Signals in the Noise: Biomarkers in Rheumatoid Arthritis

During the past decade, investigators studying rheumatoid arthritis have begun to look more closely at biomarkers of disease severity and progression. They are evaluating cytokines in the serum compartment, adipokines in the fat compartment and looking at calproteins from synovium and markers of angiogenesis. The net has been cast over barometers of blood vessel growth, including vascular endothelial growth factor, and over markers of collagen degradation like matrix metalloproteinases.

Studies have assessed the role of the oligomatrix protein in cartilage degradation, while others have focused on measures of bone turnover, bone resorption and bone formation. Novel markers, such as 14-3-3 eta, also have been studied. However, the perfect biomarker — or combination of biomarkers — has not yet emerged. Walter P. Maksymowych, MD, professor in the Department of Medicine and Division of Rheumatology at the University of Alberta in Canada, said part of the challenge lies in conducting research.

“To do a good biomarker study, you need to have very well systematically collected observational cohorts,” he said. “If you were to ask a rheumatologist, ‘How do you do a clinical trial in rheumatoid arthritis (RA)?’ most rheumatologists would say you do a 12-week placebo-controlled study, then you have an open-label crossover, you use an ACR20 to measure responses, etc. If you ask a rheumatologist, ‘How would you validate a clinically useful biomarker?’ I think you would get a lot of blank stares.”

Gregg J. Silverman, MD, professor in the Departments of Medicine and Pathology and director of the Laboratory of B Cell Immunobiology at the NYU Langone School of Medicine, suggested it is always helpful to keep an eye on the clinical utility of a new assay reported in the latest biomarker research report.

“I once gave a presentation on biomarkers and someone in the audience said, ‘Biomarkers, what are biomarkers? We used to have tests. How are these different than tests?’” he said. “You need to understand that as time has evolved, we are using them in a number of different contexts. Some biomarkers are being developed to aid diagnosis, others to give you an idea of prognosis and the chance of developing tissue injury, others for sub-setting patients with the goal of being able to better decide the optimal therapeutic regimen for an individual.”

In this Cover Story, Healio Rheumatology outlines where rheumatologists stand with biomarkers in RA and how these markers may guide therapy moving forward.

Biomarker Signatures

McArdle and colleagues recently published a study that outlined biomarker signatures in use today.

“Currently, there is a major international effort driven by the Outcome Measures in Rheumatology Clinical Trials and the Group for Research and Assessment in Psoriasis and Psoriatic Arthritis well underway in RA,” they wrote. However, they suggested that because of the heterogeneous nature of the disease, it has proven a challenge to find a single marker that is predictive of damage. “Consequently, there is great interest in identifying a panel of biomarkers that could be incorporated into a signature predictive of disease progression,” they noted.

The researchers suggested a signature could signify joint destruction. They reviewed data sets pertaining to markers of acute phase response, autoantibodies, cytokines, chemokines, adipokines, calprotectin, markers of angiogenesis, products of degradation and enzyme mediators of destruction.

Increased erythrocyte sedimentation rate and hepatocyte production of acute phase proteins, including C-reactive proteins (CRP) and acute phase serum amyloid, may be surrogate biomarkers of inflammation in the synovium, according to McArdle and colleagues. The researchers suggested investigation of acute phase serum amyloid as a correlative factor with joint damage in systemic or local inflammation may be a useful avenue for research.

Rheumatoid factor (RF) and anti-citrullinated peptide antibody (ACPA) are autoantibodies that may predict outcomes in RA, according to the authors. Erosions have been associated with these biomarkers, while positivity for ACPA and negativity for RF have resulted in progression to severe disease compared with patients who were positive for RF and negative for ACPA. Positivity for both may yield the greatest risk for disease progression. “ACPA emerge as the most useful indicator of joint damage,” they wrote.

Regarding cytokines and chemokines, interleukin (IL)-6, IL-16, IL-22, IL-33, chemokine ligand (CCL)11 and chemokine (C-X-C) motif ligand (CXCL)13 may predict radiographic progression, according to McArdle and colleagues. However, they noted certain data sets involving these cytokines and chemokines have demonstrated conflicting results and should undergo further investigation.

Adiponectin, leptin and visfatin are adipokines that have shown some association with radiographic progression, according to the McArdle and colleagues. Regarding calprotectin, they suggested a correlation between this biomarker and 10-year radiographic damage may exist. McArdle and colleagues also suggested vascular endothelial growth factor may be the most important mediator of angiogenesis and, therefore, correlate with joint damage in RA.

In terms of products of collagen degradation, collagen type II degradation product epitopes C2C and C1,2C may reflect bone and cartilage damage, according to McArdle and colleagues. Cartilage oligomatrix protein may have a similar effect, they added. CTXI and CTXII also may be associated with long-term radiographic progression.

Matrix metalloproteinase proteases (MMPs), such as MMP-1 and MMP-3, have demonstrated correlations with significant radiographic progression, according to McArdle and colleagues. Specifically, they suggested elevated levels of these enzyme mediators of destruction at baseline may yield this result.

14-3-3 eta

“The one that has been discovered that is showing pretty good promise is the so-called 14-3-3 eta,” said John R.P. Tesser, MD, of Arizona Arthritis & Rheumatology Associates. “It has been getting increasing interest over the last, I would say, handful of years, and has been showing some help in identifying patients with RA and may correlate with disease and erusive disease.”

The 14-3-3 protein is a major proinflammatory mediator, according to Maksymowych.

“It causes the expression of a lot of inflammatory molecules and molecules that lead to cartilage damage and erosion of bone,” he said. “The 14-3-3 story is just in its infancy, but I think we are going to hear a lot more about it as time goes on.”

Maksymowych said there are still hurdles to clear.

“It is a new test that deals with an aspect of biology that rheumatologists are not familiar with,” he said. “It is a completely new protein and not something that is part of our training and understanding of the pathophysiology of disease.”

Maksymowych first grew interested after meeting Aziz Ghahary, MSc, PhD, PDF, a professor in the Division of Plastic Surgery, Department of Surgery at the University of British Columbia, who studied patients with burns who developed scar tissue. Ghahary had isolated a factor that regulated collagen production through the expression of certain degrading enzymes, metalloproteinases.

“He had isolated this factor and had some evidence it was associated with the production of metalloproteinases,” Maksymowych said. “I suggested to him that he have a look at synovial fluid of patients with inflammatory arthritis because metalloproteinases are relevant to the joint degradation that occurs in inflammatory arthritis.”

They eventually found factor 14-3-3 eta is one of seven isoforms.

“There are seven different types of the 14-3-3 protein,” Maksymowych said. “It is typically present inside cells, so [it is] a chaperone protein. But one of these isoforms, 14-3-3 eta, for reasons that we still do not understand, is secreted into the extracellular fluid. It is also present in the synovial fluid.”

The protein is also present in the blood of patients with RA, according to Maksymowych.

“We also know that, just as in burn scars, it is strongly associated with metalloproteinase production, but it does a lot more than that,” he said. “It is also associated with the production of other inflammatory molecules that are typically seen in large quantities in the inflamed joints of patients with RA. The peculiarity of this specific isoform is that it is not detected in any other condition.”

Maksymowych said it is unknown what triggers production of this isoform. He added the test is different from the rheumatoid factor and anti-cyclic citrullinated peptide (CCP) tests, which detect antibodies associated with rheumatoid arthritis.

CCP, CRP and MMP-3

Leslie R. Harrold, MD, MPH, associate professor of medicine and orthopedics at the University of Massachusetts Medical School and senior medical director of Pharmacoepidemiology and Outcomes Research at Corrona LLC, and colleagues assessed the impact of anti-CCP and rheumatoid factor status on treatment response separately in a cohort of 566 patients beginning therapy with abatacept (Orencia, Bristol-Myers Squibb) and 1,715 patients who were initiating tumor necrosis factor inhibitors (TNFi). Positivity for both anti-CCP and RF results was associated with improved response rates compared with negativity for both. Response rates were defined as disease activity measures and achievement of remission. Among patients treated with abatacept, those with single-positive status were more likely to reach remission than those with double-negative status, according to the findings. No significant differences were reported for the anti-CCP or RF groups among patients treated with TNFi inhibitors. Investigation of the relationship between anti-CCP antibodies and RF revealed that seropositivity to anti-CCP antibodies alone drove responses. Rheumatoid factor failed to predict response when observed as an independent factor.

“These findings are exciting, as anti-CCP antibodies are a marker of disease severity and detectable early in the course of the disease,” Harrold said. “A better understanding of the relationship between anti-CCP antibodies and treatment response has the potential to advance patient care. Specifically, patients with RA can spend years trying different treatments until their disease is properly controlled.”

Duran-Avelar and colleagues investigated the STAT4 rs7574865 G/T polymorphism in RA disease activity and anti- (CCP) antibody levels in a cohort of 140 patients in Mexico. The study also included 150 controls. Results indicated the GT and TT genotypes, along with the T allele from rs7574865, increased risk factors for RA regardless of anti-CCP status.

“This work supports the association of the STAT4 rs7574865 polymorphism with RA and disease activity, but not with anti-CCP antibody levels in a Mexican population,” the researchers concluded.

“Many clinicians are using CCP,” Silverman said. “The issue is whether they are using them to their fullest and thinking about what a high titer vs. a low titer represents.”

For Maksymowych, it is a matter of building on a long-standing knowledge base.

“We still rely on CRP, which is about 70 [years] or 80 years old now,” he said.

Takeshita and colleagues suggested CRP and MMP-3 require further investigation as biomarkers in RA. They took serum from 24 patients with RA.

“Intense signals were observed on a sialic acid-binding lectin (Agrocybe cylindracea galectin [ACG]) and O-glycan-binding lectins (Jacalin, Agaricus bisporus agglutinin [ABA], and Amaranthus caudatus agglutinin [ACA]) by applying subnanogram levels of serum MMP-3,” they wrote. “ACG, ABA and ACA revealed differences in MMP-3 quantity, and Jacalin revealed differences in MMP-3 quality.” Because of these results, they suggested ACG-Jacalin is associated with disease activity and that glycan change at a local lesion may be a tool for assessing disease activity. In addition, the degree of serum MMP-3 alpha-2,6-sialylation may also have utility as a way of measuring disease activity, according to the results.

Tesser put the findings from Duran-Avelar and Takeshita into context.

“CRP itself is a protein that promotes inflammation,” he said. “But, it is not at all specific. You can see it elevated in rheumatic diseases, but also in infections and in cancers.”

Silverman is confident that understanding of many of these biomarkers, including CCP, is increasing.

“If a patient has high titer CCP, we have to try another therapy to provide adequate improvement of tender and swollen joints,” he said. “In general, I also think that having an appropriate disease activity measure you use routinely with each patient and target removes a lot of the uncertainty.”

Harrold said further study is needed beyond 6 months to see how seropositivity increases treatment outcomes.

It is because of the research that these biomarkers are becoming more widely used, according to Silverman. Moreover, with more clinicians employing them, understanding is increasing.

“There are a number of variations on that scene for other kinds of citrullinated proteins and other ligands,” he said. “The CCP-2 and CCP-3 commercial tests were designed to pick up lots of antibodies to different citrullinated specific proteins.”

Silverman added that investigators have begun to look at specific proteins, such as alpha enolase or vimentin.

“There have been claims that these may be more specific or focus more on disease activity,” he said. “The problem is that many individuals, even if they are CCP-2 or CCP-3 positive, will be negative for these specific antigens. So they may have some utility, but not as generally useful as the commercial test. Similarly, there is PADI4, which is peptidyl arginine deaminase 4, that has been thought to be a marker for more aggressive disease. But it is not present in everybody.”

Cytokines

Osiri and colleagues compared the effects of treatment with combination disease-modifying antirheumatic drugs or biologics on cytokines, disease activity and function. The study included serum samples from 81 patients. Serum concentrations of IL-1beta, IL-2, IL-4, IL-6, IL-8, IL-10, IL-17A, IL-23, IL-33, interferon, granulocyte monocyte-colony stimulating factor and TNF in patients treated with combination DMARD therapy were not significantly different from patients treated with biologics, according to the results. In particular, IL-33, IL-8 and IL-6 carried the strongest correlation coefficients.

“The cytokine profiles in established RA were mainly those produced from effector cells, especially IL-6, IL-8 and IL-33,” the researchers concluded. “Both IL-8 and IL-33 may be potential biomarkers and/or treatment targets in patients with late RA.”

“Taking cytokines as an example, whether it be IL-17, IL-1, IL-6, IL-12-23, none of those are used for diagnostic purposes,” Tesser said. “We have been precise as to, when you are talking about a biomarker related to a disease, what category of these considerations are you talking about. These include diagnostic, disease activity related, or prognostic coinsiderations, especially for radiographic progression.”

Gualtierotti and colleagues investigated the correlation of IL-6 in 15 patients with RA treated with tocilizumab (Actemra, Genentech) an IL-6 receptor inhibitor. They compared outcomes with those found in 40 controls. Patients with RA had higher baseline plasma levels of CRP (P = 0.0001), IL-6 (P = 0.043), sIL-6R (P = 0.003), TNF-alpha (P = 0.0001), F1+2 (P = 0.0001) and D-dimer (P = 0.002) than individuals in the control group. Four-week results indicated reductions in DAS28 (P = 0.0001), erythrocyte sedimentation rate (ESR) (P = 0.00001), CRP (P = 0.014), TNF-alpha (P = 0.006), F1+2 (P = 0.009) and D-dimer (P = 0.04) in the patient arm.

“The reduction of prothrombotic biomarkers parallels the reduction of inflammatory parameters and clinical symptoms in RA patients treated with tocilizumab, both 4 weeks after the first administration and during maintenance therapy,” the researchers concluded.

Multi-biomarker Measures

Many experts believe the way forward is to develop tools that assess a number of biomarkers. Hirata and Tanaka investigated the multi-biomarker disease activity (MBDA) score, which includes 12 serum biomarkers. These are VCAM-1, EGF, VEGF-A, IL-6, TNF-RI, YKL-40, MMP-1, MMP-3, leptin, resistin, SAA and CRP, according to their results.

“The MBDA score not only reflects disease activity in RA, but also is predictive for radiographic progression and risk of flare after drug reduction,” they wrote.

“One of the things that has generated a lot of interest, study, criticism and debate is the concept of looking at a multi-biomarker assay,” Maksymowych said. “It includes a conglomeration of 12 markers that you would maybe not look at individually. This commercially available multi-biomarker assay has demonstrated clinical significance related to several of these things we are talking about.”

These measures include disease activity or severity, which is another way of saying inflammatory response, according to Maksymowych.

“It is the inflammatory aspect of the disease that leads to the damage. It has had, as I said, a body of evidence from a number of studies, virtually all retrospective, to show this is the case.”

Maksymowych said the 12-marker assay may improve on CRP alone.

“It also has predictive value for progression radiographically,” he said. “You can parse that out. It has demonstrated statistically a significant correlative value of present radiographic disease and the ability to predict radiographic progression.”

The OMERACT International Biomarker Working Group, which is chaired by Maksymowych, has published recommendations and guidelines for the clinical validation of biomarkers. Most notably, they are conducting a study called BIODAM, which was started in 2011. It is taking place in 10 countries in western Europe, Israel and North America. The researchers are collecting biosamples every 3 months from patients with RA.

“The patients are being followed in a systematic manner, similar to what would be done in a clinical trial,” he said. “We are also collecting X-rays.”

The study will yield some 100,000 biosamples and 3,000 radiographs, according to Maksymowych.

“That will allow us to deal with rigorous biomarker validation studies not only for 14-3-3, but for others, as well,” he said. “This kind of knowledge base is crucial if we are going to move ahead with precision medicine.”

Guiding Therapy

“The only thing that matters, as far as I am concerned, about defining a subset, is whether it helps you figure out what agent the patient is more likely to respond to,” Silverman said. “If you can better craft your bets of what agent to start with, I think it will be more cost-effective and patients will have more effective therapies sooner.”

Maksymowych built on this point.

“We all want to know which patients are destined to develop the more severe disease so we know who will benefit from early intervention strategies,” he said. “This is what we should be moving toward. We identify these patients early and we get them into remission early so that we might even be able to taper them off biologics early and keep them in remission on a long-term basis.”

More studies of the kind conducted by Harrold and colleagues will serve to push the field forward, according to Harrold.

“Both RA patients and their providers would benefit from a biomarker that would identify the subsets of patients who would respond to certain specific treatments,” she said.

Tesser added, “In fact, this would be the holy grail of therapeutics in RA, or at least until we might find a cure.”

Further understanding of the impact of TNFi was provided by the findings of her study group, which may be beneficial for the clinical community overall.

“The effects of TNFi are not dependent on the ACPA antibody status,” she said. “However, the outcomes of patients receiving the T-cell costimulation modulator abatacept were dependent on ACPA status, with better responses observed in those who were positive for anti-CCP antibodies compared to those who did not have these antibodies.”

For Maksymowych, clinicians may be able to alter the natural history of the disease.

“That is the kind of strategy that is actively explored in the field right now,” he said. “What limits us is this lack of prognostic markers in our practice.”

Despite these limitations, Silverman is optimistic.

“Things are moving in the right direction,” he said. “We are making advances in understanding how genetic susceptibility translates into the autoimmunity. But, we are not good at treating the autoimmune aspects of RA. So there is room for improvement.”

Maksymowych said it is a simple matter of uncertainty.

“How exactly all these old biomarkers are going to help us is still unclear because we are getting so much recent information,” he said, referring to the vast body of information being gathered in the BIODAM study. “We have been patiently collecting this data across countries since 2011. We are going to be finished with all of that in a year. So that tells you how long this sort of initiative takes. I think this is why there are some pessimists because this is the kind of data collection and the kind of study that needs to be done to allow the field to move forward.” – by Shirley Pulawski and Rob Volansky

• References:
• Duran-Avelar MJ, et al. Clin Rheumatol. 2016 May 28. [Epub ahead of print].
• Gualtierotti R, et al. Clin Exp Rheumatol. 2016;Apr 15. [Epub ahead of print].
• Hirata S, et al. Nihon Rinsho Meneki Gakkai Kaishi. 2016;doi:10.2177/jsci.39.37.
• McArdle A, et al. Arthritis Research & Therapy. 2015;doi:10.1186/s13075-015-0652-z
• Osiri M, et al. Clin Rheumatol. 2016; May 17. [Epub ahead of print].
• Takeshita M, et al. Arthritis Res Ther. 2016;doi:10.1186/s13075-016-1013-2.

• For more information:
• Leslie R. Harrold, MD, MPH, can be reached at the University of Massachusetts Medical School, 55 Lake Ave., North Worcester, MA 01655; email: leslie.harrold@umassmed.edu.
• Walter P. Maksymowych, MD, can be reached 562 Heritage Medical Research Center, University of Alberta, Edmonton, Alberta, Canada T6G 2S2; email: walter.maksymowych@ualberta.ca.
• Gregg J. Silverman, MD, can be reached at 450 East 29th St., New York, NY 10016; email: gregg.silverman@nyumc.org.
• John R.P. Tesser, MD, can be reached at Arizona Arthritis & Rheumatology Associates, 9305 W. Thomas Rd., Suite 505, Phoenix, AZ 85037; email: jtesser1@cox.net.

Disclosures: Harrold reports she is an employee of and has stock options with Corrona LLC and receives grant funding from Pfizer. Maksymowych reports he is a co-inventor of 14-3-3, receives honoraria from Augurex and QUEST and royalty payments around the discovery of 14-3-3. Silverman reports he is a consultant for Celgene, Genentech, Eli Lily, Pfizer and Roche. Tesser reports he is a speaker for Crescendo Bioscience, makers of VECTRA DA, the MBDA Score.