Several laboratory studies were presented at the San Antonio Breast Cancer Symposium aimed at improving the understanding of the biology and potential therapeutic targets of breast cancer. Highlighted below are a few that focused in the area of triple-negative breast cancer (TNBC).
Biomarkers of Homologous Recombination Deficiency or ‘BRCAness’ in TNBC
BRCA-mutated tumors have enhanced sensitivity to DNA cross-linking agents such as platinum and PARP inhibitors due to a defective homologous recombination (HR) DNA-repair mechanism. BRCAness, or BRCA-like phenotype, refers to the traits in sporadic cancers that are shared with those of BRCA-mutated tumors, represented by c-Myc amplification, p53 mutation, loss of RAD51-focus formation, and extreme genomic instability.
TNBC has commonly been linked to BRCAness, possibly as a result of promoter methylatioin or defects in other HR proteins since mutation in BRCA1/2 is infrequent. Identification of BRCAness or HR deficiency may help to identify TNBC sensitive to platinum and PARP inhibitors.
There has been ongoing effort to develop assays applicable in clinical practice that accurately predict BRCAness. Examples of assays in development include expression of BRCA1/2 and other HR genes, BRCA1 promoter hypermethylation, array Comparative Genomic Hybridization (aCGH) classifiers by copy number variations, and functional approaches using RAD51 foci induction:
* Homologous recombination deficiency (HRD) score: Telli et al (Abstract PD 09-04) investigated whether HRD score was predictive of pathologic response to neoadjuvant platinum-based therapy for TNBC and BRCA1/2 mutation-associated breast cancer, and therefore BRCAness. The HRD score is a tumor DNA-based assay that counts the number of loss of heterozygosity (LOH) regions of intermediate size (>15 Mb and < whole chromosome) in the tumor genome. In previous studies of ovarian cancers, the HRD score was able to identify, with a high sensitivity, defects in the HR as a result of a variety of etiologies, including BRCA1/2 mutation.
The current study was performed on 55 fresh frozen tumors from patients enrolled onto a single-arm, phase II study of neoadjuvant gemcitabine, carboplatin, and iniparib. The mean HRD score for responders (pCR/residual cancer burden [RCB1]) is 16.5 and 11.4 for non-responders. The p-value for association between response to treatment and HRD score by the Mann-Whitney U test is 0.004. Further validation is needed for its use in clinical trial and practice.
* Gain in 17q25.3: Toffoli (P3-04-04) identified a recurrent gain in 17q25.3 characteristic of BRCA1-mutated or methylated TNBC by CGH-array. DNA from 130 formalin-fixed paraffin embedded (FFPE) of breast cancer of various subtypes were extracted and genomic copy-number alterations were analyzed by CGH-array (Agilent, SuperPrint G3 Human CGH 8×60K Oligo Microarrays). A genomic region (17q25.3) was found to be amplified in 90 percent of the BRCA1-mutated tumors (29/32), compared with 30 percent (6/20) of BRCA1 non-mutated TNBC.
Four of eight cases with BRCA1 promoter methylation but non-mutated TNBC were found to have the 17q25.3 amplification. Further validation is needed for its use in clinical trial and practice.
* Level of large-scale genomic instability: Stern et al (P5-02-03) compared the genomic profiles of BRCA1/2 mutated and wild type basal-like breast cancer and demonstrated that the level of the large-scale genomic instability predicted BRCA1/2 impairment with a high sensitivity and specificity in basal-like breast cancers. Further validation is needed for this assay for its use in clinical trial and practice.
Combination Strategies to Target DNA Damage Response and Repair Mechanisms
* Inhibition of HDAC in combination with PARP inhibitor or DNA-damaging agent: Bhalla et al (S3-7) presented in vitro cell line studies that treatment with pan-histone deacetylase inhibitors including vorinostat (VS) and panobinostat (PS), induced hyperacetylation, and inhibition of chaperone function of nuclear hsp90, leading to proteasomal degradation and depletion of DNA damage response pathway proteins, including ATR and CHK1, which are essential for cell cycle arrest, and BRCA1, which is essential for DNA repair via homologous recombination, leading to BRCAness.
Combined treatment with VS or PS with ABT888 or cisplatin synergistically induced apoptosis in TNBC cells with (HCC1937) or without BRCA1 mutation (MB-231 and SUM159PT cells). The authors further demonstrated that it is HDAC3 that is responsible for the deacetylation of nuclear hsp90. This study provided a rationale for HDAC inhibitors in combination with a DNA damaging agent or PARP inhibitor in the treatment of TNBC.
* Inhibition of EGFR in combination with PARP inhibitor and platinum: Dey et al (P5-19-03) presented preclinical antitumor activity of a triplet combination of a PARP inhibitor (olaparib), carboplatin, and an EGFR/VEGF inhibitor (vandetanib) in xenograft models of BRCA-wild type TNBC cell lines. The result raised the possibility of combined targeting of EGFR and PARP in TNBC.
* Inhibition of AKT in combination with radiation in p53-deficient tumors: Connolly et al (PD09-09) presented in vitro cell culture data showing that AKT inhibition with MK-2206 sensitized radiation induce an apoptotic effect in p53-deficient TNBC cells. The authors showed that MK-2206 decreased the proportion of cells arrested in the G2/M arrest following radiation, which is necessary for DNA damage repair in p53-deficient cells. This study suggested that the involvement of PI3K pathway in DNA damage response could be explored as a synthetic lethal target for p53-deficient tumors.
Combination Strategies Targeting Growth Factor Receptor and PI3K Pathway Signaling
* Inhibition of EGFR/HER3 in combination with PI3K pathway inhibitor: Tao et al (S5-7) demonstrated that treatment with an AKT inhibitor GDC0068 or a pan PI3K inhibitor GDC0941 induced activation of EGFR and HER3 in TNBC cell lines including MDA-MB-231, MDAMB-468, HCC70, HCC1143, and HCC1937. Combined therapy with MEHD7945A, a dual-action antibody that targets both EGFR and HER3, and GDC0068 or GDC0941 was superior to monotherapy in vitro and in vivo using xenograft models derived from either TNBC cell lines or patient-tumor. This study provided a rationale for combination therapy of PI3K pathway inhibition and upstream receptor tyrosine kinase inhibition.
* Inhibition of Androgen receptor in combination with PI3K pathway inhibitor: Lehmann et al (P6-05-03) previously identified six distinct molecular TNBC subtypes that include two basal-like (BL1 and BL2), an immunomodulatory (IM), a mesenchymal (M), a mesenchymal stem-like (MSL), and a luminal androgen receptor (LAR) subtype. At this meeting, the authors demonstrated that androgen receptor (AR)-positive TNBC tumors frequently possess PIK3CA kinase mutations (69.2%).
Treatment with AR antagonists and bicalutamide increased the pAKT and the combination of bicalutamide with PI3K/mTOR inhibitors GDC0980 or NVP-BEZ235 decreased tumor volume in mouse xenograft studies, providing a rationale for the combination of AR antagonists with PI3K/mTOR inhibitors.
Novel Strategies to Target Myc-Amplified TNBC
* Myc-amplification occurs in about 40 percent of TNBC and is associated with a particularly poor prognosis. To search for new “synthetic lethal” strategies to treat Myc-overexpressed TNBC, Goga et al (S3-8) performed a high-throughput shRNA synthetic-lethal screen using about 2,000 shRNAs, which target the human kinome (about 600 kinases), in human mammary epithelial cells (HMECs) that have inducible MYC over-expression.
The authors identified 13 kinases targeted by at least two shRNAs and gave more than 50 percent growth inhibition, including genes involved in metabolism (GSK3A, AMPK-A1, ARK5, HUNK, LKB1) and MAPK signaling (MAPK, MAP2K, MAP3K, JNK1).
One of the genes highlighted by the authors was PIM1, a proto-oncogene serine/threonine kinase, a co-factor for Myc. They demonstrated that overexpression of PIM1 occurs frequently in Myc-overexpressed tumors and in basal-like or TNBC and was associated with a poor prognosis. PIM1 deletion or pharmacologic inhibition induced cell death in Myc-overexpressing HMECs and TNBC cells, indicating a synthetic-lethality interaction. This study provides a rationale for the development of PIM1 inhibitors in TNBC with Myc overexpression.
In addition to the studies highlighted above, several presentations investigated tumor specimens collected at baseline and/or post-neoadjuvant chemotherapy to uncover genetic alterations for potential therapeutic targets and predictors of chemotherapy efficacy. Balko et al (S3-0) suggested that JAK2 amplification may provide additional important therapeutic targets.
In summary, there has been steady progress in the understanding of the biology of TNBC. This SABCS has pointed to several potential therapeutic strategies that should be high priorities in clinical trial development for the treatment of patients with TNBC. We look forward to next year's SABCS for additional development in this area.