Accuracy of high-resolution computed tomography in cochlear implantation

Accuracy of high-resolution computed tomography in cochlear implantation ALAN W. LANGMAN, MD, and SUZANNEM. QUIGLEY,PhD, Seattle, Washington Multichannel cochlear implants are a proven method for the auditory rehabilitation of individuals who have severe-to-profound sensorineural hearing loss. These devices typically require insertion into the scala tympani of the cochlea to provide auditory stimulations. A patent scala provides the best chance for an adequate insertion of the electrode array. Preoperative high-resolution computed tomography imaging has traditionally been used to determine the patency of the scala tympani. Its ability to accurately predict the patency of the cochlea has been questioned in several retrospective studies. A prospective study was undertaken in 28 consecutive individuals undergoing cochlear implant surgery to compare the findings on high-resolution computed tomography with the surgical findings in an attempt to determine high-resolution computed tomography's accuracy. Cochlear obstruction caused by ossification was accurately predicted in six of six individuals but overestimated in the round window region in three individuals. High-resolution computed tomography accurately predicted patent cochleas in 19 individuals. No false-negative results were encountered. In this study sensitivity of high-resolution computed tomography was 100%, and specificity was 86%. High-resolution computed tomography appears to be more helpful than previously reported for determining cochlear patency. [OTOLAR¥NGOLHEAD NECK $URG 1996;I 14:38-43,]

|ntracochlear cochlear implants are an accepted therapeutic intervention for individuals with severeto-profound sensorineural hearing loss. These devices are usually inserted into the scala tympani to provide the means to stimulate t h e auditory system. Severe-to-profound sensorineural hearing loss may be associated with obstruction of the cochlear scalae. Accurate identification of obstruction within the cochlea is critical in assisting the surgeon to anticipate events at the time of the operation and to plan possible alternative methods for the insertion of the electrode array. ~3 In addition, it may be the determining factor for choosing the ear to be implanted when the results of auditory testing are equivocal. Preoperative identification of cochlear obstruction will also enable the implant team to

From the Divisionof Otology,Neurotology& Skull Base Surgery, Virginia Mason Medical Center. Supported by a grant from the William G. Reed Fund for Research in Otolaryngology. Received for publication March 16, 1995; accepted June 9, 1995. Reprint requests: Alan W. Langman, MD, Division of Otology, Neurotology & Skull Base Surgery, Virginia Mason Medical Center, 1100 Ninth Ave., Seattle, WA 98111. Copyright © 1996by the American Academyof OtolaryngologyHead and Neck Surgery Foundation, Inc. 0194-5998/96/$5.00 + 0 23/1/66921 38

counsel the patient and/or family members of the potential impact of the obstruction. High-resolution computed tomography ( H R C T ) is the most common radiographic technique that is used to evaluate the cochlea for potential cochlear implant surgery. It allows identification of both abnormalities of the temporal bone that would be a contradiction to implantation (e.g., narrow internal auditory canal) and inner ear malformations that provide a diagnosis for the hearing loss. 4,s H R C T also allows evaluation of the patency of the basal turn of the cochlea. The accuracy of H R C T in identifying the presence and extent of obstruction within the scalae of the cochlea has been questioned in numerous studies. 6-11The reported sensitivity and specificity of H R C T in determining patency and obstruction of the cochlea have been variable. Sensitivity, which measures how accurate a test is in identifying the presence of a disease or condition, has ranged from 0% to 91%. Specificity, which measures how accurate a test is in identifying the absence of the diseased state, has ranged from 60% to 100%. Because of the possible limitations of H R C T to identify obstruction within the cochlea, it has been suggested that magnetic resonance imaging (MRI) may be more useful in identifying cochlear obstruction than H R C T . 12'13 Most studies that have investigated the usefulness

Otolaryngology -

Head and Neck Surgery Volume 114

Number 1

LANGMAN and QUlGLEY

Fig. 1. Normal a p p e a r a n c e of the basal turn of the c o c h l e a (small arrow) on an axial HRCT scan.

(large arrow) a n d

39

round window niche

Fig. 2. Axial HRCT scan of the c o c h l e a at the level of the basal turn of the c o c h l e a a n d round window niche in a patient with cochlear otosclerosis. Round window niche is totally obliterated by new b o n e growth (small arrow). Basal turn of the c o c h l e a (large arrow) is patent.

of H R C T in identifying cochlear obstruction have been retrospective and have included older generation H R C T data. 61~ We undertook this study to determine the accuracy of H R C T in a prospective fashion using state-of-the-art radiologic technology to decide whether H R C T should be abandoned in

favor of MRI in the preoperative evaluation of cochlear patency. METHODS AND MATERIAL Patients

with

severe-to-profound

sensorineural

hearing loss who subsequently underwent cochlear

40

Otolaryngology Head and Neck Surgery January 1996

LANGMAN a n d QUIGLEY

Fig. 4. Limited obstruction

(large arrow) of

the basal turn of

the c o c h l e a on an axial HRCT scan in a patient with a transverse temporal b o n e fracture

Fig. 3. Total obstruction of the basal turn of the c o c h l e a an axial HRCT scan.

(arrow) on

implant surgery between November 1990 and February 1994 were consecutively entered into this study. Only patients with no radiographic evidence of an inner ear malformation were included. All patients had an H R C T scan of their temporal bones performed on a G E 9800 scanner (Milwaukee, Wis.). These scans were obtained within 6 months of their implant surgery. The scans consisted of direct images of 1.5 mm thick contiguous slices that were obtained in both axial and coronal planes. The images were noncontrast enhanced and were photographed with a bone algorithm. All scans were reviewed by the operating surgeon before surgery, and a determination of patency or obstruction of the cochlea was made. A patent cochlea was defined as no obstruction either at the round window region or in the basal turn of the cochlea (Fig. i). If obstruction was present, its site and extension were noted (Figs. 2 and 3). Obstruction of the round window area was based on increased soft tissue or bone density in that area as compared with the normal appearance of that area, which is usually seen as

(small arrow).

black on H R C T (Fig. 1). Obstruction of the basal turn of the cochlea was defined as encroachment of the lumen by either soft tissue or bone (Fig. 4). The ear that was implanted was determined by behavioral auditory testing or by the auditory brain stem response to electrical stimulation. All patients received a Cochlear Corporation Nucleus 22-channel cochlear implant through a transmastoid facial recess approach. The cochleostomy was performed at the round window region. Intraoperatively, the presence or absence of obstruction at the round window and scala tympani was determined. Insertion length of the electrode array was also determined intraoperatively. The operative findings were then compared with the preoperative radiographic findings to determine the sensitivity and specificity of H R C T with regard to cochlear patency in this group of patients. RESULTS

The study population consisted of 28 patients, of whom 11 were male and 17 were female. They ranged in age from 2 years to 83 years, with a mean of 40 years. The degree of hearing loss was severe in 5 patients and profound in the other 23. The causes of the hearing loss in this group of patients are summarized in Table 1. Most of the patients' hearing loss was the result of an idiopathic cause. Overall, there was no relationship between the cause of the hearing loss and the finding of obstruction either

Otolaryngology Head and Neck Surgen/ Volume 114 Number 1

LANGMAN and QUIGLEY 41

Table 1. Cause of hearing loss (n = 28) Cause

No. of patients

Idiopathic Meningitis Trauma Viral Otesclerosis Noise Ototoxicity Meniere's disease

15 4 2 2 2 1 1 1

Table 2. Summary of HRCT and surgical findings [n = 28] Potent cochlea

Obstruction of cochlear lumen

Obstruction of round window niche

19 22

2 2

7 4

HRCT Surgery

Table 3. Comparison of surgical and radiographic findings {n = 28)

on H R C T or intraoperatively. The length of insertion of the electrode array ranged from 6 mm to 23.5 ram, with a mean insertion length of 19.75 ram.

Radiographic Findings The round window region and the basal turn of the cochlea were found to be patent in 19 of 28 H R C T scans (Table 2). Obstruction of either the basal turn of the cochlea or the round window niche was identified in nine individuals. This area of obstruction did not extend for more than 2 ram. All patients who had obstruction of the round window niche seen on H R C T had patent basal turns of the cochlea beyond the area of obstruction. Surgical Findings Intraoperatively, obstruction of the basal turn of the cochlea was found in two patients (Table 2). Only one obstruction was severe enough to limit the full insertion of the electrode array. Obliteration of the round window was identified in four individuals. The extent of the obstruction at the round window niche was less than 2 mm in these patients. Patency of the round window niche and the scala tympani was noted in 22 patients. Of the four patients with obstruction in the round window area, all had a patent scala tympani once the round window obstruction was removed. Comparison of Surgical and Radiographic Findings Cochlear obstruction caused by ossification was accurately predicted by H R C T in six of six individuals (true-positive) but overestimated by H R C T in the round window area in three patients (falsepositive) (Table 3). The two patients who had obstruction of the basal turn of the cochlea identified on H R C T were found intraoperatively to have obstruction. H R C T accurately predicted a patent cochlea in 19 individuals (true-negative). H R C T also accurately predicted a patent scala tympani in

Su~ew HRCT

Obstruction

No obstruction

Obstruction No obstruction

6 0

3 19

the three false-positive cases and in the four individuals with round window niche obstruction after removal of the obstruction. There were no falsenegative results.

DISCUSSION Obstruction of the cochlear scalae can be caused by fibrosis or ossification alone or by a combination of these growths. Fibrosis and ossification can result from injury to the membranous portion of the inner ear. Ossification within the cochlea has been divided into two types, metaplastic and osteoplastic, on the basis of the pathophysiologic mechanism of its formation. 14 Obstruction within the fluid-filled spaces of the cochlea can develop as a sequela of several disorders that can produce deafness. These disorders include meningitis, otosclerosis, autoimmune diseases, and temporal bone trauma. 15,~6Obstruction may occur in the round window area or anywhere within the scalae of the cochlea. It has a greater significance in patients undergoing cochlear implant surgery with an intracochlear device. If it occurs in the round window niche or in the basal turn of the cochlea, it may interfere with the insertion of the electrode array. H R C T scanning is useful in the evaluation of individuals with severe-to-profound sensorineural hearing loss. It can identify multiple potential causes for the hearing loss. 4,~ This includes malformations of the inner ear and temporal bone and diseases that affect the otic capsule. H R C T of the temporal bone is typically obtained in the preoperative evaluation of cochlear implant candidates for a diagnostic pur-

42

Otolaryngology Head and Neck Surgery January 1996

LANGMAN a n d QUIGLEY

Table 4. Comparison of HRCT and surgical findings in cochlear implantation Year

No. of ears

Patients with meningitis [%]

Jackler et al.6 Wiet et al, 7 Seicshnaydre et al? Klein et al? Seidman el al. 1° Frau et alY

1987 1990 1992 1992 1994 1994

36 26 31 40 81 81

22 NS 45 NS 39.5 37

Langman and Quigley

1995

28

14

Author

Sensitivity [%]

Specificity [%1

54 73 83 0 36 56-76 (Ave, 67) 88-94 (Ave, 91) 100

100 100 60 100 98 100 50-98 (Ave, 90) 86

NS, Not stated; Ave, average.

pose and to evaluate the potential placement site for the electrode array. Although obstruction of the cochlear lumen is no longer an absolute contraindication for cochlear implant candidacy, its presence may require a modification of the standard surgical procedure for the particular implant that is to be used. Obstruction within the cochlea may also limit the implant's benefit because of the potential inability to insert a significant number of the implant's electrodes. There appears to be a wide variability for HRCT to accurately identify obstruction at different areas in the cochlea. Reports have shown HRCT to be able to identify either obstruction or narrowing within the basal turn of the cochlea with a high degree of accuracy. 6,s HRCT seems to be less accurate in identifying the presence or absence of obstruction at the round window niche. The findings in this study are similar. All the false-positive HRCT readings in this study were related to findings of obstruction of the round window area. The accuracy of HRCT to identify obstruction of the cochlear lumen has been reported by several investigators (Table 4). In these retrospective studies, the sensitivity and specificity of HRCT to identify obstruction of the cochlear lumen have been variablel Some of this variability might be explained by the use of older radiologic technology with its limited image resolution (i.e., greater volume averaging between soft tissue and bone) and early experience with the interpretation Of HRCT of the temporal bone. The findings in this study-sensitivity and specificity of HRCT of 100% and 86%, respectively-suggest that HRCT may be more helpful in the preoperative evaluation than previously reported. The discrepancy between our findings in HRCT's sensitivity and specificity for the evaluation of cochlear patency and those previously reported may be a result of the low incidence of meningitis in our

study population (Table 4). It has been noted that the accuracy of HRCT in evaluating cochlear patency diminishes in patients whose hearing loss is caused by meningitis. 6'1°There were too few patients with hearing loss caused by meningitis in our study to adequately evaluate HRCT in this setting. Because of HRCT's reported inaccuracy for determining cochlear patency, it has been suggested that MRI may be more beneficial. 12'13'17This is based on the ability to observe a difference on MRI between soft tissue and fluid primarily on T2-weighted images. It has been demonstrated that obstruction may be visualized within ~he cochlear lumen on MRI but not demonstrated on ~n HRCT s c a n J 2 MRI may not be superior to HRCT in evaluating cochlear patency. It has been suggested that MRI is inferior to HRCT in cases of cochlear otosclerosisJ 3 At this time there is a limited experience with the use of MRI to determine cochlear patency. It appears that MRI might be most useful in evaluating cochlear implant candidates whose hearing loss is caused by meningitis, as the results from this study demonstrate that HRCT is very helpful in evaluating cochlear patency in a group consisting of individuals whose hearing loss was mostly not caused by meningitis. CONCLUSIONS

HRCT appears to be more helpful in determining cochlear patency than previously reported. It is accurate in evaluating cochlear patency in a group of individuals whose hearing loss was mostly a result of causes other than meningitis. HRCT seems to be less accurate in evaluating obstruction of the round window area. REFERENCES

1. Gantz BJ, McCabe BF, Tyler RS. Use of multichannel cochlear implants in obstructed and obliterated cochleas. OTOLARYNGOLHEAD NECK SURG 1988;98:72-81. 2. Steenerson RL, Gary CB, Wynens MS. Scala vestibuli

' Otolaryngology Head a n d Neck Surgery Volume 14 4 Number I

3.

4.

5. 6.

7.

8.

9.

cochlear implantation for labyrinthine ossification. Am J Otol 1990;11:360-3. Kemink JL, Zimmerman S, Kileny PR. Auditory performance of children with cochlear ossification and partial implant insertion. Laryngoscope 1992;102:1001-6. Shelton C, Luxford WM, Tonokawa LL, et al. The narrow internal auditory canal in children: a contraindication to cochlear implants. OTOtaRVNGOLHEADNECKSURG1989;110: 227-31. Phelps PD, Annis JAD, Robinson PJ. Imaging for cochlear implants. Br J Radiol 1990i63:512-6. Jaclder RK, Luxford WM, Schindler RA, McKerrow WS. Cochlear patency problems in cochlear implantation. Laryngoscope 1987;97:801-5. Wiet RJ, Pyle GM, O'Connor CA, et al. Computed tomography: how accurate a predictor for cochlear implantation? Laryngoscope !990;100:687-92. Seicshnaydre MA, Johnson MH, Hasenstab S, Williams GH. Cochlear implants in children: reliability of computed tomography. OTOtaRyNCOI~HEAO NECKSuR6 1992;107:410-7. Klein It, Bohndorf K, Hermes H, et al. Computed tomography and magnetic resonance imaging in the preoperative work-up for cochlear implantation. Eur J Radiol 1992;15:8992.

LANGMAN and QUIGLEY 43

10. Seidman DA, Chute PM, Parisier S. Temporal bone imaging for cochlear implantation. Laryngoscope 1994;104:562-5. 11. Frau GN, Luxford WM, Lo WWM, et al. High-resolution computed tomography in evaluation of cochlear patency in implant candidates: a comparison with surgical findings. J Laryngol Otol 1994;108:743-8. 12. Lasig R, Terwey B, Battmer RD, Hesse G. Magnetic resonance imaging (MRI) and high resolution computer tomography (HRCT) in cochlear implant candidates. Scand Audiol Suppl 1988;30:197-200. 13. Phelps PD. Fast spin echo MRI in otology. J Laryngol Otol 1994;108:383-94. 14. Kotzias SA, Linthicum FH Jr. Labyrinthine ossification: differences between two types of ectopic bone. Am J Otol 1985;6:490-4. 15. Balkany T, Gantz B, Nadol JB Jr. Multichannel cochlear implants in partially ossified cochleas. Ann Oto! Rhinol Laryngol 1988;97:3-7. 16. Ibrahim RAA, Linthicum FH. Labyrinthine ossificans and cochlear implants. Arch Otolaryngol Head Neck Surg 1980; 106:111-3. 17. Yune HY, Miyamoto RT, Yune ME. Medical imaging in cochlear implant candidates. Am J Oto11991;12(suppt):ll-7.