Abstract 1181P
Background
Diagnosing cancer earlier can support timely treatment and improve patient prognosis. Innovative diagnostic tests need to appropriately respond to unmet needs. Target Product Profiles (TPPs) are specification documents that serve as demand-signalling tools to inform developing improved tests – in terms of scientific performance, fit with clinical pathways and patient needs. There is growing interest in their use. Commissioned by Cancer Research UK, RAND Europe and the Office of Health Economics conducted a study to establish a methodological process for developing robust diagnostic TPPs for cancer and identify the types of features TPPs should consider.
Methods
A mixed-methods approach combined international literature review, interviews, workshops with UK experts and developing an economic modelling tool for TPP requirements. 129 individuals contributed (academics, funders, healthcare professionals, industry, policy, regulatory and HTA experts, health economists, and patient and public voice).
Results
Establishing specifications for combinations of core features driving a test’s value is key to TPP development. Trade-offs between different requirements must be considered. Relevant features types fall into 9 categories (e.g. unmet need, analytical performance, clinical validity, clinical utility, human factors, infrastructure, cost/economic, regulation, environmental). Fit-for-purpose TPPs require input from diverse stakeholders through desk research, consultation and modelling methods. TPP development has 2 stages: inception (establishing core working group, governance, action plan) and implementation (scoping unmet need and key test requirements, TPP drafting, consensus building).
Conclusions
This internationally-relevant, tumour-site agnostic guide is an evidence-based resource for designing TPPs for specific cancer tests and use cases. It highlights multifaceted factors and practical tips for deciding which features to consider and methods to use in TPP development. Robust TPPs are an important tool for developing improved cancer diagnostic tests, but need to align with wider innovation and health system efforts.
Clinical trial identification
Editorial acknowledgement
Legal entity responsible for the study
RAND Europe; Cancer Research UK.
Funding
Cancer Research UK.
Disclosure
All authors have declared no conflicts of interest.
Resources from the same session
1182P - Determination of tumor PSMA expression in prostate cancer from blood using a novel epigenomic liquid biopsy platform
Presenter: Praful Ravi
Session: Poster session 09
1183P - Impact of multicancer early detection (MCED) test on participant-reported outcomes (PRO) and behavioral intentions by cancer risk
Presenter: Christina Dilaveri
Session: Poster session 09
1184P - Early real-world experience with positive multi-cancer early detection (MCED) test cases and negative initial diagnostic work-up
Presenter: Candace Westgate
Session: Poster session 09
1185P - Clinical applications of a novel blood-based fragmentomics assay for lung cancer detection
Presenter: Marc Siegel
Session: Poster session 09
1186P - SmartCS-LPLLM: Enhancing early cancer detection through ctDNA methylation analysis leveraging large language models
Presenter: Li Chao
Session: Poster session 09
1187P - Molecular diagnosis of lung cancer via ctDNA and ctRNA detection on bronchoscopic fluid specimens from 31 patients: A retrospective analysis
Presenter: Vincent Fallet
Session: Poster session 09
1188P - Modeled economic and clinical impact of a multi-cancer early detection (MCED) test in a population with hereditary cancer syndromes
Presenter: Sana Raoof
Session: Poster session 09
1189P - Cancer genome interpreter: A data-driven tool for tumor mutation interpretation
Presenter: Santiago Demajo
Session: Poster session 09
1190P - Circulating tumor DNA from the tumor-draining pulmonary vein as a biomarker in resected non-small cell lung cancer
Presenter: Raphael Werner
Session: Poster session 09
1191P - Efficient lung cancer stage prediction and outcome informatics with Bayesian deep learning and MCMC method
Presenter: Maria Gkotzamanidou
Session: Poster session 09