27

The balance between prevention of thrombotic phenomena and hemorrhagic complications in the neurosurgical patient population presents an ongoing clinical dilemma. The risk of specific procedures, in the context of patient comorbidities and the associated treatments with antiplatelet and anticoagulant therapy, produces a complex decision tree with many permutations.1 Frequently there exists no solid, evidence-based guidelines on which to frame these decisions and help maximize the risk–benefit profile for a given intervention. This chapter examines the role of anticoagulant and antiplatelet agents in the context of cerebrospinal fluid (CSF) shunts, external ventricular drains, intracranial pressure monitors, and lumbar drains.

There can be substantial variability in treatment paradigms based on institutional protocols as well as surgeon preference. There is a paucity of studies with respect to these specific neurosurgical procedures in the context of antiplatelet and anticoagulant therapy. The treatment plan for patients with brain and spinal catheters who experience hemorrhagic complications is often dictated by their clinical status, the predicted clinical course, and a perceived optimal risk–benefit profile. This chapter highlights some of the clinical issues pertinent to these procedures as well as our own clinical practice-based experience and institutional guidelines. There are ongoing studies specifically addressing some of these clinical dilemmas, but there is a need for further investigation and prospective data collection to determine the optimal guidelines and protocols.

Cerebrospinal Fluid Shunts

Cerebrospinal fluid shunting as a procedure has been practiced since the 1950s. The patient population that receives these devices is primarily pediatric because hydrocephalus is the primary indication for their placement.² However, since the mid-1960s, shunts have been used to treat the diagnosis of normal pressure hydrocephalus (NPH).³ NPH primarily affects patients in the elderly population and therefore has relevance in understanding procedure-associated complications from the use of anticoagulant and antiplatelet agents.

Complications from shunt placement in all age groups include infection, malfunction, seizures, subdural hematoma, and, rarely, intracerebral hemorrhage.⁴ Patients with NPH represent a specific subpopulation that receives shunts, and patients with NPH who are on antiplatelet or anticoagulation therapy, primarily for age-related cardiovascular disease, represent a significant proportion of this group. There have been relatively few investigations that have looked at the hemorrhagic complication rates in patients undergoing ventriculoperitoneal (VP) shunt in the context of these agents.

The occurrence of intracranial ventricular catheter-related hemorrhage is so rare that it is limited to a handful of case reports in the literature, but it is likely an underreported complication of this procedure.5 In our experience 1% of patients demonstrate a significant hemorrhage along the catheter tract on immediate postoperative computed tomography (CT) or magnetic resonance imaging (MRI) scans; however, these hemorrhages are rarely symptomatic.

In addition to routine laboratory testing, it is critically important to obtain an appropriate history with respect to any potential bleeding disorder, including a family history of abnormal bleeding. This can help identify and screen for any potentially high-risk patients prone to hemorrhagic complications and help predict the ability to achieve adequate intraoperative hemostasis.⁶

Because shunts are uncommonly placed on an emergent or even urgent basis, placing them can usually be considered an elective procedure. Patients who are on anticoagulants or antiplatelet agents, therefore, should have an appropriate preoperative medical evaluation and risk stratification. There are no specific evidence-based guidelines to determine perioperative management of patients on anticoagulant and antiplatelet therapy, and many institutions create policies and protocols based on studies that may not be specific or even relevant for a neurosurgical patient.⁷

With the lack of specific guidelines for shunt surgery, our institution’s policies have been put into practice based on the most current recommendations and input from medicine, cardiology, anesthesia, and neurosurgery services. In general, patients who are on aspirin are directed to stop therapy 7 days prior to the initial shunt placement. Patients on clopidogrel stop therapy 7 days prior to initial surgery. Patients on oral anticoagulant therapy (Coumadin) discontinue medication 5 days prior to surgery; a preoperative coagulation profile must document a normal international normalized ratio (INR) prior to proceeding to the operating room. If the patient’s risk profile for cardiovascular events is high, the patient can be bridged off of the oral anticoagulant with low molecular weight heparin (LMWH) and can receive the final dose up to 24 hours preprocedure.

Shunt malfunction is an unfortunate but rather common complication in patients of all ages including those with NPH. In the NPH population, most commonly the malfunction is distal (intra-abdominal) obstruction, and revision surgery for these patients is usually limited to intra-abdominal exploration and repositioning of the distal catheter. Because there is no intracranial catheter manipulation for this procedure, patients are not routinely asked to stop aspirin, with the understanding that there is a small, but increased, possibility of an abdominal wound hematoma. Oral anticoagulants are discontinued as with primary shunt placement.

After a shunt surgery is performed, determining when it is safe to resume anticoagulation or antiplatelet therapy can be more variable. Factors specific to the surgery, postoperative clinical status, and preprocedural risk stratification are taken into consideration. Most often, antiplatelet drugs and anticoagulants will be restarted 5 to 7 days after surgery. However, certain high-risk patients are restarted on antiplatelet agents on postoperative day 2 or 3, and anticoagulants can be restarted on day 3 without using an LMWH bridge while following daily coagulation profiles. A postoperative CT or MRI scan should be obtained prior to resumption of antiplatelet or anticoagulant therapy to document that there is no significant intracranial hemorrhage.

During the perioperative period, the use of routine pharmacological deep venous thrombosis (DVT) prophylaxis is also recommended. The guidelines are generally similar for patients undergoing other intracranial procedures.1 Patients are given prophylactic doses of LMWH daily while admitted. This is stopped 24 hours preprocedure and restarted 24 hours postprocedure, provided that there are no concerns about hemorrhagic complications. In addition to pharmacological DVT prophylaxis, sequential compression devices on the lower extremities are ordered for all patients. Early mobilization and ambulation, when the patient is ready, are also important adjuncts to prevention of DVT and pulmonary embolism (PE) in patients undergoing shunts.

Another common and unfortunate complication of shunted NPH is the development of a subdural hematoma. Very few studies have specifically addressed the risks of a subdural hematoma in shunted patients on antiplatelet or anticoagulant therapy. It is thought that overdrainage and abrupt changes in CSF dynamics can precipitate a subdural hematoma. There are some data to suggest that the use of a programmable valve or flow-regulated valves may lower this risk of a subdural hematoma⁸; however, this has not been confirmed in prospective randomized studies. Some patients may initially present with chronic subdural hygromas and expanded extra-axial spaces. Again, a prudent decision may be to use a valve that limits siphoning, including fixed pressure, flow regulated, or programmable devices for these patients to avoid overdrainage and the expansion of the subdural space, which could increase the risk of precipitating a subdural hematoma.⁹

Patients are monitored closely in the postoperative period, both clinically and radiographically, as needed. There is some evidence specific to NPH patients on warfarin therapy that there is no increased risk of a subdural hematoma in the postoperative period. Additionally, patients whose warfarin was discontinued in the perioperative period and whose coagulation profile is normalized do not necessarily experience an increased risk of thromboembolic complications.⁹ The few studies that have investigated these dilemmas are limited in their design but demonstrate the importance of further investigations in this area.

External Ventricular Drains and Intracranial Pressure Monitors

The use of external ventricular drains (EVDs) and intracranial pressure (ICP) monitors is common and varies widely in the patient populations and the conditions that they treat. It is an important diagnostic and therapeutic tool in many neurosurgical disease processes. There are many causes of acute hydrocephalus in the neurosurgical patient population including acute aneurysmal subarachnoid hemorrhage, traumatic brain injury, intracerebral and intraventricular hemorrhage, postoperative hydrocephalus, and postoperative CSF leak. These devices are commonly placed at the bedside in an intensive care unit (ICU) or emergency room setting. The specific pathology for which these devices are used commonly presents in patients who are currently on antiplatelet or anticoagulant therapy. There is utility in understanding the hemorrhagic complications associated with these devices for patients who are taking antiplatelet or anticoagulant therapy, but again, little has been written or studied in this context.

As with any procedure, there is an inherent risk of hemorrhagic complication in placing an EVD or ICP monitor. This risk has been studied retrospectively and is estimated to range from 1 to 8%. Although most radiographically evident hemorrhages are not clinically significant, up to 2% can be symptomatic.^10,11 No study has addressed the risks specific to all patients on antiplatelet or anticoagulant therapy. However, patients with aneurysmal subarachnoid hemorrhage, requiring EVD placement, have yielded some interesting information and insight into understanding the risks to patients who are on therapy. Patients with acute aneurysmal subarachnoid hemorrhage frequently require placement of a bedside EVD for the development of acute communicating hydrocephalus. With today’s modern endovascular techniques, more patients are receiving definitive treatment of the offending aneurysm using coil embolization or stent-assisted coil embolization. Patients for whom a stent-assisted technique is utilized must be administered aspirin and clopidogrel perioperatively to prevent the high risk of stent-associated thrombotic complications. Kung et al11 found that the rates of radiographic hemorrhage from EVD in the stent-assisted patient group were as high as 32%. There was an 8% risk of symptomatic and clinically significant hemorrhage for these patients. The comparison group of patients receiving only coil embolization, and therefore not requiring aspirin and clopidogrel, had a symptomatic hemorrhage rate of only 0.9%.¹¹

For patients who are on antiplatelet or anticoagulant therapy and present with intraparenchymal or intraventricular hemorrhage, there often exists a need for acute CSF diversion or determination of ICP parameters to optimize best medical management of an otherwise catastrophic intracranial process. Again, there are no guidelines based on level 1 evidence, but it is generally prudent to discontinue the use of these agents, and in most cases reverse these agents prior to placement of an EVD or ICP monitor. Our institutional policy suggests that patients on antiplatelet agents be transfused with platelet products, and patients on anticoagulant therapy should be actively reversed with factors or fresh frozen plasma and vitamin K prior to the procedure. Specific details of reversal agents and their pharmacological details are presented in Chapter 14. Patients should have repeated laboratory testing prior to the procedure to ensure efficacy of the reversal and adequate platelet function and coagulation cascade. When clinical deterioration in a critically ill patient poses time constraints on the reversal process, a risk–benefit analysis should be performed, and a discussion with the patient’s family can help direct appropriate care.

A common complication in patients with intraventricular hemorrhage who require placement of an EVD for acute hydrocephalus is clot-related catheter malfunction. Although there are some bedside maneuvers that can be utilized to flush the clot and reestablish catheter patency, often flow cannot be safely reestablished, and one of the catheters or the ventricular or distal tubing obstructs repeatedly. In some cases, a second EVD can be placed contralaterally when the ICP is increasing and no CSF is draining. In less emergent situations, and when ICP is stable, another option that has become more widely accepted involves the use of intrathecal recombinant tissue plasminogen activator (rtPA).¹² A useful protocol involves removal of 6 mL of CSF prior to injection of 1 mg in 1 mL of rtPA reconstituted in preservative-free sterile saline. The rtPA is flushed into the ventricular system with 5 mL of sterile saline, and the EVD is then clamped for 1 hour. During this time, the ICP is continuously monitored. If the ICP increases above 20 mm Hg, then the drain can be opened after all other maneuvers have been attempted to lower the ICP. The rtPA can be injected every 12 hours, and serial head CTs can be obtained to monitor for clot resolution. Although still controversial, there is some evidence that this method may confer benefit to long-term outcomes.¹²

Patients who require placement of EVD or ICP monitor for CSF diversion or diagnosis of malignant intracranial hypertension are often critically ill and neurologically devastated. Such patients are at extremely high risk for development and propagation of DVT. To date, no specific studies have addressed the safety or efficacy of pharmacological DVT prophylaxis in patients with EVD or ICP monitors. Our routine practice has been to obtain a postprocedure noncontrast head CT. If there is no evidence of acute catheter- or device-associated hemorrhage, patients are started on subcutaneous heparin (SQH) 24 hours after the procedure as routine DVT prophylaxis. Unpublished preliminary data from our own retrospective review suggest that there is no increased risk of EVD or ICP monitor-related hemorrhagic complications for patients on SQH. Again, the use of pneumatic compression devices is an important adjunct in the prevention of DVT in this patient population. For patients who develop DVT or PE and have an EVD or ICP monitor in place, full anticoagulation is not recommended. A risk–benefit analysis can be performed, and in some cases a decision to remove the EVD or ICP monitor in favor of systemic anticoagulation can be made. For patients in whom systemic anticoagulation is not an option, the placement of an inferior vena cava filter may offer important protection from PE.13

Lumbar Drains and Lumbar Puncture

Lumbar drains (LDs) are used in a variety of clinical situations and are part of everyday clinical practice for most neurosurgery centers. Their utility varies widely, ranging from CSF diversion for postoperative or traumatic CSF leaks, to routine intraoperative placement for intracranial aneurysm surgery and other skull-based approaches. They are often utilized for preoperative workup and assessment of NPH patients.¹⁴ More recently, there has been some evidence supporting the placement of drains in patients undergoing aortic aneurysm stents or open repair procedures to increase postoperative spinal cord perfusion.¹⁵ Despite their wide clinical use, there are few published data regarding the safety of the procedure and of ongoing use in patients on antiplatelet or anticoagulant therapy.

The primary hemorrhage-related risks and potential complications from LD use involve lumbar epidural hematoma and overdrainage or low-pressure phenomena including the risk of developing a subdural hematoma.

Patients who undergo LD placement should be monitored closely for the development of any neurologic changes to suggest an evolving complication. A preprocedure motor and sensory exam should be well documented, and surveillance monitoring performed regularly postprocedure. The advent of CSF collections systems that limit the amount of CSF drainage per unit of time has been important in reducing the chance of overdrainage-related complications including low pressure headaches and subdural hematoma.¹⁴ Despite the increased safety margin, any unexpected change in neurologic status should prompt an appropriate clinical evaluation and radiographic study.

When considering the use of LDs in the context of patients who are on antiplatelet or anticoagulant therapy, several important points should be noted. There is no direct evidence or published data to assess the risk of hemorrhagic complications from pharmacological DVT prophylaxis in patients undergoing CSF diversion with an LD. A risk–benefit analysis can be attempted prior to initiating prophylactic dosing. A patient with severe traumatic brain injury (TBI) undergoing lumbar drainage who is bed bound has a high risk for DVT, whereas a patient who has an LD placed for NPH workup and is encouraged to ambulate and be mobile while hospitalized is at much lower risk. Some NPH protocols for LD placement include subcutaneous heparin administered twice a day, with very low complication rates.¹⁴ Although not definitive, there are also data from the vascular surgery literature to suggest that even systemic heparinization intraoperatively does not increase the risk of an epidural hematoma or subdural hematoma when an LD is placed.15

For patients who present for elective LD placement, our practice has been to manage antiplatelet and anticoagulation therapy in a manner similar to our protocol for patients who present for elective VP shunt placement. Preprocedure regimens are stopped and normal coagulation profiles documented. After the LD is removed, the patients may be restarted on the preprocedure therapeutic regimen without additional radiographic imaging unless there has been a deterioration in their clinical examination or new symptomatology. For patients who require more urgent placement of an LD, patients on antiplatelet or anticoagulant therapy can be reversed using agents and protocols appropriate for the degree of urgency.

Another common procedure, and similar in technique to the LD, is the lumbar puncture (LP). The LP is one of the most widely utilized bedside procedures in clinical medicine and its indications are plentiful. Its utility as a diagnostic test, as well as a therapeutic intervention, and the low cost and its relative safety and efficacy make it an attractive procedure. There are limited studies looking at the hemorrhagic complications specifically in patients on antiplatelet or anticoagulant therapy. As with the LD, it is important to document a normal coagulation profile and confer competent platelet function prior to performing an LP. Depending on the indication for the procedure, an appropriate reversal of any antiplatelet or anticoagulant therapy can be pursued prior to the procedure. If there is concern that a patient might have inadequate postprocedure hemostasis, the patient should be monitored closely and any change in neurologic status should be worked up. As with the LD, there is a small risk of a lumbar epidural or subdural hematoma. Although the LP is meant for CSF sampling as compared with continuous CSF diversion, a small percentage of patients will continue to drain CSF into the epidural space postprocedure and are therefore at risk for the development of overdrainage phenomena including an intracranial subdural hematoma. A lumbar epidural blood patch may be required for patients who continue to experience post-LP overdrainage symptoms.¹⁶

Other Considerations

The use of indwelling brain and spinal catheters is not limited to CSF shunts, EVDs, ICP monitors, and LDs. There are many variations on a theme, including, depth electrodes for deep brain stimulation, subdural electrodes for seizure surveillance and isolation of seizure focus, and intracranial and lumbar intrathecal drug delivery devices, to name a few. As with the aforementioned brain and spinal catheters, there are even fewer studies examining the role of hemorrhagic complications for patients with these indwelling devices. As the use of antiplatelet and anticoagulant agents will inevitably continue to increase, it will be important to define the role of these agents in the context of the wide variety of implants and devices that are utilized in the field of neurosurgery.

Although avoidance of all hemorrhagic complications in patients undergoing placement of brain and spinal implants is impractical, a thorough understanding of patients’ coagulation status and appropriate and thoughtful perioperative management of antiplatelet and anticoagulant therapies can help minimize risk. Balancing the prevention of hemorrhagic complications and thromboembolic events relies on thorough risk stratification and the judicious timing and use of pharmacological agents in the perioperative period. Close monitoring of clinical status can help identify complications early and direct care in an appropriate manner. Prospective data collection will be an important factor in characterizing the true risks for these given interventions and developing appropriate protocol driven guidelines for the future.