Objectives: To evaluate the long-term safety in patients with RA, treated with TNFi vs JAKi drugs in clinical practice.

Methods: Retrospective study of patients being followed up for RA, treated with any TNFi since 2000 or with JAKi since 2017. The following data were collected: epidemiological data, comorbidities, cardiovascular (CV) risk factors; From AR: RF, ACPA, evolution time, introduction order number, type, dose and time in treatment with TNFi and JAKi, cause of termination and its incidence rate/100 patients/year (IRPY): thrombosis, CV events, neoplasia, serious infection. Contrast of proportions z-test was used for the comparison of percentages of TNFi vs JAKi patients with a 5% significance level, and for comparisons of the incidence rate difference (IRD). Chi-square test to test whether or not the incidence rate ratio (IRR) was different from unity, with significance at 5%.

Results: 355 patients treated with some TNFi are included, as the 1st-2nd-3rd biological drug and 238 with some JAKi drug (Baricitinib/BARI: 138 [58%] patients, Upadacitinib/UPA: 70 [30%] and Tofacitinib/TOFA: 30 [12%]). The overall time of exposure to drugs was 1,328.06 years for TNFi and 539.48 for JAKi (BARI: 437.74 years, UPA: 70.52 and TOFA: 35.22). No significant differences were detected in the presence of CV risk factors. The TNFi vs JAKi group presented (Table 1): 1) MACE: 8 cases (2.2%), with a IRPY of 0.60 and no cases with JAKi. 2) Thrombosis, 4 (1.1%) cases, with IRPY 0.305 (0.08-0.78) in TNFi (pulmonary embolism [PE)]: 3 cases, Superficial venous thrombosis [SVT]: 1 case) vs 3 (1.6%) cases (PE: 1 case, DVT: 1 case and TVS: 1 case; all in the BARI group), p=0.88, with IRPY: 0.685 (0.14-2.00), with IRD: -0.379 (-1.06-3.05, p=0.27), IRR: 0.446 (0.0007-3.04, p: 0.316). However, the mean time on treatment until the appearance of thrombosis was higher in TNFi vs JAKi: 5.3 (SD: 4.1), range: 0.4-10 years vs 1.33 (0.57), range: 1-2 years (p<0.0001), as well as the mean time from RA diagnosis to thrombosis: 19.0 (13.5) vs 30.3 (5.5), p<0.0001). 3) Neoplasia, The TNFi group presented 13 (3.6%) cases (colon: 3 cases, breast: 2 cases, melanoma, sarcoma and 1 case of lymphoma, gallbladder, myelodyspasic syndrome, peritoneal carcinomatosis) vs 3 (1.2%) cases (p =0.07). 1 case in BARI of glioblastoma, melanoma and 1 case with TOFA of thyroid ca), with a IRPY of 0.993 (0.05-1.69) vs 0.685 (0.01-2.0), IRD: 0.308 (-0.72-1.34), p=0.55), IRR: 0.39-7.93, p=0.59).

4) Severe infection, The TNFi group is older (67.7 [SD: 12.3] vs 59.75 [20.5], p<0.0001), with 29 (8.1%) cases (pneumonia: 9, septic arthritis: 6, urinary tract infection: 5, osteomyelitis; 3, cellulitis: 2, others: 4) vs 4 (1.6%) in the JAKi group, p=0.0007 (in the BARI group: pneumonia 3 cases [2 of them due to COVID19] gastroenteritis: 1), with a IRPY of 1.303 (-0.18-2.79) vs 2.42 (0.85-9.48), p=0.07.

Conclusion: 1. Between both groups, no differences were detected in conventional risk factors, appearance of thrombosis or neoplasia. 2. The JAKi group did not present cases of MACE. 3. However, the TNFi did present a significantly higher number of cases of serious infection, in this case the older patients and pneumonia being the most frequent infection. 4. Highlight the overall low number of events that occurred during follow-up.

REFERENCES: NIL.

Acknowledgements: The study was supported by a research grant from the Association for Research in Rheumatology of the Marina Baixa (AIRE-MB).

Disclosure of Interests: None declared.