Shortening infusion times for infliximab administration

Sir, Infliximab is generally administered over 2 h with further 2 h of monitoring, making its management expensive, with major utilization of hospital resources. The risk of infusion reactions remains a concern. This study's aims (in accordance with our local ethics committee guidelines) were to determine: (i) the frequency of reactions and their relation to concomitant therapy; (ii) when reactions occurred; (iii) the consequences of infusion events; (iv) the incidence of adverse events in the monitoring period; and (v) the possibility of shortening infusion times incrementally. This study was performed in accordance with the guidelines of our local ethics committee, which has approved the recording and reporting of safety and clinical data of patients on biological therapies; specific patient consent was not required. As part of this units’ protocol, the first four infliximab infusions [3 mg/kg in rheumatoid arthritis (RA) and 5 mg/kg in seronegative spondyloarthropathy] are administered over 2 h and patients are monitored for a further 2 h. If tolerated, infusions 5–10 are given over 1 h with 1 h post-infusion monitoring; the 11th and subsequent infusions are administered over half an hour with half an hour of monitoring. Infusion reactions are described as acute (during the infusion) or delayed (after completion of infusion and post-infusion observation). Severe reactions constitute a drop in systolic blood pressure (SBP) to ≤80 mmHg with or without bronchospasm and with or without rash, warranting cessation of the infusion. Cessation is usually managed by stopping the infusion and administering intravenous (i.v.) hydrocortisone 100 mg, i.v. chlorpheniramine 10 mg and, if indicated, nebulized salbutamol 2.5–5 mg. Mild reactions are defined as SBP falling below baseline but remaining above 80 mmHg with or without nausea and with or without other non-specific symptoms (e.g. headache) that clinically justify only a temporary stop of the infusion. The infusion is stopped for 30 min and restarted with i.v. hydrocortisone and chlorpheniramine administered as above. Subsequent infusions in this mild reaction group are premedicated with i.v. hydrocortisone 100 mg stat and chlorpheniramine 10 mg stat indefinitely. From December 1999 to December 2003, a total of 2166 infusions have been administered (2–28 infusions/person) to a total of 234 patients (age 22–77, 164 female, 70 male). Two hundred and eight were being treated for RA [161 on concomitant methotrexate (MTX) 7.5–25 mg weekly, and 47 were on leflunomide 10 or 20 mg daily]. Twenty patients with ankylosing spondylitis (AS) and six with psoriatic arthritis (PsA) were on concomitant MTX. Fifteen of the RA patients had been on the ATTRACT study [1]; all flared on completing the study (mean time to flare 13.5–20 weeks for the four dosage regimens of the study) [2] and were re-established on infliximab, including the initial loading dose. Findings are summarized in Table 1. Summary of results a208 RA, 20 AS, 6 PsA. Mean infliximab dose 3 mg/kg for RA, 5 mg/kg for AS and PsA. bEight on leflunomide; con leflunomide; d3 on leflunomide; NA, Not Applicable. Summary of results a208 RA, 20 AS, 6 PsA. Mean infliximab dose 3 mg/kg for RA, 5 mg/kg for AS and PsA. bEight on leflunomide; con leflunomide; d3 on leflunomide; NA, Not Applicable. Of the 234 patients who received infliximab at the usual rate of 2 h, 6% (n = 13) ceased treatment due to a severe infusion reaction (all between infusions 2 and 4). Of these 13, 62% (n = 8) patients were in the RA/leflunomide group (19% of this group) and 38% (n = 5) patients were in the RA/MTX group (3% of this group; included one anaphylactic and one delayed reaction). Application of the χ2 test with (Yates’) continuity correction determined this to be a significant difference between patients on leflunomide and those on MTX (P = 0.002). No patients in the AS or PsA group developed a severe reaction. The total number of infusions administered to the 234 patients over 2 h was 845; reactions therefore occurred in 1.5% of all 2 h infusions. Six per cent (n = 11) of patients who received infliximab at the usual rate of over 2 h suffered a mild, acute reaction. Of these 11, 46% (n = 5) were in the RA/MTX group (3% of this group) and 27% (n = 3) in the RA/leflunomide group (6% of this group), 36% (n = 4) were in the AS group (20% of this group) and 18% (n = 2) in the PsA group (33% of this group) (no significant difference). This constituted 1.3% of all 2 h infusions. All patients continued with premedication for subsequent infusions and were able to follow the protocol of reduced infusion duration. Of the 221 patients who had completed the first four infusions, 171 went on to receive further infusions at a rate of 1 h. One patient (0.6%) (RA/leflunomide group) suffered a serious acute infusion reaction (total of 796 infusions administered over 1 h; reaction in 0.1% of all 1 h infusions). Two patients (1.2%) (RA/leflunomide group) suffered a mild, acute reaction (0.2% of all 1 h infusions) and continued with premedication (no significant difference). Eighty-nine patients received infusions 5–10 over 1 h and progressed to half-hour infusions. No infusion reactions were observed during these infusions (total of 525 infusions over half an hour). Of the 15 patients reinfused with infliximab following completion of the ATTRACT study, none sustained an infusion reaction. Outcome with regard to efficacy was comparable, as has been described previously [2]. Of all the severe reactions observed, none warranted intramuscular adrenaline or hospitalization. No patients sustaining severe reactions received further infliximab. No differences in disease characteristics or serology were associated with the development of an infusion reaction. It may be relevant to note, however, that a separate study by Bingham et al. [3] assessing the efficacy and tolerability of infliximab and leflunomide revealed a higher rate of induction of ANA and double-stranded DNA antibodies (although this was observed in the response group as opposed to the toxicity group, possibly due to a longer time on therapy for this former group). Our experience of infliximab administration is comparable to that of other large centres and studies [1, 4, 5] and confirms the general tolerability of infusions. Furthermore, it demonstrates the possibility of shortening infusion times incrementally. In this retrospective survey, 6% of patients (0.7% of all 2166 infusions) ceased treatment due to a severe reaction. Notably, 93% of the severe reactions occurred during infusions 2–4 and, strikingly, 64% of patients were on concomitant leflunomide therapy. Overall, for the 2166 infusions, only 1.2% resulted in either a severe or mild infusion reaction; most (89%) were during the first four infusions (none during half-hour infusions). Our approach of shortening infusion times is similar to the more limited experience outlined in a brief letter by van Vollenhoven et al. (440 infusions to 113 patients) [5]. We conclude that (i) administration of infliximab with concomitant leflunomide in patients with RA confers a significantly increased incidence of infusion-related reactions, (ii) greatest care should be taken during the first four infusions, (iii) infusion reactions are rarely associated with hospitalization and further sequelae, (iv) the monitoring period may not be necessary, (v) our incremental shortened infusion time does not seem to compromise safety if initial infusions are tolerated, and (vi) stopping and restarting infliximab (albeit in a small subgroup of patients) does not seem to confer greater toxicity risks, suggesting careful reintroduction of infliximab can be justified. The authors have declared no conflicts of interest. Academic Unit of Musculoskeletal Disease, Leeds and 1Remicare Nurse Liaison Service, Ashfield Health Care, Ashfield House, Resolution Road, Ashby de la Zouche, Leicesteshire, LE 6S IHW, UK