Visual Abstract
Abstract
Our objective was to determine the diagnostic value of prostate-specific membrane antigen (PSMA) PET/CT in staging men with newly diagnosed unfavorable intermediate-risk prostate cancer (PCa). Methods: Patients with newly diagnosed unfavorable intermediate-risk PCa, in whom PSMA PET/CT was performed as a primary staging modality, were retrospectively studied. PSMA PET/CT was performed at several diagnostic centers and reported by expert nuclear medicine physicians within 2 high-volume PCa centers. A multivariate logistic regression analysis, taking into account clinical, biochemical, pathologic, and radiologic variables, was performed to identify potential independent predictors for metastatic disease on PSMA PET/CT. Results: In total, 396 men with newly diagnosed unfavorable intermediate-risk PCa were studied. Metastatic disease was observed in 37 (9.3%) men, of whom 29 (7.3%) had molecular imaging locoregional lymph node metastases (miN1) and 16 (4.0%) had distant metastases (miM1). A radiologic tumor stage of at least T3 on MRI (odds ratio, 2.72 [95% CI, 1.27–5.83]; P = 0.01) and more than 50% positive prostate biopsies (odds ratio, 3.87 [95% CI, 1.74–8.62]; P = 0.001) were found to be independently associated with metastatic disease on PSMA PET/CT. Conclusion: Given that metastatic disease was observed in nearly 1 in 10 men with newly diagnosed unfavorable intermediate-risk PCa, PSMA PET/CT is considered to be of diagnostic value within this population. Further stratification using the radiologic tumor stage and the percentage of positive prostate biopsies could aid in identifying those patients at risk of having metastatic disease on PSMA PET/CT.
Metastatic screening using prostate-specific membrane antigen (PSMA) PET/CT in men with prostate cancer (PCa) is increasingly being adopted in urologic practice. Compared with conventional imaging techniques (such as CT and bone scintigraphy), PSMA PET/CT appears to have greater accuracy in detecting metastases, is more likely to lead to management changes, produces fewer equivocal findings, and has lower radiation exposure (1–3).
To avoid overdiagnosis, adequate selection of patients at risk of having metastatic disease is of paramount importance. The presence of metastases on PSMA PET/CT has been shown to be directly associated with prostate-specific antigen (PSA) levels, clinical tumor stages, and biopsy International Society of Urological Pathology (ISUP) grade groups (GGs) (3–6). Therefore, the decision on whether to perform PSMA PET/CT is based on the stratification of patients according to the American Urological Association risk groups (i.e., low, favorable intermediate, unfavorable intermediate, and high risk; Table 1) (7,8). Consequently, the Society of Nuclear Medicine and Molecular Imaging recommends performing metastatic screening using PSMA PET/CT in patients with unfavorable intermediate-risk and high-risk PCa (9).
American Urological Association Risk Groups for Localized PCa
Even though the diagnostic value of PSMA PET/CT in men with high-risk PCa has already been demonstrated, the clinical benefit of PSMA PET/CT in men with newly diagnosed unfavorable intermediate-risk PCa remains questionable and literature on this topic is scarce (3,4,10). Therefore, this study aimed to determine the diagnostic value of PSMA PET/CT for metastatic screening within the largest known dataset of newly diagnosed unfavorable intermediate-risk PCa patients.
MATERIALS AND METHODS
Study Design and Patient Population
A retrospective cohort study was performed at The Netherlands Cancer Institute and the Amsterdam University Medical Center. The institutional review board approved this retrospective study and waived the requirement to obtain informed consent.
Patients found to have unfavorable PCa on diagnostic prostate biopsies and on whom PSMA PET/CT was performed between January 2018 and December 2021 as a primary staging modality were retrospectively studied. Excluded from the analysis were patients with any high-risk factors, such as a PSA level of at least 20 ng/mL or a clinical tumor stage of at least T3a on digital rectal examination. Patients who had received prior PCa treatments were excluded as well.
Patient characteristics were collected from patient charts, and PSMA PET/CT results were collected from nuclear medicine reports.
MRI Acquisition and Analysis
Before undergoing prostate biopsies, the men underwent MRI. Prostate MRI was performed using a 3.0-T scanner according to the European Society of Urological Radiology MRI protocol in high-volume diagnostic PCa centers (11). All MRI scans were reviewed by experienced radiologists according to the standardized Prostate Imaging Reporting and Data System (PI-RADS) classification system, version 2.1 (12). Following Dutch guidelines, all radiologists were experienced enough to independently read prostate MR images (13).
Prostate Biopsies and Histopathologic Assessment
All prostate biopsies took place in high-volume diagnostic PCa centers within the Prostate Cancer Network Netherlands by experienced urologists with at least 5 y of experience performing prostate biopsies. The biopsies were performed transrectally or transperineally, using either MRI/transrectal ultrasound fusion software or cognitive fusion. If an MRI-positive lesion (defined as PI-RADS classification 3–5) was visible, systematic prostate biopsies were preceded by targeted prostate biopsies. If MRI had negative findings (defined as PI-RADS classification 1–2) or was not performed, only systematic prostate biopsies were obtained. Given the fact that not all hospitals were yet performing targeted prostate biopsies, some MRI-positive patients underwent only systematic prostate biopsies. Prostate biopsies were processed and reported according to the ISUP protocol by specialized uropathologists with at least 5 y of experience in assessing prostate biopsies (14). MRI-directed targeted prostate biopsies were pooled and reported as 1 biopsy core in a systematic biopsy set. Before PSMA PET/CT was performed, prostate biopsies were not reassessed. If the presence of a cribriform growth pattern was reported, then it was considered.
PSMA PET/CT Imaging and Analysis
Although PSMA PET/CT was performed at several diagnostic centers, all scans were revised and reported by expert nuclear medicine physicians within 2 high-volume PCa centers (Netherlands Cancer Institute and Amsterdam University Medical Center). PSMA PET/CT images were analyzed and classified according to the European Association of Nuclear Medicine PSMA guidelines (15). A PSMA PET/CT scan was reported as positive when the lesion was compatible with or suggestive of PCa. According to the molecular imaging classification (mi), locoregional lymph node metastases (miN1) were defined as lymph node metastases in the pelvic region, whereas distant metastases (miM1) were defined as either extrapelvic lymph node (miM1a), bone (miM1b), or visceral metastases (miM1c).
For logistic reasons, several PSMA radiotracers were used for metastatic screening during the study time frame: [18F]DCFPyL, [18F]PSMA-1007, [18F]PSMA-JK7, and [68Ga]PSMA-11. PET images were acquired from mid thigh to skull base or vertex. Incubation times and doses differed among tracers and sites, according to local protocols: a median of 99 min (interquartile range [IQR], 60–120 min) after a median dose of 280 MBq (IQR, 198–313 MBq) for [18F]DCFPyL, a median of 90 min (IQR, 90–120 min) after a median dose of 258 MBq (IQR, 208–293 MBq) for [18F]PSMA-1007, a median of 77 min (IQR, 60–90 min) after a median dose of 201 MBq (IQR, 194–265 MBq) for [18F]PSMA-JK7, and a median of 49 min (IQR, 45–60 min) after a median dose of 133 MBq (IQR, 103–154 MBq) for [68Ga]PSMA-11. PET images were combined with either a low-dose CT scan (120–140 kV, 40–80 mAs) or a diagnostic CT scan (130 kV, 110 mAs) for anatomic correlation and attenuation correction.
Statistical Analysis
Categoric variables were reported as frequency distributions and percentages, and continuous variables were expressed as medians with IQR.
To study the diagnostic value of PSMA PET/CT, the number needed to image (NNI) was assessed. The NNI was defined as the number of PSMA PET/CT imaging examinations performed for every PSMA PET/CT scan with metastatic disease (i.e., miN1 and higher) (16). Univariate and multivariate logistic regression analyses were performed to identify possible predictive variables for metastatic disease on PSMA PET/CT imaging. First, univariate analyses were performed to assess the association between clinical, biochemical, pathologic, and radiologic variables and metastatic disease on PSMA PET/CT imaging. Variables included in the model were age, PSA level, prostate volume, PSA density, radiologic tumor stage on MRI, PI-RADS assessment category, percentage of positive prostate biopsies, percentage of ISUP GG3 PCa in prostate biopsies, and cribriform growth. Univariate statistically significant variables were applied in the multivariate conditional logistic regression analysis using a backward stepwise variable selection method. For clinical applicability, the percentage of positive prostate biopsies and the percentage of ISUP GG3 in prostate biopsies were used as binary (≤50% and >50%) and categoric (≤25%, 25%–50%, 50%–75%, and >75%) variables. Subsequently, through further stratification in subgroups, precisely those patients at risk for metastatic disease on PSMA PET/CT were identified.
A P value of less than 0.05 was considered to indicate statistical significance. All statistical analyses were performed with the statistical package SPSS, version 27 (IBM), for MacOS (Apple).
RESULTS
In total, 396 men with newly diagnosed unfavorable intermediate-risk PCa were studied. Patient characteristics and PSMA PET/CT results are presented in Tables 2 and 3, respectively. PSMA-expressing metastases were observed in 37 (9.3%) men, resulting in an NNI of approximately 10. No patients were found to have miM1c. Metastatic disease was observed significantly more often in men undergoing PSMA PET/CT scans using [18F]PSMA-1007 than using the other PSMA tracers (P = 0.01; Table 3). No statistically significant differences regarding clinical variables were observed among the applied PSMA radiotracers.
Patient Characteristics of Both Entire Study Population and Subgroups With and Without PSMA-Expressing Metastases
PSMA PET/CT Results for All Unfavorable Intermediate-Risk PCa Patients by Applied PSMA Radiotracer
Univariate and Multivariate Logistic Regression Analysis
Univariate logistic regression analyses showed that initial PSA level (P = 0.001), radiologic tumor stage on MRI (P = 0.001), percentage of positive prostate biopsies (P < 0.001), and percentage of ISUP GG3 PCa in prostate biopsies (P = 0.001) were associated with the presence of metastatic disease on PSMA PET/CT (Table 2). After backward selection, 2 variables remained independently statistically significantly associated with the presence of PSMA-expressing metastases (Table 4). More precisely, a radiologic tumor stage of at least T3 (odds ratio, 2.72 [95% CI, 1.27–5.83]; P = 0.01) and more than 50% positive prostate biopsies (odds ratio, 3.87 [95% CI, 1.74–8.62]; P = 0.001) were found to be associated with metastatic disease on PSMA PET/CT.
Detection of PSMA-Expressing Locoregional Lymph Node or Distant Metastases on PET/CT Multivariate Logistic Regression Analysis
The Prostate Cancer Network Netherlands Subclassification
PSMA-expressing metastases were particularly common in men with a radiologic tumor stage of at least T3a and higher proportions of positive prostate biopsies. Especially men with more than 75% positive prostate biopsies were at increased risk of having metastatic disease on PSMA PET/CT (odds ratio, 3.88 [95% CI, 1.23–12.21]; P = 0.02; Supplemental Table 1; supplemental materials are available at http://jnm.snmjournals.org). Therefore, further risk stratification was investigated.
By further stratification, according to radiologic tumor stage and percentage of positive prostate biopsies, patients at higher risk of metastatic disease on PSMA PET/CT could be identified. Three subgroups were created: low, moderate, and high metastatic potential (Table 5). The NNIs were 32, 10, and 5 for patients with low, moderate, and high metastatic potential, respectively. Patients with high metastatic potential might benefit most from metastatic screening with PSMA PET/CT. The incidence of PSMA-expressing metastases was significantly higher in patients with high metastatic potential than in those with low metastatic potential (18.5% vs. 3.0%; P < 0.001). No significant difference was observed between patients with moderate and high metastatic potential (9.4% vs. 18.5%; P = 0.18).
Incidence of Metastatic Disease and NNI According to Prostate Cancer Network Netherlands Subclassification
DISCUSSION
We evaluated the diagnostic value of PSMA PET/CT within the largest known dataset of newly diagnosed unfavorable intermediate-risk PCa patients. Given that 9.3% of patients had PSMA-expressing metastases, resulting in an NNI of approximately 10, this study supports the diagnostic value of PSMA PET/CT within this specific yet important patient population. However, as a relatively expensive diagnostic tool with only limited availability, careful consideration should be given to who will benefit most from PSMA PET/CT (17). In this study, we showed that a radiologically more advanced stage of disease on MRI and a higher proportion of positive prostate biopsies were independently associated with metastatic disease on PSMA PET/CT.
Even though the proPSMA study has demonstrated the diagnostic value of PSMA PET/CT in men with newly diagnosed ISUP GG3 PCa, data on its diagnostic value within the intermediate-risk group, and in particular the unfavorable intermediate-risk group, are lacking (1). There are only a few reports on the incidence of PSMA-expressing metastases within the intermediate-risk PCa population, which do not elaborate on the incidence of metastases within the different prognostic subgroups (3,6,18–21). In the retrospective cohort study by Yaxley et al., the incidence of PSMA-expressing metastases was studied in 638 intermediate-risk PCa patients. Metastatic disease was identified in 5.2% of the patients (3). On the other hand, Chikatamarla et al. found higher incidence rates for metastatic disease (6). Within their intermediate-risk PCa population, 8.5% (6/71) of men had bone metastases on PSMA PET/CT. Their higher metastatic rates may be attributed to the PSMA radiotracer used, namely [18F]PSMA-1007, which is known to have higher rates of nonspecific bone lesions (22–24). This was likewise observed in our cohort, as 20.8% of men with unfavorable intermediate-risk PCa had bone metastases. Their lower observed incidence of metastatic disease, compared with our study, may be explained by the fact that Yaxley et al. and Chikatamarla et al. included both favorable and unfavorable intermediate-risk PCa patients. Therefore, the incidence of metastatic disease in their unfavorable intermediate-risk populations might be higher and hence similar to ours.
In our study, a radiologic stage of at least T3 was found to be a statistically independent prognostic factor for PSMA-expressing metastases. Similar outcomes were shown by Yaxley et al. as they reported metastatic disease in 35.2% of men with radiologic stages of at least T3a (2). However, this increased incidence was observed in a population with low-, intermediate- and high-risk patients, and specific data regarding the unfavorable intermediate-risk PCa population were lacking. The clinical benefit of the radiologic tumor stage for stratification of patients before metastatic screening may be partially explained by the improved ability of local staging by MRI compared with digital rectal examination and the predictive value of the tumor stage on the presence of nodal metastases (25,26).
Along with the radiologic tumor stage, the proportion of positive prostate biopsies was associated with the presence of metastatic disease on PSMA PET/CT in our study. This is equally the case for the proportion of ISUP GG3 in prostate biopsies and in line with preexisting literature (27). The presence of these associations may be explained by potential improper grading at the time of biopsy. The proportion of highest ISUP GG in biopsies has been shown to be a significant predictor of both up- and downgrading in radical prostatectomy—the lower the proportion, the greater the likelihood of downgrading and, consequently, an inaccurate indication (28). The fact that both targeted and systematic biopsies were taken in most men may also have influenced our unfavorable intermediate-risk group, as the original American Urological Association risk groups were based on ISUP GG from systematic sextant biopsies. Systematic biopsies are more frequently upgraded at radical prostatectomy than targeted biopsies (29).
Our study is not devoid of limitations. First, given the retrospective nature of this study, the presence of selection bias cannot be ruled out, meaning results should be interpreted with caution. Second, the lack of histopathologic reevaluation when prostate biopsies were not performed within the Netherlands Cancer Institute and Amsterdam University Medical Center may have influenced the ISUP GG and potentially our results. Furthermore, given that PSMA PET/CT scans were performed in several hospitals, scan protocols may have differed, which might have influenced our findings. In addition, multiple PSMA radiotracers were used within this cohort. Since experience with particular PSMA radiotracers is essential for interpreting scans, the variety of tracers used may have influenced the assessments (24). Finally, because of the lack of histopathologic confirmation, it remains unclear whether these PSMA-avid metastases were indeed PCa metastases. Considering the limitations discussed, future prospective studies are needed to further investigate the diagnostic value of PSMA PET/CT within this specific yet important patient population.
CONCLUSION
PSMA PET/CT is of diagnostic value in men with newly diagnosed unfavorable intermediate-risk PCa. Our data show that nearly 1 in 10 men who underwent metastatic screening with PSMA PET/CT showed metastatic disease. Especially men with more advanced radiologic tumor stages and higher percentages of positive prostate biopsies might benefit most from PSMA-based metastatic screening. On the basis of our analyses, further stratification of unfavorable intermediate-risk PCa patients using the proposed Prostate Cancer Network Netherlands subclassification could aid in identifying those patients at risk of having metastatic disease on PSMA PET/CT. However, given the paucity of data on the diagnostic value of PSMA PET/CT within this distinct patient population, further research is needed to support and externally validate our findings.
DISCLOSURE
No potential conflict of interest relevant to this article was reported.
KEY POINTS
QUESTION: Should PSMA PET/CT be recommended in men with newly diagnosed unfavorable intermediate-risk PCa?
PERTINENT FINDINGS: In a retrospective analysis with 396 newly diagnosed unfavorable intermediate-risk PCa patients, PSMA-expressing metastases were observed in 37 (9.3%) men. A radiologically more advanced stage of disease on MRI and a higher percentage of positive prostate biopsies were found to be independently associated with the presence of PSMA-expressing metastases.
IMPLICATIONS FOR PATIENT CARE: Further stratification of unfavorable intermediate-risk PCa patients using the proposed Prostate Cancer Network Netherlands subclassification could aid in identifying those patients at risk of having metastatic disease on PSMA PET/CT.
Footnotes
Published online Jun. 29, 2023.
- © 2023 by the Society of Nuclear Medicine and Molecular Imaging.
REFERENCES
- Received for publication November 16, 2022.
- Revision received February 21, 2023.
