Original article

Prospective comparison of 18F-FDG PET

with conventional imaging modalities (CT, MRI, US)
in lymph node staging of head and neck cancer
Stefan Adams1, Richard P Baum1, Tankred Stuckensen2, Klaus Bitter2, Gustav Hr1
1
2

Department of Nuclear Medicine, Johann Wolfgang Goethe University Medical Center, Frankfurt/Main, Germany
Department of Oral and Maxillofacial Surgery, Johann Wolfgang Goethe University Medical Center, Frankfurt/Main, Germany

&misc:Received 17 February and in revised form 12 June 1998

&p.1:Abstract. The aims of this study were to investigate the

detection of cervical lymph node metastases of head and
neck cancer by positron emission tomographic (PET)
imaging with fluorine-18 fluorodeoxyglucose (FDG)
and to perform a prospective comparison with computed
tomography (CT), magnetic resonance imaging (MRI),
sonographic and histopathological findings. Sixty patients with histologically proven squamous cell carcinoma were studied by PET imaging before surgery. Preoperative endoscopy (including biopsy), CT, MRI and sonography of the cervical region were performed in all
patients within 2 weeks preceding 18F-FDG whole-body
PET. FDG PET images were analysed visually and
quantitatively for objective assessment of regional tracer
uptake. Histopathology of the resected neck specimens
revealed a total of 1284 lymph nodes, 117 of which
showed metastatic involvement. Based on histopathological findings, FDG PET correctly identified lymph node
metastases with a sensitivity of 90% and a specificity of
94% (P<106). CT and MRI visualized histologically
proven lymph node metastases with a sensitivity of 82%
(specificity 85%) and 80% (specificity 79%), respectively (P<106). Sonography revealed a sensitivity of 72%
(P<106). The comparison of 18F-FDG PET with conventional imaging modalities demonstrated statistically
significant correlations (PET vs CT, P = 0.017; PET vs
MRI, P = 0.012; PET vs sonography, P = 0.0001).
Quantitative analysis of FDG uptake in lymph node metastases using body weight-based standardized uptake
values (SUVBW) showed no significant correlation between FDG uptake (3.72.0) and histological grading of
tumour-involved lymph nodes (P = 0.9). Interestingly,
benign lymph nodes had increased FDG uptake as a result of inflammatory reactions (SUVBW-range: 215.8).
This prospective, histopathologically controlled study
confirms FDG PET as the procedure with the highest
sensitivity and specificity for detecting lymph node meCorrespondence to: S. Adams, Department of Nuclear Medicine,
Johann Wolfgang Goethe University Medical Center, TheodorStern-Kai 7, D-60590 Frankfurt/Main, Germany&/fn-block:

tastases of head and neck cancer and has become a routine method in our University Medical Center. Furthermore, the optimal diagnostic modality may be a fusion
image showing the increased metabolism of the tumour
and the anatomical localization.
&kwd:Key words: Head and neck cancer Fluorine-18 fluorodeoxyglucose Conventional imaging modalities
Squamous cell carcinoma
Eur J Nucl Med (1998) 25:12551260

Introduction
Head and neck carcinomas constitute approximately 5%
of all malignancies worldwide and their frequency is increasing [1]. Squamous cell carcinomas represent the
vast majority of all malignant tumours of the head and
neck and originate in the superior alimentary and respiratory tracts [2]. About 40% of the newly discovered cases
are early-stage lesions while the remaining 60% of patients have advanced stage disease. The early stages of
head and neck tumours are usually treated by surgery or
radiotherapy alone, and advanced stages by surgery and
concomitant chemoradiotherapy [35]. Lymph node involvement is the most important prognostic factor affecting survival of patients with head and neck cancer [6, 7].
The average 5-year survival is >50% in patients without,
but only 30% in patients with cervical lymph node metastases [8]. Furthermore DNA ploidy and high proliferating activity are important parameters in the assessment
of squamous cell carcinoma, being indicative of a worse
prognosis and a high metastatic risk [9, 10]. The effectiveness of surgical treatment depends on the complete
excision of all tumour tissue, and accurate preoperative
tumour-node-metastasis (TNM) staging is therefore mandatory. Initial diagnosis and staging of head and neck
carcinomas is based on physical and endoscopic examination [2]. Conventional imaging modalities [computed
European Journal of Nuclear Medicine
Vol. 25, No. 9, September 1998 Springer-Verlag 1998

1256

tomography (CT), magnetic resonance imaging (MRI),

sonography] have been applied for the localization of
primary head and neck tumours, regional lymph node
metastases and their relationship to adjoining anatomical
structures [11]. Discrimination between reactive enlargement of lymph nodes and tumour-infiltrated nodes on the
basis of morphological criteria may be problematic [12].
Fluorine-18 fluorodeoxyglucose (18F-FDG) is a
marker of tumour viability, based upon the increased
glycolysis that is associated with malignancy as compared with most normal tissues. It has also been suggested that tumours with increased FDG uptake appear more
aggressive and are associated with a less favourable
prognosis [13]. Head and neck carcinomas have high
glycolytic activity and increased FDG uptake [14, 15].
The aims of this study were to evaluate the detection
of regional lymph node metastases of head and neck
cancer by 18F-FDG PET and to perform a prospective
comparison with CT, MRI, sonographic and histopathological findings.
Materials and methods
Patients. &p.2:The study group consisted of 60 patients (16 female and
44 males; mean age 58.310 years, range 3876 years) with histologically proven squamous cell carcinoma of the head and neck
region, all scheduled for surgery. Physical examination was performed by the head and neck surgeon. Preoperative endoscopy
(including biopsy) was performed in all patients within 2 weeks
preceding 18F-FDG whole-body PET. Informed consent was obtained from all patients.
Histopathological examination. &p.2:All resected tissues were exactly
localized and documented at each level to allow correlation between histopathological findings and preoperative imaging results.
Classification of the primary tumour and regional lymph node metastases was based on the TNM system of the International Union
Against Cancer (UICC 1992). The primary tumours were differentiated into four groups (G1 well differentiatedG4 anaplastic), as
described by Hermanek and Sobin [16].
PET technique. &p.2:PET studies were acquired on an ECAT Exact 47
whole-body tomograph (Siemens-CTI, Knoxville, Tenn., USA)
with a transaxial field of view of 16.2 cm (slice thickness 3.4 mm;
spatial resolution 4 mm). Prior to the 18F-FDG PET, patients had
been fasting for at least 12 h. Patients with known diabetes mellitus were excluded from the study, so normal glucose plasma levels
(<100 mg%) were confirmed in all patients. Sixty minutes after
intravenous administration of 370 MBq 18F-FDG, PET studies
were performed using a whole-body technique (three to five bed
positions; acquisition time per position: 15 min). In addition, static regional scans of the head and neck region with attenuation correction were acquired (20 min for emission; 10 min for transmission). The original transverse images were three-dimensionally reconstructed by filtered back-projection. Regional FDG uptake was
expressed as the standardized uptake value (SUV) and calculated
for all primary tumours and nodal lesions. The body weight-based
standard uptake value (SUVBW) was defined as follows:
Mean PET counts ( Ci/ml Calibration factor
.
Injected dose (MBq)/Body weight (g)

Only hypermetabolic lesions with strong focal uptake (SUV>2.0)

were considered malignant. However, the final decision to classify
a focus as a metastasis was based on the visual evaluation, which
also took into account the history of the patient (e.g. inflammation
in the head and neck region). Results were coded according to the
TNM classification by two experienced nuclear medicine physicians (in independent, blinded evaluations) on the screen display
of a work station (three-dimensional black-and-white as well as
colour images) utilizing the original data set. Scans were documented as color printouts and on X-ray films (laser technique).
Conventional imaging modalities. &p.2:High-resolution ultrasound
studies (7.5 MHz, linear array, Volu-Son 530 D, Kretz, Marl, Germany) of the regional lymph nodes were performed in all patients
within 2 weeks preceding 18F-FDG whole-body PET.
CT scans of the cervical region were obtained in all patients
within 1 week preceding the PET examination with a conventional
CT scanner (Somatom Plus, Siemens, Erlangen, Germany). Slice
thickness was 45 mm (continuously, without a gap). Contrast
material enhancement was achieved by intravenous administration
of 100 ml of non-ionic contrast material (Ultravist 300, Schering,
Berlin, Germany) with a power injector rate of 0.5 ml/sec.
Concerning MRI studies (Siemens Magnetom Vision, Erlangen, Germany), we first obtained non-enhanced transversal slices
with an inversion recovery (T2-weighted slices) with a slice thickness of 6.0 mm (gap 0.6 mm). In addition, coronal slices (T1weighted) were performed with a slice thickness of 6.0 mm and
with a gap of 1.5 mm. All patients had transversal T1-weighted
slices before and after intravenous administration of contrast medium [0.1 mmol of gadolinium diethylenetriamine penta-acetic
acid (Gd-DTPA)/kg body weight; slice thickness 6.0 mm, gap
0.9 mm].
Results of conventional imaging were classified preoperatively
according to the TNM classification by two experienced radiologists without any knowledge of the background of the patients.
Regional lymph nodes of the head and neck region more than
12 mm in diameter were considered pathological. In addition,
lymph nodes were also staged as tumour involved if other signs of
malignancy, such as grouping of nodes, central necrosis, shape
(e.g. spherical lymph nodes) or pathological contrast material enhancement were encountered.
Statistical analysis. &p.2:Values are given as mean SD. The sensitivity, specificity, positive predictive value, negative predictive value
and accuracy for the imaging modalities were calculated using
standard statistical formulas. Students t-test and Spearman correlation were used to examine the correlation between continuous
variables in the groups, and non-continuous variables were compared with a chi-square test. Statistical significance was assumed
when P0.05.

Results
Among the 60 patients, 15 primary tumours were localized in the tongue, 28 in the floor of the mouth, 5 in the
palate and 12 in the mandibular or maxillary region
(Fig. 1). All patients underwent surgery; 22 had T1, 15
T2, 3 T3 and 20 T4 tumours. Concerning the postoperative grading, 15 primary squamous cell carcinomas were
G1, 30 G2 and 15 G3 (Fig. 2). FDG PET identified the
primary lesions in 59 cases. In one patient with a small

European Journal of Nuclear Medicine Vol. 25, No. 9, September 1998

1257

Fig. 1. Localization of the primary head and neck tumours&ig.c:/f

Fig. 2. Staging and grading of primary squamos cell carcinomas

of the head and neck region&ig.c:/f

and well-differentiated squamous cell carcinoma (T1)

PET failed to detect the primary tumour.
FDG uptake was calculated for each primary lesion in
the attenuation-corrected regional bed position of the

European Journal of Nuclear Medicine Vol. 25, No. 9, September 1998

head and neck region. The SUVBW ranged from 2.5 to

13.8, with a mean of 4.542.8. The comparison of glucose plasma levels and FDG uptake in primary lesions
revealed no statistically significant correlation
(P = 0.42). Furthermore, there was no statistically significant relationship between primary tumour uptake of
FDG and tumour grade (P = 0.9). Interestingly, there
was a tendency towards higher SUVBW values in lymph
node metastases in patients with increased FDG uptake
in primary lesions (P = 0.04).
In our study 1284 lymph nodes were resected and
documented according to level. The preoperative lymph
node staging according to the TNM classification was
compared with postoperative histopathological findings.
Twenty-nine patients revealed no malignant lymph node
involvement (N0), whereas 31 patients demonstrated 117
lymph node metastases. There were ten patients with only one ipsilateral lymph node metastasis of less than
3.0 cm in diameter (N1) (Fig. 3). In 18 patients multiple
ipsilateral malignant lymph nodes (N2b) were diagnosed
and three patients demonstrated bilateral lymph node involvement (N2c). Based on histopathological findings
FDG PET correctly identified lymph node metastases
with a sensitivity of 90% and a specificity of 94%
(P<106). CT and MRI visualized histologically proven
lymph node metastases with a sensitivity of 82% and

Fig. 3ae. Primary carcinoma of the

tongue with one lymph node metastasis
on the left side of the neck (level 2) [T4,
N1, M0]. ac Based on increased glucose utilization, the primary tumour
(coronal and transversal views) and the
lymph node metastasis (a coronal,
b sagittal and c transversal views; arrow) are seen as areas of markedly
enhanced activity. d, e Corresponding
CT (d) and MRI (e) (T1-weighted)
slices through the same anatomical
region (transversal view). The lymph
node metastasis (arrow) demonstrates
the criteria for malignancy: size
>1.5 cm, central necrosis, spherical
node&ig.c:/f

1258
Table 1. Comparison of 18F-FDG PET and conventional imaging
modalities with histopathological findings
a) Estimates of sensitivity, specificity, positive and negative predictive values and accuracy (*P<106)&/tbl.c:&
Method

Sensitivity (%)

SpecifPPV
icity (%) (%)*

NPV
(%)*

Accuracy
(%)

PET*
CT*
MRI*
Sonography*

90
82
80
72

94
85
79
70

99
98
98
96

93
85
79
70

58
35
27
19

PPV, positive predictive value; NPV, negative predictive value&/tbl.:

b) Total number of resected lymph nodes in comparison with histopathological findings&/tbl.c:&

Fig. 4. Preoperative lymph node staging of head and neck cancer;

comparison of FDG PET with conventional imaging modalities&ig.c:/f

Method

Truepositive

Falsenegative

Falsepositive

Truenegative

Total

take of FDG in lymph node metastases and tumour grade

(P = 0.9). Interestingly, benign lymph nodes had increased FDG uptake as a result of inflammatory reactions (SUVBW range: 215.8).

PET
CT
MRI
Sonography

105
96
94
84

12
21
23
33

75
175
250
350

1092
992
917
817

1284
1284
1284
1284

Discussion

&/tbl.:
Table 2. Correlation (chi square test) of
ventional imaging modalities&/tbl.c:&

18F-FDG

PET with con-

Correlation of 18F-FDG PET with:

P-Values

CT
MRI
Sonography

0.016897
0.011566
0.000137

&/tbl.:

80%, respectively (P<106). Sonography revealed a sensitivity of 72% (P<106) (Table 1). The comparison of
18F-FDG PET with conventional imaging modalities
demonstrated statistically significant correlation (Table
2). Concerning anatomical imaging, false-positive results were mainly found in enlarged reactive lymph
nodes which decreased specificity (CT: 85%; MRI: 79%;
sonography: 70%). In the comparative analysis of images, CT was true-negative in 19 of 29 patients (66%),
whereas FDG PET showed no malignant lymph node involvement (N0) in 23 of these patients (79%). Fortythree patients were correctly staged by metabolic imaging with FDG PET. Nine patients were classified as having higher N stages (over-staged) and eight as having a
lower N stage (under-staged). Lymph node staging using
sonography agreed with histopathological findings in 25
patients; in 18 patients there were false-positive classifications (over-staged) because of reactive enlargement of
lymph nodes, whereas 17/60 patients were under-staged.
Both CT and MRI over- and under-staged about 25% of
the patients (Fig. 4). The SUVBW of lymph node metastases ranged from 2 to 11, with a mean of 3.72.0. There
was no statistically significant relationship between up-

Physical and endoscopic examination with open biopsy

to obtain tissues for cytology is considered to be the gold
standard technique for initial diagnosis of head and neck
cancer [2]. Enlarged lymph node metastases in the head
and neck can be detected by physical examination in up
to 85% of patients [17]. Palpable lymph nodes do not always indicate metastatic disease, because reactive enlargement of lymph nodes (inflammation) is common in
patients with head and neck carcinomas. Anatomical imaging modalities have significantly advanced the clinicians ability to detect and determine the extent of primary head and neck carcinomas and the presence of associated adenopathy [18, 19]. Sonography of the cervical region has been applied for the detection of lymph
node metastases with an accuracy between 70% and
89% [12, 20]. In our study population sonography was
able to localize regional lymph node metastases with a
sensitivity of 72% and a specificity of 70%, respectively.
The main limitation of this anatomical diagnostic modality is that tissue can only be visualized to a depth of 46
cm and the results are dependent on the experience of
the investigator [11]. Problems for differential diagnosis
may arise because shape and size are the criteria applied
for tumour involvement of lymph nodes. Therefore sonography is not able to distinguish between reactive enlarged lymph nodes and enlargement due to metastatic
disease.
CT is the preferred modality for the mapping of the
deep tissue, when endoscopy is used for evaluation of
the mucosal extent [21]. Furthermore, CT may be useful
for detecting regional lymph node involvement and for
determining the extent of bone destruction [22, 23]. In
agreement with findings reported by other groups [24],
in the present study CT localized malignant lymph node
involvement with a sensitivity of 82%. In staging head

European Journal of Nuclear Medicine Vol. 25, No. 9, September 1998

1259

and neck cancer CT is still the gold standard, although

MRI is gaining in importance [25]. However, in our
study MRI (specificity 79%) was not better than CT
(specificity 85%) in distinguishing between malignant
and inflamed tissues. If strict criteria of morphological
imaging are applied (size >1.5 cm, central necrosis,
grouping nodes, spherical shape) CT has been shown to
alter N staging in 20%30% of the examined patients
[26]. It should be noted that more than 40% of all lymph
node metastases are localized in nodes smaller than
1.0 cm in diameter [27, 28]. In comparison with these
results, in the present study the smallest lymph node metastasis detected by CT was only 1 cm in diameter,
whereas FDG PET was able to localize smaller lymph
node metastases (0.6 cm in diameter).
In patients with malignant regional lymph node involvement, radical neck dissection is usually performed
in order to eradicate the cervical metastatic foci. Correct
N staging is necessary to define the appropriate surgical
treatment because radical neck dissection affects both the
function and the appearance of patients. Concerning
lymph node staging, both CT and MRI correctly staged
about 50% of our patients when compared with histopathological findings, whereas metabolic imaging with
FDG PET revealed no malignant lymph node involvement in 23 of 29 patients (79%). In this study, FDG PET
detected lymph node metastases with a sensitivity of 90%
and a specificity of 94%, respectively. Similar results
concerning sensitivity and specificity for the detection of
lymph node metastases have been reported by Laubenbacher et al. [29]. It is important to emphasize that increased FDG uptake is also observed in benign lymph
nodes and peritumoral granulation tissue as a result of inflammatory reaction (activated macrophages) [30].
In agreement with findings of other groups, our study
population showed an inhomogeneous FDG uptake in
primary tumours (SUVBW: 2.513.8) and lymph node
metastases (SUVBW: 211) [15, 29, 31]. Possible explanations of this heterogeneity may be the mixture of cell
clones in tumours, different expression of the glucose
transporter gene/glycolysis-related genes or changes induced by oncogenic alteration [32, 33]. The use of other
substances labelled with positron emitters such as L[11CH3]-methionine and L-1-[11C]-tyrosine seems to be
more suitable in the evaluation of patients with palpable
lymph nodes given their potential to differentiate between malignant and inflammatory tissue [3436].
However, L-[11CH3]-methionine shows bilateral symmetrical accumulation in the salivary glands; therefore tumour detection in these regions and especially at the
submandibular and submental lymph node levels may be
problematic [37].
In patients with advanced head and neck tumours,
Valk et al. demonstrated the potential cost-effectiveness
of an FDG PET-based strategy by avoiding surgery for
non-resectable tumours. The savings from contraindicated surgical procedures exceeded the cost of PET imaging by ratios of 2:1 to 4:1, depending on the indication

European Journal of Nuclear Medicine Vol. 25, No. 9, September 1998

[38]. In the current health care environment, imaging

modalities must not only change medical management
but also demonstrate that those changes improve patient
outcome. However, additional studies are necessary to
show the potential effect of FDG PET on the management of patients with head and neck cancer.
In conclusion, this prospective, histopathologically
controlled study confirms FDG PET as the procedure
with the highest sensitivity and specificity for detecting
lymph node metastases of head and neck cancer, and this
has become a routine method in our University Medical
Center. Furthermore, the optimal diagnostic modality
may be a fusion image showing the increased metabolism of the tumour and the anatomical localization.

References
1. Parker SL, Tong T, Bolden S, et al. Cancer statistics 1996. CA
Cancer J Clin 1996; 46: 527.
2. Vokes EE, Weichselbaum RR, Lippman SM, et al. Head and
neck cancer. N Engl J Med 1993; 328: 184194.
3. Forastiere AA. Overview of platinum chemotherapy in head
and neck cancer. Semin Oncol 1994; 21: 2027.
4. Panje WR, Namon AJ, Vokes E, et al. Surgical management of
the head and neck cancer patient following concomitant multimodality therapy. Laryngoscope 1995; 105: 97101.
5. Aisner J, Jacobs M, Sinabaldi V, et al. Chemoradiotherapy for
the treatment of regionally advanced head and neck cancers.
Semin Oncol 1994; 21: 3544.
6. Sham JS, Choy D. Prognostic factors of nasopharyngeal carcinoma: a review of 759 patients. Br J Radiol 1990; 63: 5158.
7. Snow GB, Annanyas AA, Van Slooten EA, et al. Prognostic
factors of neck metastases. Clin Otolaryngol 1982; 7:
185192.
8. Whitehurst JO, Droulias CA. Surgical treatment of squamous
cell carcinoma of the oral tongue. Arch Otolaryngol 1987;
103: 212215.
9. Ensley JF, Maciorowski Z, Hassan M. Cellular DNA content
parameters in untreated and recurrent squamous cell cancers
of the head and neck. Cytometry 1989; 10: 334338.
10. Johnson GS, Williamson KD, Cramer MM, et al. Flow cytometric analysis of head and neck carcinoma DNA index and
S-fraction from paraffin-embedded sections: comparison with
malignancy grading. Cytometry 1985; 6: 461470.
11. Brinkmann G, Brix F, Beigel A. Sonography, CT, MR in soft
part growths in the head and neck region. Rntgenbltter
1990; 43: 5864.
12. van den Brekel MWM, Casteljins JA, Stel HV, et al. Cervical
lymph node metastases: assessment of radiologic criteria. Radiology 1981; 180: 457461.
13. Strauss LG, Ponti PS. The application of PET in clinical oncology. J Nucl Med 1991; 32: 623648.
14. Jabour BA, Choi Y, Hoh CK, et al. Extracranial head and
neck: PET imaging with 2-[F-18]fluoro-2-deoxy-D-glucose
and MR imaging correlation. Radiology 1993; 186: 2735.
15. Lindholm P, Minn H, Leskinen-Kallio S, et al. Influence of
the blood glucose concentration on FDG uptake in cancer a
PET study. J Nucl Med 1993; 34: 16.
16. Hermanek P, Sobin LH. TNM classification of malignant tumours, 4th edn, 2nd revision. Berlin Heidelberg New York:
Springer, 1992.

1260
17. Leicher-Dber A, Thelen U, Bleier R. Palpation and sonography of metastases of the cervical lymph nodes. Rntgenbltter
1989; 42: 195198.
18. Dooms GC, Hricak H, Crooks LE, et al. Magnetic resonance
imaging of the lymph nodes: comparison with CT. Radiology
1984; 153: 719728.
19. Mafee MF, Campos M, Raju S, et al. Head and neck: high
field magnetic resonance imaging versus computed tomography. Otolaryngol Clinic North Am 1988; 21: 513546.
20. Gritzmann N, Czembriek H, Hajek P, et al. Sonographie bei
cervikalen Lymphknotenmetastasen. Radiologie 1987; 27:
118122.
21. Archer CR, Yeager VL. Computed tomography of laryngeal
cancer with histopathological correlation. Laryngoscope 1982;
92: 11731180.
22. Zheng G, Zeng Q, Wu P, Yuan CM. Computed tomography in
the management of nasopharyngeal carcinoma. Clin Radiol
1989; 40: 2529.
23. Mancuso AA, Harnsberger HR, Muraki AS, et al. Computed
tomography of cervical and retropharyngeal lymph nodes:
normal anatomy, variants of normal, and applications in staging head and neck cancer. Part II. Pathology. Radiology 1983;
148: 715723.
24. Moreau P, Goffart Y, Collignon J. Computed tomography of
metastatic cervical lymph nodes. A clinical, computed tomographic, pathologic correlative study. Arch Otolaryngol Head
Neck Surg 1980; 116: 11901193.
25. Helmberger R, Jager L, Grevers G, et al. Computerized tomography of malignancies of the oral cavity, the oropharynx
and hypopharynx and invasiveness. Radiologe 1996; 36:
193198.
26. Stevens MH, Harnsberger HR, Mancuso AA, et al. Computed
tomography of cervical lymph nodes. Staging and management of head and neck cancer. Acta Otolaryngol 1985; 111:
735739.
27. Eichhorn T, Schroeder HG, Glanz T, et al. Histologically controlled comparison of palpation and sonography in the diagnosis of cervical lymph node metastases. Laryngol Rhinol Otol
1987; 66: 266274.

28. van den Brekel MW, Castelijns JA, Snow GB. Imaging of cervical lymphadenopathy. Neuroimaging Clin North Am 1996;
6: 417434.
29. Laubenbacher C, Saumweber D, Wagner-Manslau C, et al.
Comparison of fluorine-18-fluorodeoxyglucose PET, MRI,
endoscopy for staging head and neck squamous-cell carcinomas. J Nucl Med 1995; 36: 17471757.
30. Kubota R, Yamada S, Kubota K, et al. Intratumoural distribution of fluorine-18-fluorodeoxyglucose in vivo: high accumulation in macrophages and granulation tissues studied by microautoradiographic comparison with FDG. J Nucl Med 1992;
3: 19721980.
31. Haberkorn U, Strauss LG, Reisser C, et al. Glucose uptake,
perfusion, and cell proliferation in head and neck tumours: relation of positron emission tomography to flow cytometry. J
Nucl Med 1993; 34: 1217.
32. Field JK, Lamonthe A, Spandidos DA. Clinical relevance of
oncogene expression in head and neck tumours. Anticancer
Res 1986; 6: 595600.
33. Field JK, Spandidos DA, Stell PM, et al. Elevated expression
of the c-myc oncoprotein correlates with poor prognosis in
head and neck squamous cell carcinoma. Oncogene 1989; 4:
14631468.
34. Kubota R, Kubota K, Yamada S, et al. Methionine uptake by
tumour tissue: a microautoradiographic comparison with
FDG. J Nucl Med 1995 36: 484492.
35. Braams JW, Pruim J, Nikkels PGJ, et al. Nodal spread of
squamous cell carcinoma of the oral cavity detected with PETtyrosine, MRI and CT. J Nucl Med 1996; 37: 897901.
36. Leskinen-Kallio S, Nagren K, Lehikoinen P, et al. Uptake of
11C-methionine in breast cancer studied by PET: an association with the size of S-phase fraction. Br J Cancer 1991; 64:
11211124.
37. Leskinen-Kallio S, Nagren K, Lehikoinen P, et al. Carbon-11methionine and PET is an effective method to image head and
neck cancer. J Nucl Med 1992; 33: 691695.
38. Valk PE, Pounds TR, Tesar RD, Hopkins DM, Haseman MK.
Cost-effectiveness of PET imaging in clinical oncology. Nucl
Med Biol 1996; 23: 737743.

European Journal of Nuclear Medicine Vol. 25, No. 9, September 1998