Annals nf Otiitogv. Rhiiuilogy & Lanniialany 115(2): 124-127.© 2006 Annals Publishing Company. All righls reserved.

Magnetic Resonance Imaging Findings in the Evaluation ofTraumatic Anosmia

Jeffrey B. Wise, MD; Gul Moonis, MD; Natasha Mirza, MD

Objectives: Head trauma is a common cause of anosmia, but diagnosis is typically late, owing to more life-threateningsequelae of the injury. Herein, we describe our workup for a case of traumatic anosmia and lhe magnetic resonanceimaging (MRI) findings both at the time of injury and at the 18-month follow-up.

Methods: We present a case report and a review of the literature.

Results: A 33-year-old woman presented to our institution with achief complaint of loss of smell and taste following anoccipital blow to her head that occurred when she was hit by a car while riding a bicycle. We present the findings of MRIperformed at the time of the injury and at the 18-month follow-up. We describe the clinical progression of her disease.from symptoms of parosmic and phantosmic episodes accompanied by dysgeusia to total anosmia at the 18-monthfollow-up.

Conclusions: We advocate the use of MRI, coupled with otolaryngology consultation and formal olfactory testing, inthe diagnosis, management, and counseling of patients with anosmia sustained from head injury.

Key Words: anosmia, magnetic resonance imaging, trauma.

INTRODUCTIONHead trauma is a common cause of atiosmia. but

diagtiosis is typically late, owitig to the attention giv-en to more life-threatening sequelae of the injury. Infact, studies have cited olfactory dysfunction as oc-curring in 5% to 10% of all head traumas.'-^ Histo-pathologic changes in posttraumatic anosmia havebeen described, as well a.s basic pathophysiologicmechanisms for this type of injury, the most impor-tant of which involves the shearing of olfactory filaat their site of penetration of the cribriform plate fromthe nose.-̂ Trauma severity and site of injury, speci-fically, lateral and occipital blows, are associated witha higher degree and incidence of anostiiia. Prior stud-ies have noted the rate of at least partial recovery offunction on objective testing in patients who have headinjuries as being in the range of 30% to 40% of pa-tients, although far fewer patients demonstrate truesubjective improvement."^ Conventional tests of smelland taste futiction, such as the University of Penn-sylvania Smell Identification Test, are helpful in ob-taining a diagnosis of olfactory loss, although thesetests are limitedby their inability to offer insight intothe mechanism of injury (ie, nasal obstruction ver-sus central brain trauma). Magnetic resonance irnag-ing (MRI) with and without contrast exhibits char-

acteristic findings for patients in whom anosmia de-velops as a result of head injury, and additionally pro-vides information regarding pathophysiology. Here-in, we describe our workup for a case of traumaticanosmia and the MRI findings both at the time ofinjury and at the 18-month follow-up. We advocatethe use of MRI, coupled with otolaryngology con-sultation and formal olfactory testing, as a means todiagnose and manage cases of anosmia caused byhead injury and for use in counseling patients withthis disorder.

CASE REPORTA 33-year-old woman presented to the University

of Pennsylvania Smell and Taste Center with a chiefcomplaint of loss of smell and taste following an oc-cipital blow she received to her head when she washit by a car while riding a bicycle. At the time of theaccident, she did not lose consciousness, but she re-ported a loss of smell within a few hours of her in-jury. In the .subsequent weeks, she noticed parosmicand phantosmic episodes accompanied by dysgeusia.These unpleasant sensations occasionally were in-duced by specific odorants and at other times arosespontaneously. She denied a history of sinonasal dis-ease, prior facial trauma, or tobacco use prior to herinjury. Her only medications were oral contracep-

Froin lhe Departments of Otorhinolaryngology-Head and Neck Surgery (Wise. Mir/a) and Radiology (Moonis), University of Penn-sylvania, Philadelphia, Pennsylvania.Presented at the meeting of the American Rhinologic Society. Nushville. Tennessee. May 2-3, 2003.Correspondence: Jeffrey B. Wise. MD, Dept of Otorhinolaryngology-Head and Neck Surgery, Hospital of the University of Penn-sylvania. 5-Riivdin. 3400 Spruce Si. Philadelphia. PA 19104.

124

Wise et al, Magnetic Resonance Imaging in Traumatic Anosmia 125

Fig 1. Magneiii: resonance images obtained immediately after injury.Arrows show inLiltifocal regions of hyperintensity in bilateral orbitalfrontal regions, compalible with hcmorrhagic contusions involving bi-lateral gyru.s recti and olfactoiy bulbs and tracts, A) Sagittal T1 -weightedimage, B) Axial FLAIR (fluid-attenuated inversion recovery) image atlevel of olfactory tracts. C) Coronal TI -weighted image, enhanced withgadolinium, of brain at level of cribriform plate.

tive pills and acetaminophen used occa.sionally. Acomplete head and neck examination was performedand showed no abnormalities. Nasal examination andendoscopy demonstrated a mild left-sided septal de-viation with no evidence of mueopus or polyps pres-ent. The olfactory clefts were clear on both sides.

An MRI scan of the brain was performed in theemergency depatitnent within 4 hours of the injury.Sagittal Tl-weighted imaging followed by axialFLAIR (tluid-attenuated inversion recovery) imag-ing revealed multifocal regions of hyperintensity inthe bilateral orbital frontal regions (Fig I). These re-gions also demonstrated TI hyperintensity compa-

tible with hemorrhagic contusion. Specifically, therewas bilateral involvement of the gyrus recti and theolfactory bulbs and tracts. No additional cranial ab-nomialities were noted.

The results of the University of Pennsylvania SmellIdentification Test, a 40-item standardized forced-choice microencapsulated test, were indicative of to-tal bilateral anosmia {score 13/40). Performance onthe 12-item Odor Meinory Test was essentially atchance level (left 4/12, right 5/12). Objective testingof taste function was then performed. The percentcorrect identification of whole-mouth suprathresholdsweet, sour, bitter, and salty tastant concentrations

126 Wise et al. Magnetic Resonance Imaging in Traunialic Anosmia

Fig 2. Follow-up axial FLAIR magnetic resonance imag-ing, performed at 18 months, demonstrates bilateral inter-val re.solulion of abnormal signal (hemorrhagic contu-sions) in orbilal frontal regions, and encephalomalacicchange in olfactory apparatus and gyrus rectus (arrow).

was within normal limits (40/40). Her score on theBeck Depression Inventory-II {score 3) was unre-markable.

At 18 months' follow-up, the patient reported noreturn ot olfactory function. In fact, along with thedevelopment of total anosmia, her phantosmic andparosmic episodes also disappeared. The findings otiphysical examination remained unchanged. An MRIscan performed at the 18-month follow-up was no-table for interval resolution of abnormal signal in theorbital frontal regions. There was gliosis and malacicchange noted in the olfactory apparatus and gyrusrecti as a result of the prior trauma (Fig 2).

DISCUSSION

Patients who lose their sense of smell are oftenconfronted with a multitude of pragmatic and lifestylechallenges, ranging from employment disability (eg.natural gas workers, cooks, firefighters) to a failureto appreciate pleasing odors and tastes. In their se-ries of 750 patients who presented with smell andtaste disorders. Deems et al cited that the majority ofpatients who presented with complaints of loss oftaste actually suffered from olfactory function, "re-flecting decreased effectiveness of retronasal stimu-lation of the receptors by flavor molecules arisingfrom food during deglutition.'" '̂P-*^^ '̂ Furthermore,

28.5% of the patients in their group who had smelland taste disorders scored in the range of mild tosevere depression as measured by the Beck Depres-sion Inventory.'̂

Although no universal consensus has been reached.it is generally agreed upon that total anosmia occursin 5% to 10% of head traumas, and that as many as30% to 40% of head trauma victims experience atleast partial loss of smell function (hyposmia).*'Al-though histopathologic evaluation of human olfac-tory tissue following head trauma has been limited.3 primary mechanisms of injury have been proposed:!) shearing of the olfactory fibers from the skull baseat the cribriform plate, 2) conductive loss of olfac-tion via nasal distortion or obstruction, and 3) injuryto central olfactory tracts from brain trauma.^

Approximately one third of patients with traumaticanosmia recover at least partial function of their senseof smell within a time period of I year of their in-jury. Animal studies support the notion that neuronalolfactory tracts are capable of regeneration, as evi-denced by morphological and electrophysiologicalexamination. Kern et al.^ in an immunocytochemi-cal and histopathologic analysis of the olfactory bulbsand tracts of a patient with posttraumatic olfactorydysfunction, postulated that some degree of neuronalregeneration is possible in humans; however, theyreport that fibrosis and scarring of the cribriform re-gion often limits the reconnection between the ax-ons of regenerating neurons and the olfactory bulb,

Dysosmia. defined as the distortion or absence ofsmell, has been observed in a high proportion of pa-tients with head injury. This disorder of smell can betemporary or permanent. Doty et aH noted in a se-ries of 268 patients with head trauma that more thanone third had experienced at least some degree ofdysosmia after sustaining their injuries. The patientcomplaint of dysgeusia in our case presentation mostlikely manifested from dysosmia. rather than tastedysfunction per se. This was verified by her perfectscore on suprathreshold taste testing (40/40). As de-scribed, her dysgeusic and dysosmic episodes gradu-ally disappeared and were replaced by total bilateralanosmia. This clinical course, along with the experi-ence of Doty et al. suggests that dysosmia more likelyrepresents a neurodegenerative process, rather thana regenerative one. Kern et aP reported a patient whoexperienced dysosmic episodes following a 2-monthperiod of total anosmia, which was interpreted to rep-resent at least partial regeneration of olfactory tracts.By correlating this time course to that of previouslystudied animal models. Kern et al gave at least indi-rect evidence that some functional reconnection tothe olfactory bulb can occur.

In the radiographic evaluation of bead injuries that

Wise et al. Magnetic Resonance Imaging in Traumatic Anosmia 127

cause anosmia, the inferior frontal lobe appears tobe a common site for cortical contusion. This injuryis due to the direct force of impact during rapid de-celeration of cortical surfaces with edges of the in-ner table of the skull along the tloor of the anteriorcranial fossa. Because gray matter is much more vas-cular than white matter, contusions are much morelikely to be hemorrhagic than diffu.se axonal injury.Because of the higher resolution of MRI images andthe existence of special su.sceptibility sequences sen-sitive to the presence of blood products, MRI is thepreferred imaging technique for evaluating hemor-rhagic contusions and diffuse axonal injury. Addition-al areas of abnormality that can be missed on com-puted totnographic images, including nonhemorrhag-ic foci of injury, will be better detected on MRI scans.However, computed tomographic scanning remainssuperior to MRI in the assessment of bony injury. Inthe acute setting after hemorrhagic contusion andshearing injury to the inferior frontal lobe and olfac-tory apparatus, one observes hyperintensity on Tl-weighted images (indicating hemorrhage) and T2-weighted images (indicating edema). In the chronicsetting, there is resolution of hemorrhage and edema,but residual gliosis may be seen manifested by focalT2 hyperintensity and loss of brain substance or at-rophy. There tnay be residual hemosiderin deposi-tion, best seen as signal dropout on gradient echoimages. Additionally, the dura may adhere to the un-derlying brain by fibroglial scarring, which may bea source of seizure in posttraumatic cases.^

Magnetic resonance imaging has demon.strated ab-normalities in patients with posttraumatic olfactorydysfunction at a very high rate (88%), predominantlyin the olfactory bulbs and tracts and the inferior fron-tal lobes.'"'"'However, those studies were performedin the late stage of posttraumatic olfactory dysfunc-

tion (>3 months after injury). As discussed, posttrau-matic anosmia is most often seen in the setting offrontal contact injury. Fractures of the skull or faceare commonly associated with bilateral posttraumaticanosmia, and are shown in MRI and computed tomo-graphic studies. Yousem et al'̂ "^ found that the vol-ume of the olfactory bulb and tracts was smaller inpatients with po.sttraumatic anosmia than in patientswith residual smell function. They hypothesized thatthe source of olfactory deficit after trauma may berelated to shearing of olfactory nerves at the cribri-form plate. Direct injury to the olfactory bulb or tract.,intracerebral hematoma compressing these structures,and injury to the septal nuclei have also been impli-cated in posttraumatic olfactory deficits."'^ The lo-cation and imaging appearance of MRI abnormali-ties in our case are characteristic of hemorrhagic con-tusions involving the olfactory apparatus and the in-ferior frontal cortex. Additionally, it is conceivablethat shearing of the olfactory nerve fibers at the siteof their passage through the cribriform plate mayhave been contributory to this patient's symptoms.

In the setting of head trauma, smell dysfunction isoften underdiagnosed because of the presence of orsearch for more life-threatening injuries. We proposethat conventional MRI, performed both at the timeof injury and in follow-up, may alert the clinician tothe possibility of olfactory dysfunction, thus allow-ing for early consultation by an otolaryngologist andearly use of objective smell testing. Furthermore, ithas been shown that MRI findings obtained at thetime of injury and in follow-up allow insight intothe cause and progression of smell dysfunction, thusserving to guide patient counseling and managementof this common yet troublesome sequela of head in-jury.

REFERENCES1. Sumner D. Posl-traumatic anosmia. Brain I964;87:1O7-

20.2. Yousem DM, Geckle RJ, Bilker WB, Kroger H. Doty

RL. Posttraumatic smell loss: relationship of psychophysicaitests and volumes of the olfactory hulhs and tracts and the tem-poral lobes. Acad Radiol 1999;6:264-72.

3. Kern RC. Quinn B, Rosseau G, Farbman Al. Posl-trau-matic olfactory dysfunction. Laryngoscope 2000; 110:2106-9.

4. Doty RL. Yousem DM, Pham LT. Kreshak AA, GeckleR, Lee WW. Olfactory dysfunction in patients with head trauma.ArchNeurol 1997:54:1131-40.

5. Deems DA. Doty RL. Settle RG, et al. Smell and tastedisorders, a study of 750 patients from the University of Penn-sylvania Smell and Taste Center. Arch Otolaryngol Head NeckSurg 1991:117:519-28.

6. Costanzo RM, Becker DP. Smell and taste disorders inhead injury and neurosurgery patients. In: Meiselman HL, RivlinRS, eds. Clinical measurement of taste and smell. New York,NY: Macmilian. 1986:565-78.

7. Costan/o R. Zasler ND. Head trauma. In: Getchell TV,Doty R, Bartoshuk L, Snow J. eJs. Smell and taste in healthand disease. New York, NY: Raven Press, Iy91:711-30.

8. Grossman Rl. Yousem DM. Neuroradiology — the requi-sites. St Louis, Mo: Mosby, 1994:149-70.

9. Yousem DM. Geckle RJ. Bilker WB. McKeown DA.Doly RL. Postlraumatic olfactory dysfunction: MR and clini-cal evaluation. AJNR Am J Ncuroradiol 1996:17:1171-9.

10. Yousem DM, Oguz KK. Li C. Imaging of the olfactorysystem. Semin Ultrasound CT MR 2001:22:4.56-72.

11. Levin HS. High WM. Eisenberg HM. Impairment of ol-factory recognition after closed head injury. Brain 1985:108:579-91.

12. Potter H. Butters N. An assessment of olfactory deficitsin patients with damage lo prelronial cortex. Neuropsychologia1980:18:621-8.

13. Zusho H. Posltraumaiic anosmia. Arch Otoiiiryngol 1982;108:90-2.