The advantages of central neuraxial block are well known, but when it is used in patients receiving anticoagulant drugs it may lead to vertebral canal bleeding. The incidence of vertebral canal haematoma is low; after reviewing 1.5 million patients, Tryba estimated the risk as 1/150,000 epidural anaesthetics and 1/220,000 spinal anaesthetics. However, the risk is higher in patients who have received drugs that impair coagulation.
In 1994, Vandermeulen carried out a literature review to identify case reports of vertebral canal haematoma associated with central neuraxial block and found only 61 cases published between 1906 and 1994. Of these cases, 46 were associated with epidural techniques, 32 involving a catheter and the remainder occurred after spinal anaesthesia. 68% of the cases occurred in patients who were taking anticoagulant drugs (most commonly intravenous unfractionated heparin) or had a coagulopathy. The other common risk factor was technical difficulties with block performance. Coagulation abnormalities and/or technical difficulties were associated with over 87% of these cases. Since Vandermeulen’s review was published there have been over 70 case reports of vertebral canal haematoma associated with central neuraxial block in patients taking low molecular weight heparin (LMWH). Many surgical patients take drugs that affect normal coagulation, therefore the risk associated with each drug must be estimated (Figure 1).
Deep venous thrombosis (DVT) and pulmonary embolism are commonly associated with surgical procedures. Major lower limb or pelvic surgery and trauma put patients at high risk for these events and malignancy increases the risk 7-fold. The literature
recommends anticoagulant thromboprophylaxis for these high-risk patients.1 These patients also stand to gain most from regional anaesthetic techniques so guidelines are required to allow anaesthetists to proceed safely. Central neuraxial block significantly reduces the incidence of DVT after orthopaedic surgery but additional prophylaxis is necessary to reduce the rate to acceptable levels.
Aspirin and non-steroidal anti-inflammatory drugs (NSAIDs) impair platelet function by inhibiting platelet cyclo-oxygenase (COX). Aspirin inhibits COX irreversibly while NSAIDs do so reversibly. Therefore the antiplatelet effect of aspirin persists until a new platelet population is manufactured (at least 7 days), whereas platelet function returns to normal within 3 days after stopping NSAIDs. Despite their widespread use for many years, there have been only five case reports of vertebral canal haematoma in patients receiving aspirin or NSAIDs alone. It is safe to proceed with central neuraxial block in patients taking these drugs, a view endorsed by the American Society of Regional Anaesthesia.2
COX-2 inhibitors: celecoxib and rofecoxib are anti-inflammatory drugs that selectively inhibit COX-2, which is not expressed in platelets and therefore they do not affect platelet function. They are safe when used alone, but can potentiate the effect of warfarin by increasing the prothrombin time.
Clopidogrel is a thienopyridine derivative and is a potent antiplatelet agent. It inhibits ADP-induced platelet aggregation and binding between platelets and fibrinogen. These effects are irreversible and platelet function does not return to normal until at least 7 days after stopping the drug. In 2001, the Clopidogrel in Unstable angina to prevent Recurrent Events (CURE) trial was published. Since then there has been a move to use clopidogrel in combination with aspirin in patients with acute coronary syndrome. However, there have been many case reports of increased and even fatal surgical bleeding complications associated with clopidogrel. In one observational study in patients undergoing coronary artery bypass grafting, the re-exploration rate for bleeding in patients taking clopidogrel and aspirin was 10.4% compared with 2.2% in those taking aspirin alone.3
There have been three cases of vertebral canal haematoma associated with central neuraxial block in patients taking clopidogrel and two cases of major bleeding following lumbar sympathetic block (one was fatal). The clopidogrel datasheet states that it should be discontinued 7 days before surgery. This interval should also be observed before carrying out any central neuraxial or peripheral block. It is recommended that individual hospitals adopt policies to ensure that clopidogrel is discontinued before surgery. If an antiplatelet effect must be maintained, aspirin can be substituted safely.
Platelet glycoprotein IIb/IIIa antagonists include abciximab, eptifibatide and tirofiban. They have been recommended to prevent coronary ischaemic events in high-risk patients (NICE September 2002 see www.nice.org.uk/pdf/12_Glycoprotein_A4_summary.pdf). Central neuraxial block should be avoided until platelet aggregation has returned to normal; a minimum of 8 hours after tirofiban and eptifibatide and 24–48 hours after abciximab. The data sheets for these drugs state that they are contraindicated within 4–6 weeks of trauma or major surgery.
Unfractionated heparin has been used for many years for thromboprophylaxis and therapeutic anticoagulation. Subcutaneous thromboprophylactic doses are seldom associated with bleeding complications and are not considered to increase the risk of vertebral canal haematoma significantly. Expert opinion recommends performing central neuraxial block 4 hours after the administration of unfractionated heparin or giving the drug after the block has been carried out. Patients who have been receiving unfractionated heparin for more than 4 days should have a platelet count, because the incidence of heparin-induced thrombocytopenia is about 3%.
Therapeutic anticoagulation with heparin is a contraindication to regional block and an intravenous heparin infusion should be discontinued for 4 hours and the activated partial thromboplastin time (APTT) should be normal before attempting a block or removing a catheter.
LMWHs have longer half-lives than unfractionated heparin, which allows once daily administration. They have fibrinolytic properties as well as anti-Xa activity. There is no LMWH monitoring test for routine use. In the late 1990s in North America, over 40 cases of vertebral canal haematoma occurred following central neuraxial block in patients receiving LMWH. This may have been because North American dose guidelines recommended twice daily LMWH administration, meaning there was no ‘safe’ time to perform a block or remove an epidural catheter. A similar cluster of cases was not reported in Europe even though LMWH had been available for 4 years longer, suggesting that the once daily administration regimen used there was safer.
Current guidelines recommend waiting at least 12 hours after LMWH administration before central neuraxial block or catheter removal. Postoperative LMWH dosing provides acceptable thromboprophylaxis. It is recommended that the first dose is given within 6 hours of surgery.
If high-dose LMWH is used for therapeutic anticoagulation it takes about 24 hours for coagulation to return to normal. Therefore, an interval of 24 hours should elapse before attempting central neuraxial block or peripheral nerve block.
Fondaparinux: this new thromboprophylactic drug is a synthetic pentasaccharide, which has potent anti-Xa activity. It has a longer elimination half-life than LMWH of 17 hours in young patients and 21 hours in healthy elderly patients. It is administered 6 hours after surgery, which makes decisions regarding regional anaesthesia easier. However, its long half-life means that it should be used with caution in patients with neuraxial or peripheral nerve catheters in situ. An interval of at least 24 hours should elapse before removal of neuraxial or peripheral nerve catheters.
Warfarin is indicated for thromboembolic prophylaxis in many cardiovascular conditions. Many older patients presenting for surgery are taking long-term warfarin therapy. For all but the most minor surgical procedures, warfarin should be stopped and the international normalized ratio (INR) allowed to reduce to 1.5 or less; this normally takes about 4 days. It is often appropriate to start an alternative anticoagulant, such as LMWH, until warfarin is re-established and the INR is back in the therapeutic range postoperatively. An INR of 1.5 or less is associated with normal haemostasis and it is therefore safe to proceed with central neuraxial or peripheral nerve block. If a LMWH has been administered in place of warfarin, the recommended intervals discussed above should be observed before performing any block. Regional anaesthetic techniques should be avoided in patients with an INR over 2.0.
Peripheral nerve block and anticoagulants
Most reports of serious bleeding events associated with peripheral nerve block have followed psoas compartment (lumbar plexus) blocks or lumbar sympathectomy. Several large retroperitoneal haematomas have occurred following these techniques. One case was associated with perioperative administration of heparin and warfarin, two with LMWH and two with thienopyridine antiplatelet drugs (one with ticlodipine; one with clopidogrel that was fatal). Major bleeding was of greater significance than neural compression in all these cases.
The guidelines in Figure 2 could be applied to peripheral nerve block performance, but some regional anaesthetists would argue that this would be unnecessarily restrictive. Careful consideration of the risks and benefits of a peripheral nerve block on an individual patient is the most logical way to practise.
1 Hirsch J, Dalen J E, Guyatt G. 6th (2000) American College of Chest Physicians guidelines for antithrombotic therapy for prevention and treatment of thrombosis. Chest 2001; 119: supplement.
2 Horlocker T T, Wedel D J, Benzon H et al. Regional anesthesia in the anticoagulated patient: defining the risks. (The second ASRA Consensus Conference on Neuraxial Anesthesia and Anticoagulation.) Reg Anesth Pain Med 2003; 28(3): 172–97.
3 Yende S, Wunderink R G. Effects of clopidogrel on bleeding after coronary artery bypass surgery. Crit Care Med 2001; 29: 2271–5.
Copyright © 2004 The Medicine Publishing Company Ltd