Login to Access Video or Poster Abstract: MP44-02
Sources of Funding: National Natural Science Foundation of China?8150100539?

Introduction

Determination of metastasis is urgent in the management of bladder cancer (BC). With the existing models predicting BC metastasis, we found two deviant groups that are unpredictable with clinical parameters. One is at high risk of pN+ but actually with pN0 (good prognosis group, G) and the other one is at low risk of pN+ but actually with pN+ (poor prognosis group, P). The present study aimed to determine the gene expression differences between these two groups with SEER database and TCGA database and validate the results in a cohort from Fudan University Shanghai Cancer Center (FUSCC).

Methods

1603 patients with radical cystectomy (RC) from the SEER database were enrolled to build a multivariate model predicting BC metastasis. This model was applied in 248 patients from TCGA database to distinguish these two deviant groups (G and P). The different expressed genes of the two groups were compared by t-test. RT-qPCR was applied to validate the results in a consecutive cohort of 127 BC patients who underwent RC from FUSCC.

Results

The multivariate logistic regression model based on SEER population identified 256 patients as group G and 76 patients as group P. The similar phenomenon was observed in TCGA database with 37 patients of group G and 17 patients of group P respectively. 183 genes showed significant difference between these two groups. Finally, 18 genes achieved a significant statistical power (P<0.05) in predicting lymph node metastasis in the TCGA cohort excluding these two deviant groups. Furthermore, the RT-qPCR results of our own cohort identified C1orf198 (P=0.04) as the only gene that expressed differently in the two groups.

Conclusions

Our study suggested that C1orf198 might serve as an independent predictor of metastasis in patients with muscle-invasive bladder cancer.

Funding

National Natural Science Foundation of China?8150100539?