d Neurosurg 75:759-762, 1991
Predicting recovery of facial nerve function following injury from a basilar skull fracture ANDREW B. ADEGBIYE,M.B., F.R.C.S.(C), MOE I. KHAN,M.D., F.R.C.S.(C), AND L. TAN, M.B.,B.S., PH.D.
Departments of ClinicalNeurological Sciences (Neurosurgery) and Social and PreventiveMedicine (Biostatistics), Universityof Saskatchewan, Saskatoon, Canada v" Twenty-five patients with posttraumatic facial nerve palsy were studied. Partial recovery of function had occurred in 95 % of these )atients by 18 months after injury. At 5 months posttrauma, there was some recovery in 92.5% of those with a partial lesion compared with 10% of those with a complete lesion. This difference attains statistical significance. Complete recovery of nerve function had occurred by 10.5 months in 53.5 % of the patients; in 62% of patients with a partial lesion, complete recovery had occurred by 4 months compared with 0% in those with a complete lesion. This difference also attains statistical significance. There was no statistically significant difference in recovery of function between patients with an immediate as opposed to a delayed onset of facial nerve palsy. It was determined that the degree of palsy had a statistically significant influence on recovery of facial nerve function, whereas the time of onset did not. The data presented support a conservative approach to these injuries and it is recommended that the possibility of surgicat treatment should be entertained in patients with complete facial palsy persisting for 12 to 18 months after injury. KEY WOROS 9 skull fracture nerve function
9 facial nerve palsy
T
r~E facial nerve, by virtue of its peculiar anatomy, is more prone to damage when basilar skull fractures occur. In a concurrent but unpublished study, we have demonstrated a stronger correlation between basilar skull fractures and facial nerve palsy than with any other cranial nerve. It has been traditionally stated that the recovery of function is especially favorable with nonsurgical treatment when facial nerve paralysis is incomplete or delayed in onset. L9 Our previous study suggested that the time of onset of the palsy played little role in predicting recovery; this led us to study a larger population, albeit still relatively small, of patients with posttraumatic facial palsy.
Clinical Material and Methods Of approximately 2000 patients with closed head injuries managed at University Hospital, Saskatoon, Canada, between January, 1971, and December, 1982, 26 patients sustained facial nerve palsies. Early in the study, diagnosis was made by both clinical examination and electromyography (EMG) with excellent correlation. After an initial admission period, during which the patients were also evaluated by an otolaryngologist, they were discharged and followed at regular intervals
J. Neurosurg. / Volume 75/November, 1991
9 basilar skull fracture
9
in the outpatient clinic. The mean follow-up period was 34 weeks. Recovery of facial nerve function was assessed at outpatient visits by EMG and clinical examination during the early period of the study but later by purely clinical means. One patient who was given steroids was excluded from the study. The data were subjected to statistical analysis. Results
Patient Age and Sex and Type of Accident There were 16 males (64%) and nine females (36%). The patients ranged in age from 7 to 70 years (mean 34 years). Eleven patients were injured in motor-vehicle accidents, eight in falls, and the remaining six in domestic and other accidents. Neurological and Clinical Assessment There was a history of loss of consciousness in 19 patients. Most of the 25 patients had a Glasgow Coma Scale ~1 (GCS) score of 15 on admission; only five had a GCS score of 10 or less. All patients had clinical or both clinical and radiological evidence of a basilar skull fracture; of these, 20 had otorrhea and nine had rhi759
A. B. Adegbite, M. I. K h a n , a n d L. T a n
FIG. 1. Curve showing percentage of patients with residual deficit at various times posttrauma. N = number of patients. FIG. 3. Curve showing percentage of patients with no residual deficit at various times posttrauma. N = number of patients. norrhea. Nine patients had other skull fractures and four suffered associated facial injuries. Basilar skull fractures were demonstrated on plain skull x-ray films in two patients and on computerized tomography (CT) scans in two, but were not demonstrated by either modality in 21 patients. One patient who had a diagnosis made radiologically exhibited a petrous temporal fracture that was longitudinal in configuration.
Management All patients received antibiotics, frequently ampicillin; there were no cases of meningitis. One patient underwent transmeatal decompression of the facial nerve (and was excluded from the statistical analysis).
Prognosis for Facial Palsy
Fourteen patients had an isolated facial palsy and 11 had multiple cranial nerve palsies. Seven patients had a combined sixth and seventh nerve palsy. The facial nerve palsy was complete in 10 patients, and partial in 15. The onset was immediate in seven and delayed in 16; time of onset was unknown in the remaining two patients. In all patients with a delayed palsy, the lesion was noted within 2 weeks of the injury. The palsy was on the fight side in 10 patients and on the left in 15.
Figure 1 reveals that 95% of the patients had made some recovery by 18 months after the injury. By 5 months posttrauma, 92.5 % of the patients with a partial lesion had made some recovery, compared with only 10% of those with a complete lesion (Fig. 2). This difference was statistically significant (p < 0.001). Figure 3 shows that 53.5% of the patients had made a complete recovery by 1 4 months posttrauma. Of the patients with a partial lesion, 62% had made a complete recovery by 4 months, compared with none of the patients with a complete lesion (Fig. 4). This difference was again statistically significant (p < 0.02). Figure 5
FIG. 2. Curves showing percentage of patients with residual deficit after complete palsy (solid line) or partial palsy (broken line) at various times posttrauma. N = number of patients in each group.
FIG. 4. Curves showing percentage over time of patients with partial palsy (broken line) or complete palsy (solid line) who had no residual deficit. N = number of patients.
Cranial Nerve Palsy
760
J. Neurosurg. / Volume 75/November, 199I
Recovery of facial nerve function after injury
FIG. 5. Curves showing percentage over time of patients with some residual deficit by onset of palsy. Solid line denotes patients with immediate onset of palsy and broken line those with delayed onset of palsy, n = number of patients.
reveals no statistically significant difference in recovery of function between patients with an immediate as opposed to a delayed onset of facial palsy. Using the Cox proportional hazards regression model, it was determined that the degree of facial nerve palsy (whether complete or partial) had a statistically significant influence on recovery of facial nerve function (p < 0.0001 for partial recovery and p < 0.004 for complete recovery) whereas the time of onset did not (p = 0.47 for partial recovery and p = 0.15 for complete recovery).
Discussion
Facial Nerve Injuries with Basilar Skull Fracture Posttraumatic facial nerve palsies occur in 1.4% of patients with head injuries (unpublished data). This is the second most common cause of facial paralysis in adults and the most common cause of facial nerve lesions in children, s These lesions are usually associated with temporal bone fractures which are generally classified as longitudinal or transverse. Twenty percent of longitudinal and 40% to 50% of transverse fractures are accompanied by facial palsy. 5'7 The onset of the paralysis may be immediate or delayed. Immediate paralysis is believed to result from direct trauma to the nerve by tearing, stretching, or actual sectioning, whereas delayed paralysis is usually due to pressure of hemorrhage or edema or even granulation tissue within the fallopian canal.t The various tests commonly relied upon for diagnosis in facial nerve lesions include the pereutancous nerve excitability test, EMG of facial muscles, and Schirmer's test. Prognosis for Recovery The rule of thumb regarding prognosis for nerve recovery has been that when facial paralysis is incomplete or delayed in onset, the outlook is favorable, t'j~ J. Neurosurg. / Volume 75/November, 1991
However, our study suggests that the degree of palsy rather than the time of onset is the major predictor of recovery. This is really not surprising since the pathology of an immediate but incomplete lesion should not be expected to be the same as that of an immediate and complete lesion. The prognosis should necessarily be better in the former than in the latter. The same can be said for delayed palsies that are incomplete or complete. As our results indicate, the prognosis for recovery is relatively poor for all complete lesions, whether delayed or immediate. Surgical exploration has been widely advocated in patients where clinical and electrodiagnosfic studies suggest a complete lesion. It has even been suggested that unless prompt decompression is performed, facial nerve damage may be permanent. Brodsky, et al., 2 obtained satisfactory return of facial nerve function when surgical exploration was performed as late as 14 months after injury. They therefore questioned the prognostic value of electrodiagnostic studies in determining facial nerve recovery and in making a decision as to the necessity and timing of facial nerve surgery. Some recovery was seen in most of our patients with a complete palsy as late as 18 months after injury, although this population was relatively small. Replication of these results on a larger scale would certainly be helpful in settling this controversy.
Indications for Surgical Exploration As stated above, facial nerve decompression has been widely advocated in patients with a complete lesion and possibly also in patients with a partial but progressive lesion. We favor a less aggressive approach. The potential risks to the patient associated with the middle fossa approach to the facial nerve are not insignificant. These include conductive or sensorineural hearing loss, trauma to the labyrinth, cerebrospinal fluid leak, meningitis, intracranial hemorrhage, and temporal lobe damage. 9 As our results indicate, several of the patients in our series who initially had a complete palsy showed some recovery in time. Patients who have complete palsy 12 to 18 months after posttraumatic facial nerve palsy are candidates for some surgical correction of the palsy. Hypoglossal-facial nerve anastomosis provides good functional results in 55% to 80% of patients; 3'4 surprisingly little functional loss results from hypoglossal paralysis. 6 Other available procedures for surgical correction of facial palsy are discussed in detail elsewhere. 8 Conclusions It is our view that the role of immediate decompression for posttraumatic facial palsy is questionable at best. We prefer to adopt a wait-and-see approach in all patients, but in those with complete palsy 12 to 18 months after the injury, some form of surgical correction of the facial palsy (for example, hypoglossal-facial nerve anastomosis) would be recommended. 761
A. B. Adegbite, M. I. Khan, and L. Tan Acknowledgment The authors thank Mrs. Lorraine Osinski for her assistance in the preparation of the manuscript.
7.
References 1. Boles R: Facial, auditory, and vestibular nerve injuries associated with basilar skull fractures, in Youmans JR (ed): Neurological Surgery, ed 2. Philadelphia: WB Saunders, 1982, pp 2251-2260 2. Brodsky L, Eviatar A, Daniller A: Post-traumatic facial nerve paralysis: three cases of delayed temporal bone exploration with recovery. Laryngoscope 93:1560-1565, 1983 3. Clayton MI, Rivron RP, Hanson DR, et al: Evaluation of recent experience in hypoglossal-facial nerve anastomosis in the treatment of facial palsy. J Laryngol Otol 103:63-65, 1989 4. Conley J, Baker DC: Hypoglossal-facial nerve anastomosis for reinnervation of the paralyzed face. Plast Reconstruct Surg 63:63-72, 1979 5. Fisch U: Management of intratemporal facial nerve injuries. J Larynogol Otol 94:129-134, 1980 6. Gavron JP, Clemis JD: Hypoglossal-facial nerve anasto-
762
8.
9. 10. 11.
mosis: a review of forty cases caused by facial nerve injuries in the posterior fossa. Laryngoscope 94: 1447-1450, 1984 Lambert PR, Brackmann DE: Facial paralysis in longitudinal temporal bone fractures: a review of 26 cases. Laryngoscope 94:1022-1026, 1984 Leitner DW: Surgical correction of facial palsy, in Schmidek HH, Sweet WH (eds): Operative Neurosurgical Techniques. Indications, Methods and Results, ed 2. Orlando: Grune & Stratton, 1988, pp 705-708 May M, Klein SR: Facial nerve decompression complications. Laryngoscope 93:299-305, 1983 Mealey J: Skull Fractures in Pediatric Neurosurgery. Surgery of the Developing Nervous System. New York: Grune & Stratton, 1982, pp 289-299 Teasdale G, Jennett B: Assessment of coma and impaired consciousness. A practical scale. Lancet 2:81-84, 1974
Manuscript received May 13, 1988. Accepted in final form July 22, 1991. Address reprint requests to: Andrew B. Adegbite, M.B., F.R.C.S.(C), 30 Durham Street, Suite 440, Sudbury, Ontario P3C 5E4, Canada.
J. Neurosurg. / Volume 75/November, 1991
Predicting recovery of facial nerve function following injury from a basilar skull fracture ANDREW B. ADEGBIYE,M.B., F.R.C.S.(C), MOE I. KHAN,M.D., F.R.C.S.(C), AND L. TAN, M.B.,B.S., PH.D.
Departments of ClinicalNeurological Sciences (Neurosurgery) and Social and PreventiveMedicine (Biostatistics), Universityof Saskatchewan, Saskatoon, Canada v" Twenty-five patients with posttraumatic facial nerve palsy were studied. Partial recovery of function had occurred in 95 % of these )atients by 18 months after injury. At 5 months posttrauma, there was some recovery in 92.5% of those with a partial lesion compared with 10% of those with a complete lesion. This difference attains statistical significance. Complete recovery of nerve function had occurred by 10.5 months in 53.5 % of the patients; in 62% of patients with a partial lesion, complete recovery had occurred by 4 months compared with 0% in those with a complete lesion. This difference also attains statistical significance. There was no statistically significant difference in recovery of function between patients with an immediate as opposed to a delayed onset of facial nerve palsy. It was determined that the degree of palsy had a statistically significant influence on recovery of facial nerve function, whereas the time of onset did not. The data presented support a conservative approach to these injuries and it is recommended that the possibility of surgicat treatment should be entertained in patients with complete facial palsy persisting for 12 to 18 months after injury. KEY WOROS 9 skull fracture nerve function
9 facial nerve palsy
T
r~E facial nerve, by virtue of its peculiar anatomy, is more prone to damage when basilar skull fractures occur. In a concurrent but unpublished study, we have demonstrated a stronger correlation between basilar skull fractures and facial nerve palsy than with any other cranial nerve. It has been traditionally stated that the recovery of function is especially favorable with nonsurgical treatment when facial nerve paralysis is incomplete or delayed in onset. L9 Our previous study suggested that the time of onset of the palsy played little role in predicting recovery; this led us to study a larger population, albeit still relatively small, of patients with posttraumatic facial palsy.
Clinical Material and Methods Of approximately 2000 patients with closed head injuries managed at University Hospital, Saskatoon, Canada, between January, 1971, and December, 1982, 26 patients sustained facial nerve palsies. Early in the study, diagnosis was made by both clinical examination and electromyography (EMG) with excellent correlation. After an initial admission period, during which the patients were also evaluated by an otolaryngologist, they were discharged and followed at regular intervals
J. Neurosurg. / Volume 75/November, 1991
9 basilar skull fracture
9
in the outpatient clinic. The mean follow-up period was 34 weeks. Recovery of facial nerve function was assessed at outpatient visits by EMG and clinical examination during the early period of the study but later by purely clinical means. One patient who was given steroids was excluded from the study. The data were subjected to statistical analysis. Results
Patient Age and Sex and Type of Accident There were 16 males (64%) and nine females (36%). The patients ranged in age from 7 to 70 years (mean 34 years). Eleven patients were injured in motor-vehicle accidents, eight in falls, and the remaining six in domestic and other accidents. Neurological and Clinical Assessment There was a history of loss of consciousness in 19 patients. Most of the 25 patients had a Glasgow Coma Scale ~1 (GCS) score of 15 on admission; only five had a GCS score of 10 or less. All patients had clinical or both clinical and radiological evidence of a basilar skull fracture; of these, 20 had otorrhea and nine had rhi759
A. B. Adegbite, M. I. K h a n , a n d L. T a n
FIG. 1. Curve showing percentage of patients with residual deficit at various times posttrauma. N = number of patients. FIG. 3. Curve showing percentage of patients with no residual deficit at various times posttrauma. N = number of patients. norrhea. Nine patients had other skull fractures and four suffered associated facial injuries. Basilar skull fractures were demonstrated on plain skull x-ray films in two patients and on computerized tomography (CT) scans in two, but were not demonstrated by either modality in 21 patients. One patient who had a diagnosis made radiologically exhibited a petrous temporal fracture that was longitudinal in configuration.
Management All patients received antibiotics, frequently ampicillin; there were no cases of meningitis. One patient underwent transmeatal decompression of the facial nerve (and was excluded from the statistical analysis).
Prognosis for Facial Palsy
Fourteen patients had an isolated facial palsy and 11 had multiple cranial nerve palsies. Seven patients had a combined sixth and seventh nerve palsy. The facial nerve palsy was complete in 10 patients, and partial in 15. The onset was immediate in seven and delayed in 16; time of onset was unknown in the remaining two patients. In all patients with a delayed palsy, the lesion was noted within 2 weeks of the injury. The palsy was on the fight side in 10 patients and on the left in 15.
Figure 1 reveals that 95% of the patients had made some recovery by 18 months after the injury. By 5 months posttrauma, 92.5 % of the patients with a partial lesion had made some recovery, compared with only 10% of those with a complete lesion (Fig. 2). This difference was statistically significant (p < 0.001). Figure 3 shows that 53.5% of the patients had made a complete recovery by 1 4 months posttrauma. Of the patients with a partial lesion, 62% had made a complete recovery by 4 months, compared with none of the patients with a complete lesion (Fig. 4). This difference was again statistically significant (p < 0.02). Figure 5
FIG. 2. Curves showing percentage of patients with residual deficit after complete palsy (solid line) or partial palsy (broken line) at various times posttrauma. N = number of patients in each group.
FIG. 4. Curves showing percentage over time of patients with partial palsy (broken line) or complete palsy (solid line) who had no residual deficit. N = number of patients.
Cranial Nerve Palsy
760
J. Neurosurg. / Volume 75/November, 199I
Recovery of facial nerve function after injury
FIG. 5. Curves showing percentage over time of patients with some residual deficit by onset of palsy. Solid line denotes patients with immediate onset of palsy and broken line those with delayed onset of palsy, n = number of patients.
reveals no statistically significant difference in recovery of function between patients with an immediate as opposed to a delayed onset of facial palsy. Using the Cox proportional hazards regression model, it was determined that the degree of facial nerve palsy (whether complete or partial) had a statistically significant influence on recovery of facial nerve function (p < 0.0001 for partial recovery and p < 0.004 for complete recovery) whereas the time of onset did not (p = 0.47 for partial recovery and p = 0.15 for complete recovery).
Discussion
Facial Nerve Injuries with Basilar Skull Fracture Posttraumatic facial nerve palsies occur in 1.4% of patients with head injuries (unpublished data). This is the second most common cause of facial paralysis in adults and the most common cause of facial nerve lesions in children, s These lesions are usually associated with temporal bone fractures which are generally classified as longitudinal or transverse. Twenty percent of longitudinal and 40% to 50% of transverse fractures are accompanied by facial palsy. 5'7 The onset of the paralysis may be immediate or delayed. Immediate paralysis is believed to result from direct trauma to the nerve by tearing, stretching, or actual sectioning, whereas delayed paralysis is usually due to pressure of hemorrhage or edema or even granulation tissue within the fallopian canal.t The various tests commonly relied upon for diagnosis in facial nerve lesions include the pereutancous nerve excitability test, EMG of facial muscles, and Schirmer's test. Prognosis for Recovery The rule of thumb regarding prognosis for nerve recovery has been that when facial paralysis is incomplete or delayed in onset, the outlook is favorable, t'j~ J. Neurosurg. / Volume 75/November, 1991
However, our study suggests that the degree of palsy rather than the time of onset is the major predictor of recovery. This is really not surprising since the pathology of an immediate but incomplete lesion should not be expected to be the same as that of an immediate and complete lesion. The prognosis should necessarily be better in the former than in the latter. The same can be said for delayed palsies that are incomplete or complete. As our results indicate, the prognosis for recovery is relatively poor for all complete lesions, whether delayed or immediate. Surgical exploration has been widely advocated in patients where clinical and electrodiagnosfic studies suggest a complete lesion. It has even been suggested that unless prompt decompression is performed, facial nerve damage may be permanent. Brodsky, et al., 2 obtained satisfactory return of facial nerve function when surgical exploration was performed as late as 14 months after injury. They therefore questioned the prognostic value of electrodiagnostic studies in determining facial nerve recovery and in making a decision as to the necessity and timing of facial nerve surgery. Some recovery was seen in most of our patients with a complete palsy as late as 18 months after injury, although this population was relatively small. Replication of these results on a larger scale would certainly be helpful in settling this controversy.
Indications for Surgical Exploration As stated above, facial nerve decompression has been widely advocated in patients with a complete lesion and possibly also in patients with a partial but progressive lesion. We favor a less aggressive approach. The potential risks to the patient associated with the middle fossa approach to the facial nerve are not insignificant. These include conductive or sensorineural hearing loss, trauma to the labyrinth, cerebrospinal fluid leak, meningitis, intracranial hemorrhage, and temporal lobe damage. 9 As our results indicate, several of the patients in our series who initially had a complete palsy showed some recovery in time. Patients who have complete palsy 12 to 18 months after posttraumatic facial nerve palsy are candidates for some surgical correction of the palsy. Hypoglossal-facial nerve anastomosis provides good functional results in 55% to 80% of patients; 3'4 surprisingly little functional loss results from hypoglossal paralysis. 6 Other available procedures for surgical correction of facial palsy are discussed in detail elsewhere. 8 Conclusions It is our view that the role of immediate decompression for posttraumatic facial palsy is questionable at best. We prefer to adopt a wait-and-see approach in all patients, but in those with complete palsy 12 to 18 months after the injury, some form of surgical correction of the facial palsy (for example, hypoglossal-facial nerve anastomosis) would be recommended. 761
A. B. Adegbite, M. I. Khan, and L. Tan Acknowledgment The authors thank Mrs. Lorraine Osinski for her assistance in the preparation of the manuscript.
7.
References 1. Boles R: Facial, auditory, and vestibular nerve injuries associated with basilar skull fractures, in Youmans JR (ed): Neurological Surgery, ed 2. Philadelphia: WB Saunders, 1982, pp 2251-2260 2. Brodsky L, Eviatar A, Daniller A: Post-traumatic facial nerve paralysis: three cases of delayed temporal bone exploration with recovery. Laryngoscope 93:1560-1565, 1983 3. Clayton MI, Rivron RP, Hanson DR, et al: Evaluation of recent experience in hypoglossal-facial nerve anastomosis in the treatment of facial palsy. J Laryngol Otol 103:63-65, 1989 4. Conley J, Baker DC: Hypoglossal-facial nerve anastomosis for reinnervation of the paralyzed face. Plast Reconstruct Surg 63:63-72, 1979 5. Fisch U: Management of intratemporal facial nerve injuries. J Larynogol Otol 94:129-134, 1980 6. Gavron JP, Clemis JD: Hypoglossal-facial nerve anasto-
762
8.
9. 10. 11.
mosis: a review of forty cases caused by facial nerve injuries in the posterior fossa. Laryngoscope 94: 1447-1450, 1984 Lambert PR, Brackmann DE: Facial paralysis in longitudinal temporal bone fractures: a review of 26 cases. Laryngoscope 94:1022-1026, 1984 Leitner DW: Surgical correction of facial palsy, in Schmidek HH, Sweet WH (eds): Operative Neurosurgical Techniques. Indications, Methods and Results, ed 2. Orlando: Grune & Stratton, 1988, pp 705-708 May M, Klein SR: Facial nerve decompression complications. Laryngoscope 93:299-305, 1983 Mealey J: Skull Fractures in Pediatric Neurosurgery. Surgery of the Developing Nervous System. New York: Grune & Stratton, 1982, pp 289-299 Teasdale G, Jennett B: Assessment of coma and impaired consciousness. A practical scale. Lancet 2:81-84, 1974
Manuscript received May 13, 1988. Accepted in final form July 22, 1991. Address reprint requests to: Andrew B. Adegbite, M.B., F.R.C.S.(C), 30 Durham Street, Suite 440, Sudbury, Ontario P3C 5E4, Canada.
J. Neurosurg. / Volume 75/November, 1991
