In a recent report, the Association of the British Pharmaceutical Industry (ABPI) examines the UK’s “translational readiness gap”—the persistent failure to move cutting-edge laboratory models into the actual creation of medicines to treat patients.
The Attrition Problem and the Translational Gap
Modern drug development is a high risk endeavor. Approximately 90% of drug candidates that enter clinical trials fail to reach the market. What can be done about this? Can new technology–including AI–help increase the success rate?
UK academics have pioneered “human-relevant” models—technologies like organ-on-a-chip, 3D bioprinting, and advanced computer simulations (in silico models) that aim to mimic human systems more accurately. In the US, NIH announced launch of the Standardized Organoid Modeling Center at the Frederick National Laboratory for Cancer Research. In Europe, imSAVAR project aims to deliver human-relevant models to improve the efficacy and safety testing of immunomodulatory therapies. However, the ABPI identifies a massive “translational readiness gap” from these initiatives. While these models are scientifically brilliant, are they able to be translated into real-world drug discoveries? An ABPI report identifies key gaps and limitations that impede the ability to leverage these scientific discoveries and turn them into treatments that help real-world patients. Current gaps and limitations identified by ABPI include:
Principal Gaps and Limitations
1. Materials and biological inputs
Cell sourcing and characterisation. Limited availability of well characterised, quality controlled cell sources; choices often driven by availability rather than fitness for purpose.Stem cell expertise and maturity. Technical challenges in differentiation/maturation of induced pluripotent stem cell (iPSC) derived cells (immaturity undermines physiological relevance).Standardisation of patient derived samples. Variable collection, processing and storage practices reduce reproducibility and comparability. Consent processes and commercial use permissions are inconsistently applied.Linked clinical data. Fragmented ability to connect biological samples with associated clinical data across biobanks limits stratification and validation.
2. Scientific and technical limitations
Lack of standardisation. No sector wide ‘gold standards’ or harmonised protocols for many in vitro systems; inconsistent endpoints and assay conditions hinder comparison.Physiological complexity. Key biological features remain difficult to model reliably – notably functional vasculature and robust blood brain barrier systems.Tissue microenvironment and multi organ modelling. Challenges integrating extracellular matrices, mechanical cues and multi tissue interactions constrain systemic modelling of medicine effects.Confidence and regulatory acceptance. Limited comparative data versus pre-clinical/clinical datasets reduces confidence among industry and regulators.Regulatory landscape. International regulators show growing interest, but regulatory qualification and acceptance for in vitro approaches remain limited and require clear validation and context of use definitions.
3. Infrastructure, funding and skills
Translational pull-through. There is limited support for models developed in academia to be further developed for their use in pharmaceutical research.Biobank resourcing. High costs of storing samples and managing linked data are bottlenecks for national biobanking capacity.Fragmentation and lack of connectivity. Expertise is geographically and institutionally siloed, limiting coordinated development and standardisation.Skills gap. There is inadequate distributed training in stem cell techniques, bioengineering, multi modal assay development and data analytics needed to translate complex models.
Closing the Gap
To address this issue, the 2025 Life Sciences Sector Plan (LSSP) announced the establishment of a pre-clinical translational models hub, bringing together cutting-edge human disease modelling capabilities and essential data. ABPI argues that pharmaceutical firms should be involved earlier on in these initiatives to increase the likelihood that science moves to medicines more rapidly. They write: “Realising the full potential of these models for pharmaceutical R&D will require targeted investment in infrastructure and a coordinated national strategy to support model validation, translation and cross-sector collaboration.”
ABPI argues that by involving pharmaceutical companies early in the design phase, the UK can ensure that the next generation of models is built with the standardization, scalability, and regulatory data required to finally lower the 90% attrition rate and bring safer medicines to patients faster.
Bringing academia and industry together to address the identified gaps has the potential to be highly promising. The key would be figuring out the specifics to make this promise a reality.
