Introduction

The Phase 3 QUASAR induction study (NCT04033445) was a randomized, double-blind, placebo-controlled study that evaluated guselkumab (GUS), an interleukin-23 p19 subunit antagonist, in patients with moderately to severely active ulcerative colitis (UC) who had an inadequate response, loss of response, or intolerance to corticosteroids, immunosuppressants and/or advanced therapy (ADT, i.e., tumor necrosis factor antagonists [anti-TNFs], integrin receptor antagonists [vedolizumab], and/or Janus kinase [JAK] inhibitors [tofacitinib]). Here we report induction efficacy results for GUS compared with placebo by history of inadequate response, loss of response, or intolerance to ADT (ADT-IR).

Aims & Methods

Patients were randomized in a 3:2 ratio to receive IV GUS 200 mg or placebo at Weeks (Wks) 0, 4, and 8. The primary analysis population included patients with a baseline modified Mayo score of 5 to 9, inclusive, and a centrally reviewed endoscopy subscore ≥ 2. The primary endpoint was clinical remission at Wk 12. Symptomatic remission, clinical response, endoscopic improvement, histo-endoscopic mucosal improvement (HEMI), and endoscopic normalization were assessed at Wk 12.

Results

Of the 701 patients in the primary analysis population, 357 (50.9%) had no history of ADT-IR (of which 95.0% were ADT-naïve) ; 344 (49.1%) had a history of ADT-IR. Among the ADT-IR patients, the proportions with a history of inadequate response/loss of response/intolerance to TNF antagonists , vedolizumab, tofacitinib, and multiple ADT classes were 87.5%, 54.1%, 18.0% and 47.4%, respectively. At baseline, patients with a history of ADT-IR had longer disease duration (mean 8.9 vs 6.2 yrs), more severe endoscopic disease (endoscopy subscore of 3, 78.5% vs 57.7%), higher CRP (median 5.3 mg/L vs 3.4 mg/L and proportion of patients >3mg/L 64.1% vs 53.7%) and higher fecal calprotectin (median 1647 mg/kg vs 1589 mg/kg and proportion of patients >250 mg/kg 92.3% vs 86.9%) compared with patients without a history of ADT-IR. Greater proportions of patients treated with GUS vs placebo achieved primary and key secondary endpoints at Wk 12 in both subgroups of patients with and without a history of ADT-IR, including clinical remission, symptomatic remission, clinical response, endoscopic improvement, HEMI, and endoscopic normalization (Table 1). The treatment differences for GUS vs placebo were generally greater among patients without a history of ADT-IR compared with those with such a history.

Table 1. Primary and key secondary endpoints by history of inadequate response/loss of response/intolerance to advanced therapy (ADT-IR)

Conclusion

Induction treatment with GUS 200 mg IV resulted in greater improvements compared with placebo across key clinical, symptomatic, and endoscopic/histologic outcomes at Wk 12 among moderately to severely active UC patients with and without a history of ADT-IR.

Disclosure

Brian Bressler is an advisor/speaker and has received research support from Janssen. Steering Committee member for Janssen.
Laurent Peyrin-Biroulet reports personal fees and serves as a Steering Committee member for Janssen.
Jessica R. Allegretti is a consultant and Steering Committee member for Janssen.
Kuan-Hsiang G. Huang is an employee of Johnson & Johnson and owns company stock/stock options.
Matthew Germinaro is an employee of Johnson & Johnson and owns company stock/stock options.
Nicole Shipitofsky is an employee of Johnson & Johnson and owns company stock/stock options.
Ye Miao is an employee of Johnson & Johnson and owns company stock/stock options.
Hongyan Zhang is an employee of Johnson & Johnson and owns company stock/stock options.
Martin Bortlik is a clinical investigator and consultant for Janssen.
Tibor Hlavaty is a clinical investigator and consultant for Janssen.
Toshimitsu Fujii has received research grants and speaker’s fees from Janssen.
Brian G. Feagan received grant/research support from Janssen Research & Development LLC.; a consultant for JnJ/Janssen; a member of the speakers bureau for JnJ/Janssen; a member of the scientific advisory board for JnJ/Janssen. Steering Committee member for Janssen.
Bruce E. Sands reports grants, consulting and speaking fees and other support from Janssen, and is a Steering Committee member for Janssen.
Axel Dignass reports fees for participation in clinical trials, review activities such as data monitoring boards, statistical analysis and end point committees from Janssen; consultancy fees from Janssen; payment from lectures including service on speakers bureaus from Janssen. Steering Committee member for Janssen.

Introduction

Neoplasia detection in inflammatory bowel disease (IBD) poses a clinical challenge. We have previously reported the efficacy of a novel computer aided detection (CADe) algorithm for use in IBD (CADe-IBD) in images and videos. CADe-IBD is designed for use in high-definition white light endoscopy which may reduce operator variability which comes with chromoendoscopy. The added value of CADe-IBD is yet to be explored which this study aims to address.

Aims & Methods

Based on pilot data this study was powered as a superiority test, with 90% power for increased sensitivity of non-expert endoscopists with CADe-IBD compared to without CADe-IBD. 10 experienced and 11 less experienced endoscopists were asked to review short video clips of high definition white-light IBD endoscopy on a bespoke computer software designed by NEC Japan. The videos were watched once without CADe-IBD and then again with CADe-IBD when the videos were randomly redistributed to reduce recall bias. After each video the endoscopist was asked to select ‘polyp’, ‘no polyp’ or ‘uncertain’. If the endoscopist selected ‘polyp’ for a video they were asked to annotate the suspected abnormality.

Results

The study included 285 videos from IBD colon. 131 videos had lesions which showed a diverse range of size, morphology, histology and background mucosal inflammation. Videos were from all 3 major endoscopy processors (Pentax, Olympus and Fuji).
The standalone CADe-IBD sensitivity and specificity were 94.7% and 70.8% respectively. In comparison, experienced endoscopist sensitivity and specificity were 90.5% and 80.2% and less experienced endoscopists sensitivity and specificity of 73.7% and 82.1%. CADe-IBD sensitivity was significantly better than both experts and non-exerts (p<0.05).
With the addition of CADe-IBD, less experienced endoscopists sensitivity increased significantly to 89.2% (p <0.01) and specificity remained similar at 81.0%. Experienced endoscopist sensitivity also improved with CADe-IBD to 93.4% which was again statistically significant (p<0.05).
Sub-analysis was performed based on neoplastic or non-neoplastic histology, polypoid (Paris Is, Isp, Ip) and non-polypoid (Paris IIa/IIb/IIc) morphology and Mayo endoscopic sub-score for mucosal inflammation (Table 1).

Conclusion

CADe-IBD has a standalone performance which is superior to both less experienced and experienced endoscopists. The addition of CADe-IBD significantly improved lesion detection regardless of experience. This superior performance was observed in neoplastic lesions, non-polypoid lesions and in the presence of mild background mucosal inflammation which is of high clinical importance. This calls for a head to head comparison of CADe-IBD with conventional surveillance techniques.

References

Abdelrahim M, Siggens K, Iwadate Y, et al New AI model for neoplasia detection and characterisation in inflammatory bowel disease Gut 2024;73:725-728.

Disclosure

Professor Bhandari has received research grants or is the advisory board for Fujifilm, Boston, Olympus, Pentax, 3-D matrix, NEC (Japan) and Medtronic