BJU International (1999), 84, 1028–1031

Positron emission tomography with 18fluorine-labelleddeoxyglucose: utility in localized and advancedprostate cancerG. SANZ, J.E. ROBLES, M. GIME NE Z*, J . AROCENA, D. SANCHEZ, F. RODRIGUEZ-RUBIO, D. ROSELL,J.A. RICHTE R* and J.M. BERIANDepartments of Urology and *Nuclear Medicine, University Hospital, School of Medicine, University of Navarra, Pamplona, Spain

Objective To determine the role of the positron emission ases. However, the histopathological analysis of thesenodes showed tumour in three patients. In group Btomography (PET) with 18F-labelled deoxyglucose in

the identification of prostatic cancer in the iliac and the PET scans showed recurrence of prostate cancer(by deposits of radiotracer) more clearly than didobturator lymphatic nodes before radical prostatec-

tomy, and in the localization of relapse in patients in computed tomography (CT) in two patients (both withrecurrence in soft tissue). In one patient bone scinti-biochemical progression.

Patients and methods Twenty-one patients were divided graphy identified a lesion compatible with prostaticdisease in the bone; this was clinically confirmedinto two groups. Group A consisted of 11 men diag-

nosed with organ-confined prostate cancer, where but was not identified by PET.Conclusion PET, using deoxyglucose labelled with 18F,attention was focused on the iliac and obturator

lymphatic nodes, the results being compared with the cannot reliably identify prostatic adenocarcinoma inthe iliac and obturator lymph nodes before surgery;pathological anatomy obtained from surgical pro-

cedures. Group B included 10 patients treated by other tracers may give better results. To locate relapsesin patients with biochemical progression, PET seemsradical prostatectomy, radiotherapy or orchidectomy

and who were in biochemical progression, in whom to have better sensitivity than CT when identifyingdiseases in soft tissues and is possibly inferior to bonethe aim was to identify recurrence of the disease.

Results In none of the 11 patients of group A who had scintigraphy in detecting bony metastases.Keywords Positron emission tomography, prostateundergone radical prostatectomy were deposits of

radiotracer identified in the area of the iliac and cancer, 18F-deoxyglucose, progression, prognosisobturator nodes which would indicate node metast-

tracers such as 18F-deoxyglucose. The tracers accumulateIntroduction

as a consequence of an increase in glycolytic metabolismin areas of tumour activity. The results have been goodProstate cancer is the second commonest cause of death

from cancer in men; about a third of the tumours are in initial staging and in later monitoring to detectrelapses. PET may have potential in prostate cancer andorgan-confined at diagnosis. The identification of distant

metastases or local progression signals a radical change several studies have reported the use of PET as adiagnostic technique for this disease. ECert et al. [2]in the treatment of the disease and in its prognosis [1].

Prostate cancer is usually staged using serum PSA level, determined the sensitivity of PET in diCerentiating BPHfrom prostate cancer, and Shrene et al. [3] assessed thea DRE, CT and bone scintigraphy. CT is unable to identify

lesions of <1 cm, and bone scintigraphy, despite its sensitivity of PET in identifying soft-tissue relapses.Seltzer et al. [4] and Hoh et al. [5] evaluated the capacitybetter sensitivity in detecting bony metastases, is not

specific if there has been any recent skeletal trauma, and of PET to identify metastases after radical treatment,concluding that in patients with a PSA level ofobviously provides no information about extra-skeletal

metastases (common in such patients). >4 ng/mL, PET and CT had the same sensitivity andwere better than a monoclonal antibody scan. CarlinFor many common tumours (e.g. breast, lung, lym-

phoma and melanoma) positron emission tomography et al. [6] also investigated the capacity of PET to identifylymph node involvement before surgery; there was com-(PET) allows complete body staging after administeringplete agreement between the histopathology results andthe findings on PET.Accepted for publication 21 July 1999

1028 © 1999 BJU International

PE T WITH 1 8 F-DEOXYGLUCOSE 1029

Thus, the aim of the present study was to determine a standardized uptake value (SUV); an SUV of >2.5 wasconsidered to indicate significant pathology. From thethe role of PET, using 18F-labelled deoxyglucose [7,8], in

identifying prostate cancer in the iliac and obturator images the presence of hypermetabolic foci at the prostatelevel and in areas of lymphatic drainage was evaluatedlymph nodes before radical prostatectomy, and to locate

relapses in patients with biochemical progression. (iliac and obturator chains) analysing the body scanimage in the axial, coronal and sagittal planes.

Patients and methodsResults

The study comprised 21 patients assessed between July1996 and February 1999; the patients were divided into In none of the 11 patients of Group A were deposits of

radiotracer identified in the iliac-obturator lymph nodestwo groups. Group A included 11 patients (mean age64 years, range 56–75, mean PSA level 27.7 ng/mL, which would indicate modal metastases (Fig. 1a).

However, the histopathological analysis of these noderange 10.1–41) diagnosed with prostate carcinoma byTRUS-guided transrectal biopsy and in whom CT and chains showed nodal metastases in three patients, none

of which were identified on PET or on CT. The firstbone scintigraphy suggested organ-confined disease.Complete body staging was carried out using PET before patient (Gleason 5+3 PSA 10.1 ng/mL) had an involved

node in the left chain. The second patient (Gleason 4+5,surgical treatment, focusing on the region of the iliac-obturator lymph nodes. The images were then compared PSA 40.6 ng/mL) had disease in a node of the right

iliac-obturator chain and the third patient (Gleasonwith the pathological anatomy obtained after surgery.Group B included 10 patients (mean age 66 years, 5+3, PSA 41 ng/mL) was positive for tumour in two

nodes (one in the right and one in the left chain) andmean PSA level 15 ng/mL before PET, mean time toprogression 3 years) with prostatic carcinoma previ- thus radical prostatectomy was not completed. With so

few patients the values for sensitivity, specificity and theously treated by radical prostatectomy (three), radicalirradiation (three), both (three) or orchidectomy (one), predictive values are invalid.

In group B, PET identified relapse of adenocarcinomaand in whom relapse was suspected because they hadbiochemical progression. In these patients PET was used more clearly than did CT in three patients, in whom the

relapse was in soft tissues (Fig. 1b). In three patientsto identify relapse of the disease. The results of PET werecompared with the information from CT and scintigra- disease was identified using other techniques and in

another three PET, CT and bone scintigraphy detectedphy, and especially with the clinical course of the disease.Before PET, the patients fasted for 4–6 h, had a bladder no relapse. In the last patients bone scintigraphy ident-

ified a lesion compatible with bony metastasis and whichcatheter inserted and were confined to bed-rest, to avoidimaging at the muscular level. Initially a ‘transmission’ was clinically confirmed, but this was not identified by

PET. Again, with so few patients, the determination ofimage of 15 cm was obtained, focusing on the symphysispubis before the administration of radiotracer and again sensitivity, specificity and predictive values is invalid.after a 7-min delay. Subsequently, 370 MBq of 18F-deoxyglucose was administered, allowing 45 min for

Discussionincorporation before beginning the acquisition of theimages (emission study). Since the studies by Inaba et al. [9,10], in one of which

PET was used to assess prostatic pathology, showing anAn entire body scan was carried out starting from thecervical region to the groin (35–42 min duration), increase in blood flow in malignancy relative to that in

BPH, there have been few studies to determine the roleinjecting an 0.5 mg/kg bolus of intravenous frusemideat the start, which facilitated the renal elimination of of PET in prostatic pathology. ECert et al. [2] used PET

with 18F-glucose and concluded that it was not capableradiotracer. Halfway through the study the bladdercatheter was clamped and the bladder filled with saline of diCerentiating BPH and prostate cancer. The present

study determined the utility of PET in primary tumoursolution, to dilute the activity of 18F-deoxyglucose in thebladder and minimize the artefacts caused by retention extension and its capacity to identify lymph node involve-

ment; such information could avoid radical surgery inof radiotracer at this level. Furthermore, this provided auseful anatomical reference structure in the pelvis. selected patients. Moreover, in a patient who has under-

gone radical treatment, the early diagnosis of relapseOn completing the complete body scan, a new isolatedimage of the pelvis was acquired (15 cm) over 7 min, and its location could allow the use of more aggressive

local therapies with the aim of extending survival.focusing on the same location as in the transmissionstudy. Both transmission and emission images of the In group A, PET was unable to identify lymph node

involvement at any time, while histopathologically fourpelvis were processed together to quantify the activity ofany pathological deposit at the pelvic level and to obtain nodes were infiltrated by tumour of 5–8 mm. However,

© 1999 BJU International 84, 1028–1031

1030 G. SANZ et al.

Fig. 1. PET scans showing: a, prostaticcarcinoma with accumulation of radiotracerin the prostatic area. The area of theilio-obturator nodes shows no deposit of18F-deoxyglucose. b, metastatic deposits ofprostatic cancer in the para-aortic and iliacnodes before radiotherapy and the

b

a

disappearance of the deposit after treatment.

improvements in the associated metabolic tracers may monoclonal antibody (ProstaScint@) which targets pros-tate-specific membrane antigen. The largest study withallow lymph node involvement to be determined more

accurately; indeed, 18F-glucose does not seem to be a ProstaScint was that by Elgamal [11], which included100 patients with a mean PSA level of 56 ng/mL; theuseful radiotracer. This probably represents the current

limitation in the use of PET in identifying deposits of the results were promising, although the high PSA level didnot allow its sensitivity to be assessed. Seltzer et al. [4]iliac and obturator lymph nodes, because this tracer

accumulates in the bladder. Although this artefact was found that PET was the best technique for detecting PSArelapse in patients treated for localized prostate carci-minimized in the present study by retrograde filling of

the bladder with saline, the use of PET for prostatic noma. They compared PET with CT and ProstaScint inpatients with a mean PSA level of 3.25 ng/mL; in 17carcinoma will depend on the use of radiotracers which

are not eliminated via the bladder. patients assessed, ProstaScint was completely insensitive.Two studies [12,13] reported the successful synthesisIn group B, PET was more sensitive than CT for lesions

in soft tissues and less sensitive than scintigraphy for of 18F-ligands with a high aBnity for androgen receptorthat may be useful as radiotracers in prostate cancer.lesions in bone, as reported previously [3]. Several studies

have used monoclonal antibody to target PSA, with This high aBnity would allow the identification of allstructures in which receptors are present [14], thusmany tracers, e.g. 99mTc-CYT-351 and 111In-labelled

© 1999 BJU International 84, 1028–1031

PE T WITH 1 8 F-DEOXYGLUCOSE 1031

and blood, by positron emission tomography. J Urol 1992;obtaining complete bone scintigraphy in hormone-148: 1457dependent prostate cancer. One of these ligands [15,16]

10 Inaba T, Yamashita M, Kawwase Y, Nakahashi H,(16-b-18F-5a dihydrotestosterone) was an ideal tracerWatanabe H. Quantitative measurement of renal plasmawhen used in baboons [13]. Thus the future of PET inflow by positron emission tomography with oxygen-15-prostate cancer depends on identifying ligands with highwater. Tohoku J Exp 1989; 159: 283

sensitivity and specificity for prostatic tissues. Further11 Elgamal AA. ProstaScint may enhance identification of

studies are needed to establish the definitive role of PET prostate cancer recurrences after prostatectomy, radiationin prostate cancer. or hormone therapy: analysis of 100 patients. Prostate

1998; 37: 261–912 Liu A, Carlson KE, Katzenellenbogen JA. Synthesis of high

aBnity fluorine-substituted ligands for the androgen recep-References tor. Potential agents for imaging prostatic emission tom-

1 Watanabe H. Diagnostic imaging of the prostate. Nippon ography. J Med Chem 1992; 35: 2113–29Hinyokika Gakkai Zasshi 1991; 1982: 1–15 13 Bonasera TA, O’Neil JA, Xu M et al. Preclinical evaluation

2 ECert PJ, Bares R, Handt S, Wolf JM, Bull U, Jakse G. of fluorine-18-labeled androgen receptor ligands inMetabolic imaging of untreated prostate cancer by positron baboons. J Nucl Med 1996; 37: 1009–15emission tomography with 18fluorine-labeled deoxyglucose. 14 Pertschuk LP, Rosenthal HE, Macchia RJ. Correlation ofJ Urol 1996; 155: 994–8 histochemical and biochemical analysis of androgen bind-

3 Shreve PD, Grossman HB, Gross MD, Wahl RL. Metastatic ing in prostate cancer: relation to therapeutic response.prostate cancer: Initial findings of PET with 2-deoxy- Cancer 1982; 49: 984–932-(F-18)fluoro-d-glucose. Radiology 1996; 199: 751 15 Bonasera TA, O’Neil JP, Choe YS. Imaging the prostate in

4 Seltzer M, Narth J, Cangiano T et al. Comparison of baboons with fluorine 18-labeled androgen receptorcomputer tomography (CT), positron emission tomography ligands. J Nucl Med 1994; 35: 53(PET) and monoclonal antibody scan (MAB) for evaluation 16 Hoyte RM, Borderton K, Bryson K, Allen R, Hochberg RB,of lymph node (LN) metastases in patients with PSA relapse Brown TJ. Synthesis and evaluation of 7-iodo-after treatment for localised prostate cancer (CAP). 5-dihydrotestosterone as a potential radioligand for theProceedings of the American Urologic Association Congress, androgen receptor. J Med Chem 1994; 37: 1224–301998

5 Hoh CK, Seltzer MA, Franklin J, DeKernion JB, Phelps ME,AuthorsBelldegrun A. Positron emission tomography in urological

oncology. J Urol 1998; 159: 347–56 G. Sanz, Resident.J.E. Robles, PhD, FEBU, Consultant Urologist.6 Carlin I, Martin I, Resnick PF, Faulhaber FM. Alteration in

PET scanning technique increases accuracy in detecting M. Gimenez, Resident.J. Arocena, Resident.lymphatic spread of prostate cancer: Proceedings of the

American Urologic Association Congress, 1998 D. Sanchez, Resident.D.F. Rodriguez-Rubio, PhD, FEBU, Urologist.7 Haberkorn U, Morr I, Oberdorfer F et al. Fluorodeoxyglucose

uptake in vitro: aspects of method and eCects of treatment D. Rosell, PhD, Consultant Urologist.J.A. Richter, Chief of Department of Nuclear Medicine.with gemcitabine. J Nucl Med 1994; 35: 1842–50

8 Hoh CK, Hawkins RA, Glaspy JA et al. Cancer detection J.M. Berian, Professor and Head of Department of Urology.Correspondence: Dr G. Sanz, Department of Urology, Clınicawith whole body PET using (18) fluoro-2-deoxy-d-glucose.

J Comput Assist Tomogr 1993; 17: 582–9 Universitaria, Faculty of Medicine, University of Navarra,31080 Pamplona, Spain.9 Inaba T. Quantitative measurements of prostatic blood flow

© 1999 BJU International 84, 1028–1031