P609 A treat-to-target approach via a virtual clinic amongst inflammatory bowel disease patients with secondary loss of response to anti-TNF therapy improves clinical outcomes
Little R., Taylor K., Friedman A., Headon B., Gibson P., Sparrow M., Ward M.
Alfred Health and Monash University, Gastroenterology, Melbourne, Australia
Secondary loss of response to infliximab (IFX) or adalimumab (ADA) is common in inflammatory bowel disease (IBD). Therapeutic drug monitoring (TDM) identifies patients with sub-therapeutic drug levels who are more likely to respond to dose intensification. Delivering dose-intensified therapy is resource-intensive and may benefit from a non-conventional decision making system such as a virtual clinic (VC). We sought to determine whether enrolment in a VC following a “treat-to-target” paradigm was effective in controlling disease activity in this complex cohort of patients.
Observational study of 37 IBD patients with secondary loss of response referred to our VC between September 2013-October 2016. Dose-intensification involved shortening the interval between ADA and IFX administration to weekly or six-weekly, respectively. Patients were reviewed in our VC every 6 months with scheduled C-reactive protein (CRP), faecal calprotectin (FC), intestinal ultrasound and IFX/ADA TDM using a drug-sensitive ELISA. Response was defined as maintaining improvements in biomarkers and physician global assessment for ≥12 months after initiation of intensified therapy, including those subsequently de-escalated to standard dosing. Patients failing intensified therapy were defined as non-responders. Receiver-operator characteristic analysis was performed to identify a threshold delta increase in drug level associated with response.
86% had Crohn's disease; 62% were treated with IFX. 22 (59%) responded, 55% of whom received IFX. 11 (30%) responders were de-escalated to standard dosing (median 12 months). 15 (41%) were non-responders (median 9 months). Considering the entire cohort, FC and CRP decreased after 12 months compared to baseline (450 vs. 80μg/g; p=0.019 and 8.5 vs. 3.5mg/L; p=0.004, respectively). Subgroup analyses of biomarker and TDM are shown in Table 1. Increasing IFX drug level >3μg/mL from baseline best predicted response (area under curve 0.86, sensitivity 80, specificity 78%). No threshold was found for ADA.
Responders Non-responders p-value n Last FC (μg/g) 44 566 $<0.001$ 34 Last CRP (mg/L) 3 7 0.024 37 Baseline to last IFX level (μg/mL) 1–6.5 0.8–1.1 0.002; 0.109 10, 9 Baseline to last ADA level (μg/mL) 2.8–12.7 1.7–11.9 0.002; 0.063 10, 5 Delta IFX levels (μg/mL) 6.3 1 0.006 19 Delta ADA levels (μg/mL) 9.4 7.9 0.514 15
A novel virtual clinic model to deliver intensified anti-TNF therapy enabled recapture of response in the majority of patients, with almost one-third de-escalated to standard dosing. IFX drug levels increased in responders compared with non-responders and a threshold increase of >3μg/mL from baseline was associated with response. Non-responders treated with ADA showed similar increases in drug level to responders, suggesting that outcome was independent of ADA levels.