The Decipher Genomic Resource Information Database (GRID) platform by GenomeDx Biosciences, San Diego, was recently used to develop and validate a novel genomic signature, known as KNN51, for the prediction of lymph node involvement at the time of radical cystectomy in patients with muscle-invasive bladder cancer.

Lymph node involvement is a known risk factor warranting intensification of therapy; however, it is missed in up to approximately 25% of cases. According to a new study published by the Journal of Urology, using the KNN51 genomic signature to provide more-accurate detection and classification of disease may guide patient therapy earlier in the disease.1

Bladder cancer patients with lymph node involvement in their disease have historically been shown to be at a high risk for disease recurrence and death. Because of the limitations of current imaging techniques and clinical risk-assessment tools, however, current clinical staging can miss up to approximately 25% of these high-risk cases.

Patients with muscle-invasive bladder cancer have been shown to experience improved outcomes through treatment intensification with neoadjuvant chemotherapy (NAC). The ability to preoperatively predict patient response to NAC would allow for optimal patient management. Data from the new study demonstrate that the newly developed gene signature, if validated prospectively, could be used to guide high-risk patients toward early multimodal therapy.

Doug Dolginow, MD, GenomeDx.

Doug Dolginow, MD, GenomeDx.

“Decipher GRID has proven to be a powerful tool for identifying powerful genetic markers that are predictive of aggressive disease, and a more accurate tool than existing methods for characterizing patient risk,” says Doug Dolginow, MD, chief executive officer of GenomeDx Biosciences. “The identification of new genomic solutions, such as KNN51, that can be applied toward patients suffering from bladder cancer is another step toward improving outcomes in this underserved patient population. We look forward to continuing to validate this genomic solution in bladder cancer, and expect to bring Decipher Bladder to market in early 2017.”

The Journal of Urology study included 199 patients who underwent radical cystectomy with extended pelvic lymphadenectomy at the University of Southern California between 1998 and 2004. Whole transcriptome expression profiles were generated for all patients. The cohort was divided randomly into a discovery set (n=133) and a validation set (n=66). In the discovery set, features were identified and modeled into the KNN51 classifier for prediction of pathologic lymph node metastases. For comparison, the discovery set was also modeled into two previously described bladder cancer gene signature classifiers—the 15-gene cancer recurrence signature (RF15) and the 20-gene lymph node signature (LN20).

In the validation set, KNN51 predicted lymph node involvement with high discrimination, and significantly outperformed RF15 and LN20. In addition, only KNN51 was significant in predicting the odds of lymph node metastasis, whereas neither RF15 nor LN20 were significant predictors of lymph node metastasis in this study.

The company’s first Decipher Bladder cancer classifier tests are currently under development, with a goal of providing an assessment of tumor aggressiveness based on the patient’s unique genomic profile.

For more information, visit GenomeDx Biosciences.


  1. Seiler R, Lam LL, Erho N, et al. Prediction of lymph node metastasis in patients with bladder cancer using whole transcriptome gene expression signatures. J Urol. Published online April 19, 2016; doi: 10.1016/l.juro.2016.04.061.