Phase I Clinical Trials: Hypothesis Testing
Keith Flaherty, MD, Massachusetts General Hospital Cancer Center, Boston, Massachusetts, USA
Authored by Sushma Jonna
By applying his personal experience in the development of BRAF directed therapy, Dr. Keith Flaherty elegantly described protocol development and hypothesis testing in phase I/II trials.
To give us some context, he first shared his personal story in hypothesis generation regarding BRAF therapy in melanoma. He described his education and the different opportunities that directed him to pursue this specific field. He was deeply impacted by his clinical experiences and patient interactions. In addition, he was active in scholarly activity, including staying current with the literature and attending conferences. Finally, he was also driven by intellectual curiosity. These different aspects are all important in asking research questions. While translational science provides a framework for ideas, clinical relevance and improving patient outcomes are the ultimate considerations in forming hypotheses.
Dr. Flaherty then detailed the steps for developing a pipeline of phase I/II trails. The first step is formulating a testable hypothesis. Next, we must ask “Does the drug act like a drug?” by understanding the preclinical studies as well as the pharmacokinetics and pharmacodynamics of the particular agent. Reviewing the molecular characteristics of agents including the targeted pathways and subsequent downstream signaling is prudent. We then need to evaluate the clinical effectiveness of a drug. This analysis is critical as it impacts advancement to potentially larger clinical trials. Finally, we need to consider the factors that account for the variable effects of a drug. Specifically, the possibility for predictive biomarkers should be thoroughly assessed.
Each of these concepts was applied to the development of BRAF agents. BRAF mutations are present in 7% of cancers and 60% of patients with melanoma. This oncogenic driver was first identified as a potential target when siRNA against BRAF was shown to induce apoptosis. In preclinical models, Dr. Flaherty showed that a selective BFAF inhibitor induced responses in BRAF mutant melanoma. In preclinical studies, the optimized formulation was then carefully established. Dramatic clinical responses were seen in patients in BRAF mutant melanoma, which eventually led to larger trials. Further understanding of the downstream signaling pathways and resistance mechanisms guided novel strategies; the combination BRAF/MEK inhibition was then established.
In addition to the above steps, Dr. Flaherty also emphasized the importance of collaboration with other scientists and industry throughout the process. These connections are crucial in implementation of clinical trials and obtaining adequate resources to do so.
Authored by Charlie Kuang
Dr. Flaherty’s presentation on how to conduct phase I trials had a narrative portion not reflected in the slides, but still enormously helpful for investigators to understand in order to better navigate clinical trials. His presentation included the following advice:
- New investigational opportunities may arise from new science, and these will not necessarily be adopted by the large academic programs immediately, often because they will have robust trial programs with little room to take up such new opportunities
- One difference between a smaller drug company and a larger company is that the smaller company may be more willing to compromise at certain points along the developmental pipeline, due to the pressure they are under to deliver a product to market
The drug development process follows a pattern: formulate a hypothesis, test for phenotype of the drug on tumors, test whether the anticipated mechanism is true, test for a clinical phenotype, examine for variability in effect or loss of effect, rinse, and repeat.
Dr. Flaherty’s presentation then proceeded to describe the discovery and validation of vemurafenib for BRAF V600E melanoma, as told through the lens of the above phase I/II drug development process.
- Hypothesis: Inhibiting BRAF in BRAF mutant melanomas can inhibit tumor growth. TCGA and other data show BRAF mutation to be a common alteration in melanomas.
- Test if phenotype on tumors: Preclinical data from siRNA knockdown of BRAF causing apoptosis in cell lines supported this notion. BRAF specific inhibitor PLX4720 caused shrinkage of tumors in BRAF mutant melanoma mouse model and was no better than control/vehicle in a BRAF wild type model. This last point suggests BRAF mutation is a biomarker for response to BRAF inhibitors.
- Test if mechanism is true: PLX4720 inhibition assay showed lowest IC50 on BRAF V600E, over an order of magnitude lower than with wild type BRAF, and even more so against other non-RAF kinases. PK/PD testing in initial trials were necessary to determine best dosing.
- Test for clinical phenotype: Phase I trial produced numerous notable responses, as demonstrated by comparison of baseline and restaging PET/CT scans of four patients presented. The waterfall plot from the extension cohort of the phase I trial demonstrates an impressive 70% RECIST response rate.
- Examine for variability in effect or loss in effect: in other words, what are the mechanisms of action and resistance? This section emphasizes the importance of incorporating tumor and blood-based biomarkers into phase I/II trials, including serial tumor biopsies, blood collection for ctDNA and exosomes, and rapid autopsies.
My take home points: follow this proposed model for the drug development process, keeping in mind the pearls of advice about new investigational ideas and small versus large drug companies.
Authored by Abhishek Maiti
Dr. Flaherty’s talk was particularly informative for aspiring clinical investigators. He recapitulated his experience of developing BRAK/MEK inhibitors in melanoma. It provided a practical walkthrough of how drug development works in academia and how to collaborate with industry partners. It started while he was still in training and imatinib had recently been developed, in 2002 when the BRAF mutation was reported in the UK Cancer Genome Study. Despite skepticism from his mentors, Dr. Flaherty was convinced that the MAPK/ERK pathway could potentially be targeted to treat melanomas. Subsequently, pre-clinical studies using RNA interference confirmed that BRAF mutations were essential drivers for melanoma. After failure with one non-specific inhibitor, sorafenib, Dr. Flaherty conducted preclinical work with another BRAF inhibitor which would later be known as vemurafenib. He showed selective BRAF inhibition with vemurafenib and its activity in preclinical models. In 2006 Dr. Flaherty opened the phase I trial of vemurafenib, and patients with widely metastatic melanoma showed dramatic responses. This trial led to FDA approval of vemurafenib. However, he noted that patients also developed resistance quickly. Further work led to identification of redundant feedback loops in the MAPK/ERK pathway, which provided rationale to combine BRAF and MEK inhibitors. Subsequently, the clinical trial of combined BRAF/MEK inhibition yielded the first example of improved efficacy and reduced toxicity through combining targeted therapies. The key take-home points from his talk regarding building an academic career in clinical investigation include:
- Starting out with a hypothesis which can be tested using a valid target and novel or existing drugs
- Conducting pre-clinical work ensuring that the drug hits its target and has the desired biological effect
- Designing thoughtful early phase trials to test the clinical effect
- Determine response and resistance mechanisms, and
- Repetition of this process.
Authored by Christopher Nabel
There are certain aspects of translational research that cannot be taught in a standardized curriculum and, instead, are more amenable to sharing through discussion that reveals common themes. Many of the lectures at the Society for Translational Oncology Fellows’ Forum fall under this classification but perhaps none quite as much as Dr. Keith Flaherty’s lecture on what comprises an effective Phase I clinical trial. The early investigational space is highly interdisciplinary, as basic science, medicinal chemistry, pharmacology, and clinical trial design all converge in an attempt to test a hypothesis. Navigating these issues is critical for the success of a trial and the potential to achieve long-term clinical benefit for patients.
Dr. Flaherty’s talk communicated two major messages. In the first portion, he demystified the major stepping stones in an early-phase trial, primarily based on his professional experiences. The first step involves having a clear hypothesis that guides the scientific rationale of the clinical trial. Once the hypothesis is in place, it needs to be tested—specifically with a drug (or, for the purposes of framing with regards to the hypothesis, a tool compound to test the biology of interest). A lot of iterative target validation and medicinal chemistry then go into the next two questions: Does the drug act like a drug? And does the drug do its molecular effect? While these questions feel like somewhat of an oversimplification given the tremendous amount of work that goes into these two elements, the simplistic reality is that the drug has to work if there is any potential for further success. Beyond these foundational questions, the matter of whether the drug has any clinical effect requires additionally favorable pharmacokinetics and pharmacodynamics. Additional investigation regarding factors that account for variable effects or loss of effect of the drug can drive ancillary studies on resistance.
While there are tangible “hard stops” that need to be met as described in the prior paragraph, there are additional intangibles that are harder to quantify. Specifically in this regard, Dr. Flaherty discussed the importance of finding alignment between the trial sponsor and the study team to ensure that both parties advocate towards a common, successful endpoint for the trial. I had not considered many of the stressors that early-stage biotechs may face that provide competing interests in trials. In particular, the time pressure that comes with a limited budget was not something I would intuitively anticipate in determining the endpoints for a trial. Additionally, a company without a medically-trained representative within their leadership may lack critical insight into the implementation of a trial or recruitment of patients. Making sure that the two parties truly understand each other is critical to ensure that the early stage clinical trial is in the best position to succeed.
I lastly appreciated Dr. Flaherty’s enthusiasm in sharing his work for the translational process. What came across in his presentation is his real passion for this translational process and his desire to share it with others. By distilling a complicated, somewhat nebulous process into general themes, he made it feel accessible—and I feel slightly more empowered for having heard him talk.
Return to STOFF 2020