Clinical Trial Design Evolution: Key Trends and the Integral Role of Bioinformatics Solutions
July 10, 2025
Lakshmi Santhosh Maithel, Marie-Ange Kouassi
In recent years, clinical trial design has evolved significantly, with a rising focus on applying precision medicine to better serve specific patient groups by assessing therapy quality, efficacy, and safety. This shift is driven by rapid technological progress as well as the growing need for improved efficiency and patient-centricity in therapy research and development. Drug development organizations are now leveraging biomarker-driven studies to identify responsive patients and potential side effects early in the process, accelerating decision-making and optimizing resources such as time and costs. Stratifying patients based on omics profiles enables a more personalized approach, ensuring treatments are tailored to those most likely to benefit.
Oncology studies have become increasingly driven by biomarkers rather than by the tissue where the tumor was found. This demonstrates an understanding that cancers which share common mutations or other molecular characteristics may be responsive to the same treatments even if identified in different parts of the body. It also represents the precision medicine trend towards more targeted and eventually individualized treatments. Based on this approach, Merck successfully developed the first tissue-agnostic therapy, Keytruda (Pembrolizumab), an immunotherapy drug, in 2017. More recently, Bayer’s Vitrakvi (Larotrectinib) targets NTRK gene fusions, a known oncogenic driver mutation, in tumors across tissues.
To meet this shift in scientific conceptualization of disease heterogeneity, clinical researchers have developed the master protocol trial design which includes basket trials, umbrella, and platform trials. Master protocols allow multiple sub-groups defined by molecular profile to run simultaneously in one study. The standard fixed-sample randomized control trial does not account for such patient stratification and is also much more expensive than targeted trials. But while novel master protocols are promising in accelerating therapy development, they also present difficulties related to data management as they generate increasing volumes of data, which may also vary extensively in format. Drug development organizations need a data infrastructure that enables scientists of different technical backgrounds to access suitable bioinformatics solutions, collaborate effectively, and derive answers to their scientific questions promptly — while complying with regulatory requirements. Without the right data infrastructure, these challenges can hinder the adoption of innovative trial designs for precision medicine.
Oncology and Biomarker-Driven Trials
A 2015 publication in a journal presenting practical reviews on the latest developments in cancer research, highlights how genome sequencing has enabled a shift towards the master protocol design and also mentions five major studies pioneering this concept. Three were biomarker driven trials including the NCI-Match program specifically for patients with solid tumors and refractory lymphoma of any origin stratified by biomarkers, as well as the I-SPY 2 trial which was not biomarker driven but employed the same trial design. By 2019, a landscape analysis identified 83 master protocols undertaken over the last five years of which the majority (76) were in oncology.
Another publication on the evolution of trial design described “an exponential increase in publications in this domain during the last few years in both planned and conducted trials”. The same publication described the various categories of master protocols including the basket, umbrella, and platform design. During a basket trial, a single investigational treatment or combination therapy is used to target several diseases with a shared biomarker. Umbrella trials are adopted when multiple treatments are applied to different patient subgroups under the same disease indication. Platform trials also assess responses to multiple treatments but allows for changes in study design according to the discovery of new biomarkers or the success of one treatment, essentially making it an adaptable version of umbrella trials.
Though more biopharma companies are adopting the master protocol design, due to the advantages in cost effectiveness and de-risking clinical candidates, it was actually a leading institute for oncology that piloted molecular characterization in clinical trials. The early IMPACT study in 2007 by University of Texas MD Anderson Cancer Center focused on improving outcomes for solid tumors through targeting treatment based on genetic variants. This was followed by the IMPACT 2 study, exploring molecular profiling for metastatic cancers, and then by an international trial through the World Innovation Network. These programs made the case for master protocols based on patient stratification using biomarkers. More recently, novel trial designs are being applied even outside oncology by researchers in the UK as seen in a preprint from September 2020 which showed potential to improve trial efficiency in immune-mediated inflammatory diseases.
The use of biomarkers in oncology clinical trials enables a more tailored approach to treatment and is being used more widely due to its significant benefits. Rather than a one-size-fits-all approach, precision oncology considers the molecular and genetic characteristics of a patient’s tumor to determine the most effective treatment. This leads to improved patient healthcare outcomes in patients and also enhances their safety by minimizing unnecessary interventions.
Adaptive Clinical Trial Designs
While traditional clinical trial designs have fixed sample sizes and only one or two arms may be limited to efficiently address emerging research questions, adaptive clinical trials allow modifications to be made during the trial based on accumulating data. Such changes or decisions could include sample size re-estimation, dose adjustments, early termination, or randomization. Adaptive designs have been reported to have multiple advantages over traditional, non-adaptive designs such as improving statistical efficiency, improving the understanding of therapy effects, and being more acceptable to stakeholders as it offers increased flexibility for successful trial completion. In addition, adaptive trial designs address ethical considerations regarding safety (e.g. stopping a trial early if an intervention shows safety concerns or treatment effects). Adaptive trial designs could help companies optimize clinical trials, reduce sample size burden, and improve the likelihood of success. For patients, adaptive trials could mean faster access to innovative and safe therapeutics.
Decentralized Clinical Trials (DCTs)
Although innovative options currently exist in terms of clinical trials, there is still room for improvement. For example, expanding clinical trial accessibility to a broader range of subjects remains a challenge. There is also a need to improve diversity so that different populations are represented and their response to treatments in development is determined. Otherwise, the data outputs from clinical trials may be limited in strength and lack generalizability. This could pose a safety risk by causing unexpected effects on different populations. To overcome this as well as other challenges such as patient recruitment, engagement, and retention, decentralized clinical trials were developed where a proportion of — or all — trial activities are conducted away from traditional clinical sites. These trials may be either remote or in-person (or a hybrid between the two) and may involve home visits, digital health technologies (DHTs), and mobile clinics. Virtual clinical trials are a subset of DCTs and are conducted fully remote from enrollment, monitoring, to data collection through the use of DHTs and wearable devices. By expanding clinical trials beyond onsite visits, virtual clinical trials remove geographic barriers and increase patient access. This increased accessibility makes participation more convenient, leading to higher engagement and improved retention rates.
Artificial Intelligence (AI) and Machine Learning (ML) in Clinical Trial Design
In drug development, AI — the use of machines to simulate human behavior such as learning, reasoning, and problem-solving — offers opportunities to enhance the accuracy and efficiency of drug design and laboratory result interpretation. ML, a subfield of AI, focuses on training algorithms to learn from data, identify patterns, and make predictions. AI/ML accelerates decision-making, and the discovery of novel insights, presenting opportunities to streamline clinical trials. Applications of this technology include analyzing vast amounts of data to identify eligible patients — reducing recruitment timelines and improving diversity. For clinical trial optimization, AI and ML can assess previous trials to identify potential risks, allowing scientists to mitigate risks and avoid costly failures. AI/ML also facilitates a more patient-centric approach, improving the understanding of specific patient groups for targeted clinical trials that enhance efficacy and reduce adverse effects. However, scientists must remain cautious of AI-generated hallucination by ensuring algorithm-driven conclusions are explainable and valid. To prevent biased results and misinterpretations, it is crucial to use comprehensive and complete datasets.
Evolving Regulatory Environment
Regulatory agencies — such as the FDA — are actively updating their guidelines to accommodate new trial designs. For example, the FDA has provided recommendations on the appropriate use of adaptive designs when submitting evidence on the effectiveness and safety of a drug or biologic. For DCTs, the FDA has submitted guidance to help expedite drug development and the body of clinical evidence of new and approved drugs, increase clinical trial diversity, and promote the development of rare disease treatments. For such trials, the FDA also recommends the integration of real-world evidence (RWE) and data from diverse sources (e.g. electronic health records) as this could address gaps in existing data and validate findings.
Addressing Data Management Challenges
While it is encouraging to see the uptake of master protocol trials as an outcome of increasingly personalized medicine, there are also new challenges. A research article on trial design in precision oncology describes how the increased specificity gained from smaller subpopulations also comes with statistical challenges in drawing robust actionable conclusions. The data collected must be managed consistently across studies so that researchers can refer to data from previous trials and learn from the aggregate outcomes. Another publication on precision trials for clinicians highlights biological plausibility or underlying mechanisms, biomarker prevalence when recruiting, and sample size assumptions as factors that will need to be considered during statistical analysis.
Typically, when identifying trends in data, bioinformaticians generate and share visualizations generated on tools such as Posit or Tableau. However, using several software solutions in a non-continuous manner also makes it difficult to maintain a full chain of custody. In addition, high dimensional and multimodal data needs to be harmonized across sources, whether from internal departments or external CRO partners. Since clinical trial data is often submitted to regulatory authorities, scientists and bioinformaticians require comprehensive documentation of all activities and testing of relevant systems to ensure they operate within a GxP-validated environment.
Conclusion
As drug developers strive to make clinical development more efficient while maximizing resources, staying attuned to advancements in clinical trial innovation is crucial. Today, key clinical trial trends —such as master protocol, adaptive, and decentralized trials — combined with AI/ML and RWE are making trials faster, smarter, and more inclusive. However, biopharmaceutical and biotech companies can only progress at the pace of innovation with a cutting-edge analytics system embedded in their data infrastructure that translates disparate R&D data into actionable insights. Genedata Profiler is a flexible, GxP-validation-ready data integration and analytics platform that empowers cross-functional stakeholders developing precision medicines to collaboratively and accurately assess therapy quality, efficacy, and safety. Designed to support translational research, the platform fuels biomarker research by integrating molecular data with patient outcomes for smarter trial design positioning companies for clinical success.