Organoids have the potential to revolutionize drug research and medicine. These fascinating, miniature versions of organs grown in a lab have countless applications for health care, reducing failure rates for drug screening, enabling more effective disease modeling, and personalizing medicine. In addition, because they closely mimic real human organs and tissue, organoids may be able to reduce the amount of laboratory animals needed. No wonder laboratories and research initiatives across the world focus their attention on organoids. Oncode Accelerator is no exception: the Organoids Platform is an integral part of our innovative infrastructure.

This question comes together in so-called adaptive clinical trial designs, where the Oncode
Accelerator Patient Cohorts platform and the Organoids platform collaborate. This provides
exciting new opportunities that both Helena Viñas Gaza, program manager of the Organoid
Platform at the Prinses Máxima Centrum, and Peter van de Ven, clinical trial design expert at the UMC Utrecht, are keen to explain. We asked them how Oncode Accelerator will use and integrate the Patient Cohorts and Organoids platform.

Peter: ‘The main difference between adaptive clinical trials and traditional trial designs is that adaptive trials use the data observed in the trial to make modifications to the study based on predefined decision rules. Their flexibility and the ability to learn during the trial mean that adaptive trials are often more efficient than traditional trials. Types of adaptations include re-estimation of the required sample size, early stopping of futile treatment arms, and adaptive randomization. Adaptive trials can help address more complex research questions and their costs may be lower, and duration may be shorter compared to traditional trials. Adaptive trials that reduce participant exposure to ineffective treatments may also be regarded more ethical. Adaptive trials can help address more complex research questions, their costs may be lower, and duration may be shorter compared to traditional trials. In Oncode Accelerator, integrating information from patient cohorts, organoids and artificial intelligence in early-phase adaptive trial designs is a major goal to help improve the earlier clinical stages of cancer therapy development.’

Helena: ‘Organoids are three-dimensional, miniature versions of organs or tissues grown in vitro from stem cells. They closely mimic the structure, function, and micro-environment of real organs, providing a more accurate model for studying human biology and diseases compared to traditional two-dimensional cell cultures. While cell lines can also be easily used in high- throughput experiments, they fail to represent the cellular heterogeneity of a tumour. This complicates the translation of pre-clinical findings in cell lines in vitro towards the clinic. Hopefully, in some years organoids will replace cell lines and become the default go-to model for any pre-clinical cancer candidates, which will help to reduce the failure of drugs later in the process of clinical trials.’

Helena: ‘We are already living in a time where it is possible to grow a mini-tumour derived from patient material that can help tailor treatments and responses, although there’s still room for improvement. If we want to make organoids a reality for all patients, we will have to invest in scale-up technology with the use of robots, since growing, testing, and analysing organoids is still time-consuming. There are still some tumour types that are difficult to grow as organoids, for example blood cancers or prostate cancer. We need to dig deeper into their molecular background to understand which compounds and growth factors are essential for their culturing. Another limitation is that the drug administration in patients is quite different from the one in organoids, so we are still not able to mimic exactly the same drug exposure.’

Peter: ‘Organoids are mainly used in preclinical research, with so far only a few applications in, exclusively non-adaptive, clinical trials. However, in the future, organoids could play a significant role in both adaptive as well as more traditional clinical trials by serving as biomarkers for patient stratification and selection and improving success rates by identifying non-responders to standard treatments and likely-responders to novel agents. Successful and timely production of patient-derived organoids and identification and validation of their most informative read-outs are prerequisites for their use in clinical trials.’

Helena: ‘As of now, organoids are not yet integrated into adaptive clinical trial designs. Although some regulatory organizations are opening the door to its usage in preclinical settings (FDA Modernization Act 2.0, ed.), it’s still not part of the day-to-day trials. The Organoids Platform together with the Regulatory Innovation workstream have a dedicated PhD student aiming to investigate in which circumstances in vitro studies by themselves may already deliver sufficient evidence for the use of organoid systems to support pharmacodynamics and efficacy.

Peter: ‘We still have some challenges ahead before we are close to a meaningful integration. A major issue currently is that there is large heterogeneity in response to treatments in the
unselected population in which cancer trials are generally performed. A recent review showed organoid response to be predictive for therapy success. Upfront prediction of therapy success for an individual will allow trials to be performed in more selected populations that are more likely to respond, increasing success rates of the studies. Adaptive designs may further help to increase the efficiency of the trials through efficient dose selection, timely stopping for futility or proven efficacy and optimizing the read-out of the organoid to be used for selecting the responsive group of patients within the trial.’

Helena: ‘We are putting a lot of effort in building a well-annotated organoid biobank linked with patient data that will help researchers and industry to gain access to the best models for their experiments. We hope to build a more representative platform, in close collaboration with Patient Cohorts, for easier, cheaper and faster target identification and drug screening.’

Peter: ‘Several tasks in our work package aim to improve clinical trial methodology to allow
incorporation of preclinical information from organoids and artificial intelligence in the design and analysis of early-phase adaptive trials. We are confident that we will learn along the way how we will do this exactly.’

Peter: ‘The strength is in combining the data available in the cohorts and preclinical results from organoids to optimize the design, and including adaptive elements to keep on learning during the trial itself. Our ultimate objective is to obtain evidence on how to value, interpret and optimally use patient-derived organoids in clinical practice. Cohorts and biobanking of organoids can also be used to, for instance, develop genetic tests by building a classifier based on genetic data to distinguish patients whose derived organoids predict and those whose organoids do not predict response to a novel agent.’

Peter: ‘Organoids have the potential to help make better decisions before starting a trial, for
example regarding the dose and population, and also during a trial, for example with patient selection and possibly adaptive organoid read-out. As more and more data and information tend to become available, this should be used to design innovative early-phase studies to learn as much as possible from this approach. Early health technology and ethical considerations are important as well. All this expertise is combined in the Platforms and Workstreams of Oncode Accelerator.’

Helena: ‘There is no such thing as the perfect model, and we are never going to be able to

replicate fully what happens inside the body in vitro. Having said that, every day we get closer to more representative and better models. When the appropriate ethical and legislative regulations are in place, we will be able to use these models for better and more personalised healthcare.’