Maksim Kabakou/shutterstock.com
LONDON—Josef Smolen, MD, chair of rheumatology at the University of Vienna and former president of the European League Against Rheumatism (EULAR), expressed a “personal disappointment” in the development of useful biomarkers in the treatment of rheumatoid arthritis (RA).
Even though a good portion of his life’s work has been researching biomarkers to help with targeting therapies, he lamented that the field has been full of promises that have mostly ended up dashed.
“We have to do better, from my perspective,” said Dr. Smolen, speaking in a scientific debate at the 2016 EULAR Annual Congress. His debate counterpart—Ronald van Vollenhoven, MD, PhD, director of the Amsterdam Rheumatology and Immunology Center—was not particularly effusive about the progress of RA biomarkers, either, but painted a brighter picture.
Dr. Smolen said the hopes for biomarkers have been high.
“We want to predict what’s going on,” he said. “We want to predict new treatment targets. … We want to predict bad outcome; we want to predict response to therapy. We want to know who will respond to which agents having four biological principals, plus jakinibs,” which all yield essentially similar results.
He pointed to the example of IL-17 inhibitors. Those who don’t respond to anti-tumor necrosis factor drugs tend to have increased levels of IL-17.
“What happens if we use IL-17 inhibitors, especially in patients who fail TNF inhibitors?” Dr. Smolen said. But the results, even in these patients, were disappointing.1
“IL-17 looks like a biomarker of chronic inflammation, its levels are even elevated in patients who are inadequate responders to TNF inhibitors,” Dr. Smolen said. “But it fails as a target.”
Failed Potential Biomarkers
The list of potential biomarkers that have produced similarly dashed hopes is a long one, among them the IL-2 receptor, CD4 cells, IL-12 and IL-23.
“All of these have failed as therapeutic targets, even though we would have sworn that each one of them would bring a new era into rheumatoid arthritis treatment,” Dr. Smolen said.
The result is that, for the past 20 years, the field has relied on the same old markers: mainly swollen joint count, C-reactive protein, rheumatoid factor and anti-citrullinated protein antibody.
MBDA
One of the latest attempts to use biomarkers—the 12-component multiple biomarker disease activity score (MBDA), which is approved by the FDA for monitoring disease activity—has been called into question by some studies.
A 2016 study looked at data from the AMPLE trial of abatacept vs. adalimumab. The response graphs using DAS28-CRP are superimposable for the two drugs. But using the MBDA, adalimumab shows a greater effect, Dr. Smolen noted.
“Apparently the fact that a TNF inhibitor interferes with the pro-inflammatory cytokine shuts down some of the components in the body that are measured by the MBDA more than abatacept—even if it doesn’t mean anything clinically or radiologically,” he said.
Dr. Smolen added that patients with high disease activity should be expected to have the lowest rate of no radiographic progression. That has been shown to be the case using the Simplified Disease Activity Index (SDAI), but when using the MBDA, the results partially reverse, Dr. Smolen said.
“So is such a biomarker appropriate to inform us about disease activity and correlations?” he asked. “Disease activity assessment appears to be the far better way to go.”
The list of potential biomarkers that have produced similarly dashed hopes is a long one, among them the IL-2 receptor, CD4 cells, IL-12 & IL-23.
The Search Continues
He expects new biomarkers to emerge eventually, but they must be legitimate, he said.
“The search must go on,” he said. “We are running biomarker trials currently. And we will actually find markers. But we will have to find the right ones that really distinguish what we want to distinguish. And we need to ask the right questions.”
Dr. van Vollenhoven said he understands the frustration surrounding the slow progress in the search for biomarkers.
“You can always say, ‘Biomarkers are just around the corner,’ [but] people get tired of that,” he said. “What we were hoping for was that we could find biomarkers that could guide us more in terms of treatment.”
That doesn’t mean the field should give up the search.
The MBDA has its merits, he said. An analysis of data from the SWEFOT trial, which compared conventional with biological treatment in early RA, showed that those with a low or moderate MBDA score at baseline have hardly any radiographic progression after two years, while those with a higher score had what he called “a substantial risk,” with about one in five patients experiencing radiographic progression.
This information could help clinicians discuss patients’ prospects and potentially lead to more intense treatment in some cases, said Dr. van Vollenhoven, who reported working as a consultant for Crescendo Bioscience, the developer of the MBDA test.
Also, he said, the MBDA has been better than traditional measures at predicting which patients are more likely to respond to triple therapy as opposed to anti-TNF therapy. Those data are expected to be published in a forthcoming paper, he said.
“Biomarkers will increasingly help achieve personalized treatment in rheumatology,” he said.
Role of Dependent Variables
Session moderator Gary Firestein, MD, a rheumatologist and professor of medicine at the University of California San Diego, said that in a multi-component test, “the components should be independent. And this is one of the issues that raises a lot of questions about utility here. There are at least three or four different components of the MBDA assays that are dependent variables.” For example, if IL-6 goes down, CRP and other variables are likely to go down, as well.
Dr. van Vollenhoven said the 12 components of the MBDA were chosen out of 180 component candidates in order to find the best correlation with DAS28.
He said it’s difficult to have a composite test that matches up with traditional clinical measures while also remaining distinct from those measures.
“You can say, ‘I want the biomarker that correlates to 100% with the clinical marker—but that doesn’t tell me anything else than what you already know from the clinical marker.’ Or you can have a biomarker that’s completely different from a clinical marker, and you say, ‘Gee, I don’t think I can trust that. That seems such a strange result.’
“You can’t have it both ways.”
Thomas R. Collins is a freelance medical writer based in Florida.
References
- Genovese MC, Greenwald M, Cho CS, et al. A phase II randomized study of subcutaneous ixekizumab, an anti-interleukin-17 monoclonal antibody, in rheumatoid arthritis patients who were naive to biologic agents or had an inadequate response to tumor necrosis factor inhibitors. Arthritis Rheumatol. 2014 Jul;66(7):1693–1704.
- Fleischmann R, Connolly SE, Maldonado MA, et al. Estimating disease activity using multi-biomarker disease activity scores in patients with rheumatoid arthritis treated with abatacept or adalimumab. Arthritis Rheumatol. 2016 Apr 25.
- Hambardzumyan K, Bolce R, Saevarsdottir S, et al. Pretreatment multi-biomarker disease activity score and radiographic progression in early RA: Results from the SWEFOT trial. Ann Rheum Dis. 2015 Jun;74(6):1102–1109.