Post Operative Thrombosis Among Patients Testing Indeterminate for Lupus Anticoagulant

Background: The presence of the lupus anticoagulant (LA) is an established risk factor for thrombosis especially in the post operative setting. The risk of thrombosis among patients testing indeterminate for lupus anticoagulant is unknown. In our study we aim to estimate the incidence of postoperative thrombosis in this population and determine risk factors.

Methods: We studied adult patients undergoing LA testing within 3 months of major surgery at the Cleveland Clinic between 2008 and 2013. The International Society for Thrombosis and Hemostasis (ISTH) defines the following criteria for a positive lupus anticoagulant: 1) a prolonged phospholipid dependent screening test, 2) an inhibitor demonstrated on a mixing study with normal plasma, 3) evidence that the inhibitor is phospholipid dependent, and 4) absence of a coexisting factor inhibitor, direct thrombin inhibitor or heparin. Patients fulfilling some but not all of the ISTH criteria were considered to be LA indeterminate. Patients with previous venous thromboembolism (VTE) and hypercoagulable states were excluded. We collected data on patient demographics, surgery and VTE in the first 30 postoperative days. Patients were divided into different groups based upon the presence of concomitant malignant or rheumatological disease, types of surgery and the utilization of thromboprophylaxis. Associations between categorical variables were determined using Fisher’s exact test. Multivariate logistic regression was performed; variables included age, type of surgery and utilization of pharmacologic thromboprophylaxis.

Results: Of 791patients undergoing perioperative lupus anticoagulant testing, 176 were diagnosed LA indeterminate. The median age was 55 years with males comprising 52.3% of the population. Twenty six (14.7%) patients had a concomitant malignancy. Eighteen patients (10.2%) were diagnosed with a rheumatological illness. Seventy four (42.1%) patients underwent cardiac and vascular surgical procedures. General surgery including gastrointestinal surgery accounted for 53 (30.1%) patients. Neurosurgical patients (20) comprised 11.4% of the population studied. Fifteen (8.5%) patients underwent orthopedic surgery and fourteen (8%) patients had urologic procedures. Thirty eight (21.6%,CI 16.1-28.3%) patients developed VTE in the first 30 postoperative days. Of the patients with VTE, 16 (42.1%) had isolated deep vein thrombosis (DVT). Six patients had DVT associated with internal jugular vein (IJV) thrombosis (15.8%). Five patients (13.2%) patients had DVT associated with PE. Two patients (5.2%)had IJV DVT and PE in combination. Thrombosis was also reported at the sites of arteriovenous grafts, portal vein and in the central retinal vein. Twelve (31.5%) clots were related to the presence of indwelling central venous catheters.

No significant association between presence of cancer or rheumatological disease and incident thrombosis was identified. Of those tested for beta 2 glycoprotein antibodies and anticardiolipin antibodies, no significant association was observed between presence of post-operative thrombosis and the presence of antibodies. While only 70 patients (39.8%) received any form of pharmacological prophylaxis against VTE, there was significant reduction in the incidence of VTE in patients that received prophylaxis as compared to those that did not. A statistically significant increased odds of thrombosis was observed in the neurosurgical population as compared to the patients undergoing other surgical procedures. On multivariate logistic regression analysis, neurosurgical patients had 3.4 fold increased risk of post operative thrombosis (CI 1.3-9.3) when compared to the rest of the surgical population studied, irrespective of the utilization of thromboprophylaxis.

Conclusion: The incidence of thrombosis for patients with an incidental finding of LA indeterminate is 21.6%, comparable to that seen in the general population of patients undergoing similar procedures. This implies that standard guidelines should be used in choosing appropriate post-operative thromboprophylaxis in patients with this laboratory finding. More aggressive anticoagulant regimens do not appear to be necessary, although this remains to be confirmed in a controlled randomized study.