Heparin-resistance in AL amyloidosis: a case report

An insufficient anticoagulant effect of heparin is reported in 1–10% of on-pump cardiac surgery cases [1]. Most often, AT-III deficiency is responsible, which is commonly treated with an additional dose of heparin, an infusion of fresh-frozen plasma, or supplementation with AT-III concentrate [1]. However, non-AT-III-mediated heparin-resistance is also observed (Table 1).

Non-fractionated heparin is known for its complex pharmacokinetics and bioactivity. Since the introduction of the cardiopulmonary bypass (CPB) system in 1953, non-fractionated heparin has become indispensable to prevent thrombus formation, ultimately causing end-organ failure and death. After its intravenous administration, non-fractionated heparin binds to antithrombin-III (AT-III) potentiating its activity a 1,000-fold. Subsequently, anticoagulation is mainly established by the inhibition of clotting factors II and X [1, 2]. Variability in its anticoagulant effect, especially when using CPB, mainly depends on the baseline activity of AT-III. However, various pathways of heparin elimination are also involved. When bound to molecules other than AT-III, heparin loses its bioactivity; non-AT-III mediated heparin resistance [1, 3].

In 1967, amyloid light-chain (AL) amyloidosis was suggested to be a causative factor for a decreased activated clotting time (ACT) [4]. Although several cases of non-AT-III mediated heparin resistance have been reported since, the concept of heparin-amyloid complex-formation has not re-emerged and no specific recommendations for non-AT-III mediated heparin-resistance have been reported.

We describe a fatal case of non-AT-III-mediated heparin resistance in a patient with massive AL amyloidosis discovered at autopsy. We hypothesize that close contact between blood flow and large amounts of AL amyloid in filtration-organs could lead to the binding of administered non-fractionated heparin to AL amyloid; then, being biologically unavailable for anticoagulation.

This report aims to address the hypothetical biochemical explanation and discuss treatment alternatives for non-AT-III mediated heparin-resistance.

A 70-yr-old male patient (70 kg) presented with progressive dyspnea and retrosternal pain since six months. Because of severe aortic valve stenosis (peak-gradient 76 mmHg, mean-gradient 51 mmHg, dimensionless index 0.22, hypertrophic left ventricle with ejection fraction 60%) and significant two-vessel coronary artery disease (left descending coronary artery (LAD) and circumflex coronary artery (Cx)), he was scheduled for aortic valve replacement in combination with coronary artery bypass grafting (CABG). Previously, he was physically fit, being active in long-distance ice skating, and without relevant medical history. In the work-up for surgery, he suddenly had a thrombo-embolic event of the right retina, considered being related to arteriosclerotic disease of the right internal carotid artery (stenosis 50–69%), for which treatment with clopidogrel was initiated.

At the cardiac operation (day 1), the target ACT for initiating CPB was not achieved by two subsequent boluses heparin (300–400 IU/kg). An overview of all peri-surgical anticoagulation dosages with corresponding ACT-levels are shown in Table 2.

Suspecting AT-III deficiency, antithrombin and additional heparin were administered. Because antithrombin had been given, fresh-frozen plasma was not considered for additional administration. Nevertheless, the target ACT was still not reached. AT-III deficiency as the cause of heparin-resistance became ruled out as the antithrombin-activity levels just before and after administration of AT-III were found to be within normal limits. Because of the refractory non-antithrombin-III mediated heparin resistance of unknown etiology, it was decided to limit the surgical procedure for now to an off-pump CABG-procedure of the LAD for which an ACT > 200 was sufficient. Anticoagulation with direct thrombin inhibitors was not yet considered as it was not included in the hospital protocol. The off-pump procedure and the postsurgical-care period elapsed uneventful. A percutaneous coronary intervention (PCI) of the Cx and a transcatheter aortic valve implantation (TAVI) were scheduled separately.. Because this had already been discussed as an alternative preoperatively, the choice to switch to this was obvious when CPB appeared to be ineffective. First, the PCI was planned 11 days after the first surgery. Hematological examination took place in the week before. Based on the perioperative course and laboratory results, hyperfibrinogenemia was considered the most likely cause at that time. There were no signs of heparin induced thrombocytopenia (Table 3). Due to a communication error, it was not known to the practitioners at the PCI that heparin had no effect on the patient. Therefore, again the doses of heparin administered were insufficient to achieve the target ACT (Table 2). The PCI procedure was complicated with intracoronary thrombus formation (Cx)that induced myocardial ischaemia and ventricular fibrillation. Therefore, treatment with a GPIIb-IIIa antagonists (prasugrel) and intracoronary alteplase was initiated. The stent was not placed. Post-intervention recovery was without neurological symptoms; the hemodynamic-system needed only a little pharmacological support for optimisation.

However, 5-h later, the clinical condition of the patient deteriorated. A CT of the thorax showed a retrosternal hematoma with a venous blush and a pericardial effusion of about 2–3 cm, and cardiac ultrasound showed borderline tamponade criteria. Emergency surgery was decided and 700 ml blood was evacuated from the pericardial space. Nevertheless, the patient’s condition deteriorated further, for which chemical and manual resuscitation was needed. Urgent installation of CPB was decided. Although bivalirudin was considered, for easiness of use and handling-speed an excess dose of heparin was administered, after which ACT increased to > 999 s (maximum value). However, ACT decreased after 67 min to 281 s. Additional boluses of heparin were administered three more times, maximum ACT reached was 439 s. Further treatment with extracorporeal membrane oxygenation (ECMO) with use of direct thrombin inhibitors was considered if the patients’ clinical condition could be stabilized. However, after two hours of optimal CPB-treatment with use of high doses of vasopressor and inotropic support, there was no recovery of the patients’ hemodynamic and especially metabolic condition, and so, because of poor neurological prognosis, CPB and medical treatment were discontinued, after which the patient died. Postmortem examination showed extensive AL amyloid deposited into the space of Disse in the liver, sinusoids of the spleen and the glomeruli in the kidney (Fig. 1 and supplementary Figs. 1 and 2). Although a small amount of perivascular AL amyloid deposits were found around the coronary arteries, there were no signs of a restrictive cardiomyopathy.