Advances in endoscopic surgical techniques and instrumentation have led to an expansion in the size and diversity of skull base lesions that are amenable to endoscopic resection. However, one of the stipulations for adopting an endoscopic approach for the removal of skull base lesions is the ability to repair the resultant defect, as failed reconstructions, and the subsequent cerebrospinal fluid (CSF) leak, add significant morbidity. With this in mind, the ideal endoscopic skull base repair is technically feasible as part of the endoscopic procedure and provides a reliable and robust separation between the nasal and cranial cavities that will last over the long term. In addition, the repair should reconstruct the natural tissue barriers of the skull base, minimally affect normal sinonasal and cranial physiology, and possibly obliterate the dead space after tumor removal. The surgeon must also consider the anticipated location, size, and geometry of the bony and dural defects, as well as the anticipated volume of CSF leak. Finally, the surgeon must also consider previous sinonasal surgery, previous or planned postoperative radiotherapy, and the extent of tumor involvement of nasal structures, such as the septum and turbinates, all of which may limit the available options for reconstruction.
Endoscopic skull base reconstruction has shown excellent success rates with low perioperative and postoperative morbidity even when large defects are present. In some cases of skull base surgery, when no or a low-flow CSF leak is encountered—for example, a simple reconstruction—may be all that is required using free grafts or even alloplastic materials. A large variety of local and regional vascularized pedicle flaps can be used to reconstruct more complex defects of the skull base. Both free grafts and vascularized flap repairs usually use a multilayered closure to establish a reliable barrier between the cranial and nasal cavities. The most common points of failure of flap repairs are the dependent parts, presumably owing to increased pressure, or the most superior parts, likely owing to flap migration or retraction.
Given the multitude of options available for reconstruction, an effort has been made to define the indications and utility for each type of repair. Likely the most important factor guiding the selection of a repair method is classifying the volume and flow of CSF leak (if any) as well as the size and complexity of the anticipated defect. The CSF flow rate can be classified as no leak (no intracranial opening, or appreciable leak into the nose), a low-flow (intracranial opening but minimal flow observed or no direct communication with a cistern or ventricle), and a high-flow leak (intracranial opening, direct communication into a ventricle or cistern). Preoperative imaging can be used to predict the type of CSF leak likely to be encountered after resection of a given lesion.
When there is no CSF leak, the repair is at the surgeon’s discretion and can range from simple onlay or epidural/subdural underlay placement of a synthetic graft or repair with packing and/or dural sealant at the level of the sella. Skull base defects that are small with a low-flow CSF leak can be reconstructed with a wide variety of multilayered (or even monolayered) avascular free grafts or biosynthetic materials with high success rates and limited morbidity. Larger and more complex skull base defects (> 2–3 cm) and those that are associated with high-flow CSF leaks are best repaired with a multilayered technique using a vascularized flap of tissue ( Table 37.1 ).
|No Leak||Low-Flow Leak||High-Flow Leak|
|Synthetic dural replacement graft||Synthetic dural replacement graft||Synthetic dural replacement graft|
|Autograft (fat or mucosa)||Autograft (fascia lata or mucosa)||Autograft (fascia lata or fat)|
|No repair||Intranasal vascularized flapsExtranasal vascularized flapsFree tissue transfer|
In addition to leak type, other independent factors must be considered to help guide the reconstructive decision-making process. The extent of the skull base defect should be assessed because resections involving extended approaches often result in large and more complex defects with high-flow CSF leaks that are best managed with vascularized flaps. Specific disorders also carry an increased potential for postoperative CSF leak, and as such, the use of a vascularized pedicled flap should be strongly considered in these unique instances. Among these are meningiomas (extensive bony and dural resection with intracranial disruption of the arachnoid plane), craniopharyngiomas (often requiring expanded approaches and involving arachnoid dissection), Cushing disease (reduced healing from hypercortisolemia), and morbid obesity (possible increased intracranial pressure, also potentially present with Cushing disease). Furthermore, in patients who have had or will potentially need radiation therapy, vascularized flaps should be strongly considered, as they are more likely to withstand the effects of radiation therapy in providing a durable repair. Lastly, independent of CSF leak and radiation status, vascularized reconstructions may provide adequate coverage of exposed neurovital structures (e.g., internal carotid artery) that may be uncovered and mobilized during tumor resection.
The use of a lumbar drain as part of the reconstructive strategy deserves consideration as preoperative planning takes place. The placement of a perioperative lumbar drain has not been shown to positively affect postoperative leak rates when vascularized flaps are used. However, the use of a lumbar drain in the setting of a postoperative CSF leak has been shown to be effective as first-line therapy. In our experience, lumbar drains can provide advantages in some high-risk clinical situations. In addition to potentially enhancing the success rate in some complex, high-flow CSF leak cases, the use of drains also provides the opportunity to (1) measure opening pressures before and after repair and (2) permit the intrathecal injection of fluorescein, which may be of value in some cases. The use of lumbar drains is associated with certain attendant risks that must be balanced in any given case.
The Reconstructive Ladder
A thoughtful and systematic way to organize available options is to use the idea of a “reconstructive ladder” when considering skull base reconstruction.
When no CSF leak is encountered, no complicated reconstruction is required. It is worth mentioning that provocative testing by having the anesthesiologist raise intrathoracic pressure (simulating a Valsalva maneuver) is worthwhile to perform routinely at the end of every procedure in which a CSF leak is not obvious to ensure that a small or otherwise clinically occult leak is not present. The “no CSF leak” scenario is often encountered during routine transsellar approaches for pituitary lesions. A monolayer reconstruction aimed to simply cover the exposed diaphragm sella can be used with great success. In fact, very little is required by way of reconstruction in these cases, and some may prefer to only place a small amount of absorbable hemostatic agent into the sella at the conclusion of the procedure.
Some otherwise complex skull base procedures, including approaches to the craniovertebral junction, often do not require reconstruction of the resultant defect. As this is an entirely extradural procedure during which a CSF leak should not be encountered, an involved reconstruction is not required. When no obvious intraoperative CSF leak is encountered, hemostasis is obtained and the surgical field is thoroughly irrigated. Afterward, Valsalva maneuvers are undertaken to ensure a low-flow CSF leak has not been missed. A small amount of absorbable hemostatic agent may then be placed into the surgical defect, and the lateral aspects of the nasopharyngeal muscles and soft tissues may be approximated to the midline.
Synthetic Dural Replacement Grafts
Fundamental to the endoscopic approach to intracranial lesions is the need to perform intradural dissection. The dura is typically reconstructed by some surgeons even in resections that do not result in a CSF leak (i.e., during a sellar approach without intraoperative CSF leak). When a low-flow or high-flow CSF leak is encountered, a multilayered reconstruction is used. In either case, a synthetic dural replacement graft is often used to reapproximate the dural defect. A variety of grafts are available depending on surgeon preference; however, grafts that can be sutured offer a sturdier repair substrate and are more pliant, making them easier to place and secure. A major advantage of the use of such synthetic materials is that they are readily available and do not require additional donor site morbidity for the patient.
Autograft choices typically include free mucosa, fat, and fascia lata. These tissues were some of the first reconstructive materials described for skull base reconstruction and are still excellent options. Fascia lata grafts are harvested from an incision on the lateral thigh and offer a durable inlay or onlay material. The major drawbacks to the use of fascia lata are the presence of a permanent scar on the leg and wound-related issues, especially in young physically active patients.
Autologous fat grafts also provide a suitable inlay graft that can serve to occupy dead space and also to “cork” the defect opening. These grafts are typically harvested from the abdomen but may also be harvested from the thigh, particularly when the surgeon has already decided to harvest a fascia lata graft. Abdominal fat harvest is usually performed through a periumbilical incision to permit a less obvious scar as well as to avoid confusion with an appendectomy scar. Autologous fat grafts do not necessarily provide a watertight seal by themselves but are useful for filling large cavities left behind by resection or removal of a tumor.
Free grafts may also be harvested locally from multiple sites within the nasal cavity. Free mucosal grafts may be harvested from the septum, inferior turbinate, middle turbinate, or the nasal floor. If the middle turbinate is removed during the initial approach to the skull base, use of this mucosa for a free graft may negate further donor site morbidity. The nasal floor graft is an attractive option owing to both its ease of harvest and very low donor site morbidity ( Fig. 37.1 ). Regardless of the donor site of the free mucosal graft, it is then applied with the mucosal side toward the nasal cavity to prevent development of a mucocele and is secured in place with a tissue sealant/glue of the surgeon’s choice.
Local Pedicled Flaps
The principal workhorse of contemporary endoscopic skull base repair techniques is the Hadad-Bassagasteguy flap, or the pedicled nasoseptal flap (NSF) ( Fig. 37.2 ). First described in 2006, it has proven to significantly reduce postoperative CSF leak rates. This flap has consistent vascularity (posterior septal branch of the sphenopalatine artery), a long and robust pedicle, is easy to harvest, and offers customizability/adaptability. The flap is made by making three incisions in the nasal septal mucosa using needle-tip monopolar cautery. The first cut starts superiorly just inferior to the level of the sphenoid os and extends along the septum anteriorly, keeping 1 to 2 cm below the cribriform plate to preserve olfactory neuroepithelium. Next, an inferior cut starts from the superior margin of the choana, then extends across to the posterior margin of the vomer, and proceeds along the junction of the septum and the nasal floor over the maxillary crest. This inferior incision can be extended laterally to include the nasal floor and even the lateral nasal wall for coverage of wider defects. When this inferior incision is carried laterally, care must be taken to not incise over the soft palate. The superior limb of the NSF can be extended anteriorly as far as the junction between the septal mucosa and the vestibular skin. The two incisions are joined anteriorly by a vertical incision. Once these three incisions are completed, the flap is carefully elevated from the underlying cartilage and bone with care to preserve the posterior vascular pedicle. When elevating the flap off the face of the sphenoid sinus posteriorly, the surgeon must take care to not shear or tear the flap at this point, as this would likely injure the vascular pedicle. Once sufficiently elevated, the flap is then pushed into the nasopharynx or into a large maxillary antrostomy for more inferior approaches to avoid inadvertent damage during the remainder of the procedure.