Clinical trials are facing new and evolving challenges as costs increase and approval timelines remain lengthy. Modernisation of processes and utilisation of novel technologies can help overcome such hurdles and ensure drug development meets the needs of patients.
In November 2025, experts gathered at the Clinical Trials in Oncology and Clinical Trials in Rare Diseases conferences in Munich, Germany to discuss how key challenges can be overcome in the space.
Go deeper with GlobalData
Discover B2B Marketing That Performs
Combine business intelligence and editorial excellence to reach engaged professionals across 36 leading media platforms.
Some key themes that arose were the utilisation of technologies such as wearable biosensors and the prioritisation of patient-centric approaches. Developing artificial intelligence (AI) tools to enhance efficiency of clinical trials was another hot topic.
AI-augmented clinical trial design
AI can be implemented at multiple stages of a clinical trial, from protocol design through to data collection and analysis, and is changing the way clinical trials are run, says Eslam Katab, global clinical development manager at Sandoz.
AI can be leveraged to overcome some of the challenges faced in clinical trials, which are slower, more costly and riskier than ever, said Matteo Talotta, Biorce Europe’s clinical solutions director. Companies are now offering tailored AI platforms to support clinical trial planning and execution. Aika, Biorce’s AI-native platform is marketed as a tool that can integrate protocol design, feasibility assessment and regulatory planning to speed up timelines and cut costs.
Katab highlighted the open-source Python library PyTrial, which provides machine-learning frameworks to support clinical trial modelling and design. This tool supports patient-level predictions related to response, trial completion and toxicity, and can be used to explore trial-level outcome forecasting, he said.
US Tariffs are shifting - will you react or anticipate?
Don’t let policy changes catch you off guard. Stay proactive with real-time data and expert analysis.
By GlobalDataParticipant selection and screening is another area that can be enhanced by AI, Katab said. He cited the open-source tool TrialMatchAI as an example, which provides end-to-end patient matching. The tool processes structured and unstructured patient data, and reviews that data against the relevant eligibility criteria to determine the suitability of a potential participant. While completing these checks manually may take days or weeks, TrialMatchAI can pre-screen 1,000 patients in just hours, added Katab.
Speakers at the conference highlighted the importance of targeted participant recruitment for ensuring the success of a trial, noting that the selection of the wrong inclusion and exclusion criteria can often lead to issues. Traditionally, the approach within the industry has been to aim for approval in the broadest indication possible, to maximise market potential. However, this may not always be the most effective method.
As the standard of care evolves in oncology, the thresholds for regulatory approval and commercial success rises, said Guido Wuerth, head of corporate development, radiopharmaceuticals at Sweden-based Affibody. This means it is increasingly important to select patient populations with the greatest likelihood of responding. This is where utilisation of appropriate biomarkers in inclusion criteria can enrich patient populations and increase likelihood of trial success, added Wuerth.
Concurrently, it’s important to ensure criteria are not overly restrictive in terms of previous treatment and comorbidities, added Dariusz Adamczewski, managing director at nonprofit Children’s Tumor Foundation Europe.
Cultivating patient-centric approaches
Speakers throughout the conferences highlighted the importance of early collaboration with patients and advocacy groups.
Patient engagement during clinical trial planning is invaluable, particularly for rare diseases, said Ivo Timmermans, co-chief executive officer, Pleco Therapeutics. Gaining patient feedback on protocol designs and understanding motivations for participating in clinical trials as well as potential challenges is essential, he said.
Strategies for reducing patient burden in clinical studies were highlighted as a key focal point for improving likelihood of trial success. Identifying potentially burdensome elements of the trial and reducing in-person visits where possible can be beneficial, said Adamczewski. This is an area where advancements in remote monitoring of patients can have significant impact.
Digitalisation of drug development
The use of electronic patient reported outcomes (ePROs) is growing in popularity and patient acceptance, said Katab. Their implementation can reduce the need for in-person visits, thereby easing patient burden and potentially improving retention, he added.
Katab highlighted the observational CAPRI trial (NCT02828462) as a demonstration of how ePROs can also improve patient outcomes. This study found that additional digital remote monitoring of patients who were receiving oral cancer treatments reduced days of hospitalisation and treatment-related toxicities.
Further, wearable biosensors can enhance data collection by tracking a patient’s activity, function, and physiological parameters such as heart rate variability and peripheral oxygen saturation (SpO₂), said Katab.
Latin American (LATAM) countries have traditionally been associated with complex and unpredictable regulatory timelines and poor connectivity between cities, posing challenges for drug development, said Paola Gaglio, clinical trial manager lead LATAM at Boehringer Ingelheim. However, the region is bidding to increase its influence in the clinical trial space through digitalisation strategies and regulatory reform, she added. Among these, the implementation of electronic submissions for regulatory approval has helped to streamline processes, explained Gaglio.
Alternative clinical trial designs and AI enhancements
Researchers are increasingly looking to leverage historical real-world data as a comparator in clinical trials. Here, AI could enhance the construction of synthetic control arms to ensure the matching of key prognostic factors and allow for application of propensity weighting.
Clinical trials for rare diseases face unique challenges like the need to recruit within limited patient populations, often making placebo-controlled trials difficult to conduct. Hence, this is an area where synthetic control arms can be particularly valuable.
Acceptance of single-arm trials that use real-world data is increasing among regulators, said Channa Debruyne, global clinical development lead, late-stage and LCM oncology, at Servier. As an example, Bavencio (avelumab), co-developed by Merck KGaA and Pfizer, was granted accelerated approval by the FDA and EMA for treating metastatic Merkel cell carcinoma based on data from a single-arm trial in 2017.
The quality of the data being fed into synthetic control arms is crucial, explained Debruyne. Therefore, standardised and robust methods of data collection and storage are required to ensure high quality real-world datasets. To generate sufficient datasets, multiple registries with different parameters will need to be harmonised, which will be one of the biggest challenges for effective implementation, Debruyne added. This will require close collaboration between key stakeholders, and strengthening of partnerships between clinicians, researchers, non-profit organisations, regulatory bodies, patient associations and industry, she said.
Another major barrier is access to data needed to create synthetic control arms, said Clara Cali Mella, data strategy lead at Bayer. She said that “the data is there,” but some challenges lie in navigating complex data protection laws. Further, Mella called on more countries to embrace the sharing of anonymised patient data, providing Finland as an example, which she said could be a game changer for rare diseases.
