Abstract
Objective
To analyze the long term outcomes after surgery in tympanomastoid paragangliomas.
Study design
Retrospective study.
Methods
The charts of 145 patients with tympanomastoid paragangliomas managed between 1988 and 2013 were reviewed. The clinical features, audiological data, pre- and postoperative notes were noted. The tumors were staged according to the modified Fish and Mattox classification. The surgical approaches for all patients were formulated according to the surgical algorithm developed at our center.
Results
34 (23.5%), 46 (31.7%), 22 (15.2%), 18 (12.4%) and 25 (17.2%) patients were diagnosed to have TMP class A1, A2, B1, B2 and B3 tumors respectively. Gross tumor resection was achieved in 141 (97.2%) patients. The facial nerve was uncovered in four patients and infiltrated in three. The cochlea was found eroded in seven cases. The mean follow-up was 48.4 months. Recurrence was seen in one patient (0.7%). In the cases where the facial nerve was preserved (n = 143), the nerve function was graded as HB grade 1 in 138 patients (97%). Postoperatively, the mean AC showed an improvement in all categories except in class B2 and B3, which corresponds to the classes that include patients who underwent subtotal petrosectomy.
Conclusion
We report the long term surgical outcomes in tympanomastoid paragangliomas in the largest series published till date. It is possible to completely eradicate all types of tympanomastoid paragangliomas with minimum sequelae by choosing the correct surgical approach to achieve adequate exposure for individual tumor classes as described in our classification and algorithm.
Level of evidence
IIb.
1
Introduction
Temporal bone paragangliomas (TBPs) are benign but locally aggressive tumors that arise from various sites in the temporal bone. There are two types: (1) tympanomastoid paraganglioma (TMP), commonly known as ‘glomus tympanicum (GT)’, which are tumors that originate from the glomus bodies that lie along the Jacobson’s nerve and the Arnold’s nerve and (2) tympanojugular paraganglioma (TJP) or ‘glomus jugulare’ which originate from the paraganglia located in the adventitia of the dome of the jugular bulb or those that originate in the hypotympanum with secondary invasion of the jugular bulb . The term ‘glomus’ itself is a misnomer and hence its use must be discontinued and the above terminology must be used to replace ‘glomus’ consistently while reporting in literature.
With the advances in imaging and refinements in surgical techniques even the most difficult TMPs can be safely eradicated thereby offering a complete cure and making surgery the treatment of choice. The rates of recurrence are very low and hearing preservation can be achieved in almost all cases. The post-operative sequelae and complications are minimal and are comparable to other middle ear surgeries. Barring the one by Forest et al. there have been no other articles that comprehensively described long term outcomes of surgery in TMPs in the last 20 years. In this article we present the long term outcomes in the surgical management of patients with this rare subset of tumors at our center over a 30-year period which is presently also the largest series published in English literature thus far.
2
Materials and methods
The medical records of 145 patients with TMPs who were managed at the Gruppo Otologico, Piacenza, Rome (Italy), between December 1988 and July 2013 were reviewed. Clinical, audiological and radiological data of all the patients were documented and analyzed and tumors staged according to the modified Fish and Mattox classification ( Table 1 ). At our center all patients with TMPs undergo pre-operative HRCT with iodinate contrast medium and an MRI with gadolinium DTPA enhancement when the tumor is found to be close to the jugular bulb or the carotid. Angiography and neuroendocrine testing are not routinely performed for TMPs. The surgical approaches for all patients with TMPs are formulated according to the surgical algorithm as shown in Fig. 1 . The surgical techniques have been described elsewhere . The pre-operative and postoperative facial nerve (FN) function was graded according to the House–Brackmann (HB) grading system . Follow up consisted of clinical evaluation, hearing tests and serial CT scans. Follow up was defined as that period of time from surgery to the most recent office visit or patient contact. Most patients were followed up at our center. Four patients who were lost for follow up or were not followed up at our center were contacted by phone to get details of follow up.
| Class | Description | |
|---|---|---|
| A | Tumors limited entirely to the middle ear | |
| A1 | Tumors completely visible on otoscopic examination | |
| A2 | Tumor margins are not visible on otoscopy. Tumor may extend anteriorly up to the Eustachian tube and/or to the posterior mesotympanum | |
| B | Tumors limited to the tympanomastoid segment (middle ear cleft) of the temporal bone | |
| B1 | Tumors filling the middle ear with extension into the hypotympanum and tympanic sinus | |
| B2 | Tumors filling the middle ear with extension into the mastoid and medially to the mastoid segment of the facial nerve | |
| B3 | Tumors filling the middle ear with extension into the mastoid with erosion of carotid canal | |
2.1
Statistical analysis
Data were analyzed with a statistical software program (SPSS Statistics for Windows version 20, Chicago, IL). Continuous data were summarized as mean ± interval of confidence at 95% (IC 95%). Categorical data were presented as frequencies and percentages. Preoperative and postoperative hearing results were evaluated according to tumor class and type of surgery performed. p Values below 0.05 were considered significant.
This study has been approved by the institutional review board of the Associazione Italiana Neurotologica (AINOT).
2.2
Results and observations
Of a total of 382 patients with TBPs who were managed at the Gruppo Otologico, a quaternary referral center for otology and skull base surgery in Italy, 145 were diagnosed to have TMPs. The male:female ratio was 15 (10.3%):130 (89.6%). The age, sex and clinical presentation of the study population are presented in Table 2 . The median age of the population was 55 years (range 13 years–82 years). 93 (64%) patients presented with pulsatile tinnitus, 118 (81%) with hearing loss, and 22 (15%) with vertigo and 10 (7%) presented with ear discharge and ear pain. A pulsatile reddish mass was seen behind an intact tympanic membrane (TM) in 137 (94%) of the patients while the TM was perforated in the remaining 8 (6%) patients. Pre-operative facial nerve palsy was seen in 2 (1%) patients. The mean length of symptoms previous to presentation was 36 ± 33.1 months (range, 3–192 months). Two patients presented to us after failure of radiotherapy (RT) and three others presented to us after failure of surgery elsewhere. There were no patients associated with other paragangliomas elsewhere in the body.
| Class of TMP (n = 145) | Number | Age | Sex | Presenting symptoms | Otoscopy | Facial nerve palsy | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| < 20 years | 20–40 years | 40–60 years | > 60 years | Male | Female | Pulsatile tinnitus | Hearing loss | Vertigo | Pain, otorrhea | Pulsatile mass behind intact TM | TM perforation/previous surgery | ||||
| A (n = 80) | A1 | 34 (24%) | – | 6 (17%) | 21 (62%) | 7 (21%) | – | 34 (100%) | 25 (73.5%) | 23 (68%) | 5 (15%) | – | 34 (100%) | – | – |
| A2 | 46 (32%) | – | 8 (17%) | 22 (48%) | 16 (35%) | 5 (11%) | 41 (89%) | 17 (37%) | 33 (72%) | 6 (13%) | 2 (4%) | 45 (98%) | 1 (2%) | – | |
| B (n = 65) | B1 | 22 (15%) | 1 (5%) | 2 (9%) | 14 (63%) | 5 (23%) | 1 (5%) | 21 (96%) | 20 (91%) | 21 (96%) | 5 (23%) | 2 (9%) | 20 (91%) | 2 (9%) | – |
| B2 | 18 (12%) | – | 4 (22%) | 5 (28%) | 9 (50%) | – | 18 (100%) | 15 (83.3%) | 16 (89%) | 2 (11%) | 2 (11%) | 16 (89%) | 2 (11%) | 1 (6%) | |
| B3 | 25 (17%) | – | 4 (16%) | 9 (36%) | 12 (48%) | 1 (4%) | 24 (96%) | 16 (64%) | 25 (100%) | 4 (16%) | 4 (16%) | 22 (88%) | 3 (12%) | 1 (4%) | |
| Total | 145 (100%) | 1 (1%) | 24 (16%) | 71 (49%) | 49 (34%) | 7 (5%) | 138 (95%) | 93 (64%) | 118 (81%) | 22 (15%) | 10 (7%) | 137 (94%) | 8 (6%) | 2 (1%) | |
2.3
Tumor class and surgical procedure
34 (23.5%) patients were diagnosed to have TMP class A1 tumors, 46 (31.7%) to have class A2 tumors, 22 (15.2%) to have class B1, 18 (12.4%) to have class B2 and 25 (17.2%) to have class B3 tumors. All class A1 tumors were approached with a transcanal approach (TCA), class A2 with a postauricular transcanal approach (PA-TCA) and class B1 with a canal wall up (CWU) mastoidectomy with a posterior tympanotomy (PT) ( Fig. 2 ). 10/18 patients with class B2 tumors underwent with a CWU mastoidectomy with a PT and a subfacial recess tympanotomy (SFRT) ( Fig. 3 ). The remaining eight patients underwent a subtotal petrosectomy (STP) with middle ear obliteration (MEO) due to cochlear fistula, inadequate exposure of tumor margins or bleeding. All but one of the patients with class B3 tumors underwent an STP with MEO. In the remaining case, an infratemporal fossa approach (ITFA) type A was performed combined with a transcochlear approach (TCA) for a large class B3 tumor which had infiltrated the facial nerve which was sacrificed and reconstructed using a sural nerve graft. This patient had a previously anacoustic ear. In another patient who underwent an STP, the facial nerve was sacrificed and a facial-hypoglossal anastomosis was done later. One patient who had undergone RT elsewhere for a class B3 tumor also underwent an STP.
2.4
Intraoperative findings ( Table 3 )
Gross tumor resection was achieved in 141 (97.2%) patients. Subtotal resection (tiny tumor bit or > 98% tumor removal) was the result in four patients with B2/B3 tumors. Barring the patients in whom an STP with MEO was performed (n = 112), the ossicular chain was preserved intact in 106 (95%) patients. In the remaining 6 (5%) patients, the ossicular chain was reconstructed with autologous incus interposed between the stapes and the malleus in a second stage surgery. The facial nerve was uncovered in four patients and infiltrated in three. The cochlea was found eroded in seven; middle fossa dura and the temporomandibular joint were eroded in two cases each. There was an associated cholesteatoma in three patients and cholesterol granuloma in two patients. In two patients there was an intraoperative CSF leak which was closed using a muscle graft.
| O’Leary MJ et al. 1991, (n = 64) | Jackson CG et al. , 1989, (n = 60) | Forest JA et al. , 2001, (n = 95) | Our series, 2013 (n = 145) | |
|---|---|---|---|---|
| No. (%) | No. (%) | No. (%) | No. (%) | |
| Gross total resections (GTR) | 61 (95.3%) | 54 (90%) | 90 (94.7%) | 141 (97%) |
| Recurrence in GTR | 0 (0%) | 2 (3.7%) | 2 (2%) | 1 (0.7%) |
| Management of recurrence in GTR | – | S (n = 1)/W&S (n = 1) | S (n = 1)/W&S (n = 1) | S (n = 1) |
| Subtotal resections (STR) | 3 (4.7%) | 6 (10%) | 4 (4.2%) | 4 (2.8%) |
| Recurrence in STR | 3 (100%) | NA | 2 (50%) | 0 (0%) |
| Management of recurrence in STR | S (n = 3) | S (n = 6) | RT (n = 1)/W&S (n = 1) | – |
| Tympanic membrane perforations | 3 (4.7%) | 1 (1.7%) | 4 (4.2%) | 6 (4.1%) |
| External auditory canal stenosis | 0 (0%) | 0 (0%) | 1 (1%) | 3 (2.1%) |
| External auditory canal cholesteatoma | 0 (0%) | 0 (0%) | 0 (0%) | 1 (0.7%) |
| Middle ear cholesteatoma | 1 (1.6%) | 1 (1.7%) | 1 (1%) | 0 (0%) |
| Temporary facial nerve palsy | 0 (0%) | 1 (1.7%) | 1 (1%) | 5 a (3.5%) |
| Permanent facial nerve palsy | 1 (1.6%) | 0 (0%) | 0 (0%) | 2 b (1.4%) |
| CVA/hemiplegia due to ICA violation | 0 (0%) | 1 (1.7%) | 1 (1%) | 0 (0%) |
| Intraoperative CSF leaks | 0 (0%) | 2 (3.3%) | 0 (0%) | 0 (0%) |
| Wound infection | 0 (0%) | 1 (1.7%) | 0 (0%) | 0 (0%) |
a Recovered to HB grade 1 in 6 months.
2.5
Follow up and complications
The follow-up ranged from six to 230 months (mean: 48.4 ± 51.1 months, median: 47.2). A recurrent TMP was seen only in one patient (0.7%) after nine years for whom a revision surgery was performed and the tumor was excised completely. This patient is tumor free after five years of follow up after the second surgery. Postoperative TM perforations were seen in six (4.1%) patients and external auditory canal stenosis in thee (2.1%) patients which were all reconstructed in a second stage. A pearl-like residual cholesteatoma was found in the external auditory canal in one case (0.7%) and easily removed in the outpatient clinic.
In the immediate post-operative period, in the cases where the facial nerve was preserved (n = 143), the nerve function was graded as HB grade 1 in 138 patients (97%), HB grade II in three (2%) and HB grade III in two (1%) had grade III. Of the five facial palsies, the tumor was adherent to an uncovered facial nerve in three cases. At the last follow-up the facial palsies had recovered and all 143 patients had grade I FN function. In the patient in whom the nerve was sacrificed and reconstructed with a sural nerve graft, the nerve function was graded as HB grade III after one year. In the patient in whom a hypoglossal facial anastomosis was done, the nerve function was HB grade IV after 1 year.
2.6
Hearing results
Preoperative and postoperative hearing was analyzed according to the modified Fish classification of TMPs. Eleven patients with preoperative anacusia (seven due to cochlear erosion, three due to previous surgery elsewhere and one due to previous RT) were excluded from the hearing analysis. Two patients with incomplete postoperative records were excluded from the postoperative hearing results analysis. It can be observed in Fig. 4 that preoperative hearing worsened with progression of tumor class. Postoperatively, the mean AC showed an improvement in all categories except in class B2 and B3, which corresponds to the classes that include patients who underwent STP. There was no statistically significant difference in the improvement in AC among the classes A1 through B1 (p = 0.27, Kruskal–Wallis test).
