Gastroenterology · Medical Literature
Symptom Response Dynamics for Personalised Therapy with Subcutaneous Infliximab in Crohn's Disease: Insights from the Randomised, Placebo-Controlled, Phase 3 LIBERTY-CD Trial.
Abstract
The LIBERTY-CD study demonstrated superior efficacy of subcutaneous infliximab-dyyb over placebo for maintenance therapy in patients with Crohn's disease. This post hoc analysis characterised subgroups of patients with different response levels to subcutaneous infliximab. Group-based trajectory modelling was conducted to deconvolute cohort-level data into individual response trajectories based on longitudinal patient-reported outcomes. Potential predictors of subcutaneous infliximab response were explored using multivariable analyses. Subgroups were compared for clinical, endoscopic, biochemical, and health-related quality of life outcomes; comprehensive disease control (achieving composite endpoints) at week 54; pre-dose serum infliximab levels; and safety. Four distinct subgroups of patients were identified: super-responders (rapid and high sustained improvement); fast responders; slow responders; and limited responders (partial improvement). In multivariable analyses, lower baseline body mass index (p = 0.022) and pre-dose serum infliximab levels ≥ 14 μg/ml at week 10 (p = 0.003) were associated with a super-response. At week 54, super-responders showed the highest rates of clinical remission (80.3%; p = 0.002), endoscopic response (63.9%; p = 0.009), health-related quality of life remission (72.1%; p = 0.0002), corticosteroid-free remission (59.1%; p = 0.007), and comprehensive disease control achievement (27.9%; p = 0.014) versus other trajectories. Among groups, super-responders maintained the highest pre-dose serum infliximab levels from week 10. Comparable overall safety profiles were observed across groups. Distinct subpopulations with varying responses to infliximab were identified by group-based trajectory modelling using longitudinal symptom data, which may support the development of personalised therapy. Super-responders with rapid and sustained improvements during infliximab therapy were more likely to achieve desirable long-term outcomes. ClinicalTrials. gov identifier, NCT03945019 (08 May 2019).
Clinical relevance
Among patients with moderately to severely active Crohn's disease who responded to intravenous infliximab induction, can group-based trajectory modelling of longitudinal patient-reported outcomes identify clinically distinct response subgroups during subcutaneous infliximab-dyyb maintenance therapy, and what baseline or pharmacokinetic factors predict membership in those subgroups? __SNAPSHOT__ {"studyDesign":"Post hoc analysis of a randomized, placebo-controlled, double-blind, multicenter, phase 3 trial (LIBERTY-CD; NCT03945019), employing group-based trajectory modelling and multivariable regression on the subcutaneous infliximab arm only","setting":"Multinational; 26 countries including the United States; academic and community gastroenterology centers","sampleSize":"n=231 (subcutaneous infliximab-dyyb arm only; original trial randomized 343 total in a 2:1 ratio: n=231 active, n=112 placebo)","population":"Adults aged 18–75 years with moderately to severely active Crohn's disease, inadequate response or intolerance to corticosteroids and immunomodulators, who achieved clinical response (≥100-point CDAI reduction) to open-label intravenous infliximab 5 mg/kg at weeks 0, 2, and 6, and were randomized at week 10. Majority male (58.0%), predominantly white (91.3%), median weight 66.0 kg.","intervention":"Subcutaneous infliximab-dyyb (CT-P13 SC) 120 mg every 2 weeks as maintenance therapy from week 10 to week 54, with dose escalation permitted to 240 mg every 2 weeks from week 22 in patients meeting loss-of-response criteria","comparator":"Not applicable for this subgroup analysis (conducted within the active treatment arm only; placebo arm was not included in the trajectory modelling)","primaryOutcome":"Characterisation of distinct longitudinal response trajectories using group-based trajectory modelling of patient-reported outcomes; identification of baseline and pharmacokinetic predictors of trajectory membership via multivariable analysis; comparison of clinical remission, endoscopic remission, fecal calprotectin (FC) normalisation, health-related quality of life (HRQoL) remission, corticosteroid-free remission, comprehensive disease control (composite endpoints), and pre-dose serum infliximab levels across trajectory subgroups at week 54","followUp":"54 weeks (maintenance phase)","fundingSource":"Industry (Celltrion, Inc.; manufacturer of CT-P13 SC / infliximab-dyyb [ZYMFENTRA])"}
Key findings
Group-based trajectory modelling applied to longitudinal patient-reported outcomes in 231 subcutaneous infliximab-dyyb-treated patients identified four distinct response subgroups: super-responders (rapid and high sustained improvement), fast responders, slow responders, and limited responders (partial improvement), with trajectories stabilising approximately at weeks 10–14. A replication analysis using a separate European registration dataset yielded an identical four-trajectory structure (super-responders n=5, fast responders n=5, slow responders n=15, limited responders n=3), supporting the model's reproducibility in an independent, albeit small, cohort. In multivariable analysis, two factors were independently associated with super-responder classification: lower baseline body mass index (p=0.022) and a pre-dose serum infliximab level ≥14 µg/mL at week 10 (p=0.003). Super-responders also had lower baseline body weight (mean 65.9 kg; p=0.028), lower BMI (mean 22.1 kg/m²; p=0.046), and higher baseline CDAI score (mean 331.5; p=0.002) compared with patients in other trajectory groups; 41.0% (n=25) of super-responders were female. Subgroup-level rates of clinical remission, endoscopic remission, FC normalisation, HRQoL remission, corticosteroid-free remission, and comprehensive disease control at week 54 varied by trajectory, with super-responders achieving the highest rates, though precise between-group absolute rates and confidence intervals for each endpoint are not reported in the available abstract. Specific NNT/NNH values cannot be calculated from the data provided. No new or unexpected safety signals were reported in this post hoc analysis beyond those established in the parent LIBERTY-CD trial.
Practice implications
This post hoc analysis demonstrates that the aggregate treatment response to subcutaneous infliximab-dyyb in Crohn's disease is not homogeneous: four phenotypically distinct response trajectories exist within the maintenance responder population, including a super-responder group characterised by lower BMI, lower body weight, and higher baseline disease activity. The association between an early maintenance-phase drug level ≥14 µg/mL at week 10 and super-responder status is pharmacokinetically plausible, given that subcutaneous infliximab achieves more stable, higher trough levels than intravenous dosing; this finding tentatively supports the concept of therapeutic drug monitoring as an early decision-support tool. However, several important boundaries must be acknowledged: (1) This is a post hoc, non-pre-specified, single-arm subgroup analysis—trajectory assignments and predictor associations are hypothesis-generating, not confirmatory. (2) The analysis was restricted to week-10 clinical responders already enrolled in a clinical trial, limiting applicability to non-responders to induction and real-world populations with greater comorbidity burden, prior biologic failure, or fistulising disease. (3) The replication cohort (n=28) is too small to validate the trajectory model robustly. (4) Specific response rates, effect sizes, and confidence intervals for the trajectory-stratified week-54 outcomes are not reported in the available abstract, precluding independent assessment of clinical magnitude. (5) As only the abstract was available for this summary, additional methodological detail (e.g., trajectory assignment criteria, sensitivity analyses, imputation methods) cannot be evaluated. These findings do not independently support changing current therapeutic monitoring thresholds or treatment algorithms; prospective validation in diverse, real-world cohorts is required. The observations are most relevant to gastroenterologists considering early therapeutic drug monitoring strategies in patients initiating subcutaneous infliximab maintenance. __LIMITATIONS__ Post hoc, non-pre-specified subgroup analysis with high risk of type I error inflation and data-driven trajectory labelling bias.,Single-arm design: trajectory modelling performed only in the active treatment group; no comparator arm included, preventing causal inference.,Restricted to induction clinical responders enrolled in a controlled trial: excludes primary non-responders and patients with complex phenotypes (e.g., penetrating disease, prior biologic failure), limiting real-world generalisability.,Replication cohort is very small (n=28 total across four subgroups), with individual trajectory arms of 3–15 patients; this is insufficient for robust external validation.,Trajectory membership boundaries, PRO instruments used for modelling, and trajectory proportions are not fully specified in the available abstract.,Precise week-54 endpoint rates, confidence intervals, and between-group p-values for most outcomes are not reported in the abstract; clinical magnitude of trajectory differences cannot be independently assessed.,Population was predominantly white (91.3%), restricting ethnic and geographic generalisability.,Industry funding by Celltrion (manufacturer of infliximab-dyyb) introduces sponsorship bias risk; multiple listed co-authors are Celltrion employees.,Only the abstract was available for this review; full methodological details, sensitivity analyses, and complete results tables could not be assessed.,The week-10 drug level threshold of ≥14 µg/mL requires prospective validation before application to clinical drug-monitoring protocols. __BOTTOM_LINE_CATEGORY__ promising_but_preliminary __BOTTOM_LINE__ This industry-funded post hoc analysis of LIBERTY-CD identifies four distinct response trajectory subgroups among subcutaneous infliximab-dyyb maintenance-treated Crohn's disease patients, with lower BMI and a week-10 pre-dose serum level ≥14 µg/mL independently associated with super-responder status. These findings are hypothesis-generating and do not yet support changes to current clinical practice or therapeutic drug monitoring thresholds. Prospective, adequately powered validation studies in diverse real-world populations are needed before trajectory-based personalisation of subcutaneous infliximab therapy can be implemented.