Dr Gen Li

Phesi’s Dr Gen Li on unlocking the power of data in oncology

Dr Gen Li, founder and president of Phesi, a provider of AI-powered clinical development analytics, products, and solutions, discusses ASCO, digital patient profiles, and how digital twins and trial arms can help optimise trial design.

How can sponsors ensure that clinical development in oncology delivers on its promise to become more data led?

Sponsors should increase their focus on integrating external data sets from the industry to add to the existing body of internal knowledge to better inform trial design. This includes data from historical trials, ongoing clinical trials, electronic patient records, and epidemiological studies, alongside newer sources such as wearable and telemedicine devices, as well as health applications. But more data does not automatically equal more knowledge. Analytics are essential to extract the information needed to actually design oncology trials around the insights held within.

Moving away from long-held trial planning and execution practices to adopt a data-led approach will require buy-in at a senior level to really maximise the benefits. New skills and tools will be needed, and key ‘actors’ in the development space – from regulators to investigators and CROs, and patients – will need to collaborate more closely. In oncology, data-led trials will ultimately accelerate the clinical development pathway, reduce patient burden, and minimise amendments – or even eliminate them entirely.


Patient centricity has been repeatedly talked about as a priority in clinical development. What measures can be taken to improve patient centricity in cancer trials?

Data holds much promise in alleviating the burden on patients across all clinical trials. Whether by reducing the need for comparator arms, shortening cycle times, or reducing amendments – data can be used to make the trial experience and execution better. In cancer trials specifically, insights from digital patient data help sponsors make sure the right patients are being recruited for a trial and assigned the most appropriate potential therapy. Digital data at patient and cohort levels can be used to identify the attributes of the patients a therapy is supposed to treat before the trial even starts, to improve recruitment and enrolment success.

By leveraging historical and ongoing oncology trial data at scale, trials can be simulated and modelled accurately. This allows safety and efficacy outcomes to be predicted for many different treatments before they are administered to patients. Reducing the need for patients to take inferior comparator or placebo medicines and managing risk in this way is key, especially in oncology, for which therapies are becoming increasingly targeted.

With data we have obtained from patients in the past, it is also possible to help us to interpret results from single-arm trials more accurately. Among other benefits, this will help us to avoid predictably futile clinical trials.

Phesi recently released a catalogue of digital patient profiles. What is a digital patient profile, and how can sponsors use them?

You have spoken previously about the “digital trial arm” and the use of “digital twins” in clinical development. What are the latest developments in these areas?

In clinical development, data enables researchers to simulate and predict different outcomes in a clinical trial with greater certainty. Data is collated from similar or identical trials using the same agent, with real-world patient data, to accurately model placebo/comparator outcomes and construct digital twins and digital trial arms.

A digital twin can be used to detect early signals of trial outcomes, protect patient safety, facilitate regulatory dialogue, and enhance submissions. When a digital twin is planned and implemented in alignment with regulatory authorities, it becomes a digital trial arm, or external control arm, as the FDA calls them, reducing or eliminating the need for a placebo or a comparator arm. Thanks to oncology being such a well-researched field, there is a wealth of available data that can be employed to produce digital twins and digital trial arms for many indications.

With small-group or single-arm cancer trials becoming more common in the industry, what are some of the pitfalls sponsors should look out for?

Greater understanding of the genetic drivers of cancer is leading to more targeted therapies designed for smaller patient sub-populations, with oncology research increasingly focusing on cancers with specific genetic markers. This precision inevitably also leads to fewer and smaller suitable patient populations. With a growing number of single-arm studies and small-group trials, it is critical to have a watertight recruiting process and trial design.

Single-arm trials bring an inherent degree of uncertainty that can cause problems in trial design, patient recruitment, and data interpretation – all of which can delay patient access. Small single-arm trials are not always representative of the real-world patient population, which brings complications around interpretation of the results, delays approval, and causes difficulty in obtaining reimbursement. Clinical trial designs that are not informed by data, therefore, run the risk of developing drugs that are not efficacious for the target patient population.

You have observed that around a fifth of all trials fail. How can sponsors mitigate the risk of failure?

All too often, trials are being initiated without a solid data foundation – with many sponsors still relying on a ‘gut feel’. By using data to profile the targeted patient population and implementing predictive analytics to determine projected outcomes, protocol and trial design can be optimised and feasibility studies can be carried out.

This reduces the number of patient participants needed for the trial and minimises costly delays around recruitment – and it ensures data requirements for regulatory submissions can be met. By predicting the pathway of a clinical trial before embarking on it, chances of success are high, unnecessary delays can be avoided, costs can be reduced, and failures mitigated.

In a data-led approach, it has repeatedly proven that we can prevent a failing trial from starting.

What kind of industry insights are you hoping to see at ASCO this year? What outcomes are you hoping to achieve from the event?

ASCO is always an excellent opportunity to connect with colleagues working in the oncology space, hear about breakthrough advances in medical and scientific research, and explore the key challenges facing the industry. By getting the very latest information on innovations in biology and chemistry – and advising on the best practices for trial design and execution for these potential therapies – we can support what we hope will be tremendous advances in oncology treatments and patient outcomes.

How does Phesi see data analytics shaping the future of clinical development?

The future is here and now. As digital tools and practices become a mainstay in the life sciences space, the next one to three years hold the promise of significant breakthroughs. Digital analysis, empowered by artificial intelligence, is going to transform the planning and execution of projects – and in clinical development, this will see a move away from dated, gut-feel trial protocols towards data-driven, patient-centric clinical trials.

Smarter trials, faster cures.

About the interviewee

Dr Gen Li founded Phesi in 2007 with the aim of revolutionising the clinical trials industry. Prior to founding Phesi in 2007, Dr Li was head of productivity for Pfizer Worldwide Clinical Development, a position he assumed following Pfizer’s acquisition of Pharmacia, where he delivered the first implementation of productivity measurement for clinical development.

While at Pharmacia and Pfizer, Dr Li significantly contributed to the Centre for Medicines Research (CMR) International database for pharmaceutical R&D performance, assuring the collection of key clinical trial parameters as representative of the critical path for delivery. He was also instrumental in creating the KMR productivity mode.

Previously, he earned his PhD in Biochemistry from Beijing University, and an MBA from the Johnson Graduate School of Management at Cornell University.

About Phesi

Phesi is a data-driven provider of AI-powered clinical development analytics, products, and solutions to the biopharmaceutical industry. The company’s integrated offerings cover the entire clinical development process – from development planning and indication assessment to protocol evaluation, country and site selection, and trial implementation management.

Phesi has the world’s largest real-time clinical development database; delivering patient-centric data science that enables biopharmaceutical companies to predict and optimise clinical development outcomes in any indication. Its database consists of 485,000 completed clinical trials, 604,000 completed research projects, >4.2 million physicians and >600,000 investigator sites worldwide.

Phesi delivers data, insights, and answers, enabling smarter trials and faster cures. For more information, please visit Phesi.com.

Don’t miss your complimentary subscription to Deep Dive and our newsletter

Sign up



Your name

Your e-mail

Name receiver

E-mail address receiver

Your message