Pos1560 using survival classification and regression trees and survival random forest to investigate associations among factors for cancer progression and arthritis control in a multi-center retrospective cohort of patients with immune checkpoint inhibitor associated inflammatory arthritis

Background It is unknown whether particular presentations of immune checkpoint inhibitor-associated inflammatory arthritis (ICI-IA) are associated with better cancer responses or faster time to arthritis control. Machine learning methods have the ability to determine important factors in datasets where there are often many variables. Objectives To identify variables associated with arthritis control and cancer progression among persons with ICI-IA. Methods This study is ancillary to a retrospective observational study of 147 DMARD-treated (TNFi, IL-6Ri, or methotrexate) patients with ICI-IA seen at 6 U.S sites. Variables from the medical record included demographics, patterns of swollen joints, medications, lab values, and concomitant irAEs. Arthritis control and cancer progression were the two outcomes analyzed. survival Classification and Regression trees (sCART) were created using the rpart and partykit R packages [1] . Random Survival Forest (RSF) was performed using the R package randomForestSRC [2] . Variable importance (VI) for sCART and Relative Importance Score (RIS) were computed to assess influential variables.. Results We analyzed 147 people with ICI-IA. 69% received PD-1/PD-L1 monotherapy, and 43% had melanoma. ICI-A treatment was a TNFi in 17%, IL6Ri in 26%, methotrexate in 33%, and 24% received >1 DMARD. Median time to cancer progression was 333 (IQR 110, 811) days for the 26% that progressed. Median time to arthritis control was 109 days (IQR 32, 287) for the 93% that achieved control. For cancer progression the following were identified by both sCART and RSF as important: steroid duration, total joint count (TJC), study site, maximum steroid dose, ICI type, shoulder arthritis, >1 DMARD and number of IRAEs. For classifying arthritis control, the following variables were found to be important in both sCART and RSF: steroid duration, >1 DMARD, elbow arthritis, age, cancer type, TJC and first DMARD (Table 1). The Figure 1 shows the sCART for arthritis control. Conclusion Both methods, SCART and RSF, demonstrated the important influence of steroid duration on arthritis control and cancer progression. Machine learning methods demonstrated the potential prognostic importance of specific joint involvement for each outcome- knee for time to arthritis control and shoulder and wrist for cancer progression. References [1]Bou-Hamad I, Larocque D, Ben-Ameur H. A review of survival trees. Stat Surv . 2011;5:44-71. [2]Ishwaran H, Kogalur UB, Blackstone EH, Lauer MS. Random survival forests. Ann Appl Stat . 2008;2(3):841-860. Table 1. Demographics, Variable Importance (VIMP) from survival Classification and Regression Trees (sCART) and Relative Importance Score (RIS) from Random Survival Forest (RSF). Arthritis Control Cancer Progression Characteristic (%) Or median(IQR) sCART VIMP 1 RSF RIS 2 sCART VIMP 1 RSF RIS 2 Steroid duration, days 318 (181, 658) 12 1.0 23 1 Cancer type 6 0.49 14 First DMARD class 3 0.37 2 >1 DMARD 36 (24.5) 11 0.33 4 0.37 ICI type* 1 0.05 5 0.22 Study site 19 0.02 8 0.36 Age (years) 62.3 (52.2, 69.0) 7 0.42 10 0.04 Maximum prednisone dose (mg) 40 (20,60) 4 6 0.10 Arthritis affected Joints Knee 106 (72.1) 1 0.75 Shoulder 54 (36.7) 11 2 0.49 Wrist 82 (55.8) 0.46 0.41 Hand 111 (75.5) 0.23 0.35 Elbow 44 (29.9) 10 0.25 1 Ankle 64 (43.5) 4 0.10 Total Joint Count 3 (2, 4) 5 0.23 15 0.26 Characteristics ≤3 for sCART VIMP and ≤0.3 RIS from RSF Sex, cancer stage, number of IRAEs, Sicca, thyroiditis, colitis, RF+, RF+ or CCP+, arthritis affecting hip, feet 1 sCART VIMP the higher the number the more influential the variable 2 sRF RIS A value of 1 indicates an influential variable; values that are negative or close to 0 are “noise” variables. *ICI combination vs monotherapy Figure 1. Survival Classification and Regression Tree for Arthritis Control. Kaplan-Meier curves are displayed at the bottom for each terminal node (or group) with time on the x axis measured in days. Acknowledgements: NIL. Disclosure of Interests Deanna Jannat-Khah Shareholder of: Dr. Jannat-Khah owns shares of Walgreens Boots Alliance, AstraZeneca, and Cytodyn., Grant/research support from: Dr. Jannat-Khah has a grant from the Hospital for Special Surgery., Laura Cappelli Grant/research support from: Dr. Cappelli has a grant from the NIH (NIAMS K23AR075872) and from Bristol Myers Squibb, Pankti Reid Consultant of: Dr. Reid was a consultant for Level Ex, Grant/research support from: Dr. Reid has grant support from the following: COVID-19 Funds to Retain Clinical Scientists by the Supporting Early Career University Researchers to Excel through Disruptions Steering Committee and The University of Chicago Institute of Translational Medicine Clinical and Translational Science Award K12/KL2 Grant 5KL2TR002387-05, Jeffrey Sparks Consultant of: Dr. Sparks was a consultant for AbbVie, Amgen, Boehringer Ingelheim, Bristol Myers Squibb, Gilead, Inova Diagnostics, Janssen, Optum, and Pfizer., Grant/research support from: Dr. Sparks has received grant support from the following entities: Bristol Myers Squibb, National Institute of Arthritis and Musculoskeletal and Skin Diseases, Rheumatology Research Foundation, R.Bruce and Joan M. Mickey Research Scholar Fund, and Llura Gund Award for Rheumatoid Arthritis Care and Research, Noha Abdel-Wahab Speakers bureau: Dr. Abdel-Wahab received an honorarium for a lecture by ChemoCentryx, Consultant of: Dr. Abdel-Wahab was a consultant for ChemoCentryx, Grant/research support from: Dr. Abdel-Wahab has grant funding from the following institutions: National Institute of Allergy and Infectious Disease (NIH-K01AI163412) and University of Texas MD Anderson, cassandra calabrese Speakers bureau: Dr. Calabrese received an honorarium for a lecture from Sanofi., Consultant of: Dr. Calabrese was a consultant for Lilly, and AstraZeneca, Carlos Aude: None declared, Nilasha Ghosh Grant/research support from: Dr. Ghosh has a grant from the Hospital for Special Surgery, Karmela Kim Chan: None declared, Anne Bass Grant/research support from: Dr. Bass has grants from the following institutions: Hospital for Special Surgery, Memorial Sloane Kettering Cancer Center and Rheumatology Research Foundation.