Abstract
Purpose
The purpose of this study was to evaluate the effectiveness on function preservation and tumor control of the treatment of glomus tympanicum tumors with pre-operative embolization followed by total surgical resection.
Material and methods
We describe a series of 6 patients with a glomus tympanicum tumor who were treated in our hospital using the same technique: the day before surgery selective tumor embolization due to denaturation with 96% ethanol. Following parameters were considered: tumor classification, tumor control, clinical and audiological outcome, effectiveness of embolization, percentage of tumor necrosis and treatment complications.
Results
There were no severe complications due to embolization or surgery. Tumor blush disappeared completely in 5 patients on DSA post embolization and histologic evaluation of the resected tissue showed a median of 69.2% of tumor necrosis. Pulsatile tinnitus disappeared in all patients and 3 patients had no symptoms at all. Hearing ameliorated in 4 patients, 1 patient without hearing loss pre- treatment still had normal hearing after treatment and 1 patient’s hearing was worse after treatment. Average follow-up was 21.3 months.
Conclusions
Treatment of glomus tympanicum tumors by pre-operative embolization with ethanol and surgical resection has not been described before. Our results show that it is a safe procedure with a good long term tumor control, good clinical and audiological outcome.
1
Introduction
Paragangliomas, also named glomus tumors or chemodectoma, are rare highly vascular neuroendocrine neoplasms arising from neurectodermally derived paraganglionic cells located in the walls of blood vessels or associated with specific nerves . They have an estimated incidence of 1 in 1.3 million people . Head and neck paragangliomas comprise 3% of all paragangliomas and their most common site is the carotid body, followed respectively by the tympanic plexus, jugular bulb and vagal nerve . Jugulotympanic paragangliomas represent the most common primary neoplasm of the middle ear . They are usually benign, slowly growing, painless tumors with the potential to remain stable over years . Their growth pattern is multidirectional along the tracts of least resistance. They grow primarily in the air filled spaces of the temporal bone. Vascular lumina, neurovascular foramina, and the Eustachian tube provide routes of extra-temporal extension, into the infratemporal fossa or along the skull base . Classifications according to Fisch ( Table 1 ) or Glasscock–Jackson are most used .
| Class A (glomus tympanicum) tumors arise along the tympanic plexus on the promontory of the middle ear. The blood supply is from the tympanic artery, a branch of the ascending pharyngeal artery. Class A tumors may produce minimal erosion of the promontory. |
| Class B (glomus hypotympanicum) tumors originate in the canalis tympanicus of the hypotympanum and invade the middle ear and mastoid. The carotid foramen and canal are intact. By definition, class B tumors invade the bone of the hypotympanum, but the cortical bone over the jugular bulb is intact. |
| Class C (glomus jugulare) tumors originate in the dome of the jugular bulb and destroy the overlying cortical bone. They may spread in the following directions: inferior, along the jugular vein and cranial nerves IX, X, XI, XII; posterior, into the sigmoid sinus; superior, toward the internal auditory canal and otic capsule; lateral, to the hypotympanum and middle ear; and medial, to the jugular foramen and cerebellopontine angle. Further classification is based on the degree of erosion of the carotid canal. |
| C1 tumors erode the carotid foramen, but do not invade the carotid artery. |
| C2 tumors destroy the vertical carotid canal between the carotid foramen to the carotid bend. |
| C3 tumors grow along the horizontal portion of the carotid artery but do not reach the foramen lacerum. |
| C4 tumors grow to the foramen lacerum and along the carotid artery to the cavernous sinus. |
| Class D tumors have intracranial extension. The extension may be extradural (De) or intradural (Di). |
* Microsurgery of the skull base, U Fisch and D Mattox 1988, p 149–153.
1.1
Diagnosis
In tympanic paragangliomas symptoms are produced at an early stage, i.e. pulsatile tinnitus and conductive hearing loss . In larger tumors, sensorineural hearing loss, ear pain, ear discharge, bleeding and dizziness often occur . They commonly present as a vascular middle ear mass. They may transgress the tympanic membrane and appear as an inflammatory polyp .
Clinical diagnosis is confirmed by imaging ( Fig. 1 ). HRCT images with bone window display are very sensitive for delineating the extent of bone destruction and are therefore essential for classification and surgical planning . Magnetic resonance imaging is sensitive for defining the soft tissue involvement. T1- and T2-weighted images after gadolinium enhancement and/or fat-suppression sequences are used. On T1-weighted images, paragangliomas have a typical ‘salt-and-pepper’ appearance . MR angiographic images can be created, with identification of the tumor’s feeding vessels (most commonly the inferior tympanic branch of the ascending pharyngeal artery ) and the venous flow. DSA provides an arterial ‘map’ and flow dynamics of the blood supply and it is a sensitive diagnostic study for detecting multiple paraganglioma . Because of its invasive nature, it is only performed if followed by tumor embolization or if there is need to assess the contralateral arterial supply.
1.2
Management
The attitude toward therapy goals has changed and more and more a function preserving treatment (maintenance of cranial nerve function, hearing preservation or amelioration) is emphasized, next to a high level of tumor control. To date, surgery remains the only curative treatment option and preoperative embolization has been introduced to facilitate surgical removal . Absolute ethanol can be used for embolization as it is a sclerosing agent. It denudes the endothelial cell from the vascular wall, induces a spastic vasoconstriction with disruption of the vessel wall at the internal elastic lamina and as a result a blood clot is formed. This results in complete and permanent obliteration of the vessel lumen . Ethanol is toxic to the system and the maximal given dose is 1 mL/kg body weight/day .
The role of preoperative embolization is still a matter of debate . In this paper we will describe a group of patients with tympanic paraganglioma who were treated with preoperative embolization and surgical resection and we will examine tumor control rate, function preservation and safety.
2
Material and methods
Between 2006 and 2014, a total of 6 patients with a tympanic paraganglioma were treated with preoperative embolization with ethanol and total surgical resection with a function preserving intent at the ENT Department of the Ghent University Hospital in collaboration with the Department of Interventional Radiology. Approval of the UZ Ghent ethical committee was obtained for a retrospective review of the patient’s medical and imaging records. There were one male (age 66 years) and 5 females (mean age 58.8 years; range 54–69 years). The tympanic paragangliomas were categorized in accordance with the Fisch classification; 5 were class A. 1 patient had a class B tumor and was first treated with embolization alone but because of tumor growth, she was treated 4 years later with preoperative embolization and surgical resection.
Micro-otoscopic examination and evaluation of the cranial nerves were performed before and after treatment. Preoperative and postoperative audiograms were obtained. Pre- and postoperative Air Conduction (AC) and Bone Conduction (BC) thresholds (at frequencies 0.5, 1, 2 and 4 kHz) were recorded for each case and the pure tone average was calculated (PTA). Subtracting the individual pre- and postoperative average BC threshold from the pre- and postoperative average AC thresholds provided the values of the average Air Bone gap (ABG) at pre- and postoperative evaluations. Preoperative audiograms were compared with the postoperative audiograms for differences in AC, BC thresholds and ABG.
HRCT of the petrous bone and MRI of the inner ear and posterior fossa were carried out in all patients. One day preoperatively, DSA was carried out in all patients with subsequent embolization of the feeding vessel(s) when possible. The follow-up period was defined as that period of the time from surgery to the most recent office visit. Criteria for successful tumor control included no recurrence of symptoms, a normal microscopic evaluation of the middle ear and normal imaging when available.
2.1
Embolization procedure
Preoperative embolization started with a diagnostic DSA, performed under general anesthesia via a transfemoral Seldinger approach. After selective probing of the common, external and internal carotid artery with a 5F guiding catheter, biplanar angiography of the skull base and petrous bone was achieved using non-ionic contrast medium (Iodixanol 270, GE Healthcare). The tympanic paraganglioma was delineated by a typical homogenous hypervascular blush. After vascular mapping, a microcatheter (Marathon, EV3; Covidien) was introduced through the guiding catheter into the tumor feeder and advanced up to the petrous bone. After confirming the tumor blush on superselective contrast injection, embolization started with pure ethanol. The microcatheter was repeatedly loaded with 0.05–0.1 cm 3 of pure alcohol and subsequently slowly flushed with saline by a pump infusion at a rate of 3–6 cm 3 /h. After each injection, superselective control angiography demonstrated the gradual disappearance of the tumor blush. Once denudation of the tumor artery was achieved, the embolization was stopped. Catheter systems were then retrieved and the femoral artery sealed with an AngioSeal closure device. During the procedure, the patient was kept under heparin. The technical success of the embolization procedure was determined by the residual percentage of parenchymal staining of the tumor on post-embolization DSA.
2.2
Surgical technique
A retro-auricular tympanic access route with canaloplasty was used. Depending on the location and size of the glomus tumor, an endaural approach to the middle ear with additional mastoidectomy and myringoplasty was performed. Ossicular reconstruction was done if needed.
2.3
Procedure of histological preparation
The resected specimen was immersed into 4% buffered formalin, routinely processed and embedded in paraffin. 4 μ sections were made, stained with HE and histologically analyzed. Additional immunohistochemistry was performed, applying antibodies against S100-protein, chromogranin and synaptophysin. S100-protein demonstrates the sustentacular cells around the groups of cells displaying neuroendocrine activity as demonstrated with the antibodies against chromogranin and synaptophysin. Histologically, the resected tumors revealed a spectrum from severe congestion up to (partial or complete) necrosis. The positivity of the immunohistochemical staining was determined by the viability of the tumor cells.
2
Material and methods
Between 2006 and 2014, a total of 6 patients with a tympanic paraganglioma were treated with preoperative embolization with ethanol and total surgical resection with a function preserving intent at the ENT Department of the Ghent University Hospital in collaboration with the Department of Interventional Radiology. Approval of the UZ Ghent ethical committee was obtained for a retrospective review of the patient’s medical and imaging records. There were one male (age 66 years) and 5 females (mean age 58.8 years; range 54–69 years). The tympanic paragangliomas were categorized in accordance with the Fisch classification; 5 were class A. 1 patient had a class B tumor and was first treated with embolization alone but because of tumor growth, she was treated 4 years later with preoperative embolization and surgical resection.
Micro-otoscopic examination and evaluation of the cranial nerves were performed before and after treatment. Preoperative and postoperative audiograms were obtained. Pre- and postoperative Air Conduction (AC) and Bone Conduction (BC) thresholds (at frequencies 0.5, 1, 2 and 4 kHz) were recorded for each case and the pure tone average was calculated (PTA). Subtracting the individual pre- and postoperative average BC threshold from the pre- and postoperative average AC thresholds provided the values of the average Air Bone gap (ABG) at pre- and postoperative evaluations. Preoperative audiograms were compared with the postoperative audiograms for differences in AC, BC thresholds and ABG.
HRCT of the petrous bone and MRI of the inner ear and posterior fossa were carried out in all patients. One day preoperatively, DSA was carried out in all patients with subsequent embolization of the feeding vessel(s) when possible. The follow-up period was defined as that period of the time from surgery to the most recent office visit. Criteria for successful tumor control included no recurrence of symptoms, a normal microscopic evaluation of the middle ear and normal imaging when available.
2.1
Embolization procedure
Preoperative embolization started with a diagnostic DSA, performed under general anesthesia via a transfemoral Seldinger approach. After selective probing of the common, external and internal carotid artery with a 5F guiding catheter, biplanar angiography of the skull base and petrous bone was achieved using non-ionic contrast medium (Iodixanol 270, GE Healthcare). The tympanic paraganglioma was delineated by a typical homogenous hypervascular blush. After vascular mapping, a microcatheter (Marathon, EV3; Covidien) was introduced through the guiding catheter into the tumor feeder and advanced up to the petrous bone. After confirming the tumor blush on superselective contrast injection, embolization started with pure ethanol. The microcatheter was repeatedly loaded with 0.05–0.1 cm 3 of pure alcohol and subsequently slowly flushed with saline by a pump infusion at a rate of 3–6 cm 3 /h. After each injection, superselective control angiography demonstrated the gradual disappearance of the tumor blush. Once denudation of the tumor artery was achieved, the embolization was stopped. Catheter systems were then retrieved and the femoral artery sealed with an AngioSeal closure device. During the procedure, the patient was kept under heparin. The technical success of the embolization procedure was determined by the residual percentage of parenchymal staining of the tumor on post-embolization DSA.
2.2
Surgical technique
A retro-auricular tympanic access route with canaloplasty was used. Depending on the location and size of the glomus tumor, an endaural approach to the middle ear with additional mastoidectomy and myringoplasty was performed. Ossicular reconstruction was done if needed.
2.3
Procedure of histological preparation
The resected specimen was immersed into 4% buffered formalin, routinely processed and embedded in paraffin. 4 μ sections were made, stained with HE and histologically analyzed. Additional immunohistochemistry was performed, applying antibodies against S100-protein, chromogranin and synaptophysin. S100-protein demonstrates the sustentacular cells around the groups of cells displaying neuroendocrine activity as demonstrated with the antibodies against chromogranin and synaptophysin. Histologically, the resected tumors revealed a spectrum from severe congestion up to (partial or complete) necrosis. The positivity of the immunohistochemical staining was determined by the viability of the tumor cells.
3
Results
The most frequently reported symptoms were pulsatile tinnitus and hearing loss ( Table 2 ). None of the patients had a pattern of familiar occurrence. A reddish bulging mass in the middle ear was seen in all patients on micro-otoscopy, the mass was invading in the external ear canal in 1 patient and the mass was clearly pulsating in another patient. Cranial nerve evaluation showed normal function in all patients. Evaluation of hearing pre treatment showed normal hearing in 2 patients, conductive hearing loss in 3 patients and mixed predominant sensorineural hearing loss in 1 patient. Pre treatment audiometric AC and BC thresholds are shown in Fig. 2 .
| patient | gender | age | Fisch | pulsatile tinnitus | hearing loss | ear pressure | otalgy | otorrhea | itching ear | vertigo | tinnitus |
|---|---|---|---|---|---|---|---|---|---|---|---|
| A 2006 | F | 55 | B | + | + | + | − | + | − | − | − |
| A 2010 | B | + | + | − | − | + | − | − | − | ||
| B | M | 66 | A | + | + | − | − | + | + | − | − |
| C | F | 54 | A | + | − | − | + | − | − | − | − |
| D | F | 58 | A | + | + | + | − | − | − | − | − |
| E | F | 69 | A | − | + | − | − | − | − | − | − |
| F | F | 58 | A | + | + | − | + | − | − | − | − |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
