enough to deal with the growing complexity of trials?

The FDA’s attempt to downplay false conclusions about a drug’s effect in a clinical trial with multiple endpoints has reached its finish line. The agency has released its latest guidance on the issues of managing multiple comparisons when analyzing multiple outcomes in a single study.

The FDA announced the first draft guidelines in 2017, calling on industry to provide feedback on challenges in analyzing and interpreting data when more than one endpoint is included in the study. The purpose of the guidance was to provide a more detailed perspective on the issue than that indicated by the International Council for Harmonization (ICH) guidance for industry in the report, E9 Statistical Principles for Clinical Trials, published September 1998.

While reiterating the statistical knowledge and practices common to the various stakeholders of the clinical trials industry, the guidelines did not break new ground. Experts share their thoughts on the FDA guidelines with Clinical Trials Arena and its impact on the future design and analysis of clinical trials.

Clarity on terminal terminology

According to Dr. Daniel Normolle, associate professor of biostatistics at the University of Pittsburgh School of Public Health, the guidance has made some of the terminology for different types of parameters more explicit. The final version includes clarified distinctions between secondary and exploratory endpoints, which was requested by Novartis and Boehringer Ingelheim.

Novo Nordisk asked to see a reference to the ICH E9(R1) guidelines on estimates and their relationship to families of primary and secondary endpoints. Clinical Trials Arena has already reported on the estimation framework and its benefits for dealing with missing data.

At Merck’s request, the appendix has been expanded to include available statistical methods, which were originally incorporated into the body of the draft guidelines. According to the Regulations.gov submission portal, the FDA received a total of 20 comments and proposed changes from the pharmaceutical industry on the draft guidelines.

Overall, the guidance is informative and will be useful to the entire pharmaceutical industry, which ranges from multibillion-dollar global pharmaceutical companies to smaller-scale biotechs, points out Dr Chen Hu, associate professor of oncology at Johns Hopkins University School of Medicine. “Having common ground for people to start is something that should be welcomed,” he adds.

Repetitive but still beneficial

Although the guidelines clarify terminology and types of endpoints, they don’t add much value when it comes to more complex clinical trials, Normolle says. “I found the FDA guidelines somewhat disappointing. Frankly, it could have been written 20 years ago,” he notes.

Normolle explains that although this guide is a useful document, it is limited to very low-dimensional problems. Indeed, the guide provides scenarios where few parameters are included in a clinical trial. However, clinical trial sponsors and researchers want to see guidance that shows them how to deal with high-dimensional parameters in more complex trial designs, Normolle says. “What I took away is that you should talk to [the] FDA before doing anything important. That’s good advice, but why do you need that in a guide? ” he adds.

On the contrary, Hu says the FDA guidelines cannot be very specific to overly complex scenarios. Rather than criticizing what is missing in the guidance, it should be viewed as the first version of a document that will continue to evolve over time as this field matures. He adds that at some point the FDA should resume this conversation with industry and academia, especially statisticians.

In therapeutic areas such as oncology, there has been an increasing use of multiple endpoints, particularly in large randomized phase III trials with multiple aims and clinical benefits measured over the course of a disease. Hu said. He adds that this poses problems for correctly categorizing an entire treatment.

Even though the field of oncology has practiced much of what the guidelines state, it still has value. Hu says repeating the same advice gives others a good start and encourages them to have a more focused discussion.

Last month, the FDA released draft guidelines for including children in clinical trials. The experts said Clinical Trials Arena that while the project is useful for some stakeholders, most of the points raised are already in practice.

Related companies

Sharon D. Cole