NEW YORK – The Leukemia and Lymphoma Society recently required that its partners in the Beat Acute Myeloid Leukemia Master Trial embrace technologies, such as centralized oversight platforms, remote data monitoring, and digitized protocols, recognizing that such multicohort studies, increasingly popular in the era of precision oncology, require different management strategies than traditional trials.
In a recent publication in Therapeutic Innovation & Regulatory Science, Beat AML investigators, including Amy Burd, LLS' VP of research and strategy, laid out the approaches they've taken to streamline the precision medicine umbrella trial and pivot all stakeholders to new technologies.
"One of the biggest lessons we've learned is that with the amount of data we have … using these new technologies has become almost a necessity," Burd said. "If you approach [an umbrella trial] like it's just one study, you're going to be in trouble." The LLS is hoping that its experience implementing these new tools and decentralization strategies will influence investigators to embrace similar approaches within their precision oncology master trials.
An umbrella trial, while technically one trial with an overarching protocol, is simultaneously investigating a variety of interventions across multiple cohorts. When it comes to selecting technologies and staffing sites, an umbrella trial cannot be approached the same way as a single clinical trial evaluating one intervention. This, Burd explained, is only going to become more important for drugmakers and researchers to recognize as more investigators embrace basket and umbrella trials as efficient ways of finding new signals of therapeutic activity in biomarker-defined populations.
The Beat AML trial, launched in 2016, currently involves more than a dozen sub-studies, 16 sites, nine biopharmaceutical partners, and 1,065 patients. Within the study, researchers hope to demonstrate the feasibility of performing molecular, genetic, immunophenotypic, and biochemical testing for AML patients in fewer than seven days; assess the proportion of patients assigned to a sub-study based on molecular profiling; and evaluate clinical response rates for patients assigned to these biomarker-matched trials. Most enrolled patients are treatment-naïve and at least 60 years old, and the biomarker analysis includes locally administered metaphase cytogenetics, central next-generation sequencing conducted by Foundation Medicine, and FLT3-ITD mutation analysis performed by Invivoscribe.
Though investigators publish data from individual Beat AML sub-studies as they read out, in October 2021 they summarized the overarching progress in the trial in Nature Medicine. Among patients enrolled between 2016 and 2018, 94.7 percent successfully underwent molecular analysis within the seven-day timeframe, and 56.7 percent enrolled in a biomarker-matched sub-study. The median overall survival among enrolled patients was 12.8 months compared to 3.9 months for patients on standard-of-care therapy.
The trial's success to date, according to the investigators' latest paper in Therapeutic Innovation & Regulatory Science, is enabled by the numerous technologies that have streamlined Beat AML operationally. Burd was quick to highlight a technology called e-Source, a product manufactured by the company Protocol First, as the technology that has had the greatest influence on the trial’s efficiency.
"[E-Source] allowed sites to obtain certified copies of the medical records that corresponded with the data entry in the [case report form]," she explained. "This minimized the amount of time that was needed for data entry and the time that you need to review the data that's entered." The e-Source system, which is integrated with an e-Protocol, is accessible remotely through the cloud.
"Traditionally, a trial would give a hard copy of the protocol in a binder to the site, and when a patient came in, the investigator would look at the protocol and say, 'I need to do X, Y, and Z,'" Burd said, explaining that with the e-Protocol and e-Source, the master protocol and the sub-study-specific protocols can be accessed off-site via phones, tablets, or computers, and investigators can directly input the data as they collect it.
In addition to saving time on the data entry side, the centralized cloud-based technologies also minimize data entry errors, such as omitting entries for certain required data points. "Oftentimes [before implementing the new system], things were missed because you're looking directly at the protocol, and sometimes it's not easy to follow," Burd said, adding that since digitizing the protocol, these types of omissions have been rare.
These types of efficiencies became particularly valuable last year as the COVID-19 pandemic upended the world of clinical research, hindered travel, and halted traditional workflows. But having implemented some of these digital tools in the first several years of the Beat AML trial, LLS was already in a good place in terms of real-time, remote data monitoring. "The pandemic forced companies to move to this digitalized setup, [but] we didn't really have to change anything dramatically in terms of our operating procedures, because we were already set up to be remote," she said, noting that a lot of companies and trial sponsors during the pandemic encountered issues trying to access data without physically being at their study sites.
Even though LLS had already built remote data monitoring into their master protocol when the pandemic hit, they did make other changes to Beat AML's operations to minimize patient exposure to the virus. These changes, allowed under a March 2020 FDA guidance for clinical trial conduct during the pandemic, included shipping some of the targeted oral medications being studied within sub-protocols directly to patients' homes and incorporating virtual visits via telemedicine for follow-up visits. Simple tests such as blood draws could be done at local labs rather than requiring patients to come into the study sites. These changes may have seemed minor in the grand scheme of Beat AML's scope, but Burd said they traditionally would not have been allowed under the protocol.
Now that vaccination rates are rising and patients' risk of coronavirus exposure in the US is falling, many are wondering whether a silver lining of the pandemic may be that these emergency modifications become permanent. Patients and many physicians have embraced changes like remote enrollment and monitoring since they minimize travel for the sick and make it easier to comply with trial requirements.
Burd cautioned, however, that too much decentralization could run the risk of challenging standardized data collection, potentially confounding data analysis. The field should seek a happy medium between the rigid, centralized clinical trial structure where everything is done at a single site to complete decentralization, which could complicate things, she said.
"Some of these things are here to stay, but the challenge we'll face as the pendulum swings the other way is that it could become a challenge to find the data," she said, referencing how too many exceptions and deviations made to accommodate decentralization could skew the data. "And this is a very important piece, because at the end of the day for a [drug] trial, that's what gets you regulatory approval."
Safety monitoring is a critical part of drug trials, particularly for novel treatments with risk/benefit profiles that are not as well understood, and is a major reason patients must physically come into a study site. But there may be a middle ground on this, too, Burd suggested. "Initially, you need to make sure [the treatment] is safe, and then maybe eventually you can start to wean off how often patients come into the site, and start to do things remotely," she said.
New umbrella trials on the horizon
In light of what LLS has learned both throughout the pandemic and through streamlining Beat AML, the organization is planning to incorporate remote-friendly data monitoring and centralized oversight systems into two additional umbrella trials it is designing, one for pediatric patients with acute leukemia dubbed PedAL, and one for patients with myelodysplastic syndromes dubbed Stop MDS. Both umbrella trials will involve an overarching master protocol and sub-studies to which patients are matched based on molecular profiling.
With the Stop MDS trial in particular, Burd said that LLS is planning to use the same e-technologies that it adopted into the Beat AML design. In addition, when the Stop MDS trial launches, LLS will begin piloting a BioSticker, which is an FDA-cleared adhesive medical device roughly the size of a quarter that collects data on patients' vitals, including blood pressure and heart rate, via their chest.
The technology may even be adapted to collect data when a patient falls or vomits, Burd said. "It's a way you can monitor patients while they're at home without having them coming in directly to the site," she said.
For now, LLS chose to pilot the BioSticker in MDS, rather than AML, in part because the disease is less aggressive but also because many MDS patients are treated at community practices rather than large academic centers, making remote monitoring well-suited to the treatment setting. If the pilot goes smoothly, Burd said her team will probably implement BioStickers into Beat AML, too.
Change management challenges
With all the changes implemented into Beat AML, particularly the major technologies, Burd acknowledged it was challenging to convince every stakeholder of their value and to encourage routine adoption.
For example, in Therapeutic Innovation & Regulatory Science, Burd and co-authors described how, when implementing the centralized oversight platform, organizers had to "remind study users to upload information onto the platform, to review it there, and to use the solution as the main communication dashboard for the program." Previously, most communications were via email, which wasn't always ideal because crucial information could get buried in cluttered email inboxes. The new system, on the other hand, allows Beat AML's investigators to communicate with all 16 sites in one place.
The same resistance to change presented challenges to implementing the e-Source technology, which required site personnel to stray from the workflow they were accustomed to. "Sites may resist the new e-technologies initially because of the perceived potential to increase workload, decrease productivity, and the time required for training," the authors wrote. "This was overcome by using 'technology change management' strategies, [such as] working directly with sites to train the staff on how to use the 'systems' and stressing the benefits of the e-technologies."
At first, Burd said, she and her colleagues underestimated how much convincing those at different study sites would need to adopt the changes. "We made the assumption that of course they would want to use [the new technology] because it's more efficient," she said. "But the reality is … we don't like change, or at least most people don't. … What we learned pretty quickly was that we needed to spend time with the sites to walk them through the new technologies [and] train them on how to use them."
While this initial learning curve did present challenges, Burd ultimately said that working through these challenges improved the efficiency of the study overall, and most of all, benefited the patients.
"There are two ways you can go about master trials," she said. "You can plow forward with the traditional way clinical trials are done, or you can move forward with utilizing new technologies to make things more efficient. ... We've shown how to make that work for this type of study."