Triple-Negative Breast Cancer Time to Slice and Dice? Carsten Denkert, MD Charité University Hospital Berlin, Germany
Triple-Negative Breast Cancer (TNBC) 2018 Presentation Outline The molecular heterogeneity and clinical diversity of TNBC Pathologic and molecular characteristics of both hereditary and nonhereditary subtypes of the disease Different approaches for subtyping Potential prognostic biomarkers and therapeutic targets in TNBC PD-L1 and other immunologic biomarkers Androgen-receptor signaling The relevance of deficiencies in DNA repair Current technologies for evaluating defects in DNA repair
The Simple View Three Subtypes of Breast Cancer Triple-negative (15%) HR-positive (70%) HER2-positive (15%) Chemotherapy Endocrine therapy Anti-HER2 therapy HER2, human epidermal growth factor receptor 2; HR, hormone receptor
The Simple View Extended Subtypes of TNBC (?) Triple-negative (15%) Stem-like AR-pos Mesenchymal Immune Basallike 2 Basallike 1 Challenges to define TNBC subtypes: Different subgroups Mixture of tumor-cell features and tumor-microenvironment Subtypes are overlapping No established diagnostic assay AR, androgen receptor positive; pos, positive
Molecular Model for TNBC Biologic Features of a Heterogeneous Disease HRD, homologous recombination deficiency; METABRIC, Molecular Taxonomy of Breast Cancer International Consortium; TIL, tumor-infiltrating lymphocytes; TCGA, The Cancer Genome Atlas Denkert C, et al. Lancet. 2017;389(10087):2430-2442.
TNBC Pathologic Characteristics The classical histology High-grade IDC Triple-Negative High nuclear grade Mitoses Solid growth Event-Free Survival, % pcr no pcr High proliferation Time Since Randomization, Years 33% pcr rate IDC, invasive ductal carcinoma; pcr, pathologic complete response Cortazar P, et al. Lancet. 2014;384(9938):164-172.
TNBC Special Situations The classical histology High-grade IDC Salivary gland like tumors better prognosis, not real TNBC High nuclear grade Mitoses Solid growth Medullary BC immune activation Observed in other TNBCs as well No clearly defined entity High proliferation Minimal ER expression (1% to 9%) Borderline between luminal and TN tumors BC, breast cancer; TN, triple-negative Rakha EA, et al. J Clin Oncol. 2008;26(15):2568-2581.
Different Approaches to Molecular Subtyping of TNBC Lehmann BD, et al. J Clin Invest. 2011;121(7):2750-2767. Analysis restricted to the TNBC group. Sørlie T, et al. Proc Natl Acad Sci U S A. 2001;98(19):10868-10874. Comparison of all breast cancer subtypes. Burstein et al. Clin Cancer Res. 2015;21(7):1688-1698. Analysis restricted to the TNBC group. 6 groups: Basal-like 1 Basal-like 2 Immunomodulatory Mesenchymal-like Mesenchymal stem-like Luminal/AR 2 groups: basal-like claudin-low 4 groups: Basal-like immuno suppressed Basal-like immune activated Mesenchymal Luminal/AR Overlapping signatures in different approaches Currently no standardized diagnostic test for molecular subtyping Reduction of 2188-gene algorithm (Lehmann) to a 101-gene algorithm was successful (Ring BZ, et al. BMC Cancer. 2016;16:143.) Consistent molecular motives: Luminal/AR Immune-activated Two types of basallike tumors Mesenchymal Basis for molecular targeted therapy of TNBC
TNBC Subtypes Have Different Therapy Response Rates Lehmann groups 2011 6 groups: Basal-like 1; Basal-like 2 Immunomodulatory; Mesenchymal-like Mesenchymal-stem-like; Luminal/AR UNS BL1 BL2 IM M MSL LAR P<.04 pcr, % P<.01 pcr, % P<.011 pcr, % Lehmann BD, et al. PLoS One. 2016;11(6):e0157368. Highest response rate: Lowest response rate: BL1 BL2; Luminal/AR
Immune Activation Is Observed in Many TNBC Subtypes Lehmann groups 2011 6 groups: Basal-like 1; Basal-like 2 Immunomodulatory; Mesenchymal-like Mesenchymal-stem-like; Luminal/AR UNS BL1 BL2 IM M MSL LAR Lehmann BD, et al. PLoS One. 2016;11(6):e0157368. Lehmann BD, et al. J Clin Invest. 2011;121(7):2750-2767.
Immune Activation Is Observed in Many TNBC Subtypes Lehmann groups 2011 6 groups: Basal-like 1; Basal-like 2 Immunomodulatory; Mesenchymal-like Mesenchymal-stem-like; Luminal/AR UNS BL1 BL2 IM M MSL LAR Lehmann groups 2016 4 groups: Basal-like 1; Basal-like 2 Mesenchymal-like; Luminal/AR BL1 BL2 M LAR excluded 2016 IM = presence of immune cells MSL = presence stromal cells Immune activation TILs Lehmann BD, et al. PLoS One. 2016;11(6):e0157368. Lehmann BD, et al. J Clin Invest. 2011;121(7):2750-2767. immune genes
Immune Gene Signature: Correlation With TIL Morphology and Chemotherapy Response in GeparSixto cold hot intermediate PDL1 IDO1 PD1 CTLA4 pcr Rate, % 60 50 40 30 20 24 37 56 CXCL9 10 CD8A CCL5 CXCL13 0 60 Immune-A Immune-B Immune-C Immune-type A C B Denkert C, et al. J Clin Oncol. 2015;33(9):983-991. IGKC CD21 FOXP3 CD80 HR+ HRstrLy (%) LPBC n=481 LPBC Cases, % 50 40 30 20 10 0 50 P<.00001 19 1 Immune-A Immune-B Immune-C
TILs and Increased Chemotherapy Response Rates in Breast Cancer Combined Analysis of 6 Neoadjuvant Trials (N = 3771) 62.5 P<.0005 P<.0005 P<.0005 P<.0005 pcr Rate, % 50. 37.5 25. 19.6 27.0 43.7 28.5 48.5 49.8 38.7 32.1 30.8 31.4 Low (0% to 10%) Intermediate (11% to 59%) High ( 60%) 12.5 11.0 5.9 0. n= 1667 1369 725 759 435 172 605 512 262 260 373 273 All Patients Lum/HER2- HER2+ TNBC pcr: ypt0ypn0 Denkert C, et al. Lancet Oncol. 2018;19(1):40-50.
Univariate analysis Multivariate analysis all baseline parameters Multivariate analysis all baseline parameters and pcr Prognosis: High TILs Improved Survival in HER2+ and TNBC high TILs = better survival Disease-Free Survival low TILs = better survival HR per 10% P value All tumors 1.02 ns TNBC 0.93.011 HER2+ 0.94.017 Lum/HER2-1.02 ns All tumors 0.94.001 TNBC 0.91.003 HER2+ 0.92.004 Lum/HER2-0.99 ns All tumors 0.96.037 TNBC 0.95 ns HER2+ 0.94.041 Lum/HER2-1.01 ns high TILs = better surival Overall Survival low TILs = better survival HR per 10% P value All tumors 1.01 ns TNBC 0.92.032 HER2+ 0.94 ns Lum/HER2-1.10.011 All tumors 0.96 ns TNBC 0.91.020 HER2+ 0.92.061 Lum/HER2-1.09.044 All tumors 0.99 ns TNBC 0.95 ns HER2+ 0.96 ns Lum/HER2-1.11.014 Denkert C, et al. Lancet Oncol. 2018;19(1):40-50. ns, not significant
TNBC Strata: -TILs (low/med/ high) N = 174 R GeparNuevo Study Design Window of opportunity Durvalumab Placebo 2 weeks Window phase stopped after patient 117 (request of IDMC) Core biopsy nab-pac +Durvalumab nab-pac +Placebo 12 weeks Clinical response ECx4 +Durvalumab ECx4 +Placebo 8 weeks Samples* Samples** Samples** Samples** Surgery Primary endpoint: - pcr ypt0 ypn0 Secondary biomarker endpoints: - TILs - Ki-67 - Immune mrna marker - PD-L1 *Tissue: FFPE, fresh frozen; Liquid biopsies: full blood; plasma, serum **Tissue: FFPE, fresh frozen; Liquid biopsies: plasma, serum
GeparNuevo: Complete Cohort and Window Cohort - pcr ypt0, ypn0 80% 70% 60% 50% 40% Complete cohort n=174 (primary endpoint) Window subcohort (exploratory) n=117 53.4% P =.287* Adjusted** OR 1.53 [95%CI 0.82-2.84] P =.182 44.2% 80% 70% 60% 50% 40% 61.0% P =.052 41.4% 30% 20% 10% 0% Durvalumab Placebo N = 88 N = 86 Loibl S, et al. J Clin Oncol. 2018;36(suppl): Abstract 104. 30% 20% 10% 0% Durvalumab Placebo N = 59 N = 58 * Continuous corrected χ² test ** For stratification factor (TIL groups)
GeparNuevo: Subgroup Analysis GeparNuevo preliminary conclusion: Timing of immunotherapy might be important Ongoing translational research Validation in additional trials Loibl S, et al. J Clin Oncol. 2018;36(suppl): Abstract 104.
Genomics of TNBC: Number of Mutations in Breast Cancer Mutational Burden (Whole-Exome NGS Sequencing) P = 1.4e-14 P = 1.4e-10 1000 1000 number of mutated genes 100 10 1 23 27 43 number of mutated genes 100 10 1 27 40 41 50 G1 G2 G3 HR+ HER2+ HER2+ TNBC HER2- HR+ HR- Budczies J, et al. J Pathol Clin Res. 2015;1(4):225-238.
Mutational Burden, Therapy Response, and Prognosis MDACC cohort, n = 29 TCGA cohort, n = 101 P =.05 P =.029 Jiang T, et al. PLoS Med, 2016;13(12): e1002193.
BRCA Mutations in TNBC The majority of hereditary breast cancers show a TN-phenotype (75% to 80%) But still, most TNBC are nonhereditary BCs BRCA mutations are relevant for response to PARP-inhibitors and/or platin-therapy BRCAness : there are tumors that show a homologous recombination deficiency (HRD) without a BRCA mutation these might also respond to PARP-inhibitors/platin-therapy
HRD - Biomarker Strategies and Tumor Heterogeneity Different Assays Identify Overlapping Populations Methods for HR deficiency: NGS sequencing for BRCA In normal cells (germline) In tumor cells HDR score Genomic scars Different methods Different cutpoints HRDetect HRD based on WES GeparSixto: 19.8% gbrca mutant 30.3% tbrca mutant 66.8% HRD score high 70.5% any HR deficiency von Minckwitz G, et al. Cancer Res. 2017; 77(4 Suppl): Abstract P1-09-02.
Additional Options HRDetect Somatic mutational signatures (=mutation patterns) Physiologic readout of the DNA damage and DNA repair processes that have occurred through tumorigenesis Measured by whole-genome sequencing 98.7% sensitivity for BRCA1/2-mutated tumors Davies H, et al. Nat Med. 2017;23(4):517-525.
HRD and BRCA As Biomarkers in Clinical Trials GeparSixto Trial (neoadjuvant, early BC): Increased overall CTX response with gbrca1 mutations No carboplatin difference with gbrca1 mutations Positive carbo effect in BRCA wildtype (Increased pcr with HRD, no additional carbo effect) TNT Trial (advanced/metastatic BC): Increased carboplatin response with gbrca1 mutations (No difference with HRD assay) Vn Minckwitz G et al ASCO 2104 Von Minckwitz G, et al. J Clin Oncol. 2014;32(5s):Abstract 1005. Updated by Hahnen E, in press Tutt A, et al. Nat Med. 2018;24(5):628-637.
Mina A, et al. Onco Targets Ther. 2017;10:4675-4685. Parker JS, et al. J Clin Oncol. 2015;33(15 suppl):1083-1083. Vera-Badillo FE, et al. J Natl Cancer Inst. 2014;106(1):djt319. Additional Biomarkers: Androgen Receptors in TNBC AR is present in 13% to 37% of TNBC Detectable by immunohistochemistry Gene expression analysis under development Target for anti-androgenic therapy AR positive TNBC: Better prognosis (Consider AR testing in low-proliferative TNBC)
Additional Biomarkers PD-L1 Expression in Breast Cancer Study n %PD-L1 Positive Wang ZQ, et al. Oncotarget. 2017;8(31):51641-51651 Muenst S, et al. Breast Cancer Res Treat. 2014;146(1):15-24. Bertucci F, et al. Oncotarget. 2015;6(15): 13506-13519. Ali HR, et al. Ann Oncol. 2015;26(7):1488-1493. 441 16.5% all tumors 35.4% in TNBC 650 23.4% all tumors 112 38% of inflammatory BCs 3916 PD-L1 immune cells 6% PD-L1 tumor cells 1.7% Membranous staining; associated with better RFS in TNBC Membranous or cytoplasmic staining; associated with worst survival in all subtypes Staining compared to normal breast tissue; Correlation with increased pcr rate Evaluation on TMAs Open questions: Best cutpoints Tumor cells vs TILs Role in immunotherapy trials Currently not a biomarker for clinical practice
Take-Home Message TNBC TNBC is a clinical term, not a defined disease entity Molecular features of TNBC include: High immune infiltrate AR expression BRCA mutations / BRCAness Mesenchymal and stem-like features Basal-like phenotype Increased mutational load Each molecular alteration is found only in a subset of TNBC; there is overlap between subtypes They cannot be used for stable subtyping of TNBC, but they are the basis for new therapeutic options Denkert C, et al. Lancet. 2017;389(10087):2430-2442.