Abstract

Objectives

Delayed [⁶⁸Ga]Ga-prostate-specific membrane antigen (PSMA) positron emission tomography/computed tomography (PET/CT) images show reduced PSMA uptake in benign lesions and increased PSMA uptake in malignant lesions. This study investigated the efficacy of PRIMARY scoring on [⁶⁸Ga]Ga-PSMA PET/CT images at standard versus delayed time points and assessed the potential added value of delayed imaging in PRIMARY scoring.

Methods

A total of 140 patients with biopsy results of International Society of Urological Pathology grade groups (ISUP) 1-2 who had standard (median 60 min) and delayed images (median 138 min) with [⁶⁸Ga]Ga-PSMA PET/CT before radical prostatectomy were included. Results were confirmed in pathological reports. For diagnostic parameters, two experienced nuclear medicine physicians, who were blinded to clinical data, independently reviewed the images, and a third physician provided consensus in cases of disagreement. PRIMARY scoring was also conducted by four nuclear medicine physicians on both images, with a 1-month interval between assessments for intraobserver agreement analyses.

Results

The percentage of lesions scored as 1-2 in PRIMARY scoring decreased from 29% to 10% in delayed images compared with standard images, whereas lesions scored as 3-5 increased from 71% to 90%. Additionally, agreement between two experienced nuclear medicine physicians regarding scoring was 66% for standard imaging and 77% for delayed imaging. The number of patients with PRIMARY score 5 increased from 31 to 46 in delayed imaging. All patients were confirmed to have clinically significant prostate cancer (csPCa). Furthermore, no csPCa of ISUP grade 3 or higher was detected in patients with a delayed PRIMARY score (dPRIMARY). The sensitivity of standard PRIMARY scoring was 71%, which increased to 92% with dPRIMARY scoring, with a consistent positive predictive value of 87% for both. Intraobserver agreement Cohen kappa values for all observers were higher for delayed images than for standard images. Inter-observer agreement, assessed by Fleiss kappa, was 0.47 and 0.52 for standard images in rounds 1 and 2, respectively, and 0.61 and 0.72 for delayed images, respectively.

Conclusion

Decreased background activity and increased primary tumor uptake in delayed images improved differentiation between primary tumors and benign lesions, leading to better primary tumor identification. Enhanced reliability was also observed in both intraobserver and interobserver assessments of delayed images.

Introduction

Positron emission tomography/computed tomography (PET/CT) with [⁶⁸Ga]-labeled prostate-specific membrane antigen (PSMA) inhibitors ([⁶⁸Ga]Ga-PSMA-11 PET/CT) has emerged as a valuable modality for both staging and restaging of clinically significant prostaate cancer (csPCa) (1,2,3). However, recent studies have suggested its potential utility in the primary diagnosis of csPCa, particularly in distinguishing clinically significant csPCa from indolent forms (4,5). Notably, PSMA expression correlates positively with csPCa grade, with higher expression associated with increased disease severity and poorer prognosis (6). Molecular imaging with [⁶⁸Ga]Ga-PSMA PET/CT provides insights into disease at the molecular level, with PSMA uptake reflecting PSMA expression levels. Studies have indicated a direct association between maximum standardized uptake value (SUV_max) on [⁶⁸Ga]Ga-PSMA PET/CT and PSMA expression, with SUV_max escalating alongside higher-grade groups of csPCa (7). The PRIMARY trial was undertaken to evaluate the diagnostic roles of [⁶⁸Ga]Ga-PSMA PET/CT and multi-parametric magnetic resonance imaging (mpMRI) in discerning csPCa (8). Notably, the PRIMARY trial revealed a higher sensitivity [⁶⁸Ga]Ga-PSMA PET/CT than mpMRI, with a marked improvement observed when both modalities were combined.

The introduction of the PRIMARY score signifies a remarkable advancement in csPCa diagnosis. Incorporating factors such as the uptake pattern within the prostate gland, location within the peripheral zone, and intensity of PSMA uptake, the PRIMARY score aims to enhance diagnostic accuracy (4). Reproducibility studies have demonstrated comparable reliability to that of MRI, suggesting the potential for predicting csPCa (9). Although the PRIMARY score has been reported to achieve successful results in the diagnosis of csPCa in a selected group of patients with low-grade csPCa (10), its applicability in patients with low-grade disease remains uncertain and requires further refinement (11).

PSMA uptake varies in benign prostate lesions and in normal prostate tissue on [⁶⁸Ga]Ga-PSMA PET/CT. Within the PRIMARY scoring system, the first three scores are primarily aimed at differentiating benign pathologies from csPCa, with a focus on lesions predominantly located in the peripheral zone (4). It has been postulated that increased PSMA uptake outside the peripheral zone predominantly indicates benign conditions. Early studies using [⁶⁸Ga]Ga-PSMA PET/CT demonstrated a reduction in PSMA uptake in benign lesions and background activity on delayed images obtained 2-3 hours after injection (12,13,14,15). In contrast, PSMA uptake was observed to be constant or increased in malignant lesions. Therefore, delayed imaging may allow for a more accurate classification of benign prostate lesions, potentially improving the efficacy of the PRIMARY scoring system.

This study aimed to investigate the comparative efficacy of PRIMARY scoring on [⁶⁸Ga]Ga-PSMA PET/CT images acquired at standard versus delayed time points and to ascertain the potential incremental value of delayed imaging in PRIMARY scoring assessment.

Materials and Methods

Patient Population and Study Protocol

We assessed 140 treatment-naive patient records diagnosed with International Society of Urological Pathology grade groups (ISUP) 1 and 2 PCa by biopsy (bISUP) and who underwent radical prostatectomy (RP) in different hospitals. The indications for RP were patient preference, physician preference, and high D’Amico risk. The final pathology results were compared with preoperative [⁶⁸Ga]Ga-PSMA PET/CT imaging PRIMARY scoring data. [⁶⁸Ga]Ga-PSMA PET/CT imaging was performed within 3 months before RP. The initial informed consent form also included consent for future retrospective analysis. The study was approved by the Yeditepe University Rectorate Non-Interventional Clinical Research Ethics Committee (number: E.83321821-805.02.03-377, date: 15.03.2024) and was carried out in accordance with the Declaration of Helsinki.

Imaging Protocol

Preoperative [⁶⁸Ga]Ga-PSMA PET/CT imaging was performed with a mean dose of 240.5±67.16 MBq of [⁶⁸Ga]Ga-PSMA-11. The mean start time of scanning after injection was 59.5±15.8 min (median 60.0 min). Whole-body [⁶⁸Ga]Ga-PSMA PET/CT imaging was performed using a GE Discovery 710 PET/CT scanner with 64 CT slices from the vertex to the mid-thigh. Additional delayed pelvic imaging, which is routine in our clinic, was also acquired using 3-bed pelvic imaging at a mean of 139.1±21.3 min (median 138.0 min) after injection. Standard and delayed imaging was performed with an acquisition time of 3 min per bed position (Figure 1). Low-dose CT images were obtained for anatomical localization and attenuation correction. [⁶⁸Ga]Ga-PSMA PET/CT images were analyzed using the Advanced Workstation (v4.7) (GE Healthcare, WI, USA). The prostate region of interest was manually delineated for SUV_max measurement. The standard uptake value (SUV) calibration of the scanner was performed every 3 months as recommended by the manufacturer. The other quality control checks were regularly performed according to the recommendations of the European Association of Nuclear Medicine Guidelines (16).

PRIMARY Scoring

[⁶⁸Ga]Ga-PSMA PET/CT images were independently assessed by two experienced nuclear medicine physicians to determine the PRIMARY scores of both standard (sPRIMARY score) and delayed imaging (dPRIMARY score). To ensure an unbiased assessment, these physicians were blinded to the patients’ clinical and pathologic data. In case of disagreement in the assignment of PRIMARY scores, a third nuclear medicine physician was consulted to reach a consensus. To assess inter- and intra-observer reproducibility, initial scoring was performed on standard imaging by four distinct nuclear medicine specialists who were blinded to both each other and the patient’s clinical information. Subsequently, the same specialists repeated the procedure using only delayed [⁶⁸Ga]Ga-PSMA PET/CT images, while remaining blinded to the standard [⁶⁸Ga]Ga-PSMA PET/CT images and initial scores. Intra-observer agreement analysis involved the specialists re-assessing all images after a 1-month interval, following the same protocol as described above. In the PRIMARY scoring system used, score 2 was not divided into two subgroups; instead, both subgroups A and B were considered as group 2. Patients with a final pathological report indicating ISUP 2 or higher were diagnosed with csPCa.

Statistical Analysis

The descriptive data were used to calculate the median and mean values and the corresponding standard deviations(±). The values were analyzed using the Statistical Package for the Social Sciences, version 25.0 (IBM Corporation in Chicago, Illinois, USA). Sensitivity and specificity values for PRIMARY scoring were determined using pathologic ISUP results as a standard reference. Differences in diagnostic parameters between the sPRIMARY and dPRIMARY scores were assessed using the MedCalc proportional comparison calculator, which can be accessed at the following URL: https://www.medcalc.org. The relationship between ISUP and both sPRIMARY or dPRIMARY scores was analyzed using the Spearman correlation test. Cohen’s kappa and Fleiss kappa were determined as 5 categories of the PRIMARY scoring for both imaging. Receiver operating characteristic (ROC) curves were also calculated for 5-level sPRIMARY and dPRIMARY scores. The paired samples t-test was used to compare the SUVmax of standard and delayed images. The Related-Samples Wilcoxon signed-rank test and McNemar’s test were used to perform a comparative analysis between the sPRIMARY and dPRIMARY scores. Statistical significance was attributed to the values with a p<0.05 (two-sided).

Results

The mean age of the 140 men included in the study was 62.6±7.6 years (range, 43-80 years). The mean prostate-specific antigen (PSA) level was 8.5±7.2 ng/mL, with values ranging from 1.8 to 67.0 ng/mL. Regarding the preoperative biopsy results, 51 patients (36%) had bISUP 1, while 89 patients (64%) had bISUP 1 and 2, respectively. After RP, a 69% and 20% upgrade was observed in patients initially classified as bISUP 1 and a 20% upgrade in patients initially classified as bISUP 2. There was a 38% increase in bISUP in all patients to a higher value (Table 1).

Among the 140 patients, csPCa was detected in 34 (24%) patients with sPRIMARY scores 1-2 and in 86 (61%) patients with sPRIMARY scores 3-5 after RP (Tables 2,3). The sensitivity and specificity of sPRIMARY scoring in identifying csPCa were calculated to be 71% and 35%, respectively (Table 4). The positive predictive value (PPV) and negative predictive value were calculated to be 87% and 17%, respectively. The sensitivity of dPRIMARY scoring increased from 71% [confidence interval (CI), 63%-80%] to 92% (CI, 85%-96%) (p<0.0001) without any change in PPV. It is noteworthy that the number of patients with sPRIMARY score 5 increased from 31 to 46 with dPRIMARY score 5, all of whom were confirmed to have csPCa. Additionally, no csPCa of ISUP grade 3 or higher was detected in any of the patients with a dPRIMARY score of 1-2. The area under curves (AUCs) of sPRIMARY and dPRIMARY in the receiver operating characteristic curves (ROC) analysis were 0.616 and 0.721, respectively. The sPRIMARY and dPRIMARY scores were moderately correlated with ISUP, with Spearman-Rho values of 0.302 and 0.389, respectively.

The percentage of lesions scored as 1-2 in the PRIMARY scoring decreased from 29% in standard images to 10% in delayed images, whereas the percentage of lesions scored as 3-5 increased from 71% in standard images to 90% in delayed images (p<0.001 Table 2). In addition, the percentage of giving the same PRIMARY score by two experienced nuclear medicine physicians in standard imaging was 66%, which increased to 77% in delayed imaging (p<0.05). According to the results of Cohen’s kappa analysis at standard imaging, intra-observer agreement was 0.536, 0.798, 0.593, and 0.638 for observers 1, 2, 3, and 4, respectively. The delayed images showed higher Cohen’s Kappa coefficients of 0.697, 0.839, 0.769, and 0.740 for all observers, respectively, compared with the standard images, indicating better intraobserver reproducibility for delayed images (Figure 2). In the first assessment for PRIMARY scoring, the inter-observer Fleiss kappa coefficient was 0.47 for standard images and 0.61 for delayed images. In the second assessment, these coefficients increased to 0.52 for standard images and 0.72 for delayed images (Figure 3). Fleiss kappa coefficients were better for delayed images than for standard images in both sessions. Additionally, inter-observer Fleiss kappa coefficients were higher in the second assessment evaluation than in the first assessment evaluation for standard and delayed images.

The mean SUV_max of the focal lesion, regarded as the primary tumor in the prostate gland, was 9.3±6.1 (Table 2) using standard images. The SUV_max significantly increased to a mean value of 11.4±7 in delayed images. The SUV_max was significantly higher in delayed images than in standard images (p<0.001, Figure 4).

Discussion

The introduction of the PRIMARY score represents a remarkable advance in the initial diagnosis of PCa using [⁶⁸Ga]Ga-PSMA PET/CT (4). The incorporation of the PRIMARY risk classification scheme has the potential to standardize the interpretation of lesions and allow for clearer and more objective communication between nuclear medicine physicians and clinicians (5). Nevertheless, further refinement of the PRIMARY score is required, particularly in cases of low-risk PCa (11). The purpose of PRIMARY scoring was to identify PCa in the prostate gland. However, PSMA uptake can be high under benign conditions. In such cases, PRIMARY scoring can be challenging because it is difficult to distinguish between csPCa with low PSMA expression and benign lesions with high uptake (4).

Initial studies reported that the normal prostate gland showed heterogeneous PSMA uptake and that even patients without a diagnosis of csPCa could show PSMA uptake as high as an SUVmax of 8.3 (17). High uptake within the prostate gland may make it difficult to distinguish csPCa with low PSMA expression. However, both early and recent studies have shown that background activity decreases in [⁶⁸Ga]Ga-PSMA PET/CT images taken after 2-3 hours and higher tumor/background ratios are obtained in delayed images compared with standard images (12,13,14,15). These findings may allow for better classification of benign prostate lesions and more accurate detection of masked malignant lesions. Accordingly, studies have shown that additional lesions that can change the stage of the disease and treatment management can be detected in up to 25% of patients in delayed images due to the higher tumor/background ratios. In our study, we observed a significant decrease in the frequency of the first two scores (from 29% to 10%) in the PRIMARY scoring on delayed images, which reflects the washout of PSMA uptake from benign lesions in delayed images (p<0.001). In addition, the agreement rate between two experienced nuclear medicine physicians in assigning identical PRIMARY scores during standard imaging was 66%, a rate that increased to 77% during delayed imaging (p<0.05).

In the aforementioned studies, it was noted that not only does the background uptake decrease in delayed images, but also the PSMA uptake of the tumor increases, resulting in enhanced visibility of the tumor (12,13,14,15). Consistent with this observation, a significant increase in SUV_max of the focal lesion, which is considered the primary tumor in the prostate gland, was observed in delayed [⁶⁸Ga]Ga-PSMA PET/CT images compared with standard [⁶⁸Ga]Ga-PSMA PET/CT images in our study. Decreased background activity and increased primary tumor uptake have contributed to a clearer distinction between primary tumors and benign lesions, leading to an improved identification of the primary tumor and a better assessment of its location in the peripheral zone of the prostate gland (Figure 4). Consequently, our diagnostic parameters exhibited statistically significant improvement for delayed images compared with standard images. The sensitivity of dPRIMARY scoring increased significantly from 71% to 92% compared with sPRIMARY scoring (p<0.0001). When comparing the AUCs of the sPRIMARY and dPRIMARY scoring in the ROC analysis, the higher AUC of the dPRIMARY score indicated superior diagnostic parameters.

For all sPRIMARY scores, Fleiss’s kappa values of inter-observer reproducibility were found to be 0.47 for the first-assessment and 0.52 for the second assessment. Emmett et al. (9) found Cohen’s kappa values for inter-observer reproducibility to be 0.65 for the entire sPRIMARY scoring scale. Inter-observer Fleiss’s kappa values in standard images in our study were found to be lower compared to that of study by Emmett et al. (9) This difference may be explained by the presence of patients with bISUP 2 in their patient cohort and the kappa analysis with two observers instead of four. In our study, unlike the study conducted by Emmett et al. (9), PRIMARY scoring was also evaluated in delayed images. For the dPRIMARY scoring, the Fleiss’s kappa coefficients were 0.61 for the first assessment and increased to 0.72 in the second assessment. The decrease in PSMA uptake in benign lesions and the increase in PSMA uptake in the primary tumor over time improved both intra-observer and inter-observer consistency in dPRIMARY scoring. Higher consistency rates in late images were also observed for all four observers. In addition, it was consistently observed that the second assessment was better than the first assessment for both standard and delayed images. This result was considered to be related to the learning process. Even though the observers in this study were experienced in reading [⁶⁸Ga]Ga-PSMA PET/CT, they had not learned PRIMARY scoring before. The importance of the learning process was already emphasized by Emmett et al. (9).

In this study, as in previous studies, csPCa was detected in all patients with a PRIMARY score of 5 (4,10). Moreover, no ISUP 3-5 was detected in any of the patients with a dPRIMARY score of 1-2 after RP. This approach could have significant clinical implications and could help identify patients who should not be placed under active surveillance. he majority of patients with ISUP 1 exhibit PSMA expression and corresponding PSMA uptake; however, a subset of ISUP 2 patients lack PSMA expression and therefore show no PSMA uptake. This observation underscores the heterogeneity in prognosis among patients with ISUP 1 and 2, reflecting potential biological variability within these groups. Previously, it has been reported that the prognosis of the majority of patients with ISUP 1 is very good, 11% experience biochemical recurrence or progression, and 65% of patients with ISUP 2 do not experience recurrence (18). Additionally, it has been shown that high levels of PSMA expression may be present in some patients with ISUP 1-2 (19). The increased PSMA expression observed in patients with ISUP 1-2 may indicate poor clinical outcomes (6). Further studies on clinical outcomes are needed to clarify whether such a relationship exists or not.

The interpretation of the findings of this study requires caution due to the inherent limitations associated with retrospective studies. The patient cohort for RP introduces potential selection bias, as evidenced by the 86% incidence of csPCa, highlighting the need for careful evaluation of diagnostic parameters. However, the study focused on assessing the effectiveness of the PRIMARY SCORE in both standard and delayed images. Furthermore, the relatively elevated rate of upstaging from the bISUP might be attributable to the absence of a screening setting in the study.

Conclusion

In conclusion, when comparing standard and delayed images, a notable improvement in reliability was observed in both intra-observer and inter-observer assessments of delayed images. Furthermore, delayed images facilitated enhanced detection of csPCA patients with bISUP 1-2 that had been previously overlooked. Consequently, it can be used to distinguish individuals suitable for active surveillance in routine clinical practice.