Together, they will discuss the challenges, innovations, and future opportunities shaping the next generation of oncology clinical trials.
1. Early-phase oncology trials are no longer limited to safety endpoints. From your experience, how has their role evolved in shaping treatment strategies, particularly in immuno-oncology and targeted therapies?
Dr. Austin Duffy: For me, the story here is the improvement in drugs over the past 20 years and their likelihood of efficacy. Since my first introduction to Phase 1 studies, the field has radically shifted from the average trial evaluating new cytotoxics (in solid tumour subtypes refractory to cytotoxics) to more biologically informed, smarter treatments and, as a result of this, an emotional transition (both for the doctor and patient) from desperation/hope to almost an expectation of benefit. When this scientific advancement is paired with the numerous variables leading to very slow and patchy drug availability (either via reimbursement pathways for approved drugs or the highly selective nature of Phase 3 studies), it leads inevitably to increased enthusiasm for earlier phase, nonrandomized studies with more of a focus now on efficacy signal at that earlier phase. It has also become the case whereby Phase 1 trials – especially with respect to the emerging ras inhibitors – are addressing an unmet clinical need as the data emerges faster than the system can keep up with it.
2. Patient-derived xenograft (PDX) models are widely used in preclinical oncology. How effective are they in predicting clinical outcomes, and where do they still fall short?
Dr. Michael J. Wick: PDX are established from clinical cancer samples with the goal of creating actionable models representing the donor patient’s cancer. Studies have confirmed these models more closely mimic the pathology and heterogeneity of cancer versus cell-derived xenografts (CDX) or ex vivo derivatives. PDX developed from patients clinically refractory to chemotherapies and targeted agents often retain this resistance, providing invaluable tools to study mechanisms of drug resistance and represent patient populations at various stages of disease. Patients’ cancer driver variants are rarely not recapitulated in their corresponding model, providing a useful way to identify and drug these targets with novel therapies. While PDX are much more representative of human disease, they do come with limitations. PDX must be tested in immune-deficient hosts, limiting the ability of researchers to account for immune response contributions towards efficacy and since mice are typically utilized in these studies, significant differences in physiology prevent full translatability to human patients.
3. You have led multiple first-in-human and combination studies. What are the biggest challenges in designing early-phase trials involving novel combinations, especially in balancing efficacy signals with patient safety?
Dr. Austin Duffy: The biggest challenge is managing and mitigating risk while at the same time making the protocol manageable for patients. Also, the challenge of communicating that uncertainty to patients without contributing to their distress. There are practical considerations that differentiate studies in terms of how palatable they are to patients, for example, the need for inpatient admissions, biopsies, and the intensity of pharmacokinetic (PK) schedules. But most of these issues are fairly well standardized at this point and there is general consensus among investigators about what is required, or advisable, while also striking an appropriate balance based on the known or anticipated class-associated toxicity. The uncertainty around risk, and the way it is communicated and managed, remains the key challenge and the most important thing to get right and that will always be the case.
4. With increasing reliance on translational science, how can preclinical data be better aligned with clinical trial design to improve early-phase success rates?
Dr. Michael J. Wick: Advances in target discovery and preclinical assays have accelerated an influx of novel therapies through IND, and they are currently being evaluated clinically. While the rationale for many of these agents is scientifically sound, clinical failures are still common. One possible reason for this disparity is the lack of meaningful changes to early drug development methodology. While advances in ex vivo systems and computational analysis have led to some improvements, the use of immortalised cell lines for in vitro and in vivo efficacy testing limits understanding of the limitations of the test agent since they do not represent cancer heterogeneity or treatment profiles of patients that will receive the therapy in clinical trials. Utilising more representative tools like PDX models, especially those established from heavily pretreated patients, earlier in bench assays and in vivo testing, may provide a better understanding of the strengths and limitations of the test agent and improve early-phase success in the clinic.
5. Biomarker-driven trials are central to modern oncology. What limitations do you see in current biomarker strategies, and how can early-phase trials better integrate them for patient selection?
Dr. Austin Duffy: Tissue availability is obviously crucial to guide understanding and patient selection. This is a major practical issue for units in terms of accessing fresh tissue and even archived tissue, given the array of markers required for standard of care these days. We are often sending tissue for immunohistochemistry (IHC) to various labs and countries and there is a limit to the amount available. Getting more tissue is always the best answer but that is not always possible or practical. I am hoping that technological advances in virtual/digital pathology will help. I also hope to see over the coming years incorporation of radiologic tracers matching for example to ADCs as a way of getting around this barrier and circumventing the need for tissue.
6. What are the key challenges in developing and validating preclinical models that truly reflect tumor heterogeneity and treatment response in humans?
Dr. Michael J. Wick: Developing models of human cancer often requires the use of an artificial system; whether a plastic dish for in vitro and ex-vivo assays or an immune-deficient surrogate for xenograft studies, and there are limitations in translating results from these platforms to the clinic. Currently, in vivo modeling provides a physiologically relevant platform and the best tool for clinical translation, provided these models represent the early-phase clinical setting. Creating these models requires starting material from later-stage patients; however, obtaining samples from this patient population is challenging, and only a fraction will successfully grow out and be evaluable. Drug testing in these models also needs to be performed within a certain number of lineage passages to maintain cell heterogeneity and best represent donor patient characteristics, including clinical resistance to prior therapies. Integrating model development with patient care, especially in the late-stage and trial setting, and utilising resulting models at limited passages to better understand disease progression, provides excellent tools for the development of next-generation therapies.
7. Patient recruitment remains a major hurdle. What practical strategies have proven effective in accelerating enrollment while maintaining trial quality and diversity?
Dr. Austin Duffy: At START Dublin, we are in a more enviable situation than other sites regarding accrual. As the only Phase 1 site in a country with poor general access to new drugs, accrual has not been an issue for us. On the contrary, we have more patients being referred to START Dublin than we currently have slots open on trials. We are working diligently and forming new partnerships with pharma companies to bring more trials to patients in Ireland.
In terms of practical strategies, by far the most important is continual engagement with our referring oncologists via emails, letters and text messages to keep them in the loop. There are approximately 80 referring oncologists in our network and I put considerable effort into keeping them updated. But my overall message to sponsors who worry about accrual is to come to Ireland!
8. As preclinical platforms evolve, what innovations (e.g., humanised models, AI integration) do you believe will most significantly impact early-phase oncology development?
Dr. Michael J. Wick: The toolbox for preclinical oncology research has grown significantly in the past decade, including advances in high-throughput screening, ex vivo platforms, gene editing and computational analysis. Integrating these and other established and novel platforms to streamline target discovery, drug design and evaluation with a focus on pharmacokinetics, pharmacodynamics and efficacy will accelerate oncology early drug development and lead to improved therapies reaching the clinic faster and with a better chance of success.
9. Having worked on innovative approaches like combining immune checkpoint inhibitors with loco-regional therapies, what key lessons can be applied to future early-phase trial designs?
Dr. Austin Duffy: Producing innovative combinations – particularly cross-modality – is difficult in the industry-sponsored space where the focus, understandably, is on assessing safety and efficacy of the molecule in development and proceeding as quickly as possible. But once that initial hurdle has been crossed, there are novel spaces that sponsors could target with their drug, despite the very crowded space, by exploring gaps in standard of care (SOC), looking for combinations with SOC or different timing (e.g., neoadjuvant early phase).
10. How can collaboration between preclinical and clinical teams be strengthened to create a more seamless transition from lab to clinic?
Dr. Austin Duffy: The problem here is that the obvious answers – such as improving access to tissue – are difficult, as discussed above. In terms of making the transition more seamless, many sponsors do a good job of engaging with Principal Investigators at study initiation and conducting in-depth reviews of the science rather than delegating that step to an activation specialist who might not be as close to the science. It makes a significant difference to site staff when they have that strong engagement from people who are passionate about their drug.
Dr. Michael J. Wick: Increased communication to identify areas of need for modeling and disease characteristics, including innate and acquired drug resistance.
Improved workstreams for educating patients and care workers on the value of preclinical and translation research for next-generation drug development.
Sharing information and results between teams to improve understanding of clinical need and any limitations on model translatability.
11. Regulatory expectations for early-phase trials are evolving. How are these changes influencing innovation, particularly in adaptive trial designs and accelerated pathways?
Dr. Austin Duffy: Most regulatory evolutions have been positive, particularly the increased emphasis on widening the net for data, resulting in more backfill study slots being available for patients, which helps to increase the understanding and knowledge of drugs. Accelerated pathways are obviously good at providing access to wider patient populations for promising drugs. But one challenge that results from this approach is increasing selectivity for early-phase studies, which can place limitations on lines of therapy and even increase the number of studies focusing on untreated patients. It is understandable from the viewpoint of wanting to give a drug every chance of succeeding. However, it can also result in turning away patients whose biology has allowed them to receive multiple lines of treatment and might make them more likely to respond/benefit and contribute.
Dr. Michael J. Wick: In the preclinical space, advances and improvements in cell-based and in silico assays and computational analysis have improved target identification and early-stage validation.
Continued innovation in assay development and modeling may improve investigational drug potential and provide for accelerated IND approval.
12. Looking ahead, what major shifts do you anticipate in early-phase oncology trials over the next decade in terms of technology, trial design, and patient-centric approaches?
Dr. Austin Duffy: I presume and hope it will be on the technological side, incorporating not just AI but also, for example, virtual pathology/digital microscopy. Being so reliant on IHC – with labs distributed across multiple countries for multiple sponsors and trials – seems to be an anachronistic mismatch with the increasing refinement of the drugs themselves. So, I hope that can be improved.
Dr. Michael J. Wick: Improvements in molecular profiling assays, databases with translational drug testing outcomes and computational modeling should accelerate the identification of cancer drivers and potential therapies for patients, allowing a more precise treatment plan and reducing less useful therapies.
Thank you, Dr. Austin Duffy and Dr. Michael J. Wick, for sharing your valuable insights on the evolving landscape of early-phase oncology trials. Your perspectives on translational research, clinical innovation, biomarker-driven strategies, and precision oncology have provided meaningful insight into the future of cancer drug development.
