November 2011: Volume 37, Number 11
Antinuclear Antibody Testing Dilemmas
Does High Throughput Trump Sensitivity?
By Genna Rollins
Autoantibodies have been used as a biomarker for systemic autoimmune rheumatic diseases (SRDs) for more than a half-century and have found a solid place as a tool in diagnosing conditions such as systemic lupus erythematosus, scleroderma, and Sjögren syndrome. Recently, however, evolving technologies and inadequate understanding of the limitations of various methods have led antinuclear antibody (ANA) results to be used inappropriately as a screening test for SRDs. Given this challenge, laboratorians have a responsibility to know the ins-and-outs of their methods and educate physicians to make the best use of ANA testing, according to experts.
“Many primary care practitioners have an imperfect understanding of the diagnostic criteria for autoimmune disorders and the pitfalls that go along with ANA screening testing. There is a problem, a big problem, of false, clinically irrelevant positives in a large minority of normals. This is well-appreciated by rheumatologists, but generalists really are not that aware of it,” said John L. Carey, MD, vice-chair of pathology and laboratory medicine at Henry Ford Health System in Detroit. “This ocean of patients who do not have systemic rheumatic disease, yet test positive, destroys the positive predictive value of ANA tests. Unfortunately, this is not appreciated by clinicians and it really challenges the lab to correct a false impression created by clinically irrelevant positive tests.”
The Pros and Cons of Immunofluorescence
False positives are not the only issue labs face in educating physicians about ANA testing. Both old and new testing technologies are evolving, and evidence is emerging that could help improve the specificity of established methods. Indirect immunofluorescence (IIF) has been used since the 1950s to detect ANAs, initially using rodent organs and now, most commonly, human epithelioid laryngeal cancer cell (HEp-2) cell lines as substrates. IIF offers the advantages of providing both ANA pattern and titer information, and detecting in excess of 100 autoantibodies. The method has very good sensitivity for certain SRDs, such as lupus (85%), mixed connective tissue disease (100%), and drug-induced lupus (100%).
However, IIF has some downsides as well. It suffers from being a non-standardized manual test and having subjective result interpretation and low reproducibility. This time-consuming, low throughput test also produces its share of false negatives for certain autoantibodies, including Jo-1 (30%), Ribosomal P (40%), and proliferating cell nuclear antigens (30%). At any given time, IIF misses about 15% of lupus patients, too, but when established patients are followed over time, approximately 97–98% will have positive IIF results, according to Carey.
On the flip side, as many as 13% of healthy individuals will have positive ANA IIF results at a 1:80 dilution. Though higher titers generally are thought to be more clinically significant and indicative of SRD, evidence suggests that beyond a threshold of 1:80 or 1:160, ANA titer “has little bearing on diagnosis or disease activity,” according to Marvin Fritzler, MD, PhD, professor of medicine at the University of Calgary in Alberta, Canada.
What’s in the Pattern?
Given the sometimes limited information provided by titers, clinicians also use the staining pattern for diagnostic clues. Several IIF patterns have been linked to specific antigens and SRDs. For instance, a mitotic apparatus pattern is associated with nuclear mitotic apparatus 1, which is present in lupus and Sjögren syndrome. However, the staining pattern is not always specific for ANA or disease. As an example, a speckled pattern has been associated with certain ANAs like SS-A/Ro, but it also is found in patients with a variety of SRDs, as well as in healthy individuals.
Even with the limitations of both titer and staining patterns, some researchers are looking closer at staining pattern as a way to discriminate healthy individuals from those with SRDs. In a recent study comparing healthy subjects to patients with SRDs, Brazilian researchers found distinct pattern profiles between the two cohorts and suggested that pattern is a “critical parameter” for interpreting positive ANA IIF results (Arthritis Rheum 2011;63:191–200). The nuclear dense fine speckled pattern occurred only in healthy individuals, and at a moderate titer of 1:640 in nearly two-thirds. In contrast, nuclear course speckled, nuclear homogeneous, and nuclear centromeric patterns were found only in patients with SRDs. The nuclear fine speckled pattern was common to both groups, but tended to be of high titer in patients with SRDs and low titer for healthy individuals.
“Our daily practice suggested that the ANA pattern might play a role in the interpretation of a positive ANA test. However, we were surprised by the extent with which this parameter may guide the clinical significance of a positive ANA test,” said the study’s senior author, Luis E. C. Andrade, MD, PhD, associate professor of medicine and head of rheumatology at Universidade Federal de São Paulo, Brazil. “Our results imply that certain ANA patterns, such as the nuclear homogeneous and the nuclear coarse speckled pattern, should direct a thorough clinical and laboratory investigation to unravel a systemic autoimmune disease, especially within the spectrum of systemic lupus erythematosus. Even if no abnormality is found in an initial screening, it may be wise to follow such patients for a while because they may be undergoing an incipient stage of the disease.”
While Andrade and his coauthors called for further research to test the reproducibility of their findings, he suggested that labs could readily implement this approach to pattern recognition. “The analyst in an ordinary clinical laboratory should be able to recognize the ANA patterns that are clinically relevant and those that are most probably observed in individuals with no apparent autoimmune disease,” he commented.
In an accompanying editorial, Fritzler agreed that “this study may provide the leading edge of understanding which autoantibodies rule out the diagnosis of systemic ARD.” However, he went on to caution that as promising as the findings may be, “interlaboratory exchange of sera and the establishment of a validated reference standard producing the nuclear dense fine speckled IIF pattern are desperately needed before these observations can be put into wide clinical laboratory practice.”
Technological, Standardization Advances
Recently, some manufacturers have taken steps to automate both the analytical and pre-analytic components of IIF ANA testing, including using automatic fluorescent image analysis to provide a virtual titer, thereby eliminating the process of manually staining a series of diluted samples. At least one system simultaneously reads multiple slides and provides basic positive or negative results and pattern recognition.
Recognition of the variability of IIF and other types of autoantibody testing methods lead in the early 1980s to formation of the Autoantibody Standardization Committee in Rheumatic and Related Diseases under the auspices of the International Union of Immunological Societies, the World Health Organization, Arthritis Foundation, and the Centers for Disease Control and Prevention (CDC). In the intervening decades, the committee has overseen development of reference sera for 16 ANAs which are available through CDC to research and diagnostic laboratories and companies developing autoantibody diagnostic kits. Andrade, who chairs the committee, indicated that the group is on the verge of releasing a standard for citrullinated peptides and is in the process of establishing standards for antibodies to beta-2 glycoprotein 1.
Other standardization efforts are underway as well. For instance, the European Autoimmunity Standardization Initiative was founded in 2002 with the goal of improving diagnostics in chronic rheumatic disorders by strengthening the collaboration between clinical and laboratory scientists responsible for autoimmune diagnostics.
The Move to Higher Throughput
Even as technology, the evidence base, and standardization efforts coincide to improve the performance of IIF and other traditional assays based on hemagglutination reactions and immunodiffusion, these methods are being given a run for their money by enzyme immunoassays (EIA) and other solid phase methods like microarrays and bead-based assays. The latter have come on strong in the past 15 years, and may account for more than half the testing currently performed in the U.S., since many high-volume labs use the tests, according to David Keren, MD, medical director of Warde Medical Laboratory in Ann Arbor, Mich.
These methods offer some distinct advantages in an era of belt-tightening in healthcare. Automated and faster than IIF, they run on platforms that can be used to perform other tests and are easier for lab staff to maintain proficiency on. The quality of these tests has improved over time, and College of American Pathologists surveys show that both EIA and multiplex tests perform well in comparison to IIF, according to Keren. But they have their own set of challenges.
“The solid-phase assays such as bead and EIA assays include relatively few antigens in comparison to intact HEp-2 cells. They also don’t provide information such as patterns of reactivity and strength of titer, which rheumatologists have found useful in evaluating their patients,” explained Keren. “Further, manufacturers of solid phase assays use different combinations of antigens and even within one manufacturer, one must compare lot-to-lot variability.”
What’s it Missing?
One analysis of commercial multiplexed ANA screening kits found that they used a range of six to 15 antigen combinations (Lupus 2005;15:412–421). Depending on how they’re made, some solid-phase assays contain, in addition to specific antigens, an extract of HEp-2 cells that enables detection of a broader range of antigens. While these assays generally perform well, depending on a patient’s SRD and the type of autoantibodies he or she has, the assays will do better or worse. For instance, a recent analysis of scleroderma patients found that only 51% tested positive by multiplex assay versus 91% by IIF (Clin Rheumatol 2011 DOI 10.1007/s10067-011-1766–6). Although there was good agreement between the methods in detecting Sc170, RNP, and centromere antibodies, the multiplex assays did not detect several autoantibodies of importance in scleroderma.
“The test performs well in what it’s been designed to do, which is to detect those specific antibodies that have been put into the assay. The problem is, for certain subsets of patients, it has a high false-negative rate,” explained the study’s lead author, Victoria Shanmugam, MD, an assistant professor of medicine at Georgetown University in Washington, D.C. In the case of scleroderma, one of the missed autoantibodies, RNA polymerase III, is associated with poor outcomes, so the insensitivity of some assays in picking it up is a concern.
“Those patients are at high risk of renal crisis and can have abrupt onset of disease, presenting initially with renal crisis and not having other clinical signs like skin thickening. So the emergency physician or internist may not consider scleroderma as an underlying etiology because they’re seeing a negative ANA. That potentially could lead the patient to not being treated appropriately.”
IIF: Still the Gold Standard
For labs that are using EIA or multiplex ANA tests instead of IIF, do the benefits of higher throughput and less staff time outweigh some sensitivity and specificity problems? Not according to the American College of Rheumatologists, which in 2009 issued a position statement that declared that despite the advent of multiplex and other solid-phase assays, IIF remains the gold standard for ANA testing (See Table, below). The statement went on to call for labs using the newer methods to provide data to ordering physicians showing that their assay has the same or improved sensitivity and specificity compared to IIF.
ANA Testing Recommendations
How Does your Lab Measure Up?
A task force of clinical experts convened by the American College of Rheumatology (ACR) examined the evidence surrounding antinuclear antibody (ANA) testing and issued a position statement that was approved by the ACR board of directors in 2009.
Recommendations of the task force included:
- The immunofluorescence (IF) ANA assay should remain the gold standard for ANA testing.
- Hospital and commercial laboratories using bead-based multiplex platforms or other solid phase assays for detecting ANAs must provide data to ordering physicians on request that their assay has the same or improved sensitivity and specificity compared to IF ANA.
- In-house assays for detecting ANA, as well as anti-DNA, anti-Sm, anti-RNP, anti-Ro/SS-a, anti-La/SS-B, etc., should be standardized according to national (eg., CDC) and/or international (eg., WHO/IUIS) standards.
- Laboratories should specify the methods used for detecting ANAs when reporting their results.
Where does this leave laboratorians and clinicians? If Shanmugam and other rheumatologists worry that false-negative results primarily by EIA or multiplex methods might lead to potentially serious diseases being misdiagnosed or at least not readily identified, others see even darker clouds in complicated work-ups that can arise from false positives by IIF. “ANA frequently is performed as a screening test and unfortunately, it’s a poor screening test with low sensitivity and specificity for disease. So when it’s used that way, you get a lot of weakly to mildly positive tests and the patients get referred to rheumatologists only to find there’s no disease present,” said Charles Spencer, MD, section chief of rheumatology at Nationwide Children’s Hospital and professor of clinical pediatrics at The Ohio State University College of Medicine in Columbus. “A low-titer ANA probably costs the healthcare system $500 to $2,000 every time one is performed that’s inappropriate.” Some ANAs are used as part of the diagnostic criteria for certain systemic rheumatic diseases, but others are used to support the diagnosis of some SRDs (See Table, below).
ANAs in Diagnosing Systemic Rheumatic Disease
Some antinuclear antibodies are used as part of the diagnostic criteria for certain systemic rheumatic diseases, but others are used to support the diagnosis of some SRDs.
Antinuclear Antibody used as
ANA, anti-dsDNA, anti-SM
anti-U1RNP (high titer)
Anti-centromere, anti-topo I
|Juvenile Chronic Arthritis
Legend: anti-dsDNA: anti-double stranded DNA; anti-SM: anti-smooth muscle; MCTD: mixed connective tissue disease; anti-U1RNP: anti-U1 ribonucleoprotein; anti-topo I: anti-toposomerase I; PM/DM: poly-dermatomyositis; Secondary APAS: Secondary anti-phospholipid antibody syndrome
Source: International Federation of Clinical Chemistry and Laboratory Medicine Guidelines for Antinuclear Antibody Testing.
Wading Through Testing Challenges
Although rheumatologists are familiar with the pros and cons of the various ANA methods, they often are not the ones ordering the tests initially. Donald Bloch, MD, a rheumatologist and assistant professor of medicine at Harvard Medical School, emphasized that general complaints like musculoskeletal pain and fatigue, which many SRDs share with a variety of other illnesses, can trip-up internists. “The doctor’s faced with a patient with vague symptoms. He doesn’t know what to think in advance and the differential diagnosis could be quite large, ranging from a functional diagnosis to very severe illness. He’s looking for a hint—does his patient have an autoimmune disease?”
This problem can be compounded when the ordering physician isn’t aware of the type of ANA test that’s being performed. According to Shanmugam, initially after some labs switched to EIA or multiplex methods, it was difficult to request IIF, but this situation has since been rectified. “The lab needs to make sure that physicians like hospitalists, emergency physicians, and internists have easy access to immunofluorescence assays so that if they have a clinical suspicion for an autoimmune disease they have the opportunity to request IIF,” she suggested.
By the same token, labs offering the newer methods need to tweak their menus and educate physicians about how to utilize the tests. “Sometimes with ELISA, it’s built-in that if the ANA’s positive, then all these other tests will be done. But they’re usually not needed,” said Spencer. “So you’d want to have some barrier so that all these other tests aren’t automatically ordered, and there’s a step-wise order to the process.”
Indeed, guidelines for ANA testing by the International Federation of Clinical Chemistry and Laboratory Medicine suggest that clinicians and laboratorians might want to agree on “practical algorithms for test ordering, for rational stepwise exploration of a preliminary result at screening, and for interpretation of a positive final result.”
Many labs, Carey’s included, run an EIA if the physician orders an ANA test without specifying IIF. If the initial EIA is positive, the lab confirms the result with IIF and reports both titer and pattern. This approach, coupled with advisory comments, has worked well at Henry Ford Health System, he said. “I’ve found this very helpful, in reassuring my clinicians that after reviewing a weak positive ANA screen, that if they feel the signs and symptoms are not there for making a diagnosis for systemic rheumatic disease, they’re probably right. I’ve seen that with follow-up notes advising patients of the meaning of these positive tests.”
As the technologies and evidence base around ANA testing continues to evolve, labs will do well to keep up the conversation with physicians. As Bloch observed, “It’s our responsibility to teach clinicians not to fear a positive ANA.”