68Ga-PSMA PET/CT is a valuable tool for detecting and assessing metastatic disease burden in PCa, as well as detecting and restaging PCa recurrence after curative treatment [11]. In addition to evaluating systemic metastases, 68Ga-PSMA PET/CT can identify the intraprostatic index lesion. It is known for its high sensitivity in detecting intraprostatic tumours, due to its ability to target PSMA expression on PCa cells [12]. The intensity of tumour-related tracer uptake on 68Ga-PSMA PET/CT correlates with PSA level and Gleason score in newly diagnosed PCa [13, 14].
The PRIMARY score, which incorporates intraprostatic patterns and intensity on 68Ga-PSMA PET/CT, is an accurate and reproducible method for reporting intraprostatic PSMA PET findings in men before prostate biopsy [9]. The recently updated PROMISE framework, which aims for standardization in 68Ga-PSMA PET/CT, includes the use of PRIMARY scoring in the definition of intraprostatic lesions [15]. Akcay et al. [16] showed that the PRIMARY scoring system had a sensitivity of 77% and specificity of 58% in identifying clinically significant PCa and is a potential tool for selecting patients for active surveillance. Active surveillance is a management method allowing to postpone definitive treatment and avoid the side effects of surgery and radiotherapy for patients with low-risk PCa. [17, 18, 19]. In our study, unlike the existing literature, we analyzed the distribution of PRIMARY scoring among PCa risk groups. Our results showed that while the PRIMARY scoring had good sensitivity, it was lower than the PI-RADS scoring (82.3% vs 94.2%), and PRIMARY scoring had higher specificity than PI-RADS (34.6% vs 5.2%) in identifying patients with intermediate/high-risk prostate cancer. Therefore, the use of PRIMARY scoring alone or in combination with PI-RADS scoring in the detection or follow-up of active surveillance patients can reduce unnecessary biopsies.
In the PROMIS trial investigating the diagnostic accuracy of mpMRI in PCa, the sensitivity of mpMRI for clinically significant cancer was 93% and the negative predictive value (NPV) was 89%. The specificity of mpMRI was 41% with a PPV of 51% [6]. The low specificity and PPV of mpMRI indicate that it may lead to unnecessary biopsy. mpMRI still misses a small number of clinically significant PCa, which leads to the need for new approaches to reduce false-negative rates [20]. The PRIMARY trial found that using PSMA PET/CT in combination with mpMRI increased sensitivity and NPV for clinically significant PCa compared to the use of mpMRI alone [21]. Scheltema et al. [22] showed that PSMA-PET demonstrated higher diagnostic accuracy, sensitivity, specificity, NPV, and PPV, especially for ISUP grades 2–3 compared to mpMRI, and the combination of PSMA-PET and mpMRI improved sensitivity, specificity, NPV, and PPV for detecting ISUP grades 2–3 PCa. Using 68Ga-PSMA PET/CT and mpMRI together has the potential to accurately predict clinically significant PCa prior to biopsy, offering a noninvasive method for clinicians to gauge clinically significant PCa risk and potentially reduce unnecessary biopsies [23]. As in our study, the higher sensitivity of the PI-RADS score suggests that it is effective in correctly identifying a larger proportion of patients with intermediate and high-risk PCa. However, the low specificity and positive predictive rates of PI-RADS may lead to unnecessary biopsies if used alone in decision-making during the follow-up of active surveillance patients. In contrast, PRIMARY scoring had a more balanced performance with lower sensitivity but relatively higher specificity compared to PI-RADS.
Active surveillance (AS) is a strategy that uses PSA testing, digital rectal examination, and prostate biopsies to monitor PCa in men rather than recommending preventive treatment for low-risk localized PCas [24]. Very low-risk and low-risk patients (biopsy Gleason score of ≤ 6, a PSA level less than 10 ng/mL, and a clinical stage of T1c or T2a) with a life expectancy greater than 10 years are candidates for AS. [25]. Although the overall risk of major complications from the prostate biopsy is low, it is still an invasive technique that can cause complications like bleeding, infection, pain, lower urinary tract symptoms, urinary retention, and erectile dysfunction [26]. On this basis, interest in the potential of imaging methods to replace biopsy in the diagnosis and AS of PCa has increased. mpMRI is used in some AS protocols to confirm eligibility, monitor for changes, and guide targeted biopsies, but current data do not support the use of negative mpMRI alone to avoid scheduled biopsies [18]. The ROC Curve analysis revealed a moderately higher AUC for PRIMARY scoring (0.727, 95% CI: 0.623–0.830) compared to PI-RADS scoring (0.662, 95% CI: 0.550–0.774). Moreover, the AUC of the means of PRIMARY and PI-RADS scores was even higher (0.744, 95% CI: 0.643–0.846). This suggests that PRIMARY scoring may offer an improved ability to select patients for active surveillance and has the potential to change the re-biopsy decision, preventing unnecessary biopsies during active surveillance when used together with PI-RADS.
SUVmax values can predict the grade groups of the primary tumor and may serve as a target for biopsy sampling in selected patients, in conjunction with mpMRI [14]. Jiao J et al. [27] recommended an SUVmax cutoff value of 5.30 to discriminate between clinically significant PCa and benign prostatic disease using 68Ga-PSMA PET/CT imaging. PRIMARY score 4 represents any focal activity in the peripheral zone without minimum intensity (10). In our results, the most frequent PRIMARY score in both the intermediate/high-risk group and the low-risk group was 4 (48.2% and 51.0%, respectively) and a statistically significant difference was detected between the average SUVmax values of the intermediate/high and low-risk groups with a PRIMARY score of 4 (6.7 and 5.0, respectively) (p = 0.018). Also in our analysis, when a cut-off SUVmax value of 5.0 was determined for the PRIMARY score of 4, despite the decrease in sensitivity (82.3–67.0%), the specificity (34.6–65.3%) and the PPV (68.6–77.0%) of the PRIMARY scoring had increased. This suggests that revising the PRIMARY score of 4 with a cut-off SUVmax value may improve the specificity and PPV of the scoring system.
Our study had limitations due to its single-center retrospective nature. The ISUP degree distributions of the patient group were not homogeneous. PI-RADS scores were recorded retrospectively based on reports, which gave rise to the possibility that the evaluation was made by a single expert. Although at least two expert opinions were taken when determining PRIMARY scores, the experts' experience in the use of this scoring was limited.