Artificial Intelligence, Cancer Prevention, Gastric Cancer, Helicobacter Pylori Infection
Conditions
Brief summary
This clinical study aims to be used to implement and validate the AIDA tool in two phases: * Phase 1: Risk stratification and personalised recommendations & Model development * Phase 2: Mechanistic Model (Bioresource) development & testing
Detailed description
The AIDA objective (project) is to develop and validate a multidisciplinary AI-powered assistant that helps clinicians diagnose precancerous inflammation, suggests personalised therapeutic strategies for medical treatment and follow-up, and makes personalised recommendations for monitoring patient health status, thus contributing to gastric cancer prevention. This prospective clinical study aims to implement and validate such tool.
Interventions
BEHAVIORALHealth reccommendations
Patients are given recommendations according to their risk group, based on the model which already predicted health indicators. This information will be sent to the patient's treating physician so that treatment and recommendations are aligned with the clinical care practice based on the European Code of Cancer guidelines, the H. pylori best practices guidelines and European GIM guidelines
Sponsors
Fundación para la Investigación del Hospital Clínico de Valencia
Study design
Allocation
NON_RANDOMIZED
Intervention model
PARALLEL
Primary purpose
PREVENTION
Masking
NONE
Intervention model description
In practice, for rare diseases we chose a sample size that will allow estimation of sensitivity of the risk score to a pre-specified precision. We have chosen the sample size assuming that the resulting precancer detection model will have a sensitivity of at least 85%, and so that the lower 95% confidence interval will exceed 76%, with 80% power. The R package MKmisc estimates sample size for a proportion based on the Binomial distribution rather than a Normal approximation. In this case, we have estimated a sample size of 141 GIM + cancer and 141 GIM controls (282 in total). With 5% of samples assumed to be failing to be analysed due to a variety of technical reasons, we would need \ 300 samples in total. In addition, to train the system 150 gastric cancer cases will be recruited which will provide pathology samples images and endoscopy images that will be taken as part of the clinical practice.
Eligibility
Sex/Gender
ALL
Age
18 Years to No maximum
Healthy volunteers
Yes
Inclusion criteria
* Subjects ≥ 18 years old with a diagnosis of GIM or previous or current H. pylori infection, to whom a gastroscopy is indicated within clinical care * Availability of a signed informed consent form to participate in the study
Exclusion criteria
* Patients to whom an endoscopy is performed for the follow-up of another illness such as oesophageal varices and/or for therapy such as endoscopic dilation, feeding tube placement or endoscopic resection * Subjects with a clinical diagnosis of gastric diseases other than GIM or GC * Patients who have received antimicrobials during the four weeks prior to the endoscopy * Patients who have received proton pump inhibitors and/or bismuth-based treatments at least two weeks prior to the endoscopy * Subjects for whom clinical data are not available: H. pylori status, eradication treatment, sex, age, tobacco smoking, and first-degree family history of gastric cancer * Subjects who lack the mental capacity to understand the nature and requirements of the study and who lack the ability to give informed consent
Design outcomes
Primary
| Measure | Time frame | Description |
|---|---|---|
| Risk of developing Gastric Cancer based on medical records | At the recruitment stage | Scoring patients as low (\>30) / medium / high risk (\<6) according to: * Degree of 'healthiness' of lifestyle (1 to 10, with 10 as very low and 10 as very high) * Co-morbidities (Yes = 1 / No = 5) * H. pylori infection history (Yes = 1 / No = 5) * Previous gastric intestinal metaplasia (Yes = 1 / No = 5) * Dysplasia or atrophy related to chronic gastritis (Yes = 1 / No = 5) * Family history of cancer (Yes = 1 / No = 5) |
| H. pylori Eradication Therapy Recommendation | From 30 days after the H.Pylori positive test result to one year after the first recommendation | AI driven H. pylori Eradication Therapy Recommendation |
| GIM risk score assessment using imaging modalities | At the recruitment stage | AI driven GIM risk score assessment of pre-cancerous lesions using imaging modalities based on: QLQ C30 and STO22 EORTC questionnaire, Healthy lifestyle questionnaire (adapted from EPICs), Baseline clinical data, If H. pylori positive: adherence to treatment: yes/no; eradication yes/no, If GIM: adherence to follow-up guidelines yes/no |
Countries
Spain
Contacts
Primary ContactAna Miralles Marco, PhD
Outcome results
None listed