Revamped APS Criteria and Their Impact on Clinical Research

Updated criteria for classifying antiphospholipid syndrome (APS) will increase the specificity of identifying patients with the condition to aid clinical research, according to researchers. The revamped criteria, originally presented in draft form at the American College of Rheumatology (ACR) 2022 annual scientific meeting, were published this year in late August.

APS classification criteria were last updated in 2006 and did not incorporate the wider range of manifestations of APS, including heart valve and kidney disease. The ACR and the European Alliance of Associations for Rheumatology (EULAR) jointly commissioned the update.

The new criteria include six clinical domains (macrovascular venous thromboembolism, macrovascular arterial thrombosis, microvascular, obstetric, cardiac valve, and hematologic) and two laboratory domains (antiphospholipid antibody [aPL] tests by solid-phase assay and coagulation assay). The criteria are weighted, with items in each of the eight domains receiving a score between 1 and 7. Patients who score at least three points from all clinical domains as well as three points from the laboratory domains are classified as having APS.

Medscape spoke with two US researchers in the criteria's core planning group, Medha Barbhaiya, MD, MPH, and Doruk Erkan, MD, MPH, — both from Hospital for Special Surgery in New York — to learn more about what these new criteria mean for clinical research.

What is the difference between "classification criteria" and "diagnostic criteria"?

Medha Barbhaiya, MD, MPH

Barbhaiya: Classification criteria, with stringent definitions, are used in research settings with the goal of capturing a homogeneous group of patients with highly specific, key clinical characteristics. Classification criteria do not necessarily include all the clinical or laboratory features of that particular disease. In contrast, diagnostic criteria are used in clinical settings for the purpose of accurately diagnosing and managing diseases in patients.

Was there a need for new APS classification criteria?

Doruk Erkan, MD, MPH

Erkan: Yes, due to the limitations of the revised Sapporo APS classification criteria, and especially with the recent increase in the number of APS clinical trials, there was an absolute need for new classification criteria. In fact, based on our needs-assessment survey completed before the project started, 92% of participants reported the need for new APS classification criteria.

What were the limitations of the revised Sapporo APS classification criteria?

Erkan: The Sapporo criteria were published in 1999 and updated in 2006. Based on our improved understanding of APS, the criteria have been increasingly criticized over the past 20-plus years. These limitations include:

Lack of representation of the systemic nature of APS: Some of the relatively specific aPL-related manifestations were not part of the criteria — for instance, cardiac valve disease.
Lack of risk stratification based on additional thrombosis risk factors: For instance, approximately half of aPL-positive patients with a history of thrombosis have other non-aPL thrombosis and/or cardiovascular disease risk factors at the time of their thrombotic events.
Lack of strict definitions for criteria items, especially for microvascular diseases and aPL positivity
Need for improved definitions, especially for pregnancy morbidity.
Lack of representation of the risk profiles associated with different aPL tests: For instance, IgM anticardiolipin antibodies (aCL) or IgM anti-beta₂-glycoprotein-I antibodies (aβ₂GPI) are less commonly associated with aPL-related clinical manifestations, compared with the IgG isotypes or the lupus anticoagulant (LA) test.

Barbhaiya: In addition, there have been important advances in the methodology of classification criteria development. Since the Sapporo criteria were revised, ACR published new recommendations for development and validation of criteria sets, based on contemporary standards of measurement. These recommendations, using a combination of data-driven and expert-based approaches, have helped develop robust classification criteria across various rheumatic diseases. Additionally, addressing and minimizing various biases related to the selection of cases and controls challenges older classification criteria.

What are some of the key changes to highlight in the new criteria?

Erkan: First, the clinical domain items are expanded and better defined. In the macrovascular domain, venous thromboembolism and arterial thrombosis require risk stratification based on absence or presence of additional risk factors. Additionally, there is a new microvascular domain with very clear definitions, and with a distinction between "suspected" vs "established" findings based on definitive testing. We have included a significant update to the definition of pregnancy morbidity and have added cardiac valve disease and thrombocytopenia.

Second, in the laboratory domain we have included a distinction between IgG and IgM isotypes, and there are very clear descriptions of what "moderate"- or "high"-level aCL and aβ₂GPI ELISA tests mean.

Finally, the new classification system incorporates a weighted score–based algorithm, with higher weight corresponding to items with higher specificit

What does it mean that sensitivity was lowered compared with the previous revised Sapporo criteria?

Barbhaiya: By definition, classification criteria aim to include very stringent definitions to guarantee a high specificity of patients, and that sometimes comes at the cost of sensitivity. Our goal from the outset was to try to achieve high specificity to improve the quality of clinical research in APS. Even though there was a slight lowering of our sensitivity, compared with the previous criteria, it was still very high.

What do these new criteria mean for clinical research? Are these going to help clinical trials better study disease manifestations outside of moderate- to large-vessel thrombosis or pregnancy complications, such as kidney disease?

Erkan: The major problem with APS research has been the inclusion of a heterogeneous group of aPL-positive patients — for instance, patients with both low- and high-risk aPL and/or thrombosis risk profiles, or patients with controversial pregnancy-related manifestations. This heterogeneity makes it difficult to compare and generalize the results of studies in patients with APS. More specific classification criteria will standardize the identification of a homogeneous group of patients with APS for inclusion in clinical studies and trials, thereby enhancing study interpretation and applicability.

To answer the second part of the question, we provided a way to quantify new clinical problems that are highly related to APS, such as cardiac valve disease or thrombocytopenia in the [new] classification criteria. Additionally, we also have a better understanding of microvascular disease in aPL-positive patients, which is really a different game than macrovascular disease, from both the mechanistic and treatment point of view. The old criteria included only "small-vessel thrombosis" without getting into the details of definitions. The new ACR-EULAR criteria have a separate domain, identifying microvascular disease items and defining each of them very clearly.

Barbhaiya: In addition to novel manifestations included in the new criteria, macrovascular events such as venous or arterial thromboses, which were also included in the previous criteria, now require careful consideration of risk factors present at the time of thrombotic events. Weighting thrombotic events by the presence or absence of provoking risk factors may seem like a minor nuance of the new criteria but is critical to how thrombotic events in APS should be considered. Furthermore, quantification of thrombotic events in this way should help ensure that those patients with highly likely APS are truly the ones being included in clinical studies and trials related to potentially risky treatments, such as long-term anticoagulation.

Could these new criteria affect clinical practice?

Erkan: Ideally, classification criteria should not affect clinical practice. We have to accept the fact that there are no diagnostic criteria for APS, and making a diagnosis is a very complex equation. Diagnosis of APS should be established on the basis of clinical manifestations, aPL profile, and other risk factors, but it includes other elements assessed on a case-by-case basis by physicians. Diagnosis is not a computer model or a standardized algorithm. Classification criteria can be a guide on some level but should not be used for diagnosis.

Barbhaiya: Dr Michael Lockshin here at Hospital for Special Surgery always teaches trainees that making a diagnosis is the "art" of medicine, and I believe that holds true here. Making a diagnosis includes consideration of important factors such as individual patient characteristics, less common disease manifestations, use of all available relevant test results, and sometimes even evaluating response to treatment. When we were designing the new classification criteria, the goal was to standardize definitions and identify a cohort of patients with very high specificity for APS to improve research. However, the criteria should not affect diagnostic or treatment decisions related to the individual patient.

Follow Medscape on Facebook, X (formerly known as Twitter), Instagram, and YouTube

Comments

Commenting is limited to medical professionals. To comment please Log-in.

Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.