The British Journal of Radiology, 81 (2008), 291-298

The impact of ^^F-FDG PET/CT on assessment of nasopharyngeal carcinoma at diagnosis

^A D KING, FRCR, ^B B M A , FRACP, ^ Y Y Y A U , FRCR, "^B ZEE, PHD, ^S F LEUNG, FRCR, ""J K T W O N G , FRCR, ^ M K M K A M , FRCR, ' A T A H U J A , FRCR a n d ^A T C C H A N , FRCP

^Department of Diagnostic Radiology & Organ Imaging, ^Department of Clinical Oncology, ^Hong Kong Adventlst Hospital and ^Centre for Clinical Trials, Faculty of Medicine, The Chinese University of Hong Kong, Prince of Wales Hospital, Shatin, New Territories, Hong Kong S.A.R, China

ABSTRACT. The aim of this study was to determine whether the use of whole-body ^^Ffluoro 2-deoxy-D-glucose ('^F-FDG) positron emission tomography (PET)/CT alters staging and management of nasopharyngeal carcinoma (NPC) when compared with current staging practice. 52 patients with Stage III-IV NPC without distant metastases on chest X-ray/CT, abdominal ultrasound or bone scan were recruited for the study. Whole-body F-FDG PET/CT and MRI of the head and neck were performed. The scans were compared for extent of the primary tumour (PT), cervical nodal metastases (CNM) and distant metastases (DM). Any discordance in results was assessed with respect to staging and impact on management. MRI and ^^F-FDG PET/CT scans were discordant in 28 (54%) patients. There was discordance in the extent of PT at 28 sites; in all sites, MRI showed more extensive tumour involving the nasopharynx {n^8), skull base (n = 14), brain (n-4) and orbit (n-2). There was also variation among the extent of CNM in four nodes of the retropharyngeal region, with the nodes being positive on MRI. ^^F-FDG PET /CT did not identify any additional distant metastases but did identify a second primary tumour in the colon. The additional use of ^^F-FDG PET/CT did not "up-stage" the overall stage or change management in any patient. In conclusion, there is discordance between MRI and '^F-FDG PET/CT, and the additional use of '^F-FDG PET/CT for the current assessment of NPC at diagnosis does not appear to be justified in this cohort of patients.

Received 19 December

2006
Revised 16 May 2007 Accepted 23 May 2007 DOI: 10.1259/bjr/73751469
V 2008 The British Institute of Radroiogy

Niisophnryngeal carcinoma (NPC) has a propensity to invade structures around the skull base and spread to nodes in the neck. It is a cancer that is treated primarily by radiotherapy. The clear depiction of the boundary of the primary tumour and vital anatomical structures around the skull base are important for radiotherapy planning, especially when using techniques such as intensitymodulated radiotherapy. MRI is now the imaging modality of choice for depicting this information although, in common with all imaging techniques, the accuracy of MRI cannot be verified because this cancer is not accessible to surgical resection. At present, the standard protocol for imaging NPC in many centres uses MRI to assess the primary tuniciur and the cervical nodes; ul chest X-ray/CT scan, abdominal ultrasound and wholebody bone scan are used to detect distant metastases. Positron emission tomography (PET) using '^F-fluoro2-deoxy-[)-glucose ("*F-FDG)/CT is a widely used technique for evaluating head and neck cancers that are primarily squamous cell carcinomas. The undifferentiated form of NPC, prevalent in southern China, is "*F-nXJ avid, and therefore PET has the potential to be a valuable
Address correspondence to: Dr A D King, Department of Diagnostic Radioiogy and Organ Imaging, Faculty of Medicine, The Chinese University of Hong Kong, Prince of Wales Hospital, Shatin, New Territories, Hong Kong S.A.R, China. E-mail: king 2015 cuhk.edu.hk The British Journal of Radiology, April 2008

tool in the assessment of NPC at diagnosis. Although "^FFDG PET, even when fused with CT, cannot provide the anatomical detail necessary to replace MRI for radiotherapy planning, it does have the potential to reveal more extensive tumour at the primary site, as well as to improve the detection of cervical nodes and replace conventional imaging techniques for the detection of distant metastases. Therefore, the aim of the study was to evaluate the impact of adding '^F-FDG PET/CT to the standard imaging protocol for NPC at diagnosis to determine if this technique alters staging and management.

Methods and materials

The study was performed retrospectively on a group of 52 patients (38 males, 14 females; mean age, 50 years; age range, 31-70 years) with a new diagnosis of undifferentiated NPC (WHO Type IIT) confirmed by histology. All patients had been enrolled into a clinical trial on neoadjuvant chemotherapy for which the patient selection criteria included those with Stage III-IV disease (based on initial CT) with no evidence of distant metastases on a chest X-ray/CT, abdominal ultrasound and bone scan. Institutional consent was obtained for the clinical trial and imaging study. All patients underwent MRI of the head and neck and a whole-body ' " F - F D G PET/CT scan. The two examinations were performed within 1-20 days
291

A D King, B B Ma, Y Y Yau et al

(mean 6 days) of each other. Tumour stage was determined according to the 6th edition of the American Joint Committee on Cancer (AJCC) classification in Table 1 [1]. MR! technique MRI was performed on a 1.5 MR unit (Philips Gyroscan ACS-NT, Best, the Netherlands). In all patients, images included an axial fat-suppressed TT weighted sequence (repetition time/echo time (TR/TE) 2500/100 ms, echo train length of 15, field of view 22 cm, slice thickness 4 mm with no interstice gap, and matrix size 256 x 202), coronal T2 weighted turbo spin-echo (TR 2500 ms, TE 100 ms, echo train length 14, 22 cm field of view, 4 mm slice thickness with no interslice gap and a 256 x 202 matrix), axial Ti weighted spin-echo (TR 500 ms, TE 20 ms, 22 cm field of view, 4 mm slice thickness with no interslice gap, and a 256 X 202 matrix) and contrast-enhanced T] weighted spin-echo images using a 512 x 512 matrix in the axial and coronal planes following a bolus injection of 0.1 mmol kg~^ gadolinium dimeglumine (Schering AG, Berlin, Germany). In addition, axial fat-suppressed turbo spin-(?cho images were obtained after contrast (TR 2500 ms, TE 100 ms, echo train length 15, 22 cm field of view, 4 mm slice thickness with no interslice gap, matrix 256 x 202).

PET/CT technique PET/CT was performed on a Biograph scanner (Siemens Medical Solutions, Knoxville, TN). It consisted of a PET scanner (ECAT ACCEL, CPS (CTI PET Systems) innovation) with lutetium oxyorthosilicate (LSO) crystal detectors. The CT system was a dual-slice helical scanner (Siemens Medical Solutions, Eriangen, Germany) with a bed speed of 8 mm s~' and a reconstructed spatial resolution of 0.32 mm. The details on the scanner

specification, mcxle of operation and principle of attenuation correction have been previously described (2, 3]. In brief, the LSO PET scanner (Siemens Medical Systems) acquired the PET data in a three-dimensional mode, which had an axial resolution of 6.3 mm, an axial field of view of 16.2 cm per bed and a 3.2 cm overlap of each bed position. Patients were either fasted overnight or for at least 6 h to ensure their fasting blood glucose level was <7.5 mmol 1"^ prior to administration of the tracer. 1016 mCi of F-FDG, depending upon the body weight, was given intravenously. An uptake time of 1 h was allowed for '^F-FDG distribution within the body. Contrastenhanced whole-body CT scans were initially obtained from the roof of orbit to the pelvis witli a 5 mm slice collimation and a 3.4 mm slice interval. Tliis was followed by six to seven bed positions of PET scanning of the same region, depending on the height of the patient. Emission data were acquired for 3 min per bed and were later attenuation-corrected with the digital CT data. Image reconstruction employed ordered subsets with expectation maximization algorithm of two iteration and eight subsets. Image analysis was carried out on the Synco multimodality computer platform (Siemens Medical Solufions). A region of interest was drawn at each pathological site of tracer uptake and the standardized uptake values (SUVs) were calculated automatically by the computer using the body weight methtxl: SUV-decay corrected tissue activity (kBq mr')/injected "*F-FDG dose per body weight (kBq g ')- The maximum SUV value (SUVm,i>:) was used instead of average SUV as this was more accurate and consistent.

Data analysis The primary tumour was assessed for the presence of invasion in the nasopharynx, nasal cavity, oropharynx.

Table 1. Nasopharyngeal carcinoma stages according to the 6th edition o f t h e American Joint Committee on Cancer (AJCC) classification [1] Primary tumour TI T2 Tumour confined to the nasopharynx Tumour extends to soft tissues Regional lymph nodes
NO

Distant metastases MO Ml No distant metastases Distant metastases

NI

No regional lymph node metastasis Unilateral metastasis in lymph node(s), s6 cm in greatest dimension, above the supraclavicular fossa

T3

T2a: tumour extends to the oropharynx and/or nasal cavity without parapharyngeal extension' T2b: any tumour with parapharyngeal extension^ Tumour involves bony structures and/or paranasal sinuses

N2

T4

Tumour with intracranial extension N3 and/or involvement of cranial nerves, infratemporal fossa, hypopharynx, orbit or masticator space

Bilateral metastasis in lymph node(s), -6 cm in greatest dimension, above the supraclavicular fossa Metastasis in lymph node(s) > 6cm and/or to supraclavicular fossa N3a: >6 cm in dimension N3b: extension to the supraclavicular fossa

'Parapharyngeal extension denotes posterolateral infiltration of tumour beyond the pharyngobasilar fascia.

292

The British Journal of Radiology, April 2008

'f-FDG PET/CT and assessment of nasopharyngeal carcinoma

parapharyngeal region, skull base, paranasal sinus, brain, orbit, infratemporal fossa and hypopharynx. Where tumour invasion was present, the extent was assessed. The neck was assessed for the presence of any node from the retropharyngeal region down to the supraclavicular fossa. Nodes that were identified were then assessed for malignancy. By MRI, nodes were considered to be metastatic if (a) they had a shortest axial diameter of ^11 mm in the jugulodigastric region, 5*5 nim in the retropharyngeal region and 5^10 mm in all other regions of the neck; (b) there was a group of three or more nodes Ihat were borderline in size; or (c) there were any nodes with necrosis or extracapsular spread irrespective of size. By '^F-FDG PET/CT, nodes with an SUVn,,, of >2.5 were considered malignant. This value corresponded with the visual assessment used by Ng et al [4] in which nodes with mild accumulation (Score 1) or equivocal abnormal increased accumulation (Score 2) were considered negative, and those with probable abnormal increased uptake (Score 3) or definite abnormal accumulation (Score 4) were considered positive for nodal metastases. Distant metastases by "^F-FEXi PET/CT were judged to be malignant on the same basis as that used for nodes. Distant lesions identified by "*F-FDG PET/CT underwent biopsy or imaging follow-up to confirm the diagnosis. Interpretation of ' F-FDG PET/CT was performed by a radiologist with over 10 years' experience, including 5 years' experience in head and neck '*^F-F1DG PET/CT. Interpretation of the MRI images was performed by a radiologist with over 10 years' experience in MR imaging of the head and neck. The scans were assessed independently. In cases where a discordant result was obtained, the scans were reassessed side by side and the result reached by consensus.

nodes (Table 3). Taking MRI as the reference standard, ^**F-FE)G PET/CT underestimated the N stage in one patient (Nl to NO).

Distant metastases MRI and '^F-FDG PET/CT scans were discordant in two sites in two patients. For the discordant findings, '^F-FE)G PET/CT showed abnormal uptake at two sites: one owing to pelvic inflammation and the other to a second primary (Table 3), The stage of the distant metastases on MRI combined with conventional imaging was MO (n=51) and Ml (fi= I); the latter case was found to have a cervical soft-tissue deposit identified by both MRI and '**F-FDG PET/CT. '"^F-FDG PET/CT d'id not change the M stage but did identify a second unrelated malignancy.

Overall results and stage MRI and ^^F-FDG PET/CT scans were concordant in 24 (46%) patients and discordant in 28 (547..) patients. The overall stage on MRI was Stage I (r! = Q), Stage II (n-4), Stage III (fi-30) and Stage IV A, B and C (ii-7, 10 and 1, respectively). The additional use of "^F-FDG PET/ CT did not overestimate the overall stage and did not change the form of treatment or alter the radiation field in any patient. The use of '"F-FDG PET/CT underestimated the overall stage in four patients (Stage IVa to III (=3) and Stage ill to II (?i = l)) as a result of the discrepancy in T staging.

Discussion
Results Primary tumour The T stage by MRI was TI (tt-14), T2a (n-2), T2b (n=2), T3 (n-26) and T4 (fi-8). MRI and "*F-FDG PET/ CT scans were discordant in 34 sites in 25 patients. For fhe discordant findings, MRI showed hjmour invasion at 28 sites invoh ing the nasopharynx (TI), skull base (T3), brain (T4) and orbit (T4) (Table 2); "*F-FDG PET/CT showed tumour invasion at six sites, all involving the parapharyngeal region (T2b) (Table 3). Taking MRI as the reference standard, '^F-FDG PET/CT underestimated the T stage in nine patients (from T4 to T3 (n=3); and T3 to TI (=3), T2a (n=2), T2b (K = 1)) and overestimated the T stage in four patients (TI to T2b Studies suggest that ''^F-FDG PET/CT may be more accurate than conventional imaging in the assessment of early treatment response and detection of residual and recurrent NPC [5-10]. The role of' V F D G PET/CT in the staging and planning treatment of NPC at diagnosis is less clear, although it may have a benefit in detecting loco-regional nodes and distant metastases [4, 5, 10-15]. The aim of this study was to evaluate the impact of ^^FFDG PET/CT on the staging and management of the (i) primary tumour, (ii) cervical nodal metastases and (iii) distant metastases when compared with current staging practice,

Primary tumour Previous '^F-FDG PET/CT evaluations of the primary tumour in NPC have focused on the detection of the presence of tumour, and to our knowledge there has been no study comparing the extent of the primary tumour and its impact on staging and treatment planning. In this study '^F-FDG PET/CT did not identify any additional subsites of tumour invasion that were not identified by MRI. To the contrary, the volume of disease in the nasopharynx, skull base, brain and orbit was less on '^F-FDG PET/CT than on MRI. Within the nasopharynx, the primary tumour was detected in 100% of 293

Cervical nodal metastases The N stage of the tumour by MR! was NO (=6), Nl (0 = 13) N2 (n = 23) and N3 {n = lO). MRI and ^^F-FDG PET/CT scans were discordant in four nodes in four patients. The four discordant nodes were in the retropharyngeal region and were positive by MRI (Table 2). '^F-FDG PET/CT did not identify any additional positive
TTie British Journal of Radiology, April 2008

A D King, B B Ma, YY Yau et al Table 2. Sites of discordance in patients with positive MRI and negative ^^F-FDG PET/CT
Site No of sites/no of patients 8/8 Details

Primary tumour

Nasopharynx Skull base Brain

14/14
4/4

Cervical nodal metastases Distant sites

Orbit Retropharyngeal

111 4/4 0

Unilateral tumour by ^^F-FDG PET and bilateral tumour by MRI Invasion was not identified (6) or was less extensive (8) by ^^F-FDG PET/CT Invasion into the cavernous sinus was not identified by '^FFDG PET/CT Invasion into the orbit was not identified by ^^F-FDG PET/CT The node was obscured by the primary tumour (3) or not identified {1 small necrotic node) by '^F-FDG PET/CT

'F-FDG PET, ^^F-fluoro-2-deoxy-D-glucose positron emission tomography.

patients, aithough the tumour component in the nasopharyngeal wall was less extensive in 15% of patients. This finding would not impact upon the stage or treatment planning but suggests that ^"F-FDG PET/CT has the potential to miss small volume tumours. The main site of discordance between MRI and ^^EFE)G PET/CT was at the bony skull base. Whereas "^FFEKJ PET/CT did not identify any additional areas of tumour invasion at the skull base, it did underestimate the presence or extent of tumour invasion in 27% of patients. It can be argued that, for limited degrees of invasion, direct comparison between "^E-FDG PET/CT and MRI is difficult, and the apparent discrepancy may be technical or masked by the "shine-through" effect of '^F-FE)G PET/CT around the tumour. However, even large areas of MRI abnormality in the clivus were not detected by ^*'F-FDG PET/CT (Figure 1). In these cases, the discordance between the findings would have a significant impact on planning the extent of the radiation field. At present, the standard practice is to err on the side of caution and encompass all areas of abnormality in the field, at least the clinicai target volume (CTV), which includes all regiorrs of potential microscopic disease. Therefore, the radiation field would be larger when using MRI to plan treatment than with ^^F-FDG PET/CT. Although this study highlights the discordance between the two techniques, and the effect that this has on current treatment regimes, it cannot address the question of which technique is more accurate. Unfortunately, pathological/radiological correlation could not be obtained and evidence from the literature is scanty. MRI is known to be a highly sensitive technique for identifying tumour

invasion into the mandible but it has problems with specificity, leading to false-positive results in bone 116|. In addition, a study by Daisne et al [17], using pathological specimens as the reference standard, has shown that the volume of primary cancers is smaller on "*F-FDG PET than on MRI or CT and, of these, '^'F-FDC PET provides the most accurate assessment of the gross tumour volume. Therefore, it is possible that MRI overestimated tumour extent in the skull base because the abnormalities were caused by other processes such as bone marrow oedema and inflammation rather than tumour infiltration. If '*^F-FDG is proven to be more accurate at assessing the extent of bone marrow invasion, it opens up the possibility that '^F-FDC; TET/CT could influence radiotherapy plamiing in terms of defining the gross tumour volume (GTV), which receives a higher radiation dose than the CTV. The final region in which '*'F-FDG PET/CT undermapped the extent of tumour was within the cavernous sinus and adjacent posterior aspect of the orbits. In these cases, the volume of tumour was small but its activity was enough to produce a "spillover" effect, obscuring tho extension. Once again, F-FDG PET/CT did not have any impact on the results of current imaging, although if used alone would have downstaged NPC and underestimated the extent of disease in a region where planning the radiotherapy field is crucial to vision. Despite initial concerns about the relatively high background activity of the normal temporal lobes obscuring the intracranial tumour, all cases of tumour extension into the fliKir of the middle cranial fossa were identified by '^F-FDG PET/ CT in this study, although the numbers were very small.

Table 3. Sites of discordance in patients with positive ^^F-FDG PET/CT and negative MRI
Site No of sites/no of patients 6/6 Details

Primary tumour

Parapharyngeal spread

Extent of tumour was the same on both scans but '^F-FDG PET was unable to distinguish bulging from true invasion into the parapharynx, resulting in tumour being overstaged in four patients (Tl to T2b) Biopsy confirmed an early carcinoma in a sigmoid polyp No abnormality demonstrated on follow-up ultrasound 1 year after '^F-FDG PET

Cervical nodal metastases Distant sites

0/0
Colon Ovary

1/1

PET, ^^F-f!uoro-2-deoxy-D-glucose position emission tomography. 294 The British Journai of Radiology, April 2008

^F-FDG PET/CT and assessment of nasopharyngeal carcinoma

Figure 1. A 49-year-old male with nasopharyngeal carcinoma (NPC) invading the skull base from the basisphenoid and right peterous apex (arrows), shown by T, weighted MR images (a) pre-contrast and (b) post-contrast but not by (c,d) '^F FDG PET/CT, down to the basiocciput (arrows), as shown on weighted MR images (e) pre-contrast and (f) post-contrast but not by (g,h) FDG PET/CT.

The British Journal of Radiology, April 2008

295

A D King, B B Ma, YY Yau et a! One region in which ' " F - F D G PET/CT appeared initially to show more extensive disease than MRI was in the parapharyngea] region. However, upon direct comparison, there was no actual difference in the extent of disease, and the discordance was caused by the superior ability of MRI over "^F-FDG PET/CT 118] to depict small anatomical structures and hence distinguish a tumour that is bulging into the parapharygeal region from one that has caused direct invasion. These results reflect the controversy regarding definition of parapharyngeal invasion (T2b) and highlight the discrepancy that may occur when using different definitions in different techniques. the current study concurs with the findings of Chang et al [13]. The use of 'T-FDG PET/CT caused little discrepancy for NPC neck node staging, and the additional use of ' " F - F D G PET/CT would not have altered the stage or management in any patient.

Distant metastases PET/CT has the obvious advantage of being able to assess the whole body for metastases in one examination and, at the same time, assess the primary tiimour and cervical nodes. One limitation of this study was that patients with distant metastases on conventional imaging were excluded from ' " F - F D G PET/CT, and therefore the number of ' " F - F D G PET/CT demonstrable distant metastases would be expected to be low in this study. However, recent '"F-FDG PET studies have found an unexpectedly high rate of distant metastases, especially with N3 disease [11, 13, 14]. ' " F - F D G PET has been shown to be more sensitive than conventional imaging, and approximately 12% of patients most of whom were imaged at initial staging had distant metastasGS that were missed on conventional imaging studies [11, 13]. Unexpectedly, in this study, '**F-FDG PET/CT did not reveal any additional distant metastases despite the high number of patients with N3 disease; therefore, there was no change in the M stage or overall stage of NPC. '**F-FDG PET/CT did reveal one falsepositive finding in a patient with ovarian inflammation and, more importantly, identified a second early primary colon cancer in a sigmoid polyp. One major limitation of this study, which has already been stated in the discussion above, was the lack of pathological proof of the extent of the primary and nodal NPC. However, it is stressed that, in clinical practice, planning of the radiation field is dependent upon imaging, and therefore the results in this study reflect the actual situation encountered when managing this cancer. In summary, there is discordance between '^F-FFXi; PET/CT and MRI. However, '^F-FDG PET/CT did not identify additional subsites of invasion by the primary tumour, and the volume of disease in the nasopharynx, skull base, brain and orbit was less when compared with MRI. '^F-FDG PET/CT did not identify any additional metastatic nodes in the neck or reveal any additional distant metastases but did identify a second primary tumour in one patient. Therefore, at the current time, the additional use of '"^F-FDG PET/CT does not appear to change either the staging and treatment of NPC in the head and neck, although the role of ' " F - F D G PET/CT for the identification of distant metastases in patients with advanced loco-regional disease requires further evaluation. However, several future developments may justify the use of ^^F-FDG PET/CT in the head and neck at diagnosis. The first is the potential for '^F-FDG PET/CT to distinguish reliably between oedema and tumour invasion in the bone marrow of the skull base, which could reduce the size of the GTV for radiotherapy. The second is the use of pre-treatment ' " F - F D G PET/CT to aid the interpretation of any subsequent scans used to monitor early treatment response and detect residual cancer after treatment. The British Journal of Radiology, April 2008

Cervical nodal metastases The detection of cervical nodal metastases from head and neck squamous cell carcinomas by '^F-FDG PFT has been widely reported in the literature. When compared with conventional imaging techniques, such as CT or MRI, "*F-FDG PET often has a higher accuracy [19-22]. However, in common with all imaging modalities, ^^FFDG PET produces both false-positive and false-negative results [23-27], and the SUV of benign and malignant nodes may overlap [15, 21, 23]. The accuracy of ^'^F-FDG PET in detecting nodal metastases from NPC at diagnosis has been evaluateci previously [4, 12, 13, 15]: three of these studies showed that '^F-FDG PET improved detection when compared with MRI [4, 12] and CT [15], whereas the fourth concluded that there was little discrepancy between '^F-FDG PFT and MRI [13]. In this study, there was discordance in only four nodes, all of which were in the retropharyngeal region and were identified by MRI but not ^^F-FDG PET/CT. However, in three of these nodes, '*^F-FDG PET/CT was abnormal at this site and the discrepancy arose because the anatomical detail on '^F-FDG PET/CT could not distinguish the node from the adjacent primary tumour (Figure 2a-c). Unexpectedly, '^F-FDG PET/CT did not identify any additional malignant ncxies. One of the limitations of this retrospective study was that biopsy confirmation of the presence/absence of metastatic nodes could not be obtained, and therefore the results were dependent upon imaging criteria. Reducing the SUV,,,ax in this study would have increased the sensitivity of ^'*F-FDG PET/CT and hence the number of positive nodes detected by this technique, but this could have been at the expense of reducing specificity, leading to an increase in the number of false-positive results. Using an SUV^a^^ of >2.5 corresponded to the visual categories for negative and positi\'e nodal metastases used by Ng et al [4], a study in which the number of false-positive and false-negative NPC nodes was very low. For MRI, the criteria chosen were those widely used in diagnostic imaging of nodal metastases, although the criteria for size of retropharyngeal nodes was based on an MRI study of a normal population. To the authors' knowledge, no radiological/pathological study has ever been performed because of the inaccessibility of retropharyngeal nodes to biopsy. Despite the limitation of lack of biopsy proof in this study, the results do reflect clinical practice in which fine needle aspiration of NPC neck nodes is performed infrequently for staging purposes. In summary, using the criteria stated above. 296

PET/CT and assessment of nasopharyngeal carcinoma

Figure 2. A 55-year-old male with a metastatic right retropharyngeai node (arrow), which can be distinguished from the primary tumour in the nasopharynx (arrow heads) by (a) T2 weighted MRI and (b) T] weighted MRI post-contrast, but not by (c) '^F-FDG PET/a (circled).

References
1. Greene FL. AJCC Cancer staging; mnnual (6th edn). New York: Springer; 2(X)2. 2. Antoch G, Saoudi N, Kuehl H, Dahmen G, Mueller SP, Beyer T, et al. Accuracy of whole-btxiy dual-modality The British Journal of Radiology, April 2008

fluorine-18-2-fluc)ro-2-deoxy-D-glucose positron emission tomography and computed tomography (FDG-PET/CT) for tumor staging in solid tumors: comparison with CT and PET. J Clin Oncol 2004;22:4357-68. 3. Yau YY, Chan WS, Tam YM, Vernon P, Wong S, Coel M, et al. Application of intravenous contrast in PET/CT: does it

297

A D King, B B Ma, YY Yau et al reaiiy introduce significant attenuation correction error? J Nucl Med 2005;46:283^91. 4. Ng SH, Chang JTC, Chan SC, Ko SF, Wang HM, Liao CT, et al. Nodal metastases of nasopharyngeal carcinoma: patterns of disease on MRI and FDG PET Eur J Nucl Med Mol Imaging 2004;31:1073-108. 5. Yen RF, Chen THH, Ting LL, Tzen KY, Pan MH, Hong RL. Early restaging whole-body '"F-EDG PET during induction chemotherapy predicts clinical outcome in patients with kicoregionally advanced nasopharyngeal carcinoma. Eur J Nucl Med Mol Imaging 2005;32:n52-9. 6. Mitsuhashi N, Hayakawa K, Hasegawa M, Euruta M, Katano S, Sakurai H, et al. Clinical EDG-PET in diagnostic and evaluation of radiation response of patients with nasopharyngeal tumor. Anticancer Research 1998;18:2827-32. 7. Kao CH, Shiau YC, Shen YY, Yen RF. Detection of recurrent or persistent nasopharyngeal carcinomas after radiotherapy with technetium-99m methoxyisobutylisonitrile single photon emission computed tomography and computed tomography: comparison with 18-fluoro-2-deoxyglucose positron emission tomography. Cancer 2002;94:1981-^. 8. Peng NJ, Yen SH, Liu WS, Tsay DG, Liu RS. Evaluation of the effect of radiation therapy to nasopharyngeal carcinoma by positron emission tomography with 2"[F-18]fluoro-2deoxy-l>glucose. Clinical Positron Imaging 2000;3:51-6. 9. Tsai MH, Shiau YC, Kao CH, Shen YY, Lin CC, Lee CC. Detection of recurrent nasopharyngeal carcinomas with positron emission tomography using 18-fluoro-2-deoxyglucose in patients with indeterminate magnetic resonance imaging findings after radiotherapy. J Cancer Res Clin Oncol 2002;128:279-82. 10. Yen RF, Hong RL, Tzen KY, Pan MH, Chen THH. Wholebody F-FDG PET in recurrent or metastatic nasopharyngeal carcinoma. J Nuci Med 2005;46:770-4. 11. Yen TC, Chang JTC, Ng SH, Chang YC, Chan SC, Lin KJ, et al. The value of '^E-FDG PET in the detection of stage MO carcinoma of the nasopharynx. | Nucl Med 2005;46:405-10. 12. Yen TC, Chang YC, Chan SC, Chang JT, Hsu CH, Lin KJ, et al. Are dual-phase "*F-FDG PET scans necessary in nasopharyngeal carcinoma to assess the primary tumour and ioco-regional nodes? Eur J Nucl Med Mol Imaging 2005;32:541-8. 13. Chang JTC, Chan SC, Yen TC, Liao CT, Lin CY, Chen IH, et al. Nasopharyngeal carcinoma staging by (18)F-fluorodeoxyglucose positron emission tomography. Int J Radiation Oncology Biol Phys 2005;62:501-7. 14. Liu FY, Chang JT,' Wang HM, Liao CT, Kang CJ, Ng SH, et al. (^'^F) Fluorodeoxyglucose positron emission tomography is more sensitive than skeletal scintigraphy for detecting bone metastasis in endemic nasopharyngeal carcinoma at iniHal staging. J Clin Oncol 2006;24:599-604. 15. Kao CH, Hsieh JF, Tsai SC, Ho YJ, Yen RF, ChangLai SP, et al. Comparison of 18-fluoro-2-deoxyglucose positron emission tomography and computed tomography in detection of cervical lymph node melastases of nasopharyngeal carcinoma. Ann Otol Rhinol Laryngol 2000;109:113a-4. 16. van den Brekel MWM, Runne RW, Smeele LE, Tiwari RM, Snow GB, Castelijns JA. Assessment of tumour invasion into the mandible: the value of different imaging techniques. Eur. Radiol ]998;8:1552-7. 17. Daisnc JF, Dupre^ T, Weynand B, Lonneux M, Hamoir M, Reychler H, et al. Tumor volume in pharyngolaryngeal squamous cell carcinoma: comparison at CT, MR imaging, and FDG PET and validation with surgical specimen. Radioiogy 2004;233:93-100. 18. King AD, Teo P, Lam WWM, Leung SF, Metrewcli C. Paranasopharyngeal space involvement in nasopharyngeal cancer: detection by CT and MRI. J Clin Oncol 2000;12:397-402. 19. Myers LL, Wax MK, Nabi H, Simpson GT, Lamonica D. Positron emission tomography in the evaluation of the NO neck. Laryngoscope 1998;108:232-6. 20. Bailet JW, Abemayor E, Jabour BA, Hawkins RA, Ho C, Ward PH. Positron emission tomography: a new, precise imaging modality for detection of primary head and neck tumors and assessment of cervical adenopathy. Laryngoscope 1992;l()2:281-8. 21. Adams S, Baum RP, Stiickensen T, Bitter K, Hor G. Prospective comparison of '"p-FDG PET with conventional imaging modalities (CT, MRI, US) in lymph node staging of head and neck cancer. Eur J Nuci Med 19y8;25:1255-6t). 22. Szakall S, Esik O, Bajzik G, Repa I, Dabasi G, Sinkovics I, et al. F-FDG PET detection of lymph node metastases in medullary thyroid carcinoma. J Nucl Med 20(12;43:66-7L 23. Braams JW, Pruim J, Freling NJ, Nikkels PC, Roodenburg JL, Boerling G, et al. Detection of lymph node metastases of squanious-cell cancer of the head and neck with FDG-PET and MRI. J Nucl Med 1995;36:211-6. 24. Stoeckli SJ, Steinert H, Pfaltz M, Schmid S. Is there a role for positron emission tomography with 18F-fluorodeoxyglucose in the initial staging of nodal negative oral and oropharyngeal squamous cell carcinoma. Head Neck 2002;24:345-9. 25. Paulus P, Sambon A, Vivegnis D, Hustinx R, Moreau P, Collignon J, et al. "*EDG-PET for the assessment of primary head and neck tumors: clinical, computed tomography, and histopathological correlation in 38 patients. Laryngoscope 1998;108:1578-83. 26. Stokkel MPM, ten Broek FW, van Rijk PP. Preopcrative assessment of cervical lymph nodes in head and neck cancer with fluorine-18 fluorodeoxyglucose using a dualhead coincidence camera: a pilot study. Eur J Nucl Med 1999;26:499-503. 27. Hannah A, Scott AM, Tochon-Danguy H, Chan JG, Akhurst T, Berlangieri S, et al. Evaluation of F-fluorodeoxyglucose positron emission tomography and computed tomography with histopathoiogic correlation in the initial staging of head and neck cancer. Annals of Surgery 2002;236:208-617.

298

The British Journal of Radiology, April 2008