Background/Purpose: Patients with established Rheumatoid Arthritis (RA) are at increased risk of malignant lymphomas. We have previously demonstrated a strong association between inflammatory intensity and lymphoma risk in a historical RA cohort, but also that the lymphoma risk was increased in patients diagnosed 1997 and followed through 2006. Given the dramatic changes in treatment strategies and goals over the last decade, it may be that lymphoma risks in more recently diagnosed patients have changed. We therefore aimed at assessing lymphoma risks in a more contemporary patient population with respect to year of RA onset and duration of RA disease.

Methods: 10,367 patients with incident RA (ACR criteria, >18 years age, symptom duration ≤ 13 months) diagnosed 1997 through 2010 were identified within the national Swedish Rheumatology Register (SRQ) including information on disease characteristics, disease activity, and therapy. Each patient was matched to 5 general population comparators, by gender, age and residence (n=49,825).To identify the lymphomas, all individuals were linked to the nationwide Swedish Cancer Register. Relative risks (RR) of lymphoma were assessed using Cox models. The RA lymphomas were reviewed and reclassified (WHO classification).

Results: Overall, the risk of lymphoma in RA was almost doubled (RR= 1.7 (95% CI 1.2-2.4)), based on 45 lymphomas in RA versus 126 in the general population comparator cohort (person time 58,825 versus 283,903). When stratified by RA disease duration, a statistically significantly increased risk (p value=0.04) was noted > 5 years after RA diagnosis. When relative risks were cross-tabulated according to year of RA onset and time since RA diagnosis, we noted that this risk increase was significant only among patients in calendar period 1997-2001. Short-term risks for lymphoma were similar across calendar periods. (see Table).  

Conclusion: Overall, the risk increase for malignant lymphomas in RA patients diagnosed 1997-2010 was of a similar magnitude as that reported from historical RA cohorts. Whereas our results suggested a declining point estimates for lymphoma risk in successive calendar periods of RA diagnosis 1997-2010, this observation was confounded by differences in length of follow-up, i.e., that risks increased with RA duration. It will be an important task to monitor future risks over longer follow-up time.