College of American Pathologists
Printable Version

  Feature Story


cap today

Assay takes arthritis out of the gray zone

June 2003
William Check, PhD

Legend has it that Archimedes leapt from his bath and ran dripping through the streets of Athens crying “Eureka!” when he devised a method to differentiate pure from adulterated gold. An equally enthusiastic, though less flamboyant, reception is being accorded now to a new laboratory test that accurately differentiates rheumatoid arthritis from non-rheumatoid causes of arthritic joints.

The test, which measures antibodies to a peptide that contains the atypical amino acid citrulline, “has sensitivity similar to rheumatoid factor, but its specificity appears to be a good deal better,” says Henry Homburger, MD. Dr. Homburger is professor of laboratory medicine at Mayo Medical School and director of the immunology, antibody, and cellular immunology laboratory at Mayo Clinic, Rochester, Minn. “Data suggest that increased specificity is the major advantage of this test,” he says.

Enthusiasm for the anti-cyclic citrullinated peptide, or anti-CCP, assay can be attributed in part to the timing of its arrival. “Anti-CCP is favored because of recent clinical developments in the treatment of rheumatoid arthritis,” Dr. Homburger says. Earlier, more aggressive treatment is being advocated more widely. “With these changes in therapeutic strategy, clinicians need a stronger conviction that a patient has RA,” he says. A positive anti-CCP result can bolster that conviction.

Mark Wener, MD, director of the Immunology Division in the Department of Laboratory Medicine and a rheumatology attending at the University of Washington School of Medicine, Seattle, says anti-CCP appears to be “perhaps not quite as sensitive as rheumatoid factor, but more specific.”

“And anti-CCP antibodies are present in some RA patients who do not have rheumatoid factor. However,” Dr. Wener cautions, “anti-CCP certainly doesn’t replace history and physical examination. I wouldn’t use it as a screening test.”

Rheumatoid factor and anti-CCP tests have prognostic value as well and identify a subset of patients likely to have more severe disease. “When both are positive,” Dr Wener says, “that gives 99+ percent specificity and provides support for a clinical impression to treat these patients earlier and more aggressively with the more potent and more expensive medications now available.”

Positive reviews also come from physicians in Europe, where the anti-CCP assay was developed and first evaluated. Nicola Bizzaro, MD, vice director of the laboratory of clinical pathology at the City Hospital of St. Dona di Piave, Italy, pronounces himself “very satisfied” with this test. “When it is positive, it is very strongly positive,” he says. “I have worked in the laboratory for 25 years and never saw a test that is so discriminating. There are very few cases in the gray zone.”

Says Olivier Meyer, MD, head of the rheumatology department of the Hôpital Bichat, Paris: “I think this antibody is very useful for early diagnosis of RA. Most papers say, and we have confirmed, that it is positive in about 70 percent of patients with early arthritis—I mean less than six months’ duration.” Dr. Meyer agrees it is important to make an accurate diagnosis of RA early to initiate early treatment and improve prognosis.

Robert Morris, MD, practiced rheumatology for 25 years and is now president and laboratory director of Rheumatology Diagnostics Laboratory in Los Angeles and clinical associate professor of medicine at UCLA. “I have not seen any test quite like anti-CCP,” he says. “I am very enthusiastic about it.” When talking to clinicians, Dr. Morris finds they are often “extremely gratified” by the anti-CCP result. Sometimes a positive result convinces a reluctant patient—one with an atypical presentation—to go on aggressive therapy.

However, a prominent British rheumatologist sounds a note of moderation. “Based on published data, it looks as if anti-CCP is like rheumatoid factor but becomes positive sooner and is rather more specific,” says Deborah Symmons, MD, professor of rheumatology and musculoskeletal epidemiology at the University of Manchester. “I must admit to being slightly anxious to see commercial people trying to sell kits saying if it is positive it will be RA and if it is negative it won’t,” she says.

Dr. Symmons emphasizes the need to make a diagnosis in the context of the clinical picture. “The patient needs to have arthritis before you use a test,” she says. “So you cannot use it to screen a general population.” Even a highly specific test will generate many false-positive results if used in a population that does not have a reasonably high prior probability of having RA. Dr. Symmons finds that some nonspecialists use RF inappropriately. They might measure RF in a person with back pain and, if it is positive, send that patient to the RA clinic. She worries that anti-CCP will be similarly abused.

It has been known for some time that RA patients make

autoantibodies, including RF, antiperinuclear factor, antifilaggrin, and anti-SA. In 1996, Walther J. van Venrooij, PhD, professor of biochemistry at the University of Nijmegen, the Netherlands, discovered the molecular basis of these autoantibodies through what he calls “a series of intelligent and pragmatic guesses.” Dr. van Venrooij reasoned that, if a patient makes antibodies against her own proteins, then tolerance has to be broken. “One simple idea about breaking tolerance,” Dr. van Venrooij says, “is that you need an extra agent, a hapten.” At that time he was studying a protein that was citrullinated; that is, some of its arginine residues had been enzymatically converted to citrulline. “We tested whether citrulline might be such a hapten,” he says. “The result was positive. And it was completely

specific for RA.”

During synovial inflammation, many cells die and many key proteins are enzymatically modified, some by citrullination. If clearance of dying cells is inadequate, citrullinated proteins are released into the synovium and antibodies are made locally. These autoantibodies form immune complexes and help make inflammation more chronic. “The citrullination process is not studied well,” Dr. van Venrooij says, “so we don’t know yet why antibodies to citrullinated proteins are found only in RA.”

Development of Dr. van Venrooij’s discovery was financed by a Dutch academic consortium, STW, which holds the patent to the synthetic cyclic peptides that form the basis of the anti-CCP assay. Dr. van Venrooij himself receives no royalties; however, his research group does. Licenses to the cyclic peptides were granted to Euro-Diagnostica of Arnhem, the Netherlands, for European sales, and Axis-Shield Diagnostics of Dundee, Scotland, for U.S. sales. Through exchange of rights, both companies are marketing in the United States, Euro-Diagnostica through Inova Diagnostics and Axis-Shield through Scimedx, Diasorin, and Bio-Rad.

Specificity results of Dr. Homburger’s in-house studies on the anti-CCP assay are representative of the many published studies, most from Europe. He tested three groups of non-RA patients:

  • Normals, who had no anti-CCP false-positive results.
  • Patients previously tested for RF, among whom the anti-CCP false-positive rate was three percent, compared with a false-positive rate of almost 40 percent for RF.
  • Patients positive for antinuclear or anti-DNA antibodies, with an anti-CCP false-positive rate of about 10 percent (RF was not done in this group).

In the second group, an order for an RF test was taken as an indication

that the clinician suspected RA, which, Dr. Homburger says, makes it

“the likely target group for anti-CCP testing.” The 97 percent specificity in this group was promising. The third group was obviously weighted with patients with connective tissue diseases, who can also present with symptoms of polyarthritis.

“In the written material we send to clinicians,” Dr. Homburger says, “we will say that false-positive results with this test are possible and will be higher in patients with other autoimmune diseases, particularly connective tissue diseases. But that we would anticipate a much lower false-positive rate

than with RF.”

Dr. Homburger has evaluated both commercial kits and recommends that other laboratories do the same. “The citrullinated peptides used in all commercial kits are the same, and I don’t think you will see differences [between kits] if you compare normals to RA patients,” he says. “But you might if you broaden test subjects to those actually seen in clinical practice,” such as patients with other connective tissue diseases or other forms of arthritis. “Our comparison of these kits showed differences in the dose-response curves due to other kit components, which will be reflected in the results reported to clinicians.”

A U.S. rheumatologist who evaluated the anti-CCP assay is

Peter Schur, MD, senior physician in the Department of Medicine and medical director of the clinical immunology laboratory at the Brigham and Women’s Hospital and professor of medicine, Harvard Medical School, Boston. He verified that the sensitivity of anti-CCP is about the same as RF and that its specificity for diagnosing RA is better.

“We took the first 200 sera that came to the laboratory from the arthritis center, patients being seen either for early evaluation or for followup,” he says. It was a diverse set of patients, since his laboratory does all of the rheumatology testing, such as antinuclear antibody, anti-DNA, anti-ENA, C-reactive protein, RF, immunoglobulins, and complement. Chart review showed “few positives” in non-RA patients, he says. “We thought it performed very well.” After going through the hospital’s administrative procedures, the test is now approved for routine use.

At Charing Cross Hospital, London, Peter Charles, FIBMS, chief biomedical scientist in the Division of Immunology, has done a number of studies of anti-CCP in RA and other diseases. “We confirmed the original data from Dr. van Venrooij’s laboratory,” he says. “It is a highly specific marker for RA and doesn’t appear in many other patients. Also, it is very sensitive.” Charles calls anti-CCP “a good diagnostic test.” To verify its specificity, his laboratory recently completed an audit of all anti-CCP-positive patients. Among the 50 to 60 patients, only one was judged on review by a clinician not to have RA at that point.

Charles notes that the sensitivity of anti-CCP may depend on the population studied. “Hospital specialists tend to see the more severe end of the disease spectrum,” he says. “Those patients are more likely to be anti-CCP positive.” He compares it to HLA-DR4, a marker for disease severity reported to be present in 60 to 70 percent of RA patients seen in the hospital, but in only 30 to 40 percent in the community.

Dr. van Venrooij, too, has observed that sensitivity depends on the cohort studied. “Our experience is that every cohort is different,” he says. In very early-stage patients who will develop RA much later or more slowly, sensitivity may be lower. “What is important is to look at RF in the same cohort,” Dr. van Venrooij says. “In our experience, mostly those sensitivities are comparable.”

Anti-CCP’s accuracy, particularly its specificity, has been demonstrated in detail in work by Drs. Morris and Wener. Dr. Morris analyzed serum samples from his practice and that of Allan Metzger, MD. Among 179 established RA patients, anti-CCP was positive in 87 percent, including 11 of 24 (46 percent) who were RF-negative. It was 99 percent specific among 100 normal controls and 98 percent specific among 333 disease controls. “These values are very similar to figures quoted in the literature for the second-generation assay,” Dr. Morris says. He has worked with both kits and concludes that “there is no major difference in terms of end results.”

Dr. Wener challenged anti-CCP’s ability to differentiate RF-positive RA patients from patients who are RF-positive and have polyarthritis that mimics RA. He chose chronic hepatitis C infection, in which RF is present in a significant number of patients. He found RF in 22 of 50 random cases of hepatitis C infection; none were positive for anti-CCP.

In a second challenge, Dr. Wener looked at sera from 29 patients with the syndrome cryoglobulinemia, which is often associated with chronic hepatitis C infection and which has an arthritis that, early on, can have RA-like features. Seventy-six percent of those patients were positive for RF, some with very high titers. Only two (three percent) tested positive for anti-CCP. “Because cryoglobulin serum is sticky,” Dr. Wener says, “the prevalence of anti-CCP antibody is probably even lower than we observed.” When he treated samples to eliminate nonspecific binding, both anti-CCP positives became negative, while all patients positive for RF continued to have it.

Dr. Wener considers it important not to confuse hepatitis C infection with RA, since physicians would not want to treat hepatitis C infection with the newer anti-RA drugs that suppress the immune system.

Anti-CCP may also be able to distinguish other arthritides from RA, Dr. Morris says. One such condition is polymyalgia rheumatica, seen in the elderly, which is typically RF-negative and can present with synovitis resembling RA. Also, he says, it is “very common” for lupus patients to have polyarthritis and to be RF- and ANA-positive.

“Given the fact that polyarthritis is a common presenting feature of lupus and can be RF-positive, my guess would be that anti-CCP positivity would help make an RA diagnosis more likely in this context,” he says. Among 57 lupus patients whom he tested, three (five percent) were positive for anti-CCP. Dr. Morris lists several other conditions that may have rheumatoid factor positivity and arthritis, such as parvoviral infection, sarcoidosis, hepatitis C, and occasionally spondyloarthropathy and pseudogout. These conditions can be difficult to differentiate from rheumatoid arthritis. If under these circumstances an anti-CCP result is positive, it would strongly suggest rheumatoid disease. It would therefore be helpful diagnostically and potentially allow aggressive therapeutics to be instituted earlier for rheumatoid disease, says Dr. Morris. A negative result in this context

is not helpful.

In addition to its diagnostic value, anti-CCP may also have long-term prognostic significance in terms of radiological deterioration. “We have shown that a positive anti-CCP test on the first serum drawn after the beginning of disease predicts from a statistical point of view erosions of bone five years later,” Dr. Meyer says. It also predicted progression in the total Sharp score—joint space narrowing plus erosions.

As Dr. Wener notes, it has been known for a long time that high-titer RF (>1/160 or 1/320 in old units; >100 IU/mL in new units) is likely to be associated with worse disease. “I think a similar story will turn out to be true for anti-CCP,” he says. “When present at high levels it is likely to be associated with more disabling disease.” At the present time, he says, it is not clear where the “high-titer” cutoff should be.

Considerable work has been done on anti-CCP in early RA patients.

In unpublished data, about 20 percent of all people attending the early arthritis clinic at the University of Leiden in the Netherlands eventually developed RA, Dr. van Venrooij reports. But more than 90 percent of those who were anti-CCP-positive at the first visit developed RA in two to three years. The Leiden group developed a seven-factor model including mostly clinical factors plus anti-CCP and RF testing. Applied at the first visit, this algorithm distinguished self-limiting arthritis from persistent nonerosive arthritis from persistent erosive arthritis. “Anti-CCP added substantially to the predictive value of the model,” Dr. van Venrooij says. For instance, the difference between erosive and nonerosive disease was nine points. A positive anti-CCP result counted as three points. A positive RF test result counted as two more points. So between them, these two laboratory tests could move a patient halfway toward erosive disease (Arthritis Rheum. 2002;46: 357–365).

Dr. Wener agrees anti-CCP may be important in diagnosing early synovitis, before it is clear if a patient has RA. “[The Leiden group’s] work shows that in early synovitis, when you are still trying to sort things out, the combination of anti-CCP and RF improves the ability to predict who will have erosive arthritis,” he says. Imaging modalities that visualize surrogate markers of erosive changes, such as joint edema, also identify patients more likely to progress to erosive changes, Dr. Wener notes. “Which way is better for finding these patients—physiology or anatomy?” he asks. “Certainly physiology—anti-CCP testing—costs much less.”

Dr. Symmons attempted to identify which early arthritis cases would progress to erosions in five years using clinical factors plus RF. She achieved moderate success—positive predictive value of 61 percent and negative predictive value of 74 percent. She sent samples from her cohort, the Norfolk Arthritis Register, to have anti-CCP measured, and the Leiden investigators used these data to evaluate their model on her patients. “They were trying to predict persistence, then, given persistence, whether the patient would get erosions,” she says. “The model they had generated did not work as well on our patient group [as on the Leiden cohort], but it did work reasonably well.”

In a particularly striking finding, anti-CCP has been found to be positive in some people many years before RA onset. Dr. van Venrooij worked with investigators in Umea, Sweden, to analyze serum samples from individuals who had given blood. Of 83 patients who later developed RA, 35 percent had anti-CCP antibodies before onset of symptoms; 23 percent had IgM-RF. The sensitivity and positive predictive value of anti-CCP antibodies more than one year before symptoms were 22 percent and 70 percent; for IgM-RF they were 22 percent and 47 percent respectively. In some patients the latency between detection of anti-CCP antibodies and appearance of disease was nine years, which Dr. van Venrooij calls “quite astounding.” Perhaps further research on this phenomenon can reveal how RA is triggered and shed light on its pathogenesis.

With its superior diagnostic accuracy, the anti-CCP assay is a good
complement to contemporary advances in RA therapy. “Clinicians are embarking on more aggressive forms of therapy fairly early in the course of disease in hopes of diminishing the long-term morbidity associated with RA,” Dr. Homburger says. This treatment is not without side effects, he adds, so it is important to diagnose RA accurately before initiating treatment. For clinicians to determine a strong likelihood of RA, they need a more specific laboratory test than RF. Anti-CCP meets that need. Also important is its positivity in the majority of early arthritis patients: In the Umea study, 72 percent of patients who presented early had a positive

anti-CCP result.

Two forms of aggressive therapy are being pursued. One type consists of newer drugs, called biologicals, that block or neutralize the cytokines tumor necrosis factor-alpha or interleukin-1, which are central to the inflammation of RA. Early treatment with these agents has been shown to sharply reduce radiologic damage for up to two years. “Our decision to use biologicals is based not just on a positive anti-CCP result,” says Dr. Meyer. Several clinical criteria are also critical. If anti-CCP and other factors indicate aggressive disease, Dr. Meyer says, “we will not spend too much time on methotrexate alone. If after three to four months of methotrexate monotherapy the patient has not a very good response, we can use this argument for adding a biological.”

A second form of aggressive therapy uses early initiation of older disease-modifying antirheumatic drugs, or DMARDs, such as higher-dose methotrexate or sulfasalazine, or a combination of two DMARDs. Dr. Symmons and her colleagues have been conducting observational studies to determine whether there is a window of opportunity to make a difference using early DMARD treatment (within three months of symptom onset). (United Kingdom guidelines say not to use biologicals until a patient has failed on at least two DMARDs.) After controlling for disease severity (those with the most severe disease are most likely to get early therapy), they have seen an effect of early treatment with sulfasalazine, with regard to both disability and advanced erosions. They are now following patients started on early treatment with methotrexate.

Additional evidence for the efficacy of very early treatment with DMARDs was reported last year by Valerie Nell, MD, a fellow in rheumatology at the University Hospital, Vienna, and colleagues. Patients who started DMARD treatment within three months of symptom onset had significantly superior clinical and radiological outcomes at three years than patients who

started treatment a mean of 12 months after symptom onset. Twice as many patients had erosions at three years in the group that started treatment later.

Given the performance of the anti-CCP assay, what is its role in

clinical practice? Not surprisingly, experts differ about how it should be used. For the most part, however, consultants agree it should be used in a setting where early arthritis needs to be differentiated into RA and other etiologies.

At Mayo Clinic the test is wending its way through the peer-review process and should soon be available for clinical use. “We anticipate offering the anti-CCP test in two ways,” Dr. Homburger says. First, as a stand-alone test ordered as an individual assay. Second, as part of a connective tissue disease cascade algorithm intended mainly for primary care clinicians.

“Right now we use ANA as the ‘gatekeeper’ test,” he says. If ANA is positive at a strong level, that triggers automatic followup with tests for anti-DNA and extractable nuclear antigen. “We would in all likelihood offer anti-CCP along with ANA as gatekeeper tests,” Dr. Homburger says. “In people who present with signs and symptoms of connective tissue disease, it is often not possible to distinguish clinically between those who have early RA and those who have some other connective tissue disease, such as lupus or even scleroderma. Right now we have no test at the start of the cascade that provides much information on the likelihood that the patient has RA. We think anti-CCP is good enough to do that.”

Dr. Homburger plans to continue offering RF testing. He acknowledges that RF has limitations, but clinicians are familiar with it and can use it efficaciously. “I think for now anti-CCP will supplement the RF assay,” Dr. Homburger says. “Perhaps some years from now it may come to replace it.”

When Dr. Bizzaro began offering the anti-CCP assay, he had several meetings with local general practitioners and internists at the three hospitals his laboratory services. “I told them that the specificity is very high but the sensitivity is not so high,” he says, “so they must expect at least 30 percent of patients will be negative even if they have RA.” More important, he told them that only a minority of patients with an initial complaint of arthritis will turn out to have RA. “I have instructed my colleagues of this low probability to have a positive finding unless they carefully select patients with a high suspicion of RA,” he says. “I think I made them afraid to order the test. They are ordering it very wisely in selected patients.”

Another vote to use anti-CCP in carefully selected patients comes from Dr. Nell. “From our data, we would propose to use it in patients who we think are possible/ probable RA on clinical grounds,” she says. “If RF is low or negative, then one should do anti-CCP. But on the basis of the data obtained as yet, we don’t think you should do it as a routine test.”

Dr. Nell is referring to data from a study she conducted with her hospital colleagues Josef Smolen, MD, head of rheumatology, and laboratory director Günter Steiner, PhD, among 200 patients from their very early (less than three months’ symptom duration) arthritis clinic. IGM-RF showed a high sensitivity with a specificity of about 90 percent, while anti-CCP was less sensitive but 98 percent specific. Together, they were 99 percent specific. Twenty-five percent of RF-negative or low-titer (>20<50 U/L) patients were anti-CCP positive. “The main suggestion from these data,” Dr. Nell says, “is to use anti-CCP in addition to RF to get very high specificity above RF alone. Especially in patients with low or negative RF, anti-CCP provides additional value.”

Dr. Nell acknowledges that using such stringent criteria greatly reduces sensitivity. “You may lose some patients, but you gain specificity,” she says. When initiating early aggressive therapy, specificity is paramount.

Other practitioners use anti-CCP routinely in selected patients. Says Dr. Meyer of Hôpital Bichat: “I use the anti-CCP test routinely in the hospital at the first medical interview for every patient who has arthritis of any joint or inflammatory arthritis who doesn’t yet have a diagnosis. I want to detect quite early patients who will become definite RA. On the same serum I perform the latex test for RF.” In his experience, anti-CCP will soon be performed by all physicians, including rheumatologists, who practice outside the hospital. Dr. Meyer sees the assay becoming part of a complete workup of arthritis. Perhaps cost issues are aiding this wide adoption. “An ELISA in our country is always the same price,” he says, “about $19.”

In Dr. Wener’s practice at the University of Washington, he uses anti-CCP as a corroborative test in patients who have a positive test for RF and equivocal findings for RA. “In these patients,” he says, “I am trying to decide whether it is more likely to be RA and aggressive disease or more likely to be post-infectious or related to malignancy, or one of the other causes of RF-positive arthritis.” A positive anti-CCP result tilts his thinking more strongly toward RA. And positive tests for both RF and anti-CCP make him think about worse-prognosis disease.

Anti-CCP is helpful in a patient who has hepatitis C infection and arthritis and is RF-positive. If anti-CCP is negative, then the positive RF result and the synovitis are more likely caused by hepatitis C. On the other hand, the presence of anti-CCP antibodies at a high level make it much more likely to be RA. “Then I have to decide how to treat RA in someone with a chronic hepatitis C infection,” Dr. Wener says.

“In the field of undiagnosed polyarthritis, my own opinion is that anti-CCP is the screening test of choice,” Dr. Morris says. “We already screen with RF. We shouldn’t eliminate RF, but use anti-CCP in addition. If it is positive, you can be almost certain that the diagnosis is RA.”

Could anti-CCP eventually replace RF testing? “I think it is hard to know right now,” Dr. Wener says. “It is somewhat tied into the potency and expense of the new medications and the cost in terms of reduced suffering and the money value of making an early diagnosis and getting the disease under better control early on.” At an intermediate stage, it might be that clinicians will use RF plus anti-CCP routinely. “Rheumatoid factor has been around so long, I think it will be used for at least a number of years,” Dr. Wener says. However, anti-CCP could theoretically replace it because of its specificity. “It is important to avoid using potent drugs in patients who have arthritis that is not rapidly progressive and will do well,” Dr. Wener says. “Anti-CCP may take over, but we still need more data.”

If the early promise of anti-CCP testing is borne out, it may even fundamentally influence diagnostic practices in rheumatology. Dr. Homburger notes that American rheumatologists don’t rely much on serology to diagnose RA. “I think they will find anti-CCP helpful,” he says. “But at this stage I don’t think they are as enthusiastic about it as I am in the laboratory.”

Is anti-CCP good enough to change rheumatologists’ attitudes toward laboratory testing? “This is speculative,” Dr. Homburger says, “but I think the answer is yes.”

William Check is a medical writer in Wilmette, Ill.