Background

Despite substantial advances in cancer research, translating evidence into equitable public health action remains challenging in low- and middle-income countries (LMICs), where fragmented research ecosystems and limited institutional capacity constrain the generation and uptake of contextually relevant evidence. Strengthening locally led research and embedding implementation science within cancer control programmes are therefore critical to improving equity and care.
In alignment with the IARC@60 themes Cancer Research into Action and Science–Policy Interface for Global Change, City Cancer Challenge (C/Can) and the University of Melbourne partnered to co-develop and test a structured, adaptable model for establishing and sustaining Knowledge Hubs for cancer care in LMICs. These institutionally anchored efforts aim to enable interdisciplinary collaboration and support research-to-policy translation through locally driven pathways.

Objectives

This study aimed to:

• Synthesise global evidence on Knowledge Hub models supporting cancer and non-communicable disease (NCD) research and evidence translation;
• Develop, pilot across five LMICs, and disseminate an implementation science–informed readiness assessment instrument—the City Cancer Research Capacity and Implementation Readiness  (CASTLE) tool;
• Examine the implications of integrating Knowledge Hub development within a city-level technical cooperation model.

Methods

A mixed-methods design was employed. First, a systematic review (PROSPERO CRD42024584992) examined the design, governance, and evaluation characteristics of Knowledge Hub models across cancer and NCD programmes, with a focus on LMICs. Second, the CASTLE tool was co-developed and pilot-tested with partner institutions in Nigeria, Kenya, Chile, Argentina, and Cambodia using structured surveys, institutional assessments, and facilitated multi-stakeholder workshops. Third, findings were triangulated with implementation data from C/Can’s City Engagement Process Framework, integrating targeted research capacity development and embedded research components within city-level technical cooperation programmes.

Results

The systematic review identified four core features and enablers Knowledge Hubs: (1) clear governance and institutional anchoring; (2) cross-sectoral collaboration; (3) structured capacity-building and mentorship mechanisms; and (4) explicit pathways for evidence-to-policy translation. Notably, most identified models lacked robust evaluation frameworks and sustainable financing mechanisms, limiting scalability.

Pilot testing demonstrated high feasibility and acceptability of the CASTLE tool across the five LMIC settings. Recurrent institutional readiness gaps included limited data infrastructure, fragmented research governance processes, and insufficient protected time for researchers. Enabling conditions included co-mentorship models, community-of-practice development, and alignment with national cancer control priorities.

Triangulation with embedded city-level implementation data indicated that thematic, inter- and intra-city Knowledge Hubs—particularly in quality improvement and gender-responsive health systems—support coordinated evidence generation, multicentre collaboration, and capacity development, while systematically incorporating patient and civil society perspectives, thereby strengthening research–policy linkages and institutional coordination.

Conclusions 

The CASTLE tool provides a novel, implementation science–informed approach to systematically assess institutional readiness for locally led cancer research and support the translation of evidence into policy action in LMICs. The integration of global Knowledge Hub evidence synthesis and embedded implementation research offers a practical and scalable model for cities and national cancer programmes. These findings support the integration of Knowledge Hub models within health system strengthening initiatives. They underscore the importance of sustained partnerships and locally driven research ecosystems in advancing equitable, research-informed cancer policy and system reform.

Figure 1. Exemplary CASTLE outputs from Concepción, Chile. The left panel displays mean institutional research capacity scores across CASTLE domains (scale 1–5), with bar length proportional to the mean score. Capacity means below 3 are shown in red, and