OP0257 RISK OF HAEMATOLOGICAL MALIGNANCY IN PATIENTS WITH PSORIATIC ARTHRITIS, OVERALL AND IN RELATION TO TNF INHIBITORS - A NORDIC COHORT STUDY

Abstract

Background: Several autoimmune infammatory diseases, including rheumatoid arthritis (RA), are associated with increased risk of malignant lymphomas. There is also a longstanding concern of lymphoma development with tumour necrosis factor inhibitor (TNFi) treatment, but most studies in RA to date do not indicate an additionally increased risk. Corresponding studies in psoriatic arthritis (PsA), both with respect to the underlying risks, and risks in relation to treatment with TNFi, are limited. Data on myeloid malignancies in PsA are scarce. Objective(s): To estimate the risk of haematological malignancy overall and by lymphoid and myeloid types in TNFi treated versus (vs.) biologics-naive patients with PsA across the five Nordic countries. Additionally, we investigated the underlying risk of haematological malignancies in PsA as compared to the general population. Method(s): We identifed patients with PsA starting a frst ever TNFi from the clinical rheumatology registers (CRR) in Sweden (SE), Denmark (DK), Norway (NO), Finland (FI), and Iceland (ICE) from 2006 through 2019 (n=10 621). We identifed biologics-naive patients with PsA from a) the CRR (n=18 705, all countries) and b) the national patient registers (NPR, n=27 286, SE and DK only). To estimate the underlying risk of haematological malignancy in PsA, we randomly sampled general population comparators in SE and DK matched on year of birth, sex, and calendar year at start of follow-up, to the patients with PsA. Through linkage to the mandatory national cancer registers in all five countries, we collected information on haematological malignancy overall, and categorised into lymphoid or myeloid types. By applying a modifed Poisson regression, we estimated pooled incidence rate ratio (IRR) with 95% confdence intervals (CI) for TNFi treated vs. biologics-naive PsA and for PsA vs. the general population, adjusted for age (18-55, 56-65, 66-70, >70 years), sex, calendar period (2006-2010, 2011-2019) and country, and using robust standard errors. Result(s): We observed 40 events of haematological malignancies (during 59 827 person-years) among TNFi treated PsA, resulting in a crude incidence rate (IR) of 67 per 100 000 person-years. The corresponding IR was 91 (63 events) for biologics-naive PsA from the CRR, and 118 (172 events) for biologics-naive PsA from NPR. This resulted in a pooled IRR of 0.97 (0.69 to 1.37) for TNFi-treated vs. biologics-naive PsA patients from the CRR, and 0.84 (0.64 to 1.10) vs. biologics-naive PsA patients from the NPR. The pooled IRR of haematological malignancies in PsA overall vs. the general population was 1.35 (1.17 to 1.55). Throughout, the estimates were largely similar for lymphoid and myeloid malignancies (Figure 1). The crude IR of haematological malignancies were substantially akin across different TNFi agents. Conclusion(s): In this large five-country cohort study, we did not observe any increased risk of haematological malignancies overall, nor for lymphoid and mye-loid types, in patients with PsA treated with TNFi. By contrast, there were signals of a moderately increased underlying risk of haematological malignancies, both of lymphoid and myeloid types, in patients with PsA overall as compared to the general population. The fndings are of importance from a patient information perspective.