Characteristics of the Facial Canal–Lateral Canal Dehiscence: The Under‐recognized Third Window

Abstract

Objective

We aimed to establish normative data on the bony partition between the facial canal (FC) and the lateral semicircular canal (LSCC) and present our initial series of cases in which an FC‐LSCC dehiscence was identified based on these anatomic criteria, with or without other third window pathologies.

Study Design

Normative data: Analysis of archival otopathological human temporal bone specimens and computed tomography (CT) scans. Clinical data: Case studies.

Setting

An otopathology laboratory and a specialized otolaryngology, head and neck surgery outpatient clinic.

Methods

We measured the bony partition thickness between the FC‐LSCC in human temporal bone otopathological specimens and CT scans. The clinical study involved a series of reviews of patients with FC‐LSCC, presenting clinical data and CT images.

Results

The average thickness of the bony partition measured via CT was 0.6 mm ± 0.2 mm, whereas in otopathological specimens it was 0.56 mm ± 0.17 mm. We identified 34 patients with FC‐LSCC dehiscence. The most frequently reported symptoms were hearing loss (44%), dizziness/vertigo (44%), and tinnitus (41%). Of the patients, 15 (44%) had more than 1 site of bone dehiscence and 21 (62%) had bilateral FC‐LSCC dehiscence. We then identified 36 patients in whom only FC‐LSCC dehiscences were found and provided characteristics for this population.

Conclusion

Our study provides normative characteristics on the bony partition between the FC‐LSCC and the signs and symptoms of this third window abnormality. FC‐LSCC should be included in differential diagnoses of patients presenting with third window syndrome symptoms.

The recognition of a third window abnormality began with Minor et al in 1998, ¹ specifically with the dehiscence of the superior semicircular canal. Since then, several other third window findings have been identified, sharing similar clinical features and symptomatology. ² , ³ , ⁴ , ⁵ Nevertheless, the recognition of a non‐infectious, non‐erosive, dehiscence of the FC‐LSCC was not reported until 2007 by Bassim et al. ⁶ Until recently, only a few cases were reported. Zhang et al reported on this in 2011, ⁷ and Hernandez‐Trejo et al also referred to the co‐occurrence of an FC‐LSCC dehiscence in 2020. ⁸ No one has defined or characterized this entity more extensively.

Understanding the coexistence of opposing bony dehiscences and the attendant collection of potential otologic signs and symptoms requires understanding the embryology and the anatomy of the local region where the tympanic segment of the FC and the LSCC lie in proximity. Hernandez‐Trejo et al ⁸ stated, “The development of the FC is not an isolated process and is closely related to the ossification and complex development of the temporal bone.” Their review of 184 temporal bones by computed tomography (CT), in which FC dehiscence was identified in 51.2% of the bones, found that nearly 12% also exhibited an LSCC fistula. ⁸ Others ⁹ , ¹⁰ cite that the intraoperative identification of an FC dehiscence was uniformly found within the tympanic segment and is associated with a lateral canal fistula in cholesteatoma surgery, an association pointing to the unique proximity and relationship between the 2.

Dehiscence of the otic capsule, similar to that of the FC‐LSCC type, includes other types, such as cochlear‐FC dehiscence. ³ The proximal juxtaposition of the FC and the cochlea mirrors that of the FC and the LSCC. Song et al ⁴ evaluated the cochlear–facial dehiscence (C‐F) nerve partition width to determine the overall prevalence of radiographic cochlear–facial dehiscence. The average cochlear–facial nerve partition width of 406 assessed ears via CT imaging was 0.6 mm ± 0.2 mm. We envisioned a potentially similar opportunity to define the FC‐LSCC dehiscence. In this study, we aimed to establish normative radiologic and histopathologic data on the bony partition between the FC and LSCC and present an initial series of cases in which an FC‐LSCC dehiscence was identified.

We present radiologic and correlated histopathologic characteristics of the partition between the LSCC and the tympanic segment of the FC and present a second group of patients identified with FC‐LSCC dehiscences and their associated signs and symptoms. We document the presentation, hearing, and clinical features of these individuals.

Patient Selection

The study was conducted from January to December 2021. From the otology and neuro‐otology clinical practice of the senior author, within the Ear, Nose and Throat Clinic, PA of Minnesota, we retrospectively sought out patients who had presented to the practice that year with symptoms of third window syndrome (autophony, pulsatile tinnitus, low‐frequency conductive hearing loss, hyperacusis, and sound‐induced dizziness), ultimately identifying 70 patients. For diagnostic purposes, these patients underwent fine‐cut temporal bone CT scans. Patients who did not have had these symptoms, but who had undergone CT scans of the temporal bones were also selected as controls. The scans were then evaluated for image quality, disease process, and potential for third window pathology. Exclusion criteria included an inability to visualize the partition between the fallopian and lateral canals, fistulous erosion of the LSCC by neoplasia or disease, poor image quality, and/or imaging not acquired in an axial plane with resolution permitting high‐resolution reconstruction in coronal, Stenver, and Pӧschl planes. All data collected were immediately assigned to a unique study ID number, and individuals were de‐identified. From patient charts, we collected information such as date of birth and sex, hearing levels, signs and symptoms, and clinical features.

We accessed and examined temporal bones from the Otopathology Lab at the University of Minnesota. Of the 2200 temporal bones available within the Otopathology Lab library, the medical histories of 300 donors were reviewed, and 25 temporal bones were selected. For each temporal bone evaluated, the following criteria were required: age 18 years and older, no evident facial nerve dehiscence, no loss of structural integrity, and labeled as “within normal limits” or unremarkable.

Using the Department of Health and Human Services regulations, this study was deemed exempt from Institutional Review Board oversight following an ethical review by Advarra Inc., Columbia, MD, and meeting the criteria found in 45 CFR 46.104(d)(4) on July 22, 2022.

Clinical Evaluation

A full ENT examination was performed on all 70 individuals reported in the 2 clinical series (Tables 3 and 4). This included otoscopy or micro‐otoscopy (as indicated). Auscultation of the adjacent peri‐auricular and cervical regions was performed in cases of perceived pulsatile tinnitus. Clinical confirmation of sound sensitivity and/or hyperacusis was accomplished with the loudness comparison method using a standard set of aluminum‐alloy tuning forks presented equally to the 2 ears. The tuning forks were presented with the vibratory dipole approximately 3 cm from the ear canal. Individuals were queried as to perceived intensity and visually assessed for a lateralizing physiologic response to the tuning forks. The triggering of squinting, facial contracture, or aversive recoil from the stimulus was felt to be a reliable marker for sound intolerance and supported a conclusion of either hyperacusis or occasionally misophonia. Further assessment involved testing for a positive Tullio response (elicited by a Barany box) and, in specific cases, a positive fistula test utilizing a Siegel pneumatic aural speculum.

Table 3.

Initial Series of Patients Identified With Facial Canal–Lateral Canal Dehiscences

Age	Sex	Reason	Initial presentation	Radiologic findings
49	F	R Ménières', L hearing loss, B ear fullness, pulsatile tinnitus L	Fluctuant hearing R ear, episodic vertigo, tinnitus, imbalance	B FC‐LCD B C‐F, R PSCCD, L SSCCD
59	M	R Ménières', severe R SNHL, L mild SNHL, dizziness	Sudden hearing loss R and then hydrops	B FC‐LCD (R > L)
4	F	Severe/profound L mixed loss	L hearing loss, absent OAEs	B FC‐LCD attenuated SSCC L
87	F	Vertigo with tympanometry R, head hollowness	Imbalance, intolerance to HAs, sound sensitivity, pulsatile tinnitus	B FC‐LCD
66	M	Ear fullness, possible eustachian tube dysfunction	Ear fullness, possible eustachian tube dysfunction	L FC‐LCD B C‐F
32	M	R SNHL	R mixed HL	B FC‐LCD R C‐F
44	M	Ménières' L, hyperacusis L	L tinnitus, L SNHL, vertigo, presumed Ménières'	L FC‐LCD B C‐F, attenuated R PSCC
16	F	Tinnitus L	L tinnitus, history of MX/T	L FC‐LCD
80	F	Dizziness	Dizziness	B FC‐LCD
49	M	Ear fullness, popping, dizziness	Dizziness, ear‐popping	B FC‐LCD
45	F	R asymmetric SNHL, vertigo	Vertigo, HL, ringing	B FC‐LCD
30	F	Tinnitus, echoing	Dizziness, mixed HL, post‐MVA	B FC‐LCD B C‐F
64	F	Tinnitus, echoing, L hyperacusis	Tinnitus, echoing, L hyperacusis	B FC‐LCD B C‐F
28	M	R mixed loss, L SNHL	Pulsatile tinnitus, mixed HL	R FC‐LCD
31	F	Dizziness, ear fullness, asymmetric B mixed hearing loss	Vertigo, B mixed hearing loss	B FC‐LCD B SSCCD
20	F	B hyperacusis, pain	Hyperacusis, pulsatile tinnitus	B FC‐LCD
52	F	Dizziness, tinnitus	SNHL, L Meniere's, mixed hearing loss L ear, B hearing changes, B hyperacusis	L FC‐LCD
61	M	Tinnitus, hearing loss	Asymmetric progressive HL	B FC‐LCD B C‐F
63	M	Pulsatile tinnitus	Pulsatile tinnitus	B FC‐LCD
55	M	Imbalance, dizziness, fluttering in L ear	Exertional vertigo, sound sensitivity	B FC‐LCD
72	M	Hearing loss	Hearing loss, hyperacusis	B FC‐LCD
61	M	Fluctuant R SNHL, dizziness	Vertigo, R HL	B FC‐LCD
30	F	Dizziness, hyperacusis, autophony L ear	Pulsatile tinnitus, hearing breathing L ear, hyperacusis, dizziness	B FC‐LCD B C‐F
68	M	Dizziness, recurrent BPPV, mixed HL	Dizziness	B FC‐LCD B C‐F
59	M	Hyperacusis, R mixed HL	L mixed HL, R SNHL	L FC‐LCD
10	M	Hearing loss	Hearing loss	B FC‐LCD B C‐F R cholesteatoma
5	M	Cholesteatoma L	Hearing loss	R FC‐LCD s/p L mastoid
48	F	B pulsatile tinnitus, hyperacusis	B Pulsatile tinnitus, hyperacusis	R FC‐LCD B CF
67	M	Dizziness	Dizziness	R FC‐LCD L SSCCD
47	F	Dizziness, ear fluttering	Dizziness	R FC‐LCD B C‐F
43	F	Dizziness, pulsatile tinnitus	Dizziness, pulsatile tinnitus	L FC‐LCD
13	M	Dizziness	Dizziness, Tullio	R FC‐LCD
32	M	Hearing loss R, tinnitus	Fluctuant R SNHL	B FC‐LCD B SSCCD
15	M	Hyperacusis	Sound sensitivity	L FC‐LCD

Abbreviations: B, bilateral; BPPV, benign paroxysmal positional vertigo; C‐F, cochlear–facial dehiscence; FC‐LCD, facial canal–lateral canal dehiscence; HL, hearing loss; L, left; MVA, motor vehicle accident; PSCCD, posterior semicircular canal dehiscence; R, right; SNHL, sensorineural hearing loss; SSCCD, superior semicircular canal dehiscence.

Table 4.

Characteristics of Patients With Third Window Symptoms With Only FC‐LSCC Dehiscence

Sex	DOB (mm/dd/yyyy)	Age (y)	Ear(s)	Size (mm) coronal plane	Hyperacusis	Pulsatile tinnitus	Dizziness or vertigo	Spatial imbalance	Fluctuant hearing	Hearing Loss	Precipitant Event	Tullio	Fistula
F	4/26/1989	33	L	0.5 L	Yes	Yes	Yes	No	No	Mixed HL L	Concussion	No	No
F	6/12/1958	64	L	0.8 L	No	No	Yes	Yes	No	Mixed HL L	None	Yes	‐
F	4/3/1975	47	R	0.6 R	Yes	No	No	No	Yes	Normal HL	None	No	‐
F	10/8/1990	32	B	0.8 R, 1.0 L	No	No	Yes	No	No	Mixed HL L	None	Yes	‐
F	12/7/1951	71	L	1.1 L	Yes	No	No	No	No	No	None	‐	‐
F	7/2/1974	48	B	1.7 R, 1.2 L	Yes	No	Yes	Yes	Yes	No	None	Yes	‐
F	5/23/1967	55	L	0.8 L	Yes	No	No	No	No	Yes	Sudden HL L	No	‐
F	6/17/1978	43	L	0.6 L	Yes	Yes	Yes	No	Yes	No	None	‐	‐
F	1/6/1978	45	B	1.0 R, 0.8 L	Yes, R > L	Yes	Yes	Yes	No	Mixed symmetric B hearing loss	None	Yes	No
F	9/14/1978	42	B	1.1 R, 0.9 L	Yes, L > R	No	No	No	No	Sloping L SNHL	None	‐	‐
F	3/16/1982	40	L	0.6 L	Yes	Yes	Yes	No	No	B mild SNHL	None	Yes, L	‐
F	4/8/1975	47	B	0.8 R, 0.6 L	Yes	No	Yes	No	Yes	Severe fluctuant R mixed HL	None	No	No
F	12/7/1989	33	B	0.8 R, 1.1 L	Yes	No	Yes	No	No	Normal hearing	None	‐	‐
F	11/21/1950	72	B	0.9 R, 0.4 L	Yes, L	No	Yes	No	No	B sloping mixed mild‐to‐severe H, worse on the L side	None	Yes, L	No
F	2/17/1992	30	L	1.0 L	Yes	Yes	No	Yes	Yes	Low‐frequency mild CHL, otherwise normal	None	‐	‐
F	4/24/1979	43	L	1.0 L	Yes	No	Yes	‐	Yes	Very mild mixed hearing loss with low‐frequency wedge	None	‐	‐
F	10/12/1938	83	L	0.9 L	Yes	‐	Yes	‐	No	Mild‐to‐severe SNHL	None	Yes	‐
F	8/7/1958	64	B	1.0 R, 0.8 L	Yes	No	Yes	Yes	No	Profound R‐sided hearing loss	None	Yes	No
F	4/18/1969	53	L	0.9 L	Yes	Yes	Yes	No	No	Mid‐frequency (1000 Hz) SNHL L ear	No	Yes	No
M	2/13/1937	85	B	0.6 R, 0.7 L	No	No	No	No	No	Mild to profound SNHL R, severe to profound SNHL L	None	‐	No
M	9/29/19990	16	L	0.8 L	No	No	No	No	No	Mixed HL L	None	No	No
M	12/24/1962	60	R	1.0 R	Yes	No	Yes	Yes	Yes	Yes	none	Yes	‐
M	1/17/1958	65	B	0.7 R, 1.0 L	No	Yes	No	No	No	Symmetrical sloping mild to moderately severe SNHL	Hypertensive crisis	‐	‐
M	10/18/1963	59	L	0.7 L	No	No	No	No	No	Moderate SNHL L	None	‐	‐
M	4/29/1971	51	B	1.2 x R, 1.2 L	Yes, R	Yes	Yes	No	No	Asymmetric high‐frequency mild SNHL L	None	Yes, L	‐
M	2/12/1949	73	R	0.5 R	No	No	No	No	No	Asymmetric sloping severe B SNHL	Vigorous hike	‐	‐
M	10/20/2006	16	B	1.4 R, 1.0 L	Yes	No	Yes	No	No	No	None	‐	‐
M	7/2/2006	16	L	1.1 L	Yes	No	No	No	No	No	Concussions	‐	‐
M	9/6/1972	48	B	0.9 R, 0.8 L	No	No	No	No	No	Symmetric mild low‐frequency mixed HL	Neck injury	‐	‐
M	12/17/1959	61	R	0.9. R	Yes	No	Yes	No	Yes	Asymmetric mild R SNHL, fluctuant	None	Yes	No
M	3/1/1978	44	L	0.6 L	Yes	No	No	No	No	Low‐frequency mild mixed hearing loss	None	No	‐
M	11/9/1972	50	B	1.1 R, 0.7 L	No	No	No	No	Yes	Asymmetric L > R SNHL, mixed hearing noted previously	None	No	No
M	12/22/1953	69	R	0.7 R	No	No	Yes	Yes	No	Moderate‐to‐severe mixed hearing loss R	None	No	No
M	10/11/1953	69	L	0.8 L	No	No	Yes	‐	Yes	Moderately severe L SNHL	None	No	‐
M	8/25/1951	71	L	0.4 L	Yes	No	No	No	No	Yes, sloping mild‐to‐severe mixed hearing loss with low‐frequency conductive wedge	Pistol fired closed to ear	‐	‐
M	11/17/1969	53	L	0.5 L	Yes	No	No	No	Yes	Normal to sloping mild SNHL L ear	Exposure to power tools	‐	‐

Abbreviations: B, bilateral; HL, hearing loss; L, left; R, right; SNHL, sensorineural hearing loss.

Audiovestibular Testing

Conventional pure‐tone threshold audiometry was conducted by licensed clinical audiologists using the standard Hughson–Westlake threshold technique, following ASLHA guidelines. Testing was conducted in standard acoustic booths using either an Interacoustic AC40 Clinical Audiometer or a GSI Audiostar Pro Clinical Audiometer and TDH 50 Audiometric or Sennheiser HD 300 Professional Headphones or insert headphones. Acoustic reflexes and bone‐ and air‐conduction pure‐tone thresholds down to 250 Hz were routinely acquired.

Clinical fistula testing was performed with a Siegel aural pneumatoscope. We attempted to obtain vestibular evoked myogenic potentials (cervical and ocular) as another means of objective verification. We could not elicit consistent differentiating responses in amplitudes or latencies at 500 Hz or 4000 Hz.

Imaging

All the scans were acquired on GE BrightSpeed 16‐slice CT scanners with Adaptive Statistical Iterative Reconstruction. Images were acquired from below the mastoid bone to 1.0 cm above the petrous bone in the helical mode. The field of view (FOV) was 25 cm with a slice thickness of 0.625 mm at a 0.625 mm interval. The images were reconstructed for IAC axial planes with 0.625 mm thickness, 0.625 mm interval, and 9.6 cm FOV using the bone algorithm. Bilateral reconstructions were performed with the following parameters: 0.625 mm slice thickness, 0.310 mm interval, 9.6 cm FOV, and a bone algorithm. Pöschl and Stenver reconstructions were also performed.

Each scan was evaluated and selected, and measurements were made by the senior author. All scans were acquired with magnified coronal, Pöschl, and Stenvers reformations through the temporal bone. Both standard and gray‐scale inversion image formats were used, via a Lexmark Image viewer, to identify the narrowest distance between the membranous elements of the horizontal semicircular canal and that of the facial nerve along a vertical plane. Three measurements were made for each ear, 1 in each of the designated vertical plane views: coronal, Pӧschl, and Stenvers.

We identified potential FC‐LSCC dehiscences radiologically by assessing the narrowest interval distance between the bony lateral semicircular canal (LSCC) and the adjacent facial canal (FC) in the coronal plane. Image analysis was performed with PACS measurement software using Pixel Spacing within the DICOM header. The normal appearance of the partition between the LSCC and the most proximal segment of the adjacent FC is demonstrated in Figure 1. If an apparent “conduit” between the 2 canals was seen (Figure 2) in 2 consecutive slices, it was assessed further with the inversion imaging and then corroborated with Stenvers' and Pöschl's views. After the first author identified the series with the FC‐LSCC dehiscence, the second author (neuroradiologist) reviewed and assessed each exam to assess for interobserver reliability and concordance. Thus, this was not a blinded study.

Figure 1.

Image of the normal partition between the (A) lateral semicircular canal and (B) facial canal. The arrow is directed at the normal bony partition.

Figure 2.

Image of appearance of dehiscence between the (A) lateral semicircular canal and (B) facial canal. The arrow is directed at the region of dehiscence.

Histopathology

We measured the bony partition width between the LSCC and FC in 25 human temporal bones from the Temporal Bone Library at the University of Minnesota Twin Cities Otopathology lab (Figure 3). Each bone in the library had been previously cataloged with notes detailing medical history. Candidates' temporal bone samples were selected from the slide library for measurement. Images of the archival human temporal bone slides were acquired using a Nikon Eclipse microscope with a Nikon DS‐Fi3 camera attached for the most accurate and reproducible measurements. The images were then transferred to the NIS Elements software (Nikon). These were analyzed using the microscope at 4 to 10× magnification, with calibrations adjusted accordingly. The acquired images of the human temporal bones were subjected to serial measurements to determine the distances between the FC and the LSCC, reflecting the partitional width between the 2 entities (Figure 3). The measurements were made by identifying a line of closest proximity between the 2 structures and were made along several levels of the temporal bone, using approximately every 10th section, each 20 µm in thickness. The minimal thickness of the bony partition, reflecting the narrowest region between the LSCC and the facial nerve, was determined with several measurements in each of the selected temporal bones. These were acquired using the MEASURE feature in the NIS Elements software.

Figure 3.

Image demonstrating temporal bone measurement of the facial canal–lateral canal partition.

Results

Normative Data: Temporal Bones and CT Scans

We used temporal bone CT scans of 60 patients (male, n = 30; female, n = 30), without potential radiologic dehiscence between the FC and LSCC (controls) to determine the average width of the bony partition of the 2 canals. The average partition width was 0.6 mm ± 0.2 mm. An unpaired Student's t‐test analysis (through Excel) did not show differences in measurements of male versus female patients (P = .769) or between left and right ears, both for male (P = .382) or female (P = .850) patients. Measurements made for both sexes are listed in Table 1.

Table 1.

Temporal Bone CT Measurements (mm), Divided by Sexes

Age	Sex	R S	R C	R P	L S	L C	L P	Findings	Reason
59	F	0.6	0.5	0.4	0.6	0.6	0.4	L C‐F, L SSCCD	Hyperacusis
70	F	0.3	0.4	0.6	0.4	0.6	0.7	None	Dizziness
60	F	0.6	0.6	0.7	0.5	0.7	0.7	None	L hearing loss, pulsatile tinnitus, Ménières'
56	F	0.6	0.8	0.6	0.7	0.8	0.7	B C‐F	Vertigo, pressure L
55	F	0.5	0.4	0.5	0.7	0.6	0.6	B C‐F	Hyperacusis L
67	F	0.9	0.8	0.7	0.7	0.7	0.5	R EVA, cochlea dysplasia	R SNHL, tinnitus, fullness
63	F	0.4	0.5	0.5	0.6	0.4	0.5	None	L tinnitus, dizziness
44	F	0.4	0.5	0.4	0.6	0.5	0.3	L C‐F, B SSCCD	Vertigo, tinnitus
63	F	0.4	0.6	0.3	0.5	0.6	0.5	B SSCCD	Dizziness, hyperacusis
65	F	0.7	0.8	0.7	0.9	1	1	Thinning of L SSCC	Hearing sensitivity, popping L
73	F	0.7	0.6	0.7	0.8	0.8	0.7	TORP L ear	L hearing loss
68	F	0.8	0.8	0.9	0.9	0.9	1	None	Otitis externa L ear
47	F	0.4	0.5	0.4	0.5	0.5	0.3	B SSCCD, L C‐F	Ménières'
41	F	0.7	0.5	0.6	0.5	0.5	0.6	B C‐F	R SNHL, Ménières'
71	F	0.9	0.7	0.6	0.4	0.6	0.5	B C‐F	B hearing loss
43	F	0.6	0.6	0.6	0.4	0.5	0.5	B SSCCD	L hyperacusis, hearing loss
86	F	0.4	0.5	0.6	0.5	0.5	0.5	None	Dizziness, SNHL, dizziness
38	F	0.6	0.6	0.5	0.5	0.5	0.5	L cochlear‐carotid dehiscence	Tinnitus, dizziness, Hearing loss
44	F	0.6	0.6	0.5	0.4	0.5	0.5	B SSCCD, B C‐F	Conductive hearing loss, pulsatile tinnitus
66	F	0.9	1	0.8	0.8	0.9	1	B SSCCD	Dizziness
61	F	0.5	0.4	0.5	0.8	0.7	0.8	L C‐F	Left mixed hearing loss, vertigo
90	F	0.7	0.7	0.6	0.6	0.5	0.4	None	SNHL, tinnitus
75	F	0.7	0.7	0.6	0.5	0.4	0.5	B C‐F, L SSCCD, L OCR	B mixed hearing loss
50	F	0.7	0.7	0.7	0.8	0.8	0.6	L C‐F	R SNHL
43	F	0.8	0.8	0.6	0.6	0.5	0.4	L glomus tympanicum	Glomus
44	F	0.5	0.7	0.5	0.6	0.5	0.5	B SSCCD	L CHL
61	F	0.6	0.6	0.6	0.7	0.7	0.5	B C‐F	L tinnitus, fullness
18	F	0.4	0.3	0.4	0.5	0.4	0.4	B C‐F, L retracted TM	R SNHL, L CHL
40	F	0.6	0.6	0.5	0.4	0.4	0.4	R C‐F	R Ménières'
14	M	0.7	0.8	0.7	0.8	0.8	0.9	None	L SNHL
77	M	0.9	0.8	0.8	0.8	0.9	0.8	None	Mixed HL L
33	M	0.5	0.6	0.6	0.6	0.6	0.6	B SSCCD, R PSCCD, B C‐F, cholesteatoma	Cholesteatoma, mixed loss R ear
62	M	0.5	0.4	0.4	0.5	0.4	0.4	L mastoid (ELS)	Ménières' disease L ear, h/o ELS surgery
61	M	0.5	0.5	0.4	0.5	0.6	0.5	R CWU mastoid, TORP	CHL R, h/o mastoid surgery
29	M	0.7	0.9	0.6	0.8	0.6	0.8	Oblique R TB fracture, scar of inferior canal	Dizziness, h/o TB fracture
66	M	0.4	0.3	0.4	0.5	0.5	0.5	B fenestral otosclerosis	Mixed hearing loss
73	M	0.6	0.6	0.6	0.6	0.5	0.5	B C‐F	L hearing loss and tinnitus
59	M	0.5	0.6	0.6	0.6	0.5	0.5	B CWD mastoid, B middle ear opacification	B mixed hearing loss
75	M	0.8	0.7	0.8	0.9	1	0.8	L mastoid and middle ear effusion, L C‐F	B mixed hearing loss
40	M	0.7	0.7	0.7	0.5	0.6	0.5	None	R vestibular dysfunction
21	M	0.7	0.6	0.6	0.6	0.7	0.6	Acquired EAC stenosis L	L EAC stenosis
30	M	0.8	0.7	0.8	1	1	1	None	L ear fullness
70	M	0.5	0.5	0.5	0.6	0.5	0.4	Soft tissue in L middle ear	History of L CWD tympanomastoidectomy
66	M	0.7	0.6	0.5	0.5	0.6	0.6	None	L‐sided tinnitus
58	M	0.3	0.3	0.3	0.4	0.4	0.5	High jugular bulb	L low frequency SNHL
84	M	0.7	0.6	0.6	0.7	0.7	0.7	None	R SNHL
40	M	0.6	0.6	0.4	0.6	0.5	0.6	B C‐F	L hearing loss
74	M	0.7	0.6	0.6	0.7	0.7	0.5	None	s/p L stapes
55	M	0.8	0.4	0.5	0.7	0.6	0.7	None	R tinnitus, PLF
23	M	0.4	0.4	0.4	0.3	0.4	0.4	B SSCCD	B CHL
39	M	0.6	0.5	0.4	0.6	0.4	0.4	L C‐F	L hearing loss
44	M	0.5	0.5	0.5	0.5	0.6	0.4	R LSCC fistula	Dizziness, R hearing loss
13	M	0.9	0.8	0.8	0.8	0.8	0.7	B C‐F	B CHL
13	M	0.7	0.7	0.5	0.5	0.5	0.4	B C‐F	SNHL, seizures
15	M	0.5	0.7	0.8	0.9	1.2	1.3	Middle ear opacification	R SNHL
24	M	0.6	0.6	0.7	0.5	0.5	0.6	PORP L	B mixed hearing loss
41	M	0.4	0.4	0.4	0.4	0.4	0.4	B C‐F, L SSCCD	Dizziness, vertigo, L pulsatile tinnitus
54	M	0.6	0.5	0.5	0.6	0.6	0.4	Cholesteatoma	L ear pain
50	M	0.4	0.3	0.4	0.4	0.3	0.5	Exostoses	Exostoses
29	M	0.8	0.9	0.8	0.8	0.6	0.7	Stapes prosthesis L	Conductive hearing loss L

Abbreviations: B, bilateral; C‐F, cochlear–facial dehiscence; CHL, conductive hearing loss; CWD, canal wall down; EAC, external auditory canal; ELS, endolymphatic sacculotomy; EVA, enlarged vestibular aqueduct; L, left; OCR, ossicular chain reconstruction; PLF, perilymphatic fistula; PORP, partial ossicular reconstruction prosthesis; PSCCD, posterior semicircular canal dehiscence; R, right; SNHL, sensorineural hearing loss; SSCCD, superior semicircular canal dehiscence; TB, temporal bone; TORP, total ossicular reconstruction prosthesis.

Of the 25 temporal bones evaluated in the “Histopathology” section, 13 were from women (average 50.3 years) and 12 were from men (average age 54.8 years). All were Caucasian. The average minimum bony partition width for all temporal bones was 0.560 mm ± 0.171 mm. There was no statistical difference (Student's t‐test) between sides within the sexes, nor overall. However, comparing right temporal bone widths between sexes (M = 486 mm ± 153 mm, F = 629 mm ± 168 mm), there was a statistical difference, with P = .047, favoring a wider partition for females on that side (Table 2).

Table 2.

Morphometric Temporal Bone Partition Widths for Sexes

Partition	M (♂) L temporal bone	M (♂) R temporal bone	F (♀) L temporal bone	F (♀) R temporal bone
Average width	542 mm ± 157 mm	486 mm ± 153 mm	583.2 mm ± 204 mm	629.0 mm ± 168 mm
	Student's t‐test P = .388		Student's t‐test P = .613

Average width M	514 mm ± 154 mm	Student's t‐test P = .070
Average width F	601 mm ± 183 mm

Clinical Cases

We first identified 34 patients with third window features not limited to the FC‐LSCC region within the otology–neurotology practice. Compared to other third window pathologies, hearing loss, dizziness/vertigo, and tinnitus were the most frequently reported symptoms. Of the patients, 15 (44%) had more than 1 site of bone dehiscence, and 21 (62%) had bilateral FC‐LSCC dehiscence. The most frequent initial clinical complaints were hearing loss (44%), dizziness or vertigo (41%), tinnitus (41%), and hyperacusis or sound sensitivity (24%). In the subgroup of patients who only had the FC‐LSCC dehiscence (n = 18; 53%), hearing loss was identified in 44.4% (n = 8), dizziness and/or vertigo in 44.4% (n = 8), sound sensitivity and/or hyperacusis in 44.4% (n = 8), and tinnitus in 38.8% (n = 7). Analysis of the prevalence of hearing loss (P = .96), dizziness or vertigo (P = .68), and tinnitus (P = .77) in the groups with isolated FC‐LSCC dehiscence versus ones with multiple sites of dehiscence did not reveal significant differences between these 2 groups by Student's t‐test. However, as an initial complaint, sound sensitivity and/or hyperacusis in the isolated FC‐LSCC dehiscence group was significantly higher than in patients with multiple dehiscence sites (P = .01; Table 3). Figure 2 shows the characteristic radiologic features of FC‐LSCC.

We then sought to better characterize various FC‐LSCC features in a second group of 36 patients with clinical signs and symptoms suggestive of third window syndrome but with only radiologic FC‐LSCC dehiscence(s). The demographic information, initial diagnoses, clinical signs and symptoms, and radiologic findings are listed in Table 4. In this population, we found 69.4% (n = 25) with sound sensitivity and/or hyperacusis, dizziness or vertigo in 55.5% (n = 20), pulsatile tinnitus in 22.2% (n = 8), and spatial imbalance (sway instability) in 19.4% (n = 7).

The 2 lead authors independently reviewed temporal bone CT scans from the second cohort of 36 patients (those only potentially having FC‐LSCC dehiscences). If there did not exist a defined visual separation between the 2 entities on 2 contiguous temporal bone slices viewed in the coronal plane, then that was felt to reflect radiologic dehiscence. Using specified criteria for the width of the FC and lateral canal partition, we looked at determining inter‐rater reliability between the 2 readers as to the existence of the FC‐LSCC dehiscence in a population who had third window symptoms but no other regions of dehiscence. Utilizing this criterion, they concluded that there was dehiscence in this region for all 36 subjects, yielding a 100% inter‐rater reliability. As these individuals were identified by the senior author based on clinical features, which prompted the CT scan, the second author was not truly blinded.

Discussion

Since Minor et al reported and characterized superior semicircular canal pathophysiology, ¹ several authors have investigated and identified sites of dehiscence within the bony labyrinth. Such dehiscence sites lead to what is known as third window abnormalities. Third windows can present a myriad of signs, symptoms, and clinical findings. Conductive, mixed, and sensorineural hearing loss patterns have been identified in this family of otic capsule defects. Some may have episodic vertigo or a sense of imbalance, while others may manifest vestibular weakness or inducible vertigo with pressure or sound. They may have tinnitus, either high‐pitched or pulsatile, and frequently manifest features of autophony. Often the patients are concluded to have Ménières' disease, only to have a third window pathology be identified later. Until the recent publication of Gianoli et al, ² a series detailing patients with an FC‐LSCC dehiscence had not been reported. The inclusion of the FC‐LSCC dehiscence to the family of third window phenomena is not surprising given that the tympanic segment of the FC has the highest rate of dehiscence, from 0.5% to 76.3%, ¹¹ , ¹² , ¹³ and the LSCC, which shares anatomic proximity, is the most common site of inner ear anomalies. ¹⁴ Furthermore, the LSCC is also the last of the semicircular canals to complete ossification. ¹⁵ Building upon Gianoli et al's work, ² we set out to prospectively look for the lesions, provide normative temporal bone and radiologic data, and add to the number of individuals with this new entity. We set out to better characterize the FC‐LSCC partition by combining radiologic CT identification and measurement with temporal bone histopathology, allowing us to characterize the normal anatomy and ultimately provide guidelines for defining a unique third window location. Our CT‐based radiologic characterization of the FC‐LSCC partition demonstrated the normal range of thickness to be 0.6 mm ± 0.2 mm and was found consistent for both men and women and laterality was not a determinant. Based on our findings acquired from the histopathology, the normal range of partition thickness was nearly identical, given subtle differences in measurement capabilities and tolerances. The results from the 2 evaluations have established a region to be carefully assessed when evaluating individuals presenting with potential third window symptoms, unexplained hearing loss, and dizziness.

Once the characteristics of the region's anatomy were established, we looked to identify patients who had an FC‐LSCC dehiscence. Visual identification was enhanced using the gray‐scan inversion imaging to assess the CT scan, as noted by Schwartz et al. ¹⁶ Similar to Gianoli et al, ² where 9 of the 16 patients (56%) in their case series had other otic capsule dehiscences, 16 of our 34 patients (44%) had other recognized third window pathologies of the otic capsule. We also identified 21 (62%) of the patients had bilateral FC‐LSCC dehiscences. This suggests a common process of abnormal embryogenesis. The embryogenesis of the fallopian canal is not dependent upon the ossification of the otic capsule. Through the first 16 weeks of gestation, Declau et al ¹⁷ have shown that the primitive cartilaginous otic capsule perichondrium splits around the facial nerve and that a complex process of ossification occurs, which is dependent upon nerve ensheathment of the fibrous layer and not determined by the “simple” ossification of the otic capsule. The tympanic segment is the last of the sections to be ossified. ¹⁸ Spector and Ge ¹¹ described 2 separate ossification centers that, starting at 21 weeks gestation, complete the ossification of this segment. Ossification is completed at around 40 weeks gestation. They describe a superior “shelf,” comprising 75% of the bony canal, meeting up with an inferior ledge to complete the process. They conclude that all patterns of FC dehiscence can be explained by the failure of complete ossification or by malfusion.

Others ¹⁷ have studied the embryogenesis of the LSCC. Of the 3 canals, the lateral canal arises later and progresses at a slower rate than either of the other 2. ¹⁷ They showed the LSCC doesn't complete ossification until around 21 to 23 weeks gestation. They also found that the “bone formation around the canals tends to progress from around the vestibule towards the canal vertices and that the last region of the labyrinth to be surrounded is the posterolateral part of the lateral canal.” ¹⁷ Thus, the bony encasement of the LSCC is normally completed several weeks before that of the adjacent FC. Jeffery and Spoor ¹⁹ provide multiple linear measurements of the labyrinth but do not describe the growth of the lateral canal relative to its neighboring FC or the thickness of enveloping bone in this region. Kozerska et al ²⁰ reported uneven bone mineral distribution along the fallopian with advanced volumetric 3D visualization and concluded these regions “may mimic dehiscences.” But in their series, all were described along the lateral FC and not along the inferior aspect, which opposes the proximal segment of the LSCC.

Our reported symptoms agree with Gianoli et al. ² We found sound sensitivity to be the most common auditory complaint, with nearly 70% exhibiting this feature. Nearly 55% of them had vestibular symptoms. Tullio's and Ewald's research provides the basis for understanding the vestibular responses of third windows. ²¹ , ²² Spasic et al reported that “a dehiscence acquired in any of the semicircular canals may evoke various auditory symptoms or vestibular symptoms by creating a ‘3rd mobile window’ in the bone that enables aberrant communication between the inner ear and nearby structures.” ¹⁸ , ²³ Spasic et al ²³ reported that “a dehiscence acquired in any of the semicircular canals may evoke various auditory symptoms or vestibular symptoms by creating a ‘3rd mobile window’ in the bone that enables aberrant communication between the inner ear and nearby structures.” Given the location, it is plausible that there may be both auditory and vestibular complaints. However, unlike direct exposure of the membranous canal in SSCCD, the intraosseous dehiscence identified with the FC‐LSCC would be expected to be less impactful. The facial nerve only occupies approximately 50% of the cross‐sectional area of the tympanic segment of the FC. The juxtaposed horizontal SSC and tympanic segment of the FC are typically separated by a thin bony partition. As with other SSC dehiscences, a defect in this partition may generate third‐window symptomatology. Compared to other SSC dehiscences, this region of incompetent partitioning remains contained, albeit within the perilymphatic space. Thus, it is surmised that it would be less likely impacted by external pressures and thereby rarely expected to have a positive Hennebert's sign.

Although it would be expected that more than 1 site of dehiscence would be more symptomatic, we did not find significant differences compared with patients with isolated FC‐LSCC dehiscence. The prevalence of sound sensitivity was significantly higher in the group with isolated FC‐LSCC dehiscence compared to those patients with more than 1 site of dehiscence. We postulate that the isolated FC‐LSCC dehiscence may not result in a significant modification of the cochlear biomechanics and that the multiple sites of dehiscence effectively siphon the auditory perturbation away from the cochlea, thereby diminishing the auditory stimulation and sound sensitivity.

Our study has several limitations. Our findings reflect an examination of an exclusively Caucasian population and may not be found similarly in non‐Caucasian individuals. More importantly, it lacks an assessment of vertical sections of temporal bone specimens, which prevented a more direct analysis between our clinical and retrospective histopathological studies. We realize that there exists a potential selection bias based upon using clinical audiology features to implicate a potential third window, which could contribute to an over‐read of an FC‐LSCC dehiscence, even though it may have been more valid to conclude a region of “marked attenuation.” Although it is recognized that osteoprogerin‐mediated inhibition of otic capsular bone remodeling is associated with age‐related otic capsule bone degeneration, we did not look to identify an age‐related process. Instead, we focused on an anatomical descriptor as the primary goal. ²⁴

Conclusion

Our study provides not only normative data on the bony partition between the FC‐LSCC but also data on the characteristics of this unique abnormality and a comprehensive analysis of its clinical presentation. Our results show that dehiscence between the FC‐LSCC should be included in the differential diagnosis of patients presenting with third window symptoms.

Author Contributions

William J. Garvis, study conception and design, data collection, analysis and interpretation of the results, manuscript preparation, and supervision; Blake A. Johnson, analytical methods verification and manuscript preparation; Katherine E. Kluesner, data collection and recording, performed measurements, and manuscript editing; Stephanie M. Garvis, data collection and recording, performed measurements, and manuscript editing.

Disclosures

Competing interests

None.

Funding source

None.