Acta Oto-Laryngologica. 2014; 134: 666–671

ORIGINAL ARTICLE

Noninvasive evaluation of the effect of endolymphatic sac decompression in Ménière’s disease using magnetic resonance imaging

FANG LIU1, WEINING HUANG1, QINGHUA CHEN2, XIXI MENG3, ZHENCHANG WANG2 & YUXIA HE1

1Department of Otolaryngology, Beijing Hospital, Beijing, 2Department of Radiology, Beijing Tongren Hospital, Capital Medical University, Beijing and and 3Department of Otolaryngology, Beijing Institute of Otolaryngology, Beijing, China

Abstract Conclusions: This study is the first to demonstrate noninvasive evaluation of the effect of endolymphatic sac decompression (ESD) in Ménière’s disease using magnetic resonance imaging (MRI). Objective: To evaluate the effect of ESD for the treatment of Ménière’s disease by applying noninvasive intratympanic gadolinium (Gd) perfusion through the and three-dimensional fluid-attenuated inversion MRI (3D-FLAIR MRI). Methods: This was a prospective study. 3D-FLAIR MRI was performed with a 3 Tesla unit 24 h after intratympanic administration of Gd through the eustachian tube in five patients with intractable Ménière’s disease before and 3 months after ESD, with a 2-year follow-up on the effect of ESD. Results: Gd was present in the of the inner in all the patients, which clearly displayed the endolymphatic space on 3D-FLAIR MRI with a visible borderline between the perilymph and the . According to the normal values for the endolymphatic space, four of five patients had a ratio of more than 26% in the , and three of five patients had a ratio of more than 41% in the vestibule preoperatively. All the patients had a ratio of less than 26% in the cochlea and 41% in the For personal use only. vestibule postoperatively. ESD was effective in reducing the incidence and severity of vertigo attacks with significant improvement in 60% of patients.

Keywords: Gadolinium, 3D-FLAIR MRI

Introduction patients with Ménière’s disease fail to respond to medical management, surgical options are indicated Ménière’s disease appears to be a complex to stabilize the symptoms, which include endolym- disorder. The underlying increased pressure in the phatic sac surgery (ESS), three , the so-called endolymphatic occlusion, transtympanic gentamincin, labyrinthect- Acta Otolaryngol Downloaded from informahealthcare.com by Mrs Shirley Dempster on 08/25/14 hydrops, is associated with the classic features of omy, and vestibular neurectomy. Recently, ESS has spontaneous, episodic attacks of vertigo; fluctuating, become a favorable surgical option because it has a progressive sensorineural loss; tinnitus; and low surgical morbidity and it does not significantly aural fullness. In 1938, Hallpike and Cairns dem- impact on hearing. The main types of ESS involve onstrated the histological findings of ‘hydrops’ in endolymphatic sac decompression (ESD), endolym- patients with Ménière’s disease. Despite this well- phatic mastoid shunt (EMS), and endolymphatic sac known symptom complex, Ménière’sdisease excision (ESE), since it was first described by remains a controversial and often difficult disease Portmann in 1927. However, ESS remains contro- as regards determination of diagnosis, pathogenesis, versial and its effectiveness for the treatment of and optimal treatment [1,2]. Traditionally, when Ménière’s disease is debated [3].

Correspondence: Fang Liu, MD, Department of Otolaryngology, Beijing Hospital, No. 1 Dahua Road, Dongdan, Beijing 100730, China. E-mail: [email protected] and Qinghua Chen, MD, Department of Radiology, Beijing Tongren Hospital, Capital Medical University, No. 1 Dongjiaominxiang, Beijing 100730, China. E-mail: [email protected]

(Received 7 November 2013; accepted 9 January 2014) ISSN 0001-6489 print/ISSN 1651-2251 online Ó 2014 Informa Healthcare DOI: 10.3109/00016489.2014.885118 MRI evaluation of endolymphatic sac decompression in MD 667

Diagnosis by magnetic resonance imaging (MRI) tumor or other intracranial disease. (3) There was may play a key role in understanding Ménière’s dis- no history of diseases. (4) A test for ease without obtaining histopathologic perilymphatic fistula was performed. (5) Stage on specimens as in autopsy cases. In 2013, Uno et al. hearing was 2–3. The exclusion criteria were as reported that the changes in endolymphatic hydrops follows. (1) Patients were excluded if they had after sac surgery were examined by Gd-enhanced undergone previous surgical treatment for inner ear MRI [4]. However, the actual values for the endo- diseases, had any systemic disease requiring steroid lymphatic space were the ranges measured using therapy, had used diuretics or vasodilators within specimens of temporal bone [5,6]. There was still 2 weeks before entering the study, or had active an absence in the range of normal values of endolym- bilateral disease or had undergone any previous phatic space to allow diagnosis ‘endolymphatic hydro- destructive procedure (e.g. injections with gentami- ps’ and to accurately estimate the changes in hydrops cin). (2) Allergic conditions, a history of allergy to Gd, after sac surgery. As part of a series of related studies, and pregnancy also prohibited participation. in 2011 and 2012, Liu et al. devised a new method for noninvasive standard evaluation of normal endolym- Intratympanic Gd perfusion through eustachian tube and phatic space and endolymphatic hydrops using MRI MRI and first indicated the normal values for the endo- lymphatic space of the inner ear in healthy volunteers Standard intratympanic Gd perfusion through the [7,8]. We continued to pursue further studies in eustachian tube and MRI were performed as reported clinical subjects for the evaluation of the effect of previously [7,8]. ESD for the treatment of Ménière’s disease by apply- ing noninvasive intratympanic Gd perfusion through the eustachian tube and three-dimensional fluid- ESD attenuated inversion MRI (3D-FLAIR MRI). A simple mastoidectomy was performed under general anesthesia. The facial nerve, posterior semi- circular canal, and sigmoid sinus were identified. The Material and methods sigmoid sinus, the posterior fossa dura, and the endo- Patients lymphatic duct around the posterior semicircular canal were skeletonized. The endolymphatic sac For personal use only. A total of five patients with Ménière’s disease accord- was identified around Donaldson’s line and the exten- ing to the criteria of the 1995 AAO-HNS guidelines sion line from the . The sac itself [9], referred to the Department of Otolaryngology in and surrounding dura were then decompressed from Beijing Hospital to undergo ESD, were enrolled in the region of the otic capsule bone superiorly to this study. All the patients had failed on medical approximately 5 mm inferior to the sac. The surgeon management including a low-salt diet, diuretics, (F.L.) preferred ESD for all the patients with and vestibular suppressants. The patients’ demo- Ménière’s disease. graphics are shown in Table I. The selection criteria 3D-FLAIR MRI was performed with a 3 Tesla were as follows. (1) All the patients had a diagnosis of MRI unit at 24 h after intratympanic administration unilateral definite Ménière’s disease. (2) Brainstem of Gd through the eustachian tube. The procedure Acta Otolaryngol Downloaded from informahealthcare.com by Mrs Shirley Dempster on 08/25/14 audiometry was conducted and MRI or CT scans was carried out before ESD and 3 months after ESD. were also used to rule out cerebellopontine angle The endolymphatic space was evaluated using

Table I. Patient demographics, follow-up, PTA, and vertigo class.

PTA (dB HL) Follow-up Vertigo Patient no. Age (years) Sex Ear affected (months) Stage 6 months preop 18–24 months postop control class

1 45 M L 62 3 52 47 C 2 54 F L 58 3 65 64 B 3 52 F L 29 2 40 33 A 4 47 M R 24 3 47 50 C 5 53 F R 25 3 52 45 A Mean ± SD 50.2 ± 4.0 39.6 ± 18.8 51.2 ± 9.1 47.8 ± 11.1 Db, decibel; HL, hearing level; PTA, pure-tone average thresholds; SD, standard deviation. 668 F. Liu et al.

3D-FLAIR MRI, as described previously [7,8]. In the in the vestibule before ESD. All the patients had a basal turn of the cochlea, the sectional area of ratio of more than the normal value in the cochlea the endolymphatic space and the fluid space (sum and/or the vestibule. However, all the patients had a of the endolymphatic and perilymphatic space) was ratio of less than 26% in the cochlea and 41% in the measured and the ratio of the area of the endolym- vestibule 3 months after ESD. phatic space to that of the fluid space was evaluated on Second, among the five patients, vertigo control at images parallel to the cochlea. In the ves- 18–24 months was class A (two patients), class B (one tibule, the sectional area of the endolymphatic space patient), and class C (two patients) (Table I). For the (sum of the and ) and the fluid space purpose of this study, those patients who experienced (sum of the endolymphatic and perilymphatic spaces) class A or B results were considered to have a successful were measured and the ratio of the area of the outcome. Accordingly, the success rate was 60%. endolymphatic space to that of the fluid space was Third, of five patients, the average PTA before and evaluated on images parallel to the longitudinal axis of 2 years after ESD was 51.2 dB HL and 47.8 dB HL, the vestibule. respectively (Table I). According to AAO-HNS The severity of Ménière’s disease was evaluated by criteria, there was no significant change in PTA AAO-HNS criteria for hearing loss and vertigo [9]. The preoperatively and postoperatively [9]. worst pure-tone average (PTA) at 500 Hz, 1 kHz,2 kHz, Finally, no adverse effects of the intratympanic and 3 kHz in the 6-month period before ESD and the injection of Gd were observed. There were no serious period 18–24 months after ESD was compared. complications (such as cerebrospinal fluid leaks, facial Achangeof10dBHLormoreinPTAwasconsidered nerve palsy, meningitis, or wound infections) in any of clinically significant. Likewise, the average number of the patients. episodes of vertigo 6 months preoperatively and 24 months postoperatively was reported. Discussion The protocol of the study was approved by the Ethics Review Committee of Beijing Hospital. All ESS is widely performed to control vertigo spells and the patients gave their informed consent to participa- preserve inner ear functions in patients with intrac- tion in this study, in accordance with the suggestion of table Ménière’s disease, although there has been some the Ethics Review Committee. debate about the optional treatments (shunt, excision, and decompression) [10–12]. For personal use only. Data analysis The previous theory for EMS proposed that endo- lymph was passively drained into the mastoid from the All data were extracted from the subject charts and sac, which decompressed the hydropic inner ear. then displayed in tabular form. Statistical analysis was However, this seems irrational because of the rapid performed using the SPSS 17.0/PC program. overgrowth of the shunt by mucosa within days of Descriptive statistics were applied using mean ± SD. surgery. A histologic study of explanted Arenberg shunt found that acellular debris filled the vestibule Results around the valve in all cases and showed ingrowth of fibrous tissue to the sponge [13]. Moreover, Silastic First, on MRI scans taken 24 h after the intratympanic applied to the shunt was also coated in a fibrous scar,

Acta Otolaryngol Downloaded from informahealthcare.com by Mrs Shirley Dempster on 08/25/14 Gd injection through the eustachian tube in all the which limited the drainage potential of the biocom- patients before and 3 months after ESD, Gd was patible material [10]. In 1994, Gibson proposed that observed in the cochlea, the vestibule, and/or parts an abnormal build-up of fluid in the inner ear triggers of the semicircular canals. When the Gd entered the a response in the endolymphatic sac to secrete osmot- perilymph of the cochlea and the vestibule, the endo- ically active glycoproteins, creating a rapid drainage of lymphatic space without Gd could be seen clearly. An endolymph from the inner ear toward the sac, which example from a patient is shown in Figure 1. The causes the attacks of vertigo [14]. Based on the theory, endolymphatic space was evaluated using 3D-FLAIR destruction of function of the sac or occlusion of the MRI in all the patients (Table II). As reported earlier, duct may be the reason for improvement in vertigo the normal value for the endolymphatic space in the control. Therefore, elimination of the secretion of cochlea ranged between 8% and 26%, and that in the glycoproteins into the sac may control rapid fluid vestibule between 20% and 41% in healthy volunteers changes, which would decrease vertigo. In 1996, aged between 45 and 55 years [8]. According to the Welling et al. reported that EMS and ESE appear normal value for endolymphatic space, four of five equally effective in the treatment of Ménière’s disease. patients had a ratio of more than 26% in the cochlea, According to the results, it was presumed that the and three of five patients had a ratio of more than 41% removal of the sac would still allow hydrops MRI evaluation of endolymphatic sac decompression in MD 669

AB

2

1 1 2 3 3

CD

2

1 1 23

3

For personal use only. Figure 1. Oblique sagittal 3D-FLAIR MR images from a 52-year-old female patient (no. 3 in Table I) were taken 24 h after administration of gadolinium (Gd) before endolymphatic sac decompression (ESD) (A and B) and 3 months after ESD (C and D). (A and C) Gd appears as a hyperintense signal in the perilymphatic space of the basal, second, and apical turns of the cochlea while the endolymphatic space surrounded by Gd-filled perilymph appears as a hypointense signal. The large white ring 1 indicates the whole cochlea and the small white rings from 2 to 3 indicate the endolymphatic space in the basal turn of the cochlea. (B and D) Gd appears as a hyperintense signal in the perilymphatic space of the vestibule while the endolymphatic space surrounded by Gd-filled perilymph appears as a hypointense signal. (B) The large white ring 1 indicates the whole vestibule and the small white rings 2 and 3 indicate the enlarged utricle and saccule. (D) The large white ring 1 indicates the whole vestibule and the small white rings 2 and 3 indicate the utricle and the saccule, respectively.

formation, but there would not be rapid endolym- Some studies suggested that a functioning endo- phatic shifts to disturb the neuroepithelium, thereby lymphatic sac may have both resorptive and secretive

Acta Otolaryngol Downloaded from informahealthcare.com by Mrs Shirley Dempster on 08/25/14 preventing the episodic vertigo [10]. However, several properties, and has the ability to rapidly respond to a observations of the development of low-frequency pressure change. The highly osmolar proteins air–bone gaps, minimal deterioration of long-term secreted by the endolymphatic sac appear to occur hearing levels, after EMS and ESE, also suggest with rapid changes in the endolymphatic pressure but that inner ear micromechanics have been altered by not endolymphatic duct obstruction, which may cre- the surgeries, perhaps by creation of a third window ate an osmotic gradient to attract fluid into the phenomenon [14,15]. endolymphatic sac and duct from the surrounding In 2007, Brinson et al. reported that EMS and ESD tissues and also cause a fluid shift from the cochlea are equally effective in reducing the attacks of vertigo, and the vestibule. Some factors that may damage the with significant improvement in 67% and 66% of sac, such as infection, immunologic reaction, or patients, respectively. Moreover, the risk of postop- trauma, can upset the homeostatic function of the erative profound hearing loss is higher when the sac by disruption of the secretion or function of such endolymphatic sac is incised [12]. Although ESD is proteins, resulting in endolymphatic hydrops [16– an effective and safe procedure for patients with 18]. Two main theories were proposed as to the intractable Ménière’s disease, the rationale for the cause of vertigo attacks in endolymphatic hydrops. use of ESD remains controversial. The first, reported by Schuknecht, is the rupture of 670 F. Liu et al.

Reissner’s membranes due to overdistension of the 0.07

± endolymphatic compartment, which leads to a mix- ing of the endolymph and perilymph then causes acute vertigo attack [19]. Another theory proposed

1.56 0.31 by Shea is that the distension of the membranous

± labyrinth in the hydrops of inner ear causes neural

and saccule Ratio discharge, which resolves as the membrane distension Utricle subsides [20].

2.29 4.96 To the best of our knowledge, the results of this ± study are the first to show the effect of ESD on the endolymphatic space in Ménière’s disease by applying noninvasive intratympanic Gd perfusion through the

0.11 15.74 eustachian tube and 3D-FLAIR MRI. According to ± the normal values for the endolymphatic space, four of five patients had a ratio of more than 26% in the cochlea, and three of five patients had a ratio of more 1.86 0.41 than 41% in the vestibule before ESD. All the patients ± had a ratio of more than the normal value in the

Utricle and saccule Ratio Sum cochlea and/or the vestibule. However, all the patients had a ratio of less than 26% in the cochlea and 41% in 2.29 6.43

± the vestibule 3 months after ESD. The success rate of vertigo control was 60%. The data from the study indicate that ESD can influence the endolymph in the inner ear. ESD may impact on the endolymphatic 0.09 15.7 4

± pressure in the inner ear, changes in which make a functioning sac secrete the highly osmolar proteins, which may cause a fluid shift from the cochlea and the vestibule, and improve the resorption of endolym-

0.51 0.17 phatic hydrops, controlling attacks of vertigo. There- ± For personal use only. fore, ESD may play a role in controlling endolymphatic hydrops in the inner ear with a func- Endolymphatic space Ratio Sum tioning sac, and may stabilize the symptoms in

1.28 0.73 Ménière’s disease. ± There was no significant change in PTA between the values before ESD and 2 years after ESD. No adverse effects of the intratympanic injection of Gd Cochlea Vestibule 0.11 4.08

± were observed, moreover, there were no serious com- plications (such as cerebrospinal fluid leaks, facial nerve palsy, meningitis, or wound infections) in any

Acta Otolaryngol Downloaded from informahealthcare.com by Mrs Shirley Dempster on 08/25/14 of the patients [10–12]. 0.26 0.31 ± Preop Postop Preop Postop

Endolymphatic space Ratio Sum Conclusion

1.28 1.18 Noninvasive evaluation of the effect of ESD in ± Ménière’s disease using MRI was carried out for

Sum the first time. This is a stable and repeatable method to provide a preliminary basis to clinically determine the efficacy of ESD for endolymphatic hydrops in 4.0 4.08

± Ménière’s disease. Age (years) Sex

SD 50. Declaration of interest:

± The authors report no conflicts of interest. The authors alone are responsible Patient no. SD, standard deviation. 1234 455 54 MMean 52 F 47 4.00 F 53 3.53 M 3.42 1.08 F 6.30 1.07 3.13 1.64 0.27 1.10 0.30 1.01 4.00 0.48 0.18 3.53 0.33 3.42 0.11 6.30 0.81 3.13 0.55 0.03 1.51 0.23 0.66 0.16 15.90 0.24 14.82 0.21 7.79 16.58 12.60 4.17 0.49 18.82 8.79 5.30 0.28 15.90 6.10 0.53 0.42 14.82 5.70 0.33 16.58 12.60 3.71 18.82 0.36 3.65 4.41 0.25 7.34 0.22 0.35 0.39 Table II. Ratios of endolymphatic spaces before and 3 months after endolymphatic sac decompression (ESD). for the content and writing of the paper. MRI evaluation of endolymphatic sac decompression in MD 671

References Foundation, Inc. Otolaryngol Head Neck Surg 1995;113: 181–5. [1] Andrews JC, Bohmer A, Hoffman LF. The measurement and [10] Welling DB, Pasha R, Roth LJ, Barin K. The effect of manipulation of intralabyrinthine pressure in experimental endolymphatic sac excision in Meniere’s disease. Am J endolymphatic hydrops. Laryngoscope 1991;101:661–8. Otol 1996;17:278–82. [2] Warmerdam TJ, Schroder FH, Wit HP. Perilymphatic and [11] Durland WF, Pyle GM, Connor NP. Endolymphatic sac endolymphatic pressures during endolymphatic hydrops. decompression as a treatment for Meniere’s disease. Eur Arch Otorhinolaryngol 2003;260:9–11. Laryngoscope 2005;115:1454–7. [3] Sajjadi H, Paparella MM. Meniere’s disease. Lancet 2008; [12] Brinson GM, Chen DA, Arriaga MA. Endolymphatic mastoid 372:406–14. shunt versus endolymphatic sac decompression for Meniere’s [4] Uno A, Imai T, Watanabe Y, Tanaka H, Kitahara T, disease. Otolaryngol Head Neck Surg 2007;136:415–21. Horii A, et al. Changes in endolymphatic hydrops after sac [13] Cohen EJ, Mattox DE. Histologic and ultrastructural surgery examined by Gd-enhanced MRI. Acta Otolaryngol features of explanted Arenberg shunts. Arch Otolaryngol 2013;133:924–9. Head Neck Surg 1994;120:326–32. [5] Teranishi M, Yoshida T, Katayama N, Hayashi H, Otake H, [14] Gibson W. The effect of removal of the endolymphatic sac in Nakata S, et al. 3D computerized model of endolymphatic Meniere’s disease. Aust J Otolaryngol 1994;6:527–9. hydrops from specimens of temporal bone. Acta Otolaryngol [15] Kim SH, Ko SH, Ahn SH, Hong JM, Lee WS. Significance Suppl 2009;560:43–7. of the development of the inner ear third window effect after [6] Melhem ER, Shakir H, Bakthavachalam S, MacDonald CB, endolymphatic sac surgery in Meniere’s disease patients. Gira J, Caruthers SD, et al. Inner ear volumetric measure- Laryngoscope 2012;122:1838–43. ments using high-resolution 3D T2-weighted fast spin-echo [16] Rask-Andersen H, Danckwardt-Lilliestrom N, Linthicum FH, MR imaging: initial experience in healthy subjects. AJNR Am House WF. Ultrastructural evidence of a merocrine secretion J Neuroradiol 1998;19:1819–22. in the human endolymphatic sac. Ann Otol Rhinol Laryngol [7] Liu F, Huang WN, Wang ZC, Chen QH, Liu XH, Li SL, 1991;100:148–56. et al. Noninvasive evaluation of endolymphatic space in [17] Wackym PA, Friberg U, Bagger-Sjoback D, Linthicum FH. healthy volunteers using magnetic resonance imaging. Acta Rask-Andersen H. Human endolymphatic sac: possible Otolaryngol 2011;131:247–57. mechanisms of pressure regulation. J Laryngol Otol 1987; [8] Liu F, Huang WN, Meng XX, Wang ZC, Liu XH, 101:768–79. Chen QH. Comparison of noninvasive evaluation of endo- [18] Lee AJ, Parker DA, Gladstone HB, Hradek G, lymphatic hydrops in Meniere’s disease and endolymphatic Schindler RA. Effects of endolymphatic duct occlusion on space in healthy volunteers using magnetic resonance the adult guinea pig. Laryngoscope 1995;105:192–6. imaging. Acta Otolaryngol 2012;132:234–40. [19] Schuknecht HF. Correlation of pathology with symptoms of [9] Committee on Hearing and Equilibrium guidelines for the Meniere’s disease. Otolaryngol Clin North Am 1968;1:433. diagnosis and evaluation of therapy in Meniere’s disease, [20] Shea JJ. Classification of Meniere’s disease. Am J Otol 1993; American Academy of Otolaryngology-Head and Neck 14:224–9. For personal use only. Acta Otolaryngol Downloaded from informahealthcare.com by Mrs Shirley Dempster on 08/25/14