Making an IMMPACT on Pain Research

by Megan Talkington on 3/5/2013

At the 2002 World Congress on Pain, two researchers got to talking. Robert Dworkin of the University of Rochester, New York, US, and Dennis Turk of the University of Washington, Seattle, US, commiserated about the hodgepodge of data coming out at the meeting from clinical trials of new treatments for chronic pain.

“Bob and I were lamenting that it was so difficult to draw conclusions from a lot of the different studies, because the measures used were so different, and the designs were so different,” Turk said.

“We really needed to find a way to get some agreement in the field about how to conduct clinical trials of new treatments for chronic pain,” he said.

So the pair concocted a plan for a consensus meeting. They asked biotech and pharmaceutical companies to sponsor the meeting and to send representatives. They also invited regulators from the US Food and Drug Administration (FDA), as well as representatives from US health research funding agencies. And they brought in academic scientists with expertise in pain trials. The group—27 participants in all—met in November 2002, in Annapolis, Maryland, US.

Dworkin and Turk went into that meeting with a specific goal: to identify a standard set of endpoints for use in late-stage trials of chronic pain treatments. There were precedents for building consensus about trial endpoints in painful diseases including arthritis, cancer, and back pain, but Dworkin and Turk were interested in generating standards that would be useful in any clinical trial, for any chronic pain condition.

Their plan was to have the participants spend two days reviewing the evidence for various endpoints, or “outcome domains,” and hammering out a general agreement. Then they would write a consensus paper and be done.

“Dennis and I were pretty clear that it was a one-off kind of thing,” Dworkin said.

Ten years and 15 meetings later, what is clear is that Dworkin and Turk underestimated the demand for this kind of consensus-building forum in the pain field. The first meeting set off a string of summits called the Initiative on Methods, Measurement, and Pain Assessment in Clinical Trials (IMMPACT). The sessions have tackled subjects from trial design to statistical analysis, and have generated a long list of questions for future meetings—all aimed at boosting the quality and consistency of pain trials.

“They have systematically addressed the most important issues regarding the design of pain clinical trials,” wrote Mark Jensen, also a researcher at the University of Washington in Seattle and a participant in four IMMPACT meetings, in an e-mail to PRF. The result, said Jensen and others, has been a steady stream of consensus papers on best practices that clinical researchers are reading, and heeding.

An evolving agenda

After a decade of IMMPACT meetings, it is hard to find an expert on pain trials who has not attended at least once. Dworkin, Turk, and a steering committee keep the gatherings intimate (approximately 35-45 people), but change the roster each time. All told, well over 200 specialists have participated.

Dworkin and Turk say the participants, collectively, have steered the course. The first meeting recommended six so-called “core outcome domains”—including pain intensity, but also other endpoints like physical and emotional functioning—that investigators should consider using in any pain trial. At the second meeting, the group identified the best tools to measure those domains. “And then no one was really happy with the existing measures,” Dworkin said, so IMMPACT convened a third meeting to discuss ways to develop better measures … and so it went. (Meeting rosters and slides are available at the IMMPACT website. See the table below for a list of the resulting consensus papers.)

“It was an organic, evolving process. We would end almost every meeting with a discussion of what next year’s meeting should be,” Dworkin said.

IMMPACT consensus papers are widely read and referenced by researchers—the first two (Turk et al., 2003; Dworkin et al., 2005) have been cited over 800 times. The real effects on clinical trials are hard to quantify, but participants say they have no doubt that the meetings and papers have made a difference.

Jensen, who serves as editor in chief of the Journal of Pain, said he has observed that, in line with IMMPACT’s recommendations, researchers increasingly measure outcomes other than just pain intensity, and use a more consistent set of measurement tools. “I have not systematically tracked the uptake of all of IMMPACT’s recommendations, but anticipate that if such a review were done, one might see a ‘ripple effect’ with these recommendations having increased uptake over time,” he wrote in his e-mail.

While researchers agree that IMMPACT provides needed consensus on existing best practices, some lament that the group has generally stopped short of proposing innovative new approaches to conducting pain trials. But even getting a common set of benchmarks has been a big step for the field, said Christine Rauschkolb of Johnson & Johnson, Raritan, New Jersey, US, a frequent IMMPACT participant. “Ten years ago, we didn’t even have that,” she said.

Another area where IMMPACT made an important contribution was around the question of what constitutes meaningful pain relief. The group took the position that, even if a reduction in pain or change in another endpoint is measurable and statistically significant, that is not enough to judge a trial positive—the treatment must make a real difference in patients’ lives. To provide guidance, IMMPACT set thresholds for clinically meaningful patient improvement: It defined 50 percent reduction in pain intensity as “substantial” improvement, 30 percent as “moderately important,” and 10-20 percent as “minimally important”; the group also set specific thresholds for other pain-related outcomes (physical and emotional function and overall improvement) (Dworkin et al., 2008).

The recommendations “solidified the idea that in a clinical trial, just looking at statistical significance is not acceptable,” Turk said. “You have to look at the clinical meaningfulness.”

That concept, and the thresholds IMMPACT chose, were not new (see McQuay et al., 2003; Farrar et al., 2001). But IMMPACT’s endorsement has been “very effective” at nudging investigators to apply the standards in their trials, wrote Christopher Eccleston, University of Bath, UK, in an e-mail to PRF. Eccleston, who serves as editor of the Pain, Palliative and Supportive Care Review Group of the Cochrane Collaboration, said that the Cochrane group adopted the same standards for their systematic reviews in parallel with IMMPACT. Eccleston said that dialogue with IMMPACT, and seeing IMMPACT arrive at those standards via an expert consensus approach and high-profile discussion, reinforced Cochrane’s decision (see Moore et al., 2010, and the “Author and Referee Guidance” document at the Cochrane Pain website).

In the end, IMMPACT’s most important contribution may be its influence on the culture of analgesic trials. The meetings created a “neutral ground” where participants were not caught up in debating the fate of a particular drug, but were working to advance pain science to help get effective therapies to patients, Rauschkolb said. “We could explain as an industry where our problems and hurdles lie, and make that transparent to people on the academic side and on the regulatory side—and vice versa.” As a result, “We finally speak the same language,” she said. “That’s a major accomplishment.”

“When I started out in this area, it was a mess,” said Bob Rappaport, Director of the FDA’s Division of Anesthesia, Analgesia, and Addiction Products, and an IMMPACT participant from the outset. “People didn’t know how to study pain properly, everybody disagreed, and nobody was talking to anybody else.” IMMPACT has transformed the situation, he said, because it gives him and his colleagues at the FDA and at companies a common set of principles for how to conduct a clinical trial—principles that they are working out together.

Ready for ACTTION

In 2010, Rappaport put out a call for researchers to spearhead a new public-private partnership to improve clinical trials in pain. The program would open up the FDA’s database of existing trials to researchers, with the goal of determining how study design and conduct affect outcomes. Dworkin and Turk won the funding to direct the program, which now covers analgesics as well as anesthetics and addiction treatments, and is called ACTTION (Analgesic, Anesthetic, and Addiction Clinical Trial Translations, Innovations, Opportunities, and Networks). In 2011 IMMPACT became part of the new program’s portfolio.

Besides IMMPACT, ACTTION now comprises 13 initiatives, each with its own punchy acronym. One working group, for example, has created a standard database format for clinical trial information that will allow researchers to conduct pooled analyses across trials (Dworkin et al., 2012). Another, conducted jointly with the American Pain Society, is developing standardized diagnostic criteria (a pain “taxonomy”)—in part to facilitate comparisons between trials. In the new regime, IMMPACT continues as a yearly consensus meeting in its original form, with the exception that its funding—which originally came entirely from industry—now comes from industry, the FDA, and other sources.

Meanwhile, IMMPACT is still evolving, taking up issues that were barely on the radar in 2002. At the 2012 meeting, the group considered the role of “objective” pain measures (biomarkers) like brain imaging, quantitative sensory testing, and skin biopsy in pain drug trials. The upcoming meeting in June 2013 will look at approaches to phenotyping study subjects, including the use of patient-reported measures, clinician ratings, and quantitative sensory testing, with the goal being to identify pain mechanisms and predict treatment response.

IMMPACT participants are quick to list a number of topics that could easily keep the group busy for another 10 years. Some would like to see the initiative, which until now has focused on pharmacological treatments, shift its gaze to different modalities. Turk, for example, said he would like to have the group explore trial designs for non-pharmacological treatments like physical and cognitive-behavioral therapies. Those trials “often have lots of challenges when it comes to things like blinding that would be well worth having a meeting on,” he said.

But pain researchers—including Dworkin and Turk—will have to wait until after the June meeting to see what issue IMMPACT will take on next. If this year’s meeting is anything like previous ones, they say, the winner will emerge by the end of the two-day summit. Stay tuned.

Image credit: © Rfischia | Dreamstime.com