Abstract
The purpose of this study was to evaluate the accuracy of Glu-NH-CO-NH-Lys-(Ahx)-[68Ga(HBED-CC)] PET compared with morphologic imaging for the assessment of lymph node metastases (LNM) in patients with recurrent prostate cancer. Methods: Forty-eight patients (median age, 71 y; interquartile range, 66–74 y) with biochemical recurrence (median prostate-specific antigen level, 1.31 ng/mL; interquartile range, 0.75–2.55 ng/mL) who underwent 68Ga–prostate-specific membrane antigen (PSMA) HBED-CC PET/CT or PET/MR and salvage lymphadenectomy were retrospectively included. Institutional review board approval and written informed consent were obtained from all patients for the purpose of anonymized evaluation and publication of their data. Standardized predefined lymph node (LN) template fields (n = 10) were evaluated in 68Ga-PSMA HBED-CC PET and morphologic imaging for the presence of LNM using a 5-point-scale. Additionally, SUVmean/max and size of suspicious lesions were determined. Specificity of 68Ga-PSMA HBED-CC PET imaging for PET-positive LNs was defined by comparison to histopathology. The diagnostic accuracy of 68Ga-PSMA HBED-CC PET compared with morphologic imaging alone was assessed, and areas under the receiver-operating-characteristic curves are presented. Results: LNM were found histologically in 68 of 179 resected anatomic LN fields (38.0%). The specificity of 68Ga-PSMA HBED-CC PET and morphologic imaging was 97.3% and 99.1%, respectively. However, 68Ga-PSMA HBED-CC PET detected LNM in 53 of 68 histopathologically proven metastatic LN fields (77.9%) whereas morphologic imaging was positive in only 18 of 67 (26.9%). 68Ga-PSMA HBED-CC PET imaging performed significantly superior to morphologic imaging for detection of LNM (difference in the areas under the receiver-operating-characteristic curves, 0.139; 95% confidence interval, 0.063–0.214; P < 0.001). In 68Ga-PSMA HBED-CC PET, the mean size of PET-positive LN measured by CT or MRI was 8.3 ± 4.3 mm (range, 4–25 mm), and LNs, which were suspicious only in CT or MRI, presented with a mean size of 13.0 ± 4.9 mm (range, 8–25 mm). Conclusion: 68Ga-PSMA HBED-CC PET imaging is a promising method for early detection of LNM in patients with biochemical recurrent prostate cancer. It is more accurate than morphologic imaging and thus might represent a valuable tool for guiding salvage lymphadenectomy.
- positron emission tomography
- magnetic resonance imaging
- computed tomography
- hybrid imaging
- 68Ga-PSMA HBED-CC
- prostate cancer
- salvage lymphadenectomy
Biochemical recurrence after definitive treatment of prostate cancer (PC) occurs in up to 50% of cases (1–3). Serum prostate-specific antigen (PSA) testing is routinely used to detect PC recurrence. However, PSA testing does not allow for distinguishing between local, regional, or distant recurrence. Because the use of lesion-targeted management options including salvage lymphadenectomy or targeted irradiation therapy in patients with minimal or localized metastatic tumor load are increasingly favored, a correct identification of recurrent disease is essential for further treatment planning (4,5).
At present, accuracy of currently widely available imaging modalities is suboptimal in identifying lymph node metastases (LNM) in patients with recurrent PC. Promising results have been published for the use of ultra-small superparamagnetic iron oxide particles (6). However, so far they are not routinely available. The limited role of conventional morphologic imaging techniques such as CT or MRI in the detection of local recurrence and LNM in patients with recurrent disease is well known, with sensitivities reaching only 13%–40% (7,8). Further, the performance of CT and MR is not significantly different from what could be demonstrated in a large meta-analysis including 24 studies. Here, the pooled sensitivity was 42% and 39% and the pooled specificity was 82% and 82% for CT and MR, respectively (7).
These limitations might be overcome by molecular imaging techniques (9). For example, 11C-choline PET/CT can detect LNM in recurrent PC sensitively, however, it is not able to detect small-volume lesions (10). Furthermore, the detection rate of 11C-choline PET/CT at PSA levels below 1 ng/mL is limited, with only 36% of positive findings, and thus not regarded as appropriate for the detection of early recurrent disease (11). Specificity, on the other hand, is regarded as satisfactory, with a pooled value of 89% being reported in a recent meta-analysis summarizing 29 studies evaluating 11C-choline PET (12).
Recently, a 68Ga-labeled ligand of the prostate-specific membrane antigen (PSMA) Glu-NH-CO-NH-Lys-(Ahx)-[68Ga(HBED-CC)], often termed 68Ga-PSMA HBED-CC or 68Ga-PSMA-11, has been introduced in PET imaging for PC, with highly promising results (13–15). Because of increased expression of PSMA in PC and its metastases, 68Ga-PSMA HBED-CC PET was reported to exhibit a favorable lesion-to-background ratio with high detection rates (16). Further, recent studies evaluating 68Ga-PSMA HBED-CC PET showed substantially higher detection rates in patients with recurrent PC than reported for other imaging modalities, especially at low PSA values (<0.5 ng/mL) (17,18). However, in both studies histopathology as a gold standard was available only in a minority of cases. One recent large study performing preoperative lymph node (LN) staging using 68Ga-PSMA HBED-CC PET in primary PC, however, showed a high accuracy of 68Ga-PSMA HBED-CC PET, with a superb specificity of 99.1% (19).
Thus, the aim of our study was to evaluate the accuracy of 68Ga-PSMA HBED-CC PET compared with morphologic imaging for the detection of LNM in patients with recurrent PC after primary treatment validated by histopathology.
MATERIALS AND METHODS
Patient Population
All patients (November 2012 to June 2015) with biochemical recurrence (PSA > 0.2 ng/mL) who underwent salvage lymphadenectomy after 68Ga-PSMA HBED-CC PET/CT or PET/MR were extracted from the institution’s database. Thirteen patients in the present evaluation were partially analyzed in a previous study (18). All patients were treated initially for localized PC by radical prostatectomy and pelvic lymphadenectomy (n = 45; 93.8%) or radiation therapy (n = 3; 6.2%) with curative intent. None of the patients had received androgen-deprivation therapy within the last 6 mo before the examination.
All patients had undergone open salvage lymphadenectomy, with the extent of lymphadenectomy being dependent on preoperative imaging results, patients’ risk factors, and intraoperative findings (in some patients no tissue was left after previous lymphadenectomy or extensive fibrotic changes precluded resection of further fields not suspicious on preoperative imaging). During lymphadenectomy, tissue from each resected LN field was sent separately for histologic evaluation. Surgical specimens were processed according to standard pathology protocols.
The retrospective study was approved by the Ethics Committee of the Technical University Munich (permit 5665/13), and written informed consent was obtained from all patients for the purpose of anonymized evaluation and publication of their data. All reported investigations were conducted in accordance with the Helsinki Declaration and with national regulations.
68Ga-Labeled HBED-CC PET Application and Imaging
68Ga-labeled HBED-CC was produced as previously described (20). The 68Ga-PSMA HBED-CC-ligand complex solution (1.8–2.2 MBq per kg of body weight) was applied intravenously (mean, 154 MBq; interquartile range [IQR], 133–179 MBq), and PET acquisition was started at a mean time of 57 min (IQR, 49–63 min) after tracer injection. Thirty-one patients underwent 68Ga-PSMA HBED-CC PET/CT on a Biograph mCT scanner (Siemens Medical Solutions), and 17 patients underwent 68Ga-PSMA HBED-CC PET/MR on a fully integrated whole-body hybrid PET/MR system (Biograph mMR; Siemens Healthcare). PET/CT included a diagnostic CT (240 mAS, 120 kV, 5-mm slice thickness) in the portal venous phase (80 s after injection of the intravenous contrast agent). For the assessment of the presence of LNM in PET/MR, an axial T2 turbo spin echo sequence of the pelvis and an axial T2 half Fourier acquisition single-shot turbo spin echo of the whole body was used. Details on the complete protocol in PET/MR and PET/CT have been described previously (18,21). For PET/CT, the emission time was 4 min per bed position, whereas for PET/MR the emission time for the trunk was 5 min per bed position. All PET images were acquired in 3-dimensional mode and reconstructed by an attenuation-weighted ordered-subsets expectation maximization algorithm (4 iterations, 8 subsets) followed by a postreconstruction smoothing gaussian filter (5 mm in full width at half maximum).
Image Analysis and Quantitative Assessment
The entire rating procedure was performed using a dedicated workstation and software (Syngo MMWP and Syngo TrueD; Siemens Medical Solutions). First, images were analyzed by 1 double-trained board-certified radiologist and nuclear medicine physician with 9 y of training in oncologic imaging masked to the patient history and the extent of performed lymphadenectomy. Thus, in every patient in total and in a standardized way the following 10 anatomic fields were evaluated: right/left common iliac vessel, right/left internal iliac vessel, right/left external iliac vessel, right/left obturator fossa, presacral and other regions, for example, paraaortic LNs.
In PET, every anatomic field was rated on a 5-point-scale using the following criteria: 1, tumor manifestation (intense, focal 68Ga-PSMA HBED-CC uptake higher than liver); 2, probably tumor manifestation (68Ga-PSMA HBED-CC uptake clearly higher than background in vessels but not higher than liver); 3, equivocal (68Ga-PSMA HBED-CC faint uptake between background in muscle and vessels); 4, probably benign (68Ga-PSMA HBED-CC uptake as faint as background, e.g., equally to adjacent muscle); and 5, benign (no 68Ga-PSMA HBED-CC uptake). Hereby, anatomic images were used only for anatomic allocation of a suspicious focal increased uptake to the corresponding LN field.
In a second step, with a time interval of at least 4 wk, the diagnostic CT or MR dataset was analyzed for LNM using morphologic criteria (size, short axis; shape; and regional grouping) (7). Each anatomic field was rated on a 5-point-scale using the following criteria: 1, tumor manifestation (short-axis diameter > 10 mm); 2, probably tumor manifestation (short-axis diameter, 8–10 mm; round configuration; or regional grouping); 3, equivocal (short-axis diameter, 8–10 mm; oval configuration; and no regional grouping); 4, probably benign (short-axis diameter < 8 mm); and 5, benign (short-axis diameter < 5 mm).
For quantitative analysis, a resident with 4 y of experience in oncologic imaging noted the highest SUV in each suspicious LN field. If multiple LNs were suspicious in 1 anatomic field, only the one with the highest uptake was analyzed. To calculate SUVs, an isocontour volume of interest including all voxels above 50% of the maximum was created, covering the whole lesion volume. Within all volumes of interest, mean and maximum SUVs were measured. In addition, size (maximum short diameter in mm) of PET-positive lesions and suspicious LNs in the morphologic MR or CT dataset were measured. If multiple LNs were suspicious in 1 anatomic field, only the largest one was measured.
Statistical Analysis
All statistical analyses were performed using MedCalc software (version 13.2.0, 2014; MedCalc). Histopathologic findings of resected LNs were compared with the results of morphologic imaging (MR or CT) alone and 68Ga-PSMA HBED-CC PET imaging in a field-based manner. This approach was chosen, because exact tracking of a single LN is impossible without further guidance, for example, by radiolabeled colloid or PSMA ligands (especially in the case of normal-sized LNs because of disorientation during resection and histologic evaluation).
Overall diagnostic accuracy using field-based and patient-based data was assessed conducting receiver-operating characteristics (ROC) analyses. For both modalities (68Ga-PSMA HBED-CC PET, morphologic imaging), ROC curves were calculated, and the areas under the ROC curves (AUCs) are presented. For estimation of 95% confidence intervals (CIs) and for comparison of the modalities, the nonparametric method proposed by Obuchowski was used to account for the correlation of multiple observations within the same patients (22).
After dichotomization of the qualitative assessments (≤2, positive test; >2, negative test), sensitivities, specificities, and accuracies were estimated for both modalities. For field-based analysis, CIs were derived from logistic generalized estimating equation models (23). For estimation of sensitivities and corresponding CIs, an intercept-only logistic generalized estimating equation model was fitted to the data. The result of the dichotomized test was used as a dependent variable, and only patients with a positive histopathologic result were considered (24). To derive estimates for the specificities, a variable indicating whether a negative test result was observed was used as a dependent variable, and only patients with a negative histopathologic result were included. Accuracy was estimated in an intercept-only model with a dependent variable that indicated whether the test result and the result of the histopathologic assessment agreed. For all generalized estimating equation models, an independent correlation structure was assumed.
RESULTS
Patient Characteristics
Forty-eight patients were included in the analysis (median age, 71 y; IQR, 66–74 y). Median time between imaging and salvage lymphadenectomy was 47 d (IQR, 28–67 d). In 4 of those 48 patients, salvage lymphadenectomy was performed in an external hospital. However, in all patients detailed surgical and histologic reports were available. A median of 3 LN fields (range, 1–9) per patient were resected during salvage lymphadenectomy. The median primary Gleason score was 7 (IQR, 7–9), and median PSA level at time of PET imaging was 1.31 ng/mL (IQR, 0.75–2.55 ng/mL).
Histopathologic Results
In total, LNM were found in 68 of 179 resected anatomic LN fields (38.0%) after histopathologic workup. The metastatic LNs were located in the following regions: right common iliac vessel (n = 6), left common iliac vessel (n = 7), right internal iliac vessel (n = 4), left internal iliac vessel (n = 9), right external iliac vessel (n = 6), left external iliac vessel (n = 9), right obturator fossa (n = 7), left obturator fossa (n = 3), presacral (n = 9) or other regions, for example, paraaortic (n = 8). Details on the operated LN fields and the percentage of LNM in each LN field are shown in Table 1.
Total Number and Location of Resected LN Fields and Histopathologically Proven LNM
Imaging Findings
On a field-based analysis, specificity of 68Ga-PSMA HBED-CC PET was 97.3% (95% CI, 91.9%–99.1%) and 99.1% (95% CI, 93.9%–99.9%) for morphologic imaging, respectively. However, 68Ga-PSMA HBED-CC PET detected LNM in 53 of 68 histopathologically proven metastatic LN fields (77.9%; 95% CI, 65.1%–87.0%), whereas morphologic imaging was positive in only 18 of 67 (26.9%; 95% CI, 16.2%–41.1%). The positive predictive value was 94.6% for PET and 94.7% for morphologic imaging. Diagnostic accuracy of 68Ga-PSMA HBED-CC PET imaging was 89.9% (95% CI, 82.5%–94.4%) and for morphologic imaging 71.9% (95% CI, 61.8%–80.2%), respectively (Table 2).
Results and Diagnostic Values for Detection of LNM on Field-Based Analysis
In detail, 111 fields were free of tumor invasion after histopathologic evaluation, with 108 of them being correctly identified as negative with PET and 110 of them being correctly identified as negative with morphologic imaging. One metastatic LN field could not be evaluated in CT because of clip material located in the pelvis. Three fields in 3 patients were classified as suspicious in PET with no correlation in histopathology (false-positive), whereas in morphologic imaging 1 field was judged false-positive. Fifteen fields (in 9/48 patients) were false-negative in PET, whereas in morphologic imaging 49 fields revealed a false-negative finding.
AUC for 68Ga-PSMA HBED-CC PET imaging and morphologic imaging was 0.878 (95% CI, 0.819–0.937) and 0.738 (95% CI, 0.660–0.816), respectively (Fig. 1). 68Ga-PSMA HBED-CC PET imaging performed significantly better than morphologic imaging on a field-based analysis for detection of LNM (difference in AUCs, 0.139; 95% CI, 0.063–0.214; P < 0.001).
ROC curves for 68Ga-PSMA HBED-CC PET imaging (red) and morphologic imaging (CT/MR) (blue) for detection of LNM on a field-based analysis (P values: comparison to AUC = 0.5).
On a patient-based analysis, sensitivity, specificity, and accuracy of 68Ga-PSMA HBED-CC PET and morphologic imaging were 100%, 50%, and 93.8% versus 34.1%, 83.3%, and 40.4%, respectively (Table 3). The AUC for 68Ga-PSMA HBED-CC PET and morphologic imaging was 0.732 (95% CI, 0.583–0.850) and 0.665 (95% CI, 0.512–0.796), respectively. There was no significant difference between 68Ga-PSMA HBED-CC PET and morphologic imaging for the detection of LNM on a patient-based analysis (difference in AUCs, 0.0671; 95% CI, −0.241 to 0.376; P = 0.6701).
Results and Diagnostic Values for Detection of LNM Using Patient-Based Analysis
In 68Ga-PSMA HBED-CC PET, the mean SUVmean of suspicious LN was 9.0 ± 7.5 (range, 1.8–36.0), and mean SUVmax was 12.7 ± 10.8 (range, 2.4–51.0). Corresponding mean lesion size of 68Ga-PSMA HBED-CC PET–suspicious LN in CT or MRI was 8.3 ± 4.3 mm (range, 4–25 mm). The mean size of LN suspicious only in morphologic imaging was 13.0 ± 4.9 mm (range, 8–25 mm). Histopathologic workup of false-negative LN fields in 68Ga-PSMA HBED-CC PET revealed a mean lesion size of 4.7 ± 3.4 mm (range, 0.5–11 mm). Representative examples of correctly classified LNM by 68Ga-PSMA HBED-CC PET/CT and representative false-negative findings are shown in Figures 2 and 3.
Example of a 55-y-old patient with biochemical recurrence after radical prostatectomy (Gleason score, 7; PSA level at PET examination, 0.77 ng/mL) and a correctly classified LNM by 68Ga-PSMA HBED-CC PET imaging: a 6-mm LN is visible in right obturator fossa on CT imaging (A, arrow) that shows intense, focal and thus suspicious tracer uptake on 68Ga-PSMA HBED-CC PET (B) and PET/CT fusion imaging (C). Salvage lymphadenectomy with histologic evaluation confirmed a single LNM.
68Ga-PSMA HBED-CC PET imaging of a 55-y-old patient with recurrent PC (Gleason score, 8; PSA level at PET examination, 5.1 ng/mL). Patient presented with 2 correctly classified LNM (5 and 6 mm) behind left common iliac artery with intense, focal uptake on 68Ga-PSMA HBED-CC PET/CT fusion (B, arrow) and on maximum-intensity-projection images (C, arrow). However, paraaortal/interaortocaval LN field was negative on 68Ga-PSMA HBED-CC PET, showing no increased uptake either on maximum-intensity-projection images (C) or on axial 68Ga-PSMA HBED-CC PET/CT (E). In conventional CT imaging, only small, unsuspicious LNs could be found (D, arrow). Histopathology revealed overall 9 further LNM in this anatomic field (paraaortic).
DISCUSSION
Several first studies have demonstrated the high clinical value of 68Ga-PSMA HBED-CC PET/CT in patients with recurrent PC after primary treatment (17,18,25,26). Besides reporting first on the superior performance of 68Ga-PSMA HBED-CC PET/CT mainly in recurrent PC, Afshar-Oromieh et al. also demonstrated a high specificity of 68Ga-PSMA HBED-CC PET in a subgroup of patients with histologic verification (17). In addition, so far only 2 further studies reported on the use of 68Ga-PSMA HBED-CC PET for LN restaging with histologic evaluation of all patients (25,26). However, both studies encompass not only patients scheduled for restaging after biochemical failure but also patients undergoing primary radical prostatectomy including lymphadenectomy. Thus, to the best of our knowledge, this is the first report presenting data on the performance of 68Ga-PSMA HBED-CC PET in a homogeneous patient collective with biochemical recurrence after primary definite treatment using histopathology as a standard of reference for all cases.
Our study demonstrates that 68Ga-PSMA HBED-CC PET is a highly accurate tool in detecting LNM in recurrent PC, with a sensitivity, specificity, and accuracy of 77.9%, 97.3%, and 89.9%, respectively. Morphologic imaging was considerably less sensitive (26.9%) and accurate (71.9%), with an equal performance regarding specificity (99.1%). The latter is most likely explained by the use of strict criteria for morphologic assessment of LNM with a short-axis diameter of greater than 10 mm for rating 1 (tumor manifestation) and a short diameter of 8–10 mm and additional round configuration or regional grouping for rating 2 (probably tumor manifestation).
The recent study by Herlemann et al. evaluating 68Ga-PSMA HBED-CC PET in a mixed group of primary PC (n = 20) and recurrent PC (n = 14) showed a sensitivity of 84% and specificity of 82% for PET for all patients as well as a clear superiority in diagnostic accuracy of PET versus CT, with 77% versus 65% for the subgroup of patients with salvage lymphadenectomy (28). Taken together, both the data presented by this work and that of Herlemann et al. strengthen the value of 68Ga-PSMA HBED-CC PET for preoperative staging. The other study by Hijazi et al. investigated 68Ga-PSMA HBED-CC PET/CT also in a mixed group of patients with biochemical recurrence and primary PC, judging 8 different LN fields in imaging (25). The authors reported a high specificity of 99% similar to our data. However, they found a considerably higher sensitivity of 94% than found in our data stating that only 78% of histopathologically positive LN fields were detected. As in the work by Hijazi et al., extended LN dissection was available in only 17 of 35 patients, thus the true extent of LNM could have be underestimated.
In addition, the presented data are well comparable to a large series of patients with intermediate- to high-risk PC in the setting of primary LN staging using 68Ga-PSMA HBED-CC PET. Maurer et al. reported a sensitivity, specificity, and accuracy for 68Ga-PSMA HBED-CC PET of 78.2%, 99.1%, and 95.7%, respectively (19). Further, 68Ga-PSMA HBED-CC PET performed significantly superior to morphologic imaging alone (difference in AUCs, 0.174; 95% CI, 0.114–0.233; P < 0.001), which is also similar to the results of this study (difference in AUCs, 0.139; 95% CI, 0.063–0.214; P < 0.001).
When our results are compared with studies of other PET agents, Rinnab et al., evaluating 11C-choline PET/CT in patients with recurrent PC at a PSA level of less than 2.5 ng/mL, reported a sensitivity of 91% and specificity of 50% in a patient-based analysis (27). These values are comparable to the results for 68Ga-PSMA HBED-CC on a patient-based level in our study. However, the median PSA level of patients examined by Rinnab et al. was 2.42 ng/mL, compared with 1.31 ng/mL in our study. In addition, caution is warranted because the high sensitivity in both studies when reporting data on a patient basis is strongly influenced by imaging-derived preselection. The lack of a statistical significance for our data when presented on a patient basis comparing 68Ga-PSMA HBED-CC PET versus morphologic imaging is most likely due to the considerably smaller sample size as opposed to a field-based approach.
Using a site- or lesion-based analysis, Scattoni et al. reported a sensitivity, specificity, positive predictive value, negative predictive value, and accuracy of 64%, 90%, 86%, 72%, and 77%, respectively, for 11C-choline PET/CT (10). In contrast, another study identifying histologically confirmed LNM with 18F-fluoroethylcholine PET/CT reported sensitivity, specificity, positive predictive value, and negative predictive value of 39.7%, 95.8%, 75.7%, and 83% in a site-based analysis, respectively (28). However, in this study only 3 different sites were evaluated separately, compared with up to 10 LN fields in our study collective. In total, compared with data from literature, our results suggest that PET using 68Ga-PSMA HBED-CC seems to be more accurate than choline-based PET in the detection of LNM in recurrent PC. An improved tumor-to-background ratio facilitating the detection of suspicious lesions using 68Ga-PSMA HBED-CC PET compared with 18F-choline has already been shown (16). Indeed, we also found high SUVs in most histopathologically positive LN fields.
In our study, for 68Ga-PSMA HBED-CC PET 15 false-negative fields in 9 patients were found. Interestingly, in 4 of 9 patients the false-negative fields were located adjacent to correctly classified metastatic LN fields. Therefore, 1 explanation might be problems with correct allocation of LN fields in PET and lymphadenectomy. Nevertheless, it is known that despite increased detection using 68Ga-PSMA HBED-CC PET, microscopic lesions might still be missed (19). In addition, 3 false-positive LN fields in 3 patients were found with no histopathologic correlate. It is worth mentioning that 1 patient (PSA level 0.45 ng/mL) presented with an intense, focal 68Ga-PSMA HBED-CC uptake in the left obturator fossa with follow-up 68Ga-PSMA HBED-CC PET/MR 6 mo after operation later revealing an increasing focal uptake of 68Ga-PSMA HBED-CC (Supplemental Fig. 1; supplemental materials are available at http://jnm.snmjournals.org). Therefore, it has to be assumed that during salvage lymphadenectomy this LN was not resected.
From a clinical point of view, the high specificity provided by 68Ga-PSMA HBED-CC highly predicts that a positive 68Ga-PSMA HBED-CC PET scan will result in a histopathologically proven LNM in most patients or LN fields. In addition, compared with morphologic imaging alone, 68Ga-PSMA HBED-CC PET can detect even small LNM with a lesion size below 10 mm (mean size of 68Ga-PSMA HBED-CC–positive LN, 8.3 ± 4.3 vs. 13.0 ± 4.9 mm for suspicious LN on morphologic imaging). Because small lesions are difficult to detect intraoperatively, the intraoperative use of PSMA ligands during, for example, PSMA-radioguided surgery might become an important tool in guiding and improving resection of small prior 68Ga-PSMA HBED-CC–positive LNs. Preliminary and encouraging experience on the use of 111In PSMA imaging and therapy for radioguided surgery has recently been published (29).
There are several limitations to our study. First, despite extensive lymphadenectomy in 11 of 48 patients, in the remaining cases only LN fields adjacent to primary imaging-positive templates were resected. Thus, with regard to potential undetected distant lesions (e.g., LNM in distant templates not being positive in imaging), the positive predictive value is the only descriptive statistical value that is unaffected by this strong selection bias. Therefore, possible distant imaging-negative but histology-positive templates could have been missed, and therefore the sensitivity of the examined imaging technique might be underestimated. In addition, more robust data derived from prospective clinical trials are needed to validate the role of salvage lymphadenectomy guided by 68Ga-PSMA HBED-CC PET for progression-free survival in this highly selected patient population (5). Because salvage lymphadenectomy was performed in an external hospital in 4 of 48 patients, only detailed surgical and histology reports were available in those cases. Because of a limited patient number in each subgroup, we cannot specifically compare the diagnostic value of 68Ga-PSMA HBED-CC PET/MR with PET/CT. However, as indicated by several studies a similar performance for both PET of PET/MR and PET/CT and CT versus MR can be assumed. Regarding the diagnostic performance of PET, a recent study by Freitag et al. comparing 68Ga-PSMA HBED-CC PET/MR with PET/CT in the evaluation of LNM and bone metastases of PC showed that both LNM and bone metastases are accurately and reliably depicted by 68Ga-PSMA HBED-CC PET/MR, with low discordance compared with PET/CT including even PET-positive LNs of normal size (30). In their study, the correlation between SUVs of PET/MRI and PET/CT was linear in LNM and bone metastases. Further, there was no discordance in the number of suspicious LNs based on size criteria between CT and MRI. Additionally, a recent study by Spick et al. summarized data on more than 2,300 patients and concluded that 18F-FDG PET/CT and PET/MRI perform equally well in different cancer types (31).
Furthermore, several studies (using 11C-choline, 68Ga-DOTATOC, or 18F-FDG) have shown that PET datasets acquired by PET/MR and PET/CT are highly comparable (32–34). Regarding the diagnostic performance of CT and MR for detection of LNM, a meta-analysis by Hovels et al. showed that CT and MRI demonstrate an equally poor performance (7). In this study for CT, pooled sensitivity was 0.42 (95% CI, 0.26–0.56), and pooled specificity was 0.82 (95% CI, 0.8–0.83). For MRI, the pooled sensitivity was 0.39 (95% CI, 0.22–0.56), and pooled specificity was 0.82 (95% CI, 0.79–0.83). Further, in a recent work Giesel et al. showed in recurrent PC patients that 68Ga-PSMA HBED-CC PET was able to detect LNM in two thirds of patients who would have been missed using conventional morphologic criteria (35). This result is in line with our observation that in 68Ga-PSMA HBED-CC PET, the mean size of PET-positive LNs was considerably lower (8.3 ± 4.3 mm) than LNs suspicious only in CT or MRI (13.0 ± 4.9 mm).
Finally, for PET/MR it is known that highly specific tracers such as DOTA conjugates and PSMA ligands can cause a reduced signal around the urinary bladder and the kidneys, potentially obscuring the visibility of tumor lesions and underestimating SUVs (33,36). However, when furosemide is used as implemented as routine at our department for PSMA ligand PET imaging, this issue is significantly reduced. Nevertheless, we cannot completely exclude a potential influence on the results of the presented study.
CONCLUSION
This study shows that 68Ga-PSMA HBED-CC PET imaging is a promising method for early detection of LNM in patients with biochemical recurrence of PC after primary treatment. Its efficacy even at low PSA levels and in small-sized LNs is higher than morphologic imaging. Therefore, it could be a promising tool not only for detecting LNM but also for guiding salvage therapy.
DISCLOSURE
The costs of publication of this article were defrayed in part by the payment of page charges. Therefore, and solely to indicate this fact, this article is hereby marked “advertisement” in accordance with 18 USC section 1734. No potential conflict of interest relevant to this article was reported.
Footnotes
↵* Contributed equally to this work.
Published online Jun. 3, 2016.
- © 2016 by the Society of Nuclear Medicine and Molecular Imaging, Inc.
REFERENCES
- Received for publication February 2, 2016.
- Accepted for publication April 27, 2016.
Jump to section
Related Articles
Cited By...
- Validation of 18F-rhPSMA-7 and 18F-rhPSMA-7.3 PET Imaging Results with Histopathology from Salvage Surgery in Patients with Biochemical Recurrence of Prostate Cancer
- Detection Efficacy of 18F-rhPSMA-7.3 PET/CT and Impact on Management in Patients with Biochemical Recurrence of Prostate Cancer After Radical Prostatectomy and Before Potential Salvage Treatment
- Matched-Pair Comparison of 68Ga-PSMA-11 and 18F-rhPSMA-7 PET/CT in Patients with Primary and Biochemical Recurrence of Prostate Cancer: Frequency of Non-Tumor-Related Uptake and Tumor Positivity
- Imaging the Distribution of Gastrin-Releasing Peptide Receptors in Cancer
- 18F-rhPSMA-7 PET for the Detection of Biochemical Recurrence of Prostate Cancer After Radical Prostatectomy
- Evaluation of an Automated Module Synthesis and a Sterile Cold Kit-Based Preparation of 68Ga-PSMA-11 in Patients with Prostate Cancer
- Head-to-Head Comparison of 68Ga-PSMA-11 with 18F-PSMA-1007 PET/CT in Staging Prostate Cancer Using Histopathology and Immunohistochemical Analysis as a Reference Standard
- Can the Injected Dose Be Reduced in 68Ga-PSMA-11 PET/CT While Maintaining High Image Quality for Lesion Detection?
- Matched-Pair Comparison of 68Ga-PSMA-11 PET/CT and 18F-PSMA-1007 PET/CT: Frequency of Pitfalls and Detection Efficacy in Biochemical Recurrence After Radical Prostatectomy
- Detection Rate of 18F-Choline PET/CT and 68Ga-PSMA-HBED-CC PET/CT for Prostate Cancer Lymph Node Metastases with Direct Link from PET to Histopathology: Dependence on the Size of Tumor Deposits in Lymph Nodes
- Metaanalysis of 68Ga-PSMA-11 PET Accuracy for the Detection of Prostate Cancer Validated by Histopathology
- 68Gallium-labelled PSMA-PET/CT as a diagnostic and clinical decision-making tool in Asian prostate cancer patients following prostatectomy
- 68Ga-PSMA-HBED-CC Uptake in Cervical, Celiac, and Sacral Ganglia as an Important Pitfall in Prostate Cancer PET Imaging
- Effects of Fasting on 18F-DCFPyL Uptake in Prostate Cancer Lesions and Tissues with Known High Physiologic Uptake
- Reply: Comparison of 68Ga-PSMA-11 and 18F-Fluciclovine PET/CT in a Case Series of 10 Patients with Prostate Cancer Recurrence: Prospective Trial Is on Its Way
- Prostate Cancer Molecular Imaging Standardized Evaluation (PROMISE): Proposed miTNM Classification for the Interpretation of PSMA-Ligand PET/CT
- Impact of 68Ga-PSMA-11 PET/CT on the Management of Prostate Cancer Patients with Biochemical Recurrence
- 68Ga-PSMA-11 PET/CT Mapping of Prostate Cancer Biochemical Recurrence After Radical Prostatectomy in 270 Patients with a PSA Level of Less Than 1.0 ng/mL: Impact on Salvage Radiotherapy Planning
- PSMA-targeted polyinosine/polycytosine vector induces prostate tumor regression and invokes an antitumor immune response in mice
- Impact of 68Ga-PSMA-11 PET on Management in Patients with Biochemically Recurrent Prostate Cancer
- Seduction by Sensitivity: Reality, Illusion, or Delusion? The Challenge of Assessing Outcomes after PSMA Imaging Selection of Patients for Treatment
- 68Ga-PSMA PET/CT and Volumetric Morphology of PET-Positive Lymph Nodes Stratified by Tumor Differentiation of Prostate Cancer
- Initial Experience with Volumetric 68Ga-PSMA I&T PET/CT for Assessment of Whole-Body Tumor Burden as a Quantitative Imaging Biomarker in Patients with Prostate Cancer
- 18F-DCFPyL PET/CT in the Detection of Prostate Cancer at 60 and 120 Minutes: Detection Rate, Image Quality, Activity Kinetics, and Biodistribution
- PSMA Ligands for PET Imaging of Prostate Cancer
- Detection Efficacy of Hybrid 68Ga-PSMA Ligand PET/CT in Prostate Cancer Patients with Biochemical Recurrence After Primary Radiation Therapy Defined by Phoenix Criteria
- The Clinical Impact of Additional Late PET/CT Imaging with 68Ga-PSMA-11 (HBED-CC) in the Diagnosis of Prostate Cancer
- 68Ga-PSMA Ligand PET/CT-based Radiotherapy for Lymph Node Relapse of Prostate Cancer After Primary Therapy Delays Initiation of Systemic Therapy
- PET-Guided Stereotactic Irradiation of Prostate Cancer Lymph Node Metastases