Clinical Case Summary

References

Case Jr R, et al. (BMJ Case Rep 2019;12:e226355. doi:10.1136/bcr-2018-226355)
Anat Yerushalmy-Feler et al. , Diagnosis and Management of CMV Colitis doi.org/10.1007/s11908-019-0664-y

Disclosure

nothing to declare

Introduction

Placebo and treatment response rates in randomized ulcerative colitis (UC) trials impact clinical trial design and patients required to demonstrate efficacy. Prior efforts to predict endoscopic healing (EH) in UC trial patients using clinical measures have failed to produce meaningful predictive power. Using AI analysis of baseline endoscopy videos from the ustekinumab UNIFI registration Phase 3 trial in moderate to severe UC, we demonstrate robust accuracy to predict 8 week endoscopic healing in ulcerative colitis patients treated with placebo or drug.

Aims & Methods

300 screening endoscopy videos were obtained from UC patients receiving placebo, and 606 screening endoscopy videos were obtained from UC patients receiving ustekinumab (130mg or 6mg/kg) in the UNIFI induction trial. For all videos, visual features were extracted using EndoDINO, an AI foundation model pre-trained on the largest reported endoscopy video dataset. For each patient, 21 baseline clinical measures were also obtained as predictors. These included fecal calprotectin, albumin, C-reactive protein, lactoferrin, Geboes grading, prior biologics failure, prior steroids failure, UCEIS scores, and Mayo scores.
For each of the placebo and drug arms, we trained two binary classification models: one using the AI visual features and the other using the baseline clinical measures as predictors. Each model was trained to predict 8 week EH, defined as an endoscopy subscore of the Mayo score equal to 0 or 1. We performed 5-fold cross validation and reported the average Area Under the Receiver Operating Curve (AUROC).
This study, carried out under YODA Project # 2024-0668, used data obtained from the Yale University Open Data Access Project, which has an agreement with JANSSEN RESEARCH & DEVELOPMENT, L.L.C.. The interpretation and reporting of research using this data are solely the responsibility of the authors and does not necessarily represent the official views of the Yale University Open Data Access Project or JANSSEN RESEARCH & DEVELOPMENT, L.L.C.. The original proposal can be found: https://yoda.yale.edu/data-request/2024-0668/.

Results

For the placebo arm, our best performing model using AI visual features predicted 8 week endoscopic healing with an AUROC of 0.78, compared to 0.70 for our best performing model using the combination of 21 baseline clinical measures.
For the drug arm, our best performing model using AI visual features predicted 8 week endoscopic healing with an AUROC of 0.75, compared to 0.72 for our best performing model using the combination of 21 baseline clinical measures.
For models using baseline clinical measures, table 1 shows the importance factor for select measures across arms, calculated using XGBoost built in feature importance.

Conclusion

Our results demonstrate AI analysis of baseline colonoscopy videos with the EndoDINO AI foundation model provides robust accuracy for predicting 8 week EH in ulcerative colitis patients in both placebo and drug arms. AI placebo response prediction could be used in future trial design as covariate analysis to increase statistical power or reduce necessary sample size.
In future work, we aim to improve predictive power by combining our AI visual features with the most important baseline clinical measures through different modeling techniques.

References

Sands BE, Sandborn WJ, Panaccione R, et al. Ustekinumab as induction and maintenance therapy for ulcerative colitis. N Engl J Med. 2019;381(13):1201–1214. doi:10.1056/NEJMoa1900750

Disclosure

AK, PD, MC, and MS are employees of Virgo Surgical Video Solutions, Inc., which developed technology used in this work.

Introduction

The lack of reliable intestinal fibrosis imaging modalities hampers anti-fibrotic therapy development. Gallium-68 Fibroblast Activation Protein inhibitor (⁶⁸GaFAPi) PET/CT offers promising in vivo fibrosis visualization via Fibroblast Activation Protein (FAP) targeting. Despite reported overexpression of FAP in chronic inflammatory and fibrotic diseases, data in inflammatory bowel disease (IBD) and the potential use of ⁶⁸Ga-FAPi PET/CT as a non-invasive fibrosis biomarker remains sparse. We performed the first in-human study to investigate ⁶⁸Ga-FAPi PET/CT to detect and quantify fibrosis in IBD patients using pharmacokinetic modelling.

Aims & Methods

Symptomatic patients with active Crohn’s disease (CD) or ulcerative colitis (UC) waiting for surgical resection were invited to the study and consented. Patients underwent a 60-minute dynamic ⁶⁸Ga-FAPi PET/CT scan, followed by 12 minutes of static imaging within 12 weeks before surgery. Arterial and peripheral blood samples were obtained during the first 10 scans to validate image- versus plasma-derived input function. Two blinded readers independently assessed the PET/CT images, wherein bowel segments showing increased uptake against local background were delineated and validated by a non-blinded third reader. Bowel segment uptake, time-activity-curves (TAC) and input functions were analyzed using irreversible (2T3k), reversible two-tissue compartment models (2T4k), reference tissue models, and standardized uptake values acquired on static images 50-60 minutes post-injection (SUV_50-60). Surgical resection specimens were selected based on ⁶⁸Ga-FAPi PET/CT with focus on segments with high FAPi uptake. In the tissue samples, gene expression and protein levels of FAP were determined using quantitative polymerase chain reaction (qPCR) and immunohistochemistry. To allocate samples to a fibrotic, inflamed or combined phenotype, hematoxylin and eosin and Masson’s trichrome staining were performed for inflammatory activity and fibrosis scoring by a blinded pathologist (GH) using previously described histological inflammation and fibrosis scores.¹

Results

In total, 15 CD and 4 UC patients (mean age 39 years, median disease duration 12 years) underwent FAPi PET/CT prior to surgery. On these 19 scans, 50 bowel regions of interest were identified and 52 transmural surgical samples were acquired. FAPi tracer uptake was markedly increased in inflamed and fibrotic regions of all participants compared to reference, non-pathologic bowel segments (SUV_50-60 ranging 1.0 to 11 versus 0.3 to 1.1). Although a higher tracer activity was seen in all patients, the inflammatory bowel segments showed the highest peak in tracer activity, followed by a decline in uptake, indicating increased blood flow, vascular permeability and reversible, potentially non-specific or low-binding affinity, uptake. Contrarily, the TAC of patients with a more fibrotic phenotype, showed a low peak uptake followed by a relatively stable plateau, suggesting a more irreversible or higher binding affinity uptake component.

Conclusion

⁶⁸Ga-FAPi PET/CT shows significant promise for IBD assessment as uptake aligns with histological fibrosis scoring, enhanced FAP gene and protein expression in resected inflammatory and stenotic specimens in CD and UC patients. Pharmacokinetic analyses suggests more frequent irreversible uptake kinetics, indicating higher FAP binding affinity, in predominantly fibrotic segments compared to inflamed regions, thus underscoring its potential to differentiate fibrosis from inflammation.

References

1. Ke BJ, Abdurahiman S, Biscu F, Zanella G, et al. Intercellular interaction between FAP+ fibroblasts and CD150+ inflammatory monocytes mediates fibrostenosis in Crohn's disease. J Clin Invest. 2024 Jul 23;134(16):e173835. doi: 10.1172/JCI173835. PMID: 39042469; PMCID: PMC11324301.

Disclosure

M.L.: Consultancy/lecture fees from Abbvie, Bristol Myers Squibb, Eli Lilly, Galapagos, Janssen-Cilag, Johnson & Johnson, Medtronic, Pfizer, Takeda, Tillotts. Central reader for Alimentiv. Grants received from ZonMW, Galapagos, NFU transformation deal, and TKI (unrelated)
G.R.A.M.D.: research grants from Abbvie, Alimentiv, BMS, J&J, Pfizer, Takeda; consulting fees from Abbvie, Agomab, Alimentiv, AstraZeneca, AMT, Bristol Meiers Squibb, Boehringer Ingelheim, Celltrion, Eli Lilly, Exeliom Biosciences, Galapagos, Index Pharmaceuticals, Glaxo Smith Kline, Pfizer, Johnson and Johnson, Polpharma, Procise Diagnostics, Prometheus laboratories, Prometheus Biosciences, Protagonist, and Ventyx; speaker’s fees from Abbvie, Celltrion, Galapagos, Pfizer, and Takeda; data monitoring board: Galapagos, AstraZeneca, Seres Health (unrelated)