Classification Criteria for the Antiphospholipid Syndrome: Not the Same as Diagnostic Criteria for Antiphospholipid Syndrome

The latest consensus “guidelines” outlining revised Classification Criteria for antiphospholipid syndrome (APS) has just been published.[1] The aim of this publication was to develop new APS classification criteria with high specificity to be used in observational studies and trials, jointly supported by the American College of Rheumatology (ACR) and the European Alliance of Associations for Rheumatology (EULAR). Currently, classification of APS, for identification of homogeneous research cohorts, is based on the Sapporo criteria published in 1999,[2] as revised in Sydney in 2006.[3] The Sydney revised Sapporo criteria for APS require clinical features (thrombosis or pregnancy morbidity) and laboratory tests for antiphospholipid antibodies (aPL; i.e., lupus anticoagulant [LA], IgG/IgM anticardiolipin antibodies [aCL], and/or IgG/IgM anti-b2-glycoprotein I antibodies [aβ2GPI]), with a minimum of one positive aPL test, performed on two occasions at least 12-week apart.[3]

The 2023 ACR/EULAR APS classification criteria include an entry criterion of at least one positive aPL test within 3 years of identification of an aPL-associated clinical criterion, followed by additive weighted criteria (score range: 1–7 points each) clustered into six clinical domains (macrovascular venous thromboembolism, macrovascular arterial thrombosis, microvascular, obstetric, cardiac valve, and hematologic) and two laboratory domains (LA functional coagulation assays and solid phase enzyme-linked immunosorbent assays [ELISA] for IgG/IgM aCL and/or ab2GPI).[1] Patients accumulating at least 3 points each from the clinical and laboratory domains are classified as having APS. In the validation cohort, the new APS criteria versus the 2006 Sydney revised Sapporo classification criteria had a specificity of 99 versus 86% and a sensitivity of 84 versus 99%.[1]

This criteria set was generated by a multidisciplinary committee and was finally approved by the ACR Board of Directors and the EULAR Executive Committee. The authors further state:[1] “The criteria set were quantitatively validated using patient data and have undergone validation based on an independent data set. All ACR/EULAR-approved criteria sets are expected to undergo intermittent updates. Classification criteria are essential in clinical and basic science research because they allow investigators to study relatively homogeneous populations of patients recruited from a single or multiple research sites. In clinical settings, diagnoses are made by health care professionals evaluating an individual patient's symptoms, signs, and results of laboratory and imaging studies in order to guide therapeutic recommendations. Patients diagnosed with a particular disease may or may not fulfill classification criteria for that disease. Classification criteria, in the hands of an experienced clinician with expertise in rheumatology, may inform a diagnostic evaluation but improperly applied classification criteria may lead to misdiagnosis.” The publication also identifies caveats:[1] “The ACR is an independent, professional, medical, and scientific society that does not guarantee, warrant, or endorse any commercial product or service.”

The purpose of this Commentary is not to criticize the new Classification Criteria, but rather to expand on the differential between “Classification” criteria versus “Diagnostic” criteria. As correctly stated by the authors, “Patients diagnosed with a particular disease may or may not fulfill classification criteria for that disease” and “improperly applied classification criteria may lead to misdiagnosis.”[1] It is important to clearly identify that “Classification” criteria are for identifying “definite” APS for inclusion of patients into “observational studies and trials.” This ensures a homogeneous patient cohort for evaluation of potential therapies or improved treatments.[1] [2] [3] However, history tells us that “Classification criteria” may also be used as surrogates for “Diagnostic criteria” by many clinicians, especially as otherwise reported in the literature. As one example, Meroni and Borghi, in a recent review, summarize that “aPL are mandatory for the diagnosis but are also a risk factor for the APS clinical manifestations” and that “LA, aCL, and ab2GPI assays are the formal laboratory classification/diagnostic criteria.”[4] While we would take the intention here to advise that these assays are included in both classification and diagnostic criteria, readers may inappropriately conclude that classification and diagnostic criteria are one and the same thing. Another example is another review on diagnosis and management of APS that includes the key concepts of aPL testing, along with the definition of moderate to high titers (40 GPL or MPL or 99th percentile) of aCL or ab2GPI IgG or IgM (99th percentile),[5] essentially as taken from the Sydney criteria.[3]

Each of the authors of the current Commentary act as independent peer-reviewers for various international journals, and we often “correct” authors who refer to these “Classification criteria” as “Diagnostic criteria.” Naturally, we are not able to capture and correct every instance of this misconception, and thus, authors may continue to incorrectly ascribe the “Classification criteria” as “Diagnostic criteria.” For example, Kinoshita et al report that “The diagnosis of antiphospholipid syndrome is usually established based on clinical and laboratory findings by strictly following the 2006 Sapporo classification.”[6] Kamal Eldin et al, in their study of aβ2GPI in systemic lupus erythematosus (SLE) patients, identify that although “aβ2GPI IgA is not within Sapporo diagnostic criteria, it seems to contribute to the pathogenesis of thrombotic manifestations of SLE and may represents a useful indicator particularly when standard aPL tests are negative.”[7] Similarly, Kriseman et al “discuss the difficulty of diagnosing APS in patients presenting solely with dermatologic complaints, as these skin manifestations are not specific enough for APS to be included in the Sapporo diagnostic criteria,”[8] and Sibilia notes that aβ2GPI and autoantibodies to phosphatidylethanolamine and annexin-V. “… are not listed in the new diagnostic criteria set (Sapporo, 1999) because their usefulness needs to be validated and their detection standardized.”[9]

There are also several publications that discuss “diagnostic criteria” for APS and include reference to the classification criteria without directly linking the two, but which could be incorrectly misinterpreted by readers as being directly linked. For example, Roselli et al identify that “a thrombotic event with aPL detection is known as APS and the diagnostic criteria include the presence of LA, aCL, and ab2GPI antibodies.”[10] This is indeed true, but as these same antibodies are listed in the classification criteria later in the manuscript (“The international consensus for APS classification includes LA in the laboratory criteria, showing the highest strength of association with thrombotic complications. Furthermore, the consensus criteria also involve the detection of aCL [IgG/IgM] and ab2GPI [IgG/IgM] antibodies”),[10] readers may incorrectly conclude “classification” and “diagnostic” criteria are the same.

So, what are the essential differences in “classification” and “diagnostic” criteria for APS? Essentially, clinicians are able to “diagnose” APS using the “classification” criteria (both clinical and laboratory), but diagnosis of APS is not restricted to the items present in the “classification criteria.” In other words, the “classification” criteria establish a finite list of clinical and laboratory parameters that can be used to identify some “definite” APS manifestation for inclusion in future studies, but a broader list of both clinical and laboratory criteria are available to help diagnose APS. In terms of classification speak, many of these might have previously been identifiable as “noncriteria” clinical manifestations and laboratory tests. The latest classification publication[1] has greatly extended the clinical criteria compared with earlier Sapporo[2] and Sydney[3] classification criteria and also gives weight (or different numerical points) to different criteria.

However, of some interest, the laboratory criteria remain essentially those of the Sydney criteria, being “a positive aPL test (LA, or moderate-to-high titers of aCL or ab2GPI antibodies [IgG or IgM]) within 3 years of the clinical criterion.” The notable differences include a points-based inclusion such that persistent positivity for LA according to the International Society on Thrombosis and Haemostasis Scientific Standardization Committee (ISTH SSC) on LA/aPL criteria,[11] is no longer mandated, although persistent LA scores higher.[1] Another difference is that ISTH SSC recommendations to calculate 99^th percentiles for laboratory assays[11] [12] [13] does not appear in the new classification criteria.[1] Instead, although agreeing that LA test assessment for positivity should progress according to the ISTH guidelines, solid phase aPL assays (aCL, aβ2GPI) remain restricted to ELISA, with “moderate” and “high” positivity defined, respectively, as 40 to 79 and ≥80 units.[1] The ISTH SSC on LA/aPL have also published a guidance document for the laboratory diagnosis of APS, including the use of non-ELISA assays and warning against a fixed cutoff.[12]

The lack of inclusion of other methodologies for solid phase aPL assays is interesting and based (according to the authors) on the lack of clinical research studies using automated test systems. This is a clear separation of classification and diagnostic criteria. The restriction of classification criteria to ELISA assays for aCL and aβ2GPI is in complete odds to current diagnostic practice, as ELISA testing is essentially no longer used in most diagnostic laboratories. As an example, [Fig. 1] shows the makeup of different test platforms for aCL and aβ2GPI in current use in Australasia according to current 2023 data from the RCPAQAP (Royal College of Pathologists of Australasia Quality Assurance Program). There are quite a few different ELISA assays in use, but overall, much fewer laboratories perform ELISA-based assays compared with non-ELISA-based assays. Thus, restricting diagnostic criteria to ELISA methods for aCL and ab2GPI would mean that most patients with APS would never be diagnosed. Also, our additional concerns regarding ELISA assays can be reflected by the small number of overall users, but the large variety of ELISA methods ([Fig. 1]), which may generate greater result variability than other more modern assays. Moreover, although the door has been left open in the new classification criteria to inclusion of other test systems in the future,[1] all solid phase assays show inconsistency and variation in results. This holds true for both ELISA and automated systems, making semiquantitative reporting into low–medium–high titer very difficult.[14] [15] Thus, we need further efforts to harmonize results of solid phase assays, but restriction to ELISA only will not solve that problem. Finally, there are a large number of LA and aPL assays that may have value for diagnosing APS that are not as yet listed in the classification criteria; in particular, assays for antiphosphatidylserine/prothrombin antibodies and anti-domain I of β2-glycoprotein I,[16] [17] [18] [19] or those for Taipan snake venom time, ecarin time, and dilute prothrombin time for LA.[20] [21] [22]

Fig. 1 Methodology in current use for laboratory identification of solid phase aPL (aCL, ab2GPI) (according to current 2023 RCPAQAP [Royal College of Pathologists of Australasia Quality Assurance Program] data). ab2GPI, anti-b2-glycoprotein I; aCL, anticardiolipin antibody; aPL, antiphospholipid antibody.

In conclusion, Classification criteria can be used to diagnose APS, but clinicians should not restrict the diagnosis of APS to only these Classification criteria. Indeed, we would recommend not using classification criteria for diagnosis of APS, unless that patient is planned for inclusion in a clinical trial. Instead, clinicians should be guided by other expert guidelines, such as those from the ISTH SSC for LA/aPL.[11] [12] [13] [14] [18] [19] [20] Finally, according to the ACR Subcommittee on Classification and Response Criteria,[23] diagnosis is defined as determining the cause or nature of a disease by evaluating signs, symptoms, and supporting tests in the single patient. Therefore, diagnostic criteria are a set of signs, symptoms, and tests for use in routine clinical care to guide the clinical decision making in individual patients. Classification criteria are instead standardized definitions used primarily to create well-defined, relatively homogeneous cohorts of patients for clinical research. Thus, they are not intended to capture the entire population of potential patients, but rather the majority of those with key common characteristics of disease. It is therefore quite clear that the ultimate purpose of the use of classification and diagnostic criteria is substantially different. In view of this clear-cut concept, we propose here to exercise great caution in the diagnosis of APS and to avoid using classification criteria for this purpose, unless the intention is for the classification of already diagnosed patients for entry into clinical research. Of course, there is also a need to balance the diagnostic equation and not to overdiagnoses APS by adding too many aPL assays and aPL types to ‘manufacture’ an APS diagnosis based on weak evidence.

[Fig. 2] illustrates some of the concepts we refer to, and [Table 1] summarizes the main differences.

Table 1 Summary of main differences between classification and diagnostic criteria for antiphospholipid syndrome

Classification criteria[1] [2] [3]	Diagnostic criteria
Clinical criteria restricted to certain clinical manifestations, with weightings applied in newest Classification criteria	Broader set of clinical manifestations can be applied to diagnosis of APS
Laboratory criteria restricted to LA, aCL (IgG and IgM) and aβ2GPI (IgG and IgM). LA positivity as per ISTH SSC guidance,[11] and persistence not mandated, but scoring higher. Solid phase aPL (aCL and aβ2GPI) as measured by ELISA, with grading as moderate or high levels defined, respectively, as 40–79 and ≥80 units.	Laboratory criteria can be broader, potentially using other subtypes for solid phase aPL (aCL, aβ2GPI) such as IgA, and also more solid phase assays (e.g., aPS/PT and aDI). Additional methodologies to ELISA are available and indeed now dominating the test landscape. Potential use of 99th percentiles to define cutoff values.

Abbreviations: aβ2GPI, anti-β2-glycoprotein I; aCL, anticardiolipin antibody; aDI, anti-domain I of β2-glycoprotein I; aPL, antiphospholipid antibody; APS, antiphospholipid syndrome; aPS/PT, antiphosphatidylserine/prothrombin; ELISA, enzyme-linked immunosorbent assay; ISTH SSC, International Society on Thrombosis and Haemostasis Scientific Standardization Committee; LA, lupus anticoagulant.

Fig. 2 Distinguishing classification criteria from diagnostic criteria. APS, antiphospholipid syndrome.

Publication History

Article published online:
20 October 2023

• References

• 1 Barbhaiya M, Zuily S, Naden R. et al. ACR/EULAR APS Classification Criteria Collaborators. 2023 ACR/EULAR antiphospholipid syndrome classification criteria. Ann Rheum Dis 2023; 82 (10) ard-2023–224609
• 2 Wilson WA, Gharavi AE, Koike T. et al. International consensus statement on preliminary classification criteria for definite antiphospholipid syndrome: report of an international workshop. Arthritis Rheum 1999; 42 (07) 1309-1311
• 3 Miyakis S, Lockshin MD, Atsumi T. et al. International consensus statement on an update of the classification criteria for definite antiphospholipid syndrome (APS). J Thromb Haemost 2006; 4 (02) 295-306
• 4 Meroni PL, Borghi MO. Antiphospholipid antibody assays in 2021: looking for a predictive value in addition to a diagnostic one. Front Immunol 2021; 12: 726820
• 5 Garcia D, Erkan D. Diagnosis and management of the antiphospholipid syndrome. N Engl J Med 2018; 378 (21) 2010-2021
• 6 Kinoshita Y, Mayumi N, Inaba M, Igarashi T, Katagiri I, Kawana S. Elevated levels of antibodies against phosphatidylserine/prothrombin complex and/or cardiolipin associated with infection and recurrent purpura in a child: a forme fruste of antiphospholipid syndrome?. Dermatol Online J 2015; 21 (07) 13030
• 7 Kamal Eldin TM, Elsherif WT, Elgendy SS, Youssif MS, ElMelegy TT. Assessment of anti-β2 glycoprotein1 antibody in systemic lupus erythematosus patients. Egypt J Immunol 2021; 28 (04) 185-194
• 8 Kriseman YL, Nash JW, Hsu S. Criteria for the diagnosis of antiphospholipid syndrome in patients presenting with dermatologic symptoms. J Am Acad Dermatol 2007; 57 (01) 112-115
• 9 Sibilia J. Antiphospholipid syndrome: why and how should we make the diagnosis?. Joint Bone Spine 2003; 70 (02) 97-102
• 10 Roselli D, Bonifacio MA, Barbuti G, Rossiello MR, Ranieri P, Mariggiò MA. Anti-phosphatidylserine, anti-prothrombin, and anti-annexin V autoantibodies in antiphospholipid syndrome: a real-life study. Diagnostics (Basel) 2023; 13 (15) 2507
• 11 Devreese KMJ, de Groot PG, de Laat B. et al. Guidance from the Scientific and Standardization Committee for lupus anticoagulant/antiphospholipid antibodies of the International Society on Thrombosis and Haemostasis: update of the guidelines for lupus anticoagulant detection and interpretation. J Thromb Haemost 2020; 18 (11) 2828-2839
• 12 Devreese KMJ, Ortel TL, Pengo V, de Laat B. Subcommittee on Lupus Anticoagulant/Antiphospholipid Antibodies. Laboratory criteria for antiphospholipid syndrome: communication from the SSC of the ISTH. J Thromb Haemost 2018; 16 (04) 809-813
• 13 Devreese KM, Pierangeli SS, de Laat B, Tripodi A, Atsumi T, Ortel TL. Subcommittee on Lupus Anticoagulant/Phospholipid/Dependent Antibodies. Testing for antiphospholipid antibodies with solid phase assays: guidance from the SSC of the ISTH. J Thromb Haemost 2014; 12 (05) 792-795
• 14 Vandevelde A, Chayoua W, de Laat B. et al. Semiquantitative interpretation of anticardiolipin and antiβ2glycoprotein I antibodies measured with various analytical platforms: communication from the ISTH SSC Subcommittee on Lupus Anticoagulant/Antiphospholipid Antibodies. J Thromb Haemost 2022; 20 (02) 508-524
• 15 Chayoua W, Kelchtermans H, Moore GW. et al. Detection of anti-cardiolipin and anti-β2glycoprotein I antibodies differs between platforms without influence on association with clinical symptoms. Thromb Haemost 2019; 119 (05) 797-806
• 16 Devreese KMJ. Laboratory testing for non-criteria antiphospholipid antibodies: anti-phosphatidylserine/prothrombin antibodies (aPS/PT). Methods Mol Biol 2023; 2663: 315-327
• 17 Devreese KMJ. Laboratory testing for non-criteria antiphospholipid antibodies: antibodies toward the domain I of beta2-glycoprotein I (aDI). Methods Mol Biol 2023; 2663: 329-340
• 18 Vandevelde A, Gris JC, Moore GW. et al. Added value of antiphosphatidylserine/prothrombin antibodies in the workup of obstetric antiphospholipid syndrome: communication from the ISTH SSC Subcommittee on Lupus Anticoagulant/Antiphospholipid Antibodies. J Thromb Haemost 2023; 21 (07) 1981-1994
• 19 Vandevelde A, Chayoua W, de Laat B. et al. Added value of antiphosphatidylserine/prothrombin antibodies in the workup of thrombotic antiphospholipid syndrome: communication from the ISTH SSC Subcommittee on Lupus Anticoagulant/Antiphospholipid Antibodies. J Thromb Haemost 2022; 20 (09) 2136-2150
• 20 Moore GW, Jones PO, Platton S. et al. International multicenter, multiplatform study to validate Taipan snake venom time as a lupus anticoagulant screening test with ecarin time as the confirmatory test: communication from the ISTH SSC Subcommittee on Lupus Anticoagulant/Antiphospholipid Antibodies. J Thromb Haemost 2021; 19 (12) 3177-3192
• 21 Moore GW. Lupus anticoagulant testing: Taipan snake venom time with ecarin time as confirmatory test. Methods Mol Biol 2023; 2663: 263-274
• 22 Moore GW. Lupus anticoagulant testing: dilute prothrombin time (dPT). Methods Mol Biol 2023; 2663: 275-288
• 23 Aggarwal R, Ringold S, Khanna D. et al. Distinctions between diagnostic and classification criteria?. Arthritis Care Res (Hoboken) 2015; 67 (07) 891-897