showed excellent performance (AUROC 0.852 [95% CI 0.796-0.907]). Patients were stratified into the low-risk, intermediate-risk, and high-risk groups based on the SCRIM score, and the incidence of myositis-associated cancer increased significantly in a stepwise manner across groups. Patients in the low-risk group had a cancer incidence comparable to that of the general population (SIR 0.79), whereas those in the intermediate-risk (SIR 2.28) and high-risk (SIR 15.30) groups had significantly higher cancer incidences. CONCLUSION: The SCRIM score precisely stratified the risk of myositis-associated cancer for patients with IIM, which can be an important basis for establishing an evidence-based screening cancer strategy.