POS0604 JAKi’S SURVIVAL RATE AND PREDICTORS OF DISCONTINUATION IN A COHORT OF PATIENTS WITH RHEUMATOID ARTHRITIS

Background: After the publication of the ORAL Surveillance trial1, the European Medical Agency (EMA) recently recommended prescribing JAK inhibitors (JAKi) in patients with rheumatoid arthritis (RA) only after an adequate risk/benefit assessment. Indeed, according to EMA recommendations, patients ≥65 year-old, former or current smokers, at high risk of developing major cardiovascular events (MACEs), thromboembolic events or cancers should receive JAKi only when other treatments, such as TNF-inhibitors, are not appropriate. Unlike randomised controlled trials, real-world data seem to be reassuring promising and, in fact, the prescription rate of JAKi has dramatically risen in the last years. Objectives: The primary objective was to assess whether being at high risk according to EMA recommendations can impact JAKi’s discontinuation rate. The secondary aim was to find out other possible predictors of JAKi’s discontinuation. Methods: This was a retrospective Italian study carried out in 22 Italian centres since 2017. All patients with RA on treatment with JAKi were included in the dataset. This cohort was subdivided into two groups: “high risk patients” and “low risk” according to EMA recommendations. The first group included patients ≥65 year-old, past or current smokers and having at least one cardiovascular or neoplastic comorbidity. The following variables were collected at baseline: sex, age, disease duration (years), smoking habit, BMI, comorbidities (diabetes, hypertension, dyslipidaemia, cancer, previous MACEs), positive RF/ACPA, associated conventional DMARDs, concomitant use of prednisone and dosage (mg/day), previous use of JAKi, discontinuation reasons, time to discontinuation (days), b/tsDMARDS naïve patients, DAS28-ESR at baseline. Results: A total of 693 patients were enrolled. 48 patients were excluded due to missing data. Overall features of our cohort are summarised in Table 1 (N=645). 372 (57.7%) patients received baricitinib, 135 (20.9%) tofacitinib, 86 (13.3%) upadacitinib and 52 (8.11%) filgotinib. 21 (3.2%). 141 (21.9%) patients discontinued JAKi after a median time of 366 days (IQR 155-914). “High-risk patients according to EMA” were the majority of our cohort (n=384, 59.5%) vs 261 “low-risk patients” (40.5%). Reasons for discontinuation were primary inefficacy (n=48, 7.4%), secondary inefficacy (n=25, 3.9%), infections (n=8, 1.2%), pulmonary embolism/deep venous thrombosis (n=6, 0.9%), cancer (n=5, 0.8%), deaths (n=2, 0.3%), and other causes (n=21, 3.3%) including remission status in 1 patient. Notably, VZV infection determined JAKi’s withdrawal in 3 patients and pulmonary embolism/deep venous thrombosis in 6 patients (all treated with baricitinib). At multivariate stepwise Cox analysis, predictors of discontinuation were: prednisone dosage [Hazard Ratio -(HR)- 1.1, 95% confidence interval- (CI)- 1.05-1.17], use of selective JAKis(HR 4.1, 95% CI 1.80-9.10), and absence of RF/ACPA (HR 0.46, 95% CI 0.26-0.83). Finally, being classified as “high risk” according to EMA was not statistically associated with JAKIs’ withdrawal (HR 1.96, 95% CI 0.96-4-01). Conclusion: Our study shows that only a minority of patients discontinued JAKi (21.9%). Notably, among discontinuation‘s causes, no MACE’s were found. Being classified at “high-risk” according to EMA was not associated with JAKi’s discontinuation. Conversely, higher prednisone dosages, treatment with selective JAKi and absence of RF/ACPA were predictors of JAKi’s withdrawal. REFERENCES: [1] Ytterberg SR, et al. Cardiovascular and cancer risk with tofacitinib in rheumatoid arthritis. New Engl J Med 2022;386(4):316-326.