Ensemble-based classification using microRNA expression identifies a breast cancer patient subgroup with an ultralow long-term risk of metastases.

BACKGROUND: Current clinical markers overestimate the recurrence risk in many lymph node negative (LNN) breast cancer (BC) patients such that a majority of these low-risk patients unnecessarily receive systemic treatments. We tested if differential microRNA expression in primary tumors allows reliable identification of indolent LNN BC patients to provide an improved classification tool for overtreatment reduction in this patient group. METHODS: We collected freshly frozen primary tumors of 80 LNN BC patients with recurrence and 80 recurrence-free patients (mean follow-up: 20.9 years). The study comprises solely systemically untreated patients to exclude that administered treatments confound the metastasis status. Samples were pairwise matched for clinical-pathological characteristics to minimize dependence of current markers. Patients were classified into risk-subgroups according to the differential microRNA expression of their tumors via classification model building with cross-validation using seven classification methods and a voting scheme. The methodology was validated using available data of two independent cohorts (n = 123, n = 339). RESULTS: Of the 80 indolent patients (who would all likely receive systemic treatments today) our ultralow-risk classifier correctly identified 37 while keeping a sensitivity of 100% in the recurrence group. Multivariable logistic regression analysis confirmed independence of voting results from current clinical markers. Application of the method in two validation cohorts confirmed successful classification of ultralow-risk BC patients with significantly prolonged recurrence-free survival. CONCLUSION: Profiles of differential microRNAs expression can identify LNN BC patients who could spare systemic treatments demanded by currently applied classifications. However, further validation studies are required for clinical implementation of the applied methodology.

Comparison of the Metastasis Predictive Potential of mRNA and Long Non-Coding RNA Profiling in Systemically Untreated Breast Cancer.

Several gene expression signatures based on mRNAs and a few based on long non-coding RNAs (lncRNAs) have been developed to provide prognostic information beyond clinical evaluation in breast cancer (BC). However, the comparison of such signatures for predicting recurrence is very scarce. Therefore, we compared the prognostic utility of mRNAs and lncRNAs in low-risk BC patients using two different classification strategies. Frozen primary tumor samples from 160 lymph node negative and systemically untreated BC patients were included; 80 developed recurrence-i.e., regional or distant metastasis while 80 remained recurrence-free (mean follow-up of 20.9 years). Patients were pairwise matched for clinicopathological characteristics. Classification based on differential mRNA or lncRNA expression using seven individual machine learning methods and a voting scheme classified patients into risk-subgroups. Classification by the seven methods with a fixed sensitivity of ≥90% resulted in specificities ranging from 16-40% for mRNA and 38-58% for lncRNA, and after voting, specificities of 38% and 60% respectively. Classifier performance based on an alternative classification approach of balanced accuracy optimization also provided higher specificities for lncRNA than mRNA at comparable sensitivities. Thus, our results suggested that classification followed by voting improved prognostic power using lncRNAs compared to mRNAs regardless of classification strategy.