Visual Abstract
Keywords: CHAARTED, mHSPC, PSMA, PET/CT, bone scan, aBSI, whole-body tumor volume
Abstract
High-volume disease (HVD) and low-volume disease (LVD) definitions in metastatic hormone-sensitive prostate cancer (mHSPC) patients are based on conventional imaging (CI) (CT/MRI with bone scan [BS]) according to CHAARTED criteria. HVD and LVD definitions are associated with overall survival and are used for treatment decisions. It remains unknown how these definitions transfer to prostate-specific membrane antigen (PSMA) PET imaging. The aim of this retrospective multicenter study was to compare the CI-based disease volume criteria to PSMA PET–based volume definitions in a CHAARTED-like cohort. Methods: mHSPC patients from 5 international sites who underwent PSMA PET/CT or PSMA PET/MRI and BS within a time interval of 100 d and without initiation of a new therapy between the 2 scans were retrospectively included in the analysis. CHAARTED HVD and LVD criteria were applied to BS, CT, MRI, and PSMA PET. HVD was defined by the presence of visceral metastases or at least 4 bone metastases (with ≥1 beyond the spine or pelvis). Whole-body (WB) tumor burden was estimated with the automated bone scan index (aBSI, EXINI v2.0) on BS and with the WB PSMA PET–positive tumor volume (PSMA-TV) on PSMA PET, respectively. Results: Sixty-seven patients with paired PSMA PET and BS were included. The median prostate-specific antigen level was 54.9 ng/mL (interquartile range [IQR], 10.4–191.0 ng/mL). On the basis of CI, it was determined that 17 of 67 patients had HVD-CI (25.4%) and 50 of 67 patients had LVD-CI (74.6%). On the basis of PSMA PET, it was determined that 27 of 67 patients had HVD-PET (40.3%) and 24 of 67 patients had LVD-PET (35.8%). In total, 16 of 67 patients (22.4%) had no visible lesion or only locoregional pelvic disease (M0) with PSMA PET (M0-PET). Stage migration between CI and PSMA PET occurred in 27 of 67 patients (40.3%) by both upstaging and downstaging: 11 of 50 (22%) LVD-CI patients were HVD-PET, whereas 1 of 17 (5.9%) HVD-CI and 15 of 50 (30%) of LVD-CI patients were M0-PET. The median WB PSMA-TV and automated BS index were 248.0 mL (IQR, 54.6–1,427.0 mL) and 3.4% (IQR, 1,0–7.2%) for HVD-CI, 25.1 mL (IQR, 6.6–74.6 mL) and 0.1% (IQR, 0.0–0.2%) for LVD-CI, 141.0 mL (IQR, 47.5–458.0 mL) and 0.9% (IQR, 0.04–4.1%) for HVD-PET, and 31.5 mL (IQR, 10.1–67.9 mL) and 0% (IQR, 0–0.1%) for LVD-PET, respectively. The optimal WB PSMA-TV to stratify CI-based CHAARTED LVD-CI versus HVD-CI was 107 mL with a misclassification of 21.9%. Conclusion: Compared with CI, addition of PSMA PET leads to M0 downstaging in every third and LVD to HVD upstaging in every fifth mHSPC patient. Future HVD and LVD definitions based on PSMA PET/CT should be adjusted based on patient outcome.
International guidelines recommend that prostate radiotherapy should be added to androgen-deprivation therapy (ADT) in metastatic hormone-sensitive prostate cancer (mHSPC) patients with low-volume disease (LVD), whereas patients with high-volume disease (HVD) mHSPC who are fit for chemotherapy are considered for triple therapy (ADT plus docetaxel plus either darolutamide or abiraterone/prednisone) (PEACE-1 and ARASENS trials) (1,2). Apalutamide or enzalutamide (TITAN, ARCHES, and ENZAMET studies) are used for LVD or HVD, and abiraterone acetate and prednisone have their value in high-risk disease (3–7).
The current definitions of LVD and HVD were defined for the first time in the CHAARTED study based on a combination of prior classifications that defined tumor burden in mHSPC (4). CHAARTED and STAMPEDE underlined the importance of volume of disease for patient prognosis as assessed by conventional imaging (CI), such as CT, MRI, and bone scan (BS), whereas trials such as GETUG-AFU 15 and metaanalyses such as STOPCAP confirmed the prognostic and predictive findings from CHAARTED (8–10). HVD is defined as the presence of visceral metastases or at least 4 bone metastases (≥1 beyond the vertebral bodies or the pelvis).
Prostate-specific membrane antigen (PSMA) PET has a higher diagnostic accuracy for initial staging, biochemical recurrence localization, and advanced disease restaging than does CI (11–14). Additionally, PSMA PET offers whole-body (WB) tumor burden volume quantification (15). How to extrapolate the current disease volume definitions in mHSPC—that are based on CI only—to PSMA PET imaging remains unknown. The aim of this retrospective multicenter study was to compare the CI-based disease volume criteria to PSMA PET–based volume definitions in a CHAARTED-like cohort.
MATERIALS AND METHODS
This was an international multicenter retrospective study. PSMA PET databases from 5 institutions (UCLA, UCSF, University of Pretoria, U.K. Essen, and LMU Munich) were retrospectively screened for patients meeting the following inclusion criteria: hormone-sensitive prostate cancer, defined as no prior history of ADT, and paired planar BS and a PSMA PET/CT scan (with diagnostic CT component, 64/67 patients, 95.5%) or PSMA PET/MRI (3/67 patients, 5%) performed within 100 d or less of one another without any new interval treatment initiation. To provide a CHAARTED-like cohort (4), only patients with at least M1a disease on CI (combination of BS and CT/MRI component of PSMA PET/CT/MRI) defined as lymph node metastases above the bifurcation of the common iliac arteries were included. Anonymized Digital Imaging and Communications in Medicine and clinical data were transferred electronically to the lead investigators’ site (UCLA). This study was approved by the institutional review board of UCLA (21-000029).
Image Analysis
CI
Visual Analysis.
The CT/MRI component of PSMA PET/CT or PSMA PET/MRI was retrospectively analyzed by one central reader using Affinity 2.2 (Hermes Medical Solutions) who was masked to the respective PET component. The presence of lesions in the following 4 regions was assessed by visual analysis: prostate bed, lymph nodes, bones, and viscera. Lymph nodes were rated as positive if the short axis was greater than 15 mm or if the short axis was greater than 8 mm with loss of fatty hilum, focal necrosis, penetration of the capsule, or increased contrast enhancement.
For bony involvement, CT/MRI and BS images were evaluated (16). No SPECT/CT was used. BS (planar scintigraphy) was evaluated using EXINI version 2.0 (EXINI Diagnostics AB) for the number (≤10 bone metastases or >10 bone metastases) and localization of bone metastases with approval by visual analysis from the central reader.
CHAARTED criteria dividing HVD-CI and LVD-CI were applied: HVD-CI was defined as the presence of visceral metastases or at least 4 bone metastases (≥1 beyond the vertebral bodies or the pelvis). All other patients not meeting these criteria were assigned to LVD-CI.
Semiquantitative Analysis.
EXINI version 2.0 was used to assess the automated bone scan index (aBSI), an artificial intelligence–based method to quantitatively measure the burden of bony disease (17): aBSI represents the percentage fraction of the skeletal mass with abnormal tracer uptake that is indicative of metastatic bone disease on a conventional planar BS (17). In mHSPC patients, a mean bone scan index value of 1.0% (range, 0–13.2%) was reported in previous publications (18,19).
PSMA PET
PET/CT and PET/MRI datasets were read by one central reader using Affinity 2.2.
Visual Analysis.
The presence of PSMA-positive lesions in the following 4 regions was assessed by visual analysis: prostate bed, lymph nodes, bones, and viscera. For bone regions, localization and number of metastases (≤10 bone metastases or >10 bone metastases) were assessed.
In accordance with the CHAARTED criteria, HVD-PET was defined as the presence of visceral metastases or at least 4 bone metastases (≥1 beyond the vertebral bodies or the pelvis). All other patients not meeting these criteria were assigned to LVD-PET except if only local, pelvic lymph node tumoral burden or no PSMA-positive lesion at all was detected (M0-PET).
Semiquantitative Analysis.
Whole-body PSMA PET–positive tumor volume (WB PSMA-TV) for [68Ga]Ga-PSMA-11 PET was obtained using a patient-specific global threshold: PSMA threshold = 4.30/SUVmean × (SUVmean + SUVSD), where SUVmean and SUVSD denote the mean and SD of a spheric liver region of interest (15-mm radius) (15). WB tumor volume for [18F]PSMA-1007 was assessed using a fixed threshold of SUV 4.0 as described previously (20). A region-specific tumor subvolume was assessed for 4 regions: prostate bed (PSMA-TV-prostate), lymph nodes (PSMA-TV-LN), bones (PSMA-TV-bone), and viscera (PSMA-TV-viscera).
Statistical Analysis
Patient characteristics and study variables are presented as median and interquartile range (IQR) or frequency (percentage) unless otherwise noted. Discordance between LVD- and HVD-CI and M0-, LVD-, and HVD-PET was depicted by creating a Sankey flow diagram. Differences between volume of disease as assessed by PSMA PET and CI were compared using t tests or Mann–Whitney U tests. To compare binary outcome parameters, the χ2 test was used. To assess the PET-based volume optimal cutoff for CI-based CHAARTED HVD versus LVD, a receiver-operating-characteristic analysis was performed, and the area under the curve was assessed. The Youden J index was used to evaluate the best discriminative cutoff value. The significance level was set at a P value of less than 0.05. All statistical analyses were performed using SPSS version 29 (IBM Corp.).
RESULTS
Patient Characteristics
In total, 67 mHSPC patients (age, 69.0 y; IQR, 64.0–76.0 y) without any prior ADT and with a median prostate-specific antigen level of 54.9 ng/mL (IQR, 10.4–191.0 ng/mL) were included across the 5 institutions (flowchart shown in Supplemental Fig. 1 [supplemental materials are available at http://jnm.snmjournals.org]; patient characteristics given in Table 1). Of 67 patients, 61 (91.0%) underwent the scans at initial staging and 6 of 67 patients (9.0%) underwent scans for biochemical recurrence. The median time between BS and PSMA PET was 37.3 d (IQR, 1–98 d). Of the 67 patients, 63 (94.0%) underwent [68Ga]Ga-PSMA-11 PET/CT and 4 (6.0%) underwent [18F]PSMA-1007 PET/CT (PSMA PET: median injected activity, 163 MBq [IQR, 110.5–188.5 MBq] at a median uptake time of 61.0 min [IQR, 55.0–69.0 min]).
TABLE 1.
Patient Characteristics (n = 67)
| Characteristic | Data |
|---|---|
| Age at imaging (y) | 69.0 (range, 53.0–82.0) |
| PSA at imaging (ng/mL) | 54.9 (IQR, 10.4–191.0) |
| Indication for imaging | |
| Initial staging | 61/67 (91.0%) |
| Biochemical recurrence | 6/67 (9.0%) |
| Gleason score at initial diagnosis | |
| <8 | 23/67 (34.3%) |
| ≥8 | 43/67 (64.2%) |
| Unknown | 1/67 (1.5%) |
| Prior treatments | |
| Radical prostatectomy | 4/67 (6.0%) |
| Definitive radiotherapy | 1/67(1.5%) |
| Salvage radiotherapy | 1/67 (1.5%) |
| None | 61/67 (91.0%) |
Imaging Visual Analysis
Disease Extent on CI
Overall, 21 of 67 patients (31.3%) showed a suggestive finding in the prostate region, and 30 of 67 patients (44.8%) had at least one suggestive lymph node metastasis on CT/MRI (N1 or M1a). EXINI version 2.0 detected at least one bone metastasis in 47 of 67 patients (61.2%). In 5 of 47 patients (12.8%), the central reader’s visual analysis led to a downstaging in the amount of included suggestive lesions compared with the results of EXINI version 2.0. Of 47 patients, 34 (72.3%) had no more than 10 bone metastases, and 13 (27.7%) had more than 10 bone metastases. In total, 4 of 67 patients (6.0%) had suggestive visceral lesions in the liver visible by CT/MRI (M1c) (further specifications are available in Table 2; Supplemental Table 1). Applying CHAARTED LVD/HVD criteria to CI, of 67 patients, 50 (74.6%) had LVD-CI and 17 (25.4%) had HVD-CI.
TABLE 2.
Imaging Characteristics (n = 67)
| Parameter | CI | PSMA PET |
|---|---|---|
| Highest (mi)TNM | ||
| (mi)M0 | 0/67 (0%) | 16/67 (23.9%) |
| (mi)M1a | 9/67 (13.4%) | 9/67 (13.4%) |
| (mi)M1b | 54/67 (80.6%) | 38/67 (56.7%) |
| (mi)M1c | 4/67 (6.0%) | 4/67 (6.0%) |
| Number of bone lesions | ||
| 0 | 20/67 (29.9%) | 26/67 (38.8%) |
| 1–10 | 34/67 (50.7%) | 24/67 (35.8%) |
| ≥ 10 | 13/67 (19.4%) | 17/67 (25.4%) |
| LVD | 50/67 (74.6%) | 24/67 (55.8%) |
| HVD | 17/67 (25.4%) | 27/67 (40.3%) |
| Local disease only/no suggestive lesion on a WB level | 0/67 (0%) | 16/67 (23.9%) |
| aBSI (%) | 0.2 (IQR, 0–0.5) | |
| WB PSMA-TV (mL) | 37.8 (IQR, 7.9–128.0) | |
| PSMA-TV-prostate (mL) | 6.7 (IQR, 0–31.8) | |
| PSMA-TV-LN (mL) | 11.9 (IQR, 0–22.5) | |
| PSMA-TV-bone (mL) | 0 (IQR, 0–27.4) | |
| PSMA-TV-visceral (mL) | 0 (IQR, 0–0) |
mi = minimally invasive; TNM = tumor, nodes, metastasis.
Disease Extent on PSMA PET
Overall, 45 of 67 patients (67.2%) showed a positive finding in the prostate region that could be detected when applying the above-listed tracer-specific thresholds for PSMA-positive lesions. Of the 67 patients, 36 (53.7%) had at least one positive lymph node metastasis (N1 or M1a); of those 36 patients, 25 (69.4%) had PSMA-positive extrapelvic lymph node metastases (M1a). PSMA-positive bone involvement was detectable in 41 of 67 patients (61.2%); among these 41 patients, 24 (58.6%) had no more than 10 bone metastases and 17 (41.5%) had more than 10 bone metastases. All 4 patients with visceral metastases by CT/MRI were also positive by PSMA PET in the respective visceral area (M1c) (4/67 patients, 6.0%). Table 2 and Supplemental Table 1 have further specifications.
When CHAARTED criteria are applied to PSMA PET, out of 67 patients, 27 (40.3%), 24 (25.8%), and 16 (22.4%) were designated HVD-PET, LVD-PET, and M0-PET, respectively (Table 2). Of note, only 1 of the 67 patients (1.5%) showed no PSMA PET–positive lesions at all.
Discordance Between CI and PSMA PET
LVD and HVD migration between CI and PSMA PET occurred by both upstaging and downstaging in 27 of 67 patients (40.3%) (Fig. 1): 11 of 50 patients (22%) of LVD-CI were HVD-PET, whereas 1 of 17 patients (5.9%) of HVD-CI and 15 of 50 (30%) of LVD-CI patients were M0-PET (Table 3; Figs. 1–3).
FIGURE 1.
Change in stage between CI and PSMA PET imaging.
TABLE 3.
Contingency Table: Concordant and Discordant Volume of Disease Classifications in Included Patient Cohort Divided by PET and CI (n = 67)
| Classification | HVD-CI | LVD-CI |
|---|---|---|
| HVD-PET | 16 | 11 |
| LVD-PET | 0 | 24 |
| M0-PET | 1 | 15 |
FIGURE 3.
68-y-old patient with increased aBSI of 0.5% rated as LVD on CI with local disease only on PSMA PET. With increased prostate-specific antigen level of 74.2 ng/mL and intense tracer uptake in lumbar spine, patient was referred to PSMA PET/CT and BS without any prior therapy. PSMA PET revealed no elevated uptake in respective region in lumbar spine, also no osteolytic or osteoblastic lesion was detected in this area.
FIGURE 2.
74-y-old patient with M1a disease only on CI but HVD on PSMA PET. Contours on BS are generated automatically by EXINI version 2.0. With increased prostate-specific antigen level of 77.0 ng/mL, patient was referred to PSMA PET/CT and BS without any prior therapy.
Semiquantitative Analysis
Quantification of Tumor Burden on BS
Overall, the median aBSI was 0.2% (IQR, 0–0.5%). The median aBSI was 3.4% (IQR, 1.0–7.2%) for HVD-CI and 0.1% (IQR, 0–0.2%) for LVD-CI (Tables 2 and 3; Supplemental Table 2).
The aBSI-based cutoff to differentiate LVD-CI and HVD-CI was 0.65% (area under the curve, 0.95 [95% CI, 0.853–1.00; P < 0.001]) with a misalignment (false-positive/false-negative rate) in 2 of 67 patients (3.0%) (Fig. 4).
FIGURE 4.
Scatterplots of tumor burden quantitative metrics. (A) Optimal cutoffs of WB PSMA-TV in milliliters and aBSI in percent to differentiate CHAARTED HVD-CI (n = 17; red plots) and LVD-CI (n = 50, blue plots) defined by receiver-operating-characteristic curve analysis were 107 mL and 0.65%, respectively (yellow lines). (B) Optimal cutoffs of WB PSMA-bone-TV in milliliters and aBSI in percent to differentiate CHAARTED HVD-CI (n = 6, red plots) and LVD-CI (n = 27, blue plots) defined by receiver-operating-characteristic curve analysis were 107 mL and 0.65%, respectively (yellow lines). Logarithmic scale was used to display numeric data over wide range of values.
Quantification of Tumor Burden on PSMA PET
Overall, the median WB PSMA-TV was 37.8 mL (IQR, 7.9–128.0 mL). The median PSMA-TV-prostate subvolume was 6.7 mL (IQR, 0–31.8 mL), the PSMA-TV-LN subvolume was 0.1 mL (IQR, 0–22.5 mL), the median PSMA-TV-viscera subvolume was 0 mL (IQR, 0–0 mL), and the median PSMA-TV-bone subvolume was 0 mL (IQR, 0–27.4 mL) (Table 2).
The median WB PSMA-TV was 31.5 mL (IQR, 10.1–67.9 mL) for LVD-PET and 141.0 mL (IQR, 47.5–458.0 mL) for HVD-PET.
The optimal cutoff to differentiate LVD-PET from HVD-PET was 87.7 mL (area under the curve, 0.80 [95% CI, 0.680–0.923; P < 0.001]). Using this cutoff, 55 of 67 cases (82.1%) showed concordant findings, whereas 12 of 67 patients (17.9%) were discordant using this cutoff value.
Intermodality Cross-Comparison
WB PSMA-TV, PSMA-TV-bone, PSMA-TV-viscera, and aBSI were significantly higher in HVD patients than in LVD patients with both the CI-based classification and the PSMA PET–based classification (Table 4).
TABLE 4.
Comparison of HVD and LVD on CI and PSMA PET
| Parameter | CI (CT/MRI plus BS) | P | PSMA PET imaging | P | ||
|---|---|---|---|---|---|---|
| HVD-CI (n = 17) | LVD-CI (n = 50) | HVD-PET (n = 27) | LVD-PET (n = 24) | |||
| Last PSA before scans (ng/mL) | 151.3 (IQR, 83.2–452.7) | 36.6 (IQR, 10.1–113.5) | 0.001 | 109.2 (IQR, 39.3–275.3) | 36.6 (IQR, 9.4–186.0) | 0.153 |
| Gleason score at diagnosis | 4/17: ≤8 | 19/49: ≤8 | 0.259 | 7/27: ≤8 | 7/24: ≤8 | 0.726 |
| 13/17: >8 | 30/49: >8 | 20/27: >8 | 16/24: >8 | |||
| 1/24: unknown | ||||||
| WB PSMA-TV (mL) | 248.0 (IQR, 54.6–1,427) | 25.1 (IQR, 6.6–74.6) | <0.001 | 141.0 (IQR, 47.5–458.0) | 31.5 (IQR, 10.1–67.9) | <0.001 |
| PSMA-TV-LN (mL) | 0 (IQR, 0–26.3) | 0.3 (IQR, 0–22.9) | 0.171 | 1.9 (IQR, 0–44.5) | 8.9 (IQR, 0–23.6) | 0.016 |
| PSMA-TV-bone (mL) | 209.0 (IQR, 32.3–783.5) | 0 (IQR, 0–8.8) | <0.001 | 40.1 (IQR, 18.7–458.0) | 0 (IQR, 0–0.25) | 0.001 |
| PSMA-TV-visceral (mL) | 0 (IQR, 0–4.2) | 0 (IQR, 0–0) | 0.035 | 0.0 (IQR, 0.0–4.2) | 0 (IQR, 0–0) | 0.035 |
| PSMA-TV-prostate (mL) | 0 (IQR, 0–14.8) | 7.1 (IQR, 0.2–31.9) | 0.374 | 3.4 (IQR, 0–30.1) | 9.6 (IQR, 0–24.0) | 0.24 |
| aBSI (%) | 3.4 (IQR, 1.0–7.2) | 0.1 (IQR, 0–0.2) | <0.001 | 0.9 (IQR, 0–4.1) | 0 (IQR, 0–0.1) | 0.003 |
PSA = prostate-specific antigen.
From CI to PET
HVD-CI and LVD-CI patients had a median WB PSMA-TV of 248.0 mL (IQR, 54.6–1,427 mL) and 25.1 mL (IQR, 6.6–74.6 mL), respectively. The optimal cutoff on WB PSMA-TV to differentiate LVD-CI from HVD-CI was 107 mL (area under the curve, 0.84 [95% CI, 0.714–0.963; P < 0.001]). Using this cutoff, 53 of 67 patients (79.1%) showed concordant findings, whereas 14 of 67 patients (20.9%) were discordant (Fig. 4).
From PET to CI
HVD-PET and LVD-PET patients had a median aBSI of 3.4% (IQR, 1.0–7.2%) and 0.1% (IQR, 0–0.2%), respectively. By receiver-operating-characteristic analysis, the aBSI-based cutoff between HVD-PET and LVD-PET was 0.65% with a discordance in 23.5% of the cases (area under the curve, 0.79) (Fig. 4).
Factors Associated with Discordance Between CI and PSMA PET
Patients without any stage discordance between CI and PSMA PET (n = 40) had a significantly higher prostate-specific antigen level (at the time of imaging) and significantly higher WB PSMA-TV than did the patients with a migration between both imaging modalities (n = 27; P < 0.01, each; Supplemental Table 3).
Patients with downstaging from HVD-PET to LVD-CI (n = 11) had a lower PSMA-TV-bone than did the patients with HVD by both PSMA PET and CI (n = 16; P = 0.02; Supplemental Table 4).
DISCUSSION
In this international retrospective analysis, we used a cohort of 67 CHAARTED-like patients with paired PSMA PET and planar BS to compare the definitions of HVD and LVD as defined by CI and PSMA PET imaging.
In 40.3% of the included patients, a hypothetical migration of the classification of the volume of disease (LVD or HVD), which means a discordance, using CI or PSMA PET occurred by both upstaging and downstaging when applying CHAARTED criteria to CI and PET separately, which is in line with the current literature (21).
In this cohort, PSMA PET detected significantly more lymph node metastases than did CI, which is in line with the current literature (22–24). PSMA PET is also more sensitive than the combination of CT and BS in detecting bone metastases, and as such, 11 of 50 (22%) of the LVD-CI patients were upstaged to HVD-PET. It remains unanswered if these patients should be treated as HVD-CI or still as LVD-CI, as recently suggested (25). Of note, patients with HVD as defined by CI (HVD-CI) had a significantly higher WB PSMA-TV and aBSI than did the patients categorized as HVD when applying CHAARTED classification to PET (HVD-PET). Also, patients with HVD-CI had a significantly higher PSA than did the patients with LVD-CI, whereas HVD-PET and LVD-PET patients did not have significantly different prostate-specific antigen levels. This indicates that patients with HVD by CI have a more advanced disease than patients with HVD by PSMA PET. It is important to point out the risk of false-positive bone lesions by PSMA PET, especially in the ribs and when using [18F]-based PSMA PET tracers that are predominantly gastrointestinal tract–excreted, which could lead to a direct shift to HVD-PET when applying CHAARTED criteria. Reader training and communication are critical to limit this potential pitfall (26–28).
Inversely, CI also detected 9% more suggestive bone lesions than PSMA. As reported earlier, this may be related to the high false-positive rate of BS in the untreated disease or initial staging setting (29). SPECT/CT should be acquired whenever possible to reduce the high false-positive and false-negative rates of planar BS (29,30). Interestingly, a significant group of 22.4% patients shifted from LVD-CI to a local tumor burden only in the PSMA PET (M0-PET). This questions whether the benefit observed of local prostate radiotherapy in the LVD patients of the STAMPEDE trial may be related to some patients that were incorrectly considered metastatic (false positive of BS) and thus appropriately treated with local treatment (30). Of note, one patient (1.5%) even shifted from HVD-CI to M0-PET.
The WB PSMA volume of bone metastases in milliliters was significantly different between HVD-CI/PET and LVD-CI/PET. Barbato et al. introduced a cutoff of 40 mL to divide HVD-CI and LVD-CI, but this previous publication included only PSMA PET/CT without a corresponding BS (11). In our study, the optimal cutoff of WB PSMA-TV to differentiate the CI-based CHAARTED LVD-CI from HVD-CI was 107.0 mL. This difference between the results from Barbato et al. and our findings might be caused by the missing BS in the study by Barbato et al. When applying CHAARTED criteria to PSMA PET, the optimal cutoff to differentiate LVD-PET from HVD-PET was 87.7 mL. In both analyses, there was a misclassification between LVD and HVD in up to 22% of cases. Using the aBSI classification, a cutoff of 0.65% to differentiate HVD-CI/PET from LVD-CI/PET was determined. Future studies should investigate how to integrate WB PSMA tumor quantitative parameters into disease staging and treatment management as the prognostic value of PSMA PET tumor volume could already have been demonstrated in prior studies (21,31,32).
Similarly, the combined integration of the localization of the disease (M1a, M1b, M1c) with the volume of disease should be further investigated. Follow-up analyses of this cohort with long-term survival data are planned.
This study has limitations: First, the retrospective design and the missing outcome parameters of the included patients were limitations. Second, the included mHSPC cohorts were heterogenous as some patients were diagnosed already with a very high tumor burden at the initial diagnosis without any pretreatment, whereas other patients had only a local tumor burden as measured by PSMA PET. Third, planar BS were used to be consistent with the CHAARTED and STAMPEDE trial. Of note, many institutions are using SPECT/CT systematically to increase specificity and positive predictive value of BS. Fourth, the potential of false-positive results by PSMA PET cannot be ruled out. Fifth, there was a time interval of 100 d between BS and PSMA PET: a potential progression of disease within this time interval cannot be excluded. Sixth, we excluded patients (n = 44) that had locoregional disease only on CI. Seventh, as the visual analysis of the BS following the EXINI version 2.0 analysis was performed by one reader only, we cannot rule out false-positive BS. Eighth, we included for the PSMA PET analysis only WB PSMA-TV combined with a stratification into more than 10 and less than 10 bone metastases. Here, it would be interesting to include in further analyses the exact numbers of bone, lymph node (stratified into N1 and M1a), and visceral metastases on a per-patient basis.
CONCLUSION
A discordance between PSMA PET and CI findings could be observed when applying CHAARTED criteria to both these modalities, whereas this migration between LVD and HVD from CI to PSMA PET occurs by both upstaging and downstaging. The association of the discordances between CI and PSMA PET–based tumor burden with survival outcomes remains unknown. Integration of PSMA PET in major industry-sponsored clinical trials for secondary endpoint analyses is warranted. PSMA PET/CT–based definitions of HVD and LVD may be redefined when correlating with patient outcome.
DISCLOSURE
Lena Unterrainer reports funding from Bavarian Cancer Research Center (BZKF) and from the Munich Clinician Scientist Program (MCSP). She reports fees from Novartis (speaker), Telix (consultant), and Astellas Pharma Inc. (speaker) outside of the submitted work. Wolfgang Fendler reports fees from SOFIE Bioscience (research funding), Janssen (consultant, speaker), Calyx (consultant, image review), Bayer (consultant, speaker, research funding), Novartis (speaker, consultant), Telix (speaker), GE HealthCare (speaker), Eczacıbaşı Monrol (speaker), ABX (speaker), Amgen (speaker), and Urotrials (speaker) outside of the submitted work. Jeremie Calais acknowledges grant support from the Prostate Cancer Foundation (PCF). Jeremie Calais reports grants from support to his institution from Lantheus, Novartis, and POINT Biopharma. He also reports consulting activities (advisory boards, speaker, blinded reader) for Advanced Accelerator Applications, Amgen, Astellas, Bayer, Blue Earth Diagnostics Inc., Curium Pharma, Coretag, DS Pharma, Fibrogen, GE HealthCare, Isoray, IBA RadioPharma, Janssen Pharmaceuticals, Monrol, Lightpoint Medical, Lantheus, Novartis, Nucleus Radiopharma, Pfizer, POINT Biopharma, Progenics, Radiomedix, Radiopharm Theranostics, Sanofi, Siemens-Varian, SOFIE, and Telix Pharmaceuticals outside of the submitted work. Johannes Czernin reports the following disclosures: SOFIE Biosciences (founder), Trethera Corporation (founder), and Atkis Oncology (medical advisory board). Mike Sathekge reports the following conflicts of interest: Novartis (advisory board); Adcock, AHRI, and NuMeRI (board of directors); corporate-sponsored research with Novartis, ABX, NTP, and Point Biopharma Global, Inc. Ken Herrmann reports personal fees from Bayer, personal fees and other from SOFIE Biosciences, personal fees from SIRTEX, nonfinancial support from ABX, personal fees from Adacap, Curium, and Endocyte, grants and personal fees from BTG, personal fees from IPSEN, Siemens Healthineers, GE HealthCare, Amgen, Novartis, ymabs, Aktis Oncology, Theragnostics, Pharma15, Debiopharm, AstraZeneca, and Janssen outside of the submitted work. Thomas Hope has grant funding to the institution from Bayer, Clovis Oncology, GE HealthCare, Lantheus, Janssen, Novartis, Telix Pharmaceuticals, the Prostate Cancer Foundation, and the National Cancer Institute (R01CA235741 and R01CA212148). He received personal fees from Bayer, Cardinal Health, BlueEarth Diagnostics, Lantheus, RayzeBio, and Sanofi and received fees from and has an equity interest in Curium and AdvanCell. No other potential conflict of interest relevant to this article was reported.
REFERENCES
- 1.Fizazi K, Foulon S, Carles J, et al. Abiraterone plus prednisone added to androgen deprivation therapy and docetaxel in de novo metastatic castration-sensitive prostate cancer (PEACE-1): a multicentre, open-label, randomised, phase 3 study with a 2 × 2 factorial design. Lancet. 2022;399:1695–1707. [DOI] [PubMed] [Google Scholar]
- 2.Smith MR, Saad F, Hussain M, et al. ARASENS: a phase 3 trial of darolutamide in combination with docetaxel for men with metastatic hormone-sensitive prostate cancer (mHSPC). J Clin Oncol. 2018;36(suppl 6):TPS383. [Google Scholar]
- 3.James ND, de Bono JS, Spears MR, et al. Abiraterone for prostate cancer not previously treated with hormone therapy. N Engl J Med. 2017;377:338–351. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Sweeney CJ, Chen Y-H, Carducci M, et al. Chemohormonal therapy in metastatic hormone-sensitive prostate cancer. N Engl J Med. 2015;373:737–746. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Chi KN, Chowdhury S, Bjartell A, et al. Apalutamide in patients with metastatic castration-sensitive prostate cancer: final survival analysis of the randomized, double-blind, phase III TITAN study. J Clin Oncol. 2021;39:2294–2303. [DOI] [PubMed] [Google Scholar]
- 6.Armstrong AJ, Szmulewitz RZ, Petrylak DP, et al. ARCHES: efficacy of androgen deprivation therapy (ADT) with enzalutamide (ENZA) or placebo (PBO) in metastatic hormone-sensitive prostate cancer (mHSPC). J Clin Oncol. 2019;37(suppl 15):5048. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Sweeney CJ, Martin AJ, Stockler MR, et al. Testosterone suppression plus enzalutamide versus testosterone suppression plus standard antiandrogen therapy for metastatic hormone-sensitive prostate cancer (ENZAMET): an international, open-label, randomised, phase 3 trial. Lancet Oncol. 2023;24:323–334. [DOI] [PubMed] [Google Scholar]
- 8.Vale CL, Fisher DJ, Godolphin PJ, et al. Which patients with metastatic hormone-sensitive prostate cancer benefit from docetaxel: a systematic review and meta-analysis of individual participant data from randomised trials. Lancet Oncol. 2023;24:783–797. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Gravis G, Boher J-M, Joly F, et al. Androgen deprivation therapy (ADT) plus docetaxel versus ADT alone in metastatic non castrate prostate cancer: impact of metastatic burden and long-term survival analysis of the randomized phase 3 GETUG-AFU15 trial. Eur Urol. 2016;70:256–262. [DOI] [PubMed] [Google Scholar]
- 10.Burdett S, Boevé LM, Ingleby FC, et al. Prostate radiotherapy for metastatic hormone-sensitive prostate cancer: a STOPCAP systematic review and meta-analysis. Eur Urol. 2019;76:115–124. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Barbato F, Fendler WP, Rauscher I, et al. PSMA PET for the assessment of metastatic hormone-sensitive prostate cancer volume of disease. J Nucl Med. 2021;62:1747–1750. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Fendler WP, Calais J, Eiber M, et al. Assessment of 68Ga-PSMA-11 PET accuracy in localizing recurrent prostate cancer: a prospective single-arm clinical trial. JAMA Oncol. 2019;5:856–863. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Fendler WP, Weber M, Iravani A, et al. Prostate-specific membrane antigen ligand positron emission tomography in men with nonmetastatic castration-resistant prostate cancer. Clin Cancer Res. 2019;25:7448–7454. [DOI] [PubMed] [Google Scholar]
- 14.Hofman MS, Lawrentschuk N, Francis RJ, et al. Prostate-specific membrane antigen PET-CT in patients with high-risk prostate cancer before curative-intent surgery or radiotherapy (proPSMA): a prospective, randomised, multicentre study. Lancet. 2020;395:1208–1216. [DOI] [PubMed] [Google Scholar]
- 15.Seifert R, Herrmann K, Kleesiek J, et al. Semiautomatically quantified tumor volume using 68Ga-PSMA-11 PET as a biomarker for survival in patients with advanced prostate cancer. J Nucl Med. 2020;61:1786–1792. [DOI] [PubMed] [Google Scholar]
- 16.Evangelista L, Bertoldo F, Boccardo F, et al. Diagnostic imaging to detect and evaluate response to therapy in bone metastases from prostate cancer: current modalities and new horizons. Eur J Nucl Med Mol Imaging. 2016;43:1546–1562. [DOI] [PubMed] [Google Scholar]
- 17.Mota JM, Armstrong AJ, Larson SM, Fox JJ, Morris MJ. Measuring the unmeasurable: automated bone scan index as a quantitative endpoint in prostate cancer clinical trials. Prostate Cancer Prostatic Dis. 2019;22:522–530. [DOI] [PubMed] [Google Scholar]
- 18.Poulsen MH, Rasmussen J, Edenbrandt L, et al. Bone scan index predicts outcome in patients with metastatic hormone‐sensitive prostate cancer. BJU Int. 2016;117:748–753. [DOI] [PubMed] [Google Scholar]
- 19.Miyoshi Y, Yoneyama S, Kawahara T, et al. Prognostic value of the bone scan index using a computer-aided diagnosis system for bone scans in hormone-naive prostate cancer patients with bone metastases. BMC Cancer. 2016;16:128. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Mittlmeier LM, Brendel M, Beyer L, et al. Feasibility of different tumor delineation approaches for 18F-PSMA-1007 PET/CT imaging in prostate cancer patients. Front Oncol. 2021;11:663631. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Bodar Y, Luining W, Keizer B, et al. A prospective, multicenter head-to-head comparative study in patients with primary high-risk prostate cancer investigating the bone lesion detection of conventional imaging and 18F-PSMA-PET/CT. Urol Oncol. 2023;41:205.e17–205.e24. [DOI] [PubMed] [Google Scholar]
- 22.Donswijk ML, van Leeuwen PJ, Vegt E, et al. Clinical impact of PSMA PET/CT in primary prostate cancer compared to conventional nodal and distant staging: a retrospective single center study. BMC Cancer. 2020;20:723. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Murthy V, Sonni I, Jariwala N, et al. The role of PSMA PET/CT and PET/MRI in the initial staging of prostate cancer. Eur Urol Focus. 2021;7:258–266. [DOI] [PubMed] [Google Scholar]
- 24.Öbek C, Doğanca T, Demirci E, et al. The accuracy of 68Ga-PSMA PET/CT in primary lymph node staging in high-risk prostate cancer. Eur J Nucl Med Mol Imaging. 2017;44:1806–1812. [DOI] [PubMed] [Google Scholar]
- 25.Fizazi K. APCCC 2022: when low-volume on conventional imaging goes into high-volume on next-generation imaging in mHSPC: treat like low volume. Urotoday website. https://www.urotoday.com/conference-highlights/apccc-2022/136802-apccc-2022-when-low-volume-on-conventional-imaging-goes-into-high-volume-on-next-generation-imaging-in-mhspc-treat-like-low-volume.html. Updated September 6, 2022. Accessed December 4, 2024.
- 26.Fanti S. APCCC 2022: how to interpret and confirm lesions positive on PSMA PET but with no correlation on CT? Urotoday website. https://www.urotoday.com/conference-highlights/apccc-2022/136832-apccc-2022-how-to-interpret-and-confirm-lesions-positive-on-psma-pet-but-with-no-correlation-on-ct.html. Updated March 30, 2023. Accessed December 4, 2024.
- 27.Arnfield EG, Thomas PA, Roberts MJ, et al. Clinical insignificance of [18F]PSMA-1007 avid non-specific bone lesions: a retrospective evaluation. Eur J Nucl Med Mol Imaging. 2021;48:4495–4507. [DOI] [PubMed] [Google Scholar]
- 28.Seifert R, Telli T, Opitz M, et al. Unspecific 18F-PSMA-1007 bone uptake evaluated through PSMA-11 PET, bone scanning, and MRI triple validation in patients with biochemical recurrence of prostate cancer. J Nucl Med. 2023;64:738–743. [DOI] [PubMed] [Google Scholar]
- 29.Hope TA, Benz M, Jiang F, et al. Do bone scans overstage disease compared with PSMA PET at initial staging? An international multicenter retrospective study with masked independent readers. J Nucl Med. 2023;64:1744–1747. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Ayati N, Herrmann K, Fanti S, Murphy DG, Hofman MS. More accurate imaging is not stage migration: time to move from “Hubble” to “Webb” in hormone-sensitive prostate cancer. Eur Urol. 2023;83:6–9. [DOI] [PubMed] [Google Scholar]
- 31.Seifert R, Rasul S, Seitzer K, et al. A prognostic risk score for prostate cancer based on PSMA PET–derived organ-specific tumor volumes. Radiology. 2023;307:e222010. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Karpinski MJ, Hüsing J, Claassen K, et al. Combining PSMA-PET and PROMISE to re-define disease stage and risk in patients with prostate cancer: a multicentre retrospective study. Lancet Oncol. 2024;25:1188–1201. [DOI] [PubMed] [Google Scholar]