History

Minimally invasive cranial neurosurgery represents a history of challenges and periodic successes. For decades, the primary concern of neurosurgery has been to minimize the neurovascular impact of surgery. Adequate exposures, which have usually meant large exposures, were seen as the key to good outcomes. Skull base microsurgery developed around the belief that better outcomes are achieved by moving bone and soft tissue rather than brain and nerves. The minimally invasive neurosurgery paradigm evolved out of the recognition by many surgeons that, in many cases, they did not use or need much of the exposure provided by extensive approaches, and they began to seek alternatives to the soft tissue trauma, healing, and recovery time involved with extensive approaches. To this end, a few pioneers began to explore smaller approaches to achieve the same neurosurgical goals.

One such approach, the supraorbital transbrow craniotomy, was an adaptation of the orbitozygomatic approach first proposed by Jane and colleagues. Popularized by Reisch and Perneczky and by Jho, the technique has become widely practiced by some schools of neurosurgery but has not been accepted universally. The minimally invasive keyhole principles promulgated by these practitioners have gradually been applied to other areas of cranial neurosurgery, where they excite ongoing debate.

Neuroendoscopy has been the foundation of many minimally invasive approaches, although it too has had an on-again, off-again history. In neurosurgical terms, most of the applications of neuroendoscopy proposed to date have been focused on addressing intraventricular pathologies. The first description of a neuroendoscopy procedure was by L’Espinasse in 1910. As described by Walker, L’Espinasse cauterized the choroid plexus in 2 hydrocephalic infants with the assistance of a cystoscope. Dandy described his use of the endoscope to remove the choroid plexus, with results that were similar to those of his experience with formal craniotomy for open choroid plexectomy. The growth of neuroendoscopy was slowed by the significant limitations associated with the lighting and magnification available for early endoscopes. Furthermore, the advent of ventricular shunts and advances in microneurosurgery reduced interest in neuroendoscopy.

In the 1960s and 1970s, technological advances in lens development and the application of fiberoptics improved neuroendoscopic visualization significantly. In 1963, Guiot and colleagues used ventriculoscopy to explore a patient with a colloid cyst. In 1973, Fukushima and colleagues described the first ventricular biopsy. In 1983, Powell and colleagues reported the first endoscopic resection of a colloid cyst, and the technique was popularized throughout the late 1990s. The use of endoscopy to address selected intraventricular tumors for biopsy or resection is now a widely accepted therapeutic option.

Subsequently, the application of minimally invasive principles and endoscopy to extra-axial tumors was popularized by Perneczky and Fries. They promoted the endoscope as a means to minimize retraction on the brain and to avoid resection of the dura and bone, which they argued increased operative time and operation-related trauma. In 1998, Fries and Perneczky reported 380 endoscope-assisted cases, 242 of which involved either the subarachnoid space or the cerebral parenchyma. The investigators found that endoscopy improved visualization of perforators during aneurysm surgery and permitted exploration of the ventral brainstem, ventral side of cranial nerves, and ventral aspect of the cervical spine while minimizing retraction. They reported no complications associated with the use of the endoscope itself.

Endoscopic transsphenoidal surgery has been the largest area of recent growth in neuroendoscopy. The field was pioneered by Guiot and colleagues, although Guiot later abandoned it because visualization was poor. In the 1970s the use of neuroendoscopy was again reported as an adjuvant to transsphenoidal microneurosurgery, but it was not until 1992 that Jankowski and colleagues reported a purely endoscopic transsphenoidal approach to the sella turcica. Jho and Carrau, who are considered early pioneers of the field, reported 46 purely endoscopic endonasal procedures in 1997. This area continues to expand rapidly, with an increasing number of intradural lesions being approached through an endonasal route. This topic is explored more fully elsewhere in this issue.

The most recent intracranial application of neuroendoscopy is endoscopic resection of intraparenchymal brain tumors. In 2008, Greenfield and colleagues described the use of METRx tubular retractors (Medtronic Inc, Memphis, TN, USA) in combination with frameless stereotactic navigation for complication-free removal of 10 deep lesions. The next year, Kassam and colleagues reported the use of a nonfixed transparent conduit to remove 21 subcortical lesions with no new neurologic complications.

The forays from the ventricles into the subarachnoid and intraparenchymal spaces hardly constitute a mature field. However, these forays are an intriguing direction for minimally invasive neurosurgery. The remainder of this article considers the current state of and future applications for minimally invasive and neuroendoscopic neurosurgery beyond the intraventricular and endonasal routes.

Indications

The general indications for a minimally invasive approach are the same as those for any other neurosurgical approach to a given tumor. Often the decision to use a keyhole approach depends more on the specific pathologic condition and on the practitioner’s experience than on any other factor. Although use of the endoscope depends entirely on the specifics of a case, the addition of the endoscope is often complementary in many keyhole approaches.

The appropriate trajectory is the single most important factor in the success of a minimally invasive tumor approach. Various factors must be considered: (1) the planned extent of resection for the tumor being addressed (complete removal vs debulking vs biopsy); (2) the nature of the tumor involved, particularly the distinction between the tumor and normal brain tissue; (3) the vascularity of the tumor and whether vascular control can be achieved from the chosen trajectory; (4) the surface structures that will be penetrated and whether cosmesis or intervening structures (eg, venous sinuses or bony sinuses) may force deviation from an otherwise ideal trajectory; and (5) the depth of the tumor, which at times can prevent the lesion from being accessed via a small craniotomy. The 2-point method is a simple technique for estimating the best trajectory. A point is placed at the geographic center of the lesion, and a second point is placed where the lesion comes closest to the surface. A line drawn through these 2 points to the surface of the head constitutes the best approach trajectory, subject to modification by the previously mentioned considerations.

The absolute minimum size of the craniotomy is constrained by the instruments that will be used, but it typically is no smaller than the size of an open bipolar forceps.

Indications

The general indications for a minimally invasive approach are the same as those for any other neurosurgical approach to a given tumor. Often the decision to use a keyhole approach depends more on the specific pathologic condition and on the practitioner’s experience than on any other factor. Although use of the endoscope depends entirely on the specifics of a case, the addition of the endoscope is often complementary in many keyhole approaches.

The appropriate trajectory is the single most important factor in the success of a minimally invasive tumor approach. Various factors must be considered: (1) the planned extent of resection for the tumor being addressed (complete removal vs debulking vs biopsy); (2) the nature of the tumor involved, particularly the distinction between the tumor and normal brain tissue; (3) the vascularity of the tumor and whether vascular control can be achieved from the chosen trajectory; (4) the surface structures that will be penetrated and whether cosmesis or intervening structures (eg, venous sinuses or bony sinuses) may force deviation from an otherwise ideal trajectory; and (5) the depth of the tumor, which at times can prevent the lesion from being accessed via a small craniotomy. The 2-point method is a simple technique for estimating the best trajectory. A point is placed at the geographic center of the lesion, and a second point is placed where the lesion comes closest to the surface. A line drawn through these 2 points to the surface of the head constitutes the best approach trajectory, subject to modification by the previously mentioned considerations.

The absolute minimum size of the craniotomy is constrained by the instruments that will be used, but it typically is no smaller than the size of an open bipolar forceps.

Neuroendoscopy: approach or adjunct?

The decision to apply neuroendoscopy to a given tumor surgery requires a major distinction to be made, that is, whether the operation is endoscopically controlled, whereby the endoscope is the sole or primary means of visualization, or whether it is endoscopically assisted. In the latter, microsurgical techniques are the mainstay of dissection and tumor resection, while endoscopy is used to help visualize areas of the tumor that are otherwise difficult to see with the uniaxial view provided by the operative microscope. Endoscopy is also used to examine the tumor bed for completeness of resection. With image injection, endoscopic images can be inserted into the microscope eyepiece, allowing both to be used together. Many surgeons, however, find this technique to be distracting.

Endoscopically controlled techniques have become the approach of choice for many neurosurgeons for transsphenoidal surgery. In most of these applications, the microscope is never used during the operation. A completely endoscopic surgical approach is less commonly used during transcranial neurosurgery. The nose is lined with mucosa, and the risk of damaging structures as one passes instruments is low. Two surgeons can work in concert, introducing and removing instruments with ease. In the cranium, the corridor to the tumor is often lined with brain and neurovascular structures, which poorly tolerate even slight manipulation by passing endoscopes or instruments. Furthermore, the microscope provides an excellent view in a direct line and frees both hands to participate in the dissection. In contrast, in these applications the endoscope often occupies the space of 1 instrument and must be held by the primary surgeon. Therefore, the endoscope tends to be used in a supporting role after access is obtained through microneurosurgery.

However, with a holder or a good assistant, 2-handed work is possible through the endoscope. The instruments are introduced along the endoscope and kept just in front of its tip so that they are always in view. This technique differs from the purely endoscopic approach, in which the instruments travel through the shaft of the endoscope, as is common in intraventricular endoscopic surgery. In the endoscopically assisted technique, the endoscope can be used in multiple stages to inspect the approach, better define the anatomy, see parts of the tumor not in the direct line of vision, and inspect the tumor bed for residual tumor. Particularly, when combined with minimally invasive approaches, smaller craniotomies, and an effort to preserve overlying neural structures, the advantage of the endoscope for working in small spaces and for seeing angles perpendicular to the line of sight can be invaluable. In some instances, both intracranial extra-axial and intra-axial tumors have been removed without the use of the microscope in an endoscopically controlled fashion, but doing so remains relatively unusual.

During tumor resection, the endoscope may be valuable for removing a small amount of residual tumor in a difficult location. For example, Chang and colleagues reported the use of the endoscope to remove the last fragments of a large ecchordosis physaliphora from the clivus that could not be seen with the microscope.

Equipment

For endoscopically assisted or endoscopically controlled tumor surgery, handheld endoscopes are used and usually held in the nondominant hand or in a rigid holder. Various endoscopes are available. In the authors’ opinion, the most suitable endoscope for endoscopically assisted intracranial surgery is the Perneczky endoscope. The right-angle pistol-grip configuration of this endoscope allows it to be held comfortably in the hand for long periods. The rigid shaft of the endoscope allows precise control of its position.

The endoscope is usually held in one hand while the other hand is used for working. An assistant can also hold the endoscope, freeing the surgeon to work with both hands. If needed, the endoscope can be fixated with the use of one of many available retractor arms (see another article elsewhere in this issue). Many other types of endoscopes can be used for endoscopically assisted or endoscopically controlled work. The main issue is ergonomic: it is helpful to have an endoscope that stays out of the way so that other instruments can be used simultaneously. At minimum, an appropriately sized suction device is a necessary second instrument. If possible, 2 additional instruments offer greater surgical possibilities.

New technologies continue to improve the image produced by endoscopes. For example, with the appropriate endoscope, viewing apparatus, and image processing, a 3-dimensional image can be produced. The surgeon can then view the image with the aid of polarized glasses. As an alternative to standard-definition video, new high-definition images are available and more closely approximate the view that most neurosurgeons are familiar with from the operating microscope.

In addition to endoscopes, specialized instruments are helpful for minimally invasive cranial work, regardless of whether the work is performed with an endoscope or microscope. These instruments include narrow-shafted bipolar instruments; various slim dissectors; and single-shafted bayoneted scissors, pituitary forceps, and graspers. Angled dissectors, bipolar forceps, and a suction device that can operate around corners are highly advantageous. Without these instruments, the endoscope may reveal areas of residual tumor that remain out of reach due to the working angles available to straight-shafted instruments.

Operating technique

The principal issue for operating with the endoscope is maintaining the appropriate orientation. The novice endoscopist may find the view from the endoscope disorienting compared with the view provided by the microscope. This problem improves with experience. Familiarizing oneself with the endoscopic view during laboratory or practical courses is invaluable. Looking at known objects through the endoscope also can help one to adapt more rapidly to the endoscopic perspective.

In general, the endoscope is held in the nondominant hand and the working instrument or suction is held in the dominant hand. A 30° endoscope is most often used. This endoscope provides enough of an angle to be useful for looking around corners while still allowing the user to see straight ahead. Each time instruments are introduced into the field, the endoscope is removed so that the instruments can be followed into the head. This strategy prevents conflicts between the instruments and neurovascular structures.

The angled endoscope is readjusted alternately to inspect each direction where additional visualization is needed. Each time such an adjustment is made, the endoscope is withdrawn from the head, the angle is changed, and the endoscope is reinserted to prevent the tip from wandering into structures as the endoscope is rotated.

Tumor removal proceeds in a standard fashion, either with the use of the endoscope or with microsurgical instruments, such as scissors, a bipolar device, and dissectors. The endoscope is especially used at the end of surgery to check the completeness of resection and repair.