RC is considered as the standard therapy treatment for MIBC in the past. Meanwhile, due to the perioperative mortality and the loss of bladder organ function, more and more studies focus on the bladder preserving therapy. Recently available research evidence(9, 16, 17)indicated that the long-term prognosis after TMT was not significant inferior to that after RC. Therefore, multimodal bladder-preserving treatment strategies may be an alternative approach in the treatment of MIBC. For those MIBC patients who have a clinical complete response (cCR) to definitive therapy including neoadjuvant chemotherapy, consolidative bladder-preserving strategies can also be considered. Meanwhile, MIBC patients with high-risk factors, such as LN metastasis, are not recommended for conservative strategy. Hence, evaluation of preoperative LN status is an important part of selection criteria for multimodal bladder-preserving treatment.
A few researches focused on the imaging, pathology, and molecular characteristic of BC to predict pN+(18). Several data of systemic reviews and meta-analyses for the diagnostic accuracy for lymph node staging at first diagnosis in bladder cancer patients, demonstrated that MRI or CT featured modern sensitivity (ranging from 56% − 66%) but high specificity (ranging from 89% − 98%)(19–22). And it is often difficult to precisely ascertain metastatic LN less than 6.8 mm(23). That is to say, the risk of miss diagnosis leading to a proportion of patients being understaged. Cao R et al developed a predictive nomogram based on the EMT-related genes signature to discriminate the LN metastasis status of the Cancer Genome Atlas Urothelial Bladder Carcinoma (TCGA-BLCA) cohort, with the AUC of curve 71.7% and 75.9% in training and testing datasets, respectively(24). The genomic signature alone or combined with CT/MRI for the prediction of LN metastasis in patients with BC, which was not a convenient and accurate prediction tool. Hence, it’s necessary to improve the accuracy of detecting the LN metastasis of the newly diagnosed cases with convenience.
MIBC patients with pN + are correlated with high recurrence rates and poor survival. The recurrence free survival (RFS) and OS for patients with pN + at 5 years was 35% and 34%, and 31% and 23% at 10 years, respectively(7). Consequently, MIBC patients with pN + ought to receive positive intervention and follow-up. However, it lacks the research paid attention to the LN metastasis of non-metastatic MIBC patients.
In the present study, we developed a user-friendly predictive nomogram with preoperative variables. According to our logistic regression results, larger tumor size, overlapping lesion, young age, female, poorly differentiated histological grade, and advanced T stage, are independent risk factors for pN+. We incorporated clinically relevant variables and potentially available oncologic characteristics into the nomogram model construction. Oncologic variables, including tumor size, T stage, tumor sites in the BC wall and tumor grade could be routinely identified with imaging. Age and gender are frequently available from clinical information. Thus, our nomogram model could be easy-going to clinical application. Meanwhile, the AUC of our nomogram model was 0.704 with a satisfactory pN + predictive probability. Last but not least, we developed a risk classification system according to the scores based on the nomogram, thus all patients were divided into low-risk group (44.6%), intermediate-risk group (25.3%), or high-risk group (30.1%). Of all patients, the intermediate-risk and high-risk group accounting for about 55.4%, having a higher tendency for LN involvement, are recommended for RC and LND.
Our study also has several limitations as followed: (1) This research was a retrospective dataset based on the SEER database used to identify pN + for MIBC patients. (2) This study only included 12,269 nonmetastatic MIBC patients underwent RC were identified from the SEER database from 2004 to 2015, without external validation. Therefore, it is of great importance to validate our clinical nomogram model in multicenter clinical trials.