Abstract
Younger patients with cryptogenic stroke typically undergo testing for antiphospholipid antibodies (aPL). Patients with strongly and persistently positive aPL meet criteria for antiphospholipid syndrome (APS), and warfarin is standard of care for secondary prevention of thromboembolic events. In the absence of a clear hypercoagulable condition, antiplatelet agents are the mainstay of secondary stroke prevention. However, some patients have persistent aPL above the reference range but below APS thresholds. It is unclear whether their risk of recurrent thromboembolism justifies anticoagulation over antiplatelet therapy.
We conducted a retrospective cohort study to describe antithrombotic practice patterns and clinical outcomes in patients with weakly versus strongly positive aPL titers after cryptogenic ischemic stroke.
Methods We studied adult patients in the University of Pennsylvania Health System with cryptogenic stroke or transient ischemic attack (TIA) who had sustained aPL positivity for ≥12 weeks between March 1, 2017 and December 31, 2023. We excluded patients without follow-up in our health system, with other indications for antithrombotic therapy, on antithrombotic therapy at the time of the index event, and with active malignancy.
Our control group was the strongly positive aPL group, defined according to ACR/EULAR criteria as those with persistent: (a) lupus anticoagulant (LA) detected; and/or (b) at least one anticardiolipin (aCL) IgM, aCL IgG, anti-ß2-glycoprotein-1(aB2GPI) IgM, or aB2GPI IgG ≥40 units. Our exposure group was the weakly positive aPL group, defined as those with: (a) no LA detected; and persistent (b) aCL IgM, aCL IgG, anti-B2GPI IgM, and/or aB2GPI IgG above the upper limit of the laboratory reference range but <40 units; and (c) None of aCL IgM, aCL IgG, aB2GPI IgM, or aB2GPI IgG ≥40 units.
Our primary outcome was recommended antithrombotic therapy after sustained aPL positivity was confirmed, and whether recommendations differed by group. Secondary outcomes included bleeding and thrombotic sequelae. We described cohort characteristics and performed statistical tests using non-parametric statistics. We performed a survival analysis comparing risk of recurrent ischemic stroke (RIS) using univariate Cox proportional hazards model.
Results Forty-seven patients met inclusion criteria. Thirty patients had weakly positive aPL, and 17 had strongly positive aPL, therefore meeting criteria for APS. The groups were similar in age, sex, and race. They were also similar in rates of active smoking, hypertension, hyperlipidemia, and diabetes mellitus. Most in the weakly positive group (n=25, 83%) had only one positive aPL, most commonly aCL. Few patients in either group had prior thrombotic or bleeding events or pregnancy loss.
In the weakly positive group, antiplatelet therapy was recommended to approximately half of patients (n=16, 54%), warfarin to 30% (n=9), and a direct oral anticoagulant (DOAC) to 17% (n=5). In the strongly positive group, warfarin was recommended to most patients (n=12, 71%), DOACs to 11% (n=2), and antiplatelet agents to 17% (n=3). Warfarin was recommended less often (p=0.007) and antiplatelet agents more often (p=0.029) in the weakly positive group compared to the strongly positive group.
Median follow-up was 32 months (IQR 24,48) and did not differ by group (p=0.765). RIS occurred in 11 patients overall (23%): four weakly positive (13%) and seven strongly positive (41%). In survival analyses, RIS incidence was higher in the strongly positive group [hazard ratio 3.93 (1.14-13.55), p=0.030]. The risk of RIS did not differ by antithrombotic agent used in individual subgroups or overall. Two patients had intracranial hemorrhage, one in each group. One weakly positive patient had heavy menstrual bleeding. There were no gastrointestinal bleeds or instances of hemorrhagic shock. There were no peripheral arterial thromboses, venous thromboemboli, or myocardial infarctions, and no deaths occurred.
Conclusions In summary, though there was substantial practice variation, clinicians recommended antiplatelet agents for most patients with cryptogenic stroke and weakly positive aPL and Coumadin more commonly for strongly positive aPL. Despite less intensive antithrombotic therapy, weakly positive patients had lower rates of RIS compared to strongly positive patients, suggesting they may resemble the general cryptogenic stroke population, though further study is needed.
