Role of cerebrospinal fluid diversion in cerebrospinal fluid leak repairs and skull base surgery

Introduction

Whether, when, and how to divert cerebrospinal fluid (CSF) after skull base surgery remains a highly controversial topic, and high-quality clinical evidence to guide decisions is lacking. Both otolaryngologists and neurosurgeons encounter such scenarios in daily practice and have varying algorithms for management. The literature has demonstrated that there may be numerous practice patterns with which to successfully prevent and treat skull base CSF leaks.

Cerebrospinal fluid leak management should be carefully tailored to the cause of the skull base defect and patient comorbidities. CSF leaks can be classified into the following categories: spontaneous, iatrogenic, and traumatic. These leak types have been described in earlier chapters of this book. This chapter reviews the role of CSF diversion in each of these repairs, with particular attention to our own institutional biases. The management of each group is discussed, and a decision-making algorithm is proposed.

Spontaneous cerebrospinal fluid leaks

The association between spontaneous skull base CSF leaks and idiopathic intracranial hypertension (IIH) has been well documented in the neurosurgical and otolaryngologic literature since the 1960s. For many patients with IIH, a spontaneous CSF leak may be the first clinical symptom that prompts workup. For this patient population, the skull base defect and leakage of spinal fluid acts as a natural mechanism of CSF diversion. Patients often present via primary care referral to either a neurosurgeon or otolaryngologist after there is laboratory confirmation via β2 transferrin of CSF rhinorrhea. Neuroimaging can help clinicians to identify the origin of the leak, often revealing osseous dehiscence of the sphenoid or ethmoid in anterior CSF leaks and the tegmen tympani of the temporal bone in lateral CSF leaks.

There is a strong consensus that direct surgical repair is the primary strategy to treat spontaneous CSF leaks. Health care providers differ, however, regarding preoperative optimization of these patients and the use of intraoperative and postoperative CSF diversion techniques. Because the majority of patients with spontaneous CSF leaks likely have IIH, several considerations are necessary for treatment planning. First, it is important to acknowledge that these patients are at high risk of recurrent leakage after direct repair because of high intracranial pressures (ICPs), often obese body habitus, and other comorbidities associated with the diagnosis, such as sleep apnea. Spontaneous CSF leaks are associated with the highest rate of encephalocele formation (50%–100% of cases) and the highest recurrence rate following surgical repair (25%–87% of cases). ^, We believe that multidisciplinary management by the neurosurgeon, otolaryngologist, neurologists specializing in IIH, sleep specialists, weight loss specialists, and primary care providers is a key to successful treatment of these patients.

Acetazolamide

Several studies have reported that primary repair success rates are higher among patients with spontaneous CSF rhinorrhea who had preoperative ICP management with acetazolamide or CSF diversion. Oral acetazolamide has been shown to significantly decrease ICP within 4 to 6 hours. Postoperative acetazolamide is advocated in the 2019 International Consensus Statement on Allergy and Rhinology: Endoscopic Skull Base Surgery (ICAR: ESBS). At our institution, we connect patients with a neurologist specializing in IIH, who initiates therapy with acetazolamide 1 to 2 weeks before the operation to decrease ICP and thereby improve the possibility of a successful procedure. These neurologists also treat headaches associated with IIH; offer weight loss counseling; and prescribe other medications to decrease ICP, such as methazolamide, spironolactone, and furosemide. Neurologists are also helpful in managing rebound headaches, which can occur after CSF leak closure.

Lumbar drainage

Lumbar drains (LDs) are a controversial adjunct in the management of spontaneous CSF leaks, with many studies reporting that perioperative use of LDs is not necessary for successful repair. Other studies have found LDs useful after spontaneous CSF leak repair, not necessarily for diverting CSF but for monitoring and guiding the management of ICPs postoperatively. In a recent prospective study by McCormick et al., all patients with spontaneous CSF leaks had LDs placed after repair to monitor ICP. Patients received acetazolamide if their ICP was elevated, and ventriculoperitoneal shunting (VPS) was placed if the ICP was not adequately reduced with acetazolamide therapy. McCormick et al. argue that for patients with spontaneous CSF leaks, the LD allows for guided control of ICP postoperatively, while not being necessary for other types of CSF leaks. At our institution, LDs are seldom placed in preoperative or postoperative management because of the associated complications.

Lumbar drains, although often inserted with ease and with minimal patient discomfort, are not without risks or drawbacks, especially in obese patients. Overdrainage, the most serious complication associated with LDs, can result in severe headache, vision loss, intracranial hemorrhage, brainstem herniation, and even death. Patients with LDs require close monitoring in the intensive care unit even when their condition is otherwise stable enough for less closely monitored care or discharge home. Other complications occur with LDs as well, including breakage or malfunction of the drains, nerve root irritation, pain, bleeding, and infection. ^{, ,} In our experience, the use of LDs prevents early mobilization, increases the length of hospital stay, and requires higher levels of nursing care and hospital resources than would otherwise be necessary while not increasing the success of skull base repairs. Therefore, for spontaneous CSF leaks, LDs are not a standard component of our management algorithm. Furthermore, we have found a high rate of durable repair when using a multidisciplinary approach, which reduces the need for LDs.

Shunting

Permanent CSF diversion remains highly controversial in the management of spontaneous CSF leaks. At our institution, we avoid permanent CSF diversion in this patient population when possible, reserving this technique for patients with several failed attempts at primary surgical repair combined with adjuvant medication, lifestyle modifications, and bariatric surgery to decrease ICP. The 2019 ICAR: ESBS advocates the use of acetazolamide postoperatively and recommends VPS as an option in those who are intolerant of medication or those with refractory intracranial hypertension and recurrent leaks. Our experience with shunting in these patients suggests a high rate of shunt malfunction from slit ventricles in the setting of a ventricular shunt, abdominal catheter migration from obesity, lumbar radiculopathy in the setting of lumbar shunts, and multiple recurrent emergency department visits to manage patient anxiety and provide reassurance.