P497 The value of endoscopic healing index monitoring for guiding infliximab dosing in patients with Crohn’s disease

Background The endoscopic healing index (EHI) is a novel multi-protein serum biomarker test developed and validated to assess endoscopic disease activity in patients with Crohn’s disease (CD). Evidence for the use of EHI to guide decision-making during infliximab (IFX) treatment remains scarce. Therefore, we aimed to characterise the relationships between IFX dose, serum IFX concentrations, EHI, and endoscopic remission (ER). Methods Data were obtained from 118 biologic naïve adult patients with CD enrolled in the phase 4 TAILORIX trial (EudraCT 2011 003038 14). All patients had confirmed active CD at baseline based on clinical, biological, and endoscopic criteria. IFX and EHI (scores ranging from 0–100) were measured using a homogenous mobility shift assay (HMSA) and immunoassay, respectively (Prometheus Laboratories). First, the previously published population pharmacokinetic (popPK) model of the TAILORIX study population was revisited to describe the HMSA data. The effect of EHI, faecal calprotectin (FC), C-reactive protein (CRP), and serum albumin (ALB) on IFX clearance was evaluated. Next, a minimal continuous-time Markov model was developed to describe the time course of EHI within patients. EHI was considered as a three-stage ordinal variable (scores 0–19, 20–49, and 50–100) with the lowest score stage (0–19) indicative of ER. The course-modifying effect of IFX on EHI was assessed. Finally, a generalised linear model was used to describe the relationship between EHI and the probability of attaining ER (Simple Endoscopic Score for CD [SES-CD] ≤2). The predictive ability of EHI for ER was compared with that of FC, CRP, ALB, and IFX using a receiver operating characteristic (ROC) curve analysis. Results The revisited two-compartment popPK model described the IFX data with adequate descriptive and predictive accuracies. EHI, FC, CRP, and ALB at week (w)0 were not found to explain interpatient variability in IFX clearance. In contrast, higher IFX at w14 was associated with a higher probability of achieving EHI <20 at w14 (Figure 1). The probability of attaining EHI <20 at w14 was predicted to increase more than four-fold when IFX at w14 was targeted at 10 mg/L instead of 5 mg/L (Table 1). EHI and FC equally well predicted the probability of attaining ER at the same time point (Figure 2, Table 2). Conclusion EHI, FC, ALB, and CRP at w0 should not be considered for a priori IFX dose optimisation. Nevertheless, a posteriori IFX dose optimisation (based on IFX concentrations measurements) towards a predefined IFX concentration at w14 may lead to lower post-induction EHI scores and thus improved ER rates. An IFX target of 10 mg/L at w14 is associated with four-fold higher normalisation of EHI as compared to the commonly used target of 5 mg/L.