D. Thane R. Cody’s research while affiliated with University of Minnesota Rochester and other places

This is a report of three patients who presented at the Mayo Clinic over a two-year period. All were initially diagnosed as having Bell's palsy but were later found to have a malignant neoplasm causing the paralysis. Two of the patients had breast carcinoma metastases involving the mastoid portion of the facial nerve. The third patient had an adenocarcinoma of the deep lobe of the parotid that involved the facial nerve distal to the stylomastoid foramen. The course of the facial paralysis in the two patients with the metastatic breast disease was almost identical. It consisted of episodes of pain in the mastoid area, generally in the late evening or during the night, often awakening the patient from sleep. This was then followed by peripheral facial-nerve paralysis, sometimes partial and at other times complete. These episodes lasted from 10 minutes to several hours and then resolved completely. They recurred over several months. The patients were completely asymptomatic and normal on examination in the intervals between episodes of paralysis until it became permanent. Metastatic lesions causing facial paralysis are extremely rare in the literature. In those cases that have been reported, the paralysis was progressive from the start and in the vast majority of cases was either painless or associated with other aural symptoms such as otorrhea, hearing loss, and periauricular swelling. There are two unusual features of these two cases: 1. the initial presentation of a breast metastasis as a facial paralysis; in the first case there were no other metastatic lesions present at diagnosis, whereas the second patient had other, asymptomatic, metastatic nodules; and 2. the multiple, brief, recurring episodes of facial paralysis, which have not previously been reported as a mode of presentation of metastatic disease. The third patient was diagnosed as having Bell's palsy. A facial nerve decompression was performed, and the nerve apparently looked normal. The paralysis failed to resolve. He was later found to have adenocarcinoma of the deep lobe of the parotid that involved the facial nerve distal to the stylomastoid foramen. A report of only one similar case could be found in the literature. The sequence of events in these three cases emphasizes the importance of submitting a patient suspected of having Bell's palsy to a thorough otoneurologic examination.

An experimental model was employed to establish an endotemporal bone facial nerve paralysis in cats. 12 facial nerves were initially surgically decompressed, both bony decompression and sheath decompression, to determine if any harm was done to the nerves by these procedures. Transient harm was found in 3 of the 12 nerves. After the course of the facial paralysis without any decompression was determined, 39 cats underwent either bony or sheath nerve decompression, both immediately after injury and after a delay. The cats were followed clinically and electrophysiologically; the nerve excitability test and the amplitude and latency of muscle response evoked by nerve stimulation were used to evaluate nerve function. While immediate postinjury bony decompression resulted in slightly earlier recovery times, immediate sheath splitting significantly lengthened recovery times and worsened the electrophysiologic test results. Histologic studies confirmed the clinical and electrophysiologic results anatomically. Delayed nerve decompression, either bony or sheath, was not associated with faster recovery rates.

The study involved 42 healthy adult guinea pigs placed into two major groups. In group A, only histologic studies were performed after obliteration of the right vestibular aqueduct, the cochlear aqueduct, or both the vestibular and cochlear aqueducts. In group B, hearing changes were monitored by repeated determination of averaged temporal response thresholds after the same obliterative operations as in group A. At the conclusion of the experiments, histologic studies of temporal bone were performed and hearing changes were correlated with the severity of disease found in the inner ear. Histologic examination of 32 temporal bones from animals in which the vestibular aqueduct or both the vestibular and cochlear aqueducts were obliterated consistently demonstrated endolymphatic hydrops. The endolymphatic hydrops was usually progressive, and its severity was proportional to the length of time from the operation. Serial determination of averaged temporal response thresholds in 18 guinea pigs after these operations demonstrated a progressive deterioration in thresholds. The hearing loss was greater at the lower frequencies. The audiometric changes correlated positively with the severity of endolymphatic hydrops. No significant audiometric or temporal bone histologic changes were found after obliteration of the cochlear aqueduct.

A total of 91 cats were evaluated to determine the effectiveness and nature of closure of total tympanic membrane perforations with homografts preserved in buffered formaldehyde, 70 percent alcohol, benzalkonium in Tis-U-Sol (1:750 solution), or Cialit (1:5,000 aqueous solution). Controls consisted of unoperated ears, ears with partial and total perforations that were not grafted, ears in which the total tympanic membrane was removed and immediately replaced, and ears grafted with fresh homograft tympanic membranes. The study revealed that the removal of the tympanic annulus was necessary to prevent spontaneous regrowth of a new tympanic membrane. The percentage of successful closures of tympanic membrane perforations according to the preservative was formaldehyde 50 percent, alcohol 50 percent, benzalkonium 45 percent, and Cialit 30 percent. The rate of successful closures with fresh autograft tympanic membranes was 40 percent and with fresh homografts 25 percent. The homografts preserved in formaldehyde or alcohol were more rigid and easier to handle than were those preserved in benzalkonium or Cialit. The histologic findings were similar in ears that had developed a new tympanic membrane after grafting with preserved homograft tympanic membranes and in ears with fresh homografts or autografts. The graft served as an inert scaffold for the new tympanic membrane to grow across. Gelfoam used in the middle ear to support either a homograft or an autograft became incorporated into the new tympanic membrane and was associated with a foreign-body reaction. There was, however, no evidence of an immunologic reaction to the homograft.

In the anesthetized chinchilla and guinea pig with electrodes implanted on the dura near the auditory cortex an averaged response (temporal response) could be evoked consistently by pure tones presented to the ear contralateral to the electrode site. The response was a triphasic wave with an onset latency of 7 msec. In both animals temporal response thresholds were similar and lowest for the ear contralateral to the electrode site. Chinchilla behavioral thresholds tended to be lower than temporal response thresholds but the discrepancy was usually not greater than 10 dB. It was concluded that during anesthesia temporal response thresholds determined for the ear contralateral to the electrode site could be used with a reasonable degree of accuracy to monitor auditory acuity in these animals.

Long term results after 878 tympanoplasties using five different methods of repair were analyzed. The five groups were as follows: canal skin and fascia, single fascia, double fascia, canal skin, and canal skin and homograft tympanic membrane. 10% of the graft failures occurred 10 months or longer after operation. The highest percentage of graft takes (88%) and satisfactory hearing improvement (socially adequate level 87%, air bone gap closure to within 15 dB 76%) and the lowest incidence of depression in cochlear reserve (0.7%) was in the double fascia group. Overall hearing results achieved in the 878 tympanoplasties were poorer than anticipated. Differences in results in the five groups, although occasionally large, were not statistically significant. Differences in the percentage of successful graft takes in ears that had no active disease as compared with ears that had active disease and in primary operations as compared with revision operations were not statistically significant. Complications encountered were common and varied.

The purpose of this study was to analyze long‐term hearing results in 122 tympanoplasties with incus repositioning and in 64 tympanoplasties using a homograft incus. The mean postoperative follow‐up period in the autograft incus group was 56 months (range 18 to 95 months) and in the homograft incus group 54 months (range 28 to 87 months). Results were compared after incus placement between the tympanic membrane‐malleus complex and stapes head and placement between the tympanic membrane‐malleus complex and stapes footplate. In each group, hearing results tended to be better after malleus‐to‐stapes head placement. There was no statistically significant difference in the results achieved when similar incus placements in the two groups were compared. When the data on malleus‐to‐stapes head tympanoplasty (152 patients) and malleus‐to‐stapes footplate tympanoplasty (34 patients) in each group were combined, the results with malleus‐to‐stapes head placements were considerably better ( P = 0.01). The best hearing results were not satisfactory, however, with only 54 percent of the malleus‐to‐stapes head placements achieving a socially adequate hearing level and the air‐bone gap being closed to within 15 db on only 32 percent. Hearing results were also analyzed for three subgroups. The autograft and homograft incus groups were combined. The first subgroup consisted of 30 patients who had no additional surgery other than ossicular chain reconstruction; the second subgroup consisted of 53 patients who in addition to ossicular chain reconstruction had closure of the tympanic membrane perforation; the third subgroup consisted of 103 patients who in addition to ossicular chain reconstruction had tympanic membrane perforation repair and mastoidectomy. There was no statistically significant difference in the hearing results among the three subgroups. Eleven percent of the autograft incus group and 13 percent of the homograft incus group initially had excellent hearing results and then, at various times after the operation, developed gradual deterioration in auditory acuity, which eventually resulted in an unsatisfactory hearing level. In revision operations, the most common cause of failure was non‐attachment of the incus to the head of the stapes. Second in frequency was fixation of the ossicular chain by fibrous tissue and impaction of the incus against the posterior bony annular rim. The most serious complication was depression in cochlear reserve, which occurred in 5 percent of the 186 tympanoplasties. This analysis of the long‐term hearing results achieved with incus grafts revealed that the results were far from satisfactory.

The tack operation was done in 140 patients who had endolymphatic hydrops. Vertigo was successfully controlled in 82%. Of the first 50 patients operated on, in whom a tack with a head 0.1 mm thick was used, 84% have had satisfactory control of vertigo at the approximate five-year follow-up period. However, of the subsequent 90 patients operated on, in whom a tack with a head 0.4 mm thick was used, only 81% have had satisfactory control of vertigo. The difference in results probably was due to the tendency for perilymph fistula to develop when a thick-headed tack is employed.

The averaged vertex response evoked by pure tone stimuli was studied in awake human beings between the ages of 2 months and 76 years, and the following conclusions were reached: There was no linear relationship between age and vertex response latency or amplitude. For yet unknown reasons, an occasional child, particularly in the 2- to 3-year age range, has a vertex response that is longer in latency and larger in amplitude than the latency and amplitude usually encountered in the vertex response of patients in the age range from 2 months to 5 years and in the adult population. The characteristics of the averaged vertex response are not altered by distracting the subject's attention from the stimuli; neither are they altered in patients with functional impairment of hearing. Vertex response thresholds determined in patients with functional loss of hearing correlate accurately with their true organic thresholds of hearing. The vertex off-response is a most inconsistent phenomenon in the normal population. When present, the amplitude of the off-response was approximately a third of the amplitude of the on-response and was characterized n most subjects by a large positive components (P2) and a minor negative component (N1).

The averaged vertex response has definite but limited diagnostic uses in clinical otology. The vertex response can be used with a reasonable degree of accuracy (1) to determine objectively the thresholds of hearing in subjects who cannot be tested satisfactorily by conventional audiometric methods; (2) to differentiate conductive from sensorineural hearing loss; and (3) to differentiate organic from functional hearing impairment. Preliminary evidence suggests that the vertex response can be used to distinguish between auditory and pseudoauditory air-bone gap in patients who have been severely hard of hearing since early childhood. At the present time the vertex response has not been found to be of significant value in differentiating cochlear and retrocochlear lesions.