11

Mechanical prophylaxis, which includes compressive stockings and sequential compression devices (SCDs), is uniformly recommended for all neurosurgical patients regardless of the type of procedure undertaken. The risks are minimal with stockings and SCDs, whereas the potential gains are substantial. Although the use of SCDs has not been compared with the absence of treatment in a randomized trial, the incidence of VTE in neurosurgical patients with only mechanical prophylaxis has been studied. Naturally, the incidence of DVT in mechanical-only prophylaxis varies linearly with the intensity of screening methods (i.e., using duplex ultrasonography in a patient only when symptoms are present versus routine regular screening) from 3.2 to 43%.1–11 The average incidence is around 25% in studies with routine screening. Similarly, the exact conversion rate of DVT to PE is unknown and varies with screening technique, but is estimated to be 0.5 to 5%. PEs carry approximately an 18 to 60% mortality rate,8,12–14 and there are other well-known risks for patients treated with systemic anticoagulation, including gastrointestinal hemorrhage, skin necrosis, systemic allergic responses, and renal failure.

Table 11.1 Summary of Deep Vein Thrombosis (DVT) Prophylaxis Guidelines

* High-risk features: Surgery > 6 hours, malignancy, prolonged immobility

** Complex spine-specific high-risk features: history of VTE, malignancy, known hypercoagulability, prolonged immobilization > 2 weeks, staged procedures > 5 levels, combined anterior-posterior approaches, iliocaval manipulation during exposure, planned anesthetic time > 8 hours
Abbreviations: IVC, inferior vena cava; LMWH, low molecular weight heparin; SCI, spinal cord injury TBI, traumatic brain injury; UFH, unfractionated heparin; VTE, venous thromboembolism.

The most up-to-date guidelines recommend chemical DVT prophylaxis for surgical patients “once the risk of bleeding diminishes.”15 Thus, the crux of the issue is when does the risk of bleeding become outweighed by the risk of VTE? The answer is not straightforward, but here we review each subset of neurosurgical patients routinely encountered in practice, with final recommendations at the conclusion of each subset (summarized in Table 11.1).

Deep Vein Thrombosis Prophylaxis in Adult Craniotomy Patients

Browd et al16 reviewed the major studies that have examined the incidence of VTE and ICH in neurosurgical patients on DVT prophylaxis undergoing craniotomy. Since that review, only one other large series has been published,⁸ and it confirmed the conclusions that the incidence of VTE in patients receiving chemical DVT prophylaxis is always reduced compared with patients receiving mechanical compression only, and that the rate of VTE in patients receiving chemical prophylaxis is 0 to 18.7%, with a mean of around 15% for patients routinely screened with duplex. Thus, the relative risk reduction of DVT when using subcutaneous chemoprophylaxis is 18 to 82%, with an average of 40%.^1–11 Additionally, the rate of ICH in the nonchemically treated groups was 0 to 4.3%, compared with 0 to 10.9% in patients receiving chemical prophylaxis. When patients who were treated preoperatively with chemical prophylaxis are excluded, the ICH range drops to a more reasonable level of 0 to 2.6%. This range is still wide and is the main reason why creating a risk/benefit analysis is so difficult. Indeed, a decision analytic model dealing with this variability using both sensitivity analysis and Monte Carlo simulation¹⁷ found, rather surprisingly, that mechanical prophylaxis only yielded superior outcomes compared with either unfractionated heparin (UFH) or low molecular weight heparin (LMWH) in patients undergoing craniotomy. There were, however, significant assumptions made when “rating” the effective quality-of-life effects of DVT, PE, and ICH—thus making this conclusion rather arbitrary.

Timing and Type of DVT Prophylaxis

The timing and type of chemical DVT prophylaxis—UFH or LMWH—after craniotomy is another area of controversy. Contrary to recommendations in other surgical cohorts, routine preoperative DVT prophylaxis in neurosurgical patients is not recommended because of the unacceptably high ICH rate demonstrated in older studies.¹⁴ Furthermore, many institutions have a rather arbitrary protocol of waiting 12 to 48 hours after surgery before starting subcutaneous chemical DVT prophylaxis. This postoperative “waiting period” is intended to reduce the incidence of postoperative ICH by allowing adequate hemostasis before beginning DVT prophylaxis. Any immediate postoperative ICH on computed tomography scan would usually warrant an extension of this waiting period.

Only a handful of studies have evaluated the type and timing of postoperative DVT chemical prophylaxis. LMWH has a greater anti–factor Xa/anti–factor IIa activity, greater bioavailability, more predictable anticoagulatory effects, and longer duration of action than UFH. None of the various forms of LMWH has been routinely proven superior to another. Given this information, LMHW has replaced UFH in general surgery and internal medicine patients; however, UFH is still very popular for postoperative use in neurosurgical patients because of the older studies that suggest a possible elevated ICH rate in LMWH.

In 2000, Iorio and Agnelli¹⁸ published a meta-analysis of the LMWH risk/benefit in neurosurgical patients. They found that, on average, LMWH reduced the relative rate of DVT by 38%, which was in accordance with previous studies, with a major ICH rate of 2.2%. The authors concluded that for every 11 thrombotic events prevented by LMWH, one major nonfatal ICH would occur.

Khaldi et al8 studied the risk of hemorrhage and DVT in patients receiving UFH 24 or 48 hours postoperatively. They found no difference in VTE or ICH in either group, with a DVT incidence of 16%. Additionally, they found that 92% of DVTs occurred in the first 2 weeks after surgery, and the incidence had a linear relationship to the length of surgery, with a statistically significant increased risk of PE in surgeries lasting > 6 hours.

Recommendations

No craniotomy patient should be treated with preoperative subcutaneous heparin for the sake of improving DVT prophylaxis. All patients should receive intraoperative and postoperative mechanical DVT prophylaxis. Subcutaneous UFH chemoprophylaxis should be initiated 24 to 48 hours postoperatively to reduce the risk of DVT by 40%. If the surgery lasts longer than 6 hours or other high-risk features are present (malignancy or prolonged immobility), an initial use of or switch to subcutaneous LMWH can be considered, weighing against the slightly higher risk of hemorrhage.

Deep Vein Thrombosis Prophylaxis in Traumatic Brain Injury

Incidence of DVT in Traumatic Brain Injury

Serious head injuries are an independent risk factor for DVT. The rate of DVT in patients with moderate to severe traumatic brain injury (TBI) is up to 33%.^19–23 Different causes, such as increased circulating levels of tissue factor and von Willebrand factor, have been proposed, but the exact reasons remain unclear. Interestingly, even with properly initiated chemoprophylaxis, the rate of DVT formation in patients with severe head injury is three to four times higher than in matched controls.²³ To compound matters further, Norwood et al²⁴ reported an intracranial bleeding risk of 9.1% in patients with TBI who required craniotomies when enoxaparin prophylaxis was initiated within 24 hours of the procedure.

Timing and Type of Chemoprophylaxis in Traumatic Brain Injury

Choosing when to initiate chemoprophylaxis in the TBI patient is difficult, as is choosing the type of heparin to use. The risk of development of DVT has to be cautiously weighed against the potentially increased risk of bleeding complications; however, the literature is quite clear on a few key issues. First, withholding DVT chemoprophylaxis for > 48 hours after TBI resulted in a fivefold increase in the rate of DVT.²³ Additionally, in analyzing the results in all trauma patients together, UFH administered every 8 hours has proven noninferior to LMWH every 12 hours in a large prospective study.²⁵ This information, coupled with the results of the study by Norwood et al²⁴ on LMWH, suggests that UFH should be initiated in all trauma patients within 48 hours of admission. Additionally, because the risk of DVT is high even on chemoprophylaxis, aggressive screening with weekly Doppler ultrasonography should be considered. If chemoprophylaxis cannot be initiated, the use of an inferior vena cava (IVC) filter (as described below) should be considered. Extended DVT prophylaxis during inpatient rehabilitation can be considered in patients with TBI, but the rate of new DVT after an inpatient stay is only 2 to 5%.26

Recommendations

The use of UFH every 8 hours should be initiated in all TBI patients within 48 hours of admission. Screening with weekly Doppler ultrasound examinations should be considered. If chemoprophylaxis cannot be initiated, the use of an IVC filter should be strongly considered.

Incidence of Deep Vein Thrombosis in Traumatic Spinal Cord Injury

Incidence of DVT

The reported incidence of DVT in patients with spinal cord injury (SCI) varies widely depending on the screening method used, with rates ranging from 9 to 100%.²⁷ Notably, VTE causes nearly 10% of all deaths in the first year after SCI. The cause for such an increased risk is multifactorial, but again harkens back to Virchow’s triad of stasis, hypercoagulability, and vessel intimal injury. Additional pathophysiological mechanisms have also been proposed, including impaired circadian variations of hemostatic and fibrinolytic parameters and changes in platelet function and fibrinolytic activity.²⁸

Type and Timing of DVT Prophylaxis

Previous systematic review has suggested that LMWH is almost three times as effective as UFH in preventing DVT in the SCI population,^29,30 but the same is not necessarily true with PE. Additionally, only a handful of studies have investigated the timing of early versus late administration of subcutaneous DVT prophylaxis. One of these studies found a 10-fold increase in DVT when LMWH was started > 72 hours after injury.³¹ As for duration of prophylaxis in SCI patients, level III evidence suggests that nearly 90% of all DVTs in patients with SCI occur within the first 3 months, and studies have confirmed the utility of chemoprophylaxis for this time frame.³² If the patient has a known history of VTE, 6 months of chemoprophylaxis should be considered.

Recommendations

Patients with SCI are arguably at the highest risk for VTE of any neurosurgical subgroup. The SCI patient should have LMHW initiated within 72 hours of injury for a total of at least 12 weeks. If chemoprophylaxis with LMWH cannot be initiated, an IVC filter should be strongly considered. Chemical prophylaxis should be stopped 24 hours before any planned surgery and resumed 24 hours after.

Deep Vein Thrombosis Prophylaxis in Adult Patients Undergoing Elective Spinal Surgery

One large meta-analysis33 has been performed on studies of patients undergoing elective spine surgery. As with cranial patients, screening protocols and methods varied, making definitive conclusions difficult, but this meta-analysis found that the rate of DVT in patients undergoing elective spine surgery is 6% when no prophylaxis is used, 2% when mechanical prophylaxis is used, and < 0.01% when mechanical prophylaxis and LMWH are used. The results are similar in the cervical and the thoracolumbar spine. Patients undergoing a combined anterior and posterior reconstruction have the highest absolute rate of DVT (14–18%), but the small sample sizes in these studies prevented drawing definitive conclusions.³⁴ The incidence of PE in the published series of patients undergoing elective spine surgery was exceedingly low at 0.06%. The incidence of postoperative epidural hematoma was 0% in the group that did not receive any chemical prophylaxis compared with 0.4% in the LMWH group, with 38% of those patients developing permanent defect. Similar data have been confirmed by other, more recent literature reviews as well.³⁵ The use of preoperative IVC filters has been studied in elective spinal reconstruction patients. These important data and clinical recommendations are discussed later in this chapter.

Recommendations

For the patient undergoing elective spine surgery that does not involve a major spinal reconstruction, mechanical prophylaxis alone is sufficient DVT prophylaxis. If other high-risk features are present (malignancy or prolonged immobility) and there is no contraindication to heparin, subcutaneous chemical prophylaxis should be considered for 24 to 48 hours postoperatively. Spine reconstruction patients face different risks, which are discussed below.

Deep Vein Thrombosis Prophylaxis in Pediatric Neurosurgical Patients

Incidence of DVT in Pediatric Populations

The incidence of DVT in the hospitalized pediatric population is exceedingly low but on the rise; it approaches 0.2% in North America.³⁶ As stays in the intensive care unit are prolonged with indwelling central venous catheters, the incidence of associated DVT naturally increases.

Recommendations

As the baseline rate of VTE is so low in children, level III guidelines recommend screening only the children who are symptomatic or have multiple proven risk factors of VTE (e.g., prolonged immobility, underlying malignancy, sepsis, protracted length of indwelling central venous access) and not to treat with any postoperative chemical prophylaxis unless the child has a history of VTE.

The EVDs are typically inserted either at the bedside in the intensive care unit or in the operating room. Their insertion is associated with a hemorrhage rate of 5 to 41%,37–41 with only a 0.5 to 2% risk of the hemorrhages becoming symptomatic. EVDs are often inserted before the definitive surgery to secure the source of hemorrhage; thus, their insertion contributes little to the algorithm of when to start chemical DVT prophylaxis. New symptomatic hemorrhage from EVD removal has been reported, however, in patients on subcutaneous LMWH and aspirin after coiling of a ruptured aneurysm.⁴¹ Although this was a report of a small number of patients, it suggested that there was an increased risk with either insertion or removal of an EVD within 24 hours of a patient receiving subcutaneous DVT prophylaxis. There is, however, considerable evidence to the contrary. Hoh et al³⁹ reported a 9.2% EVD-associated hemorrhage rate even when a patient received full-systemic therapeutic anticoagulation with heparin during endovascular coiling of a ruptured aneurysm within 24 hours after ventriculostomy placement. Of these hemorrhages, only one patient (0.8%) was symptomatic. If the activated partial thromboplastin time was kept below 90, there was a 0% incidence of EVD-related hemorrhage from full systemic heparinization in their series. Thus, if a patient can tolerate a partial thromboplastin time up to 90 without increased hemorrhage rate within 24 hours, it can reasonably be concluded that starting UFH or LMWH for DVT prophylaxis 24 hours after an EVD placement would likely not result in an increased rate of statistically significant ICH. On the opposite end of the spectrum are patients receiving dual antiplatelet therapies after stent-assisted coiling. These patients suffered a 32% EVD hemorrhage rate, with a quarter of these hemorrhages becoming symptomatic.⁴² Thus, if a patient with an EVD is on dual antiplatelet therapy for a stent, subcutaneous DVT prophylaxis can be justifiably withheld.

Inferior Vena Cava Filters in Neurosurgical Patients

Theoretically, IVC filters are inserted to prevent pulmonary embolism; however, the insertion of these devices is not without risk. Complications include insertion-site hematoma (1%), IVC thrombosis that is often asymptomatic (up to 18%),⁴³ postthrombotic syndrome (7–40%), migration into the superior vena cava or right atria (2–3%), and even erosion through the wall of the IVC into the duodenum or renal collecting system.43,44 Permanent IVC filters seem to lose their initial benefit in preventing PE and actually increase the morbidity of long-term recurrence of DVTs. With these safety concerns, retrievable filters have been developed to take advantage of short-term PE prevention benefit without the long-term disadvantages in patients with temporary contraindications to anticoagulation.⁴⁴ Placing a retrievable filter in a patient requires a system to ensure the patient is not lost to follow-up, as the chances of removing the filter diminish exponentially after 6 weeks in most cases.

IVC Filters in Neurosurgery Patients

Retrievable IVC filters are routinely used in two unique situations in neurosurgery patients. The first, in which there is clear benefit, is in a neurosurgery patient with a newly diagnosed DVT who is unable to receive systemic anticoagulation because of the risk of bleeding. Absolute contraindication to systemic anticoagulation is somewhat controversial but certainly includes recent (< 48 hours) intracranial or spinal surgery, ongoing gastrointestinal bleeding, severe thrombocytopenia, severe uncontrolled hypertension, or a known bleeding diathesis. IVC filters should be placed as soon as possible in these patients to prevent pulmonary embolus.

The second use, which remains controversial, is a prophylactic usage in patients with absolute contraindications to subcutaneous DVT prophylaxis who are at high risk for VTE. The two types of neurosurgical patients that may meet this definition are those with multiple DVT risk factors preparing to undergo a complex spinal reconstruction and those with acute multisystem neurotrauma/trauma. Both of these situations are discussed below.

IVC Filters as Prophylaxis in the Adult Spine Reconstruction Patient

Major adult spinal reconstruction is regarded as a contraindication to anticoagulation; however, reconstructive spine surgery is a recognized risk factor for VTE.^45–48 Rosner et al⁴⁸ performed a pilot study of 22 patients and found decreased mortality and VTE rate in patients receiving a preoperative IVC filter for major spinal reconstruction. McClendon et al⁴⁶ followed this pilot study with a large 219-patient study in high-risk patients undergoing adult spinal reconstruction. The definition of “high risk” included any of the following: history of DVT or PE, concurrent malignancy, hypercoagulability, prolonged immobilization (bedridden > 2 weeks before surgery), staged procedures of longer than five segment levels, combined anterior–posterior approaches, iliocaval manipulation during exposure, or planned anesthetic time of more than 8 hours. Venous lower extremity surveillance duplex Doppler ultrasonography was performed on all patients in the early postoperative period and then weekly. The lower extremity DVT rate was 18.7% (41/219 patients). The documented PE rate was 3.7% (8/219 patients) as measured by computed tomography (CT) angiography. The incidence of developing VTE (whether DVT or PE) was significantly higher in patients with anesthesia over 8 hours. Comparing the DVT-to-PE conversion rate in filtered versus nonfiltered historical controls with the same high-risk criteria resulted in a statistically significant decrease in rate of PE, with an odds ratio of 3.7.⁴⁷

IVC Filters as Prophylaxis in the Neurotrauma/Polytrauma Patient

Pulmonary embolism is often a preventable cause of late morbidity and mortality after trauma, as it occurs in up to 1.5 to 9% of patients who survive their initial trauma.44 In high-risk injury patterns, including severe closed head injury and spinal cord injury with paraplegia or quadriplegia prophylactic, IVC filters carry a level III evidence of recommendation for prophylactic use.⁴⁹ In a large meta-analysis in which over half the included studies had sample sizes greater than 100 patients, the use of IVC filters in these predetermined trauma patients definitively decreased PE (0–9%) and PE-related death (0–0.8%) compared with historical controls.

Recommendations for IVC Filters in Neurosurgery Patients

Inferior vena cava filters should be used in any patient with a known DVT who is unable to undergo systemic anticoagulation for prevention of PE. IVC filters should be strongly considered for PE prophylaxis in TBI and SCI patients without DVT who are unable to receive subcutaneous DVT prophylaxis. Additionally, high-risk patients (those with previous history of DVT or PE, concurrent malignancy, hypercoagulability, or prolonged immobilization) preparing to undergo complex spine reconstruction, particularly if the general anesthetic time will be more than 8 hours, should be considered for a prophylactic filter as well.