European Review for Medical and Pharmacological Sciences 1999; 3: 135-138 Cholesterol of the petrous apex A post mortem study on temporal bones

F. SALVINELLI, F. GRECO, M. TRIVELLI, F.H. LINTHICUM JR*

Institute of Otolaryngology, “Campus Bio-Medico” University - Rome (Italy) *Department of Histopathology, “House Institute” - Los Angeles, CA (USA)

Abstract. – (OM) is an infec- matory reaction may be the result of the con- tion localized in the : mastoid, middle tamination of the middle ear by inhaled am- ear cavity, . The classification of niotic squames. In these cases the histiocytes OM includes otitis media with effusion, otititis media without effusion, and chronic otitis media. reacting to the foreign material fuse to form A rare complication of chronic otitis is choles- giant cells. Haemorrhage is a common result terol granuloma of the petrous apex. It may de- of the congestion of otitis media. It may lead velop in any aerated portion of the temporal to cholesterol granuloma. bone but most commonly develops when a pathologic process obstruct the air cell tracts to the petrous apex preventing normal aeration. Key Words: Materials and Methods Cholesterol crystals, Giant cells. We have studied the histopathological changes in temporal bones of a deceased in- dividuals, with concomitant cholesterol gran- uloma of the left petrous apex. This patient was a donor and agreed during Introduction his life to donate post mortem his temporal bones to the House Ear Institute as a contri- The acute otitis is characterized by a severe bution to a better knowledge of temporal congestion of the mucosa of the middle ear bone diseases. and the tympanic membrane. It is not gener- We have removed the temporal bones in ally realized that congestion of the mucosa is our usual way1-3. frequently also a marked feature of chronic otitis media. The fluid portion of blood, plas- ma, may leave a deposit of fibrin in the tis- sues. A fluid in the middle ear cavity Results is frequently a prominent component of the inflammatory reaction – a specific form of the Yellow nodules are found in the tympanic disease known as otitis media with effusion. cavity and mastoid in many cases of chronic In these cases mucus may be secreted by otitis media. These are composed microscopi- newly formed glands in the middle ear mu- cally of cholesterol crystals (dissolved away to cosa and may contribute to the fluid “exu- leave empty clefts in paraffin- embedded his- date”. In acute neutrophils are tological sections) surrounded by foreign prevalent. In chronic inflammation histio- body type giant cells and other chronic in- cytes (derived from monocytes of the blood), flammatory cells (Figure 1). Such cholesterol lymphocytes and plasma cells (derived from are almost always found in the lymphocytes), are the characteristic infiltrate. midst of haemorrhage in the middle ear mu- Organisms are seen very rarely in histological cosa. That haemorrhage is the cause of cho- sections of acute or chronic inflammation of lesterol granuloma has been denied by Sadè4, the middle ear. In newborn infants an inflam- who thought that the blood seen in the biop-

135 F. Salvinelli, F. Greco, M. Trivelli, F.H. Linthicum Jr

These findings are compatible with an origin of the lipid material in cholesterol granuloma from red cell membranes, in which cholesterol exists mainly in the free, not esterified state. Cholesterol granuloma is the most com- mon cystic lesion of the petrous apex, but still a rare one, occurring 30 times less frequently than acoustic neuroma9. It may develop in any aerated portion of the temporal bone but most commonly occurs in the mastoid air cells distant to a lesion that prevents normal aeration. Cholesterol granuloma of the petrous apex probably develops when a Figure 1. Cholesterol granuloma showing granulation pathologic process or trauma obstructs the air tissue (1) containing cholesterol clefts lined with foreign cell tracts to a well-pneumatized petrous body giant 4 cells (2). Loculated cyst (3) also contains apex. cholesterol clefts and . × 117 The treatment for cholesterol granuloma of the temporal bone is drainage and re-es- tablishment of adequate aeration to the in- sies of such lesions was the result of surgery. volved area. The cyst wall is composed of a fi- There can be no doubt, however, that haem- brous connective tissue. It is free of keratiniz- orrhage is a feature of cholesterol granuloma. ing squamous epithelium that characterizes Red cells are localized to the granuloma and , and complete removal of the are not usually found in the tissues. cyst is not necessary. Haemorrhage is also a frequent concomitant Infralabyrinthine, infracochlear and trans- of cholesterol granuloma in the maxillary sphenoidal approaches are most commonly antrum, and in thyroid adenomas. Cholesterol chosen for drainage of cystic lesions of the granuloma in the mastoid air cells must be petrous apex in an ear with serviceable hear- distinguished from lipid deposits of hyper ing. These lesions are frequently detected at cholesterolaemic xanthomatosis. an asymptomatic stage with today’s imaging techniques. Because the natural history of small benign cystic lesions is not well docu- mented, surgical drainage should be reserved Discussion for patients with larger lesions or with symp- toms, including , visual changes, diplopia, Cholesterol granuloma has been produced , vertigo, or facial nerve weak- experimentally by obstructing natural air ness. for patients without serviceable hearing, filled spaces in bone the humerus of cock- these lesions should be drained through a erels5 and the Eustachian tube of squirrel translabyrinthine approach10. Because other monkeys6-7. In the study of Thomas et al6 vital structures may be affected by enlarge- haemorrhages were observed to accompany ment of the cyst, delaying surgery in sympto- the cholesterol granulomas. matic patients provides no advantage. Zini These experiments, while suggesting that suggests the occipito-temporal approach as a lowered air pressure in the middle ear might direct access to the anterior and posterior be associated with cholesterol granuloma, do petrous apex without opening the dura, thus not exclude the possibility of haemorrhage preserving the facial and the cochlear- resulting from lowered pressure being a pre- vestibular functions11. cursor. Preoperative evaluation of these patients is Sadè and Teitz9 found the lipid in choles- based upon their symptoms. Patients present- terol granulomas of the middle ear to be ing with hearing loss are evaluated initially mainly cholesterol with only very small with audiometric testing, including air, bone amounts of cholesterol esters. In serum the and speech reception thresholds and speech reverse is the case: a high proportion of cho- discrimination scores. Electronystagmogra- lesterol ester is present, but little cholesterol. phy is performed in patients who complain of

136 Cholesterol granuloma of the petrous apex. A post mortem study on temporal bones imbalance or vertigo. In patients with other- shorter periods of time seem to have a better wise normal results on physical examination, prognosis and fewer long-standing deficits asymmetric hearing is next evaluated with au- then those affected longer. Patients are re- ditory brainstem response testing. If these re- minded this is a drainage procedure whose sults are abnormal, an MRI scan is indicated. goal is to decompress the lesion and provide In patients with cranial nerve involvement an aereated cavity, if possible. The goal is not other than the eight nerve, with asymmetric the removal of the lesion, and close follow-up hearing, auditory brainstem response testing may be necessary. Recurrence of the lesion is not performed, and the physician proceeds secondary to inadequate drainage is usually directly to an MRI scan. The advent of mag- heralded by the return of preoperative symp- netic resonance imaging has helped, showing toms. Follow-up MRI frequently reveals a a high signal in T1 and T2 without enhance- cholesterol granuloma cyst that remains full ment after the administration of intravenous of fluid, but the T1-weighted image is hy- paramagnetic contrast10. pointense, compared with the preoperative Patients who have normal hearing but have hyperintense image on T1 views. A return of other cranial nerve deficits that may be refer- hyperintensity on the T1 image suggest inad- able to the petrous apex may be screened equate drainage in a symptomatic lesion9. with either a high- resolution, thin-section CT Advances in radiologic imaging during the of the temporal bone or an MRI with past decade have made it possible to reliably gadolinium. If an abnormality is found, all differentiate lesions of the petrous apex pre- patients undergo air, bone, and speech recep- operatively; The development of CT scanning tion thresholds and speech discrimination au- was the first major step in imaging the tempo- diometric testing before surgery to document ral bone since the development of polytomog- hearing levels before a procedure that jeopar- raphy. CT scanning gives the surgeon the abili- dizes hearing. ty to visualize the size of the lesion and its re- Preoperatively, patients are counseled to lationship to vital structures, including the in- expect resolution of pain, if present, and the ternal auditory canal, cochlea, vestibular possibility of improvement in cranial nerve labyrinth, carotid artery, and jugular bulb. It function if it is decreased preoperatively. also helps characterize the border of the lesion Cranial nerves that have been affected for as expansive or invasive, which may differenti-

Table I. Mean values and significativity of variables.

MRI Lesion Computed tomografy T1 T2 Enhancement

Retained mucus Normal bony architecture, Hypointense Hyperintense No nonenhancing Mucocele Hypodense, expansile smooth Hypointense Hyperintense No border, nonenhancing Asymmetric Normal bony architecture Hyperintense Hypointense No pneumatization nonenhancing Cholesteatoma Loss of normal air cells, Hypointense Hyperintense No nonenhancing, isointense with CSF Cholesterol Expansile smooth border, Hyperintense Hyperintense No granuloma occasional rim enhancement, isointense with brain Metastatic lesion Destructive, indistinct border Isointense Hyperintense Yes Chordoma Aggressive bone destruction, Isointense: 75% Hyperintense Yes calcification Hypointense: 25% Chondroma Aggressive bone destruction, Hypointense to Hyperintense Yes calcification isointense Chondrosarcoma Aggressive bone destruction, Hypointense to Hyperintense Yes calcification isointense

137 F. Salvinelli, F. Greco, M. Trivelli, F.H. Linthicum Jr ate between benign lesions and malignant neo- 4) SADÈ J. The blue drum (idiopathic hematympa- plasms. MRI of the temporal bone added the num) and cholesterol granulomas. ln: Sadè J, ed. Secretory otitis media and its sequelae. New capability of characterizing the substance of York: Churchill Livingstone, 1979: 12-22. the lesion rather than its effect on bony inter- 5) BEAUMONT GD. The effect of exclusion of air from faces, allowing the surgeon to distinguish be- pneumatised bones. J Laryngol Otol 1966; 80: tween mucus, fat, cholesterol granuloma, 236-249. cholesteatoma, and neoplasm. The combina- 6) THOMAS 5, MAIN TS, SHIMADA T, L IM DJ. Experimental tion of CT, with its superior bone imaging al- cholesterol granuloma. Arch Otolaryngol 1970; gorithms, and MRI, with its enhanced tissue 91: 356-359. imaging capabilities allows the surgeon to dif- 7) KUIPERS W, V AN DEN BECK JMII, WILLARD ECT. The ef- ferentiate accurately and reliably between be- fect of experimental tubal obstruction on the mid- nign cystic lesions, normal anatomic variants, dle ear. Acta Otolaryngol (Stockh) 1979; 87: 345- and neoplastic lesions of the petrous apex. 352. 8) SADÈ I, TEITZ A. Cholesterol in cholesteatoma and in the otitis media syndrome. In: Sadè J, ed. Choles- teatoma and mastoid surgery. Proceedings of the References second international conference. Amsterdam: Kugler 1982: 125-132.

1) SALVINELLI F, G RECO F, T RIVELLI M, LINTHICUM F. Acute 9) BRACKMANN D SHELTON C ARRIAGA M. Ear surgery. otitis media. Histopatological changes. Eur Rev Philadelphia: WB Saunders 1996: 1-885. Med and Pharmacol Sci (in press). 10) HOUSE WF, DE LA CRUZ A, HITSELBERGER WE. Surgerys 2) IURATO S, BREDBERG G, BOCK G. Functional of the skull base: Transcochlear approach to the histopathology of the human audio-vestibular or- petrous apex and clivus. Otolaryngology 1978; 86: gan. Euro-Data-Hearing. Acta Otolaryngol 1983; 770-779. 95: 705-708. 11) ZINI C, GANDOLFI A, PIAZZA F, A VENDANO ARAMBULA J, 3) KARNOVSKY MJ. A formaldehyde-glutaraldehyde fix- PEREZ RAFFO G, VEGA FEIJOO S. Granuloma colesterini- ative of high osmolality for use in electron mi- co de la punta del penasco. Acta Otorrinolaringol coscopy. J Cell Biol 1965; 27: 137A-138A. Esp 1997; 48: 496-500.

138