Experience matters in clinical trials. Clinical research is a highly-regulated industry with myriad codes and guidances, and it can be dangerous if you ‘don’t know what you don’t know.’ Oncology as a therapeutic area is no different than many others: it has developed its own set of indication-specific processes and procedures, and experience in these intricacies should have a strong bearing on your choice of contract research organization (CRO). Though some of these may not be unique to oncology, and can overlap with other therapeutic areas, they are key aspects to success in oncology, and touch on many outsourcing cost drivers (e.g. patients, sites, timeline). I endeavor to share some key lessons from my last decade of oncology trials from both the CRO and sponsor side. Hopefully these will pique your curiosity and prepare you for your next oncology outsourcing relationship.
Phase I Trials
It is rare for an oncology therapeutic to begin as a Single Ascending Dose in healthy normal patients. CRO Business Development staff who tout their CRO’s Phase I unit to an oncology sponsor may have a tough time gaining the trust of that sponsor. Oncology trials start as multiple ascending doses in cancer patients, though they may not start in the precise indication they eventually intend to treat (e.g. non-Hodgkins lymphoma, bladder cancer, etc.). Phase I oncology patients have usually exhausted all other treatment options. It has also become uncommon for Phase I oncology trials to solely be a dose finding trial. Most of these trials incorporate at least an uncontrolled Phase II sample size component in more narrow indications. We have even seen pivotal Phase I trials, such as Keynote-001 that Merck used to gain approval of Keytruda (R) (pembrolizumab) in advanced melanoma. It was a 1200-patient Phase I, but a Phase I nonetheless. The next time you operationalize a Phase I study, ask yourself – what if this trial is used for registration?
Phase I/II Combo Trials
The Keynote-001 trial was an extreme example of Phase I/II combo trials in oncology, for size, efficacy, and length. But this model is the norm lately and creates operational and budget considerations. The Phase I portion of the trial begins in a handful of sites, typically in only one country. When the dose is selected the Phase II portion is opened. This requires close Project Management to ensure the Phase II site load is activated “Just in Time.” However, there can by country-specific regulatory restrictions on this approach. There is also a risk that the Phase II will never open, which can strain investigator relationships. Overall, I feel the operational advantages of a single trial outweigh the Project Management challenges, and I’m fond of this approach. I recommend thoroughly forecasting various enrollment scenarios and performing thorough country-level regulatory reconnaissance to mitigate the risk of your site activation strategy. It is also prudent to contract the Phase I portion separately from the Phase II portion of the same trial.
Academic Medical Centers
Many of the patients involved in clinical trials have been referred to Academic Medical Centers (AMC), which therefore form a large percentage of these trials. Academic centers drive study start-up timelines and effort, and prove more difficult to monitor. All National Cancer Institute-designated cancer centers in the US are required to have a Protocol Review Committee (PRC). When these operate in sequence to other steps it increases start-up times, and these PRC processes usually operate in the background with little visibility to the study team. Further, these centers have higher overhead costs but paradoxically usually have more resource problems for the study team (perhaps due to enrollment load). Though Academic Medical Centers are late to the game, their higher enrollment can make up for this lost time. Good budget models should include a variable for percent of AMC involved in the trial in order to account for these resource strains.
Oncology trials tend to be slow-enrolling (0.25-0.5 patient/site/month) which yields a protracted enrollment period (often 1-2 years solely to enroll the trial). Only a fraction of oncology patients enroll in clinical trials (though I often wonder what fraction of ‘healthy normal patients’ enroll in trials), and there is substantial competition for certain patient types. We’ve seen tremendous changes with recent approvals in immunotherapy, and those trials often boast higher recruitment rates than more traditional chemotherapeutics or targeted therapies. On top of these challenges, most oncology trials involve a survival endpoint requiring protracted survival follow-up. Good Oncology CROs have the experience to vet enrollment rate assumptions which can be invaluable. These CROs are likewise adept at correctly interpreting enrollment rates provided by KOLs.
Objective measurement of tumor growth is important to assess treatment effects. In some oncology indications, these objective tumor responses (either the rate of objective reduction or the length until objective increase) can be used for conditional regulatory approval. Different assessment criteria exist for different tumor types, and experience in navigating these criteria is vital to quality outcomes data for your trial. Experienced oncology CROs have established training courses in these response criteria (RECIST, irRECIST, Cheson, etc.) to support their operations staff.
Survival follow-up in a double-edged sword in oncology drug development outsourcing: it’s a part of your contract that you hope goes over budget. There are many tumor types where measuring tumor response can lead to conditional regulatory approval, but survival is the gold standard. These follow-up periods can be lengthy, and it is worth operational planning during the contract phase. Ensure your contracting discussions consider resource management, such as how to manage FTE over the course of such a “slow burn” period of the study. There can be implications for long-term EDC/IRT/RBM license fees (and monitoring of those data) that can drive study costs. It is also costly to keep other sites open while only a fraction of sites have patients in follow-up. All of these considerations warrant exploration of creative solutions, such as creating two studies: a main study with the first treatment period, and then a rollover protocol to allow continued treatment (e.g. beyond the first year) and capture additional survival follow-up. Though this process adds operational complexity and cost upfront, it yields benefits in the long term. Many trials have “exceptional responders,” and consideration of just these ‘exceptions’ can make these efforts worthwhile.
Oncology treatment responses are distal to the treatment, and a small fraction of patients respond to medication. This drives a need to collect many covariates that could become predictors of response. These manifest in early development as large CRF casebooks, and large numbers of biologic samples. Data volume can impact the cost of many trials: monitoring, EDC fees, statistical outputs, CSR, etc. Comorbidities and treatment toxicities yield large Adverse Event volume in most oncology populations. Multiply data volume by treatment length, and you have a major impact on trial costs.
My experience with these discussions on the CRO side indicated that sponsors were reluctant to cut data volume. My experience with these discussions on the sponsor side was that I was reluctant to cut data volume: I was not willing to potentially miss a safety/efficacy signal for this incremental cost. From the CRO angle, I felt there was also reluctance to adopt risk-based monitoring (RBM) approaches in Oncology, for similar reasons. Therefore, spend the time during the contract phase to set realistic expectations for data volume. But, I disagree with the aversion to RBM approaches, and I feel these analytics can provide tremendous value and operational efficiency.
Experience from strong oncology CROs can mitigate most operational challenges. Oncology indications have a large unmet need, and there is significant innovation in the field. The scientific excitement, an altruistic sense of the greater good, and a desire to contribute to the unmet need are all large drivers for my interest in this field. I hope these experiences aid your trials and therefore help the patients managing this disease.
Brandon Early is an oncology drug development executive with deep Oncology and Immunology experience. He currently leads and owns primary responsibility for clinical development at Heat Biologics, a cell therapy immuno-oncology company in Raleigh, NC. Brandon is also the owner and managing partner at Clinology LLC, drug development consulting group. Prior to joining Heat Biologics, Brandon led global Oncology programs at INC Research at office locations both in the US and in Europe. Brandon began his career as a Clinical Research Coordinator at the University of Virginia in the Carter Immunology Center, and Asthma and Allergic Disease Division. He can be reached at www.clinologyllc.com.