Prostate cancer (PCa) is the most common cancer in men (1), and pelvic lymph node metastasis (PLNM) is an important factor associated with poor prognosis (2). Pelvic lymph node dissection (PLND) is the “gold standard” for PLNM diagnosis in PCa; however, the indication, scope, and patient benefit of PLND remain controversial (3). The European Association of Urology (EAU) guidelines recommend PLND when the risk of lymph node metastasis is predicted to be ≥5% using the Briganti nomogram (4), while the National Comprehensive Cancer Network (NCCN) recommend PLND when the risk of lymph node metastasis is predicted to be ≥2% using the MSKCC nomogram (5). Although sentinel lymph node biopsy has a diagnostic role, standard pelvic lymph node (SPLND) dissection can only remove 38% of suspicious lymph nodes (6). Multiple studies have shown that the number of lymph nodes obtained and the positive rate are positively correlated with the extent of PLND (7, 8). Heidenreich reported that the lymph node detection rate of extended pelvic lymph node dissection (EPLND) is approximately twice that of SPLND (9). Notably, EPLND significantly increases postoperative complications compared to SPLND (10).

Preoperative diagnosis of PLNM of PCa is crucial for accurate staging and determining optimal treatment. PLNM in PCa usually appear small on magnetic resonance imaging (MRI) (11), and if a threshold of 0.8-1.0cm in short diameter is employed as a criterion, the clinical stage may be underestimated due to low detection sensitivity (12). Ultra-small iron oxide particle MRI or choline positron emission tomography-computed tomography (PET-CT) has demonstrated high accuracy in detecting PLNM (13, 14), nonetheless, the executability of these techniques is low. Therefore, accurately predicting PLNM before radical prostatectomy (RP) of PCa remains the key and challenging aspect of clinical work.

Recent studies analyzing magnetic resonance images to quantitatively grade the likelihood of extraprostatic invasion in PCa, have shown that higher extraprostatic extension (EPE) grade is associated with higher tumor stage, Gleason score, or biochemical recurrence (15). According to these results, a higher EPE grade is associated with more aggressive PCa and a poorer prognosis. Consequently, EPE grade may also be associated with PLNM in PCa.

We used preoperative MRI and clinicopathological data to assess the risk of PLNM in 317 patients with PCa undergoing RP and PLND and found EPE grade and biopsy positive rate as independent risk factors. A combination of these independent risk factors improved the accuracy of predicting PLNM in PCa.

At present, preoperative lymph node staging of PCa relies mainly on CT and MRI, but these imaging methods cannot identify PLNM. Hovels (16) found 39% and 42% sensitivity, respectively, of MRI and CT in predicting lymph node metastasis. PSMA PET-CT demonstrates low sensitivity and high specificity in predicting PLNM in PCa (17). The radiomics nomogram currently has good predictive efficacy for PLNM of PCa, but its reliability needs to be verified further (18, 19).

In this group, the metastasis rate of PLNM of PCa was 10.4%, and that of PLNM of EPE grades 0-I, 2, and 3 were, respectively, 1.4%, 18.3%, and 56.6%, with statistically significant differences between groups. The purpose of EPE grading is to stratify the risk of PCa extrinsic invasion. Studies on the efficacy of EPE grade have found the correlation of EPE grade with pathological Gleason score and clinical stage (20). The probability of lymph node metastasis also correlates highly with tumor stage and pathological Gleason score (21). Outside the prostate capsule, there is a dense lymphatic drainage network. As the tumor grows in size and breaks through the prostate capsule, the density of lymphatic vessels around the tumor increases, hastening the occurrence of pelvic lymph node metastasis of PCa (22). This may explain the study findings that the majority of patients with PLNM had high EPE grade on magnetic resonance images. Therefore, theoretically, the risk of PLNM increases significantly in patients with EPE grade 2 and 3. For such patients, surgical indications should be strictly adhered to, and the risk of PLNM must be fully explained to the patients and their families before surgery.

Clinicopathological factors PLNM of PCa with high accuracy, and PSA and Gleason scores are independent predictors of PLNM of PCa (23). Porcaro found a positive correlation between PSA level and PLNM (P = 0.012) (24), and Yiakoumos also reported PSA density as an important predictor of PLNM of PCa (25). One study based on the SEER database and the American Cancer Database demonstrated a very low risk of PLNM for PCa with a Gleason score of 6 or less and a significantly higher risk, with a Gleason score of 8-10 (26). The risk factors for PLNM of PCa include the number of positive needles and biopsy positive rate, and biopsy positive rate serves as an independent predictor (27). According to recent studies, peripheral monocyte count is also one of the best predictors of PLNM of PCa (28). In this study, the differences between positive and negative PLNM in PSA, biopsy positive rate, and biopsy GG were statistically significant, among which biopsy positive rate was an independent risk factor for PLNM of PCa.

A combination of MRI with clinicopathological indicators approach can enhance the accuracy of predicting PLNM of PCa. Multivariate models incorporating MRI tumor volume, tumor T stage, PSA, and biopsy GG demonstrated a high predictive value for PLNM in PCa (29). In this study, the AUC for predicting PLNM through EPE grade was 0.879, and the AUC for predicting PLNM increased to 0.921 upon employing the combination of EPE grade and the biopsy positive rate, with statistical significance. This study indicates that the combined biopsy positive rate and EPE grade has significant clinical value in predicting PLNM of PCa. To a certain extent, it can provide important reference for clinical treatment decisions regarding PCa. When the EPE grade is greater than or equal to level 2, it suggests that urologists should consider performing PLND for patients of PCa.

In this study, among the 33 patients of PCa in this group who had PLNM, 23 patients had PLNM on the same side as the main lesion of the tumor, 6 patients had PLNM on the contralateral side as the main lesion of the tumor, and 4 cases had bilateral PLNM simultaneously. This was quite similar to that reported by Weckermann (30). It indicates that unilateral lymph node dissection on the side where the main tumor is located alone has a relatively high risk of missing pelvic lymph node metastasis on the contralateral side.

The study also had some limitations. First, the sample size of PLNM+ cases (n=33) is small for multivariable regression, risking model over fitting, Small PLNM+ cohort inflates confidence intervals. Second, as a retrospective study, there are fixed limitations in data collection and a risk of selection bias. Thirdly, All patients came from the same center, resulting in a relatively small overall sample size, the evidence level of the research results is not high, and large-scale, multi-center, and prospective studies are needed for further verification. Lastly, MRI protocol (1.5T) may limit the assessment of EPE grade of PCa.

To conclude, there is an independent correlation between EPE grade and the biopsy positive rate with PLNM of PCa. A comprehensive evaluation of these factors is helpful in predicting the probability of PLNM of PCa before surgery, which will be helpful for urologists in deciding whether to perform PLND during RP.