⁶⁸Ga-PSMA PET/CT and Volumetric Morphology of PET-Positive Lymph Nodes Stratified by Tumor Differentiation of Prostate Cancer

Abstract

⁶⁸Ga-prostate-specific membrane antigen (PSMA) PET/CT is a new method to detect early nodal metastases in patients with biochemical relapse of prostate cancer. In this retrospective investigation, the dimensions, volume, localization, and SUV_max of nodes identified by ⁶⁸Ga-PSMA were correlated to their Gleason score (GS) at diagnosis. Methods: All PET/CT images were acquired 60 ± 10 min after intravenous injection of ⁶⁸Ga-PSMA (mean dose, 176 MBq). In 147 prostate cancer patients (mean age, 68 y; range, 44–87 y) with prostate-specific antigen relapse (mean prostate-specific antigen level, 5 ng/mL; range, 0.25–294 ng/mL), 362 ⁶⁸Ga-PSMA PET–positive lymph nodes (LNs) were identified. These patients were classified on the basis of their histopathology at primary diagnosis into either low- (GS ≤ 6, well differentiated), intermediate- (GS = 7, moderately differentiated), or high-GS cohorts (GS ≥ 8, poorly differentiated prostate cancer). Using semiautomated LN segmentation software (Fraunhofer MEVIS), we measured node volume and short-axis dimensions (SADs) and long-axis dimensions based on CT and compared with the SUV_max. Nodes demonstrating uptake of ⁶⁸Ga-PSMA with an SUV_max of 2.0 or more were considered PSMA-positive, and nodes with an SAD of 8 mm or more were considered positive by morphologic criteria. Results: Mean SUV_max was 13.5 (95% confidence interval [CI], 10.9–16.1), 12.4 (95% CI, 9.9–14.9), and 17.8 (95% CI, 15.4–20.3) within the low-, intermediate-, and high-GS groups, respectively. The morphologic assessment of the ⁶⁸Ga-PSMA–positive LN demonstrated that the low-GS cohort presented with smaller ⁶⁸Ga-PSMA–positive LNs (mean SAD, 7.7 mm; n = 113), followed by intermediate- (mean SAD, 9.4 mm; n = 122) and high-GS cohorts (mean SAD, 9.5 mm; n = 127). On the basis of the CT morphology criteria, only 34% of low-GS patients, 56% of intermediate-GS patients, and 53% of high-GS patients were considered CT positive. Overall, ⁶⁸Ga-PSMA imaging led to a reclassification of stage in 90 patients (61%) from cN0 to cN1 over CT. Conclusion: ⁶⁸Ga-PSMA PET is a promising modality in biochemical recurrent prostate cancer patients for N staging. Conventional imaging underestimates LN involvement compared with PSMA molecular staging score in each GS cohort. The sensitivity of ⁶⁸Ga-PSMA PET/CT enables earlier detection of subcentimeter LN metastases in the biochemical recurrence setting.

Biochemical recurrence (BCR) of prostate cancer (PCa) after initial therapy can herald an aggressive disease course. About 1 in 3 PCa patients who underwent local treatment with curative intent develops prostate-specific antigen (PSA) recurrence within 15 y (1,2). PSA is a highly sensitive method for detecting recurrence, and treatment of BCR is critical to prolong survival (3). However, recurrences can occur in a variety of locations including the lymphatic tissue (30.5%), skeleton (42.1%), retroperitoneum (13.7%), and viscera (13.7%) (4). Treatment varies according to the number and site of recurrence, but localizing such recurrences has been problematic with existing imaging methods such as CT and bone scanning because of their lack of sensitivity and specificity. For instance, only 11%–14% of patients with biochemical failure after radical prostatectomy have positive CT scan findings (5,6), with similar results with MRI (6–8). Therefore, detecting and localizing sites of recurrence is a strong unmet need for personalizing salvage therapy within individual patients.

Although a variety of PET agents have been developed to detect PCa recurrences, the most successful to date have been small-molecule agents targeting prostate-specific membrane antigen (PSMA) (9,10). One such agent, Glu-urea-Lys-(Ahx)-(⁶⁸Ga(HBED-CC)), also known as ⁶⁸Ga-PSMA, is particularly promising (10–12). ⁶⁸Ga-PSMA PET imaging has been shown to improve detection of lymph node metastasis and other sites of recurrence with very high positive predictive values and accuracy (10,13,14). Even though PET has a lower intrinsic spatial resolution of 3–5 mm compared with the submillimeter resolution for CT and MRI, ⁶⁸Ga-PSMA PET imaging has successfully detected nodal recurrences in two thirds of patients who would have been missed using conventional CT morphologic criteria (15). ⁶⁸Ga-PSMA PET/CT specifically benefits from the power of molecular imaging to detect small-sized but high–target-expressing lesions. One recent study demonstrated that ⁶⁸Ga-PSMA PET can achieve a sensitivity, specificity, and accuracy of 77.9%, 97.3%, and 89.9% for N staging in recurrent disease (13). A metaanalysis involving 16 ⁶⁸Ga-PSMA PET articles covering 1,309 patients reported a summary sensitivity and specificity of 80% and 97% on a per-lesion analysis (14).

Until now, most ⁶⁸Ga-PSMA PET imaging studies have focused on the occurrence of lymph node metastases without correlation to the risk profile of the patient at primary diagnosis. For instance, the performance of ⁶⁸Ga-PSMA PET studies in PCa patients with histopathologic low-risk disease is still lacking. Thus, the aim of this study was to compare the profiles of ⁶⁸Ga-PSMA PET–positive nodal metastases in patients with BCR according to PSMA uptake and CT on well, moderately, and poorly differentiated PCa at initial diagnosis.

MATERIALS AND METHODS

Patients

In this 2-center retrospective investigation, we evaluated 362 PET-positive lymph nodes in 147 consecutive patients (median age, 68 y; range, 44–87 y) with BCR of PCa (median PSA, 5 ng/mL; range, 0.2–294 ng/mL). The analysis was based on institutional databases at the University of Heidelberg and the Technical University Munich, which was randomized to reach a relatively equal cohort size. Only patients who underwent previous prostatectomy without any other oncologic disease were included in this study. Gleason scores (GS) were obtained according to the reports on prostatectomy specimen, and because of the restriction of data low-GS patients’ grading was based on either prostatectomy specimen or biopsy. Of 51 low-GS patients, grading for 32 GS was based on prostatectomy specimen and grading for 19 GS was based on biopsy. BCR was defined as 2 sequential PSA values of 0.2 ng/mL or more after prostatectomy (3). Patient characteristics are shown in Table 1. Before prostatectomy, the median PSA was 13 ng/mL (range, 2–252 ng/mL). Patients were stratified by GS as follows: low-GS (GS ≤ 6, well-differentiated PCa), intermediate-GS (GS = 7, moderately differentiated PCa), or high-GS cohort (GS ≥ 8, poorly differentiated PCa). All patients were enrolled under an investigational protocol that was approved by the investigational review board of both universities and was in accordance with the Helsinki Declaration (permit S-321/12 at the University of Heidelberg and 5665/13 at the Technical University Munich). Written informed consent for anonymized evaluation and publication of their data were obtained from all patients.

TABLE 1.

Patient Characteristics

Patient characteristic	n
Total no. of patients	147
Low-GS cohort
No. of patients	51
Median age (y)	71 (59–87)
GS	6
Median iPSA (ng/mL)	11.3 (3.2–84.3)
Median BCR PSA (ng/mL)	5.0 (0.5–41.3)
Mean time interval from RP to BCR (mo)	103
Further treatment
Radiation therapy after RP	5
Androgen-deprivation therapy  during/prior imaging	4
Unknown	20
Intermediate-GS cohort
No. of patients	48
Median age (y)	66 (52–85)
GS	7
Median iPSA (ng/mL)	12.0 (3.8–252.0)
Median BCR PSA (ng/mL)	5.1 (0.3–212.0)
Mean time interval from RP to BCR (mo)	56
Further treatment
Radiation therapy after RP	20
Androgen-deprivation therapy  during/prior imaging	12
Unknown	6
High-GS cohort
No. of patients	48
Median age (y)	66 (44–80)
GS	≥ 8
Median iPSA (ng/mL)	23.5 (2–241)
Median BCR PSA (ng/mL)	5.0 (0.2–293.7)
Mean time interval from RP to BCR (mo)	45
Further treatment
Radiation therapy after RP	15
Androgen-deprivation therapy  during/prior imaging	18
Unknown	8

RP = radical prostatectomy.

Data in parentheses are ranges.

Image Acquisition, PSMA PET Analysis, and Volumetric CT Histogram Analysis

All PET/CT examinations were performed on a Biograph 6 PET/CT or a Biograph mCT scanner (Siemens Medical Solutions). Imaging was initiated 60 ± 10 min after intravenous injection of ⁶⁸Ga-PSMA at a median dose of 2–3 MBq/kg of body weight. On the Biograph 6 PET/CT, a CT (130 keV, 80 mAs; CareDose) without contrast medium was obtained for attenuation correction of the PET scan. On the Biograph mCT, first a diagnostic CT scan was obtained in the portal venous phase 80 s after intravenous injection of contrast agent (Imeron 300; Bracco Imaging) followed by the PET scan.

Static emission scans, corrected for dead time, scatter, and decay, were acquired from the vertex to the proximal legs—requiring 8 bed positions, each taking 3–4 min. The images were iteratively reconstructed with an ordered-subset expectation maximization algorithm using 4 iterations with 8 subsets and gaussian filtering with an in-plane spatial resolution of 5 mm in full width at half maximum. For calculation of the SUV, circular regions of interest were drawn around the region of interest on transaxial slices and automatically adapted to a 3-dimensional volume of interest with e.soft software (Siemens) at a 70% isocontour. The CT scan was reconstructed with a B30 kernel to a slice thickness of 5 mm with an increment of 2.5 mm.

The threshold SUV_max for discriminating between benign and malignant lymph nodes in ⁶⁸Ga-PSMA PET was based on a blood-pool sample from 20 patients in each group and results from a recent prospective study in primary PCa evaluating the use of SUV thresholding (15,16). On these considerations, lymph nodes were considered abnormal if their uptake SUV exceeded 2.0. A maximum of 3 malignant nodes was identified in all patients. If a patient harbored more than 3 lymph node metastases, we chose the 3 most prominent lymph nodes with the highest SUV_max value as it represents PSMA uptake. All suspected lymph nodes were assigned to one of the following anatomic locations: regional lymph node metastasis including internal iliac, external iliac, obturator fossa, presacral, and distant lymph node metastasis for those around the common iliac, paraaortic, paracaval/interaortocaval, and other regions, for example, cervical and mediastinal (17).

Volumetric CT analysis was performed on ⁶⁸Ga-PSMA PET–positive nodes using semiautomated software (Fraunhofer MEVIS) that segments the nodes and automatically determines mean density, short-axis diameter (SAD), long-axis diameter (LAD), and volume (Figs. 1 and 2). Semiautomatic 3-dimensional histogram analysis was performed after the user provided a seed point in the lymph node and the software found the margins of the node. The segmentation started with a fixed-width thresholding around the seed point. To remove attached vessels, muscles, or other lymph nodes, a watershed transform was performed on the distance map of the thresholding result. The watershed transform is controlled to include and exclude features that are set according to an ellipsoid approximation of the lymph node. Quantitative and qualitative data of each patient were analyzed by 1 experienced radiologist and 1 nuclear medicine physician in each center. The volumes were evaluated by a radiologist to ensure that no extra nodal tissue was segmented and each node was manually corrected in all 3 dimensions, if necessary. The inter- and intraobserver reproducibility demonstrated a coefficient of variation of less than 5% (18). Any lymph node with an SAD above 8 mm was considered positive according to CT criteria (7,8). The size, represented by volume, SADs and LADs, and node location, was recorded for every node.

FIGURE 1.

A 66-y-old patient with PSA relapse of 9.1 ng/mL, 5 y after radical prostatectomy (GS, 9; iPSA level, 69.9 ng/mL). (A) ⁶⁸Ga-PSMA PET maximum-intensity projection demonstrates multiple positive nodes in pelvic and abdominal region. (B) Single axial PET/CT image presents 2 PSMA-positive nodes (arrows). (C) Node indicated by red arrow with SUV_max 36.32 was segmented using Fraunhofer MEVIS software. The software enables automatic quantification of dimensions based on axial (D), coronal (E), and sagittal planes (F) of CT images (SAD, 7.57 mm; LAD, 11.35 mm; volume, 0.5 mL). (G) Image shows 3-dimensional volume rendering (VR), which was also provided by the software.

FIGURE 2.

A 58-y-old patient with PSA relapse of 4.9 ng/mL, 9 mo after prostatectomy (GS, 6; iPSA, 10 ng/mL). (A) Patient presented multiple ⁶⁸Ga-PSMA PET–positive nodes as shown in PET maximum-intensity projection. (B and C) A positive node (red arrow) with PSMA uptake SUV_max 7.2 was segmented using Fraunhofer MEVIS software (SAD, 5.93 mm; LAD, 7.73 mm; and volume, 0.31 mL) Images (D–F) show segmented lymph node on each plane and volume rendering (VR) (G).

Statistical Evaluation

The evaluation of each group was based on the following parameters: measured SUV_max, volume, SAD, and LAD. To account for multiple observations within the same patients (up to 3 lymph nodes), comparisons between different risk cohorts were performed using a linear mixed model with risk group as a fixed effect and patient ID as a random effect. Descriptive statistics defining the percentage of positive PET/CT scans with nodes smaller than 8-mm SAD were performed. Statistical calculations were performed using the software SAS (Proc Mixed; SAS Institute) and IBM SPSS Statistics (IBM Corp.).

RESULTS

Among 147 patients with BCR, a total of 362 positive lymph nodes were evaluated on ⁶⁸Ga-PSMA PET/CT. The mean SUV_max of ⁶⁸Ga-PSMA uptake in malignant nodes was 14.67 (interquartile range, 11.0–17.0). One hundred thirteen lymph nodes in 51 low-GS patients, 122 lymph nodes in 48 intermediate-GS patients, and 127 lymph nodes in 48 high-GS patients were positive on ⁶⁸Ga-PSMA PET scans. Median initial PSA (iPSA) in the low-GS cohort was 11.3 ng/mL (range, 3.2–84.3 ng/mL), in the intermediate-GS cohort iPSA was 12.0 ng/mL (range, 3.8–252.0 ng/mL), and in the high-GS cohort iPSA was 23.5 ng/mL (range, 2.0–241.0 ng/mL). Median PSA values at imaging were 5.0 ng/mL (range, 0.5–41.3 ng/mL), 5.1 ng/mL (range, 0.3–212 ng/mL), and 5.0 ng/mL (range, 0.2–293.7 ng/mL) for low-, intermediate-, and high-GS patients (Table 1).

⁶⁸Ga-PSMA Ligand Uptake

The ⁶⁸Ga-PSMA ligand uptake showed a nonsignificant difference between the 3 GS cohorts (P = 0.062). Mean SUV_max of affected lymph nodes in the low-, intermediate-, and high-GS cohorts was 13.5 (95% confidence interval [CI], 10.9–16.1), 12.4 (95% CI, 9.9–14.9), and 17.8 (95% CI, 15.4–20.3), respectively (Table 2; Fig. 3A). Although the P value of uptake between the intermediate- and high-GS cohorts was less than 0.05 (P = 0.031), it is considered nonsignificant because the P value of the mixed model of the 3 cohorts was 0.05 or more (Table 3).

TABLE 2.

Mean SUV_max, SAD, LAD, and Volume Between Each Cohort

	SUVmax			SAD (mm)			LAD (mm)			Volume (mL)
Risk profile	Mean	95% CI, lower	95% CI, upper	Mean	95% CI, lower	95% CI, upper	Mean	95% CI, lower	95% CI, upper	Mean	95% CI, lower	95% CI, upper
Low GS	13.5	10.9	16.1	7.7	6.9	8.4	11.9	11.0	12.7	1.0	0.8	1.2
Intermediate GS	12.4	9.9	14.9	9.4	8.6	10.2	13.9	12.8	14.9	1.6	1.1	2.2
High GS	17.8	15.4	20.3	9.5	8.7	10.4	14.2	13.1	15.2	1.7	1.1	2.3

FIGURE 3.

Uptake and morphologic distribution in each risk group. (A) ⁶⁸Ga-PSMA ligand uptake as represented with SUV_max shows increasing trend together with risk profile. (B) SAD of ⁶⁸Ga-PSMA–positive lymph node displays that most metastatic lymph nodes are smaller than 10 mm, indicating that ⁶⁸Ga-PSMA is able to detect micrometastasis. SAD shows growing tendency alongside risk profile.

TABLE 3.

Pairwise Comparisons Between Each Cohort

	Pairwise comparisons, P*
Risk profile (I*J)	SUVmax	SAD	LAD	Volume
Mixed-model, 3 groups	0.062	0.013*	0.023*	0.110
Low GS vs. intermediate GS	0.797	0.011*	0.024*	0.150
Intermediate GS vs. high GS	0.031	0.953	0.801	0.540
Low GS vs. high GS	0.057	0.009*	0.012*	0.040

^*Mean difference is significant at 0.05 level.

Morphometric Analysis

The 3-dimensional volumetric segmentation enabled derivation of several important parameters: lymph node size represented by SAD and LAD and lymph node volume of ⁶⁸Ga-PSMA PET–positive findings. The SAD between GS cohorts demonstrated significant differences (P = 0.013) as follows: mean low-GS SAD, 7.7 mm (95% CI, 6.9–8.4 mm; n = 113); intermediate-GS SAD, 9.4 mm (95% CI, 8.6–10.2 mm; n = 122); and high-GS SAD, 9.5 mm (95% CI, 8.7–10.4 mm; n = 127) (Table 2; Fig. 3B). Pairwise comparisons for lymph node SADs showed a significant difference between low- and intermediate-GS cohorts (P = 0.011) and low- and high-GS lymph node SADs (P = 0.009) (Table 3). The difference in SAD between the intermediate- and high-GS cohorts was not significant. ⁶⁸Ga-PSMA–positive lymph nodes also showed differences in LAD (P = 0.023) according to GS cohort with a mean low-GS LAD of 11.9 mm (95% CI, 11.0–12.7 mm), mean intermediate-GS LAD of 13.9 mm (95% CI, 12.8–14.9 mm), and mean high-GS LAD of 14.2 mm (95% CI, 13.1–15.2 mm). However, as seen in Table 3, overall lymph node volumes were not significantly different among the 3 cohorts.

Among the 113 PSMA-positive lymph nodes (found in 31/51 patients) in the low-GS cohort, 75 (66%) were morphologically negative (≤8 mm) on CT and 38 (34%) were morphologically positive. In the intermediate-GS cohort, among 122 PSMA-positive lymph nodes, 53 (44%) were morphologically negative and 69 (56%) were morphologically positive. In the high-GS group, among 127 consecutive PSMA-positive lymph nodes, 60 (47%) nodes were morphologically negative and 67 (53%) were morphologically positive. Figure 4 depicts the morphologic distribution in each cohort next to the CT morphologic cutoff line of 8 and 10 mm. Therefore, among 362 PSMA-positive lymph nodes, 174 (48%) were positive based on conventional imaging criteria and 188 nodes (52%) were negative.

FIGURE 4.

Morphologic distribution of all 362 lymph nodes aligned to their SAD. Cutoff SAD was set to both 10 and 8 mm. This graph demonstrates that around two thirds of PSMA-positive nodes are not sufficiently detected based on their morphologic criteria.

Staging Results by Imaging Modality

For each risk cohort, ⁶⁸Ga-PSMA PET/CT imaging demonstrated tumor detection superior to CT. Using standard TNM staging CT would have properly staged only 57 (38%) of 147 patients. By this standard, 35 patients (69%) of the low-GS cohort, 29 patients (60%) of the intermediate-GS cohort, and 26 patients (54%) of high-GS cohort would falsely be considered as cN0 by CT criteria. Overall, ⁶⁸Ga-PSMA PET/CT led to reclassification of 90 patients (61%) from cN0 to cN1.

Location of Lymph Nodes

Among the nodes detected by ⁶⁸Ga-PSMA PET, 254 (70%) were located in distant lymphatic stations (common iliac vessel, paraaortic, paracaval/interaortocaval, and other regions, for example, cervical and mediastinal), meanwhile 108 (30%) nodes were in regional, pelvic lymph node stations (internal iliac vessel, external iliac vessel, obturator fossa, and presacral). Distant ⁶⁸Ga-PSMA PET/CT–positive lymph nodes were found in 84% of low-GS patients, 68% of intermediate-GS patients, and 71% of high-GS patients. Regional nodes were found in 16%, 32%, and 29% of low-, intermediate-, and high-GS cohorts, respectively. The most common location for a distant malignant node was the paraaortic station (17%), followed by the interaortocaval and mediastinal regions.

DISCUSSION

In this study, we have shown that ⁶⁸Ga-PSMA PSMA PET imaging is superior to conventional CT criteria independent of initial tumor differentiation of PCa. ⁶⁸Ga-PSMA ligand uptake in patients with BCR may vary according to the GS of the patient. Multiple studies have shown that the sensitivity and specificity of ⁶⁸Ga-PSMA PET/CT varies from 54% to 94% and 97% to 100%, respectively (10,12–15,19). Because of ethical and practical reasons, histopathologic verification could be obtained only in 4 of 147 patients who underwent salvage lymphadenectomy after ⁶⁸Ga-PSMA PET. In all other patients, a comprehensive standard of reference including contrast-enhanced CT, MRI, repeated ⁶⁸Ga-PSMA ligand PET/CT, and course of PSA confirming the initial suggestive lesions or showing disappearance of suspected lesions after local/systemic treatment was obtained, also suggesting the malignancy nature of the detected nodes.

The ⁶⁸Ga-PSMA uptake increases with the initial risk profile of the patient. One potential confounder in this study would be other malignancies. Thyroid, colon, kidney, and brain malignancies have also shown PSMA uptake. Therefore, in this study we intentionally excluded patients with other oncologic disease to avoid these false-positive findings. Other studies have also reported possible false-positives in the cervical, celiac, and sacral ganglia (20,21). The SUV_max of ganglia, however, was relatively low compared with actual lymph nodes metastasis. Moreover, the localization of positive PSMA uptake corresponded to well-known lymph node stations, further lowering the possibility of a false-positive.

The morphologic distribution as shown in Figure 4 illustrates that nodal staging with CT alone strongly underestimates nodal involvement. Only 48% of all ⁶⁸Ga-PSMA PET/CT–positive nodes exceeded the size criteria for metastasis on CT. The CT sensitivity calculated in this study (48%) is consistent with the existing data published (5,6). In the low-GS cohort, most (66%) positive lymph nodes were smaller than 8 mm. Regardless of their small size, these nodes demonstrated excellent uptake, with a mean SUV_max of 13.5 (95% CI, 10.9–16.1). Though it is important for patient management to detect and localize the sites of recurrence as early as possible, cross-sectional imaging is known to be insufficient. Detecting the metastases while the lesions are still small may allow earlier and individually tailored salvage therapies (3,22). A comparison of 2 main parameters, ⁶⁸Ga-PSMA PET SUV_max and SAD on CT, is shown in Figure 5.

FIGURE 5.

Relationship between SUV_max and SAD (mm). CT morphologic cutoff was set at 8 mm and PET/CT SUV_max at 2.0.

PCa patients with low initial GS are considered unlikely to recur but nonetheless do probably because of higher grade disease that is not diagnosed at biopsy. This is also 1 limitation of our study: 19 of 51 gradings were done by biopsy, which might translate into understaging by sampling error in comparison to histologic workup of the whole gland after prostatectomy. It is unclear whether PSMA PET on biochemical relapse is as useful in these low-GS patients as it is in higher GS patients who have a known likelihood of progression. Patients with local recurrence in the low-risk group typically respond well to early salvage radiotherapy (23). Therefore, the findings of this study might offer the possibility to further individualize salvage radiation and to possibly use targeted radiotherapy to recurrent lesions in the future, thus sparing unnecessary pelvic irradiation. We identified pelvic lymph node metastases in 61% of BCR patients with initial low-risk GS, which is well in line with recent literature that reported lymph node metastases as reason for BCR in 81% of patients using response to salvage radiotherapy as a standard of reference (24).

The localization of the nodal involvement is also essential to determine whether systemic or regional salvage treatments are required. ⁶⁸Ga-PSMA PET/CT imaging provides indispensable information about the location and number of metastases. In fact, because most patients had previous pelvic lymph node resections, most of the positive nodes (70%) in this study were located in more distant lymph stations. Thus, this kind of information could significantly affect potential radiation therapy planning similar to the results recently reported by van Leeuwen et al. (25).

Comparison of the 3 GS groups showed significant differences in the diameters (SAD and LAD) of detected nodes according to risk cohort. Although SAD and LAD should correlate with the volume, nodal volume did not correlate well with risk correlation. It is possible that shape (reflected better as SAD and LAD) plays a more important role than volume per se. It is also possible that the software over- or underestimates volume whereas diameter determination is more readily verified. Furthermore, these results may also be an artifact of selecting only the 3 most prominent nodes in men with multiple metastases. The time interval of BCR corresponds with the risk profile of each cohort. BCR tends to occur for low-GS patients within the longest time interval of 103 mo, meanwhile for intermediate- and high-GS patients BCR occurred within 56 and 45 mo, respectively. In such cases, the BCR-free interval and SAD of lymph node metastases may indicate the tissue growth/malignancy of the metastases in each cohort. However, more research is needed to better understand the correlation between time interval and the malignancy of the metastases.

Our results are similar to previous studies. A ⁶⁸Ga-PSMA study of 53 positive lymph nodes in intermediate- and high-risk patients (GS ≥ 7) showed an SUV_max of 12.7 (range, 2.4–51.0) with a corresponding node size of 8.3 mm (range, 4–25 mm) (13). These data are in line with our evaluation of intermediate- and high-risk cohorts. In comparison, in this study the mean SUV_max was 12.4 (range, 2–49) and SAD was 9.4 mm (range, 1.9–33.5 mm) in the intermediate-risk group, and the mean SUV_max was 17.8 (range, 2–97) and SAD was 9.5 mm (range, 2.6–29.1 mm) in the high-risk cohort.

The classification of the subgroups in this study was based on the GS. The GS was determined at the initial PCa diagnosis, thus making its relevance in a recurrence or relapse situation questionable because the patient has received prostatectomy or other therapy. However, the correlation of GS and the risk of recurrent PCa has been reported in various studies (26,27). The guidelines on PCa published by the European Association of Urology, for instance, use the classification of GS (at biopsy) as a predictor of biochemical relapse (3). However, there have been some modifications regarding the definition of GS 6, which upgraded the previous classification of Gleason 6 into Gleason 7a (26). Some patients in this study were diagnosed before 2010, thus the Gleason grading reported may not be equal with the current definition. Overall, using the GS makes our results more generalizable.

Despite the retrospective nature of this study, randomized patient selection within each GS cohort and masked analysis might help in reducing unforeseen biases. The selection of the 3 most prominent ⁶⁸Ga-PSMA–positive lymph nodes from each patient might tend to overestimate the SUV_max and lesion volumes of each group, but at the same time reduce the in-patient uptake bias. The lack of routine histologic validation is 1 limitation of this study; however, after BCR surgery is rarely an option and clinical follow-up or imaging are routinely used standards of reference in this setting.

⁶⁸Ga-PSMA PET/CT is promising as a better restaging modality for PCa patient with PSA relapse. A further investigation with a better differentiation of the recurrence subgroup could provide a more refined indication of the size and SUV_max of lymph nodes as a function of assigned risk stratification. Besides GS, other relevant parameters such as PSA doubling time, tumor invasion specimen (pT stage), and time to PSA recurrence are intriguing and could provide a better specification for patient management. However, recently developed ¹⁸F-labeled PSMA ligands allow a higher patient throughput and show promising PSMA diagnostics for both T and N/M staging in a urooncologic environment (28–31).

CONCLUSION

This retrospective study demonstrates that ⁶⁸Ga-PSMA PET imaging is clearly superior to CT for detecting recurrent nodal disease independently of initial histology. Within risk strata established at diagnosis, there are subtle differences both in uptake and in node diameters according to initial GSs. The ability of ⁶⁸Ga-PSMA PET/CT to better stage recurrent disease along with its ability to pinpoint metastatic sites make it a valuable tool in assessing patients with BCR regardless of the initial risk cohort.

DISCLOSURE

No potential conflict of interest relevant to this article was reported.