CANDETECT: Accelerating detection of upper gastro-intestinal (UGI) cancers using a multi-cancer early detection platform in primary care

Background Outcomes for the commonest upper gastro-intestinal (UGI: oesophago-gastric, pancreatic and biliary tract) cancers remain very poor worldwide.

Most UGI cancer patients have relevant symptoms and multiple GP consultations in the two years pre-diagnosis, suggesting earlier detection is possible, with opportunities to treat with curative intent, improve disease outcomes and patient experiences.

The CanTest framework provides a translational pathway for promising clinical, biomarker and imaging data, which could accelerate detection of UGI cancers in primary care by enabling rapid triage.

Aims and methods We will develop and evaluate a UGI multi-cancer early detection (MCED) Platform for primary care using three interrelated workstreams. -Workstream 1 will review evidence on using electronic health records (EHR) to identify UGI cancers.

We will then utilise national EHR databases, linking data from general practice with hospital, cancer registry and ONS mortality data, to develop a risk prediction tool for UGI cancers using standard and machine learning approaches.

We will then use similar data from East London focusing on potential health inequalities to give additional insights. -Workstreams 2.1 and 2.2 will investigate routinely available and emerging biomarkers and imaging tests and validate in-house-discovered biomarkers, before refining and validating new biomarker panels and approaches.

We will use systematic reviews, complex proteomic and multiplex assays, machine-learning algorithms, plus pragmatic studies embedded in ongoing data collection at UCL Hospital Rapid Diagnostic Centre. -Workstream 3 will run in parallel with Workstreams 1 and 2, incorporating findings into a user-friendly UGI MCED Platform with a novel clinical decision support tool embedded into commonly used primary care software for real-time risk assessment.

It will require detailed development, in co-production with GPs and patients.

An iterative cycle of development and evaluation will use multi-methods approaches to test functionality, feasibility, patient and clinician acceptability, and indicative costs and benefits working towards real-world utility by programme end.

How the results will be used Patients with non-specific symptoms will benefit from a ‘local, smart, one-stop’ diagnostic approach, integrating clinical, biomarker and imaging data for rapid triage and appropriate management in their primary care community setting.

If the UGI MCED Platform proves effective, feasible and acceptable, the results will underpin an application for a large, phase III trial.

The methodological approaches will provide a pipeline for further emerging risk-informed approaches, and this exemplar MCED Platform will be available for future refinement and augmentation with novel biomarkers, imaging and other approaches, for both UGI and other symptom complexes.