The world’s most studied diseases: four lessons from 2025

For the fifth consecutive year, breast cancer across all subtypes was the most studied disease in the world, with 1,667 recruiting trials.

Breast cancer is the top killer among women with cancer; it continues to dominate the list because researchers have a far deeper understanding of its biomarkers and how they can be used in clinical research than other indications.

However, despite this knowledge, there are still many unmet needs to be addressed, such as in triple-negative breast cancer. Triple-negative breast cancer is defined by its lack of HER2 and oestrogen and progesterone receptors, but in order to make significant progress researchers will need to focus on finding what receptors and proteins are actually present in the subtype. Nonetheless, the sustained research investment in this devastating disease is encouraging.

The remaining top five for 2025 were solid tumours, stroke, prostate cancer and non-small cell lung cancer (NSCLC). This is the same list in the same order as last year – which is partly down to COVID-19 dropping out of the list in 2024 (table 1). This suggests the clinical development sector is slowly returning to some degree of stability after the pandemic.

Obesity came in 6th place in 2024’s list and remained just outside the top five in 2025. This is largely down to soaring interest in GLP-1 drugs. An analysis of 583 recruiting trials in September 2025 showed that more than 100 diseases are now being investigated in connection with GLP-1s as a therapeutic, reflecting growing interest in obesity as a comorbidity in a wide range of diseases. Increasing GLP-1 use will also influence clinical development in other fields, as obesity is an important comorbidity and weight loss may alter study parameters, such as dosing and endpoints.

The pandemic was the most recent major disruptor for the clinical trials sector and is still having some lingering effects. But sponsors should also be looking ahead to what the next disruptor might be. The broader adoption of GLP-1s is one possibility – given it could reshape the clinical development landscape as the focus shifts towards prevention and treating clusters of related disease. This uptick in GLP-1 use indicates that obesity is the disease area most likely to enter the top five within the next one to two years.

Trial attrition – specifically, the number of trials that are terminated in phase 2 – is an important metric that can highlight a slowing down of the rate that new therapies reach the market and the rising development costs for the industry. Often, phase 2 trials are terminated due to underwhelming phase 1 data or trial design focusing on patient populations that don’t exist in large enough numbers to recruit.

This year’s analysis shows an encouraging fall in phase 2 attrition rates to a four-year low of 26%. But this is still higher than pre-pandemic levels. In 2024, 31% of phase 2 trials were terminated, up from 29% in 2023 and around 20% before COVID-19 (figure 1). (It is important to note that the rising phase 2 attrition rate during COVID was not caused by failing COVID-related development programmes.)

Despite the positive direction of travel, a quarter of trials being terminated in phase 2 is still an unacceptably high figure and will inevitably have a knock-on effect on phase 3 trials, plus a significant negative impact on ROI across the biopharmaceutical industry.

The report’s country-level analysis shows that the US continues to host both the largest number of recruiting clinical trial investigator sites overall and the largest number for each of the top five most studied diseases. China ranks second for all five most studied diseases – except prostate cancer, where Canada takes second place. China also recorded the strongest growth in investigator site numbers between 2023 and 2025, increasing by 51% compared with 42% for the US (figure 2). France, Italy and Spain complete the top five countries for recruiting investigator sites.

While the US is a competitive place for clinical trial investigator sites, it is also uneven, with some of the most well-known names facing incredibly high demand. In China, however, patients are easily recruited into trials from hospitals – there is no need to visit a primary healthcare provider first, as in the US. Investigators therefore have access to a very large number of those patients.

In 2025, sponsors remained under pressure from macroeconomic constraints, from macroeconomic constraints, pricing challenges and increased activity in regions such as China. We are also seeing the effect of renewed emphasis from regulators, including the FDA, on country-specific representation. Well-known and high-profile investigator sites are quickly saturated as sponsors seek them out.

The increase in competition, regulatory changes and increasing acceptance of digital patient data means it is vital for sponsors to be led by insights from big data and AI, rather than instinct. Considerable volumes of contextualised and real-world data now exist to power clinical data analytics that overcome these hurdles. Breast cancer, given our deep understanding of biomarkers and the subsequent stratification of clinical trials, has been a clear beneficiary of such an approach.

Sponsors should become data-focused to ensure that every aspect of patient profiling and trial design, including the most suitable countries and investigator sites, are optimised and proven to reduce the cost, patient burden and complexity of trials. For example, there is considerable opportunity to use data science to identify investigator sites elsewhere that may have a shorter clinical trial and enrolment history, but have capacity and have delivered high-quality data in previous trials. Lessons learnt from patients, as shown by a digital patient profile, have consistently proven to guide researchers to optimise trial design and shorten cycle time by working with the right investigator sites.

Further, sponsors can use digital twin analysis – a representation of typical patients in a specific population generated from a vast body of real-world patient data from identical or similar trials, allowing researchers to simulate, model and predict safety and efficacy outcomes as a historic control arm. Other use cases include external control arms to reduce the number of patients undergoing testing in trials.
As we move into a new year, we’re hopeful the industry will increasingly embrace a data-science-led approach to clinical trials and that this will be reflected in 2026’s Most Studied Diseases report.