Handbookof Olfactionand Gustation SecondEdition Revisedand Expanded

editedby RichardL. Doty Uniaersityof Pennsylaania. Philadelfihia,Pennsylaania, U. S.A.

Mancrr Drrxrn, INc. Nsw Yom. Bespr

Copyright @ 2003 by Marcel Dekker, Inc. All Rights Reserved. MARCEL

MaRCEL DnKKER,INC. DEKKER

Registernow on dekker.comanil receivea discountcoupon towards your next Dekkerpurchase. Registered users enjoy theseadvantages: y' Previewcontent online for free / QuicHy accessyour subscriptionsand downloads / Signup for email notificationsabout new productsand specialoffers in your subjectof interest Sendyour hook order to your regular hook supplier or directly to your nearestMarcel Dekken lnc. office:

Pleaseinclude quantity, ISBN, author, title, shipping and hilling addresses,telephone number/fax numher/e-mailaddress, method of shipmentand purchase ordernumbers or credit card information.

To purchaseoffprints of chaptersthat appearin any Marcel Dekker,Inc. book or reprints of articlesthat appearin any Marcel Dekker,Inc. journal: [email protected]

To inquire aboutspecial sales and bulk purchasesof Marcel Dekker,Inc. booksor journals: [email protected] 2B

Evaluationof OlfactoryDeficits by StructuralMedical lmaging

Cheng[i, RichardL. Doty, and David W. Kennedy lJniversityof Pennsylvania,Philadelphia, Pennsylvania, U.S.A. David M. Yousem TheJohns Hopkins UniversitySchool of Medicine, Baltimore,Maryland, U.S.A.

I. INTRODUCTION 1996, 1997, 1998, 1999). In this chapter we comprehen- sively review the pertinent medical literature on this gen- Olfactory dysfunction can generally be classified into (1) eral topic and detail our own experience. conductive disorders caused by interference with the accessof odorantsto the olfactory recepto(s,(2) peripheral sensorineuraldisorders resulting from injury to the olfac- II. IMAGING MODALITIES AND TECHNIQUES tory receptors (within the olfactory mucosa), and (3) cen- tral neural disordersof the olfactory bulb or tract or related Major advancesin pinpointing the anatomicaland patholog- parts of the central nervous system such as the prefrontal ical changesof many disordersof the sinonasalcavity and lobe, septal nuclei, amygdala, and temporal lobe. For brain have become possible as a result of the development medical imaging and the anatomical approach,we catego- and refinement of imaging techniques(Carter and Runge, rize olfactory dysfunction into two major groups: periph- 1988;Healy, 1992;Jagustand Eberling, 1991;Jolles et al., eral causes-sinonasal tract disorders-and central 1989;Reiman and Mintun, 1990;Shapiro and Som, 1989; causes- intracranial disorders. It is important to relate vogl, 1990; Yousem er al. 1996a,199',7b,1998). Even olfactory deficits to the appropriate anatomical and patho- though the imaging evaluationis not the diagnostic equiva- logical changes. Unfortunately, clinical olfactory testing, lent to histological study, anatomical imaging, such as high- whether psychophysical or electrophysiological, is rarely resolution computed tomography (CT) and magnetic capable of localizing the source (aside from determining resonance imaging (MRI), can not only map regional whether it is on the right or left) or identifying the specific lesions,but may also suggesta differential diagnosis(Carter causeof decreasedsmell function. and Runge, 1988; Shapiroand Son, 1989; Som and Shapiro, Modern medical imaging techniques offer a valuable 1988).On the other hand, functional imaging (PET, SPECT, means for assessingthe basis of some disorders of olfac- fMRI), which is reviewed in Chapter 12, affords one the tion. Although revolutionary changesin medical imaging potential to explore regional pathophysiologicalchanges in techniques have occurred in the last few decades,only a the living brain (Healy, 1992; Jagtst and Eberling, l99l; few articles have dealt with imaging studies related to Jolles et al., 1989; Reiman and Mintun, 1990;Yousemet al., chemosensory disorders (Doty et al., 1999: Goodspeed I997b, 1999b, c). The relevant imaging modalities which et al., 1987:'Kimmelman, 1991; Klingmuller et al., 1987; may be helpful in the evaluation of common causesof olfac- Li et al., 1994; Schellingeret al., 1983; Yousem et al., tory deficits are reviewedin this section.

593 594 Li et al.

A. Plain Radiographs C. Magnetic Resonance Imaging

Plain film radiography, i.e., the "sinus series," including MRI's multiplanar capability is especially advantageousin the Caldwell view, the Waters view, the lateral view, and the evaluation of sinonasal tract neoplasmsand brain dis- the base view, has long been a standard method of diag- orders. MRI, however, is less sensitivefor the detection of nosing nasal and paranasal sinus inflammatory disease. bony cortical abnormalities and landmarks. Soft tissue dis- Problems of overlap and nonspecific findings are impossi- crimination, on the other hand, is more clearly illustrated ble to avoid with plain films, and thus the study has been by MRI than by CT. Most soft tissue diseaseprocesses can largely replaced by CT. The most important deficit of the be accurately localized with a minor degree of tissue dif- plain film is its inability to provide the road map of the ferentiation, i.e., infection vs. tumor vs. hemorrhage.The ostiomeatal complex, which may guide endoscopic surgi- anatomical discrimination of the brain is much better using cal intervention (Zinreich et a1., 1987). Plain radiographs MRI than CT. One can use thin sections, large matrices, and conventional plain film tomography have virtually no and smaller fields of view to improve resolution, yet main- role in the imaging evaluation of olfactory dysfunction. tain, high contrastto noiseusing T1-weightedscans (T1W) or fast spin echo T2-weighted (T2W) images. T2-weighted B. Computed Tomography scanscan better delineatethe contrastbetween normal and inflammatory or neoplastic tissue (Shapiro and Som, CT is well suited to the investigation of the sinonasalcav- 1989). New phase sensitive inversion recovery pulse ities. BecauseCT scanningis as sensitiveto soft tissuedis- sequencesor standardspoiled gradient echo sequencescan easeas to bony changes,each scancan be photographedat highlight the gray-white matter differentiation and allow an appropriate window width and level to optimally see better assessmentof the hippocampus, parahippocampus, insidious soft tissue differences in attenuation and fine gyrus rectus, and entorhinal cortex regions. Segmentation bony detail. To study soft tissue, the window widths range of images to separatecortical volume from whole brain from 150 to 400 Hounsheld units. Conversely, the bony volume is customary for volumetric studies nowadays. detail is best observed at wide window settings-from For the evaluation of skull base invasion by sinonasal 2000 to 4000 Hounsfieldunits. The basicCT scanningpro- tumors, MRI is superior to CT (Paling et al., 1987). tocol should include all of the nasal cavity, paranasal Gadolinium enhanced scans are particularly useful at the sinuses, hard plate, anterior skull base, orbits, and skull base to detect dural or leptomeningeal involvement. nasopharynx. The brain should be included if central Gadolinium-DTPA, a paramagnetic contrast agent, has causesof olfactory dysfunction are suspected.The scans been widely utilized for distinguishing solidly enhancing are commonly performed in both the axial and coronal tumor from rim-enhancing inflammatory processes planes for optimal assessmentof the complex paranasal (Brasch,1992;Yogl et al., 1990). anatomy, but coronal scans are the most valuable for the With regard to the olfactory system, CT and MRI play anterior naso-ethmoid (ostiomeatal) region. Alternatively, complementary roles in evaluating sinonasal tract neo- thin sectionsin one plane with multiplanar reconstructions plasms (Shapiro and Som, 1989; Som et al., 1990). may be adequate.For practical pu{poses,slice thicknesses However, MRI is the study of choice to directly visualize of 3-5 mm are often employed. For the evaluation of the the olfactory bulbs, olfactory tracts, and intracranial causes ostiomeatal complex (the maxillary sinus ostium, of olfactory dysfunction (Klingmuller et aI., 1987; Suzuki infundibulum, uncinate process, and middle meatus), et a1.,1989; Yousem et al., 1993,1998, 1999a). 3-mm-thick coronal sections are fairly standard unless three-dimensional(3D) reconstructionsare requested.The D. Nuclear Medicine quality of the 3D imagesis improved by utilizing l-mm- thick sectioning, which is rapidly performed with the new In general, conventional radionuclide imaging plays no spiral scanners(minutes) and multidetector scanners(sec- significant role in the diagnostic work-up of patients with onds). Intravenous contrast enhancement is usually suspected sinonasal tract disease (peripheral causes of reserved for the identification of vascular lesions, tumors, olfactory deficits), except in the caseof cerebrospinalfluid meningeal or parameningealprocesses, and abscesscav- (CSF) leaks. Functional imaging studies, such as positron ities (Carter and Runge, 1988). Intrathecal contrastmay be emission tomography (PET) and single photon emission employed when cerebrospinal fluid leaks accompany the computed tomography (SPECT), are valuable in detecting olfactory deficits. High-resolution CT is the most useful alterations of regional brain function and biochemistry in and cost-effective screening tool for the evaluation of vivo (Alavi & Hirsch, 1991;Fowler et al., 1988;Jagust and sinonasaltract infl ammatory disorders. Eberling, 1991; Jolles et al., 1989, Reman and Mintun, Medical Imaging of Olfactory Defrcits 595

1990). Recent studieshave suggestedthat functional imag- airflow to the olfactory receptors. Besides the obstructive ing is more sensitivethan anatomical imaging in detecting effect, lesions located in the upper nasal vault and/or crib- abnormalities of the brain related to disorders such as riform plate region may also directly damagethe olfactory Alzheimer's disease and Parkinson's disease-conditions epithelium and olfactory neurons (Kern, 2000). The com- associated with loss of olfactory function (Jagust and mon peripheral sinonasaltract causesof olfactory deficits Eberling,1991; Jolles et a1.,1989). include infections, tumors, allergic rhinosinusitis, congen- ital or developmentalabnormalities, etc. III. BASICANATOMYAND PHYSIOLOGY OF A. Sinonasal Infectious Disease THE OLFACTORY SYSTEM

Paranasalsinusitis is a relatively common disorder affecting Since the anatomy and physiology of the olfactory system approximately 307oof the population at some time in their is discussedelsewhere in this volume (seeChapters l-9), lives (Allphin et al., 1991). One of the common symptoms we only briefly mention this topic here. The sensationof of acute and chronic paranasalsinusitis is decreasedsmell smell is induced by the stimulation of olfactory receptor sensation,which is generally reversible.The prompt diag- cells by volatile chemicals. The olfactory receptor cells, nosis and treatment of sinusitis are important for restoring i.e., the primary olfactory neurons,are encompassedin the olfactory function. Though the exact causeof chemosensory neuroepithelium, which is located at the top of the nasal dysfunction secondaryto sinusitis is elusive, alterationsin vault, the upper poftion of the nasal septum, the superior nasal air flow and mucociliary clearanceor obstruction from surface of the superior nasal turbinate, sectorsof the mid- secretoryproducts, polyps, or retentioncysts may contribute dle turbinate, and the region of the cribriform plate. to olfactory dysfunction (Loury and Kennedy, 199i). Afferent information from the receptors is transmitted by In the diagnosis and evaluation of paranasal sinusitis, the olfactory nerves, which course through the cribriform medical imaging plays an important role. At present,high- plate of the ethmoid bone to terminate in the glomeruli of resolution CT is the preferred imaging technique,preceded the olfactory bulb. In the olfactory bulb, the olfactory by nasal endoscopicexamination. Radiographic manifesta- neryes make synaptic contact with dendrites of mitral and tions of sinusitis have been well documented. In general, tufted cells. From there, the efferent neurons of the olfac- air-fluid levels are usually indicative of acute sinusitis, tory bulb give rise to fibers forming the olfactory tracts, whereas mucoperiosteal thickening or sinus opacification which lie just under the gyrus rectus region in the olfactory can be seen in acute and chronic disease.Polyps, mucous sulcus of the frontal lobes. Axons from mitral and tufted retention cysts, sinus expansion, and bony thickening of cells project to central brain limbic system components the walls of the sinus might also indicate chronicity of dis- including the pyriform cortex and adjacent corticomedial ease. CT is an excellent modality for the evaluation of amygdala (which together form the uncus), the ventral bony abnormalities, such as osteitis or remodeling. seenin striatum, the parahippocampusarea, and the anterior olfac- some inflammatory lesions. CT also will identify the tory nuclei. From these areas there are widespread inter- infundibulum, the maxillary sinus ostium, the middle mea- connections with many parts of the brain, including the tus, the uncinate process, and the individual ethmoid air mediodorsal thalamus, hypothalamus, orbitofrontal and cells that make up the ostiomeatal complex. This will help dorsolateral frontal cortex, temporal cortex, and other the functional endoscopic sinus surgeon in his ability to areasof the limbic system (see Chapters 8 and 9). plan effective surgery to restore normal mucociliary clear- ance. On the other hand, MRI is also highly sensitive for IV. PERIPHERAL CAUSES OFOLFACTORY detecting mucosal thickening and other soft tissue abnor- DISTURBANCE malities (Shapiroand Som, 1989).By and large,inflamed mucosais usually high in signal intensity on T2-weighted Sinonasal tract disease is one of the common causes of MR imagesand low in intensityon T1-weightedscans. The olfactory disturbance (Deems et al., I99l; Doty and signal intensity of the sinus secretions will vary with the Mishra, 2001). The etiology of the olfactory deficits concentrafion of protein within the sinus (Barat, 1990; among patients with nasal and paranasal sinus diseaseis Drutman et al. 1991;Shapiro and Som, 1989). due, in many cases,to nasal airway obstruction. The influ- ence of nasal obstruction on olfaction has been compre- B. Tirmors of the Nasal Cavity and Paranasal Sinuses hensivelyreviewed (e.g., Doty and Frye, l9B9; Doty and Mirshra, 2001) (see also Chapter 2l). Any cause of bilat- Neoplasmsof the sinonasaltract are uncommon. Malignant eral obstruction can decreasesmell sensationsby limiting tumors ofthe nasalcavity and paranasalsinuses account for 596 Li et al.

only 0.2-0.87o of all human malignancies(Som, 1991). A recently described imaging finding characteristic Early symptoms of sinonasal tract tumors, such as nasal of olfactory neuroblastomasis the presenceof peripheral discharge,unilateral nasal obstruction, and minor intermit- peritumoral cysts along the intracranial portion of the tent epistaxis, may simulate low-grade chronic infection. tumor. If stippled calcifications are also seen on CT, the Subsequentsymptoms depend on the tumor's location and diagnosisis assured. pattern of growth. Neoplasms arising in the upper nasal cavity and extending through the cribriform plate or into 2. Inverted Papillomas and Other SinonasalTumors the ethmoid sinuses are often accompanied by frontal headache,visual disturbances,and decreasedsmell sensa- The inverted papilloma is a relatively rare and locally tion. Almost all sinonasaltract tumors and tumor-like con- aggressivesinonasal tumor. It constitutes 0.547o of pri- ditions that grow to a large size may cause a decline in mary nasal tumors and occurs predominantly in males in olfactory acuity by interfering with patencyof the nasal air- the fifth and sixth decadesof life (Phillips et al., 1990). way or directly destroyingthe olfactory receptors.The most The most common presenting symptoms are nasal common malignancies of the sinonasal system are squa- obstruction,epistaxis, and hyposmia.Subsequent sinusitis mous cell carcinoma and adenocarcinoma,but lymphoma, and tumor extension into the sinusesand orbits can cause melanoma, adenoid cystic carcinoma, and chondrosarco- purulent nasal discharge,pain, and diplopia (Som, 1991). mas also populate the nasal cavity. Two examplesof intrin- Radiographic findings of inverted papilloma can vary sic sinonasaltract tumors relatively unique to the sinuses from a small nasal polypoid nodule to an expansile large (the olfactory neuroblastomaand the inverted papilloma, mass, which may remodel the nasal vault and extend into both of which often causehyposmia or anosmia)may serve the sinuses.orbits. or even the anterior skull base. CT and as prototypes for massesin this region. MRI are very useful in defining the location and extension of the tumor (Buchwald et al., 1990;Yousemet al., 1992) (Fig. 1.). Calcification is not uncommon in this tumor. 1. OlfactoryNeuroblastoma Other sinonasaltract tumors, such as squamouscell car- Olfactory neuroblastoma, or esthesioneuroblastoma,is a cinoma, adenocarcinoma,melanoma, etc., can also cause rare nasal tumor originating from the olfactory neuroepi- hyposmia or anosmia during their late stage. Squamous thelium lining the roof of the nasal vault and in close prox- cell carcinoma accountsfor 8O7oof paranasalsinus malig- imity to the cribriform plate. There have been less than nances,is most commonly seenin the maxillary sinus, and 300 reported casesin the world literature. Olfactory neu- usually demonstratesbone destruction at the time of pre- roblastomasoccur in all age groups with a peak incidence sentation.Adenocarcinomas occur most frequently in the in the 11-20 and 51-60 year groups. There is a slight pre- ethmoid sinus while melanomais usually seenintranasally. ponderance of the tumor in women. The incidence of Additional benign neoplasms known to affect the olfactory neuroblastomahas been estimated to range from sinonasalcavity include osteomas,enchondromas, schwan- 2 to 3Eoof all malignant inffanasal neoplasms.The most nomas, and juvenile angiofibromas. Osteomasare usually common symptoms are unilateral nasal obstruction and identified in the frontal sinus and may be a sourcefor recur- recurrent epistaxis. Hyposmia or rhinorrhea is not rent headacheand/or recurrent sinusitis. The classic story unusual. Extension into the orbit, paranasal sinuses, or of a frontal sinus osteomanarrowing the sinus opening is a anterior cranial fossa may cause vision disturbances and patient who has severesinus pain associatedwith takeoffs headache(Elkon et a1.,1979;Li et al., 1993;Newhill et al., from airplane flights. This is a benign mass,which is often 1985). In the detection and staging of olfactory neuroblas- completely invisible on MRI due to the presenceof dense toma, CT and/or MRI play an important role. Generally compact bone making up the mass.On the other hand, it is speaking, MRI is more accurate than CT in showing the easily identified on CT as a markedly hyperdense bony tumor's intracranial extent. MRI is also exquisitely useful mass protruding in the sinus. Occasionally, the osteoma for differentiating neoplasm from postobstructed secre- will result in mucocele formation and./orpneumocephalus tions because of the difference in the signal intensity as the posterior wall of the frontal sinus is breached. (secretions are bright on T2, tumor intermediate). Enchondromas are less common neoplasms of the Unfortunately, signal intensity characteristics of various sinonasalcavity which, on CT, often have a popcorn calci- sinonasal tract tumors overlap each other, so MRI cannot fication appearancedifferent from the stippled calcifica- usually predict specific tumor histology. However juvenile tion of inverted papillomas. This lesion, because of its angiofibroma can usually be distinguished from other characteristiccalcification, is best evaluatedwith CT. tumors on the basis of its high vascularity and marked Schwannomas of the fifth cranial nerve are the most enhancement. common to affect the sinonasal cavity. They will typically Medical Imaging of Olfactory Deficits 597

Figure 1 A 40-year-old woman with 3-month history of decreasingsmell sensationand left nasal obstruction. (A) Bone-targetedcoro- nal CT shows an expanded opacified left nasal cavity with bowing of the lateral nasal wall (arrows) and opacification of the left maxil- lary and both sphenoid sinuses. (B) Axial contrast-enhancedCT scan shows erosion through the left lamina papyracea (arrow) with displacement of the medial rectus and globe laterally. The differentation between tumor and obstructed secretionsis not readily apparent with CT. Histological diagnosis: nasal cavity carcinoma arising within a dysplastic inverted papilloma.

follow the course of the nerve and can expand skull base intraorbital spread. One of the advantagesof MRI is the foramina through which they travel. The signal intensity of ability to distinguish sinus neoplasm from postobstructive schwannomas varies according to the content of the dense secretions. This may be difficult by CT if the sectetions are Antoni A tissue or loose Antoni B tissue, the latter being isodense to the mass and if the malignancy does not brighter on T2W scans. Schwannomas enhance avidly, enhance dramatically. If one was forced to study the although they may have inhomogeneity to the enhancement. patient with a single modality, the literature supportsMRI Finally, one has the juvenile angiofibroma, a fascinating as the best study for the staging of sinonasalmalignancies benign neoplasm, which appearsto arise in the region of (Hunink et al., 1990; Kraus et al.,1992; Paling et al.,1987; the sphenopalatine foramen and/or the pterygopalatine Sissonet al.. 1989). fossa The lesion accounts for O.5Voof head and neck Som et al. (1991) noted that squamouscell carcinoma massesand is typically seenin adolescentmales who pre- (low in T2 intensity) could be distinguished from inflam- sent with epistaxis and/or a nasal mass (Mehra, 1989). The mation (high in T2 intensity). They compared CT to MRI lesion is highly vascular as exemplified on MRI by the sig- for mapping sinonasal tumors. They found that MRI and nal flow voids within the lesion and its marked contrast CT were equivalent in 23 of 53 patients in defining tumor enhancement.Because of its propensity for spreading via extent and that MRI was superior to CT in 26 patients. Of the canals and foramina at the skull base,MRI is probably the 4 casesin which CT was superior, subtle bony erosion the study of choice for the evaluation of this neoplasm. (2) andosteo(1)-cartilaginous(1) lesions accounted for the Embolization of these lesions will assist the surseon in "misses" on MRI. Of 60 inflammatory lesions, MRI was limiting blood loss if resection is considered. superior (Bonte et al., 1993) or equivalent (Everall et al., 1991) to CT in all cases.Inflammation (bright) and neo- plasm (intermediate) could be distinguished in 95Vo of 3. Malignant Neoplasms casesbased on T2W signal intensity. Even when the sinus CT and MRI probably play complementary roles in the secretionsbecome increasingly inspissatedand the signal evaluation of sinonasal malignancies because of CT's intensity on T2W scansdecreases, the neoplasmcan be dis- superiority in defining bony margins and MRI's superior tinguished from the obstructed secretions by its typical soft tissue resolution and ability to define intracranial or heterogeneity as opposed to the smooth homogenous 598 Li et al.

appearanceof sinus secretions. This is also true in the squamous cell carcinoma signal intensity characteristics casesof mucoceles, which may occur after or in associa- on MRI, the lesion is characterizedby a low signal inten- tion with sinus neoplasms.Additionally, MRI has shown sity on T2W scans. This is why differentiation with that most squamouscell carcinomas of the sinonasalcav- obstructed secretions which are typically bright in signal ity enhancewith gadolinium in a solid fashion as opposed intensity on T2W scansis so easy on MRI. to a peripheral rim of enhancement in sinus secretions Becauseof Som et al.'s early work depicting sinonasal and/or mucoceles.Unfortunately, lymphomas, undifferen- malignancies as hypointense on T2W scans, people have tiated carcinomas, inverted papillomas, and some sarco- come to rely on this pulse sequencefor mapping cancers mas may have identical signal intensity and enhancement (Som et al., 1990). Unfortunately, low intensity on T2W characteristicsas squamouscell carcinoma. scansis an inconstant finding in sinonasalmalignancies in Gadolinium is particularly useful for demonstrating general. Hunick et al. found that over 50%oof head and epidural or meningeal invasion of neoplasms.Often, post- neck malignancies had signal intensity on T2W scansthat contrast scans must be combined with fat suppression was brighter than muscle and isointense to brain (Hunink techniques in order to identify enhancement amidst the et al., 1990).Approximately 25Voof benign tumors had the abundantskull basefat. In one series,'l57o of patients with same intensity pattern. Lanzieri et al. (1991) also reported intracranial extension of sinonasalmalignancies had addi- that the signal intensities of tumors, mucoceles, schwan- tional information about tumor extent demonstratedwith nomas, and obstructedsecretions may show some overlap. postcontrastMRI studies (van Tassel et al., l99l). Som et al. (1991) have found that minor salivary gland Subtraction MRI of pregadolinium scans from post- massesand schwannomasmay have T2W signal intensity gadolinium scans may improve visibility of such subtle similar to that of inflammatory lesions. Minor salivary enhancement(Lloyd and Barker, 1991). It should be noted gland tumors and melanoma are the next most common that meningeal enhancementneed not necessarily imply malignancies to affect the sinonasalcavity after squamous neoplastic invasion; just as in cases of meningioma, the cell carcinoma(van Tasselet al., 1991).The minor salivary dura may enhance because of reactive fibrovascular gland tumors representa wide variety of histological types changesalone. including adenoid cystic carcinoma, mucoepidermoid car- When one encountersa sinonasal mass that is eroding cinoma, adenocarcinoma,and undifferentiated carcinoma. intracranially, one must consider carcinoma,olfactory neu- Of these tumors, adenoid cystic carcinoma is the most roblastoma, sarcomas,lymphomas, sinonasalpolyposis, common variety. Its signal intensity may be high or low on and inverted papillomas. Twelve percent of patients with T2W scans, possibly related to the degree of tubular or polyposis and mucoceles eventually erode the skull base cribriform histological pattern as well as cystic spaces, (Som et al., l99l). The pattern of bone destruction may be necrosis, and tumor cell density. Tissue specificity is not similar between malignant and benign lesions at the readily achievablewith MRI or CT. Gadolinium is of par- non-sinus bearing skull base. Bone remodeling in this ticular use with adenoid cystic carcinomas, which have a location is a rarity; a permeativepattern is the norm for all propensity for perineural spread(Graamans and Slootweg, lesions.Som et al. (1988)have suggested that a lesionwith 1989). With sinonasal cavity malignancies one should homogeneous signal intensity invading intracranially is always attempt to trace back the branchesof the fifth cra- more likely to be a malignancy, whereas heterogeneity nial nerve via the pterygopalatine fossa, foramen rotun- suggestsan inflammatory cause. Unfortunately, necrosis, dum, foramen ovale, and orbital fissures in order to hemorrhage,or calcification in carcinomas,olfactory neu- identify perineural neoplastic spread. roblastomas,or sarcomasmay cause signal heterogeneity. Adenocarcinomas of the paranasal sinuses have a Polyps generally enhancein a peripheral pattern; true neo- predilection for the ethmoid sinusesand appearmore com- plasms enhance solidly. Malignancies have a broad flat monly in woodworkers. This tumor also tends to have low base of skull erosion; benign conditions have a rounded signal intensity on T2W MRI images but may have high polypoid intracranial excrescence. signal intensity in a small percentageof cases. Squamous cell carcinomas account for 807o of Sarcomas of the sinonasal cavities are very rare, with the malignanciesto affect the paranasalsinuses and807o in chondrosarcomabeing the most common. Again, the his- the maxillary sinus. The hallmark of malignancies of the tological diagnosis is probably better suggested by CT sinonasalcavity is bony destruction, seenin approximately based on the characteristic whorls of calcification. 80Voof CT scansof sinonasalsquamous cells carcinoma at However, for staging, MRI is competitive with CT, and, initial presentation.The lesion is confined to the maxillary particularly if repeat examinations are going to be antrum in only 25Vo of cases at presentation (Lyons and required, follow-up with MRI to avoid the radiation expo- Donald, 1991). In most series documentins sinonasal sure of CT is recommended. Medicat Imaging of Olfactory Deficits 599

Melanoma is a tumor that is usually identified in the odorant to the olfactory receptor area. The senseof smell nasal cavity as opposed to the paranasalsinuses. It has is probably less than normal in many patients with cranio- been associatedwith melanosis in which there is field facial anomalies (Crysdale, 1981). Congenital develop- deposition of melanin along the mucosal surface of the mental abnormalities include choanal atresia, hereditary sinonasal cavity. Therefore, multiplicity of lesions nasal septal deviation, facial hypoplasia, cleft palate, nasal becomes a problem when dealing with melanomas' dermoids and epidermoids,cephaloceles, and gliomas, etc. Neither CT nor MRI is particularly helpful in identifying Medical imaging techniques, especially high-resolution the field "cancerization" of melanoma.When melanoma CT, play a key role to detect and evaluate the facial and contains melanin there is paramagnetismwhich causes bony changes(Barkovich et a1.,7991; Klein et a1.,1987). T1 and T2 shortening accounting for high signal intensity CT is most useful because surgical correction requires on T1W scans and low signal intensity on T2W scans identification of and closure of the osseousabnormalities. (Atlas et a1., 1990). However, an amelanotic melanoma MRI is most effective in defining soft tissue massessuch may have bright signal intensity on T2W scans.The pres- as cephalocelesand nasal gliomas. ence of hemorrhage associatedwith the melanoma, a Congenital anosmia can be associatedwith a number of common occuruencebecause of the coincidenceof epis- developmental and inflammatory conditions. Kallmann's taxis, may further obfuscatethe signal intensity pattern syndrome, also known as hypogonadotrophic hypogonadism (Yousemet al.. 1996c). with anosmia, is a congenital X-linked disorder in which the Lymphoma does occur in the paranasalsinuses and may olfactory bulbs and tracts are not formed. This is not associ- have variable signal intensity as well. It is characterizedby ated with holoprosencephaly, and the usual deficits are homogeneous signal intensity without necrosis and the related to hormonal abnormalities in the pituitary gland with associationwith cervical lymphadenopathy. the loss of senseof smell. Infertility often coexists.In 1993, Metastatic diseaseto the paranasalsinuses is extremely an MR study of the olfactory system in Kallmann's disease rare. Of the primary causesof metastasesto the sinuses, showedabsence ofthe olfactory bulbs and tracts in 17 of 18 renal cell carcinoma is probably the most common. This patients while confirming the presenceof the olfactory bulbs tumor also has a propensity for hemorrhageand may also and tracts in all 10 studied patients with idiopathic hypo- have a variable signal intensity depending upon the stage gonadotropichypogonadism (Yousem et al., 1993, 1996a). of hemorrhage. Some patients have absenceof the olfactory bulbs and tracts without Kallmann's syndrome. It is unclear whether this rep- C. Allergic Reactions resents congenital absenceor whether an inflammatory con- dition early in infancy destroys the olfactory bulbs and tracts. Allergic rhinitis is a common upper airway condition Certain viruses have a propenslty for injuring the olfactory affecting about 30 million Americans with peak prevalence system.A recent study has noted the incomplete formation of in the age group from 35-54 years (Baroody and Naclerio, olfactory sulci in patients with congenital anosmia as well as 1991). Hyposmia or anosmia is common with allergic a variable percentage of aplastic olfactory bulbs, ffacts, and rhinitis, mainly caused by nasal obstruction by polyps or tubercles (Di Rienzo et al., 2002). Still others may have con- inflamed mucosa, which limit accessof inspired air to the genital absenceof senseof smell on the basis of early head roof of the nasal vault (Cowart et al., 1993). The diagnos- trauma where the ciliary nerves as they crossedthe cribiform tic work-up begins with a careful history, which attempts plate may be sheared and the olfactory system is affected. to identify offending allergens. Skin testing of specific Infectious causesmay also affect the senseof smell in early antigens is often used to confirm the diagnosis. Medical childhood, usually secondarilyto viruses.In thesecases one imaging studies play a supplementary role in the evalua- seesthe olfactory bulbs and tracts; but they are not functional. tion of sinonasal airway status and differential diagnosis. Holoprosencephaly is a congenital, multiple midline CT and MRI are also important for detecting any compli- malformation disorder that has a known association with cations such as sinusitis, mucoceles,and aggressivepolyps sensorydeficits of vision and olfaction. Although variable in patients with allergic rhinitis. Rounded excrescences amounts of aplasia and hypoplasia of the olfactory appura- and enlargementof ostia are seenin the airway of patients tus may be identified, the most common MR finding is with polyposis. complete absenceof the olfactory bulbs, occurring in 92Vo of patients. A high association with absenceof the olfac- D. Congenital or Developmental Abnormalities tory nerves and tuberclesis also seen.There does appearto be some, albeit poor, differentiation of the olfactory sulci It is generally acceptedthat normal variations in the nasal and gyri recti, which were absentonly in a little over half anatomy may play a role in preventing the access of an of the subjects (Barkovich and Quint, 1993). 600 Li et al.

E. Other Peripheral Causes measure to assessnaso-ethmoid trauma (Daly et al., 1990;Kassel, 1988). It is estimated that 30 million Americans have used cocaine and 5 million use it regularly (Gregler and Mark, 1986). Intranasaluse of cocaine and heroin has reached V. CENTRAL CAUSES OFOLFACTORY epidemic proportions in the United States. Although DISEASES hyposmia or anosmiahas been suggestedto occur often in cocaine abusers, few studies using quantitative mea- There are numerous CNS disorders that are associated sures of olfactory function have confirmed such reports. with olfactory dysfunction. The most common types fall in A sole study on this topic reported that of I I cocaine the categories of degenerativeneuropsychiatric disorders, abusers who underwent detailed olfactory testing, only hereditary conditions, trauma, and central neoplasms. Of one was found to be anosmic and another had mild olfac- course, in some disorders the involvement of both periph- tory discrimination dysfunction (Gordon et al., 1990). eral and central neural processesmay occur. These authors note that most cocaine abusers do not develop permanent olfactory dysfunction. If, in fact, A. Alzheimer's Disease olfactory disturbance occurs as a result of heavy cocaine use, it could be due to associatedconductive disorders, It has been well documentedthat olfaction is significantly nasal airway obstruction, alteration in sinonasal aerody- altered in Alzheimer's (AD). Nearly all studiesof olfactory namics, damage to the olfactory epithelium, damage to function in patients with AD have reported decreasedsmell the central olfactory system, or osteolysis of the cribri- relative to age-matchedcontrols (see Chapter 23). These form plate (Kuriloff, 1989). studies demonstratemarked impairment of smell function Concerning the conductive disorders, severalreports of in early AD, whether measuredby identification, discrim- osteolytic sinusitis and extensiveosteocartilaginous necro- ination, or thresholdsensitivity (Doty, 1991; Doty et al., sis of the nasal septum in cocaine abusers have been 1987;Serby et al., 1991). described (Newman, 1988; Schweitzer, 1986\. Erosion of Recent neuropathological studies have correlated well nasal septal cartilage is a known complication of cocaine with these clinical f,rndings.The anterior olfactory nuclei abuse. Within the differential diagnosis for cartilaginous in AD patients contain senile plaques, neurofibrillary tan- destruction, one should include Wegener'sgranulomatosis, gles, granulovascular degeneration, and cell loss syphilis, leprosy, lymphoma, rhinoscleroma (a klebsiella (Averback, 1983; Esiri and Wilcock, 1984). The olfactory infection), and fungal invasion. CT, preferably in the coro- bulbs also show involvement (Esiri and Wilcock, 1984: nal plane, can provide excellent views of septal perfora- Ohm and Braak, 1987), as does nasal sensory epithelium tion, osteolysis, and sinusitis. (Jafek et al., 1992). In addition, central olfactory struc- To evaluate intracranial disorders associated with tures, especially the amygdala and the entorhinal, pyri- cocaine, MRI is the study of choice. Vasculitic infarcts, form, and temporal cortices, are frequently affected by hypertensivehemorrhages, and white matter ischemic foci Alzheimer's disease(Harrison, 1986; Pearsonand Powell, may be seen with MRI. Recently Tumeth and colleagues 1989). Besides the above findings, devastating nerve cell demonstrated multifocal cortical deficits with a special loss and gliosis in the region of the hippocampal formation predilection for the frontal and temporal lobes on SPECT have been observed at autopsy in AD patients (Ball et al., perfusion brain scansin chronic cocaine abusers(Tumeth 1985;Hyman et al., 1984). et al., 1990). Similar findings have been reported by others Neuroimaging has played an important role in detecting (Holman et al., l99l1, Kolow et al., 1988). These findings some of the pathological changesof AD patients in vivo, may suggest a central basis for some cases of cocaine- and its uses are growing, both for clinical evaluation and as related decreased olfaction. Some studies also have a researchtool. Early CT studies in AD patients demon- revealed that cerebral atrophy developsin chronic cocaine strated generalized enlargement of the ventricular system abusersandthat the severity correlateswith the duration of and sulci (George et al, 1981; Naser et a1., 1980). Several abuse(Pascual-Leone et al., 1991). reports have noted that ventricular and sulcal enlargement Anosmia or hyposmia is a frequent sequelaof high- correlatewith the severityof AD (Albert et al., 1984;George level midface fractures in which the olfactory nerves et al., 1983). However, these findings are not specihc and may be severed at the level of the cribriform plate have relatively weak conelations. de Leon and colleagues (Kassel, 1988; Mathog, 1992). Because ethmoid com- (1989) have emphasizedthe rate of change in ventricular plex and cribriform plate fractures are difficult to detect size with repeatedCT scans as an important index in the on plain radiographs,thin-section coronal CT is the best diagnosisof AD. Recently,several investigators have recog- Medical Imaging of Olfactory Deficits 601 nized that CT andlor MRI delineation of atrophic changesin regional oxygen, and glucose metabolism, which may pro- the temporal lobe and the hippocampus with enlargement of vide evidence supportive of the diagnosis of AD (Jagust hippocampal-choroidalfissures strongly supportthe diagno- and Eberling, l99l). The above-mentioned structural sis of AD (deLeon et al., 1988;George et al., 1990;Kesslak atrophic changes by CT and MRI are also supported by et a1.,7991;Kido et al., 1989). functional imaging studies (McDonald et al., 1991; McDonald and colleagues(1991) reviewedMRI scansin Ohnishi et al., 1991). The major findings of functioning 22 patients with early-onset AD. The results showed that imaging studies in patients with AD are abnormal regional patients with AD were significantly more likely than age- cerebral blood flow pattern and flow reduction. The com- matched controls to have MR evidence of periventriculm mon sites of blood flow reduction are in the temporopari- hyperintensities on T2W scans.This study suggestedthat the etal region and the frontal areas.In one report (Bonte et al, increased frequency of periventricular hyperintensities may 1993), sevenpatients with possible diagnosis of AD stud- have a relationship to the diseaseprocess. Our own experi- ied by SPECT showed only frontal flow abnormalities. Is ence with MRI studies of AD patients is that most of the this an early imaging finding which may suggesta patho- cases with AD have, in addition to ventriculomegaly and physiologic basis to explain the decreasingsmell sensation sulcal widening, significantly reducedvolume of the tempo- in AD? Of course, more studies are neededfor further dis- ral lobe and slight atrophy ofolfactory bulbs. (Fig. 2). covering the nature of AD. We believe that early and cor- Besides CT and MRI, SPECT and PET techniques are rect diagnosis of AD in vivo by neuroimaging techniques also useful for evaluating regional cerebral blood flow, will be possible in the near future. There is a dose-related association between apolipopro- tein E-4 (APOE-4) allelic frequency and the development of AD (APOE-2 may confer protection)' Recent studies have shown a decline in resting parietal, temporal' and pre- frontal PET glucose metabolism in cognitively intact patients with APOE-4. It remains to be seen whether this, and/oran analogousfMRI study, may serve to be a predic- tor of development of AD. Recently some investigators have used dynamic con- trast susceptibility contrast imaging MR to try to duplicate the nuclear medicine flow studies.Indeed they have found that relative values of temporoparietal regional cerebral blood volume (as a percentageof cerebellar CBV) were reduced by a factor of 2OVobilaterally in the patients with Alzheimer disease compared to normals. Using left and right temporoparietal rCBV as index measures,specificity was 96Voand sensitivity was 95Vain moderately AD and 88/o in mild AD (Hanis, 1998).

B. Parkinson's Disease

Odor detection and identification are significantly impaired in Parkinson's disease(PD) patients (Doty et al., 1988, 1995; Montgomery et al., 2000) (see Chapter 23). It is unclear whether the olfactory dehcits in PD and AD share the sarne cause. Not surprisingly, PD research into the causeof smell dysfunction has focused on dopaminer- Figure 2 A 60-year-old woman with Alzheimer's disease. gic changes.Brooks and colleagues (1991) have demon- UPSIT scores revealed severe bilateral anosmia. (A) Normal strated by using PET that patients with PD show olfactory bulbs are seen (arrows) on coronal MR. Dilation of the significantly reduced mean uptake of l8F-dopa in the cau- olfactory sulci (arrowheads) reflects generalized atrophy. (B) in the posterior part of the Coronal MR scan through the temporal lobes shows temporal date and putamen, especially horn enlargement and atrophic changes, slightly worse on the putamen. Previous functional imaging studies have also rieht side. indicated a reduction of striatal dopamine storage in PD. 602 Li et al.

PET technique with l8F-dopa in pD patienrs has also olfactory dysfunction, but the exact mechanismfor hypos- demonstrated reduced basal ganglia activity (Alavi and mia in HD patients remains to be worked out. Hirsch, 1991). However, the olfactory deficit is unrelated to severity of motor or cognitive symptoms and is not D. Korsakoff'sPsychosis improved by l-dopa therapy (Doty et al., 1992), so the underlying causes of olfactory pD dysfunction in still Patientswith Korsakoff's psychosis (Kp) exhibit impaired requires more study. odor detection, identification, and intensity estimation CT scanning has little role in establishing the diagnosis (Joneset a1.,1975; Mair et al., 1986).An animal model of PD other than to exclude mass lesions in the brain. In study has shown that the behavior of rats recovedng from general, CT shows no specific striatal abnormalities and pyrithiamine-induced thiamine deficiency share several occasionally only mild, nonspecific ventricular and sulcal important features with KP patients, including the impair- enlargement. pD The major feature of on MRI appearsto ments observed for smell, hearing, learning, and memory be a trend towards a decreasedwidth of the pars compacta (Mair et al., 1991). The mechanism of hyposia and/or of the substantia (Braffman nigra et al, 1989). There is a dysosmia in patients with KP is unclear and still under lateral to medial gradient of loss of the normal signal of the investigation. Olfactory perception may be selectively pars compacta as well as volume loss. Moderate or marked impaired in KP by the diencephalonlesions that are char- cortical atrophy tends to occur more frequently in pD acteristic of this disease.Degeneration in the mediodorsal patients than in controls. MRI may occasionally show thalamic nucleus (the common neuropathologicallesion in abnormal decreasedT2W intensity in the putamen and to a KP) and atrophy in the prefrontal areasmay also causethe lesser degree in the caudate nuclei and substantia nigra, olfactory dysfunction (Mair et al., 1986). suggestiveof iron deposition (Drayer et al, 1986). A quantitative neuropathological study of the human cerebral cortex has shown that the number ofcortical neu- C. Huntington'sDisease rons in the superior frontal lobe in chronic alcoholic patients is significantly reduced (Happer et al., 1987). Patients with Huntington's disease (HD) evidence olfac- Chronic alcoholism is also associatedwith smaller vol- tory dysfunction (Doty, 1991; Moberg et al., 1987). umes of cortical white and gray matter relative to controls Neuropathological studies in HD have demonstratedpre- (Pfefferbaum et a1.,199 5). Traditional neuropsychological mature neuronal cell death and reactive gliosis occurring tests and functional imaging studies have also demon- most markedly in the head of the caudatenuclei bilaterally strateddisturbances of frontal-lobe function and metabolic (Myers et al., 1991; Vonsattel et al., 1985). A loss of deficits in patients with KP (Joyce and Robbins, l99l; 1A-807o of the striatal neurons may occur before func- Kopelman, 1991;Metter et al., 1989). tional impairment is obvious. Similar but less extensive Brain CT scans have demonstrated that KP patients changesalso affect the putamen.Later, atrophy of the cere- show more pronouncedthird and lateral ventricular dilata- bral cortex occurs as well. All of theseprogressive atrophic tion and wider interhemispheric fissures than matched changescan be identified on CT and MRI scans,especially groups of normal controls and non-Korsakoff alcoholics in the caudate nuclei, where the volume of the caudate (Jacobsonand Lishman, 1990; Ron, 1983; Ron et al., head decreases and the intercaudate distance increases 1982). Shrinkage in the frontal lobes and cerebellum (Simmons et al., 1986; Starksteinet a1.,1989). Increased appears to be especially pronounced (Jacobson and signal intensity in the putamen and globus pallidus has Lishman, 1990).A MRI study (Jerniganet al., 1991) has been described in the juvenile form of Huntington's dis- revealedthat patients with KP show widespreadreductions ease,and frontal atrophy is usually present. in gray matter volumes in addition to CSF increases,with PET studies of patients with HD have consistently the greatestreductions observedin diencephalic structures. demonstratedhypometabolism in the caudatenuclei, often The volume losses that best differentiate the KP patients before the development of atrophy on CT (Hayden et al., from the alcoholic controls included lossesin anterior por- 1986). SPECT studies involving HD patients have also tions of the diencephalon,mesial temporal lobe structures, revealed decreaseduptake in the caudatenuclei, including and orbitofrontal cortices (areasinvolved in olfaction per- the caudatesof one patient with early diseaseand no evi- ception). Several other studies (Donnal et al., 1990; dence of atrophy on MRI (Nagel et al., 1988; Reid et al., Gallucci et al., 1990; Squire et al., 1990;Victor, i990) have 1988). Thus, functional imaging with PET or SpECT may also demonstratedthat MRI is highly sensitivein detecting contribute to the early diagnosis of HD. reversible diencephalon (medial thalamic) and mesen- Theoretically, the input of caudate/putaminalfibers to cephalic (periaqueductal) lesions. In addition to general- the limbic system and striatum may be altered, leading to ized cortical and cerebellar vermian atrophy seen on CT Medical Imaging of Olfactory Deficits 603 and MR, recent reports have noted the presenceof high evaluatedthe volume of the temporal lobes in schizophre- signal intensity areasin the periaqueductalgray matter of nias by a quantitative MRI study. The results showed that the midbrain (40Vo),the paraventricular thalamic regions the volume of temporal lobe gray matter was 20Vo smaller (46Vo),the mamillothalamic tract, and in tissue surround- in the patients than in the control subjects,and lateral ven- ing the third ventricle on T2W MR scans (T2WI). tricular volume was 677olarger in the schizophreniagroup Reversible thalamic lesions in the dorsal medial nuclei than in the control group. Schizophrenic patients tend to have also been reported. These areas may or may not have smaller hippocampi that matched controls. schizo- enhance(in some casesthe enhancementmay be dramatic, phrenias are also reported to have cavum septum pellu- almost sarcoid-like) and may be associatedwith mamillary cidum more frequently than controls. In a recent study, body atrophy. Mamillary body enhancementmay be the Turetsky et al. (2000) reported that patients with schizo- sole manifestation of Wernicke's encephalopathy.Myelin phrenia exhibited 23%o smaller olfactory bulb volume degeneration, mamillary body volume loss, intracellular bilaterally than comparison subjectsby a quantitative MRI edema,and microglial proliferation are seenpathologically study. (but may be presentin alcoholics without Wernicke's). MRI findings in patients with KP may enable early E CongenitalAnosmia diagnosis of the disease,which may have a positive effect on both treatment and prognosis (Gallucci et al., 1990). Congenital anosmia, which traditionally has been defined as anosmia present from a patient's earliest recollection, E. Schizophrenia has been recognizedfor centuries.The most common form of congenital anosmia is Kallmann's syndrome or olfac- Impaired olfactory function has been reported in schizo- tory dysplasia, which is characteized by hypo- phrenics,especially males (seeChapters 23 and 24). These gonadotropic hypogonadism and anosmia (Kallmann et olfactory deficits, which are not of the samemagnitude as al., 1944; Lieblich et al., 1982). The incidence of those seenin AD and PD, are perhapsnot unexpectedgiven Kallmann's syndrome is about 1:100,000 in men and the occurrenceof olfactory hallucinationsas symptomsin a 1:50,000 in women. There has been increasing interest in number of patients with schizophrenia and the evidence the pathology, pathophysiology, and genetics of this disor- linking both to temporal lobe dysfunction (Rausch et al', der. Pathological and surgical studies of patients with 1977; Roberts, 1988). Neuropathologicalstudies in schizo- Kallmann's syndrome have shown agenesisof the olfac- phrenic patientshave reported neuronal loss in the entorhi- tory bulbs (DeMorsier and Gauthiet 1963; Males et al., nal region and prefrontal cortex, gliosis in the basal limbic 1973). Laboratory findings include decreasedserum folli- structures of the forebrain, and atrophy in temporolimbic cle-stimulating hormone and luteinizing hormone as well structures (Benes et al., 1986; Falkai et al., 1988). as decreasedurinary gonadotropins(Lieblich et al., 1982). Neurophysiological function studies (including regional In medical imaging studies,CT is a limited tool for the cerebral blood flow, brain electrical activity mapping, and demonstration of sinonasal and intracranial abnormalities regional metabolic activity in the brain) in patients with in patients with congenital anosmia (Klein et al., 1987; schizophrenia have demonstrated prefrontal cortex and Moorman et a1., 1984). Surface coil MRI is the optimal temporal lobe dysfunction (Mesulam, 1990). Functional modality to reveal the intricate details of the olfactory imaging, such as PET or SPECT, in the study of schizo- bulbs, tracts, and rhinencephalonin vivo. Klingmuller and phrenia is limited and inconclusive. However, functional colleagues(1987) have clearly demonstratedthe olfactory imaging has provided some evidence that certain schizo- sulci in a normal control group by MRI, but not in the phrenic patientshave decreasedblood flow and metabolism patients with olfactory dysplasia. More recently, the in the frontal lobes (hypofrontality) (Alavi and Hirsch, authorshave studied two caseswith Kallmann's syndrome r99r). by MRI. Both showed no olfactory bulb at all and flatten- Anatomical imaging findings have basically paralleled ing of the gyrus recti (Yousem et al, 1993,1996a); frontal the neuropathological changes in the brains of patients and temporal lobe volumes were normal (Fig. 3). with schizophrenia. The most consistent finding on both In a mixed population of patents with congenital anos- CT and MRI is an increase in the size of the cerebral ven- mia, we found olfactory bulb and tract absence(68-847a) tricular system, especially in the frontal and temporal and hypoplasia (16-327o) in all 25 cases studied. Eight horns, and corresponding decreases in cerebral tissue, individuals had Kallmann's syndrome (hypogonadotropic especially in the prefrontal cortex and in medial lem- hypogonadism with anosmia). Temporal andlor frontal porolimbic structures (Mesulam, 1990; Suddath et al., lobe volume loss were noted in 5 individuals, mild in all 1989;Young et al., 1991).Suddath and colleagues(1989) but one individual. We concluded that congenital anosmia 604 Li et al.

G. Head Thauma

Craniofacial trauma can alter olfactory ability through one of severalmechanisms: (1) damageto the nose,sinuses, or both with resultant mechanical obstruction to odorants,(2) shearing of olfactory filaments as they course through the cribriform plate, (3) contusion to the olfactory bulb. and (4) contusionor shearinginjury ofthe cerebralcortex. par- ticularly the frontal and temporal lobes (see Chapter 30). The incidence of anosmia or hyposmia after head trauma has been reported quite variably from2 to 38Va,(Deems et a1., l99l; Doty et al., 1997; Hagan, 1967; Leigh, 1943; Levin et al., 1985; Schechterand Henkin, 1974; Summer, 1964; Ztsho, 1982) and increases with the severity of injury (Levin et al., 1985; Summer, 1964). However, even a minor injury can sometimesresult in anosmia or hypos- mia (Schechterand Henkin,1974; Summer 1964). Recent evidence has shown that the location of the hematoma or contusion of the brain after head trauma is one of the most important factors leading to olfactory dysfunction (Costanzoand Zasler, 1991; Doty et al., 1997; Levin et al., 1985; Yousem et al., 1996b). Specifically, diminished olfactory discrimination has been confirmed in patients with prefrontal lesions (Potter and Butters, 1980). Animal studieshave shown that the prefrontal olfactory areaplays a prominent role in the fine and specific discrimination of odors (Tanabe et al., 1975). Besides prefrontal lesions, temporal lobe structuresare also involved in the odor pro- cessing of odor perception (Rausch and Serafetinides, 1975; Rauschet a1.,1977). Indeed frontal or temporal lobe hematomasor contusionsare now believed to be one of the most common causesof olfactory dysfunction after head injury (Costanzoand Zaster, 1991 ; Doty et al., 1997; Levin et al., 1985;Schellinger et al., 1993;Yousem er al., 1996b) (Fie.a). It has been established that plain skull radiography plays only a small role in the evaluation of head trauma (Masters et al., 1987). CT cunently is the study of choice when diagnostic imaging is necessary after acute head trauma (Cohen, 1990; Kelly Figure 3 (A) Coronal 500/20 scan from normal volunteer et al., 1988). CT can detecr (64-year-old woman with normal smell function) demonstrates subarachnoid hemorrhage, fractures, and intraventricular normal olfactory bulbs (anows). (B) Coronal 500117 scan of blood, lesions for which MRI is less sensitive acutely. CT 27-year-old woman with congenital anosmia without can be performed with close patient monitoring in a rapid Kallmann's syndrome shows extremely atrophic olfactory fashion. However, MRI is superior to CT in the detection bulbs (arrows). (C) Coronal 500/14 scan of 29-year-old male and characteization of subacute injuries, hemorrhage patient with Kallmann's syndrome evidences absence of olfac- outside the subarachnoidspace as in subdural hematomas, tory bulbs tracts and with flattened gynrs rectus (arrow) on the cortical contusion, and shearing injuries. MRI is exquis- right side, but with normal-appearing gyrus rectus on the left itely sensitive to diffuse axonal injuries leading to side. demyelination. MRI is also useful in the follow-up of brain contusion and/or hemorrhage, thereby eliminating the or hyposmia appears to be an olfactory bulb tract phenom- radiation exposure associated with CT (Cohen, 1990; enon rather than a central process (Yousem et a1., 1996a). Zimmerman et al., 1986). Medical Imaging of Olfactory Deficits 605

tory test scores (Doty et al., 1997b; Yousem et al., 1996b). Abnormalities on MR in patients with posttrau- matic olfactory dysfunction occur at a very high rate (88Vo),predominantly in the olfactory bulbs, tracts, and inferior frontal lobes. Qualitative and quantitative assess- ments of damage show little correlation with olfactory tests probably due to multifocal injury, ciliary nerve dam- age, and the constraints of small sample size.

H. Brain Tirmors

The incidence of chemosensory changes caused by intracranial tumors has rarely been investigated.In a study of 750 consecutivepatients presenting with chemosensory disorders to the University of Pennsylvania Smell and Taste Center, only two cases (0.3%) were induced by brain tumors (Deemset al., 1991).In one study anosmia was reportedly present in 19 of the 26 cases of Foster- Kennedy syndrome (retrobulbar optic neuritis, central scotoma, optic atrophy on the side of the lesion and con- tralateral papilledema usually occurring in tumors of the frontal lobe of the brain which press downward) (Jarus and Feldon, 1982).Bakay (1984) emphasizedthat loss of smell perception is one of the first signs of olfactory meningiomas. In general,tumors or other destructivelesions involving the olfactory bulb, tract, or prefrontal lobe may cause olfactory deficits. Temporal lobe tumors usually cause olfactory hallucinations. It is estimatedthat approximately 20% of the tumors of the temporal lobe produce some form of olfactory disturbance (Furstenberg et al., 1943). The presence of olfactory deficits correlates more with the location of tumors than the histology (Fig. 5).

I. AcquiredlmmunodeficiencySyndrome

Figure 4 A 20-year-oldwoman with posttraumaticanosmia. Olfactory deficits of patients with human immunodefi- (A) A smallolfactory tract is seenon the right side(arrow), but ciency virus (HIV) infection have been reported (Brody noneis seenon the left. Severeinferior frontal lobe encephalo- Heald et al., 1998).These authors suggest that malacia is soon on this coronal TIW MR scan. (B) et al., 1991; Encephalomalaciais well seenon the T2W MR scan where impaired olfaction might serve as a marker of early central hyperintensesignal (S) has replacedthe inferior frontal lobes nervous system HIV involvement. The principal (wheresmell processing occurs). histopathological abnormalities in the brain of acquired immunodeficiency syndrome (AIDS) patients are in the At at our institution, 25 patients with posttraumatic subcortical structures,predominantly in the central white smell dysfunction were evaluatedby olfactory testing and matter, deep gray structures including the basal ganglia, MR. Quantitative and qualitative gradings for olfactory the thalamus, and the brain stem (Petito et al., 1986; Price bulb, tract, subfrontal region, hippocampus, and temporal et al., 1988).Everall et al., (1991)have found that the neu- lobe damagecomelated with olfactory testresults. Twelve ronal numerical density in the frontal cortex is signifi- patients were anosmic, 8 had severe impairment, and 5 cantly lower in HIV patients than in controls-a loss of were mildly or impaired. Olfactory bulb and tract (88Vo about38Vo of neuronsin the superior frontal gyrus in AIDS of patients), subfrontal (60Vo), and temporal lobe (32Vo) patients. This may account for the olfactory deficits in injuries were found but did not coffelate well with olfac- these patients. 606 Li et al.

Therefore, the possibility of peripheral cause of olfactory deficits in AIDS patients also has to be taken into account in cerlain cases.

J. Multiple Sclerosis

Multiple sclerosis (MS), a markedly debilitating neurological disease,affects millions of Americans in the prime of their lives. Though the influence of MS on the sense of smell has long been controversial, recent MRI studies (Doty et al., 1991, 1999) have demonstratedthat the olfactory function in patients with MS is closely corre- lated with the number of demyelinating plaques within central olfactory processing areas of the brain, as deter- mined by MRI (Fig. 6, 7). A strong negative relationship (Spearmanr : -0.94) was found between scores on the University of Pennsylvania Smell Identification Test (UPSIT) and the number of plaques within the inferior frontal and temporal lobe regions (Doty et al., 1997). A close associationwas present, longitudinally, between the remission and exacerbationof plaque numbers and UPSIT scores,with lower UPSIT scoresocculring during periods of exacerbation(Doty et al., 1999).

K. Other Central Causes

There are also reports of olfactory dysfunction in hypochondriasis, amyotrophic lateral sclerosis, epilepsy, and migraine (Doty et al., 1991b; Mott and Leopold, 1991).Although the pathogenesisof olfactory dysfunction in these disorders is still unclear, it appearsthat a central mechanism is involved, rather than a peripheral one.

VL OVERVIEWANDDISCUSSION Figure 5 Temporallobe massin a 62-year-oldwoman with olfactoryhallucinations. (A) T2W MR scanreveals a relatively It is apparentfrom the studiesreviewed in this chapter and well-definedright temporallobe masswith mild masseffect. (B) the information presented elsewhere in this volume that Contrast-enhancedT1W MR image showsperipheral enhance- olfactory dysfunction can be due to numerous causes. ment of the tumor with a satellitenodule laterally.Sulci are Once an olfactory disorder has been recognized, the most effacedand the temporalhorn is obliterated. important step in the diagnostic processis to determine the site of the lesion, i.e., anatomical localization. Neuroradiological study has found that patients with Unfortunately, cur:rentclinical olfactory testing is unable HIV infection show widened cortical sulci, enlarged ven- to localize the site of morphological changes(Doty et al., tricles, cerebral atrophy, and brain stem atrophy when 1984). Modern medical imaging techniquescan be of great comparedwith controls(Brun et al., 1986;Elovaara et al., value in the anatomical classification and localization of 1990;Post et al., 1988).Opportunistic infections and CNS the common causes of olfactory dysfunction (Li et al., lymphoma may be superimposal on these changes. The 1994). The most common source of olfactory dysfunction pathogenesis of the olfactory deficits of AIDS patients is the peripheralpathway (Goodspeedet al., 1987; Mott needs further investigating but most likely will relate to and Leopold. 1991).In the evaluationofperipheral causes, diseasein the prefrontal lobe. In addition to CNS changes, the "sinus series"radiographs offer limited information. At sinusitis in HlV-infected patients is common and severe. present, high-resolution CT, especially coronal scans, is Medical Imaging of Olfactory Deficits 607

Figure 7 Axial TzW MRI scansfrom a severelymicrsomic (UPSIT: 20) 50-year-oldman with an 8-yearhistory of MS. The placeof sectionin (A) is 6 mm belowthat of (B). Notethe promi- Figure 6 A 55-year-oldMS patientwith no significantolfac- nent plaques(10 X 5 mm each)within the posteriorpart of the tory dysfunction,as measuredby the UPSIT.Axial T2W MRI white matter of the gyrus rectusof the L and R subfrontallobe scanshows no obviousplaques in theinferior frontal and tempo- regions(arrows I and2,respectively), and the relatively high signal ral loberegions (A). Numerousplaques were identified in supra- intensityplaques in the subtemporallobe regions(arrows 3 and4). periventricularregions (B). abnormalities of the brain parenchyma revealed by neu- roimaging studies in patients with AD, KP, or schizophre- the study of choice to look at the bony sinonasalstructures nia involve central brain areas that contain netrons of and the ostiomeatal complex. CT can also provide impor- olfactory projections including areasof the prefrontal lobe, tant information as a road map, which may be neededfor temporal lobe, hippocampus,and thalamus. surgical treatment. Recent studies have provided a clear physiological MRI possessesspecial ability in soft tissue discrimina- explanation for decreasedolfactory function in patients tion and offers multiplanar capabiTities.In the evaluation with MS (Doty et al., 1997, 1998, 1999). Current studies of the central causesof olfactory disturbances,MRI has a from our laboratory suggest that MS, with its relatively paramount role. Neuroimaging studies of patients with discrete focal regions of demyelination lesion, may be of olfactory deficits related to neuropsychiatric problems value in studying brain regions involved in sensory per- have revealed interesting findings and possibly clues for ception in addition to olfaction. understandingsome of the links between olfactory deficits It is much more difficult to explain the olfactory dys- and pathophysiological changes of the brain. The neu- function in PD patients,and presentlyimaging studieshave roimaging findings of patients with AD, KP, or schizo- been of little use in clarifying this matter. Loss of olfaction phrenia share some similarities. Thus, almost all of the in these patients may be related to factors with dopamine 608 Li et al. and dopamine receptors, although, as noted earlier, no RET'ERENCES return of function accompaniest--dopa treatment. In addi- tion, pathological changesin the areasof putamen and cau- Alavi, A., and Hirsch, I. J. (1991). Studies of central nervous sys- date nuclei, which have fibers connected with limbic tem disorders with single photon emission computed tomo- systemand striatum, may contribute to the loss of the sense graphy and positron emission tomography: Evolution over the of smell. In this hypothesis,the olfactory dysfunction in PD past 2 decades.Semin. Nucl. Med.2l:58-81. Albert M, Naser M. A, Levine H. L, et al. (1984). Ventricular size patientsmight sharea similar etiology to patientswith HD. in patients with presenile dementia of the Alzheimer's type. In congenital disorders, such as Kallmann's syndrome, Arch. Neurol. 4l :1258-1263. the cause of anosmia can be seen on MRI studies as the Allphin AL, Strauss M, and Abdul-Karin F. W (1991). Allergic absenceof olfactory bulbs (Yousem et al., 1993,1996a). fungal sinusitis: Problems in diagnosis and treatment. Other congenital abnormalities, such as choanal atresia Laryngoscope101:815-820. and meningoencephaloceles,also can be detectedby imag- Atlas S. W, Braffman B. H, LoBrutto R, Elder D. E, Herlyn D ing studies (Klein et al., 1987; Moorman et al., 1984). (1990). Human malignant melanomaswith varying degreesof In the categoriesof headtrauma and brain tumors. imag- melanin content in nude mice: MRI imaging, histopathology" ing studieshave shown strong links betweenolfactory dys- and electron paramagnetic resonance. L Comput. Assist. function and the location of the damaged brain. The Tomog. 14:.547-554. (1983). histology of the tumor or traumaticinjury is lesscritical than Averback, P. Two new lesions in Alzheimer's disease. Lancet 2:7203. its location (Costanzoand Zasler, 1991; Jamrs and Feldon, Bakay L. (1984). Olfactory meningiomas: The missed diagnosis. 1982; Schellingeret al., 1983;Yousem et al., 1996b). JAMA251:53-55. Hyposmia or anosmia induced by occupational or acci- Ball M. J, FishmanM, Hachinski V et al. (1985).A new definition dental exposure to toxins, as well as that induced by of Alzheimer's disease: A hippocampal dementia. lnncet l: intranasaluse of drugs such as cocaine, has been tradition- 14-16. ally thought to be due to damage to the peripheral path- Barat J L (1990). Mucoceles of the sphenoidalsinus. Report of six ways. However, one study has suggestedthat olfactory casesand review of the literature. J. Neuroradiol. l7: 135-151. deficits caused by occupational exposure to toxins may Barkovich A J, Vandermark P, Edwards MSB, and Cogen P H have both peripheral toxic and CNS effects (Schwartz (1991). Congenital nasal masses:C T and M R imaging fea- et al., 1989).Imaging studies haye shown CNS complica- turesin 16 cases.A./R 156:587-598. tions in cocaine abusers(Holman et al., 1991; Kalkow et Barkovich A J, and Quint, D. J. (1993). Middle interhemispheric fusion: an unusual variant of al., 1988;Pascual-Leone et al., 1991;Tumeth et al., 1990), holoprosencephaly. A,INR 14:431440. and one report ofanosmia as a sequelaofhydrogen sulfide Baroody F. M, and Naclerio RM (1991). Allergic rhinitis. In (H2S) inhalation suggestedthe loss to be due to central Smell and Taste in Health and Disease, Getchell TV et al. brain damage (Tvedt et al., 1991). (Eds.). Raven Press,New York, pp. 529-552. Benes F. M, Davidson J, and Bird E. D (1986). Quantitative cytoarchitectural studies of the cerebral cortex of schizo- VII. SUMMARY phrenics. Arch. Gen. Psychiatry 43:31-35. Bonte B. L, Tintner R, Weiner MF, et al. (1993). Brain blood flow Medial imaging is an essential part of the evaluation of in the dementias: SPECT with histopathologic correlation. patients with olfactory disorders. In the assessmentof the Radiology I 86:36I -365. peripheral causesof olfactory deficits, medical imaging stud- Braffman B. H, GrossmanR. I, GoldbergH. I, et al. (1989).MR ies, especially CT and/or MRI, can reveal anatomical infor- imaging of Parkinson disease with spin-echo and gradient- mation and structural changes,suggest differential diagnosis, echo sequences.A"/R 152:159-165. and provide the road map that may be needed for surgical Brasch R. C (1992). New directions in the development of MR intervention. On the other hand, in the evaluation of the cen- imaging contrastmedia. Radiolagy 183:1-11. Brody D, (1991). tral causes,MRI, f[4RI, PET, or SPECT can provide infor- Serby M, Etenne N, et al. Olfactory identifica- tion deficits in HIV infection. Am. J. Psychiatry L48:248-250. mation elucidating the links between olfactory dysfunction Brooks D. J, Ibanez Y Playford E. D, et al. (1991). Presynaptic and the structural or functional changesin the livine brain. and postsynaptic striatal dopaminergic function in neuroacan- thocytosis: A positron emission tomographic study. Ann. Neurol. 3O:166-l'71. ACKNOWLEDGMENTS Brun B, Boesen F, Gerstoft J, et al. (1986). Cerebral computed tomography in men with acquired immunodeficiency syn- Supported,in part, by Grants ROl DC04278, ROI drome. Acta Radiol. 2'l :385-387 . 4G11496,ROl DC 02974,and POI DC0016l (R. L. Buchwald C, Nielsen L. H, Ahlgren P, et al. (1990). Radiologic Doty, PrincipalInvestigator). aspectsof inverted papilloma. Eur. I. Radiol. 10:1 34-139. Medical Imaging of Olfactory Deficits 609

Carter B. L, and Runge V. S (1988). Imaging modalities for the Doty R. L, BartoshukL. M, and Snow J. B Jr (1991). Causesof study of paranasal sinuses and nasopharynx. Otolaryngol. olfactory and gustatory disorders. ln Smell and Thste in Clin. North Am. 2l:395420. Health and Disease, Getchell TV et al. (eds.). Raven Press, Cohen W (1990). Recent developments in the imaging of neu- New York. pp.449462. raxis trauma.Cum Opin. Radiol.2:34-39. Doty R. L, Kimmelman C. P, and Lesser R. P (1992a). Smell and CostanzoR. M, and Zasler N. D (1991). Head trauma.ln Smell taste and their disorders.In Diseases of the Nenous System: and Taste in Health and Disease, Getchell TV et al. (Eds.). Clinical Neurobiology,Yol. l, Asbury AK, Mckhann GM, and Raven Press,New York, pp. 7 ll-:730. McDonald WI (Eds.), WB. Saunders, Philadelphia, pp. Cowart B. J, Flynn-Rodden K, McGeady S. J, and Lowery L. D 390-403. (1993). Hyposomia in allergic rhinitis. J. Allergy CIin. Doty, R. L, Stern M. B, Pfeiffer C, et al. (1992b). Bilateral olfac- Immunol. yl:l4l-lJl. tory dysfunction in early stage medicated and unmedicated Crysdale WS (1981). Otorhinolaryngologic problems in patients idiopathic Parkinson's disease. J. Neurol. Neurosurg. with craniofacial anomalies. Otolaryngol. Clin. North Am. Psychiatry55: I 38-142 14:145-155. Doty, R. L, Bromley, S. M, Stern, MB (1995). Olfactory testing as Daly B. D, RussellJ. L, Davidson M. J, and Lamb J. T (1990). an aid in the diagnosis of Parkinson'sdisease: development of Thin section computed tomography in the evaluation of naso- optimal discrimination criteria. Neurodegeneration4;93-9'l . ethmoid trauma. Clin. Radiol. 4l : 1261-12'7 2. Doty R. L, Li C, Mannon L. J, and Yousem D. M (1997a). De Morsier G, and Gauthier G (1963). La dysplasie olfacto-gen- Olfactory dysfunction in multiple sclerosis [published erra- rtale. Path. Biol. ll:(l)267-272. tum appearsin N Engl J Med 1997; 337('l):5071. N. Engl. J. de Leon M. J, McRae T, Tsai JR, et al. (1988). Abnormal cortical Med. 336:1918-1919. responsein Alzheimer's diseaselinked to hippocampal atro- Doty R. L, Yousem D. M, Pham L. T, et al. (1997b). Olfactory phy. lnncet 2:391-392. dysfunction in patients with head trauma. Arch. Neurol. de Leon M. J, GeorgeA. E, ReisbergB, et al. (1989).Alzheimer's 54:1131-1140. disease:Longitudinal CT studies of ventricular change. A"/R Doty R. L, Li C, Mannon L. J, and Yousem D. M (1999). 153:1257-1262. Olfactory dysfunction in multiple sclerosis:Relation to longi- DeemsD. A, Doty R. L, SettleR. G, et al. (1991).Smell and taste tudinal changesin plaque numbers in central olfactory struc- disorders: A study of 750 patients from the University of tures.N eurolo gy 53: 880-882. Pennsylvania Smell and Taste Center. Arch. Otolaryngol. Drayer B. P, Olanow W, Burger R et al. (1986). Parkinson plus Head Neck Surg. ll7:519-528. syndrome diagnosis using high field MR imaging of brain Di Rienzo, L., Artuso, A., and Colosino, C. (2002). Isolated ir on. Radiol ogy | 59:493498. congenital agenesis of olfactory bulbs and tracts in a child Drutman J, Babbel RW, HarnsbergerHR, et al. (1991). Sinonasl without Kallman's syndrome. Ann. Otol. Rhinol-Larygol. polyposis. Semin. Ultrasound CT MR 12:561-574. lll:657-660. Elkon D, Hightower S. I, Lim ML, et al. (1979). Esthesio- Donnal J. R Heinz E. R, and Burger P. C (1990). MR of neuroblastoma.Cancer M:1087 -1094. reversible thalamic lesions in Wemick's syndrome.A./NR I l: Elovaara I, Poutiainen E, Raininko R, et al. (1990). Mild brain 893-894. atrophy in early HIV infection: The lack of association with Doty RL (1991). Olfactory dysfunction in neurodegenerativedis- cognitive deficits and HlV-specific intrathecal immune orders. In Smell and Taste in Health and Disease, Getchell response.I. Neurol. Sci. 99:.12l-136. TV et al. (Eds.).Raven Press, New York, pp. 735-751. Esiri M. M, and Wilcock G. M (1984). The olfactory bulbs in Doty R. L, Frye R (1989). Influence of nasal obstruction on smell Alzheimer's disease.J. Neurol. Neurosurg.Psychiatry 47i56--60. function. Otolaryngol. Clin. North Am. 22:3974ll. Everall I. P, Lutherat P. J, and Lantos P. L (1991). Neuronal loss Doty, R.L. & Mishra, A. (2001). Influences of nasal obstruction, in the frontal cortex in HIV infection. Lancet 337:1119-1121. rhinitis, and rhinosinusitis on the ability to smell. Falkai P, Bogerts B, and Rozumek M (1988). Limbic pathology Laryn g o s c op p e | | | :409423. in schizophrenia: The entorhinal region - a morphometric Doty R. and Practice of Rhinology, Goldman J (Ed.). John Wiley s1rtdy.Biol. Psychiatry 24:515-521. & Sons,New York, pp.761-785. Fowler J. S, Hoffman E. J, Larson S. M, et al. (1988). Positron Doty RL, Shaman P, and Dann M (1984). Development of the emission tomography: A new approach to brain chemistry. University of Pennsylvania Smell Identification Test: A stan- JAMA 206:2704-2710. dardized microencapsulated test of olfactory function. FurstenbergA. C, Crosby E, and Farrior B (1943). Neurologic P hys iol. B ehav. (M onog r ) 32:489-502. lesions which influence the senseof smell. Arclr. Otolaryngol. Doty R. L, Reyes P. F, and Gregor T (1987). Presenceof both 38:529-530. odor identification and detection deficits in Alzheimer's dis- Gallucci M, Bozzao A, Splendiani A, et al. (1990). Wernicke ease.Brain Res.Bull. 18:597-600. encephalopathy: MR findings in five patients. A"/R Doty R. L, DeemsD. A, Stellar S (1S88).Olfactory dysfunction 155:1309-1314. in parkinsonism: A general deficit unrelated to neurologic GeorgeA. E, deleon M. J, FerrisS. H, et al. (1981).Parenchymal signs, disease stage, or disease duration. Neurology 38; CT correlates of senile dementia (Alzheimer's disease):Loss 1237-1244. of grey-white discriminality. AJN R 2:205-213. 610 Li et al.

George A. E, de Leon MJ, Rosenbloom S, et al. (1983). Jacobson, R. R, and Lishman, W. A (1990). Cortical and dien- Ventricular volume and cognitive deficit: A computed tomo- cephalic lesions in Korsakoff's syndrome: A clinical and CT graphic study. Radiology 149:493498. scan study. Psychol. Med. 20:63-7 5. George A. E, de Leon M. J, Stylopoulos L. A, et al. (1990). CT Jafek, B. W, Eller, P. M, Johnson, E. W et al. (1992). diagnostic features of Alzheimer disease: Importance of the Ultrastructural changes of the olfactory epithelium in choroidal/hippocampal fi ssurecomplex. A"/NR I I : 101-107. Alzheimer's disease.Am. J. Rhinol. 6:219-225. Goodspeed,R. B, Gent, J. F, Leonard G, et al. (1987). The preva- Jagust,W. J, and Eberling, J. L (1991). MRI, CT, SPECT,PET: lence of abnormal paranasal sinus x-rays in patients with Their use in diagnosing dementia. Geriatrics 46:28-j5. olfactory disorders. Conn. Med. 5 1:1-3. Jarus, G. D, and Feldon, S. E (1982). Clinical and computer Gordon, A. S, Moran, D. T, Jajek, VW, et al (1990). The effect of tomographic findings in the Foster Kennedy syndrome.An. J. chronic cocaine abuseon human olfaction. Arch. Otolaryngol. Opht halmo l. 93:3 l7 -322. Head Neck Surg. 116;7415-1418. Jernigan, T. L, Schafer, K, Butters, N, and Cermak, LS (1991). Graamans,K, and Slootweg, P. J. (1989). Orbital exenterationin Magnetic resonanceimaging of alcoholic Korsakoff patients. surgery ofmalignant neoplasmsofthe paranasalsinuses. The Neurcp sy chop hannaco lo gy 4 :17 5-186. value of preoperative computed tomography. Arch. Jolles, P. R, Chapman, P. R, and Alavi, A (1989). PET, CT, and Otolaryngol. Head Neck Surg. 115:977-980. MRI in the evaluation of neuropsychiatric disorders: Current Gregler, L. L, and Mark, H (1986). Medical complications of applications. J. Nucl. Med. 30:1589-16. cocaine abuse.N. Engl. J. Med.3l5:1495-l5OO. Jones, B. P, Moskowitz, H. R, Butters N, et al. (1975). Gussack,G. S, and Hudgins,P. A. (1991).Imaging modalitiesin Psychophysical scaling of olfactory visual and auditory stim- recurrent head and neck tumors. ktryngoscope l0l:ll9-124. uli by alcoholic Kosakoff's patients. Neuropsychologica 13: Hagan, P. J (1967). Post-traumatiacanosmia. Arch. Otolaryngol. 387-393. 85:107*lI 1. Joyce,E. M, and Robbins,T. W (1991).Frontal lobe function in Happer, C, Kril, J, Daly J (1987). Are we drinking our neurones Korsakoff and non-Korsakoff alcoholics' planning and spatial away? Br Med. J.294:534-536. working memory. N euro p syc ho lo g ica 29:7 09-7 23. Harris, G. J, Schlaepfer,T. E, Peng, L. W, Lee S, Federman,E. B, Kallmann, F. J, Schoenfeld,W. A, and Barrera, S. E (1944). The and Pearlson G. D (1994). Magnetic resonanceimaging eval- genetic aspectsof primary eunuchoidism. Am. J. Ment. Defic. uation of the effects of ageing on grey-white ratio in the 48:203-221. human brain. Neuropathol. Appl. Neurobiol.2O 29V293. Kassel, E. E (1988). Traumatic injuries of the paranasalsinuses. Hanis, GJ, Aylward, E. H, PeyserC. E, et al. (1996). Single pho- Otolaryngol. Clin. North Am. 2l:455493. ton emission computed tomographic blood flow and magnetic Kelly, A. B, Zimmerman,R. D, SnoW R. B, et al. (1988). Head resonance volume imaging of basal ganglia in Huntington's trauma: Comparison of MR and CT-experience in 100 disease.Arch. Neurol. 53:316-324. patients.AJNR 9:699J08. Harris, G. J, Lewis, R. F, Satlin,A, et al. (1998). Dynamic sus- Kem, R. C. (2000). Chronic sinusitis and anosmia: pathologic ceptibility contrast MR imaging of regional cerebral blood changes in the olfactory mucosa. lnryngoscope I l0: volume in Alzheimer disease: a promising alternative to toTl-to77. nuclear medicine. -17 AJNR 19:.1727 32. Kesslak,J. P,Nalcioglu, O, and Cotman,C. W (1991). Qualification Harrison, P. J (1986). The pathogenesisof Alzheimer's disease: of magnetic resonance scans for hippocampal and parahip- Beyond the cholinergic hypothesis. J. Roy. Soc. Med. pocampal atrophy in Alzheimer's disease.Neurolog 4l: 5l-54. 79:347-351. Kido, D. K, Caine, E. D, LeMay M, et al. (1989). Temporal lobe Hayden, M. R, Martin, W. R. W, Stoessl,A. J, et al. 0986). atrophy in patients with Alzheimer's disease: A CT study. Positron emission tomography in the early diagnosis of A"/NR l0:551-555. Huntington'sdisease. Neurology 36:888-894. Kimmelman, C. P (1991). Medical imaging of smell and raste Heald A. E, Piper C. F, and Schiffman SS (1998). Tasteand smell disorders. ln Smell and Tastein Health and Disease, Getchell complaints in HlV-infected patients. AIDS l2:166i-1674. TV, et al. (Eds.). Raven Press,New York, pp. 471479. Healy B (1992). From form to function - better imaging techniques Klein, V. R, Friedman, J. M, Brookshire, G. S, er al. (1987). extend study of living system (from NWI) . JAMA 267 : 2863. Kallmann's syndrome associatedwith choanal atresia. Clin. Holman B. L, Carvalho P. A, MendelsonJ, et al. (1991). Brain Genet. 3l:224-227 . perfusion is abnormal in cocaine-dependentpolydrug users:A Klingmuller, D, Dewes W, Krabe T, et al. (1987). Magnetic reso- study using Technetium-99m-HMPAO and ASPEC. J. Nucl. naneeimaging of the brain in patients with anosmia and hypo- Med.32:1206-1210. thalamic hypogonadism (Kallmann's syndrome). J. Clin. Hunink, M. G, de Slegte, R. G, Gerritsen, G. J, Speelman, H Endocrinol. Metab. 65 :58I-584. (1990). CT and MRI assessmentof tumors of the nose and Kolkow, N. D, Mullani, N, Gould, K. L, et al. (1988). Cerebral paranasal sinuses, the nasopharynx and the parapharyngeal blood flow in chronic cocaine users: A study with positron spaceusing ROC methodology. Neuroradiology 3?:2ZO-225. emission tomog. Br J. Psychiatry L52:641-648. Hyman, B. T, Van Hoesen, G. W, and Damasio, A. R (1934). Kopelman, M. D (1991). Frontal dysfunction and memory Alzheimer's disease:Cell-specific pathology isolates the hip- deficits in the alcoholic Korsakoff syndrome and Alzheimer- pocampel formation. Science 225:1 168-l l7 0. type dementia. Brain ll4:ll'7-13'1 . Medical Imaging of Olfactory Delicits 611

Kraus, D. H, Lanzieri, C. F, Wanamaker, J. R, Little, J. R, and Mehra,Y. N; Mann, S. B; Dubey,S. P; Suri S. (1989).Computed Lavertu P (1992). Complementary use of computed tomogra- tomography for determining pathways of extension and a phy and magnetic resonanceimaging in assessingskull base staging and treatment system for juvenile angiofibromas. Ear lesions.Laryngoscope lO2:623-629. N ose Throat J. 68:576-589. Kuriloff, D. B (1989). Nasal septalperforation and nasal obstruc- Mesulam, M. M (1990). Schizophreniaand the brain. N. Engl. J. tior7.Ot olaryngoL CIin. N orth Am. 22:333-350. Med.322:842-845. Lanziei, C. F., Shah, M., Krauss, D., and Lavertu, P. (1991). Use Metter, E. J, Kempler D, Jackson C, et al. (1989). Cerebral glu- of gadolinium-enhanced MR imaging for differentiating cose metabolism in Wernickes's. Broca's. and conduction mucoceles from neoplasms in the paranasal sinuses. aphasia.Arch. Neurol. 46:27-34. Radiology 178:.425428. Moberg, P. J, Pearlson, G. D, Speedie,L. J, et al. (1987). Leigh, A. D (1943). Defects of smell after head injwy. ktncet Olfactory recognition: Differential impairments in early and 1:38-40. late Huntington's and Alzheimer's dtsease. J. Clin. Exp. Levin, H. S, High, W. M, and Eisenberg, H. M (1985). N europsy chol. 9 :650-664. Impairment of olfacory recognition after closed head injury. Montgomery, E. B, Koller, W. C, LaMantia, T. J, et al. (2000). Brain 108:579-591. Early detection of probable idiopathic Parkinson's disease: Li C,Yousem, D. M, Hayden RE, and Doty RL (1993). Olfactory Development of a diagnostic test battery. Movement Disord. neuroblastomas:MR evaluation.A"/NR 14:1167-l l7 1. 15:467473. Li C, Yousem DM, Doty RL, and Kennedy DW (1994). Moorman, J. R, Crain B, and Osborne D (1984). Kallmann's syn- Neuroimaging in patients with olfactory dysfunction. A,/R drome with associatedcardiovascular and intracranial anom- 162:411418. alies.Am. J. Med. 77:369-372. Lieblich, J. M, Rogol, A. D, White, B. J, et al. (1982). Syndrome Mott, A. E, and Leopold, D. A (1991). Disorders in taste and of anosmia with hypogonadism Kallmann's syndrome) An. .L smell Med. Clin. North Am.'15:1321-1353. Med.73:506-519. Myers, R. H, Vonsattell, J. P, Paskevich, P. A, et al (1991). Loury, M. C, and Kennedy,D.W (1991). Chronic sinusitisand Decreasedneuronal and increased oligodendroglial densities nasal polyposis. In Smell and Taste in Health and Disease, in Huntington's diseasecaudate nucleus. J. Neuropathol. Exp. Getchell TV et al. (Eds.).Raven Press, New York, pp. 517-528. Neurol. 50:729-742. Lloyd, G. A, and Barker, P. B. (1991). Subtraction magnetic res- Nagel, J. S, Johnson,K. A, Ichise M, et al. (1988) Decreased onance for tumors of the skull base and sinuses:a new imag- iodine-I23 IMP caudate nucleus uptake in patients with ing technique. J. Inryngol. Otol. lO5:628-631. Huntington's disease.Clin. NucI. Med. 13:486496. Lloyd, G. A. S, Lund, V. J, Phelps,P. D, Howard, D. J. (1987). Naser, M. A, Gebhardt C, and Levine, H. L (1980). Decrease MRI in the evaluation of nose and paranasal sinus diseases. computerized tomography numbers in patients with presenile Br. J. Radiol. 60:957-968. dementia.Arch. Neurol. 37:401409. Lyons, B. M, Donald, P. J. (1991). Radical surgery for nasal cav- Newbill, E. T, Johns, M. E, Cantrel,l R. W (1985). ity and paranasalsinus tumors. Otolaryngol. Clin. North Am. Esthesioneuroblastoma:Diagnosis and management. South. 24:1499-1521. Med. J. "18:275-282. Mair, R. G, Doty, R. L, Kelly, K. M, et at. (1986). Multimodel Newman, N. M, Diloreto, D. A, Ho, T. I, et al. (1988). Bilateral sensory discrimination deficits in Korsakoff's psychosis. optic neuropathy and osteolytic sinusitis. Complications of N europ sy c ho Io gic a 24:83 | -839. cocaine abuse.JAMA 259:72--7 4. Mair, G. R, Knoth, R. L, Rabchenuk, S. A, et al. (1991). Ohm, T. G, Braak H (1987). Olfactory bulb changes in Impairment of olfactory auditory, and spatial serial reversal Alzheimer's disease.Acta Neuropathol. (Berlin) 73:365-369. learning in rats recovered from pyrithiamine-induced thi* Ohnishi,T, Hoshi, H, NagamachiS, et al. (1991) Regionalcere- amine deficiency. Behav. Neurosci. fi5 :360-374. bral blood flow study with I23J-JMPin patients with degener- Males, J. L, Townsend, J. L, and Schneider,R. A (1973). ative dementia. AJNR 12:513-520. Hypogonadotrophic hypogonadism with anosmia Paling, M. R, Black, W. C, Levine, P. A, et al. (1987). Tumor Kallmann's syndrome.Arch. Int. Med. 13I:501-507. invasion of the anterior skull base:A comparison of MR and Masters,S. J, McClean, P. M, Arcarese,J. S, et al. (1987). Skull CT studies.J. Comput. Assist.Tomogr. I 1:824-830. x-ray examinations after head trauma. Recommendationsby Parker, G. D, and Harnsberger,H. R (1991). Clinical-radiologic multidisciplinary panel and validation study. N. Engl. J. Med. issuesin perineural tumor spread of malignant diseasesof the 316:84-91. extracranial head and neck. RadioGraphics ll:383-399. Mathog, R. H (1992). Atlas of Craniofocial Trauma. W.B. Pascual-LeoneA, Dhuna A, and Anderson D (1991). Cerebral SaundersCo., Philadelphia, pp. 377-379. atrophy in habitua cocaine abusers:A planimetric CT study. McAlister, W. H, Lusk R, Muntz, H. R (1989). Comparisonof Neurology4l:34-38. plain radiographs and coronal CT scans in infants and chil- Pearson,R. C. A, and Powell, T. P. S (1989). The neuroanatomy dren with recurent sinusitis. AJR 153:1259-1264. of Alzheimer's disease.Rev. Neurosci. 2:l0l-122. McDonald, W. M, Krishnan, K. R, Doraiswamy, P. M, et al Petito, C. K, Cho E-s, Lemann W, et al. (1986). Neuropathology (1991). Magnetic resonance findings in patients with early- of acquired immunodeficiency syndrome (AIDS): An autopsy onset Alzheimer's disease.Biol. Pstchiatn 29:799-810. review. -/. N europathol. Exp. N eurol. 45 :635446. 612 Li et al.

Pfefferbaum, A., Sullivan, E. V., Mathalon, D. H., Shear, P. K., Serby M, Larson P, and Kalkstein, D (1990). Olfactory sensein Rosenbloom,M. J., and Lim, K. O. (1995). Longitudinal psychoses.Biol. Psychiatry 28:830. changes in magnetic resonance imaging brain volumes in Serby, M, Larson, P, and Kalkstein, D (1991). The nature and abstinent and relapsed alcoholics. Alcoholism: Clinical and course of olfactory deficits in Alzheimer's disese. Am. ./. Experimental Res earch 19:ll7 "l -119 l. Psychiattl 148:357-360. Phillips, P. P, Gustafson,R. O, and Facer, G. W (1990). The clin- Shapiro, M. D, and Som, P. M (1989). MRI of the paranasal ical behavior of inverting papilloma of the nose and paranasal sinusesand nasal cavity. Radiol. Clin. North Am.27: 44'1475. sinuses: Report of 112 cases and review of the literature. Simmons,J. I PastakiaB, Chase,T. N, et al. (1986). Magnetic lnryngoscope 100:463-469. resonance imaging in Huntington's disease. Aru. J. Post,M. J. D, Tate,L. G, Quencer,R. M, et al. (1988).CT, MR, Neuroradiol. T:25-28. and pathology in HIV encephalitis and meningitis. AJR 151: Sisson, G. A Sr, Toriumi, D. M, and Atiyah, R. A. (1989). 373-380. Paranasal sinus malignancy; a comprehensive update. Potter, H, and Butters, N (1980). An assessmentof olfactory Lary ng os c o p e 99 :l 43-150. deficits in patients with damage to prefrontal cortex. Som, P. M (1991). Tumors and tumorlike conditions. In Head Neuropsycho lo gica 18621428. and Neck Imaging,Znd ed., Som PM (Ed.). Mosby-Year Book Price, J. L (1985). Beyond the primary olfactory correx: Inc., St. Louis, pp. 169-227. Olfactory-related areasin the neocortex, thalamus and hypo- Som, P. M, Shapiro,M. D, Biller, H. F, et al. (1988). Sinonasal thalamus. Chem. Senses lO:239-258. tumors and inflammatory tissues: Differentiation with MR Price, J. L (1990). Olfactory system. In The Human Nervous imaging. Radiology 167:803-808. Systenx,Paxinos, G. (Ed.). Academic Press, Inc., San Diego Som, P. M, Dillon, W. P, Curtin, H. D, et al. (1990). Hypointense pp.979-998. paranasalsinus foci: Differential diagnosis with MR imaging Price, R. W, Brew B, Sidtis J, et al (1988). The brain in AIDS: and relationto CT findings.Radiology 17677'l-181. Central nervous system HIV-I infetion and AIDS dementia Squire, L. R, Amaral, D. G, and Press,G. A (1990). Magnetic res- complex. Science 239:586-592. onance imaging of the hippocampal formation and mammil- Rausch R, and Serafetinides,E. A (1975). Specific alteration of lary nuclei distinguish medial temporal lobe and diencephalic olfaction in humans with temporal lobe lesions. Ncrrrre amnesia.J. Neurosci. 10:3109-31 17. 255:557-558. Starkstein, S. E, Folstein, S. E, Brandt J, et al. (1989) Atrophy in Rausch R, Serafetinides, E. A, Crandall PH (1977). Olfactory Huntington's disease: A CT:scan study. Neuroradiology memory in patients with anterior temporal lobectomy. Cortex 31:156-159. 13:445452. Suddath, R. L, Casanova,M. F, Goldberg, T. E, et al. (1989). Reid, I. C, Besson,J. A. O, Best, P.Y et al. (1988). Imaging of Temporal lobe pathology in schizophrenia. A quantitative mag- cerebral blood flow markers in Huntington's diseseusing sin- netic resonanceimaging study.Am. J. Psychiatry 146:464472. gle photon emission computed tomography. J. Neurol. Sumner, D ( 1964). Post-traumatic anosmia. B rain 8'7:lO7 -120. N eurosurg. P sychiatry 5 | :1264-1268. Suzuki, M, Takashima I Kadoya M, et al. (1989). MR imaging Reiman, E. M, and Mintun, M. A (1990). Positron emission of olfactory bulbs and tracts. A-INR 10:955-957. tomography (editorials). Arch. Intem. Med. 150:729-731. Tanabe,T, Jino M, TiakagiSF (1975). Discremination of odors in Roberts, G. W (1988). Abnormalities in brain sfucture in schiz- olfactory bulb, pyriform-amygdaloid areas,and orbitofrontal ophrenia. Curr Opin. Psychiatry 1:83-89. cortex of the monkey. J. N europhysiol. 381284-1296. Ron, M. A (1983).The alcoholicbrain: CT scanand psycholog- Tumeth, S. S, Nagel, J. S, English RI, et al. (1990). Cerebral ical findings. Psychol. Med. Monograplr, Supplement 3. abnormalities in cocaine abusers: demonstration by SPECT Ron, M. A, Acker W' Shaw GK, et al. (1982). Computeized perfusion brain scintigraphy.Radiology 176:821-824. tomography of the brain in chronic alcoholics: A survey and Turetsky, B. I, Moberg, P. J, Yousem, D. M, et al. (2000). fbllow-up study. Brain 105:497-514. Reduced olfactory bulb volume in patients with schizophre- SchechterPJ, and Henkin RI (1974). Abnormalities of taste and nia. Am. J. Psychiatry 157:828-830. smell after head trauma. J. Neurol. Neurcsurg. Psychiatry 37: Tvedt, B. Skyberg, K, Aaserud, O, et al. (1991). Brain damage 802-810. causedby hydrogen sulfide: A follow-up study of six patients. SchellingerD, Henkin RT, and SmirniotopoulosJG (1983). CT Am. J. Int. Med.20:91-101. of the brain in taste and smell dysfunction. AJNR 4:7 Van Tassel, P, Lee, Y. Y. (1991). Gd-DTPA enhanced MRI for 52-754. detecting intracranial extension of sinonasal malignancies. ./. Schwartz,B, Doty, R. L, Frye, R. E, et al. (1989). Olfactory func- Comput.As si st.Tomo gr. 15:387 -392. tion in chemical workers exposedto acrylate and methacrylate Victor, M (1990). MR in the diagnosis of Wernicke-Korsakoff vapors.Arn. J. Public Health 79:613418. syndrome.A"/R 1315-13 16. Schweitzer, V. G (1986). Osteolytic sinusitis and pneumomedi- Vogl, T, Dresel S, Bilaniuk LT, et al. (1990) Tumors of rhe astinum: Deceptive otolaryngologic complications of cocaine nasopharynxand adjacentareas: MR imaging with Gd-DTPA. abuse.Lary n g o s c op e 96:2O6-21O. AJNR l1:187-194- Medical Imaging of Olfactory Deficits 6t3

Vonsattel, J. P, Myers, R. H, Stevens, T. J, et al. (1985)' imaging during odor stimulation: Preliminary data. Neuropathological classification of Huntington's disease' J. N euroradiolo gy 204:833-83 8. Neuropathol. Exp. Neurol. 44:559-5'1'7. Yousem,D. M., Geckle,R. J., Bilker, W. B. & Doty, R. L. (1998). Young AH, Blackwood DHR, Roxborough H, et al. (1991). A Olfactory bulb and tract and temporal lobe volumes: magnetic resonance imaging study of schizophrenia: Brain Normative data across decades. Ann. NY Acad. Sci. structure and clinical symptoms. Br J. Psvchiatry 158-164. 855:546-555. Yousem, D. M, Fellows, D. W Kennedy, D. W, et al. (1992). Yousem, D. M., Geckle, R. J., Bilker, W B., Kroger, H. & Doty' Inverted papilloma: MR evaluation. Radiology 185:501-505. R.L. (1999a). Posttraumatic smell loss: Relationship to psy- Yousem, D. M, Li, C, Turner, W. J. D, et al. (1993). Kallmann's chophysical tests and volumes of the olfactory bulbs and syndrome: MR evaluation of olfactory system. AJNR tracts and the temporal lobes. Acad. Radiol.6:264-272. 14:839-843. Yousem, D. M., Maldjian, J. A., Hummel, T., Alsop, D. C.' Yousem, D. M, Geckle, R. J, BilkerW, McKeown DA, Doty RL. Geckle, R. J., Kraut, M. A., Doty, R. L. (1999b). The effect of (1996a). MR evaluation of patients with congenital hyposmia age on odor-stimulated functional magnetic resonanceimag- or anosmia. AJR 166:439443 ing. Am. J. Neuroradiol. 20:600-608. Yousem, D. M, Geckle R. J, BilkerWB, McKeown DA, and Doty Yousem,D. M., Maldjian, J. A., Siddiqi, F., Hummel, T., Alsop, RL. (1996b). Posttraumatic olfactory dysfunction: MR and D. C., Geckle, R. J., Bilker, W. B., and Doty, R. L. (1999c)' clinical evaluation. AJNR l'z :ll'7 1-ll7 9 Gender effects on odor-stimulated functional magnetic reso- Yousem,D. M., Li, C., Montone, K.T., et al. (1996c). Primary nanceimaging . Brain Res.818:480-487. malignant melanoma of the sinonasal cavity: MR imaging Zimmerman, R A, Bilaniuk, L. T, Hackney, D. B, et al. (1986). evaluation.RadioGraphics 16: 1 101-1 I 10. Head injury: early results of comparing CT and high-field Yousem,D. M., Geckle,R. J.,Doty, R. L. & Bilker, W. B. (1997a). MR. A,/R 147:121 5-1222. Reproducibility and reliability of volumetric measures of Ztnreich, S. J., Kennedy,D. W., Roenbaum,A. E., et al. (1987). olfactory eloquent structures.Acad. Radiol. 4:264-269. Paranasalsinuses: CT imaging requirements for endoscopic Yousem,D. M., Williams, S. C. R., Howard, R. O., Andrew, C., surgery.Radiology 163:769--7 7 5. Simmons,A., Allin, M., Geckle, R.J., Suskin, D., Bullmore, Zusho, H. (1982). Post-traumatic anosmia. Arch. Otolaryngol. E. T., Brammer, M. J. & Doty, R. L. (1997b). Functional MRI 108:90-92.