REMAP-CAP master protocol has lessons for studying drug cocktails
BioCentury is providing this story for free given the urgent need for information about the COVID-19 crisis. For more analysis, sign up for our daily email
The U.K.’s REMAP-CAP master protocol is poised to teach drug developers a way to cut through the mess of individual trials for drug combos in diseases where drug cocktails are all but inevitable. If successful, it will also serve as one more salutary lesson of the broad-reaching benefits of preparing for pandemics during the good times.
The case for master protocols is already receiving a significant boost during COVID-19, with nine adaptive master protocols under way already, and the design selected as the major vehicle for expediting testing by industry consortia such as COVID R&D (see “COVID R&D Consortium Launches I-SPY Adaptive Trial with Quantum Leap”).
However, REMAP-CAP is the only one with a multifactorial design, which means it studies drug cocktails, rather than evaluating multiple individual treatments side by side. As such, it’s designed to answer a much broader range of questions.
“The protocol is designed to explore the effect of treatment regimens rather than single agents, and it creates quite an efficient structure for doing so,” Elliott Levy, SVP of global development at Amgen Inc. (NASDAQ:AMGN), told BioCentury. “I find that very exciting because it’s possible we won’t find in the existing armamentarium any therapies that have a large treatment effect.”
“What we’ll need to do to have a meaningful clinical outcome is combine therapeutic agents that may have a relatively modest effect alone,” said Levy.
The trial shares the benefits of other adaptive master protocols: faster readouts and fewer patients, trials and other resources spent. Importantly for patients, the trial employs shared control arms, so more patients receive active drugs (see “Master Protocols Emerge as a Critical Clinical Tool Against COVID-19”).
“What we’ll need to do to have a meaningful clinical outcome is combine therapeutic agents that may have a relatively modest effect alone.”
In addition, REMAP-CAP allows investigators to determine which broad classes of drugs are beneficial for which patient groups, which combinations of drug classes have additive or synergistic effects, and which specific drug combinations from within those classes are most effective.
The ramifications should be clear for fields such as immuno-oncology, non-alcoholic steatohepatitis (NASH) and some other viral infections where the need for combination treatments is already evident. The trial design provides a road map for investigators willing to forgo individualized trial designs and allow direct comparisons with competitors to get meaningful results faster.
The trial was set up four years ago in preparation for, rather than in response to, a pandemic. That investment, by the European Commission and several other government agencies, could pay off, therefore, not only in COVID-19, but far beyond.
Scott Berry, a member of the REMAP-CAP steering committee and co-founder and president of Berry Consultants, which specializes in adaptive trials, believes the tide could turn for this trial design.
“The positive silver lining of the pandemic might be that it accelerates the way in which we learn in all diseases because of how the community did it in COVID-19,” Berry told Biocentury.
Benefits beyond COVID-19
COVID-19 patients in the REMAP-CAP trial are randomized to receive a therapeutic or standard of care in each of up to seven therapeutic classes: antiviral, immunomodulator, steroid, macrolide, anticoagulant, convalescent plasma and vitamin C.
Patients can receive up to seven therapies, but some qualify only for certain classes, based on co-morbidities. Combinations vary between patients.
As understanding of the disease biology unfolds, the trial can add or drop classes and therapies, employ effective therapies as active controls, and use early readouts to allocate patients to treatment arms.
The benefit of this structure, and the statistical methodology behind it, is that it doesn’t stop at determining which individual treatments work better than SOC. It points to the combinations of therapies that are most effective, and indicates which therapeutic classes or combinations of classes are useful for patients in a certain disease stage.
The trial is enrolling severely ill and moderately ill hospitalized patients, and the two groups may be given different regimens.
According to Berry, the U.K.’s PRINCIPLE trial will likely incorporate this design when it branches out from a single treatment arm. “I know the PRINCIPLE trial in the U.K. for mild, ambulatory patients started off by looking at hydroxychloroquine but is planning to add therapies in a factorial way,” he said.
“When the COVID-19 pandemic is over I think we will absolutely have a jump in usage in other diseases.”
He thinks the uptake will go further as investigators become more fluent in the design and its advantages.
“The pandemic has accelerated the science of platform trials so when the COVID-19 pandemic is over I think we will absolutely have a jump in usage in other diseases,” said Berry.
Berry said oncology is one of the few therapeutic areas where adaptive platform trials were deployed before the pandemic, notably the I-SPY 2 breast cancer trial and the NCI-MATCH basket trial (see “Back to School: Haste, Not Waste”).
Amgen’s Levy said myeloma presents an ideal setting because there are multiple therapeutic candidates, virtually every patient receives a regimen, and the number of patients and sites relative to the number of potential clinical trials is small.
NASH is another, given that the biology points to a need for combos that can address the inflammatory, fibrotic and metabolic components of the disease (see “New Partners for NASH”).
“In general, a multifactorial, multidomain adaptive trial would be well suited to a disease for which there are multiple therapeutic candidates; patients receive therapeutic regimens rather than single agents; and there are limitations on clinical trial capacity,” said Levy.
The advantages of the trial design for patients are clear, but for some companies, it raises concerns.
Participating in an adaptive trial surrenders some control over the trial design. In individual trials, companies can tailor their trials to increase odds of success for their specific drug, and they lose that level of control in a master protocol.
The adaptive trial design also tests competing products in parallel, so therapeutics can be more directly compared.
The counter argument is that getting an early signal on where a drug stands in the competitive landscape is valuable to companies, as well.
REMAP-CAP launched in 2016 in anticipation of a flu pandemic, and has been evaluating treatments for community-acquired pneumonia between flu outbreaks.
The protocols are different for the community-acquired pneumonia and pandemic patient strata, but they use the same trial sites, basic statistical design and data collection methodology.
Having the protocol and trial sites up has yielded benefits during COVID-19, where resources are stretched and major hospitals receive hundreds of trial proposals (see “U.K. Sets the Bar for COVID-19 Master Protocols”).
Berry told a webinar hosted by Berry Consultants two weeks ago that since the trial’s “sleeping strata” for the pandemic was activated at the end of February, over 100 sites were enrolling COVID-19 patients, and he expected that to rise to over 200 in the coming weeks or month.
REMAP-CAP has branched out in the molecule classes, therapeutics and disease states since the COVID-19 strata was activated.
“Now each patient contributes to learning for five or six different questions, plus we’re treating them better.”
The first design focused on severely ill COVID-19 patients; a second group of moderately ill hospitalized patients was added over the last few weeks.
“Many of these therapies might be more effective before the disease gets severe, and you may prevent the cascade that leads to severe disease if you intervene early. There’s the potential for a bigger impact on outcome by treating early and there’s a larger population of patients so that leads to a larger learning effect in the model,” Berry told BioCentury.
The trial contains at least 11 therapeutic candidates and seven therapeutic classes, and Berry said more molecules are in the planning phases.
The trial’s antiviral arms include remdesivir, hydroxychloroquine, Kaletra lopinavir/ritonavir, or a combination of the latter two; the immunomodulator arms include Anakinra, interferon or IL-6; and each of the other five domains have a single therapeutic arm.
Now that remdesivir has proven clinical utility in a separate trial, it will become the SOC arm against which other treatments are compared, as long as the drug remains available.
“I believe we’re still randomizing to standard of care because a number of sites don’t have remdesivir,” said Berry. For sites that do have remdesivir access, he believes it will become the SOC arm in the antiviral domain.
The model is statistically powered to determine how each of those individual therapies measures up against SOC, and is also powered to learn the effect of certain treatments that aren’t expected to interact with others, such as vitamin C. If there are multiple treatments with no interactions modeled, the trial can produce similarly powered answers to whether each drug works with fewer patients than separate individual trials.
“You can investigate multiple questions in one patient. Now each patient contributes to learning for five or six different questions, plus we’re treating them better,” Berry said.
Further analysis of the coronavirus crisis can be found at https://www.biocentury.com/coronavirus.