Department of Radiology, Charité-Universitätsmedizin Berlin, Berlin, Germany
Received date: July 20, 2016; Accepted date: August 09, 2016; Published date: August 16, 2016
Citation: Schreiter V, Gericke M, Heimann U, Steffen I, Stelter L, et al. (2016) Comparison of [68Ga]Ga-PSMA-HBED-CC PET versus Whole-Body Bone Scintigraphy for the Detection of Bone Metastases in Patients with Prostate Cancer. J Nucl Med Radiat Ther 7:302. doi: 10.4172/2155-9619.1000302
Copyright: © 2016 Schreiter V, et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Visit for more related articles at Journal of Nuclear Medicine & Radiation Therapy
Aim: To compare [68Ga]Ga-PSMA-HBED-CC PET and (99m)Tc-DPD bone scintigraphy for the detection of bone metastases from prostate cancer. Methods: [68Ga]Ga-PSMA-HBED-CC PET/CT and (99m)Tc-DPD bone scintigraphy in 19 men with histopathological proven prostate cancer were compared to each other for the sensitivity/specificity, accuracy, positive predictive value (PPV) and negative predictive value (NPV) for the detection of bone metastases. Results: According to the standard of reference lesion-based analysis of [68Ga]Ga-PSMA-HBED-CC PET and (99m)Tc-DPD bone scintigraphy reached a sensitivity of 45.6%/34%, specificity of 86.4%/81.4%, accuracy of 60.5%/ 51.2%, positive predictive value of 85.5%/76.1%, and negative predictive value of 47.7%/41.4%, respectively. Conclusion: [68Ga]Ga-PSMA-HBED-CC PET could detect significantly more bone metastases in prostate cancer than (99m)Tc-DPD bone scintigraphy.
[68Ga]Ga HBED-CC PSMA; (99m)Tc-DPD scintigraphy; Bone metastases; Detection rate; Prostate cancer
PPV: Positive Predictive Value; NPV: Negative Predictive Value; PSMA: Prostate Specific Membrane Antigen; 18Fcholine: 18F-fluoromethylcholine; PET/CT: Positron Emission Tomography/Computed Tomography; LEHR: Low-Energy High- Resolution Collimator; BS: Bone Scan; PSA: Prostate Specific Antigen; SD: Standard Deviation; HU: Hounsfield Units; RRP: Radical Retropubic Prostatectomy; LNE: Lymphadenoctomy; RTX: Radiotherapy; AHT: Anti-Hormonal Treatment
68Ga-labeled Glu-urea-Lys(Ahx)-HBED-CC ([68Ga]Ga HBED-CC PSMA) has recently emerged as a new radiotracer. It is a promising candidate for staging and re-staging prostate cancer using hybrid imaging technology such as positron emission tomography/computed tomography (PET/CT) [1-3]. Prostate-specific membrane antigen (PSMA) is expressed by almost all prostate cancer cells and is used as a target for imaging and therapy [4-6].
Imaging with [68Ga]Ga-PSMA-HBED-CC has shown advantages over 18F-fluoromethylcholine (18F-choline) PET/CT - the molecular imaging modality currently used in patients with suspected recurrence of prostate cancer based on laboratory chemistry tests: [68Ga]Ga- PSMA-HBED-CC PET/CT improves detection of prostate cancer and its metastases due to higher specific radiotracer uptake rates within the abnormal lesions and lower tracer uptake in surrounding normal tissue. Detection is especially improved in patients with small increases in prostate-specific antigen (PSA) levels . In contrast, the value of [68Ga]Ga-PSMA-HBED-CC PET/CT vis a vis technetium-99m-3,3- diphosphono-1,2-propanedicarboxylic acid ((99m)Tc-DPD) bone scintigraphy (BS) remains to be determined. The latter is currently used for detection of bone metastases in patients with high-risk prostate cancer. Prostate cancer is the most common cancer of men in western industrialized countries . In 2012 alone, there were 28,000 prostate cancer deaths in the United States . Prostate cancer tends to metastasize to the skeletal system . Most skeletal prostate cancer metastases are osteoblastic, but other types of bone metastases, such as osteolytic or mixed types, can also occur .
BS with (99m)Tc-DPD detects the pathological osteoblastic response adjacent to metastatic bone lesions, making it a highly sensitive and cost-efficient whole body staging modality for bone metastases associated with prostate cancer . The aim of this study was to compare the sensitivity, specificity, accuracy, and the positive/negative predictive value (PPV/NPV) of [68Ga]Ga-PSMA-HBED-CC PET and (99m)Tc-DPD BS in the detection of different types of bone metastases in prostate cancer patients.
We performed a database search for patients with prostate cancer who underwent standardized [68Ga]Ga-PSMA-HBED-CC (68Ga- DKFZ-11) PET/CT and (99m)Tc-DPD (Teceos®, CIS bio GmbH, IBA Molecular, Germany) BS for prostate cancer staging at our institution. The [68Ga]Ga-PSMA-HBED-CC PET/CT was done as additional imaging modality depending on the patients risk of having metastases or after unclear findings in former imaging. Inclusion criteria were: PET/CT and BS performed no more than one month apart; both examinations performed according to the institutional standard protocols; no relevant changes in tumor surrogate parameters, no relevant clinical deterioration, and no cancer-specific treatment of patients between the two imaging examinations. Following these criteria we included 19 patients in our analysis.
The clinical data documented for each patient included histopathologically proven prostate cancer if available, Gleason score evaluation, PSA levels, and clinical stage at time of prostate cancer diagnosis as well as prior imaging by [68Ga]Ga-PSMA-HBED-CC PET/CT and (99m)Tc-DPD BS. Clinical parameters were used to categorize cancer into low, medium and high-risk. Clinical data were retrieved from the patients’ digital hospital files and by telephone interview, if needed.
All reported investigations were conducted in accordance with the Helsinki declaration and with our national regulations. The institutional review board approved this retrospective study under the number EA1/354/14. All patients gave written informed consent and agreed to evaluation and publication of their anonymized data.
Planar (99m)Tc-DPD BS was performed 1 h and 2.5 h to 3 h after intravenous (IV) injection of approx. 700 Megabecquerel (MBq) (99m)Tc-DPD. Planar images were acquired using a hybrid dual-head Siemens Symbia TruePoint T6 SPECT/CT system (Siemens Medical Solutions, Erlangen, Germany) with a scan speed of 10 cm/min and a 256 × 1024 matrix. When diagnostic confidence in interpreting planar scans was poor, an additional SPECT/CT scan was obtained.
Technical parameters for the gamma camera were as follows: crystal thickness of 9.5 mm, 53.3 cm axial by 38.7 cm diameter SPECT field of view (FOV) and 6-slice CT scanner. Acquisition was performed in a 128 × 128 matrix, zoom 1.0, pixel size of 4.8 mm, and non-circular orbit (auto contour) with 30 projections over 180° (continuous acquisition). A low-energy high-resolution collimator (LEHR) was used with dual-energy windowing and a lower scatter energy window (15% width) for scatter estimation.
PET/CT was performed with [68Ga]Ga-PSMA-HBED-CC (=68Ga- DKFZ-PSMA-11). Tracer synthesis is described in detail in . The PET scan was initiated approx. 1 h after IV injection of approx. 120 MBq [68Ga]Ga-PSMA-HBED-CC. Ten to twelve bed positions with 90 sec per position were scanned with a 144 × 144 acquisition matrix and 576 mm FOV. The whole-body CT examination was performed immediately before the PET acquisition was started. Generally, the whole-body scan was performed from mid-thigh to the skull base.
The CT scan included in the PET/CT examination was acquired without contrast medium or, if indicated for whole-body staging, with 80 ml to 120 ml of Ultravist 370 (Bayer HealthCare Pharmaceuticals, Berlin, Germany) injected as a bolus at a rate of 2-3 ml/s. A venous contrast phase (70 sec after contrast agent administration) was acquired with 16 mm × 1.5 mm reconstructed slice thickness and used for attenuation correction of the PET dataset. CT examinations were performed on a Philips Gemini TF ToF 16 scanner (Philips Medical Systems, Cleveland, USA) using the following technical parameters: 120 kV; 100 mAs to 250 mAs with dose modulation (D-DOM).
One physician of nuclear medicine and radiology (>9 years of experience) and one radiologist specialized in hybrid imaging (>5 years of experience) assessed whole-body CT scans for the presence of bone metastases in consensus. All bone metastases were classified morphologically into osteoblastic, osteolytic, or mixed types as well as in terms of their localization using the CT data.
Attenuation-corrected [68Ga] PSMA HBED-CC PET images and planar (99m)Tc-DPD BS were interpreted for the presence of bone metastases with the readers blinded to the results of CT and other imaging modalities. Only lesions evaluable in all imaging modalities were interpreted.
Criteria for bone metastasis in PET and BS were a pathological SUV max and lesions in the CT were rated by their morphology. When tracer uptake was inconclusive in terms of lesion characterization, the readers additionally used CT and SPECT/CT datasets for assessment. Size of bone metastases and Hounsfield units (HU) were measured by one observer using the CT data. CT and clinical parameters as described above served as standard of references.
Demographic data are given as mean, range and standard deviation (SD). Lesion-based results were compared for PET and BS using the McNemar test. Patient-based results were compared using the twosided Fisher’s exact test for unpaired samples. Statistical significance was accepted at p<0.05. Statistical analysis was performed using the SPSS software package (SPSS 20.0.0.; SPSS Corporation, Chicago, USA).
Nineteen patients (mean age: 62.3 years, range: 53.5-76.1) with histologically proven prostate cancer met our criteria and were included in the analysis. Most patients underwent prostatectomy, some with additional lymphadenectomy and/or additional local radiotherapy. Table 1 summarizes the treatment histories of all patients prior to [68Ga]Ga-PSMA-HBED-CC PET/CT and (99m)Tc-DPD BS.
|Pat. No.||1st line therapy||2nd line therapy||3rd line therapy||PSA (ng/ml)||Therapy after|
|1||RRP1, LNE2, RTX3||0.7||Cyberknife for bone metastases|
|2||RRP, LNE||Salvage RTX||Bicalutamide*||1.2||none|
|6||RRP, LNE, RTX||0.5||AHT|
|8||RRP, LNE||9.6||RTX for bone metastases|
|9||GnRH agonist*, Bicalutamide||n.a.||RTX for bone metastases|
|12||GnRH agonist*, Zoledronic acid||n.a.||none|
|13||RRP, LNE||RTX||3.5||AHT, RTX for bone metastases|
|14||RRP, Zoledronic acid||1.4||None|
|15||RRP, LNE||59.8||Cyberknife for bone metastases|
|17||RRP, LNE, Bicalutamide||RTX||9.5||none|
|19||RRP||2.5||AHT, Zoledronic acid|
1radical retropubic prostatectomy; 2lymphadenectomy; 3radiotherapy; 4anti-hormonal treatments. *Systemic regimens (e.g., chemotherapy, antihormone therapy, alpharadin therapy)
Table 1: Treatment histories of prostate cancer patients prior to imaging.
All patients with histopathological examination had medium- to high-risk prostate cancer, except one patient with low-risk prostate cancer, with a mean Gleason score of 8.1 (median 8.0; range: 6-10). The mean initial PSA level at diagnosis was 27.2 ng/ml ± 24.4 SD (range: 3.4 ng/ml to 80 ng/ml). Mean PSA level prior imaging by [68Ga]Ga- PSMA-HBED-CC PET/CT and (99m)Tc-DPD BS was 13.8 ng/ml ± 20.3 SD (range: 0.5 ng/ml to 60 ng/ml). The CT scans revealed a total of 162 bone lesions. Fifty-eight lesions were rated as benign bone lesions (e.g. osteoids) and 104 lesions were classified as bone metastases.
Of the latter, 93.3% (n=97) were classified as osteoblastic, 4.6% (n=5) as osteolytic, and 1.9% (n=2) as mixed metastases (Figure 1). Applying PET scans, 47 lesions were rated as malignant as compared to 35 lesions in BS. Compared to the CT findings, PET had a sensitivity of 45.6%, specificity of 86.4%, accuracy of 60.5%, PPV of 85.5%, and NPV of 47.7% for the detection of bone metastases (Figure 2).
BS with (99m)Tc-DPD had 34% sensitivity, 81.4% specificity, 51.2% accuracy, PPV of 76.1%, and NPV of 41.4%, compared to CT (Figure 2). A subtype analysis of morphologic types of bone metastases for each imaging modality was not performed due to the uneven distribution of lesion types.
Sensitivity for the detection of bone metastases was significantly higher for [68Ga]Ga-PSMA-HBED-CC PET than for (99m)Tc-DPD BS (p<0.002), while specificity was not significantly different. A total of 49 bone metastases in 4 patients were false-negative based on [68Ga]Ga- PSMA-HBED-CC PET versus 58 metastases in 6 patients based on (99m)Tc-DPD BS. All false-negative bone metastases were classified as osteoblastic metastases with small diameter (<10 mm) and high attenuation (670 HU to 1052 HU) in CT. On a per patient-analysis of sensitivity and specificity of PET and BS compared to CT-findings, we found a sensitivity of 92% and specificity of 83% for PET, respective 82% and 83% for BS. Patient-based analysis of false-negative bone metastases using the two-sided Fisher’s exact test revealed a trend towards a lower rate of false-negative radiotracer uptake rates in [68Ga]Ga-PSMA-HBED-CC PET compared to (99m)Tc-DPD BS (p=0.013). All patients with false-negative bone metastases in PET and/or BS had a history of systemic treatment. Overall, 9 patients of the study population underwent systemic treatment prior to imaging (antihormone therapy, chemotherapy, etc.) (Table 1). Patient-based analysis of the potential gain in diagnostic information achieved with either PET or BS showed the following results: 13 true-positive bone metastases in 6 patients detected by PET (Figure 3) compared to one additional lesion detected by BS. Conversely, (99m)Tc-DPD BS had a diagnostic gain of one lesion in one patient compared with PET (Figure 4). In this case, the lesion in the 10th rib was missed by PET imaging because abnormal tracer uptake of the lesion was obscured by normal tracer uptake in the liver. After [68Ga]Ga-PSMA-HBED-CC PET the therapeutic approach was changed in 58% (11/19) patients. Of these, 6 (54.5%) received additional systemic treatment, 4 (36.4%) had a RT of bone metastases and 1 (9.1%) was treated by a combination AHT and RT of the bone (Table 1). In [68Ga]Ga-PSMA-HBED-CC PET/CT there were found additionally 4 primary prostatic lesions in 4 patients and 19 lymph node metastases in 7 patients. There were no distant visceral metastases in any imaging modality.
Bone metastases are very common in patients with prostate cancer . An autopsy study of men who died of prostate cancer suggests that bone metastases are present in up to 80% of cases . The strong tendency of prostate cancer to spread to the skeletal system has been attributed to different mechanisms including reduced cell adhesion [14-16]. In addition, a complex tumor/microenvironment interaction appears to be involved. Many morphologic and metabolic imaging modalities have been studied for the detection of bone metastases in men with prostate cancer, but no consistent picture has emerged [17-20]. To our knowledge, this is the first study to provide data on the sensitivity, specificity, accuracy, and positive/negative predictive value of [68Ga]Ga-PSMA-HBED-CC PET for the detection of bone metastases in prostate cancer.
Using CT and clinical parameters as standard of reference, our lesion-based analysis yielded 45.6% sensitivity, 86.4% specificity, 60.5% accuracy, 85.5% PPV and 47.7% NPV for [68Ga]Ga-PSMA-HBED-CC PET. Our findings suggest that PET has significantly higher sensitivity than BS although (99m)Tc-DPD BS is generally known for high sensitivity in the detection of bone metastases. The low sensitivity of 34% for the detection of bone metastases by BS in our study might be attributable to the deficiency of our standard of reference.
In general, 18F-FDG PET is considered to be superior to (99m)Tc- DPD BS in the detection of general bone metastases detection . Nevertheless, the role of 18F-FDG PET/CT in the detection of bone metastases from prostate cancer remains unclear and seems to be dependent on histology, tumor progression, morphology, and previous treatment of bone metastases . Tumor progression and aggressiveness of prostate cancer defined by rising PSA levels result in a sensitivity rate of 77% for the detection of bone metastases by 18FFDG PET . Sclerotic bone metastases are associated with less 18FFDG uptake than other morphological types of bone metastases . Similar results have been reported for bone metastases from other primaries imaged by 18F-choline PET/CT [20,25]. The reasons for reduced uptake appear to be different for 18F-FDG and 18F-choline . Our results also suggest reduced tracer uptake kinetics in completely sclerotic bone metastases using the new radiotracer [68Ga]Ga-PSMA-HBED-CC.
It has already been discussed in literature whether the detection rate is affected by functional and morphologic changes in bone metastases following systemic treatment [22,25]. For instance, Fogelman and coworkers describe bone metastases becoming sclerotic after antihormonal therapy due to an intense osteoblastic healing process .
As shown in our study population, most prostate cancer bone metastases are of the osteoblastic to sclerotic morphologic type anyway. These lesions are difficult to differentiate from a treatment response of bone metastases. On CT scans-our imaging standard of reference-osteoblastic lesions can hardly be classified into active malignant lesions and regressive lesions with residual bone transformation. This problem is amplified for metabolic imaging as shown in the study of Israel et al.  using 18F-FDG for the detection of bone metastases.
The limitations of CT as a standard of reference for treated bone metastases is certainly an important reason for the 49 false-negative bone metastases in [68Ga]Ga-PSMA-HBED-CC PET. Another possible explanation is the fact that all metastases missed in our study by [68Ga]Ga-PSMA-HBED-CC PET were small (<10 mm diameter) and completely sclerotic (670 HU to 1052 HU) in CT. Beheshti et al.  found that sclerotic bone metastases with attenuation above 825 HU measured by CT did not take up 18F-choline. Ruf et al.  described another influence on tracer uptake kinetics in lesions: the maximum standard uptake value (SUVmax) was found to be distorted by partial volume effects (PVE) for lesions <21.4 mm in mathematical analysis and for lesions <25 mm in visual analysis. Hence, the dependence of lesion size to PVE seems to be much larger than assumed in the literature before (<15 mm) .
Limitations of this study are the retrospective study design, the small sample size, and the standard of reference. Nevertheless, the retrospective study design allowed us to obtain initial results with no additional radiation exposure or discomfort from additional examinations. Large sample sizes in future studies would be interesting to further stratify the influence of bone metastasis morphology on imaging by [68Ga]-HBED-CC PSMA PET. As already mentioned, it is complicated to define an optimal standard of reference for bone metastases. All known imaging modalities are limited in the differentiation of treated bone metastases. A biopsy of all bone lesions with histopathological validation is ethically not feasible and not very practical. It would be a heavy interference into the patient’s integrity. An optimal approach is currently not available. A partial solution would be the increased use of supplemental clinical parameters such as PSA to assess treatment response. Therefore, a reasonable amount of bone metastases should be monitored with PSA samples taken at regular intervals in a selected cohort of patients with clear PSA level to tumor burden correlation. The statistical results are reliable regarding the difference in sensitivities between both methods. Despite the small patient number there was enough statistical power for this result of the study. The exact values of sensitivity, specificity, PPV and NPV deriving from this study have to be considered very critically. For defining exact values the standard of reference is too weak and the patient number too small.
Despite these limitations, our study of a new promising radiotracer for the detection of bone metastases in prostate cancer has shown a clear benefit of [68Ga]Ga-PSMA-HBED-CC PET. The new radiotracer clearly detected additional bone metastases in comparison to the current standard in nuclear medicine, (99m)Tc-DPD BS. Further studies should evaluate the potential use in evaluating the response to systematic treatment. Therefore, the problem of the standard of reference of verifying bone metastases after systemic has to be solved in a practicable way.
V. Schreiter is participant in the Charité Clinical Scientist Program funded by the Charité Universitätsmedizin Berlin and the Berlin Institute of Health.
Make the best use of Scientific Research and information from our 700 + peer reviewed, Open Access Journals