Keywords
skull base - complications - skull-base surgery - mortality
Introduction
Surgery of the skull-base forms is a challenging segment in otorhinolaryngology practice. Treatment is undertaken for complex and varied pathologies. Advances in diagnostic radiology, interventional radiology, neuronavigation, nerve monitoring, surgical instrumentation, and postsurgical care have transformed both the scope and the safety of modern skull-base surgical practice. Nevertheless, complications are a frequent reality.
The exact definitions of what constitutes skull-base surgery and what should and should not be included under this label remain contentious. The gamut of complications can vary widely depending on the case mix in a particular practice. Previous reports have documented the individual complications of endoscopic, anterior external, lateral, and pediatric skull-base surgery.[1]
[2]
[3]
[4]
[5]
[6]
[7]
This report attempts to standardize and compare surgical complexity and surgical complications across the entire surgical spectrum of skull base surgery (endoscopic anterior, open anterior, and open lateral) as encountered at a contemporary otolaryngology led skull-base surgery practice. Surgical procedures are classified as per approach and complexity[8]
[9] and complications classified as per a patient oriented, standard reporting format.[10]
Methods
A retrospective chart review of all patients with skull-base procedures was undertaken at the Department of Otolaryngology and Head–Neck Surgery under the care of the senior author from January 2002 to March 2015. Appropriate Institutional permissions were obtained. Follow-up data were collated from chart review, follow-up records, follow-up reviews, and telephonic follow-ups.
The following criteria were used to define the study population:
Inclusion criteria: All patients of who underwent surgery for lesions of the skull-base were included. These included external, endoscopic as well as combined approaches.
Exclusion criteria: Patients with less than 6 months follow-up were excluded from the study.
The charts and follow-up records of patients in the study population were evaluated for demographic data, diagnosis, the site of lesion, surgical approach, surgery undertaken, and complications encountered.
Categorization of Surgical Procedures
Surgical approaches were categorized as endoscopic endonasal, external anterior, and external lateral.
Surgical procedures were further graded as per the surgical complexity. Endoscopic endonasal procedures have been previously classified for degree of complexity by Snyderman et al.[9] As per this classification, level I procedures are the initial sinus access procedures which are precursors to subsequent skull-base procedures (e.g., endoscopic sphenoid and frontal sinus access and sphenopalatine artery ligation). The specific skull-base procedures are classified as level II–V in graded degrees of complexity progressing along level II (skull-base bone exposure, minor dural leak closure, intrasellar pituitary), level III (significant dural exposure, extradural intracranial procedures, extrasellar pituitary), level IV (intradural procedures, major carotid exposure), and level V (vascular malformations and highly vascular tumors, aneurysms). Drawing on this above classification, a similar grading was undertaken for external anterior and external lateral procedures and the same is indicated in [Table 1]. Highly vascular tumors were classified as level V by Snyderman et al but highly vascular tumors with appropriate embolization have been classified as level IV in this analysis as the majority of such tumors (juvenile angiofibroma and glomus jugulare tumors) are currently managed with preoperative embolization prior to surgery.
Table 1
Categorization of complexity of surgical procedure and severity of complication
|
Categorization of complexity of skull-base surgical procedure (applicable to all approaches, i.e., endonasal endoscopic, external anterior, and external lateral)
|
|
Level I
|
Precursor surgical procedures preceding skull-base access
|
|
Level II
|
Skull-base bone exposure, minor dural exposure/dural leak closure; intrasellar pituitary
|
|
Level III
|
significant dural exposure, extradural intracranial procedures, extrasellar pituitary
|
|
Level IV
|
Intradural procedures, major carotid exposure, highly vascular tumors (appropriate embolization)
|
|
Level V
|
Vascular malformations, aneurysms, highly vascular tumors (no or inadequate embolization)
|
|
Categorization of surgical complications
|
|
Minor complication
|
A complication which does not require treatment or delay discharge
|
|
Intermediate complication
|
A complication which delays discharge or which requires or prolongs treatment but does not cause permanent disability
|
|
Major complication
|
A life threatening complication or one which causes permanent disability or which compromises the result of surgery
|
|
Death
|
–
|
Surgical Techniques and Approaches
The surgical procedure was tailored depending on the site and extent of lesion. Presurgical embolization, neuronavigation, cranial nerve monitoring, and postsurgical ventilation were undertaken as per requirement. All surgeries were undertaken under the care of the otolaryngology service and led by the senior author (A.T.). Collaboration with neurosurgery was undertaken as and when appropriate and particularly for level IV procedures. Level V procedures are routinely undertaken under the neurosurgery service and do not form part of this particular dataset.
Categorization of Complications
The complications were classified as minor, intermediate, major, and death as in accordance with the codes prepared for the British Association of Otolaryngologists for reporting on surgical complications ([Table 1]).[10]
Expected sequelae, such as anosmia, following a traditional anterior cranio–facial resection and grade III facial weakness following a facial nerve transposition for a glomus jugulare excision were not rated as complications. All patients who underwent planned facial nerve transposition in the infratemporal approaches were counseled about postoperative facial paralysis. Recovery to grade III or better by 9 months was classified as an expected sequelae and not a complication. A grade IV or worse status at the 9-mont review was listed as a major complication.
New lower cranial nerve palsies inclusive of X nerve loss or section was corrected at surgery or in the early postsurgical period with a thyroplasty procedure.[11] In keeping with the criteria listed in [Table 2], well rehabilitated speech and swallowing postsurgery was classified as an intermediate complication (i.e., requires treatment but does not cause permanent disability).
Table 2
Listing of complications as per categorization skull-base surgical procedures
|
Surgical approach and complexity
|
Surgical complications
|
|
Anterior skull base
|
|
Endoscopic endonasal skull-base
(n = 120)
|
Level II
(n = 55)
|
• CSF rhinorrhea repair
|
Intermediate graft donor site infection (2), postsurgical seizures (1), valproate toxicity (1)
|
|
Level III
(n = 53)
|
• Meningocoele. enchephalocoele excision
• Pituitary (extrasellar) and clival tumors
• Optic nerve decompression
• Posttraumatic
• Skull-base fibrous dysplasia
• Nose and PNS tumors
• Mucocele with extensive dural exposure
• Allergic fungal rhinosinusitis -intracranial
|
Minor transient CSF leak (1), Intermediate CSF leak requiring repair (1), postsurgical seizures (1), secondary hemorrhage (1), Major extradural hemorrhage (1)
|
|
Level IV
(n = 12)
|
• Clival chordoma/olfactory groove meningioma/olfactory Neuroblastoma (intradural)
• Petrous cholesterol granuloma with major ICA exposure
• Allergic fungal rhinosinusitis with major ICA exposure
|
Minor secondary hemorrhage (1), Major meningitis (1), carotid blow-out and hemiparesis (1)
|
|
External anterior skull-base
(n = 84)
|
Level II
(n = 23)
|
• Osteoplastic flap
• Frontal sinus CSF rhinorrhea
• Maxillary swing
• Nose, PNS, Infratemporal fossa tumors
|
Minor transient CSF leak (1)
|
|
Level III
(n = 28)
|
• Ant craniofacial resection
– Nose and PNS tumors
– Orbital tumors, intracranial–extradural extension
– Invasive fungal granuloma with intracranial–extradural extension
• Maxillary swing
– Juvenile angiofibroma Radkowski's IIIA
|
Minor transient VI n palsy (2), transient III n palsy (1), Intermediate, medial canthus fistula (1), CSF leak requiring repair (1), Major palatal perforation (1)
|
|
Level IV
(n = 33)
|
• Maxillary swing/anterior craniofacial resection
– Juvenile angiofibroma Radkowski's IIIB (ICA exposure/cavernous sinus involvement)
– Nasal dermoid with intracranial extension
– Nose and PNS malignancy with ICA exposure with intradural extn./ICA abutment
|
Intermediate seizure (1), Major extradural hemorrhage (1), palatal perforation (1), death pulmonary thromboembolism (1)
|
|
External lateral skull-base
(n = 125)
|
Level II
(n = 16)
|
• CSF otorrhoea–transmastoid repair/middle cranial fossa repair
• CSF otorrhoea–subtotal petrosectomy
• Skull-base osteomyelitis with medial extn.
|
Minor, transient CSF leak (1), Intermediate postoperative CSF leak repair (1)
|
|
Level III
(n = 64)
|
• Infratemporal fossa approach of Fisch's types A/B
– Middle ear meningioma
– Jugular foramen nonparaganglioma tumors
– Endolymphatic sac tumor
• Anterior petrosectomy-
– Cholesteatoma, skull-base osteomyelitis
• Orbitozygomatic subtemporal-infratemporal approach
– Infratemporal tumors
– Invasive fungal granuloma
– Juvenile angiofibroma IIC/IIIA
|
Intermediate X nerve palsy (5),[a] seizures (1), Major extradural hemorrhage (1), VB/VC nerve section (3)
|
|
Level IV
(n = 45)
|
• Infratemporal fossa approach of Fisch's types A, B, and C
– Tympanojugular paraganglioma
• Orbitozygomatic subtemporal-infratemporal approach
– Juvenile angiofibroma IIIB/ICA exposure
– Nasopharynx malignancy (ACC)
• Subtotal temporal bone resection
– Temporal bone malignancy
• Petrous apex drainage/debridement
– Skull-base osteomyelitis extensively abutting petrous ICA
• Extensive petrous apicitis/tuberculosis
• Translabyrinthine/transotic approach
– Vestibular shwannoma/petrous cholesteatoma/cerebellopontine angle tumors
• Retrolabyrinthine presigmoid approach
• Vestibular neurectomy
• Vascular loop decompression
|
Minor transient CSF leak (2), transient VI nerve palsy (1), Intermediate CSF leak (1), meningitis (1), transient III, IV, VI palsy (2), X–XII paresis/paralysis (5)*, Major VII n palsy HB grade IV–VI (6), weak breathy voice (2), encephalocoele (2), vision loss (1), cochleo-vestibular deficit (4), death carotid blow-out (2), amphotericin toxicity–renal failure (1)[b]
|
Abbreviations: ACC, Adenoid cystic carcinoma; CSF, cerebrospinal fluid; extn. extension; ICA, Internal Carotid Artery; inv., Involvement; plum., not present in the document; PNS, Paranasal sinus.
a Well rehabilitated speech and swallowing postsurgery was classified as an intermediate complication.
b Death due to amphotericin B nephrotoxicity, classified as a nonsurgical complication.
Results
Clinicosurgical Profile of Patients Undergoing Surgery
From January 2002 to March 2015, 342 patients were operated at our otorhinolaryngology and head–neck surgery referral centre for various lesions of the skull-base. Thirteen patients were excluded from this analysis as per the exclusion criteria and 329 patients constitute the final study group.
Details of Surgery
Of the 329 patients included in the study, 204 (62%) patients underwent surgery on the anterior skull-base and 125 (38%) on the lateral skull-base. Among the 204 anterior procedures 120 (59%) were purely endoscopic and 84 were open or external approaches. Endoscopic assistance was used in 17 of the 84 patients with external anterior approaches. There was intracranial intradural involvement in 21 cases (6.4%). The details of surgical procedures as per complexity of surgery are detailed at [Table 2].
Complications Encountered
Among the 204 anterior skull-base procedures, a total of 23 surgical complications were noted (11%). The endoscopic anterior skull-base group had 12 complications in 120 patients (10%) with three major complications and no deaths. The external anterior skull-base group had 11 complications in 84 patients (13%) with three major complications and one death.
Among the 125 lateral skull-base (external) procedures, a total of 41 surgical complications were encountered (33%). The 41 complications included 4 minor, 16 intermediate, 19 major complications, and 2 surgical related deaths. One other death was classified as nonsurgical related (amphotericin B toxicity related acute renal failure) and not included in the analysis of surgical related complications. A significantly lower complication rate was noted in the anterior skull-base approaches (endoscopic + external) than in the lateral skull-base approaches (p < 0.001).
Deficits which were anticipated and expected prior to surgery were classified as sequelae rather than complications. Fifteen out of 18 cases with juvenile angiofibroma (IIIB) excised by the lateral subtemporal–infratemporal approach had section of the maxillary nerve, and were so classified.
Among 21 patients with normal presurgical facial nerve function and facial nerve transposition as part of a Fisch's type A excision, 15 recovered to grade III by the 9-month follow-up, and six had residual function worse than grade IV. The 15 patients with recovery to grade II and III were classified as sequelae rather than a complication.
12 patients had deterioration of lower cranial nerve function following surgery on the jugular foramen. Of them, 10/12 recovered function with compensation, speech, and swallowing therapy and a thyroplasty (n = 9). No patient required long-term tube feeds or a long-term tracheotomy. Two patients had residual weak voices, one of these despite a thyroplasty. Transient speech and swallowing dysfunction which was ultimately well rehabilitated was classified as an intermediate complication (requires treatment but does not cause permanent disability)
Carotid blowouts occurred on three occasions, one intraoperative, and two in the first postoperative week. The intraoperative blow out was successfully treated with limited final morbidity but both patients with postoperative blow outs died.
Analysis of Complications Depending on Surgical Approach and Surgical Complexity
[Table 3] and [Fig. 1] depict the quantum of complications for differing surgical approaches and complexity of surgery (surgical levels). Level II procedures (endoscopic/anterior external/lateral external) were associated with occasional minor and intermediate complications but no major complications or deaths. Level III procedures led to some major complications but no deaths. All three deaths were associated with level IV procedures.
Fig. 1 Quantum of complications with increasing levels of surgical complexity.
Table 3
Surgical complications as per surgical approach and complexity
|
Level II
|
Level III
|
Level IV
|
|
Minor/intermediate
|
Major/death
|
Minor/intermediate
|
Major/death
|
Minor/intermediate
|
Major/death
|
|
Anterior skull-base
|
|
Anterior endoscopic endonasal, n = 120
|
4/55
|
0/55
|
4/53
|
1/53
|
1/12
|
2/12
|
|
Anterior external, n = 84
|
1/23
|
0/23
|
5/28
|
1/28
|
1/33
|
3/33
|
|
Anterior endoscopic vs. anterior external Chi-square test, p-value
|
0.63
|
–
|
0.16
|
0.64
|
0.46
|
0.47
|
|
Lateral skull-base
|
|
Lateral external n =125
|
2/16
|
0/16
|
6/64
|
4/64
|
12/45
|
17/45
|
|
Anterior external vs. at external Chi-square test, p-value
|
0.35
|
–
|
0.25
|
0.60
|
0.006
|
0.004
|
|
Anterior (endoscopic + external) vs. lateral Chi-square test, p-value
|
0.40
|
–
|
0.73
|
0.26
|
0.004
|
0.003
|
A trend is noted for major complications to be more manifest with increasing levels of surgical complexity (level II–IV; p < 0.05). Anterior skull-base endoscopic and external approaches did not differ in terms of complication rates across all levels of surgical complexity ([Table 3]). Lateral approaches are noted to have similar complication rates as anterior approaches for level II and level III procedures but to have a greatly increased complication rate for level IV procedures.
Discussion
Surgical procedures on the skull-base have significant implications for the patient and there are occasions when some surgical morbidity is considered near-inevitable. It is naturally vital that surgical morbidity to be minimized and numerous individual advances in pretreatment radiology, adjuvant embolization, surgical instrumentation, anesthesia, and postsurgical intensive care have all contributed to cumulatively improve surgical care very significantly.[12] The present evaluation from a contemporary practice includes all these individual aids and reports on the quantum of complications in the current era of skull-base surgery. Complications have been categorized by a clear and unambiguous classification system[10] ([Table 1]) which is oriented toward patient outcomes rather than surgical procedures and is easily applied to diverse surgical procedures. Additionally, the large spectrum of skull-base procedures is categorized as per complexity and complication rates quantified for each subcategory.
The complication rate with anterior skull-base surgery is noted to be relatively lesser (11%, total complications; 3%, major complications; 0.5%, death). Lateral skull-base surgery is noted to have a significantly higher complication rate (33%, total complications; 15%, major complications, 1.6%, death; p < 0.001). The vast majority of lateral skull-base complications are consequent to level IV procedures and level II and III procedures of the lateral skull-base are noted to be no more morbid than similarly categorized anterior skull-base procedures.
An integrated and unified grading of the complexity of skull-base surgical procedures is proposed ([Table 1]). The classification is inspired from the widely quoted classification for endoscopic skull-base procedures from the Pittsburg's group[9] and is adapted to be applied to the entire spectrum of skull-base procedures (endoscopic, open anterior, open lateral). Progressing from level II to level IV, a statistically significant progression is noted across all three surgical approaches of an increasing incidence of major complications and mortality ([Table 3]). The progressive increase in complication rates are noted along increasing surgical levels is validation of the strength and appropriateness of the proposed categorization.
The original principle of skull-base surgery at its initiation was to “remove bone and so spare the brain.” The initially developed transtemporal and infratemporal fossa procedures were developed as per this guiding principle.[13]
[14]
[15] Wide bony dissection in the skull-base enabled excellent access to the dura and adjoining neurovascular structures with negligible brain retraction. Wide exposure further enabled closer access to the surgical site with shorter and proximally held instruments, thus leading to more precise and tremor free surgical work on the critical skull-base.
The advent of the endoscope with its illumination, magnification, and angled telescopes has added a new dimension to skull-base surgery. The endoscope provides for superior access and illumination but the surgical field is deeper and surgical dissection with longer instruments and narrow corridors can be technically challenging. Innovations in techniques, such as the “two surgeons-four hands” technique, the routine use of a vascularized nasoseptal flap for augmenting dural repairs, and new instrumentation (neuronavigation, intraoperative Doppler, endoscope holders, microdebrider, coblation, cavitron ultrasonic surgical aspirator [CUSA]) have enabled the development of a safe and effective procedure.[16]
[17]
[18]
[19] Centers with significant experience have noted of reduced morbidity and reduced duration of hospital stay with the endoscopic approach as compared with the external approach[2]
[3]
[4] but apprehensions have been expressed of a higher rate of cerebrospinal fluid (CSF) leaks with endoscopic procedures than with open procedures, especially so with procedures with higher grades of complexity.[2]
[20]
As is noted in this analysis, approximately 42% of our throughput in skull-base today is either totally endoscopic or endoscopic assisted. On initial evaluation, endoscopic procedures seem to have lower complication rates than external procedures (9 vs. 35%, p < 0.001) but this is a consequence of endoscopic procedures being limited to the safer areas of the anterior skull-base. When looking at the anterior skull-base in isolation and evaluating complications, specifically with regard to the complexity of surgery, comparable complication rates are noted for endoscopic and external procedures of comparable surgical complexity ([Table 3]).
A distinction has to be made between sequelae and complications. As a simple distinction, sequelae are expected adverse events, while complications are adverse events which can be potentially prevented. The differentiation between the two would probably be contentious. It would be expected that the informed consent would include a discussion of both but especially emphasize the inevitability/near-inevitability of the sequelae.
This analysis finds a much greater incidence of complications with lateral skull-base surgery. Additionally, this subgroup has a significant morbidity classified as sequelae rather than complications. Surgery on the jugular foramen is often associated with facial nerve dysfunction consequent to its surgical rerouting, and also lower cranial nerve dysfunction. Similarly, it is our experience that in patients with Radkowski's stage IIIB juvenile angiofibroma surgically accessed by the orbitozygomatic subtemporal–infratemporal approach, the VB nerve in its canal is an invariable casualty in the attempt to access tumor positioned medial to it in the Vidian canal complex, inferior and superior orbital fissures, paracavernous area, and sphenoid. In our practice, patients are specifically counseled regarding these deficits, and these are, therefore classified as sequelae rather than complications.
Rehabilitation and correction of residual deficits is an important part of skull-base practice. Such rehabilitation is best if undertaken immediately. In situations of facial nerve paralysis, our preference is to undertake immediate partial rehabilitation by upper eyelid loading (gold weight implantation) but nerve repair procedures take longer to be effective. Swallowing and speech rehabilitation in our practice is primarily undertaken by unilateral cricothyroid approximation (type IV thyroplasty) on the paralysed side.[11] The deficits of a “high” vagal paralysis consequent to jugular foramen lesions include not only the horizontal vocal fold misalignment consequent to recurrent laryngeal nerve (RLN) paralysis but also additional deficits of vertical vocal fold misalignment, vocal fold hypotonia, and sensory loss consequent to superior laryngeal nerve paralysis (SLN). In this situation of a high vagal paralysis, cricothyroid approximation is noted to better correct these deficits than a type I thyroplasty as it better corrects the deficits of vocal fold hypotonia and vertical misalignment and also partially corrects for the horizontal plane glottic gap.[11] Early interventions with cricothyroid approximation have contributed to minimizing swallowing morbidity in this report, with it being downgraded in many instances from being a major complication to an intermediate complication (i.e., a complication which requires treatment but does not cause permanent disability).
Conclusions
Many advances, surgical and nonsurgical, have contributed to the decreasing trends toward complications with skull-base surgery. Attention to rehabilitative procedures has further contributed toward this trend. In a contemporary otolaryngology led and primarily extradural skull-base practice, the overall total complication rate is currently noted at 19%, and the major complication rate at 8%. Mortality continues to be a reality and is noted at 1%. Much greater complication rates are noted with lateral than with anterior skull-base surgery, particularly so with level IV procedures.
The categorization of surgical procedures and complications as proposed here enables an estimation of surgical morbidity and complication rates as per the complexity of the procedure. These should prove useful for informed consent and also in informed decisions regarding the risk-benefit of surgical procedures in individual situations.
