Background: Cancer is a relevant risk [1] and may precede, follow, or coincide with the diagnosis of systemic sclerosis (SSc). Risk factor identification has yielded heterogeneous results [2], with an unclear role for immunosuppressants.

Objectives: The EUSTAR Project CP154 aimed to define risk factors for cancer at different SSc timepoints, explore the impact of immunosuppressants, analyze disease trajectories and survival, and evaluate machine learning models to identify patients warranting a more precise cancer screening.

Methods: We conducted a case-control study within EUSTAR, including SSc patients with/without cancer matched 1:1 for age and SSc duration. Cancers were defined as interceptable if they were synchronous (i.e., diagnosed within 3 years before or after the first non-Raynaud symptom) or subsequent (>3 years after SSc); otherwise, cancers were categorized as previous (>3 years prior to SSc). Comparisons, multivariable logistic regression, mixed effect modeling, and survival analysis were performed. Two ensemble machine learning models (Random Forest and XGBoost) were developed and tuned to predict interceptable cancers from clinical, serological, and treatment data.

Results: We included 295 cancer patients and 293 controls from 27 EUSTAR centers: 89% female; median age 67 years (IQR 58–75); median disease duration 12 years (IQR 6–19); 34% anti-topoisomerase (ATA) and 43% anticentromere (ACA) positive. Diffuse SSc was found in 26%, digital ulcers in 24%, interstitial lung disease (ILD) in 35%, pulmonary arterial hypertension (PAH) in 7.5%, esophageal involvement in 57%. Cancer temporal distribution was as follows: synchronous in 82 patients, previous in 55 (median 9 years before SSc), subsequent in 158 (median 11 years after SSc). Breast cancer was the most common malignancy (32%), followed by lung (16%), gynecological (8%), colorectal (7.5%), and hematological cancers (7%). Advanced-stage malignancies were diagnosed in 37% of cases. Systemic treatments included chemo- (38%), antihormonal (23%), targeted (7%) and immunotherapy (2.6%). On multivariable analysis, risk factors for synchronous cancers included smoking (OR 2.0; 95% CI 1.1-3.5; p=0.02), diffuse SSc (OR 1.9; CI 1.1-3.4; p=0.03), anti-POLR3 (OR 2.6; CI 1.2-5.7; p=0.01), PM/Scl (OR 4.7; CI 1.2-19; p=0.03), RNP (OR 6.8; CI 1.6-29; p=0.01), and C-reactive protein >6 mg/L (OR 1.9; CI 1.1-3.5; p=0.03). Digital ulcers were protective (OR 0.5; CI 0.2-0.9; p=0.03). Patients with synchronous cancers were also more frequently negative for both ATA and ACA (39% vs 23%, p=0.005), especially in diffuse SSc (22% vs 5.5%, p<0.001). Risk factors for interceptable cancers were similar to those of synchronous cancers. Patients with previous cancer showed no difference compared to cancer-free patients, except for more anti-POLR3 (16% vs 6.8%) and less ulcers (15% vs 28%; both p<0.05). Among immunosuppressants, cyclophosphamide (CYC) was associated with subsequent cancers (OR 2.2; CI 1.1-4.8; p=0.04), while mycophenolate mofetil (MMF) showed a negative association (OR 0.5; CI 0.3-0.9; p=0.03). Compared to cases without malignancy, breast cancers were more frequently ATA and ACA negative, particularly in diffuse SSc (18% vs 5.5%; p<0.001), with anti-POLR3 (20% vs 6.8%; p<0.001) or PM/Scl (9.5% vs 1.7%; p=0.001). Patients with lung cancer were more frequently male (25% vs. 10%, p=0.006), smokers (54% vs. 26%, p<0.001), ATA+ (60% vs. 34%, p=0.001), with ILD (56% vs. 33%, p=0.003), lower FVC (p=0.015) and DLCO (p=0.002). At the end of follow-up, 545 (93%) patients were alive. After adjusting for age and duration, cancer significantly reduced overall survival (HR 2.4; CI 1.1-5.0; p=0.02): in particular, interceptable (HR 3.3; CI 1.5-7; p=0.002) and, especially, subsequent (HR 4.2; CI 1.9-9.5; p<0.001) cancers showed the greatest impact. Longitudinal analysis of 380 patients (median follow-up 5 years [2–10]) revealed no differences in the disease trajectories (incidence or trend of manifestations), except for a higher incidence of muscle atrophy in cancer patients (OR 1.95). Random Forest and XGBoost machine learning models were trained and tuned to identify interceptable cancer cases, demonstrating good accuracy (73% vs 79%) and performance (AUC _ROC 0.88 both). Random Forest had higher sensitivity (98% vs 90%) and precision (94.5% vs 77%) but lower specificity (44% vs 66%). Key predictors included baseline ILD, digital ulcers, esophageal involvement, telangiectasia, and high CRP while MMF was protective in both models.

Conclusion: Emerging risk factors for cancer in systemic sclerosis include autoantibodies, their interplay with disease features, and the selective risk associated with CYC, but not other immunosuppressants like MMF. Malignancy significantly influences SSc survival, with interceptable cancers exerting the greatest impact. We propose personalized screening strategies, supported by machine learning algorithms, to improve early cancer detection in this population.

REFERENCES: [1] Elhai M, Ann Rheum Dis. 2017

[2] Carbonell C, Autoimmun Rev. 2022

Acknowledgements: On behalf of EUSTAR co-authors.

Disclosure of Interests: Antonio Tonutti: None declared, Francesca Motta: None declared, Stefano Erba: None declared, Brigitte Granel: None declared, Britta Maurer: None declared, Cristiana Sieiro Santos: None declared, Duygu Temiz Karadağ: None declared, Elena Rezus: None declared, Elisabetta Zanatta: None declared, Fabiola Atzeni: None declared, Francesco Benvenuti: None declared, Gábor Kumánovics: None declared, Gabriella Szucs: None declared, Gianluca Moroncini: None declared, Gonçalo Boleto: None declared, Leila Caillault: None declared, Masataka Kuwana: None declared, Massimiliano Limonta: None declared, Maurizio Cutolo: None declared, Montserrat Bordoy Pastor: None declared, Oliver Distler 4P-Pharma, Abbvie, Acceleron, Acepodia Biotech, Aera, Alcimed, Altavant, Amgen, AnaMar, Anaveon AG, Argenx, AstraZeneca, Blade, Bayer, Boehringer Ingelheim, Calluna (Arxx), Cantargia AB, Catalyze Capital, Corbus, CSL Behring, Galderma, Galapagos, Glenmark, Gossamer, Horizon, Janssen, Kymera, Lupin, Medscape, MSD Merck, Miltenyi Biotec, Mitsubishi Tanabe, Nkarta Inc., Novartis, Orion, Pilan, Prometheus, Quell, Redxpharma, Roivant, EMD Serono, Topadur, and UCB, OD has a patent issued for “mir-29 for the treatment of systemic sclerosis” (US8247389, EP2331143). OD is a co-founder of CITUS AG, 4P-Pharma, Abbvie, Acceleron, Acepodia Biotech, Aera, Alcimed, Altavant, Amgen, AnaMar, Anaveon AG, Argenx, AstraZeneca, Blade, Bayer, Boehringer Ingelheim, Calluna (Arxx), Cantargia AB, Catalyze Capital, Corbus, CSL Behring, Galderma, Galapagos, Glenmark, Gossamer, Horizon, Janssen, Kymera, Lupin, Medscape, MSD Merck, Miltenyi Biotec, Mitsubishi Tanabe, Nkarta Inc., Novartis, Orion, Pilan, Prometheus, Quell, Redxpharma, Roivant, EMD Serono, Topadur, and UCB, BI, Kymera, Mitsubishi Tanabe, UCB, Radim Bečvář: None declared, Roberto Giacomelli: None declared, Rossella De Angelis: None declared, Serena Guiducci: None declared, Sophie Blaise: None declared, Stefano Stano: None declared, Valeria Riccieri: None declared, Yasser El Miedany: None declared, Carlo Selmi: None declared, Maria De Santis: None declared.

© The Authors 2025. This abstract is an open access article published in Annals of Rheumatic Diseases under the CC BY-NC-ND license ( http://creativecommons.org/licenses/by-nc-nd/4.0/ ). Neither EULAR nor the publisher make any representation as to the accuracy of the content. The authors are solely responsible for the content in their abstract including accuracy of the facts, statements, results, conclusion, citing resources etc.