With continued advancements in endoscopic approaches for skull base surgery, understanding the risks and potential complications of these techniques is critical in planning surgical access and counseling patients. Extensive transclival operations, coronal plan approaches lateral to the carotid artery, and endoscopic orbital approaches have allowed more lesions to be accessed through less-invasive approaches. But each of these techniques is associated with an evolving risk profile. We discuss general considerations in minor and major complications for endoscopic skull base surgery, with a subsequent focus on orbital complications and endoscopic orbital surgery.
Minor Complications in Skull Base Surgery
Endonasal surgery uses the sinonasal corridors to access the skull base. Access through the nasal cavity allows for a minimally invasive approach, but it also comes with an associated cost to the normal function of the sinonasal cavity.
Postoperative sinusitis and synechia formation are perhaps the most common minor complications from endonasal skull base surgery. After a comprehensive disruption of the normal sinonasal anatomy, some degree of postoperative crusting develops in many patients. This crusting is frequently debrided in the clinic to prevent sinusitis and synechia formation. A literature review of skull base sinonasal outcomes demonstrated a 50% incidence of significant postoperative crusting, with 40% of patients demonstrating sinusitis symptoms of nasal drainage and obstructive symptoms. Although these symptoms can be relatively benign, they can significantly affect quality of life. Synechia formation after endonasal sinus surgery has been reported in 5% to 28% of patients, and results from skull base surgery would presumably be similar. Notably, delayed mucocele formation can occur when a sinus becomes obstructed from scarring of the outflow tract postoperatively and has been reported in 3% to 8% of cases.
Postoperative epistaxis is a relatively common consideration after endoscopic sinonasal surgery. And although most postoperative epistaxis is mild, severe hemorrhage requiring operative control is well defined and typically stems from an arterial source. Classically nasal epistaxis can be managed with nasal packing; however, in the fresh postoperative setting, particularly after a skull base resection, aggressive packing must be approached cautiously to avoid intracranial complications from improperly placed packs. Furthermore, it is worth noting that the majority of reported epistaxis events occur 2 to 4 weeks postoperatively. In a recent review, Zimmer and Andaluz reviewed more than 400 endoscopic pituitary surgeries, demonstrating a 4.1% rate of postoperative epistaxis. They noted that of the 18 patients, 11 were treated with in-office cauterization, packing, or intranasal hemostatic agents, whereas 5 required a return to the operating room and 2 required embolization. Similarly, Thompson et al. reported a 3% incidence of postoperative epistaxis in their single-institution cohort. Although the majority of episodes of epistaxis were controlled with packing, 5 of 14 events required control in the operating room. These data confirm that, overall, postoperative epistaxis is relatively uncommon after endoscopic skull base surgery and frequently can be managed with conservative measures. However, some patients require operative control, particularly in cases of arterial hemorrhage.
Nasal deformities such as saddle nose have been reported after skull base surgery. This is particularly identified after nasoseptal flap and subsequent septectomy. In one major report on these nasal deformities, the authors highlight a nearly 6% overall incidence of nasal dorsal collapse. The authors noted these deformities were associated with nasoseptal flap use (15% of patients who underwent nasoseptal flap) and highlight several potential explanations, including electrocautery, contracture scar forces, overaggressive septectomy, and postoperative radiation as potential implicating factors. Soudry et al. performed a retrospective review demonstrating a less than 1% rate of saddle deformity. Although these nasal deformities are not life threatening, they are challenging to repair and can have significant impacts on the patient’s social and functional status. We speculate that preservation of the entire septal attachments to the anterior premaxilla may help prevent this complication.
Using the sinonasal corridor for access to the skull base has several advantages, but one notable disadvantage is the potential disruption to the olfactory system. Postoperative hyposmia has been well documented and evaluated by several studies. Several technical concepts have been suggested to potentially improve olfaction outcomes, including the preservation of the septal olfactory strip and preservation of the middle turbinates when possible. Results from a variety of studies demonstrate a wide variety of results, ranging from no significant dysfunction, to temporary impairment, to significant permanent olfactory disturbance. Some studies have reported rates of long-term olfactory disturbance up to nearly 30%. A prospective study of 42 patients who underwent baseline and periodic postoperative testing (University of Pennsylvania Smell Identification Test) demonstrated that patients undergoing pituitary surgery with rescue flap elevation showed no evidence of olfactory dysfunction, whereas patients with a nasoseptal flap showed temporary dysfunction. A recent evidence-based review and recommendation on olfactory function after endonasal skull base surgery was published by Greig et al. They concluded that the body of evidence was heterogeneous, but routine transsphenoidal surgery with rescue flaps and at least one middle turbinate preserved likely leads to limited long-term olfactory dysfunction. However, they also concluded that nasoseptal flap harvest and potentially electrocautery likely lead to increased olfactory dysfunction.
Major Complications in Skull Base Surgery
The most common major complication after endoscopic skull base surgery is postoperative cerebrospinal fluid (CSF) leak. Breaching the dural layer of the skull base (and underlying arachnoid) typically results in a visible CSF Leak. Definitive reconstruction after surgical extirpation is critical to separate the intracranial contents from the sinonasal space and prevent infectious complications. Although various reconstructive approaches have been proposed, the nasoseptal flap has emerged as the workhorse, vascularized reconstructive tool for multilayered reconstruction ( Fig. 39.1 ). Consensus retrospective data have generally agreed that, intraoperatively, small, low-flow CSF leaks can be repaired with layered free graft approaches, whereas large, high-flow leaks should be repaired with a vascularized flap. Postoperatively the patient must be observed for CSF rhinorrhea. The incidence of postoperative CSF leak after endonasal skull base surgery ranges significantly based on the surgical subsite. In general, data suggest sellar defects have the lowest incidence, followed by cribriform, suprasellar, and then clival defects, which are generally regarded as the most difficult to repair.
Identification of a postoperative CSF leak is typically signified by clear rhinorrhea with challenge (leaning forward) or increasing pneumocephalus on computed tomography scanning. When this is identified, several strategies exist for management. For very low-flow, small persistent postoperative CSF leaks, bedrest and pressure reduction with acetazolamide (Diamox; Zydus Pharmaceuticals) can be considered. However, CSF diversion, typically with a lumbar drain, is often added to a conservative regimen for several days to allow the leak to scar and heal. Nevertheless, the majority of postoperative CSF leaks require surgical re-exploration with revision reconstruction. Often this requires minimal adjusting of the existing reconstruction, but the team needs to be prepared for a complete revision. Unfortunately, there is not an abundance of consensus data on when to choose a conservative versus operative approach, but one systematic review highlighted that the majority of cases (62%) required operative revision.
The critical importance of successful skull base reconstruction after endonasal approaches cannot be overemphasized. Postoperative meningitis is highly correlated with reconstructive failure and persistent postoperative CSF leak. Lai et al. showed the risk of meningitis is directly related to postoperative CSF leaks with an odds ratio of 92. In the absence of a CSF leak, the risk of meningitis and intracranial infectious complications approached zero. Persistent CSF leaks have been associated with up to a 21% incidence of meningitis and increased rates of reoperations and major complications.
Postoperative meningitis or other infections sequelae are potentially devastating complications of endonasal skull base surgery. Fortunately, rates of postoperative meningitis and other intracranial infections are low, ranging from 0 to 10% depending on the study evaluated. However, the complications of meningitis can be devastating, with studies highlighting up to 13% associated mortality.
As the field of endoscopic skull base surgery has evolved and new surgical approaches and techniques have developed, the major limit in the extent of dissection remains the cranial nerves. Postoperative cranial neuropathy is typically associated with significant morbidity. An exquisite knowledge of the anatomic structures, high-resolution preoperative cross-sectional imaging, intraoperative stereotactic navigation, and neurophysiologic monitoring are critical components of safe endonasal surgery and limiting risk to the cranial nerves. Fortunately, cranial nerve injuries are rare and highly correlate with the anatomic location of the target lesion and the aggressiveness of the lesion. For example, in one study of cavernous sinus tumors, new postoperative cranial neuropathies developed in nearly 12% of patients with nonpituitary adenoma pathologies, while none of patients with the adenomas had this complication. Another study highlighted an 8.7% incidence of postoperative cranial neuropathy after resection of clival chordoma. Although data are limited, experience suggests malignant pathologies that require more aggressive resection, and tumors invading the cavernous sinus along the course of VI cranial nerve VI, or those invading the optic canal appear to have increased risk profiles. There are limited options for treatment of postoperative cranial nerve deficits. Many palsies are transient and may respond to steroids. There are a variety of surgical and corrective prism options for ophthalmoplegia, and consultation with a neurophthalmologist is often beneficial. For trigeminal dysfunction, symptomatic patients may require consultation with a neurologist to discuss medical management including anticonvulsants. For lower cranial neuropathies, involvement of the speech and language pathologist can help rehabilitate a functional swallow and speech.
Although cranial neuropathies result in significant morbidity for our patients, the most feared complication in endonasal surgery is a vascular injury of the carotid or basilar system. As with the cranial nerves, exquisite knowledge of the surgical anatomy and preoperative imaging are critical to prevent these catastrophic events. Vascular injuries remain uncommon in experienced surgical hands, and a recent literature review highlighted a 0.34% rate of arterial injury. In the event of an arterial injury ( Fig. 39.2 A ), the surgical team must act fast. Controlling the bleeding with pressure and packing intraoperatively are critical to prevent exsanguination. A crushed muscle patch has been proven to expedite hemostasis ( Fig. 39.2 B). These injuries are high-stress situations for which there are several proposed training models, including live, synthetic, and cadaveric, that allow the participant to practice the surgical and psychomotor skills needed to control these catastrophic complications. Once the bleeding is controlled intraoperatively, the patient is typically taken straight to the angio-interventional suite where, if feasible, carotid stenting can be performed. However, frequently carotid sacrifice must be performed for definitive control ( Fig. 39-2 C, D).