An.. 0101RlUool1Ary1l(follOl:1992
FACIAL CANAL DEHISCENCE: HISTOLOGIC STUDY AND COMPUTER RECONSTRUCTION HARuo TAKAHASHI, MD
ISAMU SANDO, MD, DMSC
KYOTO, JAPAN
PrrrSBURGH, PENNSYLVANIA
The incidence, location, shape, and dimensions ofdehiscences in the facial canal to the middle ear space were studied in 160 temporal bones obtained from 129 individuals 24 gestational weeks to 109 years of age at death by means of light microscopy and our computer reconstruction and measurement method. Dehiscences were observed in 95 of the 129 individuals (74%) and in 119 of the 160 bones (74%). The incidence was found to be quite high among fetuses and newborns,lowest in individuals in their twenties and then again quite high in the geriatric population (:x2 test, "1.2 = 5.45 and 4.41, p < .05). The most frequent site of dehiscence was the oval window area, particularly in its posterior half (57% of all ears) on the inferior to inferomedial aspects of the canal; these dehiscences were clearly demonstrated in reconstructed images. The incidence of dehiscence in the area of the cochleariform process was 16%, and all these dehiscences were on the lateral to superolateral aspect ofthe canal. The second genu area and the mastoid portion were sites ofdehiscence in 21% and 18%, respectively, of specimens; more than half of the dehiscences in the second genu area and mastoid portion were on the lateral to anterolateral and posterior aspects of the facial canal, respectively. The shape of the dehiscence tended to be oval in the oval window area, but rather irregular in the other areas. The dehiscences ranged from 0.4 to 2.64 rom in length, from 0.12 to 1.59 rom in width, and from 0.03 to 1.87 mm' in surface area. The proximity of these dehiscences to the field of otologic surgery is stressed. KEY WORDS -
computer three-dimensional reconstruction, facial canal dehiscence, temporal bone histopathology.
INTRODUCTION
in this study were from 160 temporal bones obtained from 129 individuals 24 gestational weeks to 109 years old at death. None of the individuals had a history of congenital anomaly or osteolytic diseases such as systemic bone disease, malignant external otitis, or cholesteatoma. All specimens were processed for histologic study according to the technique described by Sando et al l 2 to provide serial 20- to 3D-lAm-thick horizontal (105 bones) or vertical (55 bones) sections.
The facial nerve is one of the most significant and vulnerable structures in the temporal bone. This nerve is at risk during surgical procedures involving the inner ear and/or middle ear, and especially so when there is dehiscence ofthe facial canal on its middle ear side, because such dehiscence increases the possibility of unexpected postoperative facial nerve complications. Having precise information about the facial canal dehiscences is considered of prime importance to otologic surgeons in avoiding such complications, but although there have been several reports on this subject.l-? to our knowledge none has provided detailed topographic data or three-dimensional (3-D) views and dimensions of such dehiscences - information that would be of very great help to otologic surgeons in picturing the anatomy of the area as they may encounter it at operation.
For this study, facial canal dehiscence was defined as an opening in the bone between the facial canal and any middle ear space, without interposition ofa major anatomic structure such as the tensor tympani muscle or the stapedius muscle, and with only a thin layer of connective tissue and the middle ear mucosa between the facial nerve and the middle ear space. The bony opening in the facial canal from which the branch of the superficial petrosal artery exits to the stapes was not included in this definition.
In the study reported here, an attempt was made to clarify the age-related incidence, and to detail the location, shape, and dimensions of facial canal dehiscences by both light microscopy and the computeraided 3-D reconstruction and measurement method we developed.l'U!
Light Microscopy. Every 10th histologic section was studied by light microscopy. The overall incidence of dehiscence and age-related incidence were determined, and the location of each dehiscence was described, not only in the longitudinal but also in the cross-sectional plane (ie, posterior part of the oval window area and on the inferior aspect of the canal).
MATERIALS AND METHODS
The materials (serial histology sections) we used
From the Elizabeth McCullough Knowles Otopathology Laboratory, Division of Otopathology, Department of Otolaryngology, University of Pittsburgh School of Medicine, Eye and Ear Institute of Pittsburgh, Pittsburgh, Pennsylvania. Dr Takahashi is currently at the Department of Otolaryngology, Kyoto University Faculty of Medicine, Kyoto, Japan. This study was supported by research grant 5 RO I DCOOI23from the National Institutes of Health/National Institute on Deafness and Other Communication Disorders. REPRINTS -!sarou Sando, MD, DMSc, Division of Otopathology, Eye and Ear Institute of Pittsburgh, 203 Lothrop St, Pittsburgh, PA 15213.
925
Downloaded from aor.sagepub.com at East Carolina University on July 6, 2015
926
Takahashi & Sando, Facial Canal Dehiscence 100.-----------------------
--
100
80
o~
w
60
o z
Fig 1. Incidence of facial canal dehiscence as function of age. n - number of ears in each age group.
w
C
40
o Z
20
o
fetus
80 (AGE) 9
(n)
wereobservedin95 of the 129 cases (74%) and in 119 of the 160 temporal bones (74%). Figure 1 shows how the age-related incidence of the dehiscences, which was found to be quite high among fetuses (90%) and newborns (85%), showed a gradual decrease to a nadir in individuals in their twenties (58%) and then increased again in middle-aged and geriatric populations (67% to 100%). The differences were statistically significant in incidences of dehiscence 1) in fetuses and children younger than 9 years versus in adolescents and young adults (between 10 and 29 years of age; X? test, X2 = 5.45, p < .05) and 2) in adolescents and young adults versus the geriatric population (over 70 years of age; X2 test, X2 = 4.41, p < .05). A total of 186 facial canal dehiscences were identified in 119 (74%) of the total 160 ears examined. Table 1 shows the distribution of dehiscences along the facial canal and with regard to the dominant aspect in cross section of the facial canal. In the cochleariform process area, dehiscences were observed in 26 ears (16% of the total ears), and all of these dehiscences were on the lateral or superolateral
TABLE 1. SITES OF 186 FACIAL CANALDEHISCENCES IN 119 OF 160 EARS
Area Involved Cochleariform process area Oval window area Anterior only Posterior only
No. of Dehiscences 26
Percentage of All (160) Ears
97
Aspect* Lateral to superolateral (26/26,100%)
16% 61%
7
4% 49%
79
Both Second genu area
34
21%
Mastoid portion
29
18%
11
Lateral (13/18, 72%) Inferior to inferomedial (75/90, 83%)
7%
*Dominant aspect of dehiscence in cross section of facial canal.
Downloaded from aor.sagepub.com at East Carolina University on July 6, 2015
Lateral to anterolateral (19/34,56%) Posterior (16/29,55%)
Takahashi & Sando, Facial Canal Dehiscence
927
LeA
Fig 2. Photomicrograph (anterior-posterior view) of left facial canal and surrounding structures (vertical section) of 4-year-old boy showing dehiscence (between arrowheads) on inferior aspect of oval window area. FN - facial nerve, LeA - lateral semicircular canal ampulla, SF - stapes footplate, SH stapes head, RWN - round window niche, UM - utricular macula, V - vestibule.
yM-. /
V
1mm
aspect ofthe facial canal. In 97 ears (61% of the total ears) a dehiscence was present in the oval window area, and most of these dehiscences (90 of 160 or 56%) were in the posterior half of this area (posterior to the center of the stapes footplate). In 83% of the ears in which the dehiscence was in the posterior half of the oval window area (42% of the total ears), the dehiscence was on the inferior to inferomedial aspect of the facial canal in cross section (Fig 2). In contrast, a dehiscence was found in only 18 ears (11% of the total ears) in the anterior halfof the oval window area; further, in 72% (13 ears) in which the dehiscence was in this area it was on the lateral aspect of the facial canal. A dehiscence was present in the second genu area or in the mastoid portion in 34 ears (21% of the total ears) and in 29 ears (18% of the total ears), respectively. Approximately half (56%) of the dehiscences in the second genu area were on the lateral to anterolateral aspect of the facial canal, opening directly into the facial recess (Fig 3); 24% of them were on the anterior aspect of the canal, opening into the tympanic cavity; and the remaining 20% were on the medial or posterior aspect of the canal, opening into the mastoid air cells. Of dehiscences in the mastoid portion of the facial canal, about half (55%) were on the posterior aspect of the canal and opened into the mastoid air cells; 34% were on the lateral aspect and opened into the facial recess; and the remaining 11% were on the medial or anterior aspect of the facial canal and opened into the tympanic sinus or tympanic cavity, respectively.
Three-dimensional Reconstruction. Facial canal dehiscences were found in a total of34 sites in the 20 temporal bones reconstructed; 5 (15%) were found in the cochleariform process area, 24 (70%) in the oval
window area, 3 (9%) in the second genu area, and 2 (6%) in the mastoid portion (Table 2). Figure 4 shows the length and location of the 24 facial canal dehiscences in the oval window area relative to the stapes footplate. As Fig 4 shows, most of the dehiscences (21 of 24 or 88%) were located posterior to the center of the stapes footplate. The dehiscences tended to be oval in shape (15 of 24 or 63%) in the oval window area (Figs 5 and 6A), but of various shapes in other areas (Fig 6B). It was also found that most of the dehiscences (21 of 24 or 88%) were on the inferior to inferomedial aspect of the canal (Figs 5B and 6A). Table 2 also shows the ranges and averages of the dimensions of the 34 facial canal dehiscences reconstructed and measured. Dehiscence length ranged from 0.4 mm to 2.64 mm; the longest dehiscence and the greatest mean value for dehiscence length were in the oval window area. Dehiscence width ranged from 0.12 mm to 1.59 mm; dehiscences in the mastoid portion were widest. Dehiscence areas ranged from 0.03 mm 2 to 1.87 mm-; the largest dehiscence in this series was in the oval window area, but the average dehiscence size was greatest in the mastoid portion. DISCUSSION
The gross incidence of facial canal dehiscence as determined in the present study (74%) is the largest ever reported, although reported incidences vary considerably (7% to 58% ).1-7 As might be expected, light microscopic studies. reveal a higher incidence of dehiscence (15% to 58% )2,5-7 than the other methods of observation such as gross dissection (15%)1 or operative microscopic examinations during surgery (7%).3 It is notable that we found the incidences of
Downloaded from aor.sagepub.com at East Carolina University on July 6, 2015
Takahashi & Sando, Facial Canal Dehiscence
928
anterior
posterior
-
1 mm
Fig 4. Schematic representation of length and location in relation to stapes footplate of24 reconstructed facialcanal dehiscences in oval window area. Note that most dehiscences are posterior to median of stapes footplate.
facial canal are often located not only just posterosuperior to the stapes, but also on the inferior to inferomedial aspect of the canal- the aspect that faces the surgical field. There have been several reports in the literature that the most frequent location of facial canal dehiscence is adjacent (superior) to the oval window.2-4,6-9 Nager and Proctor,9however, pointed out that dehiscences are particularly likely posterior to the oval window. In the present study, this was clearly shown by the results of computer reconstruction and measurement. Some interesting findings of the present study were that 1)mostdehiscences were on the superolateral or lateral aspect of the facial canal when they occurred in the anterior tympanic portion (cochleariform process area and anterior half of the oval window area), but 2) in the posterior part of the tympanic portion, the most frequent site of dehiscence was on the inferior to inferomedial aspect of the canal, and 3) the lateral or posterior aspect was most frequently involved by dehiscence in the peripheral parts of the facial canal (the second genu area and mastoid portion).
Fig 3. Photomicrograph (superior-inferior view) of right facial canal (horizontal section) of 57-year-old man showing dehiscence (between arrowheads) in second genu area. Note that facial nerve (FN) is widely exposed to facial recess (FR). cr -chorda tympani nerve, MC-mastoid air cells, ME - middle ear cavity, SM - stapedius muscle, TM - tympanic membrane.
dehiscence in the second genu and the mastoid portion to be particularly greater than those revealed by previous similar histologic studies? What we found of great significance to otologic surgeons is that some 40% of all dehiscences were on the inferior to inferomedial aspect of the canal in the posterior half of the oval window area, where they may readily be encountered during middle and/or inner ear surgery. Particularly during stapes surgery, surgeons should be aware that dehiscences of the Area Involved Cochleariform process area Oval window area
We have two comments regarding these results. First, the findings show that surgeons should be watchful for dehiscences, not only during stapes surgery, but also when manipulating the tympanic cavity during tympanoplasty, during cochlear implant surgery through a facial recess approach, and
TABLE 2. DIMENSIONS OF 34 FACIAL CANAL DEHISCENCES No. of Length Width
Area
Dehiscences
(mm)
(mm)
(mm2 )
5
0.84:t 0.91 (0.4 to 2.43) 1.41 :t 0.74 (0.42 to 2.64) 0.92 :t 0.34 (0.49 to 1.34) 1.12:t 0.04 (1.08 to 1.17)
0.53 :t 0.38 (0.23 to 1.14) 0.54:t 0.26 (0.12 to 1.06) 0.23 :t 0.02 (0.2 to 0.27) 1.43 :t 0.16 (1.27 to 1.59)
0.47:t 0.46 (0.07 to 1.27) 0.66:t 0.59 (0.03 to 1.87) 0.15:t 0.03 (0.12 to 0.21) 1.52:t 0.04 (1.48 to 1.56)
24
Second genu area
3
Mastoid portion
2
Values are mean ± SD; values in parentheses are ranges.
Downloaded from aor.sagepub.com at East Carolina University on July 6, 2015
Takahashi & Sando, Facial Canal Dehiscence
929
superior
superior
posterior'
A
A
anterior
inferior
anterior
posterior .
medial
anterior
B Fig 5. Computer-reconstructed images (left ear of3-yearold girl). TP - tympanic portion of facial canal. A) Lateral view of facial canal and dehiscence (dashed line) on inferior aspect of facial canal (behind canal in this view) in posterior part ofoval window area. G - geniculate area of facial canal, LP -labyrinthine portion of facial canal, MP - mastoid portion of facial canal, SF - stapes footplate. B) Close-up view of dehiscence from inferior direction. Note that dehiscence is oval.
during cortical mastoidectomy, because the dehiscences tend to face every surgical field just mentioned. Second, it is interesting to discuss this variety in cross-sectional location of dehiscences in relation to development of the facial canal. The origin of the facial canal has been reported to be a sulcus on the lateral surface ofthe primordial otic capsule. 9,15 With regard to other structures in the otic capsule, I) the anterior part of the tympanic portion of the facial canal is located superolateral or lateral to the cochlear otic capsule, 2) the posterior part of the tympanic portion is located inferior to the otic capsule of the lateral semicircular canal, and 3) the canal is located lateral to the otic capsule ofthe posterior semicircular canal at the second genu to the mastoid portion. It may be noticed that each aspect of the facial canal on which dehiscence is most likely is located on the opposite side ofthe facial canal from the otic capsule. Reichert's cartilage is thought to constitute the side of the facial canal opposite to the sulcus of the otic capsule.? Thus, each aspect of the canal just mentioned is the area that is completed (closed) in cooperation with Reichert's cartilage. These observations lead us to speculate that minute discordance of Reichert's cartilage with the otic capsule when com-
lateral
B Fig 6. Computer-reconstructed images (right ear of 59year-old woman). TP - tympanic portion offacial canal. A) Lateral view of facial canal with dehiscences on superolateral aspect of cochleariform process area (between arrowheads), on inferomedial aspect of posterior half of oval window area (large arrow; dashed line indicates that dehiscence is on opposite side of facial canal from that viewed), and on lateral aspect of second genu area (small arrow). PE - pyramidal eminence, MP mastoid portion of facial canal, SF - stapes footplate. B) Close-upview from superiordirection ofirregularly shaped dehiscence (hatched area) in cochleariform process area of facial canal. LP -labyrinthine portion of facial canal, ST-stapes.
pleting the facial canal might be one cause of the dehiscence in these areas. It is interesting to note that the highest incidence of facial nerve dehiscence is in the oval window area, in which many derivatives of Reichert's cartilage (superstructure of the stapes, pyramidal eminence, etc) are located. The age-related variability in the incidence of facial canal dehiscence seems to fit well with these observations. Dietzel 16 reported a higher incidence of facial canal dehiscence in infants and children 5 to 14 years old, but he did not mention noting a higher incidence in the geriatric population. The fact that ossification ofthe facial canal usually continues until the end of the first year of life 9 may be one reason for the high incidence of facial canal dehiscence in the fetal and newborn periods; this might also, in combination with the immunologic immaturity of this period, account for the comparatively high incidence of facial nerve paralysis due to acute otitis media in
Downloaded from aor.sagepub.com at East Carolina University on July 6, 2015
930
Takahashi & Sando, Facial Canal Dehiscence
infancy and early childhood.!? On the other hand, facial canal dehiscence has also been reported to be seen more often in ears with good pneumatization than in those with poor pneumatization.V This leads us to speculate that even in the same individual, the condition of the facial canal might change with age; a new dehiscence might appear as pneumatization progresses, even if the canal was completely closed early in life. Although mastoid pneumatization has been reported to be completed around 5 to 10 years of age,18 we believe it is also reasonable that air cell development may continue until old age.l? In fact, Goto 20 reported that histologically, mastoid pneumatization in terms ofbone formation and resorption can be noted into old age, although radiographic observation shows that the pneumatized space appears to stop increasing in size around adolescence. These findings do not appear to contradict the results of the present study, which show that the incidence of dehiscence again tends to increase in later years after reaching its lowest point in younger individuals.
Our computer-aided 3-D reconstruction and measurement method permitted precise delineation of the shape and dimensions (not only the length, but also the width and surface area) of the dehiscences. Thus, our method was able to confirm partially both Morikawa's report 1that dehiscences are irregularly shaped in most cases, and the reports ofHough- and Kaplan 4 that dehiscences are usually oval. In fact, as shown by our study, dehiscences in the oval window area are usually oval, but dehiscences identified in other anatomic sites are irregular in shape. Regarding the dimensions of dehiscences, there have been several reports in the literature on the length of dehiscences,I,6,7,9 but no reports on the width or surface area. Baxter? reported the most detailed data on the length ofdehiscences in each area of the facial canal; the means reported were slightly lower than ours, except for dehiscences in the cochleariform process area. This discrepancy may have resulted from a difference in measurement between the two-dimensional and 3-D techniques.
REFERENCES 1. Morikawa M. Surgical anatomy of temporal bone - Part II. Anatomy oftympanic and petrous portion of temporal bone [in Japanese]. Pract Otol (Kyoto) 1925;20:37-82. 2. Guild SR. Natural absence of part of the bony wall of the facial canal. Laryngoscope 1949;59:668-73. 3. Hough NO. Malformations and anatomical variations seen in the middle ear during the operation for mobilization ofthe stapes. Laryngoscope 1958;68:1337-79. 4. Kaplan 1. Congenital dehiscence of the fallopian canal in middle ear surgery. Arch Otolaryngol 1960;72: 197-200. 5. Futakawa A. The pneumatization of the tympanum and the relationship to the facial canal [in Japanese with English abstract]. Nippon Jibiinkoka Gakkai Kaiho 1965;68:845-55. 6. Nagakura M. A histoanatomical study of the facial nerve and facial canal [in Japanese with English abstract]. Nippon Jibiinkoka Gakkai Kaiho 1966;69:1629-45. 7. Baxter A. Dehiscence of the fallopian canal. An anatomical study. J Laryngol OtoI1971;85:587-94. 8. Sataloff RT. Embryology and anomalies of the facial nerve and their surgical implications. New York. NY: Raven Press, 1991:117-21. 9. Nager GT, Proctor B. The facial canal: normal anatomy, variations and anomalies. II. Anatomical variations and anomalies involving the facial canal. Ann Otol Rhinol Laryngol 1982; 91(suppl 97):45-61.
to. Takagi A, Sando I. Computer-aided three-dimensional reconstruction and measurement of the vestibular end-organs. Otolaryngol Head Neck Surg 1988;98:195-202. 11. Takagi A, Sando I. Computer-aided three-dimensional reconstruction: a method of measuring temporal bone structures
including the length of the cochlea. Ann Otol Rhinol Laryngol 1989;98:515-22. 12. Sando I, Doyle WJ, Okuno H, Takahara T, Kitajiri M, Coury WJ III. A method for the histopathological analysis of the temporal bone and the eustachian tube and its accessory structures. Ann 0001 Rhinol LaryngoI1986;95:267-74. 13. Takahashi H, Sando I, Takagi A. Computer-aided threedimensional reconstruction and measurement of the round window niche. Laryngoscope 1989;99:505-9. 14. Takahashi H, Takagi A, Sando I. Computer-aided threedimensional reconstruction and measurement of the round window and its membrane. Otolaryngol Head Neck Surg 1989; to 1: 517-21. 15. Anson BJ, Donaldson JA. Surgical anatomy ofthe temporal bone. 3rd ed. Philadelphia, Pa: WB Saunders, 1981:250-7. 16. Dietzel VK. Uber die Dehiszenzen des Facialiskanals. Z Laryngol Rhinol Otol Ihre Grenzgeb 1961;40:366-79. 17. Kettel K. Facial palsy of otitic origin. With special regard to its prognosis under conservative treatment and the possibilities of improving results by active surgical intervention. An account of 264 cases subjected to reexamination. Arch Otolaryngol 1943;37:303-48. 18. Altmann F. Normal development of the ear and its mechanics. Arch OtolaryngoI1950;52:725-66. 19. Williams G. Developmental anatomy of the ear. In: English GM, ed. Otolaryngology. Philadelphia, Pa: JB Lippincott, 1991:35-7. 20. Goto T. The pneumatization of middle ear and sinus and its clinical significance [in Japanese with English abstract]. Nippon Jibiinkoka Gakkai Kaiho 1954;56:1028-49.
Downloaded from aor.sagepub.com at East Carolina University on July 6, 2015
FACIAL CANAL DEHISCENCE: HISTOLOGIC STUDY AND COMPUTER RECONSTRUCTION HARuo TAKAHASHI, MD
ISAMU SANDO, MD, DMSC
KYOTO, JAPAN
PrrrSBURGH, PENNSYLVANIA
The incidence, location, shape, and dimensions ofdehiscences in the facial canal to the middle ear space were studied in 160 temporal bones obtained from 129 individuals 24 gestational weeks to 109 years of age at death by means of light microscopy and our computer reconstruction and measurement method. Dehiscences were observed in 95 of the 129 individuals (74%) and in 119 of the 160 bones (74%). The incidence was found to be quite high among fetuses and newborns,lowest in individuals in their twenties and then again quite high in the geriatric population (:x2 test, "1.2 = 5.45 and 4.41, p < .05). The most frequent site of dehiscence was the oval window area, particularly in its posterior half (57% of all ears) on the inferior to inferomedial aspects of the canal; these dehiscences were clearly demonstrated in reconstructed images. The incidence of dehiscence in the area of the cochleariform process was 16%, and all these dehiscences were on the lateral to superolateral aspect ofthe canal. The second genu area and the mastoid portion were sites ofdehiscence in 21% and 18%, respectively, of specimens; more than half of the dehiscences in the second genu area and mastoid portion were on the lateral to anterolateral and posterior aspects of the facial canal, respectively. The shape of the dehiscence tended to be oval in the oval window area, but rather irregular in the other areas. The dehiscences ranged from 0.4 to 2.64 rom in length, from 0.12 to 1.59 rom in width, and from 0.03 to 1.87 mm' in surface area. The proximity of these dehiscences to the field of otologic surgery is stressed. KEY WORDS -
computer three-dimensional reconstruction, facial canal dehiscence, temporal bone histopathology.
INTRODUCTION
in this study were from 160 temporal bones obtained from 129 individuals 24 gestational weeks to 109 years old at death. None of the individuals had a history of congenital anomaly or osteolytic diseases such as systemic bone disease, malignant external otitis, or cholesteatoma. All specimens were processed for histologic study according to the technique described by Sando et al l 2 to provide serial 20- to 3D-lAm-thick horizontal (105 bones) or vertical (55 bones) sections.
The facial nerve is one of the most significant and vulnerable structures in the temporal bone. This nerve is at risk during surgical procedures involving the inner ear and/or middle ear, and especially so when there is dehiscence ofthe facial canal on its middle ear side, because such dehiscence increases the possibility of unexpected postoperative facial nerve complications. Having precise information about the facial canal dehiscences is considered of prime importance to otologic surgeons in avoiding such complications, but although there have been several reports on this subject.l-? to our knowledge none has provided detailed topographic data or three-dimensional (3-D) views and dimensions of such dehiscences - information that would be of very great help to otologic surgeons in picturing the anatomy of the area as they may encounter it at operation.
For this study, facial canal dehiscence was defined as an opening in the bone between the facial canal and any middle ear space, without interposition ofa major anatomic structure such as the tensor tympani muscle or the stapedius muscle, and with only a thin layer of connective tissue and the middle ear mucosa between the facial nerve and the middle ear space. The bony opening in the facial canal from which the branch of the superficial petrosal artery exits to the stapes was not included in this definition.
In the study reported here, an attempt was made to clarify the age-related incidence, and to detail the location, shape, and dimensions of facial canal dehiscences by both light microscopy and the computeraided 3-D reconstruction and measurement method we developed.l'U!
Light Microscopy. Every 10th histologic section was studied by light microscopy. The overall incidence of dehiscence and age-related incidence were determined, and the location of each dehiscence was described, not only in the longitudinal but also in the cross-sectional plane (ie, posterior part of the oval window area and on the inferior aspect of the canal).
MATERIALS AND METHODS
The materials (serial histology sections) we used
From the Elizabeth McCullough Knowles Otopathology Laboratory, Division of Otopathology, Department of Otolaryngology, University of Pittsburgh School of Medicine, Eye and Ear Institute of Pittsburgh, Pittsburgh, Pennsylvania. Dr Takahashi is currently at the Department of Otolaryngology, Kyoto University Faculty of Medicine, Kyoto, Japan. This study was supported by research grant 5 RO I DCOOI23from the National Institutes of Health/National Institute on Deafness and Other Communication Disorders. REPRINTS -!sarou Sando, MD, DMSc, Division of Otopathology, Eye and Ear Institute of Pittsburgh, 203 Lothrop St, Pittsburgh, PA 15213.
925
Downloaded from aor.sagepub.com at East Carolina University on July 6, 2015
926
Takahashi & Sando, Facial Canal Dehiscence 100.-----------------------
--
100
80
o~
w
60
o z
Fig 1. Incidence of facial canal dehiscence as function of age. n - number of ears in each age group.
w
C
40
o Z
20
o
fetus
80 (AGE) 9
(n)
wereobservedin95 of the 129 cases (74%) and in 119 of the 160 temporal bones (74%). Figure 1 shows how the age-related incidence of the dehiscences, which was found to be quite high among fetuses (90%) and newborns (85%), showed a gradual decrease to a nadir in individuals in their twenties (58%) and then increased again in middle-aged and geriatric populations (67% to 100%). The differences were statistically significant in incidences of dehiscence 1) in fetuses and children younger than 9 years versus in adolescents and young adults (between 10 and 29 years of age; X? test, X2 = 5.45, p < .05) and 2) in adolescents and young adults versus the geriatric population (over 70 years of age; X2 test, X2 = 4.41, p < .05). A total of 186 facial canal dehiscences were identified in 119 (74%) of the total 160 ears examined. Table 1 shows the distribution of dehiscences along the facial canal and with regard to the dominant aspect in cross section of the facial canal. In the cochleariform process area, dehiscences were observed in 26 ears (16% of the total ears), and all of these dehiscences were on the lateral or superolateral
TABLE 1. SITES OF 186 FACIAL CANALDEHISCENCES IN 119 OF 160 EARS
Area Involved Cochleariform process area Oval window area Anterior only Posterior only
No. of Dehiscences 26
Percentage of All (160) Ears
97
Aspect* Lateral to superolateral (26/26,100%)
16% 61%
7
4% 49%
79
Both Second genu area
34
21%
Mastoid portion
29
18%
11
Lateral (13/18, 72%) Inferior to inferomedial (75/90, 83%)
7%
*Dominant aspect of dehiscence in cross section of facial canal.
Downloaded from aor.sagepub.com at East Carolina University on July 6, 2015
Lateral to anterolateral (19/34,56%) Posterior (16/29,55%)
Takahashi & Sando, Facial Canal Dehiscence
927
LeA
Fig 2. Photomicrograph (anterior-posterior view) of left facial canal and surrounding structures (vertical section) of 4-year-old boy showing dehiscence (between arrowheads) on inferior aspect of oval window area. FN - facial nerve, LeA - lateral semicircular canal ampulla, SF - stapes footplate, SH stapes head, RWN - round window niche, UM - utricular macula, V - vestibule.
yM-. /
V
1mm
aspect ofthe facial canal. In 97 ears (61% of the total ears) a dehiscence was present in the oval window area, and most of these dehiscences (90 of 160 or 56%) were in the posterior half of this area (posterior to the center of the stapes footplate). In 83% of the ears in which the dehiscence was in the posterior half of the oval window area (42% of the total ears), the dehiscence was on the inferior to inferomedial aspect of the facial canal in cross section (Fig 2). In contrast, a dehiscence was found in only 18 ears (11% of the total ears) in the anterior halfof the oval window area; further, in 72% (13 ears) in which the dehiscence was in this area it was on the lateral aspect of the facial canal. A dehiscence was present in the second genu area or in the mastoid portion in 34 ears (21% of the total ears) and in 29 ears (18% of the total ears), respectively. Approximately half (56%) of the dehiscences in the second genu area were on the lateral to anterolateral aspect of the facial canal, opening directly into the facial recess (Fig 3); 24% of them were on the anterior aspect of the canal, opening into the tympanic cavity; and the remaining 20% were on the medial or posterior aspect of the canal, opening into the mastoid air cells. Of dehiscences in the mastoid portion of the facial canal, about half (55%) were on the posterior aspect of the canal and opened into the mastoid air cells; 34% were on the lateral aspect and opened into the facial recess; and the remaining 11% were on the medial or anterior aspect of the facial canal and opened into the tympanic sinus or tympanic cavity, respectively.
Three-dimensional Reconstruction. Facial canal dehiscences were found in a total of34 sites in the 20 temporal bones reconstructed; 5 (15%) were found in the cochleariform process area, 24 (70%) in the oval
window area, 3 (9%) in the second genu area, and 2 (6%) in the mastoid portion (Table 2). Figure 4 shows the length and location of the 24 facial canal dehiscences in the oval window area relative to the stapes footplate. As Fig 4 shows, most of the dehiscences (21 of 24 or 88%) were located posterior to the center of the stapes footplate. The dehiscences tended to be oval in shape (15 of 24 or 63%) in the oval window area (Figs 5 and 6A), but of various shapes in other areas (Fig 6B). It was also found that most of the dehiscences (21 of 24 or 88%) were on the inferior to inferomedial aspect of the canal (Figs 5B and 6A). Table 2 also shows the ranges and averages of the dimensions of the 34 facial canal dehiscences reconstructed and measured. Dehiscence length ranged from 0.4 mm to 2.64 mm; the longest dehiscence and the greatest mean value for dehiscence length were in the oval window area. Dehiscence width ranged from 0.12 mm to 1.59 mm; dehiscences in the mastoid portion were widest. Dehiscence areas ranged from 0.03 mm 2 to 1.87 mm-; the largest dehiscence in this series was in the oval window area, but the average dehiscence size was greatest in the mastoid portion. DISCUSSION
The gross incidence of facial canal dehiscence as determined in the present study (74%) is the largest ever reported, although reported incidences vary considerably (7% to 58% ).1-7 As might be expected, light microscopic studies. reveal a higher incidence of dehiscence (15% to 58% )2,5-7 than the other methods of observation such as gross dissection (15%)1 or operative microscopic examinations during surgery (7%).3 It is notable that we found the incidences of
Downloaded from aor.sagepub.com at East Carolina University on July 6, 2015
Takahashi & Sando, Facial Canal Dehiscence
928
anterior
posterior
-
1 mm
Fig 4. Schematic representation of length and location in relation to stapes footplate of24 reconstructed facialcanal dehiscences in oval window area. Note that most dehiscences are posterior to median of stapes footplate.
facial canal are often located not only just posterosuperior to the stapes, but also on the inferior to inferomedial aspect of the canal- the aspect that faces the surgical field. There have been several reports in the literature that the most frequent location of facial canal dehiscence is adjacent (superior) to the oval window.2-4,6-9 Nager and Proctor,9however, pointed out that dehiscences are particularly likely posterior to the oval window. In the present study, this was clearly shown by the results of computer reconstruction and measurement. Some interesting findings of the present study were that 1)mostdehiscences were on the superolateral or lateral aspect of the facial canal when they occurred in the anterior tympanic portion (cochleariform process area and anterior half of the oval window area), but 2) in the posterior part of the tympanic portion, the most frequent site of dehiscence was on the inferior to inferomedial aspect of the canal, and 3) the lateral or posterior aspect was most frequently involved by dehiscence in the peripheral parts of the facial canal (the second genu area and mastoid portion).
Fig 3. Photomicrograph (superior-inferior view) of right facial canal (horizontal section) of 57-year-old man showing dehiscence (between arrowheads) in second genu area. Note that facial nerve (FN) is widely exposed to facial recess (FR). cr -chorda tympani nerve, MC-mastoid air cells, ME - middle ear cavity, SM - stapedius muscle, TM - tympanic membrane.
dehiscence in the second genu and the mastoid portion to be particularly greater than those revealed by previous similar histologic studies? What we found of great significance to otologic surgeons is that some 40% of all dehiscences were on the inferior to inferomedial aspect of the canal in the posterior half of the oval window area, where they may readily be encountered during middle and/or inner ear surgery. Particularly during stapes surgery, surgeons should be aware that dehiscences of the Area Involved Cochleariform process area Oval window area
We have two comments regarding these results. First, the findings show that surgeons should be watchful for dehiscences, not only during stapes surgery, but also when manipulating the tympanic cavity during tympanoplasty, during cochlear implant surgery through a facial recess approach, and
TABLE 2. DIMENSIONS OF 34 FACIAL CANAL DEHISCENCES No. of Length Width
Area
Dehiscences
(mm)
(mm)
(mm2 )
5
0.84:t 0.91 (0.4 to 2.43) 1.41 :t 0.74 (0.42 to 2.64) 0.92 :t 0.34 (0.49 to 1.34) 1.12:t 0.04 (1.08 to 1.17)
0.53 :t 0.38 (0.23 to 1.14) 0.54:t 0.26 (0.12 to 1.06) 0.23 :t 0.02 (0.2 to 0.27) 1.43 :t 0.16 (1.27 to 1.59)
0.47:t 0.46 (0.07 to 1.27) 0.66:t 0.59 (0.03 to 1.87) 0.15:t 0.03 (0.12 to 0.21) 1.52:t 0.04 (1.48 to 1.56)
24
Second genu area
3
Mastoid portion
2
Values are mean ± SD; values in parentheses are ranges.
Downloaded from aor.sagepub.com at East Carolina University on July 6, 2015
Takahashi & Sando, Facial Canal Dehiscence
929
superior
superior
posterior'
A
A
anterior
inferior
anterior
posterior .
medial
anterior
B Fig 5. Computer-reconstructed images (left ear of3-yearold girl). TP - tympanic portion of facial canal. A) Lateral view of facial canal and dehiscence (dashed line) on inferior aspect of facial canal (behind canal in this view) in posterior part ofoval window area. G - geniculate area of facial canal, LP -labyrinthine portion of facial canal, MP - mastoid portion of facial canal, SF - stapes footplate. B) Close-up view of dehiscence from inferior direction. Note that dehiscence is oval.
during cortical mastoidectomy, because the dehiscences tend to face every surgical field just mentioned. Second, it is interesting to discuss this variety in cross-sectional location of dehiscences in relation to development of the facial canal. The origin of the facial canal has been reported to be a sulcus on the lateral surface ofthe primordial otic capsule. 9,15 With regard to other structures in the otic capsule, I) the anterior part of the tympanic portion of the facial canal is located superolateral or lateral to the cochlear otic capsule, 2) the posterior part of the tympanic portion is located inferior to the otic capsule of the lateral semicircular canal, and 3) the canal is located lateral to the otic capsule ofthe posterior semicircular canal at the second genu to the mastoid portion. It may be noticed that each aspect of the facial canal on which dehiscence is most likely is located on the opposite side ofthe facial canal from the otic capsule. Reichert's cartilage is thought to constitute the side of the facial canal opposite to the sulcus of the otic capsule.? Thus, each aspect of the canal just mentioned is the area that is completed (closed) in cooperation with Reichert's cartilage. These observations lead us to speculate that minute discordance of Reichert's cartilage with the otic capsule when com-
lateral
B Fig 6. Computer-reconstructed images (right ear of 59year-old woman). TP - tympanic portion offacial canal. A) Lateral view of facial canal with dehiscences on superolateral aspect of cochleariform process area (between arrowheads), on inferomedial aspect of posterior half of oval window area (large arrow; dashed line indicates that dehiscence is on opposite side of facial canal from that viewed), and on lateral aspect of second genu area (small arrow). PE - pyramidal eminence, MP mastoid portion of facial canal, SF - stapes footplate. B) Close-upview from superiordirection ofirregularly shaped dehiscence (hatched area) in cochleariform process area of facial canal. LP -labyrinthine portion of facial canal, ST-stapes.
pleting the facial canal might be one cause of the dehiscence in these areas. It is interesting to note that the highest incidence of facial nerve dehiscence is in the oval window area, in which many derivatives of Reichert's cartilage (superstructure of the stapes, pyramidal eminence, etc) are located. The age-related variability in the incidence of facial canal dehiscence seems to fit well with these observations. Dietzel 16 reported a higher incidence of facial canal dehiscence in infants and children 5 to 14 years old, but he did not mention noting a higher incidence in the geriatric population. The fact that ossification ofthe facial canal usually continues until the end of the first year of life 9 may be one reason for the high incidence of facial canal dehiscence in the fetal and newborn periods; this might also, in combination with the immunologic immaturity of this period, account for the comparatively high incidence of facial nerve paralysis due to acute otitis media in
Downloaded from aor.sagepub.com at East Carolina University on July 6, 2015
930
Takahashi & Sando, Facial Canal Dehiscence
infancy and early childhood.!? On the other hand, facial canal dehiscence has also been reported to be seen more often in ears with good pneumatization than in those with poor pneumatization.V This leads us to speculate that even in the same individual, the condition of the facial canal might change with age; a new dehiscence might appear as pneumatization progresses, even if the canal was completely closed early in life. Although mastoid pneumatization has been reported to be completed around 5 to 10 years of age,18 we believe it is also reasonable that air cell development may continue until old age.l? In fact, Goto 20 reported that histologically, mastoid pneumatization in terms ofbone formation and resorption can be noted into old age, although radiographic observation shows that the pneumatized space appears to stop increasing in size around adolescence. These findings do not appear to contradict the results of the present study, which show that the incidence of dehiscence again tends to increase in later years after reaching its lowest point in younger individuals.
Our computer-aided 3-D reconstruction and measurement method permitted precise delineation of the shape and dimensions (not only the length, but also the width and surface area) of the dehiscences. Thus, our method was able to confirm partially both Morikawa's report 1that dehiscences are irregularly shaped in most cases, and the reports ofHough- and Kaplan 4 that dehiscences are usually oval. In fact, as shown by our study, dehiscences in the oval window area are usually oval, but dehiscences identified in other anatomic sites are irregular in shape. Regarding the dimensions of dehiscences, there have been several reports in the literature on the length of dehiscences,I,6,7,9 but no reports on the width or surface area. Baxter? reported the most detailed data on the length ofdehiscences in each area of the facial canal; the means reported were slightly lower than ours, except for dehiscences in the cochleariform process area. This discrepancy may have resulted from a difference in measurement between the two-dimensional and 3-D techniques.
REFERENCES 1. Morikawa M. Surgical anatomy of temporal bone - Part II. Anatomy oftympanic and petrous portion of temporal bone [in Japanese]. Pract Otol (Kyoto) 1925;20:37-82. 2. Guild SR. Natural absence of part of the bony wall of the facial canal. Laryngoscope 1949;59:668-73. 3. Hough NO. Malformations and anatomical variations seen in the middle ear during the operation for mobilization ofthe stapes. Laryngoscope 1958;68:1337-79. 4. Kaplan 1. Congenital dehiscence of the fallopian canal in middle ear surgery. Arch Otolaryngol 1960;72: 197-200. 5. Futakawa A. The pneumatization of the tympanum and the relationship to the facial canal [in Japanese with English abstract]. Nippon Jibiinkoka Gakkai Kaiho 1965;68:845-55. 6. Nagakura M. A histoanatomical study of the facial nerve and facial canal [in Japanese with English abstract]. Nippon Jibiinkoka Gakkai Kaiho 1966;69:1629-45. 7. Baxter A. Dehiscence of the fallopian canal. An anatomical study. J Laryngol OtoI1971;85:587-94. 8. Sataloff RT. Embryology and anomalies of the facial nerve and their surgical implications. New York. NY: Raven Press, 1991:117-21. 9. Nager GT, Proctor B. The facial canal: normal anatomy, variations and anomalies. II. Anatomical variations and anomalies involving the facial canal. Ann Otol Rhinol Laryngol 1982; 91(suppl 97):45-61.
to. Takagi A, Sando I. Computer-aided three-dimensional reconstruction and measurement of the vestibular end-organs. Otolaryngol Head Neck Surg 1988;98:195-202. 11. Takagi A, Sando I. Computer-aided three-dimensional reconstruction: a method of measuring temporal bone structures
including the length of the cochlea. Ann Otol Rhinol Laryngol 1989;98:515-22. 12. Sando I, Doyle WJ, Okuno H, Takahara T, Kitajiri M, Coury WJ III. A method for the histopathological analysis of the temporal bone and the eustachian tube and its accessory structures. Ann 0001 Rhinol LaryngoI1986;95:267-74. 13. Takahashi H, Sando I, Takagi A. Computer-aided threedimensional reconstruction and measurement of the round window niche. Laryngoscope 1989;99:505-9. 14. Takahashi H, Takagi A, Sando I. Computer-aided threedimensional reconstruction and measurement of the round window and its membrane. Otolaryngol Head Neck Surg 1989; to 1: 517-21. 15. Anson BJ, Donaldson JA. Surgical anatomy ofthe temporal bone. 3rd ed. Philadelphia, Pa: WB Saunders, 1981:250-7. 16. Dietzel VK. Uber die Dehiszenzen des Facialiskanals. Z Laryngol Rhinol Otol Ihre Grenzgeb 1961;40:366-79. 17. Kettel K. Facial palsy of otitic origin. With special regard to its prognosis under conservative treatment and the possibilities of improving results by active surgical intervention. An account of 264 cases subjected to reexamination. Arch Otolaryngol 1943;37:303-48. 18. Altmann F. Normal development of the ear and its mechanics. Arch OtolaryngoI1950;52:725-66. 19. Williams G. Developmental anatomy of the ear. In: English GM, ed. Otolaryngology. Philadelphia, Pa: JB Lippincott, 1991:35-7. 20. Goto T. The pneumatization of middle ear and sinus and its clinical significance [in Japanese with English abstract]. Nippon Jibiinkoka Gakkai Kaiho 1954;56:1028-49.
Downloaded from aor.sagepub.com at East Carolina University on July 6, 2015
