Why Do Clinical Trials of Pain Drugs Fail or Succeed? An Interview with Robert Dworkin.

Pain clinical trials

Pain clinical trials expert Robert Dworkin discusses how researchers can improve the design of clinical trials of new pain drugs. Image credit: bearsky23/123RF Stock Photo.

Published May 24, 2021

Editor’s note: This interview originally appeared on the IASP Pain Research Forum and has been lightly edited for RELIEF.

At the International Association for the Study of Pain (IASP) 2021 Virtual World Congress on Pain, to take place June 9-11 and June 16-18, IASP will present awards to honor the achievements of up-and-coming as well as more established investigators (these awards were originally to be presented at the 2020 World Congress on Pain in Amsterdam, which was canceled due to the COVID-19 pandemic). In advance of the meeting, we spoke with each of the winners. Here, in this interview, we chat with Robert Dworkin, PhD, the winner of IASP’s John D. Loeser Distinguished Lecture Award. Read more about the award here.

Dworkin is professor of Anesthesiology and Perioperative Medicine, Neurology, and Psychiatry; and professor in the Center for Health + Technology at the University of Rochester Medical Center, School of Medicine and Dentistry, New York, US. He is also director of the Analgesic, Anesthetic, and Addiction Clinical Trial Translations, Innovations, Opportunities, and Networks (ACTTION) public-private partnership with the US Food and Drug Administration. His research interests involve the development of improved treatment and preventive interventions for neuropathic and musculoskeletal pain, and the identification of factors that increase the assay sensitivity and informativeness of clinical trials of pain treatments.

Here, Dworkin speaks with freelance writer Kayt Sukel about his career in the pain field, the biggest challenges in running clinical trials, and the achievements of ACTTION. Below is an edited transcript of their conversation.

What inspired you to pursue a career in pain research?

I have a PhD in research psychology. Originally, I was doing research on personality and psychopathology. I was always drawn to using biological approaches in my work, but that initial research eventually led me to look at coping behaviors. I was interested in how people cope with diagnoses – for multiple sclerosis, Huntington’s disease, and other conditions in which patients experience great uncertainty about the future. In doing that work, many people talked about their pain and how they coped with that. It became a bridge into pain research. I’ve been working in this area now for 40 years.

Robert Dworkin

Robert Dworkin

At what point did you start focusing on optimizing clinical trials?

I started thinking about clinical trials 25, maybe 30 years ago. At that time, it became clear to many of us that we didn’t necessarily understand why we were getting some of the negative results we were getting in trials. We’d do a clinical trial of a drug or other treatment that we had a very good reason to believe would be better at relieving pain than placebo, and it just didn’t work out that way. [A placebo, such as a sugar pill, has no active components. Clinical trials often compare an active drug to a placebo to test the effectiveness of the drug].

Sometimes you’re pretty confident that it’s a false negative. If you have a lot of early trials that show efficacy, but then you do the larger confirmatory trial and the treatment doesn’t do any better than the placebo, it’s important to understand why. Bob Rappaport, a neurologist at the Food and Drug Administration [FDA], and others were interested in understanding for these trials: Was it really a negative trial of a drug that truly lacks efficacy? Or is something else going on?

Not only was this a source of consternation for pharmaceutical companies who were spending a lot of money to conduct these trials, but it was also a big issue for academic researchers like me and the regulatory agencies. What do you make of these negative results? What can we learn from them? That was the beginnings of the Initiative on Methods, Measurement, and Pain Assessment in Clinical Trials (IMMPACT) project – and that was the origin of my interest, and certainly the FDA’s interest, in setting up various kinds of partnerships to optimize the design of clinical trials.

The objective of the whole process, by conducting more informative and efficient clinical trials, is to improve the efficacy and safety of treatments for pain.

What are some of the biggest challenges in conducting an informative clinical trial?

It seems to be getting harder and harder to show separation from the placebo – or when you can show separation, the effect is much smaller than one might like. This is true not only for pain medications but for psychiatric drugs, too. It’s even true for drugs that we know are effective and have been approved by the FDA or the European regulatory agency, the European Medicines Agency [EMA]. Everyone has their own hypothesis of why this is the case, but no one really knows the reason. My own feeling is that at least part of the explanation is that clinical trials are being done differently than they were 25 years ago.

When I started doing trials for chronic pain in the 1990s, a lot of the trials were done at academic medical centers. Today, more and more are being done at dedicated clinical research operations, which often recruit potential participants from the community rather than from clinical practice. Such individuals either may have milder pain, or might have failed to respond to all the treatments they’ve tried. In the 1990s, to do a trial to test a treatment for fibromyalgia, we put an advertisement in the local Sunday paper. By Monday morning, we had 200 phone calls from people interested in participating. But the last trial I did of fibromyalgia – we didn’t even think about advertising in the paper. You’d only get five or six phone calls, and most patients wouldn’t be eligible for the trial, for one reason or another.

Importantly, we have efficacious treatments that we didn’t have back then. There are three drugs approved in the US for fibromyalgia, for example. Patients are prescribed those, and if they do well enough on them they would not be inclined to participate in a trial. That leads to a smaller reservoir of potential participants, which can make it harder to get these trials done.

Today, we also see a greater response in the placebo groups in clinical trials, and, possibly, in the active medication groups, too. It may be that the public is better informed about these new drugs and clinical trials. They have a sense of optimism about advances in medicine. So when a person does enroll in a clinical trial, even though they may know there is only a 50/50 chance they will receive the active drug, that optimism, their expectations, and perhaps the attention they get from the study staff come together to contribute to them feeling better over the course of three months – the typical length of chronic pain clinical trials. That can make it harder to get the separation you need from placebo in the trial.

How did you originally get involved with ACTTION?

IMMPACT, which started in 2002, was a smaller, less formal organization. Bob Rappaport had the idea of putting out a grant application for a larger, more formal FDA-sponsored public-private partnership to help improve clinical trials. About a dozen years ago, my colleague and close friend Dennis Turk at the University of Washington and I, along with others who were involved with IMMPACT, applied for this initial grant for ACTTION. We received the grant, which started in 2011, and the work has been incredibly gratifying ever since.

Can you describe some of the improvements in clinical trials that ACTTION has helped to foster?

One way of thinking about what we try and do is provide an evidence base for clinical trials of treatments for acute pain and chronic pain. Dennis and I want whatever we do to be relevant to trials of all types of pain treatments, and, more generally, to clinical research about pain and its causes and consequences. So we try to make general recommendations for how to do clinical trials better, in a more evidence-based way.

We’ve devoted a lot of effort to coming up with diagnostic criteria for the major acute and chronic pain conditions to help investigators figure out how you define the condition you are studying. You can’t just say, ‘We’re doing a clinical trial for a treatment for low back pain’ and then take anyone who says they have low back pain. We need a more standardized and systematic way to diagnose low back pain to determine who should be eligible for the trial.

Another major effort has involved analyzing published trials from the FDA database to look at factors that may influence the separation we see between drug and placebo. The idea is that if you can identify something that makes it harder to show an effective drug is truly effective, then you want to consider that in designing your trial. For example, most drug trials will have participants fill out a daily diary of their pain a week or so before the trial and we see that, with some people, those numbers bounce all over the place. It turns out, if you remove patients who have these excessively variable baseline pain ratings in various chronic pain conditions, your separation between drug and placebo increases. We aren’t sure why this is, but this can make a difference in whether a trial of a truly efficacious treatment could give us a negative result.

Trials also rely on the standard zero to 10 pain scale to measure pain, but most of the time we don’t do much else to help patients figure out which number represents what they feel. The question was raised about whether we should train patients when they do their pain ratings so that they do them more thoughtfully and consistently. We developed a pretty intense training program, so instead of a patient just quickly dashing off a number in the chaos between dinner and bedtime, they are encouraged to go off by themselves and really think about the number that represents the pain they feel so we can have a more accurate pain rating.

What advances in pain clinical trials do you hope to see in the next five to 10 years?

I think we can do better than the zero to 10 scale in trials; it’s been around for decades. There’s got to be a way to make it easier for patients to tell us how much pain they have and how it is affecting them.

Much of the work we’ve done has looked at how to show benefit when a drug is effective, but we also need to think more about safety, risk, and tolerability. It doesn’t matter for the patients how good a drug is in reducing headache, for example, if it makes us so sleepy that we can’t work. In the future, we want to think about more comprehensive outcome measures so we can not only say that this drug is effective for pain, but that it’s safe and patients can tolerate it well. ACTTION is going to devote more time to that kind of benefit-risk calculus.

Another thing we are thinking about is precision medicine and personalized pain treatments. How can we identify which patients are more likely to respond to physical therapy? To a drug? To acupuncture? We are devoting quite a bit of our attention to those questions because they are so important and timely. But this also taps into the benefit-risk analysis. That is because we don’t just want to know that this person will get relief from this drug for their low back pain; we want that person to respond without debilitating side effects or an increased risk of serious adverse events like heart attack. We need to take into account, as we pursue research into personalized pain treatments, not just the benefits of these interventions but also the potential risks.

Finally, also in terms of maximizing overall benefit while potentially avoiding treatment risks, we’d like to focus on prevention of chronic pain; we don’t know enough about it. Why does one individual who throws out their back while playing tennis develop a chronic condition while another person with the same kind of injury gets better in a few weeks? Once we can identify who is at most risk for chronic pain, we could potentially intervene and prevent the problem from ever occurring in the first place. It might be a drug, physical therapy, or a psychological treatment. This preventive piece is going to be an incredibly important focus in the future – I think just as important as precision medicine itself.

How do you feel about winning the John D. Loeser Distinguished Lecture Award?

It’s a huge honor – I’m thrilled and delighted. I can think of no one in the pain community who better exemplifies the critical importance of a truly biopsychosocial perspective in understanding and treating pain than Dr. Loeser, and it has been very gratifying to be able to collaborate with him. But this work really is a team effort and an international effort. Over the years, Dennis and I have worked with dozens and dozens of people who have contributed to all these efforts, both back when we were focusing on IMMPACT and now with ACTTION. This award really should be shared by all the people who have collaborated with us over all these years.

Kayt Sukel is a freelance writer based outside Houston, Texas.