Log in Register

Login to your account

Username *
Password *
Remember Me

Create an account

Fields marked with an asterisk (*) are required.
Name *
Username *
Password *
Verify password *
Email *
Verify email *
Captcha *

Captcha Image Reload image challenge

Oral Medicine


Authors : Dr. Suraj Agarwal .


The past 15 years have witnessed dramatic changes in the imaging of a variety of head and neck disorders. Although initially the emphasis of imaging was on morphologic analysis, the areas of most active recent investigation have involved metabolic and functional imaging techniques. The primary goal of head and neck imaging has always been to provide anatomic information that accurately maps a lesion, especially in areas where the clinician cannot directly use palpation.

However, the early results of the new metabolic imaging techniques suggest that they may be more useful than cross-sectional anatomic imaging in improving the radiologist’s ability to accurately identify recurrent tumor, monitor treatment, and identify cervical metastases. This article will review the promising metabolic techniques and discuss their potential clinical applications.


The technique in which initial investigations havebeen promising in the area of head and neck oncology isthallium-201 single photon emission tomography (Tl-201SPECT). Initial investigations have shown that Tl-201SPECT may be beneficial in attempting to differentiate recurrenttumor from posttreatment changes, and may beuseful for both predicting and assessing the response ofpatients treated with nonsurgical organ preservationtherapy.1-7


The primary mechanism of thallium entry into the cell isbelieved to be coupled to the sodium/potassium ATPasepump located in the cell membrane.8 This system activelytransports potassium into the cell in exchange for sodium,thereby creating a high intracellular potassium concentration.Thallium-201 is believed to act similarly to potassium and competes with potassium for intracellular transportacross the cell membrane via the sodium/potassium ATPasesystem (potassium analogue).8,9,10,11 The reason for theelevated thallium-201 uptake in tumor has not been entirelyelucidated but is likely due to the increased cellularproliferation of neoplastic cells compared to adjacentnormal tissue.8,12,13


The primary disadvantage of thallium-201 imaging ofthe upper aerodigestive tract is the background uptake ofthallium-201 by salivary and thyroid glands, as thallium-201 is normally taken up by the parotid, submandibular, andthyroid glands.14 The uptake within the parotid andsalivary glands can be especially problematic since thesaliva will also have increased activity. Pooling of the salivain the floor of the mouth may prevent accurate evaluation oflesions involving the oral cavity and may reduce theconspicuity of tumors in this area. This activity may bereduced by placing gauze in the patient’s mouth and drying the oral mucosa. Another alternative in patients who haveno medical contraindication to the administration ofatropine is to use this drug to reduce salivary function priorto imaging.

Clinical Applications

Several studies have demonstrated that SCC is thalliumavid, and previous investigators have suggested thatthallium-201 may be able to detect lesions larger than1 cc.14 As a result, it has been suggested that thallium-201may be useful in the evaluation of patients with unknownprimary tumors in a manner similar to FDG PET scanning.Isolated cases of the successful use of thallium-201 forlocalizing unknown primary tumors have been reported.7

However, larger series that have prospectively evaluated therole of thallium-201 in identifying unknown primary tumorshave not yet been performed.Initial results have shown that thallium-201 SPECT maybe helpful in attempting to differentiate recurrent tumorfrom post-treatment changes and may be more accurate thanCT or MR imaging for making this important distinction. Valdes-Olmes et al. reported ahigher diagnostic accuracy for thallium-201 SPECT (sensitivity= 93%, specificity = 78%) than for CT/MR imaging(sensitivity = 76%, specificity = 30%).7Leitha et al. reported a sensitivity and a specificity of95.3% and 78.4%, respectively, for the ability oftechnetium-99m-MIBI (hexakis-2-methoxyisobutyl isonitrile)SPECT to differentiate recurrent head and neck SCCfrom post-treatment changes.15105 If these results are substantiated,PET FDG, thallium-201, and technetium-99m-MIBIimaging all compare favorably for detecting tumors inpatients with head and neck SCC.

There may be potential advantages to using thallium-201SPECT rather than FDG PET, especially when imagingnasopharyngeal carcinoma and other skull base tumors.Unlike FDG, the background uptake of thallium-201 bynormal brain is low. As a result, the conspicuity ofnasopharyngeal and skull base tumors may be reduced onPET-FDG, as these tumors are directly adjacent to thenormal high brain FDG activity. However, because of the low background activity in the normal brain, these tumorstend to be well delineated on thallium-201 imaging.1Initial reports have suggested that thallium-201 SPECTmay be useful in predicting the response to definitiveradiation therapy and for assessing the response within 6weeks after the completion of nonsurgical organ preservationtherapy. Nagamachi et al. demonstrated that quantitativemeasurements of thallium-201 uptake may be useful inpredicting the response of SCC of the upper aerodigestivetract to radiotherapy alone.4

Mukherji et al. suggested thatTl-201 SPECT may be a reliable technique for assessing theresponse to treatment performed within 6 weeks after thecompletion of combined chemotherapy and radiation therapy.2As previously mentioned,due to a high false-negative rate, FDG PET has been shownto be unreliable in predicting the treatment response whenperformed within 1 to 2 months after completion of radiationtherapy.16 Thus, initial results suggest that thallium-201 maybe able to predict radioresponsiveness and assess the successof treatment earlier than can FDG-PET or cross-sectional imaging. However, further studies are needed to confirmthese early results.



Recent in vitro and in vivo investigations have suggestedthat proton magnetic resonance spectroscopy (1H-MRS)may be helpful in providing metabolic information aboutabnormalities in the extracranial head and neck. Severalstudies have found that assessment of the relative levels ofcholine (Cho) and creatine (Cr) can provide information thatmay help differentiate SCC from nonneoplastic tissue andmay provide information that helps assess the response tononsurgical organ preservation treatment protocols.17,18,19,20Elevation of the Cho/Cr ratio appears to be a consistentfinding for SCC and has been demonstrated in cell culturesand from tissue samples from both the primary tumor siteand metastatic lymph nodes.17


The majority of 1H-MRS tissue characterization in theupper aerodigestive tract has focused on two metabolites,Cho and Cr. Cho is a nutrient that is present in most foods, isabsorbed from the diet, and serves as a precursor of twoimportant molecules: acetylcholine and phosphatidylcholine.21Cho is also believed to be an important constituent inthe phospholipid metabolism of all cell membranes.22TheCho molecule is transported intracellularly by a low-affinitysystem with its mechanism coupled to the synthesis ofphosphatidylcholine.23 Once intracellular, Cho is initiallyphosphorylated by Cho phosphotransferase to form phosphocholine.This latter molecule is the precursor ofphosphatidylcholine, which is formed by the Kennedypathway.24 The N-methyl [+/-N(CH3)3] group of Cho andCho-based metabolites is detected spectroscopically at 3.2ppm.25 The total Cho resonance peak is believed to becomposed of Cho, phosphocholine, phosphatidylcholine,and glycerophosphocholine. Elevation of this peak is notspecific for SCC and has been demonstrated in severaltumors. Increased levels of Cho are thought to be related to increased cellular membrane phospholipid biosynthesis andare, therefore, believed to be an active marker for cellularproliferation.26,27,28,29,30

Cr is thought to play a role in the maintenance of energymetabolism. Phosphocreatine acts as a store for high-energyphosphates in the cytosol of muscles and neurons and is alsobelieved to buffer cellular adenosine triphosphate andadenosine diphosphate.21,22 It may be obtained from thediet or synthesized de novo within the liver, kidneys, andpancreas by precursor molecules, which include arginine,glycine, and S-adenosylmethionine.21,31 Cr is convertedto phosphocreatine by the enzyme creatine kinase. Becausephosphocreatine is a high-energy phosphate compound, ithas been suggested that it may help sustain the level ofadenosine triphosphate in energy-demanding tissues such asactively proliferating tumors. Spectroscopically, the Cr resonance is observed at 3.02 ppm, and it is believed to becomprised of both phosphocreatine and Cr.31

Clinical Applications

The importance of identifying and isolating consistentand reproducible 1H-MRS metabolic markers for malignanciesof the upper aerodigestive tract rests not so much in theability of 1H-MRS to initially diagnose tumor, but in itsability to identify early recurrent tumor and to monitortumor response during therapy.32,33,34 Because bothsurgery and radiotherapy result in scarring and soft-tissueedema, physical examination and CT/MR imaging often areinadequate in detecting early tumor recurrence after treatment.
Fig1: A- Before Treatment B – After Treatment
1H MRS showing loss of the Cho peak at 3.2 ppm in a node from HNSCC
that responded to treatment

Radiation therapy causes significant erythema andinduration within the treated area, and this inflammatoryresponse may be exacerbated with the addition of concomitantchemotherapy.35 In most sites, this inflammatoryresponse progresses to end-stage fibrosis (scar) within 3 to 4months after the completion of radiation therapy. 36, 37 These treatment-associated changes often prevent thoroughendoscopic evaluation and reduce the ability to accuratelydetect persistent tumor that is unresponsive to therapy. Thistask is made even more difficult since persistent tumors areoften situated deep to the mucosal surfaces and are not oftendirectly visualized on physical examination. Although CTand MR imaging can be used to complement the clinicalexamination following treatment, their role is somewhatlimited due to their inability to confidently differentiatepost-treatment changes from early tumor recurrence.35 Inaddition, the expected post-treatment imaging findings differsomewhat with the type of treatment given and with thelocation and extent of the primary tumor.36, 37

Differentiation of Recurrent Tumor fromPost-treatment Changes

Early results suggest that the 1H-MRS may be used todifferentiate recurrent tumor from post-treatment changes inpatients with a malignancy of the upper aerodigestive tractwho have undergone previous treatment.Elevation of theCho/Cr ratio in an indeterminate mass seen on CT or MRimaging is suggestive of recurrent tumor, whereas a ratioclose to 1 is suggestive of posttreatment changes.16 Despitethe encouraging early 1H-MRS results, it is unlikely thatmicroscopic foci of tumor can be completely excluded by1H-MRS alone, regardless of field strength. Thus, indeterminatemasses without the characteristic spectral tumormarkers still warrant close clinical and radiographicfollow-up, regardless of the MRS findings.

Prospective Treatment Monitoring

One of the most promising clinical applications for 1HMRSis the noninvasive treatment monitoring of patientswith SCC of the upper aerodigestive tract who are treatednonsurgically. The accurate assessment of the primary tumor’sresponse to radiation therapy and/or chemotherapy isextremely important, as early identification of an unresponsivemalignancy may indicate the need for either additionaltherapy or surgical (salvage) resection. Previous studieshave shown that cross-sectional imaging may be helpful inpredicting tumor response by demonstrating a reduction intumor size. However, these serial studies are most accuratewhen performed at least 3 to 4 months after completion ofnonsurgical organ preservation treatment.

By comparison,1H-MRS has the potential to be a much earlier indicator oftumor response, not by measuring volumetric changes but bymeasuring the internal chemistry of the treated tumor cells.The non-invasive nature of 1H-MRS allows frequent intervalstudies to be performed both during and following treatment,

and the utility of 1H-MRS in prospective treatment monitoringfor SCC is currently being investigated. Preliminary investigations have shown that progressivereduction in Cho concentrations during treatment is indicativeof a tumor response.38

In fact, some investigatorshave suggested that alterations in 1H-MRS may be able topredict the response prior to changes in tumor volume.39 Thus, persistent elevation of the Cho/Crratio suggests a nonresponsive tumor, while resolution of theCho peak is indicative of a tumorresponse . Noinformation is available regarding the specific time intervalfollowing the initiation of treatment when one would expectto see resolution of spectral markers in tumors that are beingsuccessfully treated. Large-scale prospective studies arewarranted to determine if 1H-MRS can be a reliabletechnique for assessing the response during treatmentmonitoring.

Limitations andImaging Strategies

The upper aerodigestive tract is an inherently difficultarea to perform MRS. To obtain highly resolved spectra, avery homogeneous magnetic field is required in the area tobe examined (i.e., ≤0.1 ppm deviation in the magnetic field).Numerous structural interfaces exist in this region that resultin large magnetic field inhomogeneities. In many cases,these large field inhomogeneities cannot be corrected by theuse of magnetic field shim. Susceptibility artifactfrom theparanasal sinuses and pharyngeal airway often prohibitsanalysis of low-volume tumors located adjacent to thesestructures. Susceptibility artifact from bone prevents spectral analysis of small to moderate-sized tumors that directlyabut the skull base or mandible. Pulsations from the carotidartery result in rhythmic field inhomogeneities and mayprevent spectral analysis of lesions that involve thepoststyloidparapharyngeal space.16

Motion artifact is much more problematic in the upperaerodigestive tract than in the brain or extremities. Normalrespiration usually does not interfere with analysis ofprimary parapharyngeal, nasopharyngeal, tonsillar, or masticatorspace masses. However, masses that arise in the upperairway often partially obstruct the airway and causedyspnea. In addition, patients with upper airway tumorsoften have difficulty handling their secretions. This usually results in an exaggerated swallow and motion artifact. Theresult of these problems is increased motion that mayprevent adequate shimming prior to spectral analysis. Thus,the role of 1H-MRS for evaluating advanced laryngeal orhypopharyngeal carcinomas is very limited, especially if thepatient is tracheostomy dependent.16

Fat contamination of the proton spectra is another majorproblem when performing one-dimensional 1H-MRS in theextracranial head and neck. The fat planes surrounding the extension of tumor on CT or noncontrast T1-weighted MRimaging. However, these fat planes can result in contaminationof proton spectra by a dominant lipid peak that prevents identification of the Cho or Cr peak.16 When evaluatinglesions that abut fat, every attempt should be made to centerthe voxel directly over the lesion. In these cases, it isespecially important to minimize voxel size withoutexcessively increasing acquisition time. However, despiteproper placement of the voxel over a tumor, the intrinsicallyhigh fat content of the tissue may often result in sufficient fatcontamination to obliterate the Cho and Cr resonances andrender 1H-MRS useless.16

Today, 1H-MRS may be routinely performed on patientswith non-obstructing head and neck tumors. However,because voxel size and acquisition time are competingfactors, the study parameters have to be individualized,depending on the patient’s clinical status and the location ofthe lesion in question. Thus, a smaller voxel size with alonger acquisition time could be used in a cooperativepatient with a non-obstructive mass such as a masticatorspace lesion. However, a larger voxel size, which allows areduction in acquisition time, may be a better strategy in apatient with a mass that extends into the airway, withresulting mild to moderate dyspnea.

In the future, the bestMRS results will likely be obtained by using dedicatedspectroscopy coils developed solely for use in the extracranialhead and neck. extension of tumor on CT or noncontrast T1-weighted MRimaging. However, these fat planes can result in contaminationof proton spectra by a dominant lipid peak that prevents identification of the Cho or Cr peak. When evaluatinglesions that abut fat, every attempt should be made to centerthe voxel directly over the lesion. In these cases, it isespecially important to minimize voxel size withoutexcessively increasing acquisition time. However, despiteproper placement of the voxel over a tumor, the intrinsicallyhigh fat content of the tissue may often result in sufficient fatcontamination to obliterate the Cho and Cr resonances andrender 1H-MRS useless.16

Today, 1H-MRS may be routinely performed on patientswith non-obstructing head and neck tumors. However,because voxel size and acquisition time are competingfactors, the study parameters have to be individualized,depending on the patient’s clinical status and the location ofthe lesion in question. Thus, a smaller voxel size with alonger acquisition time could be used in a cooperativepatient with a non-obstructive mass such as a masticatorspace lesion. However, a larger voxel size, which allows areduction in acquisition time, may be a better strategy in apatient with a mass that extends into the airway, withresulting mild to moderate dyspnea. In the future, the bestMRS results will likely be obtained by using dedicatedspectroscopy coils developed solely for use in the extracranialhead and neck.



An area of active investigation is the role of newintravenous contrast agents for evaluation of metastaticdisease. Over the past 15 years, superparamagnetic ironoxide compounds have been investigated primarily to detectthe presence of liver metastases. However, there has beengrowing interest in investigating the ability of dextrancoatedultrasmallsuperparamagnetic iron oxide (USPIO) todetect metastatic nodal disease.


AMI 227, Combidex TM (Advanced Magnetics, Inc.,Cambridge, MA), is a dextran-coated monocrystallinesuperparamagnetic iron oxide particle with an average particle size of 18 nm.19,40,41,42 Two major factors determine the physiologic distribution of these particles: their size and their surface properties. Iron particles with a diameter between 80 and 100 nm are trapped in liver and spleen, but not in lymph nodes and bone marrow.43,44

However, ultrasmall iron particles, which are less than 20 nm in diameter, can traverse the capillary walls, enter the interstitial spaces, and eventually appear in the lymphatics.(Fig 2) These particles are then phagocytized by macrophages within lymph nodes.45,46 Thus, in a normally functioning lymph node, USPIO are trapped due to the normal phagocytic ability of the node. If tumor replaces portions of the node, such phagocytosis of the USPIO will not occur. Superparamagnetic iron oxide particles cause local MR field effects.

Therefore, because of the functional nodal alterations caused by tumor, USPIO allows differentiation of a metastatic node from a normal or reactive node. This assessment is based on alterations of nodal signal intensities rather than on nodal morphology (size, shape, or architecture). Compared to its noncontrast signal intensities, a normally functioning node after the administration of USPIO demonstrates markedly reduced T1-weighted and T2-weighted signal intensities due to both T2 relaxation and magnetic susceptibility effects that result from the uptake of the iron particles. By contrast, a metastatic node does not have a signal intensity loss on postcontrast images because the node’s macrophages have been replaced and so fail to capture the iron particles. Because USPIO is administered intravenously, this technique has the potential to evaluate all of the lymph nodes throughout the body. This is asignificant advantage over traditional lymphography and sentinel node imaging, since these techniques only image lymph nodes in the primary drainage pathway of the injection site.
Fig2: Pre & Post USPIO images of lymph nodes in Benign & Metastatic Disease


The current recommendation is to perform MR imagingfrom the skull base to the thoracic inlet before and after theintravenous administration of USPIO. Currently, the mostcommonly used MR parameters used to evaluate thedistribution of USPIO are a (TR) of 500 msec, a (TE) of 20msec, and a flip angle of 20°. The optimum time forpostcontrast imaging appears to be between 24 and 36 hoursafter intravenous USPIO administration.47

Clinical Applications

Detection of Nodal Metastases

Identification and staging of nodal metastases is one ofthe most important prognostic procedures for patients withSCC of the upper aerodigestive tract, and both treatment and outcome are directly affected by information regarding thepresence of nodal metastases. Several studies have been performed to evaluate thediagnostic accuracy of USPIO in detecting metastatic lymphnodes. The initial multicenter phase II trial of USPIOdemonstrated a sensitivity of 84% and a specificity of 95%in detecting metastatic lymph nodes.42

The recentlycompleted phase III multicenter European and U.S. studiesconfirmed the results of the phase II studies. These trialsreported similar diagnostic accuracy in detecting metastaticdisease on postcontrast studies. The sensitivity and specificityof the U.S. and European trials in detecting metastases inindividual lymph nodes was as follows: 96% and 87%(U.S.), 95% and 86% (European).48,49 If a lymph nodeloses virtually all of its signal intensity on the postcontrastMR imaging study, it is very likely that the node containsmetastatic disease. Conversely, if there is no drop in signalintensity, the node is likely reactive. However, a problemarises when only part of the node loses signal intensity. Atpresent, it has been suggested that if half of the node loses signal intensity on the postcontrast study, there is about an80% likelihood that the node contains tumor.48,49,50These results suggest that USPIO MR contrast agents maybecome important adjuncts in evaluating metastatic nodaldisease for SCC of the upper aerodigestive tract.

Primary Site Evaluation

Investigations are now underway to determine the abilityof USPIO to evaluate the primary tumor site. Followingadministration of USPIO, the pituitary gland, cavernoussinuses, nasal turbinates (concha), and pharyngeal mucosaall show increased T1-weighted signal intensity. Thesestructures appear slightly dark on T2-weighted images dueto the T2 shortening effect of iron oxide particles. Earlyresults suggest that the T2-weighted signal intensity of aprimary tumor in the head and neck does not substantiallychange from that of pre-contrast T2-weighted imaging. Sincethe normal pharyngeal mucosa and nasal turbinates loseT2-weighted signal intensity following USPIO administration,it is possible that administration of USPIO mayimprove the conspicuity and margins of SCC of the upperaerodigestive tract.48However, further investigations arenecessary to determine the utility of USPIO for evaluation ofthe primary site.



Monoclonal antibodies (MAbs) directed against tumorspecificand tumor-associated antigens can be used forselective tumor targeting. This imaging technique is calledradioimmunoscintigraphy (RIS). MAbs can be produced tobind to specific targets and can be labeled with radionuclidesthat emit gamma rays. Thus, specific tumors can bevisualized using gamma cameras. Because RIS identifiesbiologic antigenic targets on the tumor cell, this techniquediffers fundamentally from anatomic imaging modalities.

The potential of RIS depends on the tumor antigen, theMAb, the radionuclide, the tumor, and the imagingtechnique.The hybridoma technology introduced in 1975 by Ko¨hlerand Milstein made it possible to develop MAbs directedagainst specific cellular antigens51 The ideal antigenictarget for RIS is one that is expressed on the cell surface byall tumors that are to be studied. Unfortunately, suchtumor-specific antigens have only been found in experimentallyinduced tumors and not in so-called spontaneoustumors. Most identified antigens in human tumors aretumor-associated antigens, which are present on both tumor and normal tissues. Expression of a target antigen in normaltissue is acceptable for RIS when this normal tissue is poorlyaccessible to MAb or when it is localized at a site outside the anatomic region of interest.

The uptake of a MAb in a tumor depends on the antigenrecognized by the MAb, as well as on the molecular size ofthe MAb. An intact MAb is a large immunoglobulinmolecule with a molecular weight of 150 kDa. Such largemolecules have a limited ability to penetrate a tumor and theresidence time of such an intact MAb in blood is long,resulting in low tumor/nontumor uptake ratios. For RIS, theuse of smaller MAb fragments such as F(ab′)2 (mol.wt. ≈ 100 kDa), Fab (mol. wt. ≈ 50 kDa), and Fv (mol.wt. ≈ 25 kDa) can be advantageous. Such smaller fragmentshave better penetration than whole immunoglobulin andtherefore have the potential for more rapid tumor targeting, ahigh tumor/nontumor ratio, and earlier scan times afterinjection. The radionuclides most commonlyused in RIS are 131I (half-life: 8 days), 123I (half-life: 13hours), 111In (half-life: 68 hours), and 99mTc (half-life: 6hours).

Efforts to develop RIS for the detection of primarytumors and lymph node metastases in patients with SCC ofthe head and neck are currently underway. These investigationshave led to the following initial conclusions: 131I and123I labels are not favorable, because the detachment of 131Ifrom MAbs (dehalogenation) will result in activity uptake inthe thyroid, which may disturb imaging of SCC of the headand neck in this region. Similar considerations can be madefor detached 111In, which can accumulate in lymphocytespresent in the lymph nodes and liver.

In addition, the highcost (111In, 123I), patient radiation exposure (111In), limitedavailability (123I), long half-life (131I), and poor imagingqualities (131I) are other drawbacks to these radionuclides. 99mTc appears to have the most favorable characteristics forRIS in the upper aerodigestive tract. These favourable characteristics include minimal radiation delivery to thepatient (short half-life time and low gamma energy),favorable characteristic for gamma camera imaging (high photon abundance), low cost, and wide availability. Recentdevelopments have also simplified the labeling methodsfor 99mTc.


The first images are usually acquired within a fewminutes after injection of the radioimmunoconjugate. Theseearly images serve as a baseline reference for subsequentimages. For most intact MAbs or fragments labeled with99mTc, the time for optimal imaging is 4 to 24 hours afterinjection. The later images are whole-body images andimages of the head and neck region. SPECT images areoften acquired in addition. Axial, coronal, and sagittalsections can be used to separate activity in the tumor fromuptake in normal tissues including the blood vessels.

Early Results

It is only during the past decade that MAbs have becomeavailable with a high specificity for SCC of the head andneck and a restricted reaction pattern on normal tissues.Among the best-characterized MAbs are MAb E48 and U36,which both react with SCC of the head and neck as well aswith normal squamous epithelium. These MAbs wereadministered to 49 patients with SCC of the head and neckwho were at risk for having neck node metastases and whowere scheduled to undergo neck dissection(s).51,52,53 Patientsreceived 1 to 50 mg E48 F(ab′)2 fragment (15patients), E48 intact IgG (24 patients), or U36 intact IgG (10patients) labeled with approximately 740 MBq 99mTc byintravenous injection. All 44 primary tumors in 49 patientswere visualized by RIS. Thirty-two patients underwentunilateral neck dissection, and 17 patients had bilateral neckdissection. A total of 318 nodal levels were histopathologicallyevaluated. Fifty-one of the 66 operated sides containedmetastases in a total of 77 levels.

RIS detected lymph nodemetastases in 42 of the 77 levels (sensitivity, 55%) and in 35of 51 sides (sensitivity, 69%). For CT and MR imaging,these figures were 60% for nodal level and 84% for neckside. Interpretation of RIS was correct in 276 of 316 levels(accuracy, 87%) and in 47 of 65 sides (accuracy, 72%). Theaccuracy of CT and MR imaging was, per level, 86%and 88%, respectively, and per side, 82% and 77%,respectively. The paraffin sections of the metastatic lymphnodes not detected by RIS were reexamined histopathologically.51,52,53 The missed lymph nodes all were less than2 cm in diameter and contained small tumor deposits(micrometastases) or metastases with large proportions ofnecrosis, keratin, or fibrin deposits. Furthermore, theseexaminations showed that the smallest tumor-involvedlymph nodes detected by RIS had diameters of 5 mm and 9 mm in the axial plane, with a tumor burden of 50%.

In their present form, these initial studies suggest that RISis not yet a preferred method for evaluating cervical nodalmetastases. However, the immunohistochemical analysesperformed during the initial RIS studies revealed that MAbsE48 and U36 had targeted all tumor-involved lymph nodes,as well as deposits that had not been visualized by RIS.Thus, MAbs were identifying lymph node metastasesdespite not being visualized on gamma camera imaging.

Future Trends

Investigations are now underway to improve the accuracyof RIS imaging. Ongoing investigations are aimed atreducing the time that MAbs reside in the blood pool. Oneconcept is to use pretargeting strategies that provide rapidclearance from the bloodstream and result in high tumor/non-tumorMAb ratios. Another area of active investigationis the integration of MAbs into PET technology. For thispurpose, MAbs must be labeled with positron-emittingradionuclides, among which 124I and 89Zr are particularlysuitable. Studies in tumor-bearing mice have shown that89Zr-labeled MAbs can easily be visualized with a PETcamera, allowing detection of tumors as small as 50 mg.54On the basis of these preliminary studies, it can beanticipated that antibody-PET may eventually emerge assuperior to conventional antibody-SPECT in sensitivity,resolution, and quantification.


The integration of new metabolic imaging techniques hasled to dramatic changes in the evaluation of a variety of headand neck disorders. Because of the rapid pace of developments in the areas of metabolic and functional imaging, it islikely that currently unsuspected and unknown futureapplications will fall within the purview of radiologists.55 Itis imperative that our specialty continue to pursue newinvestigations that were once believed to be outside therealm of imaging.56

  1. Mukherji SK, Gapany M, Phillips D, et al. Squamous cellcarcinoma of the upper aerodigestive tract: the ability of thallium-201SPECT to detect recurrent tumor. Am J Neuroradiol 1999;20:1215–1220.
  2. Mukherji SK, Gapany M, Neelon B, McCartney W. Evaluation ofthallium-201 single photon emission computed tomography forpredicting early treatment response in patients with squamous cellcarcinoma of the extracranial head and neck treated with non-surgicalorgan preservation therapy: initial results. J Comput Assist Tomogr2000;24:146–151.
  3. Gapany M, Grund FM. Thallium-201 imaging for upper aerodigestivetract cancer. In: Mukherji SK, Castelijns JA, eds.Modern Head and Neck Imaging. Heidelberg: Springer-Verlag, 1998;107–110.
  4. Nagamachi S, Hoshi H, Jinnouchi S, et al. Tl-201 SPECTfor evaluating head and neck cancer. Ann Nucl Med 1996;10:105–111.
  5. Nagamachi S, Jinnouchi S, Flores LG, et al. The use of Tl-201SPECT to predict the response to radiotherapy in patients with headand neck cancer. Nuc Med Commun 1996;17:935–942.
  6. Gregor T, Valdes-Olmes RA, Koops W, et al. Preliminary experiencewith thallous chloride Tl 201-labeled single photon emissioncomputed tomography scanning in head and neck cancer. ArchOtolaryngol Head Neck Surg 1996;122:509–514.
  7. Valdes-Olmos RA, Balm AJM, Hilgers FJM, et al. Tl-201 SPECTin the diagnosis of head and neck cancer. J Nucl Med 1997;38:873–879.
  8. Waxman AD. Thallium-201 in nuclear oncology. In: Freeman LM,ed. Nuclear Medicine. Annual 1991. New York: Raven Press, 1991;193–207.
  9. Gehring PJ, Hammand PB. The interrelationship between thalliumand potassium in animals. J PharmacolExpTher 1967;155:187–201.
  10. Lebowitz E, Greene MW, Greene R, et al. Thallium-201 for medicaluse I. J Nucl Med 1975;16:151–155.

More References Are Available On Request