Pair-wise HRD score difference across patients. A) The pair-wise HRD score differences were shown across 23 patients ordered by their subtype. As control, we computed the pair-wise HRD score difference between any two samples from different patients. The median level of the control is shown as a dash line. B) The pair-wise HRD score differences were shown across 14 patients with low HRD score ordered by their subtype. As control, we computed the pair-wise HRD score difference between any two samples with low HRD scores from different patients. The median level of the control is shown as a dash line. C) The pair-wise HRD score differences were shown across 9 patients with high HRD score ordered by their subtype. As control, we computed the pair-wise HRD score difference between any two samples with high HRD scores from different patients. The median level of the control is shown as a dash line.

Abstract Purpose: The 3-biomarker homologous recombination deficiency (HRD) assay measures the number of telomeric allelic imbalances, loss of heterozygosity, and large-scale state transitions in tumor DNA and combines these metrics into a single score that reflects DNA repair deficiency. The goal of this study is to assess the consistency of these HRD measures in different biopsies from distinct areas of the same cancer. Experimental Design: HRD scores, BRCA mutation status, and BRCA1 promoter methylation were assessed in 99 samples from 33 surgically resected, stage I–III breast cancers; each cancer was biopsied in three distinct areas. Homologous recombination repair (HR) deficiency was defined as either high HRD score (≥42) or tumor BRCA mutation. Results: Eighty-one biopsies from 32 cancers were analyzed. Tumor BRCA status was available for all samples, HRD scores for 70, and BRCA1 methylation values for 76 samples. The BRCA1/2 mutation and promoter methylation status and HR category showed perfect concordance across all biopsies from the same cancer. All tumors with BRCA1/2 mutations or promoter methylation had high HRD scores, as did 17% (4/24) of the BRCA1/2 wild-type and nonmethylated tumors. The HRD scores were also highly consistent between different biopsies from the same tumor with an intraclass correlation coefficient of 0.977, indicating that only 2.3% of the variance is attributed to within-tumor biopsy-to-biopsy variation. Conclusions: These results indicate that within-tumor spatial heterogeneity for HRD metrics and the technical noise in the assay are small and do not influence HRD scores and HR status. Clin Cancer Res; 23(5); 1193–9. ©2016 AACR.

Individual component scores (LOH-, TAI-, LST-Score) for each sample in triplicates, duplicates or singles according to evaluable data.

Genomic profiles of tumor samples at 3 biopsy sites. Examples are shown for tumors with similar (top) and divergent (bottom) biopsy triplets from the same tumor. The red arrows indicate regions of significant genomic divergence. The yellow line indicates regions of LOH (allele dosage of 0) and non-LOH (allele dosage of 1). HRD scores are in the bottom right of each profile.

List of direct and indirect targets of agents from each drug category.

XLSX file - 42KB, Canonical pathways altered by FL and ICD ERBB4.

PDF file - 89KB, Genes altered by expressing ERBB4 isoforms.

XLSX file - 41KB, Gene list altered by FL and ICD ERBB4 by transcriptional prolifing.

Final MTS dataset used for analysis.

Supplementary Materials and Methods and Supplementary Figure Legends

Supplementary Figure 1. Effect of panel A drugs on TNBC cell growth in 96-well format. Supplementary Figure 2. Effect of panel B drugs alone and in combination with panel A drugs on TNBC cell growth. Supplementary Figure 3. Summary of growth inhibition by pairwise drug combinations. Supplementary Figure 4. Hierarchical clustering of cell lines based on response to drug treatments and based on expression of drug targets. Supplementary Figure 5. Principal component analysis to associate patterns of target expression with patterns of drug response. Supplementary Figure 6. Assessment of superadditivity of drug combinations. Supplementary Figure 7. A: Comparison of crizotinib versus XL-184 dose response curves when combined with ABT-263 in the six TNBC cell lines. Supplementary Figure 8. ABT-263 and crizotinib induce apoptosis in TNBC cells. Supplementary Figure 9. Expression of MET and AXL in a cohort of 59 TNBC samples from the TCGA breast cancer cohort relative to the median expression of each gene in all breast cancer samples in TCGA.

Supplementary Table 1. Single doses selected for each drug from panel A. Supplementary Table 2. Layout of the two 384-well masterplates containing Panel B drugs. Supplementary Table 4. List of Top Synergistic Drug Combinations. Supplementary Table 5. Combination Index (CI) and Dose Reduction Index (DRI) values for the top drug combinations.

Supplementary Figure 1. Flow diagram of patients and samples used in analysis. Blue boxes represent sub-cohorts used for various parts of the study.

Low-throughput validation dataset.

Revised Supplementary Methods and Legends

PDF file - 135KB, Validation of transcriptional microarray results by qRT-PCR in MCF10A and T47D cells. A) Full-length (FL) and ICD (B) ERBB4 microarray gene validation in MCF10A cells. Total RNA was isolated from Vector-MCF10A, CYT-1 MCF10A and CYT-2 MCF10A cells after 2h exposure to 100ng/ml NRG1 (Vector+, CYT-1+ and CYT-2+) in serum-free OptiMEM media, or from ICD Vector, CYT-1 and CYT-2 MCF10A cells grown in serum containing medium. C) T47D ERBB4 3'-UTR KD, pInducer20 ERBB4 validation of genes from MCF10A microarray in a luminal cell background. T47D cells were starved for 24hrs and stimulated with NRG1 for 2hrs in the presence of DOX (24hrs). Graphs show relative expression of genes as measured by qRT-PCR normalized to GAPDH.

Supplementary Figure 2. HFI mutations affecting individual recurrent genes in pre-treatment samples. Only genes affected in a minimum of 10% of the cohort are shown.

XLSX file - 285KB, Predicted activated pathways by ICD ERBB4 versus VEC.

Figure S5, Related to Figure 6. A, Assessment of SRC mRNA levels in MDA-MB-231 SRC-ORF, EMPTY-ORF, or parental cell lines. Data is normalized in RPL19 levels. B, Crystal violet staining assays on day 5 post-transfection confirms the ability of c-SRC to rescue miR-34a-induced anti-tumor growth. C-D, Examples of Pearson correlation analysis indicating MDA-MB-231 cells were not similar to BT-549 and MDA-MB-436 cells, and therefore were not included in the initial K-Means clustering analysis (results shown in D). E, The fold knockdown by miR-34a as compared to the miR-Scr treatments of the indicated genes in Clusters 1-3 in both BT-549 and MDA-MB-436 cells using a 2-fold change cut-off. None of the genes in Cluster 4 were downregulated by miR-34a (data not shown). F, Schematic of the KEGG pathway (hsa04510: Focal Adhesion) with miR-34a downregulated genes highlighted in red. G-H, Represents further analysis of the miR-34a gene signature in breast cancer. G, Indicates correlation analyses of miR-34a target genes in TNBC patients from Metabric data. H, Confirmation of prognostic importance of mIR-34a gene signature using a PROGgeneV2 algorithm on the TCGA data set.

XLSX file - 169KB, ChIP Seq-RNA Overlapping genes.