The Reality of Clinical Trials – Part 2
Bess Stillman, MD details how the Regulatory process needs to be modernized to better serve both cancer patients and Sponsors. We extend the concepts to brain cancer trials.
![](https://substackcdn.com/image/fetch/w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F01f52592-807c-4306-8dd8-ce2d8ce6a4f2_1452x1260.heic)
Frustrating Aspects of Cancer Clinical Trials
In our original “The Reality of Clinical Trials – Part 1” post (see here), we featured some eloquent and poignant writing from Bess Stillman, MD whose husband has been battling recurrent squamous cell cancer of the Head & Neck. Dr. Stillman is better positioned than most to understand and navigate the vicissitudes of modern clinical trials, as she writes in a very recent public post (see here).
The same is true of the MissionGBM team, and yet we all struggle with a common set of frustrations:
Most (>80%) of the registered clinical trials in brain cancers should not be open and recruiting patients due to the poor quality of pre-clinical science used as justification for the trial. Continuing to subject patients, particularly rGBM/HGG patients, to low quality trials with no scientific basis only guarantees that the trial will fail, and perpetuate the perception among investors and other stakeholders that clinical development in brain cancer is “radioactive”.
Clinical trial sites at the major Brain Tumor Centers (“BTC”) are incentivized to recruit and enroll patients for trials open at their BTC, which means that patients will frequently meet a Clinical Trial Coordinator on their first post-neurosurgery visit and prior to a discussion of the overall treatment plan with the Neuro-Oncologist. This behavior is confusing and often perceived by the patients as self-serving and disingenuous. In some cases, it can also be harmful. We have had to coordinate rescue plans for more than a dozen cases in which the patient should never have been enrolled in the trial due to (i) ineligibility according to the published Inclusion or Exclusion criteria; or (ii) a prior medical history that clearly indicated the patient would very likely experience a Serious Adverse Event (SAE).
Clinical trial records on ClinicalTrials.gov tend to be out-of-date, incomplete, difficult to navigate, and in many cases, contain misleading information.
Individual clinical trial sites and the associated Investigators often apply a set of patient selection criteria that is somewhat different than the published Inclusion and Exclusion criteria, which is terribly frustrating to potential trial enrollees.
Both the Investigators and the Sponsor know that the best scientifically rational course of treatment for the patient involves a combination therapy, but they are prevented from doing so because the Regulatory agencies insist on a series of monotherapy RCTs, each with its own SoC Control arm.
For brain cancer patients with aggressive clinical timelines, every day counts so the above complications serve as unnecessary roadblocks to clinical trial participation.
And Investigators wonder why only 8-10% of eligible patients ever enroll in a clinical trial? The problem should be obvious and solvable, if the Regulatory and clinical studies framework can be modernized for critical diseases with aggressive clinical courses (like brain cancers).
Look, we get it. The embarrassment of approving bevacizumab (Avastin®) for GBM based on data from a single arm trial in which the readout criteria for the Primary Endpoint was poorly understood (once again, poor quality science in NO) at the time has haunted the FDA for years. But that cannot be an excuse for a reluctance to work with the Neuro-Oncology community and Sponsors to design a more flexible clinical trial environment that has a fair chance of identifying and approving therapies, including combination protocols, that can actually make a difference for brain cancer patients.
Towards Better Trials for Brain Cancers
We have a few thoughts on the matter, and we know that we are not alone in supporting some of the following proposals. Numerous Sponsors, investors, and Neuro-Oncologists have spoken to us extensively about the following topics.
Allow Trials to Incorporate External or Historical SoC Controls. There are simply not enough patients to insist on an active Control arm utilizing SoC for each pivotal clinical trial. The SoC treatment regimens have been around long enough to have yielded consistent mOS and other trial parameter data over many trials. Further, no prospective patient wants to be randomized into the SoC Control arm as they are very aware that SoC is not a winning strategy.
Permit Combination Trials That Do Not Require a Factorial Design. By factorial design, we mean the Regulatory requirement that necessitates multiple arms designed to evaluate Combination-versus-the Parts data. Again, there are simply not enough patients to meet this requirement in a fair number of brain cancers. Brain cancers tend to be heterogenous, aggressive and invasive, and we believe the #Data indicates that combination protocols will be required to make progress towards effective treatments. Instead, Regulatory agencies should consider (i) demanding rigorous, pre-clinical #Data in the IND/IDE package that establishes the MoA biology, pharmacokinetics and safety pharmacology of the Combination-versus-the Parts in suitable pre-clinical models; and (ii) incorporating sophisticated statistical and data analysis techniques capable of determining if the proposed combination protocol is valid and working. A Regulatory insistence on IND/IDE packages that emphasize Point (i) would eliminate a fair number of the poor quality protocols that currently lead to open trials, and thus, siphon away the very limited number of patients into trials that have no chance of providing any benefit other than clinical trial revenue to the BTC and yet another published paper describing a failed trial.
Encourage Cooperative Platform Trials, but Be Rigorous About the Candidates Selected for the Trial. Platform trials are an idea that has been around in rare disease communities for a long time, and we are generally supportive. When the number of patients is small, it seems reasonable to encourage patient coordination amongst potentially competing clinical sites in order to ensure that promising clinical development candidates are efficiently tested with an adequate number of trial enrollees. For platform trials to work, some key criteria must be met:
Investigators and trial sites must agree to the cooperative framework established by the Platform Trial Organization (PTO).
Regulatory agencies must approve a Master Trial Protocol that they agree can result in an approved therapeutic agent, if the trial data is supportive.
Sponsors must be sufficiently attracted to the Master Trial Protocol and the professional capabilities of the PTO that they become willing to cede trial execution and reporting responsibility to the PTO.
The Candidate Selection Committee for the platform trial must demand rigorous pre-clinical data from the Sponsor before a Candidate is allowed to enroll patients.
Patients must feel that they are willing to be enrolled in a trial that is based on sound science and clinical trial design.
In practice, it has historically been challenging to execute platform trials because one or more of the preceding criteria is compromised, usually Criteria #3 and #4. Criteria #3 can be difficult to satisfy because every Sponsor is nervous about giving up control of a clinical trial to a PTO that does not “own” the Candidate, and may be a rather inexperienced, academic organization that is prone to poor communication and quality control. But the real Achilles heel is Criteria #4 owing to the strong incentive that a PTO’s Candidate Selection Committee has to “feed the beast” by allowing lower quality Candidates to be included in the platform trial in order to help pay for the expensive infrastructure that the PTO has established.
Sponsors with high quality Candidates that have undergone extensive pre-clinical and early clinical development work do NOT want their Candidates to be associated with a platform trial that has shown itself willing to select poor quality Candidates.
The Global Coalition for Adaptive Research (GCAR, see here) was established as a non-profit organization to encourage and facilitate adaptive clinical platform trials in a variety of disease areas including GBM. GBM-AGILE was launched and has been operating for about five years (see here and here). Unfortunately, GBM-AGILE has yet to achieve its promise. For many in the brain tumor community, the effort has suffered from admitting Candidates of questionable scientific justification into the GBM-AGILE framework, often from thinly-funded Sponsors who do not have sufficient resources to prosecute well-designed pre-clinical and early clinical studies (we have seen the pre-clinical and Phase 1/2 data packages). In addition, the GBM-AGILE protocol was not designed to readily accommodate combination trials, which are vitally important in high grade brain cancers like GBM. We remain hopeful about the platform trial approach, but only if the Candidate Selection Committee is (a) able to upgrade its disciplined analysis of potential Candidates based on scientific #Data; and (b) design a Master Trial Protocol that can handle combination trials.
Discourage the Publication of “Science Lite” Clinical Trials. Academic science and medicine is a “publish or perish” business. Careers are built upon a publication record that can emphasize quantity over quality, particularly in a field like Neuro-Oncology in which little advancement has occurred over decades. If the IRB at an institution is willing to green light a study without first demanding reasonable scientific justification, and often desperate patients at that BTC can be convinced to enroll, it is possible to generate a significant volume of work. However, publication of low quality work should be another matter. Not every study is of the quality or impact to be reported in Nature, Science, Cell, New England Journal of Medicine, or even Journal of Neuro-Oncology. But all journals should nonetheless demand strong scientific rationale before publishing a clinical study. Doing so would send a clear message that publication will only be granted to those studies that are thoughtful, scientifically rigorous and focused on the patients. No science; no ink.