Historical Background

The first neuroendoscopic procedure was performed in 1910 by Vincent Darwin L’Espinasse, a Chicago urologist, who used a cystoscope to fulgurate the choroid plexus in two infants with hydrocephalus.² One of the patients died; however, the other lived for 5 years. Walter Dandy performed several procedures for hydrocephalus. In 1918, he performed the first open ventriculostomy by fenestrating the lamina terminalis through a transfrontal approach.³ In 1922, he unsuccessfully attempted a choroid plexectomy using the endoscope and at that time coined the terms “ventriculoscope” and “ventriculoscopy.”⁴ Fay and Grant were the first to photograph the inside of the ventricular system in 1923. Interestingly, the exposure time for their images ranged from 30 to 90 seconds, accentuating the extremely poor lighting and visualization of neuroendoscopy at that time.⁵

This first successful endoscopic third ventriculostomy (ETV) was performed in 1923 by William J. Mixter in a 9-month-old hydrocephalic patient. Utilizing a urethroscope, he was able to pass through the foramen of Monro, visualize the third ventricle and cerebral aqueduct, and make a hole in the floor of the third ventricle connecting it to the interpeduncular cistern. He subsequently enlarged the hole with side-to-side mechanical motion up to a diameter of 4 mm.⁶ The next successful ETV did not appear in the literature until 1935, when Scharff reported his initial results using a novel endoscope. Scharff described several modifications including an irrigation system that kept the ventricles open, a mobile cautery tip, and a moveable operating tip that could perforate the floor of the third ventricle.⁷ Around the same time, resection of the choroid plexus was still a popular treatment for hydrocephalus. Tracy Putnam modified a cystoscope for use in the ventricular system, specifically designing it to cauterize choroid plexus in order to treat children with hydrocephalus.⁸ In one of his series of 42 patients, which was the largest at that time, 17 patients (40%) responded favorably while 10 patients (25%) died perioperatively.⁹ In 1947, McNickle described a modified technique of performing a percutaneous third ventriculostomy utilizing a 19-gauge needle to puncture the floor of the third ventricle. Although he initially used an endoscope for the procedure, he later abandoned its use in favor of using x-rays and tactile feedback for localization.¹⁰

Despite the advances made by a few pioneering surgeons, neuroendoscopic procedures were not widely adopted during the first half of the 20th century.^11–13 Technology was a significant limiting factor as lighting and visualization were very poor and the setup of the endoscopic lenses was complicated and cumbersome.^11,12 In 1952, Nulsen and Spitz first reported creation of a shunt diverting cerebrospinal fluid (CSF) from the ventricular system to the jugular vein to successfully treat communicating hydrocephalus.¹⁴ This began a monumental shift in the hydrocephalus treatment paradigm toward shunt diversion. Although a select few surgeons would continue performing neuroendoscopic procedures to treat hydrocephalus, the majority of neurosurgeons adopted CSF shunt diversion as standard practice due to the low complexity and safety of the procedure.¹²

During this hiatus in the clinical use of neuroendoscopy, critical advances in technology would eventually lead to revisiting the role of neuroendoscopy in neurosurgery. First-generation endoscopes utilized conventional lenses, which had a fixed refractive index. This required the lenses of the endoscope to be carefully placed in series in order to relay an image. In 1966, Hopkins and Storz revolutionized lens technology by developing the SELFOC lens. These lenses used gradient-index glass, which allowed adjustment based on the lens radius.¹⁵ Compared to conventional lenses, a wider field of vision could be viewed while preserving the illumination.^12,15,16 In 1969, Smith and Boyle developed the charge-coupled device, which allowed optical data to be converted into electrical current.¹² This precluded the need for large relay stations and allowed for the development of smaller endoscopes. Perhaps most important was the development and improvement of fiberoptic cables that occurred between the 1950s to 1970s. Using fiberoptic cable technology, light could be separated from the rest of the endoscope and emitted from the tip without emitting significant amounts of heat or losing luminescence.^12,15

With improved technology, use of the endoscope was revisited for hydrocephalus treatment. In 1978, Vries reported a series of five patients who he had treated with ETV utilizing a fiberoptic endoscope.¹⁷ Although all of his patients eventually required a shunt diversion procedure, the patients did well postoperatively demonstrating that ETV was technically feasible with current technology. In 1990, Jones et al. reported 50% successful outcome from ETV in 24 patients with noncommunicating hydrocephalus with 8% morbidity and no mortality.¹⁸ In 1994, the same group reported an even better success rate (61%) in a series of 103 patients.¹⁹ Given the technological advances in endoscopic illumination and optics, the demonstrated safety of the procedure with second-generation endoscopes, and the clinical efficacy reported in the literature, ETV was widely adopted in the 1990’s by the neurosurgical community to treat hydrocephalus.^1,12

Patient Selection

Selection of the appropriate patient for ETV requires an understanding of CSF physiology, detailed study of radiographic anatomy, an appreciation for the location of CSF obstruction, and knowledge of results for various age groups and etiologies of hydrocephalus.^3,20 The CSF circulation is analogous to a hydraulic system. CSF flows sequentially through each chamber starting in the choroid plexus until being reabsorbed in the arachnoid villi. Pathologic obstruction can occur at any point in the circuit or at the end. ETV treats hydrocephalus by creating an internal shunt from the third ventricle to the interpeduncular cistern in the cortical subarachnoid space (Fig. 96-1). If the pathologic obstruction lies between the third ventricle and the cortical subarachnoid space, then ETV will likely be an effective treatment. If the pathologic obstruction lies downstream from the cortical subarachnoid space, then ETV will be of no benefit.²⁰

FIGURE 96-1 Schematic of bulk flow model of CSF hydraulics from the point of production in the choroid plexus until its absorption and incorporation into the systemic circulation. A, Aqueductal stenosis is portrayed along with the resultant triventriculomegaly. B, ETV creates a bypass from the third ventricle to the cortical subarachnoid space and results in resolution of the triventriculomegaly.

(Adapted from Rekate H. Selecting patients for endoscopic third ventriculostomy. Neurosurg Clin North Am. 2004;15(1):39.)

Pathologic Considerations

The etiology of the pathologic obstruction should be determined when considering the most appropriate treatment for hydrocephalus. The most responsive etiologies of hydrocephalus to ETV are tumors of the tectal region and congenital aqueductal stenosis.^21,22 Tectal gliomas or pineal region tumors can compress the cerebral aqueduct, obstructing CSF flow from the third ventricle to the fourth ventricle. Likewise, congenital aqueductal stenosis also results in impaired flow from the third ventricle to the fourth ventricle. It can present in infants, children, adolescents, or adults.^23,24 The later it presents, the better chance that ETV will successfully treat the hydrocephalus. Care must be taken not to confuse congenital aqueductal stenosis with narrowing of the cerebral aqueduct that occurs secondary to inward displacement of the brainstem. This scenario can be present in infants or young children with increased pressure of the superior sagittal sinus leading to decreased terminal absorption of CSF. When the sutures are open, this can lead to global increases in CSF volume, including in the ventricular system and the basal cisterns. An increase in the CSF volume in the prepontine and interpeduncular cisterns can produce compression of the brainstem leading to secondary compression of the cerebral aqueduct, mimicking aqueductal stenosis.²⁰ In these cases, the site of obstruction lies distal to the interpeduncular cistern and placement of a ventricular shunt, not ETV, would be the indicated intervention.^20,25

Nontectal/pineal region tumors producing hydrocephalus can also benefit from ETV if properly selected.^26–28 Prior to surgical intervention, the exact point of CSF obstruction needs to be determined. If the point of obstruction occurs along the course of CSF flow between the third ventricle and the interpeduncular cistern, then ETV may be useful. If the point of obstruction lies outside this course, then shunting should be considered. However, tumor resection may be the only treatment needed to cure the hydrocephalus. For example, in patients with noncommunicating hydrocephalus secondary to mass effect from fourth ventricular tumors, we prefer surgical resection of the tumor to correct hydrocephalus. An intraventricular catheter can be placed for intracranial pressure monitoring and temporary CSF drainage as needed prior to the tumor resection. Others advocate performing an ETV prior to tumor resection.^29,30 Although this option bypasses the site of CSF obstruction, these patients need to be carefully selected due to the risk of upward herniation of the superior cerebellar vermis which can result in midbrain compression.²⁰

Congenital abnormalities can be associated with hydrocephalus. Approximately 80% to 90% of patients with myelomeningocele have hydrocephalus requiring intervention.^31,32 The localization of CSF obstruction in these patients is often more complex than in patients with obstruction secondary to a tumor. For example, Chiari II patients can have up to four pathologic sites of obstruction.³³ If there are multiple sites of obstruction, ETV may not effectively bypass the entire site and hence be unsuccessful. However, not all Chiari II patients have the same patterns of obstruction. For example, ETV has been shown to be successful in a small cohort of adult Chiari II patients, implying a more simplified pattern of obstruction between the third ventricle and the interpeduncular cistern.³⁴ Thus, the role of ETV in cases of congenital malformations associated with hydrocephalus is not clear and consideration of intervention must be done on a case-by-case basis.

Vascular pathologies such as subarachnoid hemorrhage (SAH) and intraventricular hemorrhage (IVH) can produce hydrocephalus. The mechanism of hydrocephalus in the acute setting is usually decreased CSF absorption from obstruction in the arachnoid granulations and a mechanical outflow obstruction from clot within the ventricular system. However, over time as the blood resolves, obstruction at the level of arachnoid granulations and within the ventricles may not be the long-term cause of hydrocephalus. Once the inflammatory process progresses, the point of obstruction usually becomes localized to the basal cisterns, at the interface of the spinal subarachnoid/cranial subarachnoid spaces.²⁰ In the setting of SAH, external ventricular drainage is commonly performed to temporize the patient. In some SAH patients, there is minimal residual obstruction and the external ventricular drainage can be weaned off without further intervention. Other patients with SAH may require permanent CSF diversion. In patients without prior surgical intervention, ETV can be considered if there is no anatomic contraindication. In patients with a prior surgical intervention, ETV may be of more limited use.³⁵ For example, fenestration of the lamina terminalis during open surgery for a ruptured aneurysm is direct, safe, and potentially avoids the need for a repeat operation to treat hydrocephalus.³⁶ Patients with IVH can similarly show a clinically significant, time-dependent change in the localization of CSF obstruction. In the series by Siomin et al., a higher rate of ETV success was associated with previous history of ventriculoperitoneal shunt placement post-hemorrhage.³⁷

Post-infectious hydrocephalus can also require CSF diversion. Obstruction resulting from infection can be localized to the basal cisterns or to the arachnoid granulations.²⁰ If the obstruction is in the basal cisterns then ETV can be considered. If the obstruction is in the arachnoid granulations, then ETV would not be beneficial. However, determination of the exact site of obstruction in these patients can be challenging and is not always possible. Another consideration in post-infectious patients is that there can be significant meningeal thickening of the leptomeninges and third ventricular floor. Thus, creation of a hole in the floor of the third ventricle can be more challenging and have a higher risk of closure.³⁷ History of infection has been shown to be a negative predictor of ETV success.^22,38 However, in properly selected patients ETV can be successful in treating post-infectious hydrocephalus.^20,37

The pathophysiology of idiopathic normal-pressure hydrocephalus (iNPH) has been the subject of much debate. Although some patients exhibit asymmetry between the third and fourth ventricles, suggestive of an associated late-onset idiopathic aqueductal stenosis, others exhibit global enlargement of the ventricular system. Indeed, the pathophysiology of iNPH is likely unrelated to a mechanical obstruction, but rather an abnormality of ventricular compliance and CSF flow.^37,39 It is known that certain patients with iNPH respond favorably to ETV. Gangemi and colleagues reported the largest series to date of iNPH patients treated with ETV. Out of 110 patients, improvement was seen in 69.1% of patients, no change in 21.8%, and continued deterioration was seen in 9.1%.³⁹ This compares favorably to other reported outcomes from shunting, although no randomized comparison has yet been performed.^39,40 ETV may be a more physiologic solution and the problems associated with overdrainage have not been seen, which make it a promising alternative to shunt diversion in select iNPH patients.³⁹

Other Considerations

There are factors other than pathology that have been shown to impact success of ETV for hydrocephalus. Age is one of the most important factors in predicting success of ETV in the pediatric population. Kulkarni and co-workers performed a multicenter, retrospective analysis on 618 pediatric patients who had undergone ETV.²² A training set of 455 patients was used to identify predictors of ETV success and the remaining 163 patients were used to internally validate those predictors. In their analysis, age was the most significant predictor of ETV success. Younger patients (<1 year old) had a higher chance of failure compared to older patients, and infants (<1 month old) in particular had the highest likelihood of failure.²² This confirmed previous reports and added weight to those that discourage ETV in infant patients.^41,42 In adult patients, there does not appear to be an association between age and risk of failure.³⁹