Association of chromosome 7 aneuploidy measured by fluorescence in situ hybridization assay with muscular invasion in bladder cancer

Abstract Background The preoperative prediction of muscular invasion status is important for adequately treating bladder cancer (BC) but nevertheless, there are some existing dilemmas in the current preoperative diagnostic accuracy of BC with muscular invasion. Here, we investigated the potential association between the fluorescence in situ hybridization (FISH) assay and muscular invasion among patients with BC. A cytogenetic‐clinical nomogram for the individualized preoperative differentiation of muscle‐invasive BC (MIBC) from non‐muscle‐invasive BC (NMIBC) is also proposed. Methods All eligible BC patients were preoperatively tested using a FISH assay, which included 4 sites (chromosome‐specific centromeric probe [CSP] 3, 7, and 17, and gene locus‐specific probe [GLP]‐p16 locus). The correlation between the FISH assay and BC muscular invasion was evaluated using the Chi‐square tests. In the training set, univariate and multivariate logistic regression analyses were used to develop a cytogenetic‐clinical nomogram for preoperative muscular invasion prediction. Then, we assessed the performance of the nomogram in the training set with respect to its discriminatory accuracy and calibration for predicting muscular invasion, and clinical usefulness, which were then validated in the validation set. Moreover, model comparison was set to evaluate the discrimination and clinical usefulness between the nomogram and the individual variables incorporated in the nomogram. Results Muscular invasion was more prevalent in BC patients with positive CSP3, CSP7 and CSP17 status (OR [95% CI], 2.724 [1.555 to 4.774], P < 0.001; 3.406 [1.912 to 6.068], P < 0.001 and 2.483 [1.436 to 4.292], P = 0.001, respectively). Radiology‐determined tumor size, radiology‐determined clinical tumor stage and CSP7 status were identified as independent risk factors of BC muscular invasion by the multivariate regression analysis in the training set. Then, a cytogenetic‐clinical nomogram incorporating these three independent risk factors was constructed and was observed to have satisfactory discrimination in the training (AUC 0.784; 95% CI: 0.715 to 0.853) and validation (AUC 0.743; 95% CI: 0.635 to 0.850) set. The decision curve analysis (DCA) indicated the clinical usefulness of our nomogram. In models comparison, using the receiver operator characteristic (ROC) analyses, the nomogram showed higher discriminatory accuracy than any variables incorporated in the nomogram alone and the DCAs also identified the nomogram as possessing the highest net benefits at wide range of threshold probabilities. Conclusion CSP7 status was identified as an independent factor for predicting muscular invasion in BC patients and was successfully incorporated in a clinical nomogram combining the results of the FISH assay with clinical risk factors.