Prostate cancer (PCa) is the most common type of cancer in men, with an incidence of one in six in developed countries (1). Prostate-specific membrane antigen (PSMA) is a type II transmembrane glycoprotein that is used as a radiopharmaceutical in positron-emission tomography/computed tomography (PET/CT) scans (2, 3). It can be expressed in various anatomical sites, such as the salivary glands, renal tubules, glial cells of the central nervous system, the small intestine and the prostate gland, where high levels of PSMA are expressed in the presence of PCa (4–6). The extracellular domain of PSMA has a high affinity for urea-based ligands (7), which has led to the development of various ligands labeled with radioisotopes for diagnostic purposes, such as gallium-68 [⁶⁸Ga] and fluorine-18 [¹⁸F], and for theranostic purposes, such as lutetium-177 [¹⁷⁷Lu] (8, 9).

Currently, PSMA PET/CT in combination with various agents constitute one of the best imaging diagnostic methods for patients with a high or very high risk of PCa with suspected metastasis (10). In particular, [¹⁸F] labeled PSMA has demonstrated safety, expected tissue biodistribution, and significant PCa-specific uptake (11). Evidence shows a detection rate for PCa with biochemical recurrence of up to 86%, with a sensitivity of 61.4% and a specificity of 88.3% for [¹⁸F]PSMA PET/CT (12, 13). Additionally, [¹⁸F]PSMA-1007 generates images with higher spatial resolution than [⁶⁸Ga]Ga-PSMA-11 and [¹⁸F]PSMA-DCFPyL, improving the tumor-to-background ratio, increasing the radiopharmaceutical decay half-life (110 min), and achieving better lesion delineation at the pelvic level (14).

Importantly PSMA PET/CT yields a semiquantitative measurement of the standardized uptake value (SUV) of the radiotracer, which is useful in oncology for detecting primary or metastatic sites of PCa, differentiating between benign and malignant tumors, comparing lesions seen on CT scans, planning treatment, monitoring the response to treatment and/or detecting recurrent disease (15). Standardized reporting systems, particularly the PSMA-RADS (16) proposed by the American College of Radiology or the E-PSMA (17) supported by the European Association of Nuclear Medicine, were used to describe these results.

The aim of this study was to determine the agreement between the PSMA-RADS and E-PSMA standardized reporting systems in the classification of [¹⁸F]PSMA-1007–uptaking lesions identified on PET/CT scans in patients with PCa and post-prostatectomy with suspected recurrent disease (local recurrence, regional nodal involvement and distant metastases), based on biochemical recurrence, while also exploring the correlation between lesion size and tracer uptake.

In the retrospective analysis, a total of 32 postoperative PCa patients who underwent [¹⁸F]PSMA-1007 PET/CT were included. The mean (± SD) age of the patients was 69.98 ± 8.27 years (range 53.27–86.32 years). The characteristics of the patients are detailed in Table 2. All patients received care from the Urology Oncology Service, and [¹⁸F]PSMA-1007 PET/CT scans were performed because of BCR. No patient met the exclusion criteria, which included noncontrast PET/CT images showing movement artifacts, the presence of diuretic treatment during the PET/CT scan, and kidney or liver disease.

In addition, we considered a total of n = 149 lesions with avid uptake of [¹⁸F]PSMA-1007; the mean (± SD) number of lesions per patient was 4.66 (± 3.49, range 0–11). Figure 2 shows the [¹⁸F]PSMA-1007 uptake (SUVmax) of lesions per patient, with the size of the lesions depicted in mm, grouped by anatomical site. The distribution of the avid [¹⁸F]PSMA-1007 uptake sites and the characteristics (size and SUVmax) of the lesions are detailed in Table 3. Notably, of the 32 patients analyzed, 2/32 (6.25%) did not present with lesions with avid [¹⁸F]PSMA-1007 uptake. The largest metastatic PCa lesions, in terms of size, were found within the surgical bed, whereas lesions were found at other anatomical sites (p < 0.001); moreover we found no evidence of a difference in the SUVmax according to anatomical site (p = 0.219).

The Spearman correlation coefficients (r) between lesion size (mm) and SUVmax for [¹⁸F]PSMA-1007 overall and grouped by lesion anatomical site are shown in Table 4. When considering all the lesions, a correlation was identified (r = 0.358, p < 0.001), but when grouping the lesions by anatomical site, a correlation was observed only for the lymph node lesions (r = 0.516, p < 0.001).

Of the total number of lesions (n = 149) among the 30 patients, only those corresponding to 28 patients were classified with both the PSMA-RADS and the E-PSMA systems (n = 137); the E-PSMA was not used for two patients whose uptake levels in the spleen were higher than those in the parotid gland, making it impossible to determine the miPSMA expression of the associated lesions. Table 5 shows the frequency and percentage of each of the lesion classification categories based on the PSMA-RADS and the E-PSMA classification systems; 78/137 (56.93%) of the lesions were classified as PSMA-RADS-5, and 79/137 (57.66%) of the lesions were classified as E-PSMA-5. Additionally, substantial agreement was detected between the two systems (κ = 0.70, p < 0.001), but there was disagreement between the PSMA-RADS and E-PSMA classification systems in 24/137 (17.52%) of the lesions; of note, 18/137 (13.14%) were classified as PSMA-RADS grade 4 but E-PSMA grade 3.

We observed that in 5 of the 24 discordantly classified lesions, E-PSMA yielded a higher classification level than did PSMA-RADS. These cases included degenerative changes in the cervical vertebral body in one patient, a high-uptake bladder abscess in another patient, thickening of the rectum with high uptake in a third patient and a pair of nodes in the final patient with a short-axis diameter of 8 mm that were morphologically preserved yet demonstrated radiotracer uptake.

Among the other 19 of the 24 discordant lesions between the systems, PSMA-RADS produced a higher classification than did the E-PSMA system. These lesions had the following characteristics: a lymph node lesion with a short-axis diameter of 7 mm that demonstrated morphological changes and a miPSMA score of 3, and 18 lesions with a PSMA-RADS grade of 4 but an E-PSMA grade classified as 3, all of which lacked morphological changes with miPSMA uptake of 1 (that is, uptake between that of the blood pool and the spleen). Of these 18 lesions, seven were in lymph nodes with a short axis diameter ranging from 3 to 6 mm, seven were in bone, three were in the surgical bed and one was in the bladder.

PSMA PET/CT could substantially impact the clinical management of PCa patients based on its diagnostic accuracy (23). Recent studies have used [¹⁸F]PSMA-1007 to detect PCa, identify the presence of metastases before treatment (24) and to determine BCR after treatment (25). Ongoing initiatives are underway to evaluate the reliability and practicality of standardized reporting systems for PSMA PET/CT scans in PCa patients (26), providing external validation. The use of lesion classification systems like PSMA-RADS and E-PSMA, which rely on criteria such as location, morphology, size, and radioisotope uptake (SUVmax), aims to mitigate potential confirmation bias. Nevertheless, incorporating the patient’s clinical history during the study may inadvertently lead to focusing on specific body systems or organs, potentially neglecting other areas. This approach could lead to the omission of small lesions, highlighting the critical importance of conducting a thorough and impartial evaluation of all potential abnormalities.

Our study expands the knowledge about the presence of metastases in patients with PCa post-prostatectomy, indicating that the largest metastatic PCa lesions, in terms of size, were found within the surgical bed when compared to lesions in other anatomical sites (p < 0.001). Regarding lymph node lesions, our study revealed a positive correlation (r = 0.516, p < 0.001) between the size of these lesions and their [¹⁸F]PSMA-1007 uptake. Furthermore, we identified substantial agreement between the PSMA-RADS and E-PSMA classification system scores among all lesions (κ = 0.70, p < 0.001), and the greatest disagreement between the two scores occurred mainly among lymph node lesions.

In cases where E-PSMA resulted in a higher classification level than PSMA-RADS for lesions, these differences can be attributed to the fact that the PSMA-RADS provides the flexibility to categorize lesions that are typically benign with and without uptake, apart from identifying likely malignant nonprostate lesions. Additionally, the PSMA-RADS system does not incorporate the criterion of an 8 mm size for lymph node lesions as proposed in the E-PSMA system. Regarding lesions with discordant classifications where PSMA-RADS assigned a higher classification than did E-PSMA, these disparities can be elucidated by considering that the E-PSMA system accounts for uptake levels based on reference values (miPSMA) and suggests a standardized size criterion of 8 mm for lymph node lesions.

Some points to consider regarding these discordances in lesion classification include the fact that the uptake of PSMA ligands has been reported in a variety of benign conditions associated with osteoblastic activity, including osteoarthritis, degenerative changes, fibrous dysplasia, consolidation fractures and post radiotherapy (27). In addition, bone changes on CT scans are frequently absent. The uptake of PSMA ligands may be associated with endothelial cell neovascularity as well as the high permeability of inflammatory cells (28). Moreover, Luo et al. (29) reported excellent performance in the detection of PCa using [¹⁸F]PSMA-1007 PET/CT, which was validated by histopathology. Additionally, they determined that an optimal SUVmax threshold of 8.3 could be applied to identify PCa lesions through [¹⁸F]PSMA-1007 PET/CT.

Regarding discordant lesions in the lymph nodes, recent studies, such as the study by Gottlieb et al. (30), revealed that histopathologically, the tumor burden in lymph node lesions is associated with BCR-free survival time in PCa patients. Furthermore, Schwartz et al. (31) described that the stratification of lymph node lesions based on size and the anatomical region can hinder the serial follow-up of the lesions and potentially provide discordant findings in their classification. However, there is controversy regarding whether the size of the short axis or the long axis should be measured. On the one hand, the Prostate Cancer Clinical Trials Working Group (32) proposed a long axis ≥ 20 mm as a strict criterion for a lesion in a metastatic lymph node in PCa; on the other hand, Hövels et al. (33) suggested that lesions in lymph nodes with a short axis > 8 mm in the pelvis and > 10 mm outside the pelvis can be considered malignant in PCa. Finally, Schwartz et al. (31) reported that when only the short axis of the lymph node lesion is considered, there is a better and more significant correlation with the tumor burden, and they recommend adopting the measurement of the short-axis diameter for lymph node lesions in radiological practice.

This study has certain limitations. First, this was a cross-sectional study, so causality cannot be inferred. Second, the results are based on data from a single center with a small sample size, and from both adult patients (n = 32) and lesions (n = 149) among a population with post-prostatectomy PCa and suspected recurrent disease; therefore, the results should be interpreted with caution. Third, histopathological confirmation of the lesions was lacking but we define recurrent disease (local recurrence, regional nodal involvement and distant metastases) based on BCR post prostatectomy (at least two prostate-specific antigen (PSA) values that are 0.2 ng/mL or higher).

In future perspectives, radiomics emerges as a promising tool for the detection and categorization of PCa lesions (34). Radiomic signatures, generated through the amalgamation of information from PET/CT and PSMA, hold the potential to provide complementary insights into the detection and localization of PCa lesions, as well as predicting PSMA-RADS or E-PSMA outcomes. To advance our understanding, it is necessary to carry out longitudinal multicenter studies that allow lesion follow-up and assessment of their response to treatment, especially in cases where lesions are discordant between the PSMA-RADS and E-PSMA classification systems, particularly with an emphasis on histopathological confirmation.

Our study expands the knowledge about the presence of recurrent disease (local recurrence, regional nodal involvement and distant metastases) based on BCR in patients with PCa and postprostatectomy, revealing that the largest lesions were located in the surgical bed, while the lesions with the highest uptake of [¹⁸F]PSMA-1007 were located in the lymph nodes. Regarding lymph node lesions, our study generates evidence of a positive correlation between the size of these lesions and [¹⁸F]PSMA-1007 uptake. Furthermore, we identified that although there was substantial agreement between the two lesion classification systems (PSMA-RADS and E-PSMA), there were discrepancies, mainly among the lymph node lesions. From clinical practice, our study suggests that lymph node lesions require special attention in their classification to ensure a correct diagnosis and for better decision-making regarding their respective management. Finally, it is necessary to carry out longitudinal multicenter studies that allow lesion follow-up and assessment of their response to treatment, including their histopathological confirmation, especially those lesions that were discordant between the PSMA-RADS and E-PSMA classification systems.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

The studies involving humans were approved by Research and Research-Ethics committees of the HRAEB (registration numbers: CI-HRAEB 019-2023 and CEI-021-2023). The studies were conducted in accordance with the local legislation and institutional requirements. The ethics committee/institutional review board waived the requirement of written informed consent for participation from the participants or the participants’ legal guardians/next of kin given the retrospective nature of the study. The confidentiality of the data was meticulously maintained, and all procedures adhered to the pertinent guidelines and regulations governing this research.

MM-Á: Conceptualization, Formal analysis, Methodology, Project administration, Software, Supervision, Validation, Writing – original draft, Writing – review & editing. HE-P: Conceptualization, Formal analysis, Methodology, Project administration, Supervision, Validation, Writing – original draft, Writing – review & editing. JC-L: Conceptualization, Data curation, Methodology, Writing – original draft, Writing – review & editing. ER-M: Conceptualization, Formal analysis, Methodology, Project administration, Software, Supervision, Validation, Writing – original draft, Writing – review & editing.