Introduction

There is a steadily increasing interest among patients in obtaining medical information from electronic sources. ChatGPT, a conversational agent utilizing artificial intelligence, has been demonstrated as a reliable source of information for patients before a colonoscopy (1) as well as in various medical purposes.

Aims & Methods

The aim of this study is to determine the ability of ChatGPT 3.5 to issue recommendations after a screening colonoscopy for colorectal cancer. This is a single-center observational cross-sectional study conducted in 2023. Only colonoscopies for screening purposes were included. Colonoscopies performed for gastrointestinal bleeding, follow-up of chronic inflammatory bowel diseases, post-operative follow-up of colorectal cancer, or in patients under 40 years old were excluded. Colonoscopy reports and corresponding pathology reports were entered into a conversation with ChatGPT 3.5. The colonoscopy report includes the quality of colonic preparation, withdrawal time, results, and specimens sent for histology. ChatGPT 3.5 was asked to determine the time interval for the follow-up colonoscopy using the Consensus of the American multisociety taskforce of 2020 (2). The answers from ChatGPT were compared to the recommendations, and were validated by two gastroenterologists. The answers were categorized as “correct”, “underestimated” if the interval recommended by ChatGPT is shorter than the one provided by the consensus, and “overestimated” if the interval recommended by ChatGPT is longer than the one provided by the consensus.

Results

A total of 1259 colonoscopies were considered for analysis. The average age at the time of colonoscopy was 61.6 ± 11.0 years old. Among the participants, 688 (54.6%) were women and 517 (45.4%) were men.
ChatGPT® issued a correct recommendation in 892 cases (70.8%). The surveillance interval was underestimated in 260 (20.7%) cases and overestimated in 107 (8.5%) cases.
The accuracy of the recommendations for each identified category is represented in the table below.

Table 1. Accuracy of the recommendations issued by ChatGPT 3.5 according to the ​colonoscopy result.

Conclusion

ChatGPT appears to be a reliable tool for issuing recommendations after a screening colonoscopy. It is limited by the recognition of poor preparation requiring closer surveillance. Artificial intelligence could be exploited and integrated into colorectal cancer screening programs. It could also be used by a patient to remind them of their next colonoscopy, except in cases of poor colonic preparation.

References

1. Lee TC, Staller K, Botoman V, Pathipati MP, Varma S, Kuo B. ChatGPT Answers Common Patient Questions About Colonoscopy. Gastroenterology. 2023;165(2):509-511.e7. doi:10.1053/j.gastro.2023.04.033
2. Gupta S, Lieberman D, Anderson JC, et al. Recommendations for Follow-Up After Colonoscopy and Polypectomy: A Consensus Update by the US Multi-Society Task Force on Colorectal Cancer. Gastrointest Endosc. 2020;91(3):463-485.e5. doi:10.1016/j.gie.2020.01.014

Introduction

Vedolizumab (VDZ), a gut-selective anti-lymphocyte trafficking drug indicated for ulcerative colitis (UC), is administered intravenously at Weeks 0, 2, and 6, then every 8 weeks (Q8W). Dose escalation to every 4 weeks (Q4W) may recapture response to treatment in ~50% of patients with inadequate response. Real-world outcomes and factors associated with VDZ dose escalation from Q8W to Q4W remain underexplored.

Aims & Methods

Adults with moderate to severe UC initiating VDZ Q8W maintenance were prospectively followed in a Canadian patient support program (2015-2023). Two cohorts were included: 1) patients escalated to Q4W during maintenance after week 14 infusion (Q8W>Q4W maintenance) and 2) patients maintained on Q8W treatment for ≥2 years without escalating (Q8W maintenance). Rates of clinical remission (partial Mayo Score [PMS]<3) were described by prior biologic experience. Associations between baseline characteristics and clinical remission 3 months after Q4W dose escalation were assessed using Wilcoxon rank sum test and Fisher’s exact test. Treatment persistence at 12 and 24 months after initiation and dose escalation were evaluated in Q8W>Q4W maintenance using Kaplan-Meier survival analysis.

Results

373 patients (68% bio-naïve) and 300 patients (72% bio-naïve) were included in the Q8W>Q4W maintenance and Q8W maintenance cohorts, respectively. Median age at treatment initiation was 47 years in Q8W>Q4W maintenance and 48 years in Q8W maintenance. Both cohorts had a median baseline PMS score of 6. In Q8W>Q4W maintenance, median disease duration before treatment initiation was 4 years and median follow-up was 19 months. In Q8W maintenance, median disease and follow-up duration were 5 years and 35 months, respectively. 14 weeks post-treatment initiation, rates of remission in Q8W>Q4W maintenance were 49% (bio-naïve) and 42% (bio-experienced), and in Q8W maintenance were 72% (bio-naïve) and 70% (bio-experienced) (Table 1). Among patients in Q8W>Q4W maintenance not in remission when dose escalating, 40% achieved clinical remission 3 months after dose escalation. These patients were older at treatment initiation (49 vs. 41 years, p=0.043) and had longer disease duration (6 vs. 3 years, p=0.029) than those who did not achieve remission. In Q8W>Q4W maintenance, 12- and 24-month treatment persistence after initiation was 83% and 48% (bio-naïve) and 79% and 51% (bio-experienced), respectively. After dose escalation, 12- and 24-month treatment persistence was 48% and 39% (bio-naïve) and 53% and 45% (bio-experienced).

Conclusion

Patients with UC who underwent VDZ Q8W>Q4W dose escalation achieved lower rates of remission than those remaining on Q8W, regardless of prior bio-experience. Age at treatment initiation and disease duration were significantly associated with remission among those who escalated to Q4W.

Disclosure

JFL: AbbVie, BMS, Janssen, Frenesius-Kabi, Pfizer, Sandoz, Takeda.
EJB: Participation in advisory boards or speaker services: Abbvie, Janssen, Takeda, Pfizer, Merck, Amgen, Pendopharm, Jamp, Fresenius, Kabi, Bausch Health, Celltrion, Eli Lilly, BMS, Grant/Research: Janssen, Abbvie, Any other investment or relationship that could be judged by a reasonable and knowledgeable participant to have the potential to influence the content of the training activity, Abbvie, Janssen, Takeda, Pfizer, Merck, Amgen, Pendopharm, Jamp, Fresenius, Kabi, Bausch Health, Celltrion, BMS.
AHS: Advisory Board Membership/Consultant: Abbvie, Amgen, BioJAMP, BMS, Celltrion, Fresenius Kabi, Janssen, McKesson, Mylan Pharmaceuticals, Organon, Pendopharm, Roche, Pfizer, Sandoz, Takeda, Viatris, Speakers’ Bureau: Abbvie, Amgen, Ferring, Fresenius Kabi, Janssen, Organon, Pfizer, Sandoz, Takeda, Research Grants: Abbvie, Arena, Celgene/BMS, Genentech, Janssen, Roche, Takeda.
PW: Personal fees from AbbVie, Ferring, Biogen, and Janssen
MJA, AW, RW: were employees of Takeda Canada Inc. at the time of the study.
RM, JW, CP: were employees of Pentavere Research Group Inc. at the time of the study.
LPB: Consulting from AbbVie, Alimentiv, Alma Bio Therapeutics, Amgen, Applied Molecular Transport, Arena, Biogen, BMS, Celltrion, CONNECT Biopharm, Cytoki Pharma, Enthera, Ferring, Fresenius Kabi, Galapagos, Genentech, Gilead, Gossamer Bio, GSK, HACPharma, IAG Image Analysis, Index Pharmaceuticals, Inotrem, Janssen, Lilly, Medac, Mopac, Morphic, MSD, Norgine, Novartis, OM Pharma, ONO Pharma, OSE Immunotherapeutics, Pandion Therapeutics, Par’Immune, Pfizer, Prometheus, Protagonist, Roche, Sanofi, Sandoz, Takeda, Theravance, Thermo Fisher, Tigenix, Tillots, Viatris, Vifor, Ysopia, Abivax, Samsung, Ventyx, Roivant, and Vectivbio; grants from Takeda, Fresenius Kabi, and Celltrion; and lecture fees from Galapagos, AbbVie, Janssen, Genentech, Ferring, Tillots, Celltrion, Takeda, Pfizer, Sandoz, Biogen, MSD, Amgen, Vifor, Arena, Lilly, Gilead, Viatris, and Medac.
BB: Advisor/Speaker: Ferring, Janssen, Abbvie, Takeda, Pfizer, BMS, Merck, Sandoz, Organon, Lifelabs, Celltrion. Advisor: Alimentiv, Gilead, Iterative Health, AMT, Celgene, Merck, Amgen, Pendopharm, Eli Lilly, BMS, Fresenius Kabi, Mylan, Viatris, Bausch Health, Celltrion Healthcare, BioJamp Pharma, Eupraxia. Research support: Janssen, Abbvie, GSK, BMS, Amgen, Genentech, Merck. Stock Options: Qu Biologic.

Introduction

Recurrent flares cause disease progression and have a pronounced impact on the quality of life of people with Inflammatory Bowel Disease (IBD). Models using clinical characteristics only moderately predict flares and therefore early intervention or even prevention of flares is difficult in clinical practice. With the rise of remote monitoring platforms such as myIBDcoach, which capture besides clinical disease activity, modifiable lifestyle and psychosocial risk factors and patient-reported outcome measures (PROMs), there is a potential to harness real-world data to increase flare prediction. We build a predictive model to predict flares of IBD. The aim of this study was to validate the model in an independent dataset using the Dutch National IBD core outcome set including PROMS.

Aims & Methods

429 patients from the Patient-Reported Infections Questionnaire [PRIQ] cohort were selected to create the updated model. Data for these patients included both baseline clinical data and PROMs related to lifestyle and psychosocial factors from the myIBDcoach telemedicine platform (ClinicalTrials.gov, NCT02173002) from November 2022 till June 2024. The myIBDcoach data was captured monthly or every 3 months from the platform based on disease activity. Patients were followed for at least 12 months. A flare was defined as clinical disease activity (using the Monitor IBD At Home questionnaire) and one or more of the following: fecal calprotectin > 250g/g, endoscopy, MRI, or CT showing signs of disease activity. Associations between flares and baseline clinical variables (Montreal classification, demographic data and drug use) alone, flares and individual PROMs variable categories (subjective health, psychosocial or lifestyle variables) from the myIBDcoach platform, and between flares and the baseline clinical variables in incorporation with myIBDcoach data were estimated using stepwise group-Lasso logistic regression (G-LASSO), which were evaluated with performance matrices using accuracy, area under the curve (AUC), sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV).

Results

The performance of the G-lasso regression model with baseline variables and myIBDcoach PROMs (psychosocial, lifestyle, and medication factors) was better (accuracy: 75%, ROC-AUC: 81%, sensitivity: 71%, specificity: 77%, PPV: 80%, NPV: 69%) than that of the model with baseline data alone (accuracy: 67%, ROC-AUC: 70%, sensitivity: 70%, specificity: 64%, PPV: 71%, NPV: 63%). Based on these results, a higher ROC-AUC suggests that integrating PROMs may improve discriminative capacity.

Conclusion

The inclusion of subjective health and modifiable lifestyle and psychosocial data improved flare prediction in contrast to clinical characteristics alone, which was evidenced in the model performance matrices (accuracy, AUC, sensitivity, specificity, PPV, NPV) while reducing uncertainty and misclassification

References

1. Lalisang RCA, Adriaans G, de Jong M, Van der Meulen-de Jong A, Romberg-Camps M, Mahmmod N, Markus-de Kwaadsteniet T, Dijkstra G, Haans J, Stamm C, Vanwersch R, Jonkers D, Almeida RJ, Pierik MJ, MyIBDcoach Study Group. Can lifestyle and psychosocial factors predict flares of IBD; an exploratory study using telemedicine. J Crohn’s Colitis. 2020;14:S059-60.
2. M. J. de Jong et al., “Telemedicine for management of inflammatory bowel disease (myIBDcoach): a pragmatic, multicentre, randomised controlled trial,” The Lancet, vol. 390, no. 10098, pp. 959–968, Sep. 2017, doi: 10.1016/S0140-6736(17)31327-2.

Disclosure

This project and the submitted abstract titled "Advancing Predictive Modeling of Inflammatory Bowel Disease (IBD) Flares: A Data-Driven Approach Using Lifestyle and Psychosocial Factors from a Remote Monitoring Platform" were supported by Takeda. Takeda provided funding for the research work that formed the basis of this abstract.

Introduction

Wee1 negatively regulates the G2/M checkpoint of the cell cycle, inhibiting cells with DNA damage from entering mitosis and inducing apoptosis. Although efficacy of Wee1 inhibitors as single agents or in combination with other agents has been reported in a variety of carcinomas and the association between Wee1 inhibitors and tumor immunity has been reported, the effect of Wee1 inhibitors on the tumor immune microenvironment in colorectal cancer (CRC) has not been clarified. Since p53 and KRAS are involved in cell cycle regulation at the G1/S checkpoint, mutations in them may increase dependence on the G2/M checkpoint for DNA repair, making Wee1 inhibitors more effective.

Aims & Methods

This study aimed to clarify the association between Wee1 expression and clinicopathological features in CRC and the effect of Wee1 inhibitors on the tumor immune microenvironment in CRC. Study 1: We performed immunostaining on 152 CRC cases surgically resected at our hospital between May 2013 and February 2015, and classified them into positive and negative groups according to the degree of Wee1 expression and investigated the association between Wee1 expression and clinicopathological features. Study 2: We measured Wee1 expression levels by Realtime PCR on 10 human-derived CRC cell lines and 2 mouse-derived CRC cell lines. We examined the effect of Wee1 inhibitor (AZD1775) on cell proliferation using Timelapse system. We examined the effect of Wee1 inhibitor on cell cycle using Cell Cycle Analysis Kit. We examined the effects of Wee1 inhibitor on apoptosis induction were examined using the Apoptosis Detection Kit. Study 3: We created the CRC orthotopic transplantation mouse model by transplanting the CRC cell line MC38 derived from C57BL/6 mice into the cecum of mice, comparing the tumor sizes of the control and treatment groups treated with Wee1 inhibitor for 21 days, and evaluating the effect of Wee1 inhibitor on the tumor immune microenvironment by immunostaining and RNA sequencing of mice tumor specimens.

Results

Study 1: No significant association was found between Wee1 expression and clinicopathological features or prognosis. However, overall survival rate was significantly higher in Wee1 positive and KRAS-mutant patients compared to Wee1 negative and KRAS wild-type patients. Study 2: Real time PCR showed that Wee1 expression was higher in most CRC cell lines than in normal colon mucosa. Wee1 inhibitor attenuated cell proliferation in a concentration-dependent manner in all CRC cell lines, and the effect was independent on Wee1 expression level. Cell cycle analysis showed that the percentage of cells in G2/M phase increased after exposure to Wee1 inhibitor, especially in cell lines with both p53 and KRAS mutations. In addition, the number of apoptotic cells increased after exposure to Wee1 inhibitor, but in the cell lines with KRAS mutation and p53 wild type, the number of apoptotic cells did not increase. Study 3: In the orthotopic transplanted mouse model, tumor volume was significantly smaller in the treated group compared to the control group, and immunohistological examination and RNA sequencing of mice tumor specimens showed enhanced tumor immunity and suppressed stromal reaction.

Conclusion

Wee1 inhibitor attenuated CRC cell proliferation through modulating cell cycle, especially in cell lines with p53 mutation and attenuated tumor growth via enhancement of tumor immune response and suppression of stromal reaction in orthotopic transplanted mouse model. Wee1 inhibitor AZD1775 could be the promising treatment option for CRC.

Introduction

The prevalence of musculoskeletal disorders (MSD) among gastroenterologists is estimated between 39 to 89%. Factors associated with MSDs among endoscopists include a high number of procedures (over 20/week), prolonged procedure times (over 16 hours/week), the age of the endoscopist (> 40 years vs ≤ 40 years), and female sex.
This study aimed to evaluate the prevalence of MSD among French gastroenterologists and identify risk factors of MSD

Aims & Methods

A computerized, anonymized questionnaire was e-mailed to all members of French society of endoscopy (SFED) over a one-month period. Demographic data were collected, as well as the level of endoscopic activity performed, the quantity and volume of endoscopies performed per week, the presence of pain associated with the practice of endoscopy, and an assessment of stress.

Results

485 gastroenterologists responded to the questionnaires, the majority were male (55%), with a median age of 48 years (IQR 37–60), and 45.6% practiced endoscopy for more than 20 years. 74.3% of respondents worked full-time, mostly in private practice (47% in private practice, 13% in mixed public-private practice). Endoscopy was performed for a median of 15 hours per week (IQR 10–18.8), with 24 procedures per week (IQR 15–30).
76.3% of participants had already experienced joint and/or musculoskeletal pain related to work. Pain complaints mainly concerned the lower back (55%), shoulders (36% left, 37% right%), neck (50%), and wrists (30% both side). More than half of the participants (55.9%) reported that the pain persisted after procedures, with 17.8% experiencing nocturnal pain and about a quarter reporting pain lasting more than 3 months.
The impact of MSD on workload was evaluated. 5.1% of the gastroenterologists reported loss of efficiency, 3,8% adapted their job, 2.4% decreased their activity, and 2.1% had work stoppage. 35% declared to have management of their MSD (28.1% painkiller, 23.2% reeducation, 3.5% surgery). To reduce discomfort, 74.7% of respondents reported using an adjustable-height examination table, and half adjusted the height and distance of the screen.
In multivariate analysis, factors associated with MSD were, femal sexe (OR 1.85 (1.15-3.01) p = 0.012, number of endoscopy (OR 1.05 (1.01-1.11) p=0.03, activity > 20 years (OR 2.42 (0.89, 6.47) p=0.079.) The risk of MSD was higher after 16.5 hours of endoscopy/ week (Se 35%, Sp 83%).

Conclusion

Musculoskeletal pain among French gastroenterologists is common and concerning. Therefore, it is essential to develop preventive programs to support their health and ensure the quality and safety of care.