MR evaluation of vestibulocochlear anomalies associated with large endolymphatic duct and sac

BACKGROUND AND PURPOSE

Large endolymphatic duct and sac (LEDS) is one of the most common anomalies seen in patients with congenital sensorineural hearing loss (SNHL), and is known to occur with other inner ear findings. Our purpose was to use high-resolution T2-weighted fast spin-echo (FSE) MR imaging to describe the features and prevalence of specific anomalies that occur in association with LEDS.

METHODS

We retrospectively reviewed MR images of the inner ear obtained in 63 patients with LEDS and in 60 control subjects. We evaluated each image for features of cochlear and vestibular dysplasia, including deficiency of the cochlear modiolus, gross cochlear dysmorphism, asymmetry of the cochlear scalar chambers, enlargement of the membranous vestibule, gross vestibular dysmorphism, and abnormality of the semicircular canals (SCC).

RESULTS

Cochlear anomalies were present in 76% of ears with LEDS. Modiolar deficiency, gross dysmorphism, and scalar asymmetry were seen in 94%, 71%, and 65% of abnormal cochleas, respectively. Vestibular abnormalities were present in 40% of ears with LEDS. Simple enlargement, gross dysmorphism, and distortion of the lateral SCC were seen in 84%, 16%, and 32% of abnormal vestibules, respectively.

CONCLUSION

Coexistent cochlear anomalies, vestibular anomalies, or both are present in most ears with LEDS, and appear as a spectrum of lesions, ranging from subtle dymorphism to overt dysplasia. The presence of coexistent anomalies in LEDS affects treatment decisions and prognosis. Newer techniques of high-resolution FSE MR imaging provide a means of exquisite characterization of LEDS, as well as more sensitive detection of associated vestibulocochlear anomalies.

Sensorineural hearing loss after occlusion of the enlarged vestibular aqueduct

OBJECTIVE

This study aimed to report the hearing results of endolymphatic sac occlusion in patients with enlarged vestibular aqueduct syndrome.

STUDY DESIGN

The study design was a multiinstitutional retrospective case series.

SETTING

The study was conducted at tertiary otologic referral centers.

PATIENTS

The study included 10 previously unreported patients with progressive sensorineural hearing loss and vestibular aqueducts greater than 1.5 mm in diameter on computerized tomography.

INTERVENTION

Occlusion of the enlarged vestibular aqueduct was performed by means of a transmastoid surgical approach. Either intraluminal endolymphatic sac obliteration (five patients) or extraluminal extradural endolymphatic sac obliteration (five patients) was accomplished with temporalis fascia.

MAIN OUTCOME MEASURES

The postoperative pure tone average (PTA) and speech discrimination scores were compared with the preoperative levels using conventional audiometry.

RESULTS

Nine of 10 patients experienced some degree of sensorineural hearing loss. The median change in PTA was a loss of 21 decibels (dB), and 50% of the patients experienced a sensorineural hearing loss greater than 25 dB. Postoperative change in PTA ranged from +10 dB to -59 dB. The median change in speech discrimination score was a loss of 27.5%. Only one patient had an improvement in both speech discrimination score and pure tone averages after surgery. Patients who underwent extraluminal occlusion had a median PTA loss of 12 dB, and patients who underwent open sac occlusion had a median PTA loss of 34 dB. These were not statistically different.

CONCLUSION

In this series of 10 patients, 5 had a greater than 25 dB decrease in hearing after occlusion of the enlarged vestibular aqueduct. Surgical occlusion of the enlarged vestibular aqueduct showed no significant benefit in hearing preservation. The otologic surgeon is alerted to the potential for severe sensiorineural hearing loss after occlusion of the enlarged vestibular aqueduct.

Ultra high resolution nonenhanced fast spin echo magnetic resonance imaging: cost-effective screening for acoustic neuroma in patients with sudden sensorineural hearing loss

The financial burden for the evaluation of patients for acoustic neuroma in an otolaryngology practice is substantial. Patients with sudden sensorineural hearing loss represent a portion of that population seen with unilateral, asymmetric auditory symptoms who require investigation for acoustic neuroma. For these patients, gadolinium-enhanced magnetic resonance imaging is the diagnostic gold standard. Auditory brain stem response testing has been used in the past as a screening test for acoustic neuroma, but its apparent sensitivity has fallen as the ability to image smaller acoustic neuromas has improved. Fast spin echo magnetic resonance imaging techniques without gadolinium have been shown to be as effective in the detection of acoustic neuroma as contrast-enhanced magnetic resonance imaging. Limited nonenhanced fast spin echo magnetic resonance imaging now provides an inexpensive alternative for high-resolution imaging of the internal auditory canal and cerebellopontine angle. Fast spin echo magnetic resonance imaging can now be done at a cost approximating auditory brain stem response testing while providing the anatomic information of contrast-enhanced magnetic resonance imaging. Cost analysis was done in the cases of 58 patients with sudden sensorineural hearing loss by comparing the costs for routine workup and screening of acoustic neuroma with the cost of fast spin echo magnetic resonance imaging with the use of screening protocols based on literature review. The potential cost savings of evaluating patients with sudden sensorineural hearing loss with fast spin echo magnetic resonance imaging for acoustic neuroma was substantial, with a 54% reduction in screening costs. In an era of medical economic scrutiny, fast spin echo magnetic resonance imaging has become the most cost-effective method to screen suspected cases of acoustic tumors at our institution by improving existing technology while reducing the cost of providing that technology and eliminating charges for impedance audiometry, auditory brain stem response testing, and contrast-enhanced magnetic resonance imaging.

Combined transarterial and transvenous embolisation of jugulotympanic paragangliomas

SUMMARY

Paragangliomas of the jugular region present a greater challenge to the surgeon than paragangliomas in other locations. Because of the vascular nature of the tumour, bleeding can be substantial not only from arterial inflow to the tumour, but also from venous bleeding, if not embolised prior to surgery. Six patients were treated with combined transarterial and transvenous embolisation followed by surgical resection. In each instance, the diagnosis of jugulotympanic paraganglioma was suspected based on MR features. Substantial reduction in loss of blood appears to result from the combined transarterial and transvenous embolisation approach.

Bone cement reconstruction of the ossicular chain: a preliminary report

OBJECTIVE

To determine the feasibility and efficacy of using a bone cement, Oto-Cem, to reconstruct the ossicular chain.

STUDY DESIGN

Prospective clinical trial on nine consecutively chosen adult patients with ossicular chain defects.

PATIENTS AND SETTING

Nine patients with ossicular chain defects involving the long process of the incus were treated at the Carolina Ear and Hearing Clinic. The ossicular chain was reconstructed using bone cement by itself or in conjunction with a stapes prosthesis.

MAIN OUTCOME MEASURES

Preoperative audiograms were compared with audiograms 3, 6, and 12 months after reconstruction.

RESULTS

There was a mean pure-tone average (PTA) improvement of 15 dB in patients undergoing incus to stapes suprastructure reconstruction with the bone cement. The incus to mobile footplate reconstruction (using a stapes prosthesis attached to the newly reconstructed incus) resulted in a 34-dB PTA postoperative improvement. Two of the three patients with incus to oval window repairs experienced a 10-dB improvement in PTA. One of the three patients experienced a loss in speech discrimination and a 2-dB loss in PTA.

CONCLUSIONS

Despite the limited number of patients, this preliminary study demonstrates the effectiveness of Oto-Cem in reconstructing a foreshortened incus. There was a substantial hearing improvement in all but one patient in the incus to stapes or the incus to footplate categories.

'Leave me alone' lesions of the petrous apex

PURPOSE

When troublesome MR imaging findings are noted in the petrous apex, the radiologist must determine if the area in question needs surgical therapy. Two nonsurgical entities, asymmetric fatty marrow and fluid-filled petrous air cells (trapped fluid), can be noted on conventional brain MR images and confused with pathologic lesions. Our observation that radiologists do not always confidently define the nonsurgical petrous apex lesions precipitated this investigation.

METHODS

Twenty-three patients with either asymmetric fatty marrow (six) or unilateral effusion in a pneumatized petrous apex (17) on MR images were studied. Eighteen patients underwent high-resolution temporal bone CT. For all patients, the medical charts were reviewed retrospectively and/or the surgical and clinical follow-up findings were reviewed with the referring physician.

RESULTS

In the patients with asymmetric fatty marrow, MR signal intensity followed fat on all sequences. The questioned apex in the patients with trapped fluid showed mixed MR signal characteristics (low to high T1 signal, high T2 signal). CT scans confirmed nonexpansile air-cell opacification.

CONCLUSION

Asymmetric fatty marrow in the petrous apex and petrous air-cell effusions have characteristic MR and CT features that facilitate their correct diagnosis. Effusions with intermediate or high T1 signal are most frequently confused with cholesterol granulomas. In those patients, long-term CT follow-up may be helpful to confirm their stability.

Surgical rehabilitation of facial nerve paralysis

Facial nerve paralysis is a rare and often temporary complication of otologic and neuro-otologic surgery. Associated ocular complications must be avoided to prevent visual loss. When reanastomosis or grafting of the facial nerve is not possible, the VII-XII anastomosis offers reliable reinnervation of the distal facial nerve. For patients whose paralysis has been present for many years, the temporalis muscle transfer provides excellent results. This approach has been proven to give reproducible, aesthetic results with the fewest number of operations.

Medial antebrachial cutaneous nerve: a new donor graft for repair of facial nerve defects at the skull base

When grafting a facial nerve defect after resection of a skull base cancer, the use of the greater auricular nerve is generally contraindicated because of concern of malignant involvement. In the past, the sural nerve was used as a donor graft for reconstruction of the facial nerve. We have found a sensory branch of the median nerve of the upper arm, the medial antebrachial cutaneous (MAC) nerve, to be a suitable option for facial nerve grafting. The MAC nerve provides a good diameter match for the facial nerve and has branching to allow reconstruction of the distal facial nerve in the parotid bed. The length is adequate to graft from the brainstem to the distal facial branches. Loss of sensation is limited and well tolerated by the patient. The MAC nerve has been used in grafting in 15 patients with skull base disease and facial nerve defects. No complications have been encountered, and the functional return appears to be similar to other grafts. In our practice the MAC nerve is a valuable option for facial nerve repair at the skull base.

Surgical exposure of the petrous internal carotid artery: practical application for skull base surgery

When exposing the horizontal petrous carotid artery in preparation for intrapetrous carotid bypass, the surgeon has no definite landmarks to localize the perimeter of the cochlea. The results of this study provide a practical, consistent, and safe method to maximize carotid artery exposure while minimizing cochlear injury. We measured the carotid-cochlea distance (mean, 4.3 mm) and the carotid-cochlear angle (mean, 10.8 degrees) in 33 temporal bones in which the extended middle fossa approach had been performed. We correlated this distance to the width of a Sheehy weapon knife, which can be easily measured intraoperatively. Twenty-five temporal bones were imaged prior to surgical exposure using a new computed tomography (CT) protocol that can be used for preoperative assessment of the carotid-cochlear anatomy. The carotid-cochlea distance and carotid-cochlear angle measured on CT are compared with postsurgical measurements.

Stapedectomy vs stapedotomy. Do you really need a laser?

OBJECTIVE

To compare the effectiveness of different techniques of stapes surgery in improving the hearing of individuals with otosclerosis.

METHODS AND DESIGN

Large and small fenestra techniques, as well as the instrument used to make the fenestra (drill or laser), were compared with regard to effectiveness and rate of side effects. The charts of 875 patients who underwent primary stapedectomies performed by members of the House Ear Clinic, Los Angeles, Calif, were reviewed. Patients who underwent stapedectomy for reasons other than otosclerosis and those with inadequate post-operative bone conduction threshold data were excluded. A group of 550 patients met the criteria. This group was broken into categories depending on the technique of stapedectomy and the instrument used to create the fenestra. The techniques were then compared using air-bone gap closure at different frequencies, pure tone average, and the rate of significant side effects.

RESULTS

The study indicated that small fenestra stapedotomy and large fenestra techniques have similar rates of closure of the air-bone gap. Small fenestra stapedotomy has a slightly lower rate of postoperative sensorineural hearing loss, especially at higher frequencies. With regard to the small fenestra technique, there was no significant difference in either postoperative air-bone gap closure or postoperative sensorineural hearing loss, regardless of whether the fenestra was created by laser or microdrill.

CONCLUSIONS

While we did find a statistically significant difference between the large and small fenestra techniques in postoperative sensorineural hearing loss at higher frequency, the difference is small and is probably not clinically significant. Therefore, we find that similar good results can be obtained by the experienced surgeon using either the large or small fenestra technique. Similarly, we found the laser and microdrill to be equally safe and effective in the creation of the fenestra.

Overlapping thin-section fast spin-echo MR of the large vestibular aqueduct syndrome

PURPOSE

To evaluate a high-resolution, thin-section fast spin-echo MR imaging technique of the inner ear to identify the large vestibular aqueduct syndrome seen on temporal bone CT scans.

METHODS

We retrospectively reviewed the temporal bone CT scans of 21 patients with hearing loss and enlarged bony vestibular aqueducts by CT criteria. High-resolution fast spin-echo MR imaging was then performed on these patients using dual 3-inch phased-array receiver coils fixed in a temporomandibular joint holder and centered over the temporal bones. MR imaging included axial and oblique sagittal fast spin-echo sequences. The diameter of the midvestibular aqueduct on CT scans and the signal at the level of the midaqueduct on MR images were measured on axial sequences, then compared. High-resolution MR imaging with the same protocol was performed in 44 control subjects with normal ears, and similar measurements were taken.

RESULTS

The average size of the enlarged bony vestibular aqueduct on CT scans was 3.7 mm, and the average width of the signal from within the enlarged aqueduct on MR images was 3.8 mm. Statistical analysis showed excellent correlation. MR images alone displayed the enlarged extraosseous endolymphatic sac, which accompanies the enlarged aqueduct in this syndrome. Five ears in three patients with enlarged bony vestibular aqueducts on CT scans showed no evidence of an enlarged endolymphatic duct or sac on MR images. An enlarged endolymphatic sac was seen on MR images in one patient with a bony vestibular aqueduct, which had normal measurements on CT scans. MR imaging alone identified a single case of mild cochlear dysplasia (Mondini malformation). In the 88 normal ears studied, the average size of the endolymphatic sac at its midpoint between the common crus and the external aperture measured on MR images was 0.8 mm (range, 0.5 to 1.4 mm). In 25% of the normal ears, no signal was seen from within the vestibular aqueduct.

CONCLUSION

Thin-section, high-resolution fast spin-echo MR imaging of the inner ear is complementary to CT in studying patients with the large vestibular aqueduct syndrome, as MR imaging better displays the soft tissue and fluid of the membranous labyrinth.

Low-cost high-resolution fast spin-echo MR of acoustic schwannoma: an alternative to enhanced conventional spin-echo MR?

PURPOSE

To determine whether unenhanced high-resolution T2-weighted fast spin-echo MR imaging provides an acceptable and less expensive alternative to contrast-enhanced conventional T1-weighted spin-echo MR techniques in the diagnosis of acoustic schwannoma.

METHODS

We reviewed in a blinded fashion the records of 25 patients with pathologically documented acoustic schwannoma and of 25 control subjects, all of whom had undergone both enhanced conventional spin-echo MR imaging and unenhanced fast spin-echo MR imaging of the cerebellopontine angle/internal auditory canal region. The patients were imaged with the use of a quadrature head receiver coil for the conventional spin-echo sequences and dual 3-inch phased-array receiver coils for the fast spin-echo sequences.

RESULTS

The size of the acoustic schwannomas ranged from 2 to 40 mm in maximum dimension. The mean maximum diameter was 12 mm, and 12 neoplasms were less than 10 mm in diameter. Acoustic schwannoma was correctly diagnosed on 98% of the fast spin-echo images and on 100% of the enhanced conventional spin-echo images. Statistical analysis of the data using the kappa coefficient demonstrated agreement beyond chance between these two imaging techniques for the diagnosis of acoustic schwannoma.

CONCLUSIONS

There is no statistically significant difference in the sensitivity and specificity of unenhanced high-resolution fast spin-echo imaging and enhance T1-weighted conventional spin-echo imaging in the detection of acoustic schwannoma. We believe that the unenhanced high-resolution fast spin-echo technique provides a cost-effective method for the diagnosis of acoustic schwannoma.

Fast spin echo magnetic resonance imaging: clinical application in screening for acoustic neuroma

The advent of magnetic resonance imaging has greatly improved our ability to diagnose acoustic tumors, but it is a relatively expensive imaging modality. In the present climate of medical cost restraints, methods that reduce costs but maintain quality are extremely desirable. We report a new magnetic resonance imaging technique that uses fast spin echo without gadolinium. It provides ultrahigh-resolution images of the internal auditory canal and cerebellopontine angle. The sensitivity of this technique for the detection of acoustic tumors is equivalent to conventional gadolinium-enhanced magnetic resonance imaging, but the global cost is comparable with that of brain stem audiometry. In our practice fast spin echo magnetic resonance imaging has replaced brain stem audiometry as a screening modality to evaluate most acoustic tumor suspects. Also, the intricate detail of the internal auditory canal anatomy provided by this technique is useful in planning surgical removal of acoustic tumors.

Fast spin Echo Magnetic Resonance Imaging: Clinical Application in Screening for Acoustic Neuroma

The advent of magnetic resonance imaging has greatly improved our ability to diagnose acoustic tumors, but it is a relatively expensive imaging modality. In the present climate of medical cost restraints, methods that reduce costs but maintain quality are extremely desirable. We report a new magnetic resonance imaging technique that uses fast spin echo without gadolinium. It provides ultrahigh-resolution images of the internal auditory canal and cerebellopontine angle. The sensitivity of this technique for the detection of acoustic tumors is equivalent to conventional gadolinium-enhanced magnetic resonance imaging, but the global cost is comparable with that of brain stem audiometry. In our practice fast spin echo magnetic resonance imaging has replaced brain stem audiometry as a screening modality to evaluate most acoustic tumor suspects. Also, the intricate detail of the internal auditory canal anatomy provided by this technique is useful in planning surgical removal of acoustic tumors.

Advanced techniques in magnetic resonance imaging in the evaluation of the large endolymphatic duct and sac syndrome

The purpose of this report is to compare temporal bone computed tomography (CT) to high-resolution magnetic resonance (MR) imaging using a novel thin-section fast spin echo (FSE) pulse sequence in identifying and characterizing patients with large vestibular aqueduct syndrome. Sixteen patients with sensorineural hearing loss and a CT diagnosis of large vestibular aqueduct(s) underwent high-resolution fast spin echo magnetic resonance imaging with dual, 3-in phased array receiver coils centered over the external auditory canals. Magnetic resonance imaging parameters included axial and oblique sagittal fast spin echo with an effective slice thickness of 1 mm contiguous. Thirty-eight patients with 76 normal inner ears who underwent MR imaging using this technique had their endolymphatic duct measured. MR alone identified the enlarged endolymphatic sac seen along with the large endolymphatic duct in all cases. Three cases (five inner ears) with enlarged bony vestibular aqueducts on CT showed no evidence of endolymphatic duct or sac enlargement on MR. MR alone identified a single case of mild cochlear anomaly in conjunction with an enlarged endolymphatic duct and sac. In the normal population the size of the normal endolymphatic duct at its midpoint measured from 0.1 to 1.4 mm. Thin-section, high-resolution fast spin echo MR imaging of the inner ear may be superior to CT in the evaluation of patients with the large vestibular aqueduct syndrome.

Ossiculoplasty with Polymaleinate Ionomeric Prosthesis

With the continued concern over the possible transmission of viral infections through homologous middle ear implants, there is increasing pressure to develop a truly biocompatible alloplastic middle ear prosthesis. The polymaleinate ionomer, which has been used in dentistry as a filling and luting material for more than 15 years, has recently been used to construct total and partial ossicular replacement prostheses. In an attempt to evaluate these new implants, a multicenter prospective clinical trial was initiated. To date, 92 patients have undergone implantation. The follow-up interval ranged from 3 months to 22 months. Although it is premature to discuss the long-term results, the preliminary surgical experience and audio-metric data with these implants are reviewed. From a surgical perspective, the ionomeric prostheses were easily contoured with a diamond burr and were not prone to shattering. Preliminary follow-up audiometric data were available on 80 patients (59 partial ossicular replacement prostheses and 21 total ossicular replacement prostheses). Of the 59 partial ossicular replacement prostheses the air-bone gaps (average of 500 Hz, 1 kHz, 2 kHz and 3 kHz) were as follows: 0 dB to 10 dB, 15 (25%) of 59; 11 dB to 20 dB, 20 (34%) of 59; 21 dB to 30 dB, 11 (19%) of 59; and greater than 30 dB, 13 (22%) of 59. Of the 21 total ossicular replacement prostheses the air-bone gaps were as follows: 0 dB to 10 dB, 6 (29%) of 21; 11 dB to 20 dB, 6 (29%) of 21; 21 dB to 30 dB, 5 (24%) of 21; and greater than 30 dB, 4 (19%) of 21.

Ossiculoplasty with polymaleinate ionomeric prosthesis

With the continued concern over the possible transmission of viral infections through homologous middle ear implants, there is increasing pressure to develop a truly biocompatible alloplastic middle ear prosthesis. The polymaleinate ionomer, which has been used in dentistry as a filling and luting material for more than 15 years, has recently been used to construct total and partial ossicular replacement prostheses. In an attempt to evaluate these new implants, a multicenter prospective clinical trial was initiated. To date, 92 patients have undergone implantation. The follow-up interval ranged from 3 months to 22 months. Although it is premature to discuss the long-term results, the preliminary surgical experience and audiometric data with these implants are reviewed. From a surgical perspective, the ionomeric prostheses were easily contoured with a diamond burr and were not prone to shattering. Preliminary follow-up audiometric data were available on 80 patients (59 partial ossicular replacement prostheses and 21 total ossicular replacement prostheses). Of the 59 partial ossicular replacement prostheses the air-bone gaps (average of 500 Hz, 1 kHz, 2 kHz and 3 kHz) were as follows: 0 dB to 10 dB, 15 (25%) of 59; 11 dB to 20 dB, 20 (34%) of 59; 21 dB to 30 dB, 11 (19%) of 59; and greater than 30 dB, 13 (22%) of 59. Of the 21 total ossicular replacement prostheses the air-bone gaps were as follows: 0 dB to 10 dB, 6 (29%) of 21; 11 dB to 20 dB, 6 (29%) of 21; 21 dB to 30 dB, 5 (24%) of 21; and greater than 30 dB, 4 (19%) of 21.

Unilateral acoustic tumors: how often do they recur after translabyrinthine removal?

Unilateral acoustic tumors rarely recur after total translabyrinthine removal. Review of the patient records of the House Ear Clinic revealed five recurrent tumors, for an approximate incidence of 0.3%. A questionnaire was mailed to 857 patients who were at least 9 years postoperative and failed to find any additional recurrent tumors. No preoperative or intraoperative factors were identified to predict recurrence. The average time interval from initial removal to recurrence was approximately 10 years. Flow cytometric analysis did not reveal any fundamental differences between the recurrent acoustic tumor group and a larger group of 112 acoustic tumors. Based on observed growth rates of the recurrent acoustic tumors, a single gadolinium-enhanced magnetic resonance image 5 years after surgery is advised. To prevent recurrence, a margin of normal-appearing proximal eighth cranial nerve should be removed and the nerve stump cauterized.

Modified retrosigmoid approach: use for selected acoustic tumor removal

The authors have used a modified retrosigmoid (suboccipital) approach for removal of acoustic tumors in selected patients who have good preoperative hearing and whose tumor does not reach the brain stem or extend to the lateral third of the internal auditory canal. This report presents the surgical technique and results for 15 acoustic neuromas removed by members of the House Ear Clinic between 1986 and 1991 using this approach. The technique differs importantly from the standard suboccipital approach. A mastoidectomy with decompression of the sigmoid sinus allows forward retraction of the sigmoid sinus, enabling tumor removal without cerebellar retraction. Also, replacement of the craniotomy flap prevents adherence to the dura of the scalp, which may prevent postoperative headaches. Tumor size ranged from 0.8 cm to 4.0 cm (mean, 1.9 cm). At 1 year or more postoperative, 13 of the 14 patients with follow-up available had a House-Brackmann (H-B) facial nerve grade I, and one patient had H-B grade II. Three patients retained good hearing ( < or = 30 dB SRT and > or = 70% speech discrimination) postoperatively, and 57% retained at least measurable hearing. For a patient with good preoperative hearing and a tumor that is medially based, involving the cerebellopontine angle but not extending to the brain stem or the lateral end or the internal auditory canal, the authors will continue to use the retrosigmoid approach for tumor removal.

Surgical management of iatrogenic facial nerve injuries

Surgical management of an iatrogenic facial nerve injury represents a significant challenge for the otologic surgeon. The decision to perform facial nerve grafting is a difficult one and is based on the extent of injury to the nerve. We conducted a review of 22 patients who had sustained iatrogenic facial nerve injuries during otologic surgery that required surgical exploration. The facial nerve was transected more than half its diameter in 13 of the patients. All of these patients' nerves were repaired either with direct reanastomosis of the facial nerve or with a cable nerve graft. The transection was less than 50% in nine of the patients in the study group. Eight of these patients underwent only decompression of the facial nerve. No patient with a neural repair (direct anastomosis or cable graft) had better than a House grade III result. All of the patients undergoing direct anastomosis of the nerve obtained a House grade III result. The most common result in patients undergoing cable nerve grafting was a House grade IV. The only patients with normal or near-normal facial nerve function (House grade I or II) had only decompression of the facial nerve. Five of the eight patients undergoing decompression had results similar to those undergoing cable nerve grafts. We conclude that acceptable results can be obtained when the facial nerve is repaired by direct anastomosis or a cable nerve graft. These results are comparable with those of patients treated with decompression only. When in doubt as to the extent of injury, it is preferable to repair the facial nerve, because the extent of injury may be underestimated.