Researchers have created and validated a classifier that can predict whether recurrence risk is high or low in patients with localized, clear-cell renal cell carcinoma (RCC) and deemed the process to be more feasible for clinical practice than previous gene-expression signatures (Lancet Oncol. 2019 Mar 14. Epub ahead of print).

“Identification of high-risk localised renal cell carcinoma is key for the selection of patients for adjuvant treatment who are at truly higher risk of recurrence,” explained Jin-Huan Wei, MD, Department of Urology, First Affiliated Hospital, Sun Yat-sen University, Guangzhou, Guangdong, China, and colleagues.

“We developed a classifier based on single-nucleotide polymorphisms (SNPs) to improve the predictive accuracy for renal cell carcinoma recurrence and investigated whether intratumour heterogeneity affected the precision of the classifier,” they continued.

Using paraffin-embedded specimens from 227 patients with localized, clear-cell RCC (ie, the training set), Dr Wei and colleagues conducted a retrospective analysis of 44 SNPs potentially tied to disease recurrence. These 44 SNPs were identified via the Cancer Genome Atlas (CGA) Kidney Renal Clear Cell Carcinoma dataset (n = 114; SNPs, 906,600).

This led to the development of a 6-SNP–based classifier. Dr Wei and colleagues evaluated intratumor heterogeneity in 2 other regions within the same tumors in the training set during their validation of the 6-SNP–based classifier in an internal testing set (n = 226), independent validation set (n = 428 patients treated at 3 hospitals in China), and CGA set (n = 441 patients who underwent resection between 1998 and 2010 for localized, clear-cell RCC).

The primary end point of the study was recurrence-free survival (RFS), and the secondary end point was overall survival.

Despite finding intratumor heterogeneity in 48 (23%) of 206 cases in the internal testing set with complete SNP information, Dr Wei and colleagues observed that the predictive accuracy of the 6-SNP–based classifier was similar among the 3 different regions of training set (areas under the curve [AUC] at 5 years, 0.749; 95% CI, 0.660-0.826 in region 1; AUC at 5 years, 0.734; 95% CI, 0.651-0.814 in region 2; and AUC at 5 years, 0.736; 95% CI, 0.649-0.824 in region 3).

Independent of age, sex, or tumor stage, grade, or necrosis, the 6-SNP–based classifier accurately prognosticated the RFS of patients in 3 validation sets (hazard ratio [HR], 5.32; 95% CI, 2.81-10.07 in the internal testing set; HR, 5.39; 95% CI, 3.38-8.59 in the independent validation set, and HR, 4.62; 95% CI, 2.48-8.61 in the CGA set; all P <.0001).

A nomogram, developed through the combination of the classifier and clinicopathologic risk factors (ie, tumor stage, grade, and necrosis), had a predictive accuracy significantly higher than that of each variable alone (AUC at 5 years, 0.811; 95% CI, 0.756-0.861).

“Our six-SNP-based classifier could be a practical and reliable predictor that can complement the existing staging system for prediction of localised renal cell carcinoma recurrence after surgery, which might enable physicians to make more informed treatment decisions about adjuvant therapy,” said Dr Wei and colleagues.

“Intratumour heterogeneity does not seem to hamper the accuracy of the six-SNP-based classifier as a reliable predictor of recurrence. The classifier has the potential to guide treatment decisions for patients at differing risks of recurrence,” they concluded.—Hina Khaliq