This autoimmune disorder is characterized by antibodies directed against specific glycoproteins (β2-glycoprotein-I antibodies) within the cellular phospholipid bilayer [3]. Specifically implicated is the oxidized form of β2GPI, a complement control protein which may be antigenic in selected individuals. The proposed pathogenesis involves a two-hit process of endothelial injury followed by thrombotic response to oxidative stress injury. Several additional interrelated mechanisms include activation of coagulation factors, monocytes, endothelial cells, and platelets. Antibodies directed against annexin II on endothelial cells may lead to both activation and local injury, thus promoting a procoagulant environment. Recurrent fetal loss and increased thrombogenicity have been linked to complement C5a-mediated inflammation [4, 5].

Antiphospholipid antibody syndrome has been associated with both arterial and venous thromboembolism. Venous thrombosis is three times more prevalent compared to arterial events. Arterial events however must be promptly recognized due to their potentially devastating presentation as part of the so-called catastrophic antiphospholipid antibody syndrome . This syndrome requires multisystem organ failure (≥3 organs) occurring within 1 week combined with histopathologic evidence of arterial occlusion. When identified, patients should be screened by transesophageal echocardiography for nonbacterial thrombotic endocarditis (Libman–Sacks endocarditis) which may serve as a potential source for recurrent embolism [6]. This form of endocarditis may occur in up to one third of patients with primary antiphospholipid syndrome and half of patients with secondary antiphospholipid syndrome in the setting of systemic lupus erythematosus.

Venous thrombosis may present as the typical lower extremity deep vein thrombosis with pulmonary embolism or may involve atypical venous locations such as the cerebral, splanchnic, renal, or gonadal venous circulations. Prompt anticoagulation is a necessity whereby risk of propagation and recurrence in the absence of anticoagulation can be as high as eightfold [7, 8]. Epidemiologic data supports a differential risk relationship between the presence of lupus anticoagulant, antiphospholipid antibodies, and risk of thrombosis [9]. The Leiden Thrombophilia Study reported a 3.6-fold increased risk for deep vein thrombosis in individuals positive for lupus anticoagulant (LA), while this risk increase to 10.1 for patients positive for both LA and either antiprothrombin or anti-β2-glycoprotein-1 antibodies [10]. Patients who are “triple positive” for lupus anticoagulant, antiphospholipid antibodies, and β2-glycoprotein-1 antibodies are at the highest thrombotic risk .

Laboratory testing for antiphospholipid antibody syndrome is broadly based on the use of clot-based assays for lupus anticoagulant and enzyme-linked immunosorbent assay (ELISA)-based assays for antiphospholipid or β2-glycoprotein-1 antibodies [5]. For lupus anticoagulant testing , the general approach includes multiple clot-based assays. This approach balances the desire to maximize case identification yield without compromising specificity. The general test repertoire includes the dilute Russell Viper Venom Time (dRVVT), aPTT with platelet neutralization procedure, and Staclot aPTT. The thrombin time and reptilase time are used to search for heparin or direct thrombin inhibitors which degrade test accuracy. Each assay combines the use of mixing studies to differentiate between assay inhibition (which would point to a lupus inhibitor) and factor deficiency. “Assay inhibition ” is defined as clotting times which remain prolonged despite equal volume mixing with pooled donor plasma. If the equal volume mixing study normalizes the clotting time, then a “factor deficiency” is considered. When the assay shows inhibition, then the final step involves adding excess phospholipid to form a “sink” for the antibody. This final step results in a significant shortening of the clotting time and implies a phospholipid-dependent mechanism for clotting assay time prolongation consistent with a lupus anticoagulant. Our approach to this testing is to perform each step in the process including the screening, mixing, and confirmatory steps for consistency and quality assurance. If the confirmatory step is negative, then a specific factor inhibitor is sought. “Factor inhibitors” however present clinically with a bleeding phenotype in contrast to the thrombosis phenotype associated with a lupus anticoagulant. In general, all of the anticoagulants (heparins, warfarin, and direct factor inhibitors) have the potential for degrading test sensitivity and specificity and should be avoided if feasible at the time of testing.

To establish the diagnosis of antiphospholipid antibody syndrome, positive testing for either a lupus anticoagulant or antiphospholipid antibody is sufficient. Between 30 and 40 % of patients will have both a lupus anticoagulant and antiphospholipid antibody (by ELISA). If positive, both the clotting assays and the ELISA must be repeated to establish persistence at least 12 weeks or later [2]. An important caveat for ELISA testing of antiphospholipid antibodies (IgG, IgM) and β2-glycoprotein-I antibodies is that modest titers of these antibodies do not constitute a firm diagnosis. To support the diagnosis of antiphospholipid antibody syndrome, confirmatory titers must exceed 40 units. It has been our experience that providers often accept a weakly positive titer (<40 units) as diagnostic of this syndrome. Due to the lifelong implications of this diagnosis, we recommend strict application of the diagnostic criteria so as not to mislabel individuals .

If confirmed, the next step is to determine whether the patient has primary or secondary antiphospholipid syndrome. Patients with primary disease are more likely female with a mean age of 40 years [11]. For older patients with a new diagnosis of antiphospholipid antibody syndrome, a search for secondary causes is important. Relevant secondary causes include connective tissue diseases, both solid tumors and hematologic malignancies, medications, and infections particularly viral. Most viral-induced antibodies are not thought to carry a thrombotic propensity.

Given high risk of recurrence, anticoagulant therapy is required. Warfarin therapy remains the cornerstone of management. Randomized warfarin trials comparing INR targets of standard intensity (2.0–3.0) to high intensity (3.0–4.0 or 3.0–4.5) resulted in similar or improved efficacy for those patients randomized to the standard intensity INR targets [12, 13]. Higher-intensity nomograms have the limitation of greater bleeding outcomes. Therefore, current clinical practice is to use the standard INR goal for most patients. An important point is that these trials included a minority of patients with arterial thrombosis (24 and 11 %, respectively). Whether a higher target for those patients with arterial thrombosis would be beneficial is not known. We would typically recommend aspirin therapy added to warfarin therapy .

For patients with a baseline prolongation of the prothrombin time, this assay cannot be reliably used to monitor warfarin therapy. Fortunately, only 5–10 % of patients have this baseline abnormality. For those rare patients with baseline prothrombin time inhibition, we favor monitoring with factor Xa or factor IIa (thrombin) assays aiming for a therapeutic target of between 20 and 30 % protein activity. It would be attractive to use novel anticoagulants, including factor Xa inhibitors and direct thrombin inhibitors, as an alternative to warfarin particularly in this setting. The use of these agents has not been well studied, and recommendations regarding their use must be tempered by concerns regarding potential failures. We have recently identified three consecutive patients with antiphospholipid antibody syndrome suffering arterial events who failed these direct factor inhibitor therapy [14]. Since the publication of this small cohort, we have identified two additional patients who failed novel anticoagulant therapy.

For patients suffering acute arterial occlusion in the setting of antiphospholipid antibody syndrome, we recommend fibrinolytic therapy over immediate bypass grafting whenever feasible. In this setting of acute arterial occlusion, bypass grafting has a high likelihood of early thrombotic failure regardless of the type of conduit used. We further recommend aggressive anticoagulant therapy during the acute and convalescent time frame. Prolonged use of subcutaneous low-molecular weight heparin at the twice-daily therapeutic dose has been quite useful in such patients. This form of therapy is advantageous due to immediate therapeutic levels, lack of monitoring in the setting of an inhibited aPTT (lupus anticoagulant), and generally low rate of major bleeding. Conversion to warfarin requires clinical judgment and does not need to be done promptly. For example, we often continue outpatient therapeutic low-molecular weight heparin for several months after an acute event before entertaining this conversion .

HIT is characterized by the development of platelet-activating, noncomplement-fixing IgG antibodies directed against heparin–platelet factor 4 (H-PF4) complex [16]. Heparin is a negatively charged polysaccharide which complexes with PF4 unfolding and exposing neoepitopes within the protein. These neoepitopes may be highly antigenic and induce antibody formation in a small subset of patients receiving heparin. The degree of antigenicity corresponds to the net negative charge of the complex which is directly related to the size of the heparin molecule exposure. As such, unfractionated heparin carries a much higher risk of developing HIT (5 %) compared to low-molecular weight heparin (<1 %) or pentasaccharide therapy (fondaparinux—negligible risk) [17, 18]. Once formed, the heparin-PF4 antibody complex binds and activates platelets through interaction with the Fc receptor domain on the platelet surface. Platelet activation results in the release of procoagulant microparticles, platelet aggregation, leukocyte binding through platelet P-selectin expression, and coagulation factor activation. Tissue factor expression with cytokine release from monocytes and macrophages has also been reported [19]. This process ultimately leads to platelet sequestration and thrombocytopenia. Whereas endothelial cells are decorated with heparan sulfate and platelet factor 4, direct antibody-mediated vascular injury further promotes thrombus formation .

The incidence of HIT ranges between 1 and 5 %, depending on the type of heparin used, the route of administration, associated comorbidities, and clinical settings [15]. Patients undergoing major joint replacement surgery appear to be at particular risk [20]. Vascular surgeons may be consulted by orthopedic colleagues for acute limb ischemia postoperatively and should keep this entity in mind. Up to 20 % of patients who undergo vascular surgical procedures develop heparin-associated antibodies, which are associated with a greater than twofold increased risk for thrombotic complications [21]. By comparison, outpatient LMWH therapy carries such a low risk that regular platelet monitoring is not recommended .

When considering the diagnosis of HIT, it is important to define the pretest probability of disease [15]. The Warkentin “4 T” probability assessment tool for HIT is well validated and quite useful for this purpose. This tool includes an assessment of platelet counts relative to baseline, timing of thrombocytopenia in relation to heparin initiation, thrombosis assessment, and exclusion of other causes (Table 25.1). Each of these variables is given a weighted score which is then summed. Depending on the summed score, patients are deemed low, intermediate, or high pretest probability of having HIT. For those patients at low risk (score ≤ 3), the clinician should search for other causes of thrombocytopenia. These patients have <5 % probability of having HIT such that testing for the heparin-PF4 antibody will be of low yield. The use of heparin in such patients may be pursued unless the degree of platelet decrement precludes its use. For patients at intermediate to high risk of HIT, heparin must be discontinued and confirmatory testing with ELISA for heparin-PF4 antibodies pursued. While waiting for these test results, most patients should receive a direct thrombin inhibitor such as argatroban. For patients with negative assay results, we recommend careful clinical reevaluation and request for hematology consultation. The heparin-PF4 ELISA carries a high negative predictive value approaching 90 % for the exclusion of clinical HIT. For equivocal test results, we recommend repeat testing and if still equivocal, obtaining a serotonin release assay. The serotonin release assay carries high specificity but lower sensitivity compared to the heparin-PF4 ELISA. If the serotonin release assay is negative, the clinical diagnosis is effectively excluded. In the setting of intermediate to high pretest probability of disease, a positive assay result confirms the diagnosis of HIT.

In the clinical assessment of HIT, there are several caveats worth noting. First, there is an entity termed “nonimmune HIT” or type I HIT which occurs within 4 days of heparin exposure and results in only a modest and temporary platelet count decline (100–150,000). Patients with this type of HIT are not at risk of thrombosis, and the key is that platelet counts recover despite continued heparin exposure. This is the most common cause of heparin-associated platelet count decline but is harmless. The mechanism may include platelet “agglutination” from heparin–platelet binding. Second, it is important to scrutinize the platelet counts in total whereby a 50 % decline may still be a normal platelet count if the baseline counts were ≥300. Third, thrombocytopenia will include a 50 % drop from baseline but only rarely results in severely reduced platelet counts. The mean platelet count from published cohorts is approximately 60,000 (range 20,000–100,000). For patients with platelet counts less than 20,000, other causes should be sought particularly drug-induced thrombocytopenia due to non-heparin drugs. Exclusion of pseudothrombocytopenia is accomplished by simply repeating the phlebotomy using citrate in place of EDTA.

HIT may present as an isolated asymptomatic fall in platelet counts (“isolated HIT”) without further difficulties [15, 22]. However up to 50 % of patients developing HIT will suffer a thrombotic complication within the ensuing 30 days (“HIT with thrombosis, HITT”). The risk of thrombosis persists long after heparin is discontinued whereby heparin-PF4 antibodies may circulate for 100 days or longer postheparin discontinuation. With heparin reexposure during this time window, abrupt thrombocytopenia with thrombosis may occur (“rapid onset HIT”). This clinical presentation accounts for up to 30 % of HIT cases and does not represent an amnestic response but rather is due to persistently circulating antibodies. For this reason, it is important to draw a baseline and follow up CBC within the first 24 h of starting heparin in all patients exposed to heparin within the past 100 days. “Delayed onset HIT” is an infrequent (3–5 %) but uncomfortable subtype of HIT which occurs several days after stopping heparin typically when the patient may have already been discharged from the hospital. It is important to recognize this entity whereby treating a new thrombus with heparin in such patients can have devastating consequences.

Venous thrombosis exceeds arterial thrombotic by nearly 4:1. Phlegmasia cerulea dolens with associated digital gangrene may mimic arterial occlusion and must be kept in the differential diagnosis. Other thrombotic complications reported in patients with HIT include overt disseminated intravascular coagulation with hypofibrinogenemia, bilateral adrenal hemorrhage and necrosis, and cutaneous necrosis at sites of heparin injection (heparin skin necrosis) [10, 11, 23–25].

Appropriate platelet monitoring is essential for early diagnosis of HIT [15, 26]. A baseline platelet count should be obtained prior to heparin initiation. In heparin-naïve patients, the platelet count should be monitored every other day beginning 5 days after starting heparin and continued for 14 days or until heparin is discontinued whichever is shorter [15]. In patients with previous heparin exposure, platelets should be monitored with the start of unfractionated heparin use and continued as for heparin-naïve patients.

Prompt recognition of HIT , immediate cessation of all forms of heparin, and rapid initiation of a direct thrombin inhibitor are essential to reduce morbidity and mortality [22]. Approximately 50 % of patients with HIT characterized by thrombocytopenia alone (“isolated HIT”) will experience thrombosis within the ensuing 30 days if a direct thrombin inhibitor is not initiated [27]. The use of warfarin alone in these patients is not protective, and warfarin loading after a diagnosis of HIT can result in severe thrombotic complications including acute arterial occlusion [27, 28]. Parenteral direct thrombin inhibitor therapy is the cornerstone of initial treatment. There are currently three available agents for this purpose: argatroban, bivalirudin, and desirudin. Of these only argatroban is FDA approved for this indication. For patients with renal disease, argatroban is an ideal agent whereby it is cleared by the liver. For patients with liver disease, bivalirudin would be preferred. Both of these agents are attractive given their relatively short half-life. These agents are managed by monitoring the aPTT. Both agents to some degree also prolong the PT. Therefore, with concordant warfarin initiation, we recommend aiming for International normalized ratio (INR) values of 4–6 before stopping the direct thrombin inhibitor. The INR can then be reassessed 4 h later to insure that a therapeutic target has been achieved (INR 2.0–3.0). We recommend continued anticoagulation for a minimum of 3 months depending on the clinical situation. This will allow complete clearance of the circulating antibody. Warfarin initiation should be delayed until the platelet count rebounds to values exceeding 150,000, and large initiating doses should be avoided. Alternative agents which may be useful include fondaparinux (parenteral indirect factor Xa inhibitor), dabigatran (oral direct thrombin inhibitor), rivaroxaban, and apixaban (oral direct factor Xa inhibitors). Clinical experience with the novel anticoagulants in this setting is limited, and they are not FDA approved for such indications .

For many patients with vascular disease, heparin reexposure is required for additional procedures particularly cardiopulmonary bypass [29]. In our published experience, this can be safely accomplished once the patient has cleared the heparin-PF4 antibodies. In this case, a brief isolated rechallenge for purposes of the bypass run has been safe. We recommend restarting a direct thrombin inhibitor as soon as feasible postoperatively to avoid prolonged heparin reexposure .