Infliximab Is Associated With an Increased Risk of Serious Infection in Patients With Psoriasis in the U.K. and Republic of Ireland

Results From the British Association of Dermatologists Biologic Interventions Register (BADBIR)

Z.Z.N. Yiu; ; D.M. Ashcroft; I. Evans; K. McElhone; M. Lunt; C.H. Smith; S. Walton; R. Murphy; N.J. Reynolds; A.D. Ormerod; C.E.M. Griffiths; R.B. Warren; BADBIR Study Group


The British Journal of Dermatology. 2019;180(2):329-337. 

In This Article


In total, 3843 participants were included in the analysis, with 3421 participants included in the nonbiologic systemic cohort and 422 participants included in the all-lines infliximab cohort up to October 2016. Overall, 105 biologic-naïve participants were started on first-line infliximab therapy. The baseline demographic, anthropometric and disease characteristics of the participants are listed in Table 1. The total and median follow-up time for all lines of infliximab was 941·1 person-years and 1·49 person-years [interquartile range (IQR) 2·50 person-years], respectively. For biologic-naïve patients on infliximab the total follow-up time was 238·87 and the median follow-up time was 1·84 person-years (IQR 2·70 person-years) and for the nonbiologic cohort the total and median follow-up time was 6419·24 person-years and 1·51 person-years (IQR 1·84 person-years), respectively.

Crude Incidence Rates for Serious Infections Overall

The incidence rate for serious infections in the nonbiologic cohort was 14·18 per 1000 person-years (95% CI 11·54–17·41), with the incidence rate for the methotrexate only cohort at 11·98 per 1000 person-years (95% CI 8·82–16·27). The crude incidence rate in the entire infliximab cohort was 47·82 per 1000 person-years (95% CI 35·70–64·04), and 58·61 per 1000 person-years (95% CI 34·71–98·96) for the biologic-naïve infliximab cohort.

The most common serious infections coded using MedDRA high level terms that were experienced by participants on either nonbiologic systemic therapy or infliximab were lower respiratory tract infections, followed by skin and soft tissue infections and urinary tract infections (Table S2; see Supporting Information). The crude incidence rates for lower respiratory tract infections and skin and soft tissue infections were higher for infliximab (Table 2). The median hospital inpatient stay was 3 days (IQR 6·0) for nonbiologic therapies and 2 days for infliximab (IQR 9·0).

Propensity Score Weighted Models for the Risk of Serious Infections

The inverse probability treatment weighted (IPTW) logistic regression model for infliximab vs. nonbiologic therapies achieved good balance, removing expected bias for most of the variables (Figure 1 and Table S3; see Supporting Information), which suggested a reduction of confounding from these variables.

Figure 1.

Forest plot showing the reduction in expected percentage bias for the individual covariates after inverse probability treatment weighted propensity score weighting.
PASI, Psoriasis Area and Severity Index; BMI, body mass index; COPD, chronic obstructive pulmonary disease.

Infliximab showed a statistically significant increase in the risk of serious infection compared with nonbiologic systemic therapies overall [adjusted hazard ratio (HR) 1·95, 95% CI 1·01–3·75; Table 3]. The proportionality assumption was not met for this model, and therefore a split in a priori defined follow-up time was performed. The risk of serious infection was significantly higher in the first 6 months (adjusted HR 3·49, 95% CI 1·14–10·70) and between 6 months and 1 year (adjusted HR 2·99, 95% CI 1·10–8·14) but not significantly higher between 1 and 2 years (adjusted HR 2·03, 95% CI 0·61–6·79) compared with the nonbiologic cohort (Table 3). The adjusted estimate for biologic-naïve participants on infliximab was also higher than nonbiologic therapies but this was not statistically significant (adjusted HR 1·37, 95% CI 0·50–3·74).

Sensitivity Analysis

Infliximab (all lines) had a statistically significant increase in the risk of serious infection compared with methotrexate (adjusted HR 2·96, 95% CI 1·58–5·57; Table 3). The proportionality assumption was met for this alternative model.