Molecular Imaging and Biology

, Volume 8, Issue 1, pp 43–48

Positron Emission Tomography/Computed Tomography with F-18-fluorocholine for Restaging of Prostate Cancer Patients: Meaningful at PSA < 5 ng/ml?


    • Department of Nuclear Medicine, PET Center LinzSt. Vincent’s Hospital Linz
  • Albert Dirisamer
    • Department of RadiologyUniversity Hospital Vienna
  • Wolfgang Loidl
    • Department of UrologySt. Vincent's Hospital Linz
  • Franz Stoiber
    • Department of UrologySt. Vincent's Hospital Linz
  • Bernhard Gruy
    • Department of Nuclear Medicine, PET Center LinzSt. Vincent’s Hospital Linz
  • Silke Haim
    • Department of Nuclear Medicine, PET Center LinzSt. Vincent’s Hospital Linz
  • Werner Langsteger
    • Department of Nuclear Medicine, PET Center LinzSt. Vincent’s Hospital Linz
Original Article

DOI: 10.1007/s11307-005-0023-2

Cite this article as:
Heinisch, M., Dirisamer, A., Loidl, W. et al. Mol Imaging Biol (2006) 8: 43. doi:10.1007/s11307-005-0023-2



According to reports, re-staging of patients suffering from prostate cancer by positron emission tomography (PET) using C-11-choline has failed to produce positive findings at a PSA level of < 5 ng/ml. Hence, the purpose of our study has been to determine whether this is true also for PET/CT using F-18-fluorocholine (FCH PET/CT) or whether it is possible to obtain true positive results by FCH PET/CT even at lower PSA levels.


In 34 patients with prostate cancer who had undergone initial therapy (radical prostatectomy n = 31, radiotherapy n = 3), a PET/CT scan was performed using F-18-fluorocholine (FCH) during follow-up in case of demonstrable or rising PSA levels. Current PSA levels were determined in all patients at the time of examination.


Median PSA in FCH positive patients was 6.1 ng/ml (mean PSA 17.1 ng/ml), median PSA in FCH negative patients was 2.3 ng/ml (mean PSA 3.4 ng/ml), respectively (p < 0.05). In eight of 17 examinations (47%) with PSA < 5 ng/ml, at least one FCH-positive focus was detected. So far the findings could be confirmed by correlating imaging methods (CT and/or MR), biopsy/histology and the course of the disease, respectively, in seven of the eight FCH-positive cases with PSA < 5 ng/ml, so that a true positive FCH PET/CT finding was obtained all in all in seven of 17 (41%) examinations with PSA < 5 ng/ml. In four of these seven FCH PET-positive patients with PSA < 5 ng/ml, adjuvant hormonal therapy was administered at the time of the examination or prior to the examination.


In re-staging patients with prostate cancer, FCH PET/CT is able to yield true positive findings even at PSA < 5 ng/ml. Therefore, FCH PET/CT should not be restricted to patients with PSA > 5 ng/ml.

Key words

F-18-fluorocholinePET/CTRestaging of prostate cancerProstate specific antigen (PSA)


In prostate cancer, recurrences are not uncommon after initial curative therapeutic approach (radical prostatectomy or radiation) [1, 2]. Recurrence is chiefly revealed by a rise in the prostate-specific antigen (PSA). However, localizing the recurrence presents a major challenge. Although basically several imaging methods are available, they altogether fail to produce optimal results [36]. Not a single one of these imaging methods is able to reliably demonstrate local recurrences, malignant lymph nodes and skeletal metastases. Apart from imaging methods, the velocity of increase in PSA is used to distinguish local recurrence from distant metastases. [7, 8]. Positron emission tomography (PET) using C-11-choline suggests itself as a promising method to localize biochemical recurrence after prostate cancer [9, 10]. Apart from C-11-choline, F-18-labeled choline has been available since recently. DeGrado et al. developed 18-F-fluoromethyl-dimethyl-2-hydroxyethyl-ammonium [1114]. Hara et al. introduced F-18-fluoroethylcholine [15]. The benefit of F-18-labled tracers versus C-11-labeled choline is the longer half life of the former, which is crucial in particular for establishing a satellite concept (one cyclotron supplies several institutes). To our knowledge, no studies exist in which PET using F-18-labled choline to localize a recurrence after prostate cancer identified by PSA has been clinically investigated. Moreover, another work reports that no positive finding could be obtained by PET using C-11-choline at a PSA level of <5 ng/ml [9]. Considering these facts, the question arises whether the use of PET with 18-F-fluorocholine (FCH) in patients with prostate cancer in the follow-up after initial therapy should be restricted to patients with PSA > 5 ng/ml or whether the application of this method makes sense in case of PSA levels of < 5 ng/ml as well.

Material and Methods

From October 2003 to November 2004, 45 patients with prostate cancer were referred to our department for 18-F-fluorocholine PET/computed tomography (CT) (FCH PET/CT) because of demonstrable or rising PSA levels after prostatectomy or radiotherapy in order to localize the recurrence. After having been informed accordingly, all patients consented in writing to undergo the examination. From among these patients, those were included in the study (n = 34), from whom a PSA value was available which 1) had been determined at our own laboratory and 2) in whom the time interval between PSA determination and PET/CT scan did not exceed one week. The mean age of the patients was 69.9 years (between 52.6 and 86.8 years). Thirty-one patients had previously undergone prostatectomy (10 of these patients had received adjuvant external radiation), in three patients radiation was performed as primary therapy (external radiation in one patient, brachytherapy in two patients). Sixteen patients were under hormonal therapy at the time of examination.

Radiochemical Synthesis of F-18-fluorocholine

F-18-fluorocholine (FCH) was synthesized from N,N-dimethylethanolamine (DMAE) and F-18-fluoromethyl triflate in a fully automated solid phase fluoromethylation reaction at the Radiopharmaceutical Production Facility of ARGOS Zyklotron GesmbH Linz. The radiochemical purity as well as the DMAE concentration was determined by HPLC and GC [16]. The average concentration of DMAE in the final product did not exceed 1 μg/ml and the radiochemical purity was >99%. The pyrogenicity and isotonicity of the product was tested before application.

PET/CT Acquisition Protocol and Processing

4.07 MBq/kg bodyweight FCH (IASOCHOLINE®, ARGOS Zyklotron GesmbH Linz, Austria) was administered by intravenous injection after the patients had been fasting for at least 12 hours. Imaging was performed on an integrated PET/CT system (Discovery LS®, GE Medical Systems, Milwaukee, USA). A dynamic PET acquisition in the pelvic region during eight minutes (1 min frames), starting one minute post injection was performed first, followed by a static acquisition from thigh to base of skull (starting 14 to 19 min p.i., mean 17,6 min p.i.), four minutes emission per bed position. Attenuation correction CT scan: 140 kV, 80 mA, 0.5 s/revolution. All PET scans were acquired in 2D mode, the images were reconstructed using an iterative reconstruction algorithm, CT data were used for attenuation correction. Both, coronal, transaxial, sagittal slices and MIP (maximum intensity projection) images were interpreted on an eNTEGRA® workstation (GE medical systems, Milwaukee, USA).

PET/CT Interpretation

Every focal tracer accumulation deviating from the physiological distribution of the tracer was regarded as being positive, in particular, of course, if accumulated in the prostatic fossa, in lymph nodes and in the skeletal system. The classification of FCH PET/CT-positive or FCH PET/CT-negative was made solely on the basis of the PET finding whereas the CT set of data was merely used to localize a FCH-positive lesion; CT criteria (such as, e.g., size of the lymph nodes to determine their dignity or osseous scleroses in the sense of osteoplastic metastases) were not taken into account in the evaluation. For differential diagnosis of focal ureter activity versus pathological lymph nodes in the pelvis, the dynamic frames were used: Accumulations presenting themselves from the very beginning in the dynamic acquisition phase were interpreted as lymph nodes, those occurring in a later frame of the acquisition as tracer depot in the ureter.

Accumulations in the prostatic fossa were interpreted the same way: Accumulations that were visible right from the beginning of dynamic acquisition were considered to be a local recurrence, activities occurring later during dynamic acquisition were believed to be tracers in the urinary bladder.

PSA Determination

PSA in the serum was determined at our own laboratory using a sandwich immunoassay and applying direct chemiluminescence technology (ADVIA Centaur®, Bayer AG, Leverkusen, Germany). The lower detection limit of the test (analytical sensitivity) is 0.01 ng/ml.


Statistics were calculated using SPSS 11.5.


Median PSA in FCH positive patients was 6.1 ng/ml (mean PSA 17.1 ng/ml), median PSA in FCH negative patients was 2.3 ng/ml (mean PSA 3.4 ng/ml), respectively (p < 0.05).

In eight of 17 examinations (47%) with PSA < 5 ng/ml, at least one FCH-positive focus was diagnosed. So far, in seven of the eight FCH-positive cases with PSA < 5 ng/ml, malignancy could be confirmed by either correlating imaging methods (CT and/or MR), biopsy/histology, or the course of the disease, respectively. This means that altogether at least one true positive FCH PET/CT finding was obtained in seven of 17 (41%) patients with PSA < 5 ng/ml. The detailed results of these seven cases are summarized in Table 1. As regards the remaining patient, no evidence could be provided so far whether the lesion identified by FCH PET/CT was true positive. FCH PET/CT in this patient showed a bone metastasis, however neither biopsy nor other imaging methods have been performed so far. Figure 1 shows a FCH PET/CT scan of a patient after radical prostatectomy because of a prostate carcinoma pT3a pN0 M0 G3. It depicts a focus of approx. 15 mm in size of enhanced choline metabolism in the prostatic fossa. The PSA level at the time of the FCH PET/CT scan was 1.98 ng/ml. The subsequently performed biopsy revealed material from a prostate carcinoma G3, Gleason Score 9 (4 + 5).
Table 1

Results from patients with positive FCH PET/CT scan at PSA < 5 ng/ml


PSA (ng/ml)a

Hormonal therapya

Localization of FCH positive lesions

Mode of validation

S. G.



Prostate fossa, bone

Bone metastases by CT and MRI

H. K.



Lymph nodes, bone

Bone and lymph node metastases by serial CT scans (rapid progression)

P. W.





S. J.


Bicalutamide 150 mg

Prostate fossa, lymph nodes, bone

CT (all lesions)

R. P.


Bicalutamide 50 mg; Gosereline 10.8 mg

Prostate fossa

Biopsy/histology: local recurrence

K. K.



Prostate fossa

Histology: local recurrence

S. H.



Lymph node

second FCH PET/CT after six months: marked increase in size and uptake

aat the time of the FCH PET/CT scan.

bBicalutamide until 3 months prior to the FCH PET/CT scan.
Fig. 1

FCH PET/CT (transaxial slices) of a patient after radical prostatectomy (A: PET image; B: corresponding CT scan; C: corresponding PET/CT fusion image). The PSA level at the time of examination was 1.98 ng/ml. It shows intensive focal uptake in the prostate fossa. Biopsy/histology confirmed the presence of a local recurrence of the prostate carcinoma.


The determination of one or several tumor markers is of crucial importance in the follow-up of numerous tumors. If the presence of or rise in tumor markers is detected in the follow-up, the imaging methods are challenged to identify the precise site of the recurrence so as to decide on the further therapeutic regime. In this clinical situation—re-staging because of elevated tumor markers—PET using 2-deoxy-2- [F-18] fluoro-D-glucose (FDG) has proven to be a highly efficient method. In the re-staging of colorectal cancer because of elevated CEA level, for example, FDG-PET has been acknowledged as the method of choice [17, 18], equally in differentiated thyroid cancer in case of an elevated thyroglobuline level and negative I-131 scan [19]. In prostate carcinoma, however, the employment of FDG-PET has turned out to be limited in value. Results have been particularly disappointing as regards the diagnosis of recurrences [20]. On the other hand, alternative tracers, among which, apart from acetate, choline appears to have a considerable potential, have been developed. Choline is transported into cells, phosphorylated, thus trapped within the cells and used for synthesis of phospholipids. It could be shown that malignant cells have elevated levels of choline and an up-regulation of the activity of choline kinase [21]. These findings led to the development of positron-labeled choline analogues for PET imaging of cancers. PET using C-11-choline yielded promising results both in preoperative staging and in re-staging, in prostate cancer [9, 10, 22]. However, C-11-labeled PET tracers have the disadvantage of a short half-life, and the necessity of an onsite cyclotron presents an obstacle to their wider application in clinical routine. This drawback could be eliminated by the development of F-18-labeled choline [1115]. Yet clinical studies on FCH PET in prostate cancer are still rare and, to our knowledge, none of them has investigated the role of FCH PET in re-staging patients with prostate cancer [23].

Considering the detection limits inherent in the various imaging techniques, it is not surprising that frequently no morphologically detectable correlate of the recurrence can be found in case of low tumor marker levels. In order to make efficient use of examination methods, some of which are very costly, investigations have been made in various tumors to determine from which level of tumor marker onward it appears to be meaningful to employ a given imaging method [2427]. De Jong et al. were unable to obtain positive findings in the re-staging of patients with prostate cancer using C-11-choline PET at PSA levels <5ng/ml [9]. Picchio et al. report positive findings using C-11-choline PET even at lower PSA levels, however, without providing data as regards numbers [10]. In our group of patients, at least one FCH-positive focus could be found, after all, in eight of 17 examinations with PSA < 5 ng/ml; seven of these were confirmed by correlating imaging methods (CT and/or MR), by means of biopsy/histology or the course of the disease. The reason why we—contrary to de Jong—did obtain true positive results in several cases at PSA < 5 ng/ml may primarily be due to the composition of our group of patients. In at least one of the five FCH PET-positive patients with PSA < 5 ng/ml, a dedifferentiation of the prostate cancer had apparently occurred: FCH PET/CT showed bone and lymph node metastases although the PSA level was as low as 0.03 ng/ml. During the next months, the disease progressed rapidly, developing metastases in the lungs, liver, suprarenal gland, bones and lymph nodes. Actually, in this case we are not seeing a true positive FCH PET finding at low PSA level but a “false” negative PSA level due to dedifferentiation. In four of the seven FCH PET-positive patients with PSA < 5 ng/ml, adjuvant hormonal therapy was administered at the time of the examination or prior to the examination. It is well known, that in patients who have already seen hormonal therapy, the magnitude of the PSA level is likely to be suppressed and may not correlate well with tumor size or metabolism. Moreover, although there are reports according to which choline uptake decreases after initiating a hormonal therapy [28], we do not know whether the influence on choline metabolism and on PSA level happens in parallel. It cannot be ruled out that the FCH PET signal is influenced less strongly than the PSA level. In the study by deJong, patients receiving adjuvant hormonal therapy were excluded.

FCH Exhibits More Favorable Physical Properties Compared to C-11-choline

Hara et al compared C-11-choline with F-18-labeled choline in 16 patients with prostate cancer (this study group used F-18-fluoroethylcholine) and found out that the F-18-labeled tracer gave a slightly higher quality of image than the C-11-labeled tracer in terms of spatial resolution. This observation was explained by the shorter positron range of F-18 as compared to C-11 [15]. This might have played a role in the detection of small lesions such as, e.g., a small local recurrence or small malignant lymph nodes. Another difference was the type of scanner that had been employed. While our department works with an integrated PET/CT scanner, conventional PET scanners were used by de Jong. PET/CT fusion images may be helpful in distinguishing physiological uptake in the intestines, for example, from pathological uptake such as in a lymph node or a skeletal metastasis in the pelvis by providing more precise information on the anatomical relationship.

Performing dynamic acquisition is most helpful in FCH PET [29]. Contrary to C-11-choline, FCH is eliminated via the kidneys. Basically, there is a danger of confusing “hot” urine in the urinary bladder with a local recurrence as well as the danger of not being able to discriminate between a local recurrence close to the urinary bladder and the bladder itself and thus to demonstrate it. However, since the uptake of tracer in malignant lesions occurs promptly whereas hot urine in the urinary bladder appears later (approx. 5.–8. min. p.i.) [14], there is a time window in which it is possible to distinguish between simple urine and a malignant lesion. Considering the numbers of de Grado quoted above, it would basically also be possible to perform an early static acquisition rather than a dynamic one, e.g. in the first four minutes p.i. But since, according to our experience, the earliest moment at which the tracer appears in the urinary bladder varies considerably, we prefer dynamic acquisition. Furthermore, it has turned out to be useful to determine first from the sequence of the one minute images from which frame x onward activity is visible in the urinary bladder (and/or the ureters) and then to generate a “condensed” image from the raw data of the frames 1 to x - 1. The advantage of this image versus the 1 min images is the count rate, which is several times higher. Hara et al. solved the problem of renal elimination of the tracer by continuous bladder irrigation using an urinary catheter to eliminate the bladder radioactivity. He reports, however, that this procedure was very uncomfortable for the patients [15]. We therefore clearly prefer dynamic images over a urinary catheter. Figure 2 shows a biopsy proven local recurrence of a prostate carcinoma. The focal uptake is merely visible in the dynamic sequence but not on the static scan of the trunk, in which the lesion cannot be discriminated from the urinary bladder.
Fig. 2

FCH PET/CT of the same patient as in Fig. 1. A: Dynamic PET acquisition, coronal slices. Time sequence from top to bottom. Starting already from the first frame, focal uptake in the lesser pelvis corresponding to a local recurrence is demonstrated (arrow). B: Fusion image of a PET frame of the dynamic acquisition and CT. The local recurrence is clearly visible since the urinary bladder is still empty. c: PET/CT fusion image of the static acquisition (from the base of the skull to the proximal femur) following the dynamic scan 15 min. p.i.. As soon as the urinary bladder has filled up, the local recurrence can no longer be demonstrated reliably because it is superimposed by active urine in the bladder.


In re-staging patients with prostate cancer, FCH PET/CT is able to yield true positive findings even at PSA < 5 ng/ml. Therefore, FCH PET/CT should not be restricted to patients with PSA > 5 ng/ml.

Copyright information

© Academy of Molecular Imaging 2005