|Home | About | Journals | Submit | Contact Us | Français|
This article reports on the presentation, diagnosis, management, and treatment outcomes of lesions of the endolymphatic sac in patients treated at a tertiary neurotology referral center. It summarizes survival results in the largest series groups and presents a new diagnostic entity of pseudotumor of the endolymphatic sac. The study includes retrospective review of all patients diagnosed with lesions of the endolymphatic sac within our practice between 1994 and 2005 as well as review of the literature. The primary outcome measure was survival, and the secondary outcome measure was disease-free survival following definitive resection. Postoperative complications were assessed. Survival characteristics of the largest reported case series groups were reviewed. Five cases of endolymphatic sac lesions were identified. Of these, three were true endolymphatic sac tumors and two were inflammatory pseudotumors of the endolymphatic sac. All three of the endolymphatic sac tumors patients survived (100%), and two of the three had disease-free survival (67%). Two of three patients maintained persistent facial paresis postoperatively. Both patients with benign pseudotumors survived (100%). Our study concluded that endolymphatic sac tumors are rare neoplasms of the temporal bone that, although locally aggressive and invasive, have excellent prognosis for survival with complete resection. We report a new entity of pseudotumor of the endolymphatic sac that mimics true sac tumors in every respect on presentation but which is non-neoplastic in origin.
Endolymphatic sac tumors (ELSTs) are rare tumors of the petrous temporal bone. The first reported case of a tumor arising from the endolymphatic sac was discovered during sac decompression for presumed unilateral Ménière's disease in 1984.1 Classified as mastoid papillary tumors of unknown origin, these tumors were synthesized into a new, distinct clinicopathological entity by Heffner in 1989.2 Initially described as a low-grade papillary adenocarcinoma, their histological appearance and apparent lack of metastatic potential has since persuaded most practitioners to reclassify them as papillary adenomas. ELSTs can arise sporadically or in association with von Hippel-Lindau (VHL) disease.
Although benign, ELSTs can be locally destructive. They present with hearing loss, tinnitus, facial nerve weakness or paralysis, and vertigo, and can be lethal.3 Computed tomography (CT) imaging demonstrates erosion of the posterior petrous temporal bone with occasional intratumoral calcification. Magnetic resonance imaging (MRI) tumor signal is isointense to the brain and demonstrates gadolinium enhancement and heterogeneous signal intensity from intratumoral calcification and vascularity.4,5
We present our experience with ELSTs with respect to evaluation, diagnosis, and management. We reviewed the English language literature regarding ELSTs and compiled all papers reporting three or more cases of ELST within a given practice, in an attempt to define common characteristics of treatment and identify those management features consonant with successful or unsuccessful outcomes. We incorporated our results into the existing literature and summarized survival results of these major studies.
In addition, we report and discuss two cases of benign inflammatory lesions of the endolymphatic sac mimicking true ELSTs. The initial presentation and imaging of these pseudotumors of the endolymphatic sac are described in the context of the three true ELSTs in our practice, and we discuss limitations in diagnostic differentiation between tumors and pseudotumors using clinical and radiographic characteristics alone.
A census was performed within our practice for all benign and malignant neoplasms of the temporal bone from 1994 to 2005. In total, 134 intrinsic neoplasms of the temporal bone were identified. This figure excludes acoustic neuromas, meningiomas, carcinomas of the external auditory canal, cholesterol granulomas, and diffuse osseous lesions of the temporal bone.
By far the most common tumors were paragangliomas. In total, 105 glomus jugulare and glomus tympanicum tumors accounted for 78% of temporal bone neoplasms. The remaining tumors were comprised of six schwannomas of the jugular foramen, five facial nerve schwannomas, five hemangiomas, four metastatic foci to the temporal bone, two middle ear adenomas, two chondrosarcomas, and five lesions of the endolymphatic sac. Fig. Fig.11 presents the relative incidence of each tumor type within our practice.
A retrospective chart review of our patient database was performed after study approval was obtained from our Institutional Research Committee. We identified all patients with lesions of the endolymphatic sac. Demographic information including sex, side of lesion, age of patient, and follow-up intervals was obtained and summarized. Chart abstraction of clinical presentation, radiographic evaluation, management, and follow-up was performed for each patient, and synopses were generated.
A literature review of all case series and case reports of ELSTs in the English language was performed using the U.S. National Library of Medicine online database (PubMed). Articles containing series of three or more newly reported cases of ELST were selected for eligibility. Articles that reported study period, cohort size, patient demographics, tumor characteristics, presenting symptoms, hearing status, primary and adjuvant management, follow-up interval, and survival were chosen for review, and these data were derived from each study and compiled.
Five patients were identified with and treated for lesions of the endolymphatic sac at our institution between 1994 and 2005. Histopathological diagnosis of papillary adenoma of the endolymphatic sac was confirmed for three patients. Two patients had clinical and radiographic presentation consistent with ELST, but final histology revealed no evidence of papillary adenomatous proliferation, exhibiting only a benign inflammatory matrix within a normally appearing endolymphatic sac epithelium. These cases were deemed benign inflammatory pseudotumors of the endolymphatic sac.
Patient demographics and presenting signs and symptoms for all five patients are presented in Table Table1,1, with patients with true ELSTs denoted as 1, 2, and 3 and patients with pseudotumors denoted as A and B. There was a sex disparity with four women and one man. Three tumors were on the left, two on the right. Ages ranged from 17 to 58 years at the time of diagnosis and treatment. One patient had a known family history of VHL disease and was diagnosed with the disease postoperatively, marking her endolymphatic sac tumor as the harbinger of her VHL disease.
The most common presenting complaints were aural, with hearing loss present in four of five patients, followed by tinnitus, aural fullness, and imbalance in two patients each. The symptoms of pulsatile tinnitus, otalgia, otorrhea, vertigo, and facial paresis were present in one patient each. No other cranial neuropathies were diagnosed either at the time of presentation or following treatment.
Patient 3 presented in a unique fashion with chronic middle and external ear infection and obliteration of the ear canal by a friable mass, histologically demonstrating a papillary complex of cuboidal and columnar epithelium. Radiographic (CT, MRI, magnetic resonance angiography [MRA], magnetic resonance venography [MRV]) presentation was remarkable for extensive destruction of the lateral skull base and demonstration of a massive gadolinium-enhancing lesion centered on the left temporal bone and extending intracranially with medial displacement and compression of the cerebellum, brainstem, and fourth ventricle; superior extension into the middle fossa and enveloping the superior petrosal sinus; inferior extension into the jugular foramen; and thrombosis and loss of flow in the left transverse and sigmoid sinus. Fig. Fig.22 demonstrates the preoperative appearance of this patient's tumor, as seen on MRI, CT, and angiography.
Of interest is the similarity in presentations of patients with true ELSTs and those with sac pseudotumors. As seen in Table Table1,1, there is no obvious distinction in clinical presentation differentiating patients with true tumors from those with pseudotumors. These pseudotumors also mimicked true neoplasms radiographically, and only intraoperatively was there a suggestion of differences between the two.
Patient A demonstrated erosion of the petrous bone behind the internal auditory canal (IAC) and into the vestibule and semicircular canals on CT, with MRI revealing a 2-cm gadolinium-enhancing mass posterior to the IAC with extension into the labyrinth and labyrinthine enhancement. Intraoperative identification was made of a fibrous lesion eroding the semicircular canals and vestibule, and labyrinthectomy and skeletonization of the IAC were performed to completely resect the lesion. Frozen section histological examination revealed only dense fibrous tissue, and final histopathological diagnosis confirmed a benign endolymphatic sac containing only dense fibrous tissue and foci of chronic inflammation.
Patient B also demonstrated scalloped bone erosion along the operculum and posterior petrous face on CT, with MRI showing a gadolinium-enhancing lesion filling the extent of the endolymphatic sac along the posterior petrous apex and abutting the labyrinth. Fig. Fig.33 presents the CT and MRI radiographic findings in this patient, which are also representative of those of Patient A. A fibrous lesion was found along the posterior fossa dura encroaching on but not invading the labyrinth, and therefore a hearing conservation approach was used. The lesion infiltrated extensively posteriorly onto the sigmoid sinus, and a transjugular approach, with resection of the posterior fossa dura and sigmoid sinus, was necessary for complete resection. Final histology showed a diffuse inflammatory infiltrate within a normally appearing endolymphatic sac, without epithelial proliferation or papillary or cystic cytoarchitecture.
Table Table22 details the treatment approaches and outcomes for our patient cohort. With regards to treatment, all five patients underwent primary surgical resection of their lesions. The approaches for patients 1, 2, 3, and A were hearing ablative as all presented with hearing loss and all had tumor eroding the bony labyrinth. In addition, patients 3 and A had nonserviceable hearing. Patient B presented with normal hearing and intact otic capsule intraoperatively, and hearing was preserved with unchanged thresholds and speech discrimination postoperatively. All patients had gross total resection of tumor, with patient 3 retaining microscopic residual attached to the neural structures of the medial jugular foramen and encasing the internal carotid artery at the completion of surgery. Fig. Fig.44 illustrates the papillary-cystic nature of this patient's tumor, representative of the general histological appearance of ELSTs.
All patients with true ELSTs received adjuvant radiotherapy, with patients 1 and 2 receiving fractionated external beam radiotherapy and patient 3 receiving proton beam therapy. All patients are alive at the time of manuscript submission. Follow-up intervals range from 10 to 144 months, with the three tumor patients under routine surveillance with MRI.
Following evaluation of our group's patients, we reviewed all case series in the English literature of three or more patients with ELSTs.5,6,7,8,9,10,11 Seven eligible studies reported sufficient patient information regarding specific treatments, follow-up periods, and survival characteristics—no evidence of disease (NED), alive with disease (AWD), died of disease (DOD), or died of other cause (DOC)—to be included in the analysis, and to this we added the results of our present study. Table Table33 summarizes demographics, tumor characteristics, management, and treatment outcomes reported for each of these studies.
These eight groups comprised 49 cases. Study periods ranged from 11 to 39 years, with two articles presenting compilations of VHL patients having undergone hearing conservation surgery without specifying incidence reporting period.10,11 Patient ages at the time of presentation or treatment ranged from 17 to 75 years, although most patients clustered in the 30- to 50-year age group. There was a gender bias, with two thirds of the patients being women and one third of them men. Twelve patients, or 24%, had VHL, but this variable was not controlled in our analysis and incidence is heavily biased from the two compiled studies of VHL patients undergoing hearing conservation surgery. Reported follow-ups were highly variable and ranged from 0 to 261 months.
Reported tumor sizes ranged from 0.5 to 6.0 cm, with only half of the studies reporting. Hearing loss was the most common reported presenting symptom, occurring in nearly every reported patient. The most commonly involved nerve was the facial nerve, with preoperative facial paresis or paralysis in 43% of patients. In patients with larger tumors or in those who delayed presentation for decades after onset of initial symptoms, multiple cranial neuropathies were present including trigeminal (five patients) and glossopharyngeal/vagal nerves (five patients).
Surgical resection was the primary modality of treatment in all but one patient, with adjuvant radiotherapy applied in 14 patients (29%). Two thirds of patients underwent hearing ablative surgical approaches, whereas the remaining one third underwent hearing conservation surgery. Of these, nine patients had VHL and had their tumors diagnosed relatively earlier secondary to surveillance audiometry.
Total tumor resection was achieved on initial surgery in 40 patients. The remaining nine patients with subtotal tumor resection were divided into two groups. One patient each from three groups had massive tumors larger than 5 cm with significant cranial nerve and vascular involvement, making total resection without significant functional impairment impossible.8,9 Six patients from one study underwent planned subtotal resection with postoperative radiotherapy.5
Overall survival characteristics were 74% NED, 20% AWD, and 4% DOD. Within all groups, two patients died of disease within the reporting period. One patient died of esophageal bleeding during the course of palliative radiation therapy for presumed metastatic renal cell carcinoma to the skull base. The second patient presented with four prior subtotal resections of presumed paraganglioma followed by 54-Gy radiotherapy. The patient had a 5.5-cm recurrent tumor with extensive invasion to the jugular foramen and cavernous sinus and encasement of the internal carotid artery. Attempted radical resection was not successful.
The differential diagnosis for ELSTs includes all intrinsic temporal bone neoplasms (most commonly paraganglioma) as well as metastatic papillary thyroid carcinoma, metastatic renal cell carcinoma, and choroid plexus papilloma, the latter three of which are similar in appearance to ELSTs histologically.
ELSTs are highly vascular and are comprised of papillary cystic structures lined with a simple cuboidal or columnar epithelium. Siderophages and cholesterol clefts are seen, as are clear cells (vacuolated cells). Nuclear pleomorphism is not pronounced, and mitoses are rare.
Immunohistochemistry and special staining may aid in differentiation of papillary tumors of questionable origin. ELSTs usually stain positive for cytokeratin, vimentin, and epithelial membrane antigen, as well as stain on periodic acid-Schiff (diastase sensitive).2,12,13 Some articles have also reported sensitivity to glial fibrillary acid protein.9,14 However, most authors have had poor tumor reactivity to glial fibrillary acid protein.7,15,16 Papillary thyroid metastasis to the temporal bone may be differentiated by positive reaction to thyroglobulin immunohistochemistry. Transthyretin has been shown to exhibit differential expression in choroid plexus papillomas with little to no expression in ELSTs.15
The nature of invasive lesions to the temporal bone can be confusing when attempting to differentiate them by clinical and radiographic means. By sheer probability, a vascular skull base tumor eroding the temporal bone is likely to be a paraganglioma (and likely a glomus jugulare) (Fig. 1). Glomus tumors typically present with pulsatile tinnitus, which is an uncommon presenting symptom in our cohort. Large glomus tumors as well as large ELSTs can both present as pink or purple masses encroaching on the middle ear and external auditory canal, as seen in one of our patients as well as in nearly half of those patients in one case series.9 Glomus tumors exhibit a characteristic “salt and pepper” tumor appearance on MRI, but this heterogeneity in signal reflects the vascularity of such tumors and is not pathognomonic. The heterogeneity in signal seen in large ELSTs—arising from hypervascularity as well as intratumoral hemorrhage and/or calcification—can often mimic glomus tumors in this respect.17,18 This is not necessarily problematic, as management would proceed similarly for either histology: preoperative embolization followed by total tumor resection via the appropriate lateral skull base approach.
Controversy exists in differentiating aggressive papillary tumors of the posterior fossa as papillomas and papillary carcinomas of the choroid plexus versus of the endolymphatic sac. As mentioned, transthyretin immunohistochemistry can distinguish choroid plexus tumors in some cases.15 Radiographic features also discriminate the two tumor types: choroid papillomas of the cerebellopontine angle cistern extend from the foramen of Luschka, and therefore tumors that clearly do not extend into the fourth ventricle are unlikely to be of the choroid plexus and are more likely to be of the endolymphatic sac.19
Further differentiation can be made based on angiographic findings. Although large ELSTs may exhibit multiple blood supplies from external carotid, internal carotid, and vertebrobasilar systems, there are no ELSTs supplied solely by the internal carotid and posterior circulation.18,20 In contrast, choroid plexus papillomas, as intra-axial tumors, exhibit blood supply solely from the posterior cerebral and posterior inferior cerebellar arteries.19,21 Two cases in our cohort underwent preoperative embolization, and both were supplied by branches of the occipital and ascending pharyngeal arteries, without involvement of the internal carotid or vertebrobasilar systems. This is consistent with the known blood supply of the endolymphatic sac.22
VHL disease affects ~1 in 39,000 people.23 The current literature suggests that approximately one third of all ELSTs are associated with VHL disease.8,9,10,24 VHL disease is inherited in an autosomal dominant fashion. It encompasses a variety of neoplasia, both benign and malignant, including renal cell carcinomas; central nervous system and retinal hemangioblastomas; pheochromocytomas; and cysts of the kidneys, pancreas, and epididymis. The gene responsible for VHL disease is a tumor suppressor that has been mapped to chromosome 3p25.25 The VHL protein has a role in oxygen sensing, and its loss of function mutation can increase expression of hypoxia-inducible factor, subsequently leading to tumorigenesis.26 In VHL disease, it is believed that tumors arise when both an inherited germline mutation and a loss-of-function mutation of the wild-type VHL gene are present.27
Even though temporal bone lesions were described in patients by Lindau in 1926, the association of these tumors with VHL disease was not made until recently.28 This clinical association has been confirmed at the molecular level with mutations in the VHL gene identified in endolymphatic sac tumors in VHL patients.29 Approximately 10% of patients with VHL disease have ELSTs, and ~30% of VHL patients with ELSTs have bilateral tumors.11,28 This variable phenotypic expression may be a reflection of VHL gene function secondary to the type of mutation present. Indeed, VHL disease has been found to have phenotypic expression consistent within members of a family, thus implying a singular type of mutation. This is the basis of VHL family typing, with type 1 being without pheochromocytomas and type 2 being with pheochromocytomas. There is further subclassification of type 2 into type 2a, which has low risk for developing renal cell carcinoma, and type 2b, which has high risk for developing renal cell carcinoma. Type 1 families usually have deletion or truncation mutations, whereas type 2 families usually have missense mutations.30
If a family history of VHL disease exists, or if the diagnosis of VHL disease is made in the absence of an ELST, then early routine audiologic screening can allow early tumor detection and the possibility of hearing preservation surgery should ELST develop. Positive identification of tumor on MRI with gadolinium is necessary before surgery: To date, surgical exploration in VHL patients with audiovestibular symptoms but without MRI abnormalities has not been documented and is not recommended.
More than 175 case reports of ELSTs have been described in the literature.8,9,10,16,18,24 The majority of these are single case reports of a practice group or university. As the majority of these case reports do not disclose the population size of their patient base, it is difficult to assess the true incidence of these tumors. Our 49 reviewed cases comprise nearly one third of all reported cases of ELSTs.
An interesting finding concerns the bias for ELSTs to afflict women more than men. Our practice cohort was comprised of two women and one man, and this gender disparity has been reported in other studies, with Feghali et al reporting 3 of 3 cases in women,7 Benecke et al reporting 4 of 5 cases in women,6 and Hansen and Luxford reporting 10 of 14 cases in women.8 This gender bias appears in VHL patients with ELSTs as well, with women being affected twice as often as men.11 The overall female-to-male ratio is 2:1 in our review of the literature. The significance of this finding has not yet been established.
The most common presenting symptom was sensorineural hearing loss, although the magnitudes of such loss could not be tabulated across series due to variability in reporting. This significantly affected surgical decision making, as 32% of patients underwent hearing conservation procedures while 68% underwent hearing ablative procedures. In our cohort, all three of our ELST patients underwent hearing ablative procedures, with two having serviceable hearing preoperatively. Patient 1 had tumor directly invading the vestibule, which was noted intraoperatively. Patient 2 had facial nerve involvement preoperatively and clearly stated her priority of maximizing facial nerve outcome over hearing preservation.
The majority of patients reported to have undergone hearing conservation surgery in the literature were VHL patients in the two VHL-only studies.10,11 Three out of 4 patients in one study had bilateral disease, with all undergoing successful hearing preservation surgery on one side. One patient had prior hearing ablative surgery for a large contralateral ELST, one patient is being monitored for a small stable contralateral ELST, and one patient had successful hearing preservation surgery for removal of a small ELST on the contralateral side as well.11 All five patients in the second VHL study reported stabilization in hearing levels following retrolabyrinthine resection of tumor.10
These two groups of patients are unique in that all had tumors under 2 cm and were undergoing active surveillance and cranial imaging for hemangioblastoma as part of their VHL disease management. There was relatively little delay between onset of audiovestibular symptoms and identification of tumor.
An earlier study presented four patients who underwent subtotal resection for hearing conservation with planned adjuvant radiotherapy; of these four, two patients had subsequent regrowth of tumor requiring re-resection.5
Despite the benign histological nature of these tumors, complete resection appears crucial for ensuring success. As mentioned previously, half of the patients in one study group undergoing hearing conservation approaches with subtotal resection followed by adjuvant radiation therapy had regrowth of tumor.5
In a later study, nearly all patients presented with late findings—five with facial nerve (FN) paresis, three with tumor extension into the middle ear, and one with hydrocephalus.9 Six of seven patients required FN resection or rerouting intraoperatively. Despite the advanced stage at presentation, complete tumor resection was achieved in six patients, none of whom showed recurrence. The seventh patient had extensive dural involvement making complete resection impossible, but remains AWD over 5 years.
In another study, 12 of 14 patients underwent complete tumor removal on initial surgery, and all 12 remained disease free up to 9 years following resection.8 Two patients with initial incomplete resections have had subsequent recurrences: One is alive but with growing disease and the other is DOD.
In our study, patients 1 and 2 underwent total resection on initial surgery and continue to be tumor free. Patient 3 had a 6-cm tumor with extensive involvement of neural and vascular structures in all directions within the posterior and middle fossa. Microscopic residual was treated with proton beam therapy. Apart from facial paralysis secondary to facial nerve transposition, he is free of symptoms and without evidence of regrowth of residual tumor 1 year following resection.
Adjuvant radiotherapy was used on 14 patients (29%). Six patients with total resection followed by radiotherapy have had no recurrence, including both from our group, with 1to 12 years of follow-up.6,8 Three patients with subtotal resection followed by radiotherapy have had no progression of disease from 2 to 8 years follow-up.5 However, three patients with subtotal resection followed by radiotherapy have had regrowth of tumor.5,8 One patient with reported complete resection followed by radiotherapy had recurrence.7 No information was available for the final patient.5
We found no reported cases of radiation therapy and/or stereotactic radiotherapy used as the primary modality of treatment for ELSTs. There have been sporadic uses of stereotactic radiosurgery following subtotal tumor resection, but each of these has resulted in tumor recurrence.5,31,32,33
Total tumor resection is clearly the treatment of choice, as only one patient with reported complete resection had subsequent recurrence. In some tumors, total resection cannot be achieved without risk of catastrophic loss of function or death, and in these patients subtotal resection may be warranted. Patients who have subtotal resection may benefit from postoperative radiotherapy, but there still remains a roughly 50% risk of tumor regrowth, and therefore close surveillance is warranted as re-resection may be necessary. Stereotactic radiotherapy has shown no increased benefit above standard fractionated radiotherapy in survival or recurrence rates, and subtotal resection followed by stereotactic radiotherapy has uniformly resulted in tumor regrowth in the published literature.
There are currently no reported cases of spontaneous metastatic dissemination of ELSTs in the literature. Recently, however, two reports have surfaced describing metastatic disease following subtotal resection. The first reported case of ELST drop metastasis was in 2002, with dissemination onto the ipsilateral cerebellar convexity beyond the original tumor site in a patient who had undergone previous subtotal resection and radiotherapy.31 In 2005, a second case of drop metastasis of an ELST was reported, this time to the spine, manifesting after multiple subtotal resections and three courses of stereotactic radiosurgery.32
These seminal reports serve to illustrate the importance of complete tumor removal on initial resection to minimize both recurrence and metastatic seeding. The oncologic principle of complete tumor extirpation on primary resection is certainly applicable to ELSTs, despite their benign histology and absence of spontaneous metastasis.
We discuss the interesting findings of patients A and B, who presented with clinical and radiographic evidence of lesions mimicking ELSTs but without histopathological evidence of neoplasm. One patient had a 3-month history of complete unilateral hearing loss and symptoms of Ménière's syndrome, including tinnitus, fullness, vertigo, and imbalance. The other patient had a 2-year history of unilateral aural fullness with intact hearing. In both cases, inner ear symptomatology prompted further evaluation, which eventually included gadolinium-enhanced MRI, and in both cases contrast-enhancing lesions were clearly identified within the fossa of the posterior petrous bone at the native location of the endolymphatic sac. Both pathological specimens demonstrated fibrous infiltration of the sac, with erosion of the otic capsule in patient A and diffuse dural and lateral sinus involvement in patient B. Histopathology demonstrated only benign fibrous and chronic inflammatory components within normal sac epithelium in both cases. This new disorder represents a non-neoplastic pseudotumor of the endolymphatic sac secondary to chronic localized inflammation, the etiology of which remains unknown in both patients.
Prior histopathological studies of human temporal bones have demonstrated perisaccular fibrosis with normal endolymphatic sac epithelium in Ménière's disease patients, a nonspecific finding but indicating localization of inflammatory activity to the sac.34,35,36,37 MRI of the in vivo temporal bones of these patients was not available to correlate to these findings.
A recent study of 12 patients with assorted auditory and/or vestibular symptoms demonstrates gadolinium enhancement of the endolymphatic sac on MRI.38 These patients were treated medically, without surgical intervention of the sac, again precluding the opportunity to corroborate imaging with histological diagnosis. Only two patients had repeat MRI, and both had normalization—loss of enhancement—of the sac on repeat imaging.
To our knowledge, patients A and B are the first two case reports correlating aural symptomatology, contrast enhancement of the endolymphatic sac on MRI, and histological verification of acute or chronic inflammation of the sac without evidence of neoplasm. As ELSTs have become more widely recognized and as implementation of high-resolution MRI becomes more widespread, we are now more likely to identify gadolinium-enhancing lesions of the endolymphatic sac in general, and therefore the rate of false-positive diagnoses of ELSTs, based on clinical and radiographic features, may increase.
The pathophysiology behind our patients' endolymphatic sac pseudotumors is unclear. However, recognition of these lesions as a distinct entity may allow more comprehensive medical and histological evaluation following resection, and this may aid in understanding their etiology and natural history.
Noninvasive methods of differentiating ELSTs from benign inflammatory pseudotumors may gain importance in the future. At the moment, however, the only means of definitive confirmation—or refutation—of neoplasm within the sac is by complete excision and histological examination of the entire sac itself because ELSTs arise from the rugose or intratemporal portion of the sac. Given the aggressive and potentially life-threatening nature of ELSTs, we feel that the burden of proof must continue to be in refuting the existence of neoplastic disease via complete surgical excision of all possible ELSTs, despite this newly recognized potential for false-positive diagnosis and negative exploration.
ELSTs are histologically benign yet destructive, highly aggressive lesions. They show excellent response to primary surgical resection, with or without adjuvant radiotherapy. Complete tumor removal on initial resection is crucial. Hearing preservation should not take precedence over complete tumor removal, as adjuvant radiotherapy does not ensure against tumor recurrence, which can be devastating and lethal. In addition, drop metastases following subtotal tumor resection have now been reported. In patients with VHL disease, regularly scheduled audiometry and surveillance MRI are vital to early detection of ELSTs, which can optimize the opportunity for hearing preservation without compromising tumor control.
Pseudotumors of the endolymphatic sac represent a new diagnostic entity characterized by non-neoplastic inflammation and gadolinium enhancement of the sac on MRI. These lesions cannot be differentiated from true ELSTs on clinical and radiographic appearance alone; therefore, complete excision and examination of the entire sac is necessary to definitively exclude neoplasm. Surgical resection remains the treatment of choice for all gadolinium-enhancing lesions of the endolymphatic sac.