Lipochoristomas (Lipomatous Tumors) of the Acoustic Nerve Sandy S. Wu, MD; William W. M. Lo, MD; Donald L. Tschirhart, MD; William H. Slattery III, MD; Joseph N. Carberry, MD; Derald E. Brackmann, MD

● Context.—Lipochoristomas (lipomatous choristomas) are rare tumors of the acoustic nerve (cranial nerve VIII/vestibulocochlear nerve) within the internal acoustic canal and sometimes the cerebellopontine angle, and are histogenetically believed to be congenital malformations. Their clinically indolent behavior has recently prompted a more conservative management protocol in a quest for maximal nerve/hearing preservation. This approach contrasts sharply with that for the common internal acoustic canal/cerebellopontine angle tumors, the neuroepithelial neoplasms (acoustic schwannomas and meningiomas), which behave more aggressively and have more prominent clinical manifestations. Owing to their rarity, the clinicopathologic features of cranial nerve VIII lipochoristomas have been obtained mainly through case reports. Objective.—We present the clinicopathologic features of 11 cases of lipochoristomas of cranial nerve VIII. Design.—The 11 cases were documented between 1992 and 2003. We performed complete clinical reviews with

histologic, histochemical, and immunohistochemical analyses of formalin-fixed, paraffin-embedded tumor samples. Results.—The patients were 8 men and 3 women with hearing loss of the right ear (5 patients) or the left ear (6 patients). No patient had bilateral tumors. All lipochoristomas histologically possessed mature adipose tissue admixed with varied amounts of mature fibrous tissue, tortuous thickwalled vessels, smooth muscle bundles, and skeletal muscle fibers, the latter verified with immunohistochemistry. Conclusions.—The histomorphologic and immunophenotypic evidence showed that these tumors are better characterized as choristomas than as simple ‘‘lipomas,’’ as they have been labeled in the past. Their overall nonaggressive clinical nature in addition to the characteristic radiologic and histomorphologic findings are important clinicopathologic features for the pathologist to recognize and differentiate, especially during frozen section evaluations, in order to direct the neurosurgeon to a more appropriate conservative therapeutic intervention. (Arch Pathol Lab Med. 2003;127:1475–1479)

N

Complete pathologic evaluations are uncommon.11 Results from larger series with thorough histomorphologic analyses have yet to be documented. Our study represents the experience at our institution during the past decade. The clinicopathologic features of 11 cases of lipochoristomas of the IAC are described. (During the same 10-year period, 2537 cases of acoustic schwannomas and meningiomas were identified from our surgical pathology archive; 11/ 2537 5 0.43%.) The histomorphology, immunohistochemistry, and clinical behavior of the tumors are cataloged. A comparison is made with the information available in the English language literature.

eoplasms of the internal acoustic canal (IAC) and cerebellopontine angle (CPA) often involve the acoustic nerve (cranial nerve VIII [CNVIII]/vestibulocochlear nerve). The vast majority are of neuroepithelial nosology, comprised mostly of acoustic schwannomas (approximately 80%–90%) and meningiomas (approximately 10%),1–4 while uncommon tumors include epidermoid/dermoid cysts, arachnoid cysts, and a variety of rare benign and malignant neoplasms. Lipochoristomas (lipomatous choristomas), previously known as lipomas, are rare lesions of the IAC/CPA, comprising 0.1% of all IAC/CPA tumors.1–11 Many have been discovered incidentally. They are slowgrowing lesions with clinically indolent behavior and have been surmised to be congenital malformations rather than neoplasms.12 Clinical features of CNVIII lipochoristomas have mostly been examined in case studies and extensive reviews.5,7 Accepted for publication June 2, 2003. From the Departments of Pathology (Drs Wu, Tschirhart, and Carberry) and Radiology (Dr Lo), St Vincent Medical Center, Los Angeles, Calif; and the House Ear Clinic, Los Angeles, Calif (Drs Slattery and Brackmann). Reprints: Sandy S. Wu, MD, Department of Pathology, St Vincent Medical Center, 2131 W 3rd St, Los Angeles, CA 90057-0992 (e-mail: [email protected]). Arch Pathol Lab Med—Vol 127, November 2003

MATERIALS AND METHODS A complete review of medical records and radiographic films was performed. The pathology specimens included formalinfixed, paraffin-embedded tissue blocks from the surgically excised lipochoristomas or from tissue biopsies. Serial sections were produced from each block, and hematoxylin-eosin–stained slides were made. Additional trichrome histochemically stained slides and immunohistochemical analyses (S100, glial fibrillary acidic protein, smooth muscle actin, muscle-specific actin, desmin, CD34, and epithelial membrane antigen) were made in 7 of the 11 cases for which the individual cell types needed to be validated further. The antibodies were produced by Cell Marque (Hot Springs, Ark), available as prepackaged kits made and distributed by Ventana Medical Systems, Inc (Tucson, Ariz). All imLipochoristomas of the Acoustic Nerve—Wu et al 1475

Figure 1. Lipochoristoma (arrow) of the right eighth cranial nerve appearing hyperintense on T1-weighted image (with gadolinium). Figure 2. Few mature adipocytes intimately admixed among neurons and myelinated nerve fibers (hematoxylin-eosin, original magnification 3100). Figure 3. Abundant mature adipocytes with nerve twigs (hematoxylin-eosin, original magnification 3100). 1476 Arch Pathol Lab Med—Vol 127, November 2003

Lipochoristomas of the Acoustic Nerve—Wu et al

Therapeutic Interventions

Table 1. Immunohistochemistry Panel

Marker

Antibody

Antibody Concentration, ng/mL

S100 Polyclonal, rabbit anti-human 10 GFAP Polyclonal, rabbit anti-human 10 SMA Monoclonal, clone alphaSM1 1 MSA Monoclonal, clone HHF35 0.5 Desmin Monoclonal, clone NCL-DE-R-11 1 CD34 Monoclonal, clone QBEnd/10 1 EMA Monoclonal, cone Mc5 1 * GFAP indicates glial fibrillary acidic protein; SMA, smooth muscle actin; MSA, muscle-specific actin; and EMA, epithelial membrane antigen.

munoreactions were performed according to the manufacturer’s guidelines, using the ES Automated Slide Stainer from Ventana in combination with Ventana DAB (diaminobenzidine) Detection Kits. Tissue slices mounted on polylysine-treated slides were deparaffinized (708 oven 3 2 hours), immersed in xylene, bar-coded for automated processing, and serially rehydrated in progressively weaker ethanol solutions (100% to 80%) and lastly in water. The slides were finally immersed in Ventana Wash Solution. Inhibitor solution was applied, followed by primary antibodies (Table 1) and biotinylated secondary antibodies. No pretreatment was needed for any of the immunoreactions. Avidin-streptavidin enzyme conjugate was then used, followed by detection with the stable chromogen DAB and copper DAB enhancer.

RESULTS Demographic and Clinical Information A summary of clinical information is shown in Table 2. The 11 patients were all adults, 23 to 50 years of age. Five tumors were located at the right vestibulocochlear nerve, and 6 were on the left side. All patients were symptomatic prior to the initial diagnosis of intracranial tumor. Hearing loss was the most common presenting sign or symptom, followed in decreasing order by tinnitus, vertigo, dizziness, headache, and hyperacusis. Symptoms of facial nerve involvement were not evident in any patient. None had clinical stigmata of neurofibromatosis. Imaging Studies Preoperative computed tomography and/or magnetic resonance imaging was performed on all patients. Lipochoristomas were, in general, hyperintense with brain on T1-weighted magnetic resonance images due to their fat content (Figure 1). However, in cases in which the tumor possessed a low fat content, the T1-weighted images were isointense or mildly hypointense with brain. Confirmation with fat suppression sequence, and if necessary, by computed tomography, was performed for verification. Lipochoristomas with high fat contents (hyperintense T1weighted images) also showed no appreciable enhancement after intravenous administration of a gadolinium chelate, while those with principally nonfatty elements enhanced appreciably after gadolinium chelate administration.

A variety of therapeutic interventions were performed according to the neurosurgical standard of care at the time of diagnosis (Table 2). These included complete excision of tumor and nerve, and partial excision of tumor with nerve-sparing biopsy of tumor only and intraoperative frozen section evaluation. Gross Anatomic Features Cochlear branch of CNVIII involvement was identified in all 11 patients (all of whom underwent craniotomy), while vestibular branch involvement was identified in 5 of the 11. All tumors were contained within the IAC, and no tumor extension to the CPA was found. Facial nerve involvement within the IAC was not evident. Maximal tumor size varied from 0.4 cm to 1.0 cm. No brain or inner ear involvement was evident in any patient. All grossly visualized tumors were shown to have arisen from within the cranial nerve, never a separate tumor impinging on the nerve branches. The tumors were pink, ivory/pale gray to tan-maroon, and had a soft to rubbery postfixation consistency. Microscopic and Immunohistochemical Features Other than the normal components of CNVIII, which include myelinated nerve branches, glial cells, neurons, and small-caliber thin-walled vessels, all lipochoristomas examined histologically possessed varied amounts of mature adipocytes (Figures 2 and 3). Other components of lipochoristomas, some verified by immunohistochemistry (Table 3), included tortuous and/or thick-walled vessels (Figure 4), smooth muscle cells not associated with vessels, bland mature hyalinized fibrous tissue (Figure 5), and skeletal muscle fibers (Figure 5). These nonnative components were present in varied proportions in different tumors and were not all present in all tumors (Table 3). The nonnative components always occurred in the myelinated neurons, none in the glial portions of CNVIII. A histomorphologic feature of importance in correlating with the gross appearance was the intimate intermingling of some native components of CNVIII (myelinated nerves and neurons) with the nonnative components (thickwalled vessels, smooth muscle cells, skeletal muscle fibers, and fibrous tissue). The nonnative components were not a demarcated/separate entity impinging on the native myelinated neurons and ganglions. Clinical Follow-up Evaluations All patients recovered well postoperatively. Patients were followed up to 7 years postoperatively. No recurrence was identified in follow-up radiographic studies of the patients with completely extirpated tumors. For patients whose tumors were partially excised or from which biopsies were taken, no radiographically detectable growth was evident. Varied signs and symptoms were noted at follow-up evaluations 1 or more years after the initial evaluation (Table 2). Patients who have had conservative management (partial resection or biopsy; Table 2,

← Figure 4. Tortuous, thick-walled vessels (hematoxylin-eosin, original magnification 3100). Figure 5. Skeletal muscle fibers adjacent to bland hyalinized fibrous tissue (hematoxylin-eosin, original magnification 3100). Arch Pathol Lab Med—Vol 127, November 2003

Lipochoristomas of the Acoustic Nerve—Wu et al 1477

Table 2. Clinical Presentation, Therapy, and Follow-up Evaluations Patient Age, y/ No. Sex

Presenting Symptoms (Duration)

Site

Treatment

1 33/M Unknown R/C 2 50/F D, HL (12 y) R/C 3 39/M T, HL (30 y) L/C, Ve 4 39/M T, HL (6 y) L/C, Ve 5 23/M Unknown R/C 6 46/M T, hyperacusis R/C, Ve 7 34/M HL (10 y) L/C, Ve 8 49/F T, HL (9 mo) R/C, Ve 9 27/F T, HL (8 mo) L/C 10 36/M HL (6 mo) L/C 11 43/M HL (6 mo) L/C * D indicates dizziness; HL, hearing loss; T, tinnitus;

Follow-up Evaluation

Excised C Unknown Excised C and Ve Complete HL, increased D, 6 y postop Excised C and Ve Complete HL, increased T, 4 y postop Excised C and Ve Complete HL, increased D and T, 1 y postop Excised C Unknown Excised C and Ve Complete HL, increased D and T, 1 y postop Excised Ve, partially excised C Baseline H, 3 y postop Excised Ve, partially excised C Baseline H, same T, 1 y postop Excised C None to date (,1 y postop) Excised C None to date (,1 y postop) Excised C None to date (,1 y postop) R, right; C, cochlear branch; Ve, vestibular branch; postop, postoperatively; and H, hearing.

Table 3. Histology and Immunohistochemical Results Patient No.

Tumor Size, cm

1 2 3

0.7 0.6 0.6

Histology

Immunohistochemistry

A, V, SM, SkM, F A, V, SM, F A, V, SM, F

None Positive for SMA in SM and V Positive for SMA in SM and V Negative for EMA 4 0.5 A None 5 0.4 A, SM, SkM, F Positive for SMA in SM Positive for MSA and desmin in SkM Negative for EMA 6 0.8 A, V, SM, F Positive for SMA in SM and V 7 0.4 A, SM, SkM, F Positive for MSA and desmin in SM and SkM Positive for CD34 in endothelium/lymphatics 8 0.4 A, V, F None 9 0.7 A, V, SM, SkM, F Positive for SMA in SM and V Positive for desmin in SkM and SM Positive for CD34 in endothelium/lymphatics 10 0.6 A, V, SM, F Positive for SMA in SM and V Positive for desmin in SM Negative for MSA Positive for CD34 in endothelium/lymphatics 11 1.0 A, V None * A indicates adipocytes; V, thick-walled vessels; SM, smooth muscle; SkM, skeletal muscle; F, fibrous tissue; SMA, smooth muscle actin; and EMA, epithelial membrane antigen. Note.—Immunostaining for S100 was performed also on the specimens from patients 2, 3, 5 through 7, 9, and 10, all showing positive staining in the nerves. Staining for glial fibrillary acidic protein was performed on patient 9 and was positive in the glial portion.

patients 7 and 8) had similar or better overall outcome than patients who had followed a more aggressive protocol (complete excision, patients 2, 3, 4, and 6). COMMENT The demographic and clinical findings from our cases revealed features similar to those of prior studies.5–11 Imaging findings were similar as well,1,5,6,8–12 namely, varying degrees of hypoattenuation on computed tomography and hyperintensity on T1-weighted magnetic resonance imaging, depending on the predominance of adipose component. Facial nerve involvement within the IAC was a frequent observation in past case reports and reviews,5 but was not seen in our patient population. Therapeutic intervention for lipochoristomas of CNVIII has changed during the last decade, from early complete tumor excisions in the past10 to a more recent conservative approach for nerve preservation,1,5–7 owing to the slow growth of the tumor and the lack of success in hearing conservation by surgical resection. Presently at our institution, a conservative treat1478 Arch Pathol Lab Med—Vol 127, November 2003

ment strategy is applied, after establishing the diagnosis with reasonable confidence (magnetic resonance imaging with fat suppression sequence or open biopsy with intraoperative frozen section evaluation). Indeed, follow-up evaluations in 6 patients up to 7 years after the initial diagnosis allowed us to verify the validity of this approach with moderate certainty, although statistical rigor could not be applied owing to the limited number of patients available for this study. The normal gross and microscopic anatomy of CNVIII is well known,13 with the cochlear and vestibular branches each measuring 1.7 to 2.0 cm in length.14 The native elements include glia, myelinated nerve, neurons (ganglions), and small thin-walled vessels. A neuroglial-neurolemmal (glial-schwannian) junction is present, located 1.0 to 1.3 cm from the brainstem in men and 0.7 to 1.0 cm in women, with the vestibular branch junction usually a little more distally located than the cochlear branch.15 The nonnative elements observed in our study include adipocytes, thick-walled tortuous vessels, smooth muscle fibers not asLipochoristomas of the Acoustic Nerve—Wu et al

sociated with vessels, skeletal muscle fibers, and fibrous tissue, all of which appear mature and bland, showing no histomorphologic evidence of atypicality, dysplasia, or malignancy. All nonnative elements, except for skeletal muscle fibers, have been described in past case reports.1,6,10,16–21 Metaplastic cartilage was also observed in 1 case reported previously,11 but was not present in our specimens. Mature adipocytes are the main element consistently present in all tumors, both in this study and in all past reports. Therefore, the descriptive name lipochoristoma is befitting for this tumor in terms of histopathology. The histogenesis of lipochoristomas is of great academic interest. The consensus within the neurosurgical/clinical community is that lipochoristomas of the vestibulocochlear nerve are the maldevelopmental derivatives of the meninx primitiva,1,2,12 the primitive meningeal mesenchyme believed to be derived from either or both of the mesectoderm (neural crest) and mesoderm.22 Abnormal resorption of the inner meninx during embryogenesis is conjectured to be the epigenesis of intracranial lipomas.2,12 However, the validity of this hypothesis has yet to be shown in cases of lipochoristomas of CNVIII. Recently, retrovirus-mediated lineage analyses in the inner ears of chickens were performed.23,24 Injections of markers for mesenchymal clones on embryonic days 2.5 through 5.5 identified primitive nonneuronal cells with fibroblastic/schwannian and endothelial cord morphologies within the CNVIII ganglion and nerve. This finding indicates that mesenchymal elements are integral components of CNVIII, present very early in embryonic development. It is tempting to surmise that lipochoristomas in human CNVIII may actually have arisen from these embryologic remnants. Gross tumor features and histopathologic evidence in favor of the latter hypothesis include the intimate intermingling of nonnative mesenchymal components (adipocytes, fibrous tissue, smooth muscle, and skeletal muscle) with the native neural components (myelinated nerve and neurons). This arrangement indicates that the tumor is most likely endogenous to the nerve and accounts for the failure of hearing conservation by surgical resection. If lipochoristomas of CNVIII are derived from primitive meningeal remnants, they should occur as extranervous tumors impinging on CNVIII, with more clearly demarcated tumor borders, not unlike the gross and histomorphologic features of a peripheral schwannoma. The latter is clearly not evident from this or prior studies.

Arch Pathol Lab Med—Vol 127, November 2003

In summary, lipochoristomas of CNVIII are slow-growing tumors, likely developmental anomalies endogenous to the CNVIII parenchyma instead of meninx-derived, and should be treated conservatively once a definitive diagnosis is made, either from intraoperative frozen section evaluation or imaging studies. References 1. Zamani AA. Cerebellopontine angle tumors: role of magnetic resonance imaging. Magn Reson Imaging Clin N Am. 2000;8:491–512. 2. Lalwani AK. Meningiomas, epidermoids, and other nonacoustic tumors of the cerebellopontine angle. Otolaryngol Clin N Am. 1992;25:707–728. 3. Martuza RL, Parker SW, Nadol JB, Davis KR, Ojemann RG. Diagnosis of cerebellopontine angle tumors. Clin Neurosurg. 1985;32:177–213. 4. Brackman DE, Bartels LJ. Rare tumors of the cerebellopontine angle. Otolaryngol Head Neck Surg. 1980;88:555–559. 5. Tankere F, Vitte E, Martin-Duverneuil N, Soudant J. Cerebellopontine angle lipomas: report of four cases and review of the literature. Neurosurgery. 2002; 50:626–632. 6. Zimmermann M, Kellermann S, Gerlach R, Seifert V. Cerebellopontine lipoma: case report and review of the literature. Acta Neurochir (Wien). 1999;141: 1347–1351. 7. Bigelow DC, Eisen MD, Smith PG. Lipomas of the internal auditory canal and cerebellopontine angle. Laryngoscope. 1998;108:1459–1469. 8. Greinwald JH Jr, Lassen LF. Lipomas of the internal auditory canal. Laryngoscope. 1997;107:364–368. 9. Inoue T, Maeyama R, Ogawa H. Hemifacial spasm resulting from cerebellopontine angle lipoma: case report. Neurosurgery. 1995;36:846–850. 10. Saunders JE, Kwartler JA, Wolf HK, Brackmann DE, McElveen JT. Lipomas of the internal auditory canal. Laryngoscope. 1991;101:1031–1037. 11. Christensen WN, Long DM, Epstein JI. Cerebellopontine angle lipoma. Hum Pathol. 1986;17:739–743. 12. Truwit CL, Barkovich AJ. Pathogenesis of intracranial lipoma: an MR study in 42 patients. AJR Am J Roentgenol. 1990;155:855–864. 13. Gruskin P, Carberry JN. Pathology of acoustic tumors. In: House WF, Luetje CM, eds. Acoustic Tumors. Baltimore, Md: University Park Press; 1979:85–148. 14. Pool LJ, Pava AA, Greenfield EC. Acoustic Nerve Tumors: Early Diagnosis and Treatment. 2nd ed. Springfield, Mass: Charles C Thomas; 1970. 15. Schuknecht HF. Pathology of the Ear. Cambridge, Mass: Harvard University Press; 1974. 16. Cohen TI, Powers SK, Williams DW. MR appearance of intracanalicular eighth nerve lipoma. AJNR Am J Neuroradiol. 1992;13:1188–1190. 17. Kawaguchi S, Sakaki T, Hirabayashi H, Hashimoto H, Shimogawara T. Eighth cranial nerve lipoma manifesting as intractable vertigo. Neurol Med Chir (Tokyo). 1995;35:818–821. 18. Mattern WC, Blattner RE, Werth J, Shuman R, Bloch S, Leibrock LG. Eighth nerve lipoma. J Neurosurg. 1980;53:397–400. 19. Olson JE, Glasscock ME, Hill Britton B. Lipomas of the internal auditory canal. Arch Otolaryngol. 1978;104:431–436. 20. Rosenbloom SB, Carson BS, Wang H, Rosenbaum AE, Udvarhelyi GB. Cerebellopontine angle lipoma. Surg Neurol. 1985;23:134–138. 21. Smith MM, Thompson JE, Thomas D, et al. Choristomas of the seventh and eighth cranial nerves. AJNR Am J Neuroradiol. 1997;18:327–330. 22. O’Rahilly R, Mueller F. The meninges in human development. J Neuropathol Exp Neurol. 1986;45:588–608. 23. Fekete DM, Wu DK. Revisiting cell fate specification in the inner ear. Curr Opin Neurobiol. 2002;12:35–42. 24. Lang H, Fekete DM. Lineage analysis in the chicken inner ear shows differences in clonal dispersion for epithelial, neuronal and mesenchymal cells. Dev Biol. 2001;234:120–137.

Lipochoristomas of the Acoustic Nerve—Wu et al 1479