Background Infections are a common adverse event (AE) in chronic inflammatory arthritis patients (CIA), such as RA, PsA and AS, treated with biological therapies (BT). Their frequency varies during the treatment and the risk for a recurrence has not been well established in clinical practice

Objectives To analyze the incidence rate of different types of infection during biological therapies in CIA patients, their distribution over time and their recurrence

Methods 290 CIA patients treated with 615 BT, from 1999 to 2012 were included. All AEs developed during BT were recorded. We analyzed infectious AE with at least 10 recurrences. Global incidence rates (IR) of AE were computed as the number of events per 1,000 patient-years. IRs were also computed in multiple time intervals after starting BT to evaluate its distribution over time and stratified by recurrence (i.e. first event on a patient or recurrent event). The analysis of the recurrence of AEs has been based on a right-censored Cox proportional hazards regression model. Failure times were defined as the time intervals between the BT start and the first AE. When recurrent events were found, failure times were defined as the time interval between the recurrent event and the previous event occurrence. The stratification variable used in the Cox model was defined as the number of previous AEs found in order to evaluate how much individuals presenting a first AE were more prone to have recurrent events

Results Of the 615 analyzed BT, 680 infectious AE were recorded. Of these, 485 were respiratory infections (RI) with an incidence rate IR 305.32; 73 urinary infections (UI) (IR 45.96); 67 gastrointestinal infections (GI) (IR 42.18) and 55 oral cavity infections (OC) (IR 34.62). In the normalized distribution of RI/1,000 patients-year over time, the probability of a first RI was higher between the first 48 wks of BT. When all infections (first and recurrences) were analyzed, a homogenous distribution was found. In the Cox model, a previous RI was related with an increased risk for a recurrent event (P=7.99e-03, HR=1.08 [95% CI 1.02-1.14]). The first UI has had a less homogenous distribution in time, but similarly to RI, with a higher risk to presentation between the initiation of the BT and the 48 wks of follow up. In this particular AE, the probability to a recurrence was very significant (P=1.67e-10; HR=1.96 [95% CI 1.59-2.41]). GI events distribution showed peak in the first 12 wks, but we didn't find an increased risk of a recurrence (P=6.81e-02; HR=1.55 [95% CI 0.97-2.49]). OC first events showed two peaks, at 12-24 wks and 96-192 wks. Risk for recurrence was significant in this AE (P=5.21e-06; HR=2.12 [95% CI 1.53-2.92])

Conclusions In our series of CIA patients treated with biological therapies, we found that the risk for the first infection was higher at the beginning of treatment. The distribution of the AE over time showed a homogeneous tendency in all type of infections, except for gastrointestinal events that shows a higher incidence during the first 12 wks of BT. The ratio between recurrent and first events showed a growing tendency for respiratory and urinary infections, mainly from 48 wks of treatment. Regarding to recurrences infections, we found a higher recurrence rate in urinary and oral infections and minimal risk for gastrointestinal events

Disclosure of Interest None declared

DOI 10.1136/annrheumdis-2014-eular.5736