Predicting Outcomes of Hormone and Chemotherapy in the Molecular Taxonomy of Breast Cancer International Consortium (METABRIC) Study by Biochemically-inspired Machine Learning

Abstract

Genomic aberrations and gene expression-defined subtypes in the large METABRIC patient cohort have been used to stratify and predict survival. The present study used normalized gene expression signatures of paclitaxel drug response to predict outcome for different survival times in METABRIC patients receiving hormone (HT) and, in some cases, chemotherapy (CT) agents. This machine learning method, which distinguishes sensitivity vs. resistance in breast cancer cell lines and validates predictions in patients; was also used to derive gene signatures of other HT (tamoxifen) and CT agents (methotrexate, epirubicin, doxorubicin, and 5-fluorouracil) used in METABRIC. Paclitaxel gene signatures exhibited the best performance, however the other agents also predicted survival with acceptable accuracies. A support vector machine (SVM) model of paclitaxel response containing genes ABCB1, ABCB11, ABCC1, ABCC10, BAD, BBC3, BCL2, BCL2L1, BMF, CYP2C8, CYP3A4, MAP2, MAP4, MAPT, NR1I2, SLCO1B3, TUBB1, TUBB4A, and TUBB4B was 78.6% accurate in predicting survival of 84 patients treated with both HT and CT (median survival ≥ 4.4 yr). Accuracy was lower (73.4%) in 304 untreated patients. The performance of other machine learning approaches was also evaluated at different survival thresholds. Minimum redundancy maximum relevance feature selection of a paclitaxel-based SVM classifier based on expression of genes BCL2L1, BBC3, FGF2, FN1, and TWIST1 was 81.1% accurate in 53 CT patients. In addition, a random forest (RF) classifier using a gene signature ( ABCB1, ABCB11, ABCC1, ABCC10, BAD, BBC3, BCL2, BCL2L1, BMF, CYP2C8, CYP3A4, MAP2, MAP4, MAPT, NR1I2,SLCO1B3, TUBB1, TUBB4A, and TUBB4B) predicted >3-year survival with 85.5% accuracy in 420 HT patients. A similar RF gene signature showed 82.7% accuracy in 504 patients treated with CT and/or HT. These results suggest that tumor gene expression signatures refined by machine learning techniques can be useful for predicting survival after drug therapies.

Predicting ionizing radiation exposure using biochemically-inspired genomic machine learning

“…Redundancy increases the possibility of overfitting, thereby reducing the generalizability of these models to predict responses in independent datasets. We address this limitation with the information theory-based criterion for gene selection known as minimum redundancy maximum relevance (mRMR) 11– 13 , which ranks genes according to shared mutual information between expression levels and radiation dose (relevance), and by minimizing mutual information shared by expression values of these and other genes (redundancy) 11, 12 . mRMR outperforms ranking criteria based solely on maximizing relevance 12 .…”

“…In contrast with heuristic approaches like differential expression, we only consider genes with evidence of a relationship to radiation response, which significantly limits the number of model features. Biochemically-inspired genomic machine learning (ML) has been used to derive high performing gene signatures that predict chemotherapy and hormone therapy responses 13– 15 . From an initial set of mRMR-derived biochemically relevant genes, wrapper approaches for feature selection 16 are used to find an optimal set of genes that predict exposure to radiation.…”

Predicting Response to Platin Chemotherapy Agents with Biochemically-inspired Machine Learning

“…Previous validation of patient data for other drugs validated with other datasets 6,24 using biochemically inspired machine learning have had better performance than those reported here. We investigated the possibility that disease and molecular heterogeneity in platin-treated patients may have affected the accuracy of our results.…”

“…SVM models were also derived using the cisplatin and/or carboplatin-treated TCGA (The Cancer Genome Atlas) bladder urothelial carcinoma patients, using post-treatment time to relapse as a surrogate criterion for different GI 50 resistance thresholds (as performed in Mucaki et al [2017] 24 ; Supplementary Table S3). Similar trends to cell line SVMs are apparent: POLQ is frequently included in models with recurrence threshold of longer duration, while FEN1 is a marker of resistance, when time to relapse is shorter.…”

“…This was due primarily to a lack of patients treated with this drug combination in the TCGA bladder dataset (N=11). Individual signatures for several of these drugs have been derived and analyzed using the patient data from METABRIC (Molecular Taxonomy of Breast Cancer International Consortium) 24 . A reasonable approach to predicting combination chemotherapy would first determine the probability of sensitivity or resistance to individual drugs, accounting for the misclassification rate by each (defined as d 1 , …, d k ).…”