In 1971, the first criteria set for classification of systemic lupus erythematosus (SLE) was published by the Diagnostic and Therapeutic Committee of the American College of Rheumatology (ACR; formerly, the American Rheumatism Association) (1). Establishing these criteria involved 52 rheumatologists submitting information about the clinical and laboratory features of >800 patients. These patients comprised various rheumatologic subgroups: unequivocal SLE, probable SLE, definite rheumatoid arthritis (RA), and other, nonrheumatic diseases. Combinations of these variables were analyzed before the final construct of 14 manifestations with 21 items in the criteria was reached. This yielded a sensitivity of ∼90% and a specificity of 99% against RA and 98% against other diseases. These criteria allowed uniform categorization of lupus patients for the purpose of research.

Eleven years later, the criteria were revised by a process similar to that used in 1971. Features of the 177 SLE patients and 162 controls were subjected to computerized statistical methodology to yield a combination with the greatest utility (2). The criteria were reformatted into 11 categories, some with subcomponents, and newer immunologic tests, such as determinations of antinuclear antibody (ANA), anti–double-stranded DNA (anti-dsDNA), and anti-Sm antibody were added. It was decided that at least 4 criteria need to be met, either serially or simultaneously, for a patient to be classified as having lupus. This achieves a sensitivity of 96% and specificity of 96%. Other groups have since published their evaluations of these criteria in studies of various cohorts (3, 4).

In 1997, further changes were made to the criteria but in the immunologic disorder category only (5). These changes were the subject of the present study. Table 1 outlines the immunologic disorder criterion in the 1982 criteria and in the 1997 updated version. In contrast to the 1982 modifications, the 1997 amendment was a consensus statement with no evaluation, to our knowledge, published to date. The impact on patient selection, and sensitivity and specificity of the new criteria set have not been evaluated.

Table 1. Immunologic disorder criterion in the 1982 SLE criteria and the 1997 update*
  • *

    See refs. 2 and5. SLE = systemic lupus erythematosus.

1982 criteria
 a) Positive LE cell preparation OR
 b) Anti-DNA: antibody to native DNA in abnormal titer OR
 c) Anti-Sm: presence of antibody to Sm nuclear antigen OR
 d) False-positive serologic test for syphilis known to be positive for at least 6 months and confirmed by Treponema pallidum immobilization or fluorescent treponemal antibody absorption test
1997 update
 1. Delete item 10(a) (“Positive LE cell preparation”), and
 2. Change item 10(d) to “Positive finding of antiphospholipid antibodies based on 1) an abnormal serum level of IgG or IgM anticardiolipin antibodies, 2) a positive test result for lupus anticoagulant using a standard method, or 3) a false-positive serologic test for syphilis known to be positive for at least 6 months and confirmed by Treponema pallidum immobilization or fluorescent treponemal antibody absorption test.”

The changes in the immunologic disorder category consist of removal of the LE cell item and addition of the antiphospholipid antibody (aPL), anticardiolipin antibody (aCL), and lupus anticoagulant (LAC) items. The false-positive serologic test for syphilis feature in the immunologic disorder category in the 1997 version is unchanged from 1982. The other 10 categories in the criteria are unchanged.

Without validation of these changes, the removal of the LE cell item and the addition of the aCL and LAC items raise concerns that the clinical characteristics of patients recruited in lupus research may be altered. The aim of the present study was to evaluate whether the 1982 and 1997 criteria select different groups of patients for inclusion in lupus research.

Between 1992 and 2002, an inception cohort of lupus patients followed up in a prospective study was identified. Inception was defined as a diagnosis of SLE made at or within 12 months of the first clinic visit. All patients recruited into the cohort fulfilled at least 4 of the 11 1982 revised criteria. In this study, only those criteria present at the point of inception were analyzed, rather than criteria accumulated during the disease course, because these features determine patient recruitment. Patient assessments involve a clinical consultation and serologic studies that incorporate the appropriate immunologic, hematologic, and biochemical parameters relevant to SLE and include LE cell testing and evaluation of the LAC as determined by prolongation of the activated partial thromboplastin time (APTT). The aCL was routinely tested in all patients from 1992 onward by enzyme-linked immunosorbent assay. Therefore, between 1992 and 2002, results regarding LE cells, aCL, and LAC were available for this cohort, and the frequency of positivity for these parameters in the inception cohort was determined.

Patients meeting only 4 criteria at inception, 1 of which was the immunologic disorder category, were analyzed further. Moreover, only patients who had been tested for all 3 of the parameters of interest (LE cells, aCL, and LAC) were included. To examine the impact of removal of the LE cell test from the immunologic disorder criterion in 1997, we removed this parameter from the data on patients who met just 4 criteria, to determine whether this altered their recruitment. The final evaluation was to examine the impact of the addition of aCL and LAC to the criteria. Patients who were positive for aCL and/or LAC but negative for LE cells, and therefore met the 1997 criteria, were analyzed.

In the inception cohort, 154 patients were identified. Of this group, LE cells were positive in 79 patients (51.3%), aCL in 35 (22.7%), and LAC in 16 (10.4%). Forty-two of the 154 patients met only 4 criteria, and in 36 of these 42, 1 of the criteria was the immunologic disorder category. These 36 patients were the subject of the following analysis.

To evaluate the impact of the removal of LE cell positivity from the criteria, we removed the LE cell data for these 36 patients and found that 2 would no longer be classified as having lupus since the LE cell was their only positive feature in the immunologic disorder criterion. These 2 patients therefore would meet just 3 criteria. If we added the results for aCL and LAC in these 2 patients, they did not regain their classification, because both were aCL and LAC negative.

To evaluate the influence of including aCL and LAC in the criteria, we applied the 1997 criteria to the 36 patients to evaluate those who were aCL and/or LAC positive but negative for LE cells. We found 1 patient in this category; this individual was also positive for another feature included in the immunologic disorder criterion (anti-dsDNA).

The frequency of LE cells identified in this inception cohort, within 12 months of diagnosis, was 51.3%, which is within the range quoted in previous reports (48–92%) (1, 4, 6). This reflects the presence of LE cells at a single point in time only, and early in the disease. When disease features accumulate over time, the prevalence will be greater. In addition, therapy may influence expression of LE cells. Thus, analyzing the entire University of Toronto cohort, we find that at some point in their disease, a higher number of patients (694 of 1,032 [67.2%]) have been LE cell positive. Similarly the frequencies of aCL and LAC in this inception cohort are lower than those found in the entire cohort (372 of 740 [50.3%] positive for aCL, 551 of 1,011 [54.5%] positive for LAC).

The change in criteria did not produce a significant change in this population of lupus patients, since removal of the LE cell parameter altered the classification of only 2 members of the inception cohort (1.3%). If a larger proportion of patients were affected by the revision in the criteria, the clinical characteristics of those subjects would merit analysis to describe the change in the cohort that would be caused by the removal of such patients. Inclusion of aCL and LAC appears to have no impact in this group. Further comparisons of the 1982 and 1997 criteria in other cohorts would be of great interest, but may be limited by the requirement for a population of patients with results for LE cells, LAC, and aPL simultaneously available over a long period. The declining use of the LE cell test and the relative novelty of aPL and LAC testing may make this a difficult proposition, as is suggested by the lack of published data analyzing this change.

Alternatively, some authors have sought to evaluate the relative value of individual criteria. Somogyi et al found that among the 1982 criteria, the LE cell positivity ranked highest in contributing to the probability of SLE (7). Discoid lupus, ANA, and anti-DNA followed in consecutive order. Using receiver operating characteristic curves, Manu found that the immunologic disorder criterion in the 1982 set offered the highest discriminatory value for the diagnosis of SLE (8). Similarly, Edworthy et al (9) noted the immunologic disorder category to be the most discriminatory. These analyses serve to highlight the crucial role played by the immunologic findings in the criteria set and the need to validate any changes in this category.

The removal of LE cells as an immunologic feature addressed in the criteria came about due to declining use of this test concurrent with the advent of the ANA test. The diminishing popularity of the labor-intensive LE cell assay was highlighted during the compilation of the 1982 revised criteria, in which it had been performed in only 79 of 177 (45%) patients (2). The sensitivity of the LE cell test was shown to decrease from 92% in 1971 to 73% in the 1982 analysis. Offering a more sensitive (96–98%) but much less specific (49–72%) alternative, the ANA test has been readily available and widely used in the last 3 decades, and it has been demonstrated that ANA results are responsible for contributing to greater sensitivity of criteria sets in which they are included (6, 10).

Introducing aPL testing by standard methods to the criteria raises concern about specificity. Anticardiolipin antibody can be found in up to 16% of RA patients and may also be found postinfection (11). The implication of this is the potential inclusion of a varied population in a purported lupus cohort. The 1997 criteria update states that testing for aPL must be performed by standard techniques, but this issue warrants clarification (12). Methodology for LAC measurement is better defined in the antiphospholipid syndrome criteria (13), but there remain problems with uniformity of assays (14, 15). At the Seventh International Symposium on Antiphospholipid Antibodies (16), participants attempted to address the issue of standardization. These problems also limit comparison between cohorts.

Some limitations of our study warrant discussion. Detection of LAC was limited by use of only the prolonged APTT. For the purposes of a long-duration prospective study, this simple and cost-effective method, used for decades in this cohort, has allowed consistency of results for comparison over time. Other methods of LAC evaluation are relatively recent and are now being tested in our cohort. The clinical utility of the finding of a prolonged APTT has been demonstrated in its strong correlation with venous and arterial thromboses (17).

The impact of excluding LE cell positivity from the SLE criteria could be demonstrated here. Adequate assessment of the impact of including aCL and LAC, however, necessitates a prospective analysis of patients as they accumulate criteria, to define whether these aPL influence recruitment. These data are not available for our cohort.

In conclusion, based on this analysis in patients who meet 4 (but no more than 4) of the ACR criteria for SLE, excluding LE cell positivity from the criteria set has some impact. These criteria changes should be tested prior to application, with their sensitivity and specificity ideally assessed in patients with SLE, primary and secondary antiphospholipid syndrome, and other connective tissue diseases.

  • 1
    Cohen AS, Reynolds WE, Franklin EC, Kulka JP, Ropes MW, Shulman LE, et al. SL: Preliminary criteria for the classification of systemic lupus erythematosus. Bull Rheum Dis 1971; 21: 6438.
  • 2
    Tan EM, Cohen AS, Fries JF, Masi AT, McShane DJ, Rothfield NF, et al. The 1982 revised criteria for the classification of systemic lupus erythematosus. Arthritis Rheum 1982; 25: 12717.
  • 3
    Passas CM, Wong RL, Peterson M, Testa MA, Rothfield A. A comparison of the specificity of the 1971 and 1982 American Rheumatism Association criteria for the classification of systemic lupus erythematosus. Arthritis Rheum 1985; 28: 6203.
  • 4
    Yokohari R, Tsunematsu T. Application, to Japanese patients, of the 1982 American Rheumatism Association revised criteria for the classification of systemic lupus erythematosus. Arthritis Rheum 1985; 28: 6938.
  • 5
    Hochberg MC, for the Diagnostic and Therapeutic Criteria Committee of the American College of Rheumatology. Updating the American College of Rheumatology revised criteria for the classification of systemic lupus erythematosus [letter]. Arthritis Rheum 1997; 40: 1725.
  • 6
    Trimble RB, Townes AS, Robinson H, Kaplan SB, Chandler RW, Hanissian AS, et al. Preliminary criteria for the classification of systemic lupus erythematosus (SLE): evaluation in early diagnosed SLE and rheumatoid arthritis. Arthritis Rheum 1974; 17: 1848.
  • 7
    Somogyi L, Cikes N, Marusic M. Evaluation of criteria contributions for the classification of systemic lupus erythematosus. Scand J Rheumatol 1993; 22: 5862.
  • 8
    Manu P. Receiver operating characteristic curves of the revised criteria for the classification of systemic lupus erythematosus [letter]. Arthritis Rheum 1983; 26: 10545.
  • 9
    Edworthy SM, Zatarain E, McShane DJ, Bloch DA. Analysis of the 1982 ARA lupus criteria data set by recursive partitioning methodology: new insights into the relative merit of individual criteria. J Rheumatol 1988; 15: 14938.
  • 10
    Levin RE, Weinstein A, Peterson M, Testa MA, Rothfield NF. A comparison of the sensitivity of the 1971 and 1982 American Rheumatism Association criteria for the classification of systemic lupus erythematosus. Arthritis Rheum 1984; 27: 5308.
  • 11
    Merkel PA, Chang Y, Pierangeli SS, Convery K, Harris EN, Polisson RP. The prevalence and clinical associations of anticardiolipin antibodies in a large inception cohort of patients with connective tissue diseases. Am J Med 1996; 101: 57683.
  • 12
    Tsutsumi A, Ichikawa K, Atsumi T, Matsuura E, Koike T, Krilis SA. Use of various methods for anticardiolipin detection in the updated American College of Rheumatology revised criteria for the classification of systemic lupus erythematosus: comment on the letter by Hochberg [letter]. Arthritis Rheum 1998; 41: 13267.
  • 13
    Wilson WA, Gharavi AE, Koike T, Lockshin MD, Branch DW, Piette J-C, et al. International consensus statement on preliminary classification criteria for definite antiphospholipid syndrome: report of an international workshop. Arthritis Rheum 1999; 42: 130911.
  • 14
    Exner T. Conceptions and misconceptions in testing for lupus anticoagulants. J Autoimmun 2000; 15: 17983.
  • 15
    Jacobsen EM, Barna-Cler L, Taylor JM, Triplett DA, Wisloff F. The evaluation of clotting times in the laboratory detection of lupus anticoagulants. Thromb Res 2001; 104: 27582.
  • 16
    Pierangeli SS, Stewart M, Silva LK, Harris EN. An antiphospholipid wet workshop: 7th International Symposium on Antiphospholipid Antibodies. J Rheumatol 1998; 25: 15660.
  • 17
    Abu-Shakra M, Gladman DD, Urowitz MB, Farewell V. Anticardiolipin antibodies in systemic lupus erythematosus: clinical and laboratory correlations. Am J Med 1995; 99: 6248.