Reclassification of obstetric antiphospholipid syndrome under the 2023 ACR/EULAR criteria: risk of under-recognition in patients with non-criteria antibodies and clinical implications.

OBJECTIVE: To characterize the clinical, serologic, and treatment features of patients with obstetric antiphospholipid syndrome (OAPS), determine the proportion meeting the laboratory domain of the 2023 American College of Rheumatology/European Alliance of Associations for Rheumatology (ACR/EULAR) classification criteria, and examine patients identified through non-criteria antibodies.

METHODS: We retrospectively reviewed consecutive women with a prior clinical diagnosis of OAPS at our center and assigned de-identified codes P1-P45. Of the 45 patients, 41 had sufficiently complete historical serologic documentation for laboratory-domain reclassification; P10, P17, P34 and P42 were retained in the descriptive cohort but treated as not assessable in the evaluable-case analysis. Two investigators independently scored each evaluable patient across the six clinical domains and two laboratory domains; classification required a cumulative score of ≥3 in both the clinical and laboratory domains. For patients with concomitant systemic lupus erythematosus (SLE), obstetric or hematologic manifestations attributable to SLE were excluded from APS scoring.

RESULTS: A total of 45 patients were included, all female, with a mean age of 33.7 years (43 primary APS, 2 SLE-associated). Thirty-two patients (71.1%) had a history of pregnancy loss, and 20 had recurrent miscarriage. Anticardiolipin (aCL) IgG and IgM were positive in 9 and 14 patients, respectively; anti-β2-glycoprotein I (anti-β2GPI) IgG and IgM were positive in 2 and 16 patients, respectively; lupus anticoagulant (LAC) was positive in 2 patients. Fourteen patients (31.1%) carried non-criteria antibodies, including 7 with non-criteria antibodies only. Among the 41 patients with evaluable serologic data, 7/41 (17.1%) definitively met the ≥3-point laboratory threshold and 11/41 (26.8%) met it under the upper-bound assumptions. In the conservative sensitivity analysis, which counted P10, P17, P34 and P42 as below threshold, the corresponding estimates were 7/45 (15.6%) and 11/45 (24.4%). Patients with isolated IgM positivity, IgA-only antibodies, single-positive LAC, or non-criteria antibodies alone did not meet the threshold. An exploratory comparison found no statistically significant difference in documented standardized intervention between live-birth and further-loss cases [28/36 [77.8%] vs 6/9 [66.7%]; two-sided Fisher's exact p = 0.666].

CONCLUSION: In this single-center retrospective series of 45 patients, many patients were identified through non-criteria antibodies or low-weight serologic profiles and therefore did not meet the 2023 laboratory-domain threshold. Because of the small sample, these findings are descriptive and hypothesis-generating and should not be generalized. Nevertheless, they support using the 2023 criteria primarily to define homogeneous research cohorts rather than as a stand-alone basis for diagnosis or treatment decisions. Whether the obstetric phenotype requires a separate classification approach remains to be determined.