The traditional randomised controlled trial (RCT) model has been used for decades in drug development. Although it is considered the gold standard of trial designs, sponsors are increasingly using alternative trial designs, especially those for developing cell and gene therapies.

The unique nature of cell and gene therapies, how they are administered, the complex dosing schedule, and the specialised patient population can make it difficult to use a traditional RCT model.

Innovative trial designs, like single group assignment, adaptive, basket, umbrella and platform trials, allow flexibility to be built into a trial, which experts describe as crucial in running cell and gene therapy trials.

Chief medical and scientific officer at UK stem cell charity Anthony Nolan, Dr. Robert Danby, said: “Prospective RCTs have traditionally been the gold standard to evaluate the efficacy and safety of new therapies. For emerging cell and gene therapies, modern trial techniques like adaptive trials offer a promising alternative. These modern trial designs could offer patients new and better treatments sooner and do this without compromising on the quality of data required for new therapies to be approved.”

Single-group assignment trials far outweigh other designs

Despite data showing cell and gene therapy trials continue to be designed as RCTs, there is agreement amongst the industry professionals that RCTs may not be the best model, says Dr. Odelia Chorin, rare disease paediatrician and clinical geneticist at the Safra Children’s Hospital, Sheba Medical Center.

GlobalData’s Pharmaceutical Intelligence Centre shows that the single-group assignment trial design is by far the most used approach for cell and gene therapies.

GlobalData is the parent company of the Clinical Trials Arena.

The main reason for using alternative trial designs is to ensure trials are providing robust and interpretable data, says Amy Raymond, PhD, Executive Director Therapeutic Strategy Lead, Cellular and Genetic Medicines, at contract research organisation (CRO) Worldwide Clinical Trials.

While single assignment trials have been the leading design in cell and gene therapy trials in the past decade, Erin Griner PhD, associate director in clinical research methodology at Worldwide Clinical Trials says she is mostly seeing adaptive designs being used by sponsors.

“The thing that everybody in the process wants is optimal, efficient data, but the question is how do we get there?” Raymond asks.

Ultimately, the design is heavily influenced by the specific characteristics and requirements of both the therapy and the patient population rather than the indication itself, explains Neta Shanwetter Levit, Clinical Operations Lead, PhaseV, a company that develops machine learning (ML) technology to optimise clinical trial design and analysis.

“We are seeing a gradual increase in the number of adaptive trials and basket trials. Platform trials are not as frequent in this area yet, but are starting to increase,” Shanwetter added.

Dr. Beatrice De Vos, chief medical officer for EXO Biologics, a biotech developing exosome-based cell therapies, said that the flexibility to change protocols is the most helpful element of alternative trial designs, especially at early stages.

“The data collected in an adaptive trial is impacting the development immediately, impacting your next steps in clinical trial phases. That is different to the classic model, so I am very much in favour of the adaptive design,” De Vos added.

Therapy areas can also impact trial design

There is some differentiation between the choice of trial designs based on the indication. GlobalData’s Pharmaceutical Intelligence Centre shows following single assignment and RCTs are more common in oncology studies than in other indications while adaptive trials are mostly used to study treatments for genetic disorders.

“The challenges in rare disease studies are the limited patient population and the scarce data,” Shanwetter said. “Therefore, adaptive trials are particularly useful in rare diseases due to their flexibility. Basket trials are frequently used in oncology to test therapies targeting specific genetic mutations across different types of cancers.”

Griner agrees saying the basket trial design is used more often in oncology trials.

Despite it being difficult to randomise a gene therapy trial, in ophthalmology, sham controls are commonly used, Raymond says. But synthetic controls are also more commonly used in oncology while rare diseases are often studied in open-label trials and biomarkers are used to measure endpoints, she adds. Chorin agrees that given the lack of comparator, the industry needs to use biomarkers to measure clinical change.

Given how many different types cell and gene therapies, the design of a trial should be tailored on a drug-by-drug basis and there isn't a one-size-fits-all solution, says Shanwetter.

Alternative designs allow for alternative controls

Another benefit of alternative designs is they allow sponsors to compare the treatment arm with real-world or natural history data using synthetic arms.

Natural history data is collected from patients who received no intervention and allows investigators to have a baseline to measure the drug’s efficacy. Such approaches have been used in notable gene therapy approvals, Raymond said.

“Zolgensma [pivotal trials] used natural history data as a control, which saw the PIs coming together and doing that research. That in my opinion is what really enabled this gene therapy trial to succeed was using that baseline comparison,” Raymond explained.

Novartis’ gene therapy Zolgensma (onasemnogene abeparvovec) was approved for treating spinal muscular atrophy (SMA) in May 2019.

The ability to use existing data as a control is particularly helpful in rare disease trials where there are few patients, which often leads to trials being terminated due to low recruitment. Additionally, patients also might be more willing to participate if they know they will receive the study drug, Griner adds.

“For some rare diseases, we are seeing 80% to 90% of patients being willing to try a gene therapy that is as yet unproven because it gives them some hope. Having a trial design that is an open-label study really increases interest and as a result recruitment,” Griner adds.

De Vos says it is possible to match historical data to a degree that is similar to classical placebo, but acknowledges in statistical terms, it is not a head-to-head comparison. De Vos says EXO Biologics is running a Phase I/II trial (NCT06279741) of using a single group assignment design. The Phase I portion of the trial, which is evaluating EXOB-001, an exosome-based cell therapy for premature babies with bronchopulmonary dysplasia, is using natural history data as a comparator while the Phase II stage will be a standard RCT model.

The only challenge with using natural history data as a control, Griner explains, is that in some diseases there is very little available.

Regulators have been mostly supportive

Overall, Griner says that agencies have been supportive of studies with alternative designs, adding that the US Food and Drug Administration (FDA) has also released guidance on confirmatory evidence.

De Vos said that there is also guidance available in Europe but that it should be read very carefully. “[However] having gone through all these guidelines, companies should not refrain from coming up with new designs, rather than sticking to the old fashioned ones to extract the most out of their observation in terms of new product development,” De Vos concluded.

Raymond says the FDA and EMA are often aligned, but not always, recalling instances where a trial has required region-specific protocols, and in some cases region-specific primary endpoints.

“One thing the regulators need to be looking at is having a more unified framework with alternative trial designs. It’s not a small ask but it should be a real priority as it would be helpful to the industry as a whole,” Raymond concluded.