Plasma ESR1 mutations and outcome to first-line paclitaxel and bevacizumab in patients with advanced ER-positive/HER2-negative breast cancer

BACKGROUND

ESR1 mutations have been identified as mechanism for endocrine resistance and are also associated with a decreased overall survival. We assessed ESR1 mutations in circulating tumor DNA (ctDNA) for impact on outcome to taxane-based chemotherapy in advanced breast cancer patients.

METHODS

ESR1 mutations were determined in archived plasma samples from patients treated with paclitaxel and bevacizumab (AT arm, N = 91) in the randomized phase II ATX study. Samples collected at baseline (n = 51) and at cycle 2 (n = 13, C2) were analyzed using a breast cancer next-generation sequencing panel. This study was powered to detect a benefit in progression-free survival (PFS) at six months for patients treated with paclitaxel/bevacizumab compared to historical trials with fulvestrant. PFS, overall survival (OS), and ctDNA dynamics were exploratory analyses.

RESULTS

PFS at six months was 86% (18/21) in patients with an ESR1 mutation detected and 85% (23/27) in wildtype ESR1 patients. In our exploratory analysis, median progression-free survival (PFS) was 8.2 months [95% CI, 7.6-8.8] for ESR1 mutant patients versus 8.7 months [95% confidence interval (CI), 8.3-9.2] for ESR1 wildtype patients [p = 0.47]. The median overall survival (OS) was 20.7 months [95% CI, 6.6-33.7] for ESR1 mutant patients versus 28.1 months [95% confidence interval (CI), 19.3-36.9] for ESR1 wildtype patients [p = 0.27]. Patients with ≥ two ESR1 mutations had a significantly worse OS, but not PFS, compared to those who did not [p = 0.003]. Change in ctDNA level at C2 was not different between ESR1 and other mutations.

CONCLUSIONS

Presence of ESR1 mutations in baseline ctDNA might not be associated with inferior PFS and OS in advanced breast cancer patients treated with paclitaxel/bevacizumab.

Abstract P1-09-20: An optimized workflow to analyze ESR1 mutations in both circulating cell-free and circulating tumor cell DNA by digital PCR

Background

In metastatic breast cancer (MBC) patients ESR1 mutations (mESR1) in cell-free DNA (cfDNA) have been related to endocrine therapy (ET) resistance. Such mutations might also be detectable in circulating tumor cells (CTCs). Mutation detection in small amounts of cfDNA and in CTCs in a background of leukocytes is highly challenging. The current study evaluated how to reliably investigate mESR1 status in such minute amounts of cfDNA and in DNA from CellSearch-enriched CTCs.

Materials & Methods

Plasma (200 µL) and matched CellSearch-enriched CTC fractions of 7 healthy blood donors (HBD) and 29 MBC patients at baseline and after ET (≥ 5 CTC/7.5 mL) were evaluated. cfDNA was isolated from plasma with the QIAamp CNA kit and CTC-enriched DNA with the AllPrep kit (Qiagen). mESR1 status in both cfDNA and CTC-enriched DNA fractions was compared with or without whole genome amplification (repli-g WGA, Qiagen) or ESR1 target specific amplification. Quantitative PCR (qPCR) for wild type (WT) ESR1 was used to control the number of WT copies loaded into the chips for digital PCR (dPCR) analysis. The variant allele frequencies (VAF) of hotspot mutations for ESR1 (D538G, Y537S, Y537C and Y537N) were evaluated with mutation-specific Taqman assays by chip-based dPCR (QuantStudio 3D, Thermo Fischer Scientific).

Results

To allow inclusion of as many samples as possible, we successfully downscaled the volume of required plasma from 1 mL to 200 µL as this resulted in the same VAF. Sample-type specific thresholds for mESR1 presence were established (2% for the cell-free plasma samples, at which percentage all HBDs were negative, and 0.5% for the CTCs to allow identification of one mutated CTC-specific copy in a background of ~1,000 leukocytes).

WGA was unable to adequately amplify fragmented cfDNA, resulting in a too low DNA yield. However, locus-specific target pre-amplification of a 136 bp fragment covering all 4 different mutations followed by mutant specific dPCR performed well for both cfDNA and CTC DNA, but only if the loading of the pre-amplified product into the dPCR chips was optimized by qPCR for the number of WT ESR1 copies.

The most optimal results for dPCR data interpretation were obtained after: 1. including at least one positive sample in each dPCR session; 2. using a “safe loading window”, 3. loading and reading chips at least twice in QuantStudio 3D ; 4. critically evaluating the contribution by a non-specific “comet effect”; and 5. after loading the data in the software, performing at least two independent data analyses to exclude intra-observer variations.

Summary

Here we describe our workflow to assess mESR1 in a limited amount of plasma cfDNA or CellSearch enriched CTC DNA. This workflow has been successfully used to investigate the mESR1 VAF status in DNA from matched CTC DNA and cfDNA of MBC patients before start of 1st line endocrine therapy and at progression (see also abstract number 851017).

Citation Format: Vitale SR, Sieuwerts AM, Helmijr J, Beije N, van der Vlugt – Daane M, Foekens JA, Sleijfer S, Jansen MPHM, Martens JWM. An optimized workflow to analyze ESR1 mutations in both circulating cell-free and circulating tumor cell DNA by digital PCR [abstract]. In: Proceedings of the 2016 San Antonio Breast Cancer Symposium; 2016 Dec 6-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2017;77(4 Suppl):Abstract nr P1-09-20.

Abstract P5-08-51: SIAH2 protein expression is inversely correlated with the ER status and outcome to tamoxifen therapy in metastatic breast cancer patients

Introduction: In a previous study we observed a positive correlation between Seven in Absentia Homolog 2 (SIAH2) and Estrogen Receptor (ER) mRNA levels. Additionally, high SIAH2 mRNA levels were related to a favorable progression-free survival (PFS) after first-line tamoxifen. In contrast, others showed high SIAH2 protein levels in ER-negative breast cancer associated with an unfavorable relapse-free survival. In this study, we investigated the above discrepancy between SIAH2 protein and mRNA findings and evaluated the prognostic and predictive value of SIAH2 protein in breast cancer patients.

Patients and methods: Tissue microarrays (TMAs) of formalin-fixed, paraffin-embedded primary breast tumors were immunohistochemically stained for SIAH2 protein. The TMAs contained core specimens of 759 patients with early disease and of 245 ER-positive patients with advanced disease treated with first-line tamoxifen. SIAH2 protein staining was scored for its intensity and proportion positive cells and subsequently evaluated for its relationship with metastasis-free survival (MFS) and PFS in uni- and multivariate analyses including traditional prognostic or predictive factors, respectively.

Results: The proportion SIAH2-positive cells had a relationship with MFS and PFS, whereas staining intensity and a previous described score for SIAH2 combining intensity and proportion were not related with clinical outcome. Based on these results, tumors with more than 20% positive cells were considered as SIAH2-positive. In early disease, 267 patients (35%) had SIAH2-positive tumors, which were further characterized by decreased expression of ER at protein and mRNA levels (P <0.001 and P = 0.003, respectively). These SIAH2-positive tumors correlated with significant unfavorable MFS in lymph node negative, ER-positive breast cancer patients, but only in univariate analysis. In advanced disease, 86 patients (35%) had SIAH2-positive tumors which was associated with an unfavorable PFS after first-line tamoxifen in both uni- and multivariate analyses (HR = 1.45; 95% CI, 1.07 to 1.96; P = 0.015).

Conclusions: SIAH2 protein expression is especially observed in ER-negative tumors and has no additional prognostic value in breast cancer. The proportion SIAH2-positive cells in ER-positive tumors can be used as biomarker to predict tamoxifen treatment failure in breast cancer patients with advanced disease. Future studies should establish if expression of certain microRNAs explain the observed discrepancy in SIAH2 mRNA and protein levels.

Citation Format: van der Willik KD, Timmermans MM, Look MP, Reijm EA, van Deurzen CHM, den Bakker MA, Westenend PJ, Martens JWM, Berns EMJJ, Jansen MPHM. SIAH2 protein expression is inversely correlated with the ER status and outcome to tamoxifen therapy in metastatic breast cancer patients. [abstract]. In: Proceedings of the Thirty-Eighth Annual CTRC-AACR San Antonio Breast Cancer Symposium: 2015 Dec 8-12; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2016;76(4 Suppl):Abstract nr P5-08-51.

High protein expression of EZH2 is related to unfavorable outcome to tamoxifen in metastatic breast cancer

BACKGROUND

Metastatic breast cancer (MBC) is a highly heterogeneous disease with great differences in outcome to both chemo- and endocrine therapy. Better insight into the mechanisms underlying resistance is essential to better predict outcome to therapy and to obtain a more tailored treatment approach. We have previously described that increased mRNA expression levels of Enhancer of Zeste homolog (EZH2) are associated with worse outcome to tamoxifen therapy in MBC. Here, we explored whether this is also the case for EZH2 protein expression.

PATIENTS AND METHODS

A tissue microarray (TMA) was created using formalin-fixed, paraffin-embedded estrogen receptor (ER)-positive primary breast tumor tissues of 250 MBC patients treated with first-line tamoxifen. Quantity and intensity of EZH2 expression were determined by immunohistochemistry (IHC) and both were used to generate and group scores according to a previously described method for scoring EZH2.

RESULTS

In total, 116 tumors (46%) were considered to be EZH2 positive. The presence of EZH2 protein expression was significantly associated with progression-free survival (PFS) in both univariate [hazard ratio (HR) 1.51, 95% confidence interval (CI) 1.17-1.97, P = 0.002] and multivariate analysis including traditional factors associated with tamoxifen outcome (HR 1.41, 95% CI 1.06-1.88, P = 0.017). Considering quantity irrespective of intensity, tumors with >50% EZH2-positive cells had the worst PFS (HR 2.15, 95% CI 1.42-3.27, P < 0.001), whereas intensity alone did not show a significant association with PFS. Application of other methods of scoring EZH2 positivity resulted in a similar significant association between the amount of EZH2 positive cells and PFS.

CONCLUSION

In addition to EZH2 mRNA levels, these results suggest that protein expression of EZH2 can be used as a marker to predict outcome to tamoxifen therapy. This provides new rationale to explore EZH2 inhibition in the clinical setting and increases the possibilities for a more personalized treatment approach in MBC patients.

Abstract P3-06-01: Hotspot mutations in PIK3CA are predictive for treatment outcome on aromatase inhibitors but not for tamoxifen

Introduction: PIK3CA is the most frequent (30%) mutated oncogene in breast cancer and may lead to an activation of the PI3K/AKT/mTOR-pathway. Cell lines resistant to tamoxifen have an activated PI3K-pathway. Phase III clinical trials show substantial benefit when mTOR-inhibitors are added to aromatase inhibitor treatment. On the other hand, patients with PIK3CA exon 20 mutation expression signature show better treatment outcome after adjuvant tamoxifen therapy. To address this controversy, we evaluated the PIK3CA mutation status in 1423 primary breast cancer specimens for its relationship with prognosis and treatment outcome after first-line endocrine treatment.

Methods: Hotspot mutations in exon 9 and 20 of PIK3CA were detected by multiplex snapshot analyses. Mutation status in ER-positive tumors was related to metastasis free survival (MFS) in 292 untreated lymph node negative (LNN) patients and time to progression (TTP) in patients with metastatic disease treated with first-line tamoxifen (N = 482) or aromatase inhibitors (AIs; N=103). Whole genome mRNA and miRs expression profiling was performed in the latter patient subset to develop a PIK3CA gene signature in 64 specimens. This was validated in 28 independent ER-positive specimens.

Results: We could evaluate 1371 specimens and detected 437 hotspot mutations for PIK3CA (32%). Mutations in exon 20 were detected in 256 patients (59%), of which 40 cases with a H1047L (16%) and 216 with a H1047R (84%) mutation. Mutations in PIK3CA exon 9 were discovered in 174 patients (41%), with E542K and E545K mutations in 59 (34%) and 105 (60%) cases, respectively, as the most prevalent ones. Finally, 6 patients had PIK3CA double mutations for both exon 9 and 20 and one patient had a E542K and E545K mutation.

Evaluation of the untreated LNN patients for prognosis showed no relationship between MFS and PIK3CA mutations (HR = 1.07 [95% CI: 0.73–1.56]; p = 0.73), neither for exon 9 nor exon 20 compared to wild-type. In the multicenter cohort of 482 patients with advanced disease treated with first-line tamoxifen no link with treatment outcome and PIK3CA mutation status was observed (HR = 1.13 [95% CI: 0.92–1.38]; p = 0.24). However, patients with advanced disease treated with first-line AIs (N = 64) showed a significant longer TTP for patients with a PIK3CA mutation compared to wild-types (HR = 0.46 [95% CI: 0.25–0.87]; p = 0.017).

Expression profiles of mRNAs and miRs were integrated and resulted in signatures for PIK3CA status discovered by pathway (17 genes) and expression analyses (10 genes and 9 miRs). Validation and comparison with published signatures in the 28 independent tumors showed that our 10-genes signature had the highest PIK3CA prediction accuracy (75%), with 60% sensitivity and 78% specificity. Moreover, this signature better associates with TTP (HR = 0.38 [95% CI: 0.20–0.72]; P = 0.003).

Conclusion: Mutations in PIK3CA are not prognostic value in untreated ER-positive LNN patients, not predictive in ER-positive patients with advanced disease treated with first-line tamoxifen therapy, however, are predictive for favourable outcome after first-line AIs. Moreover, we propose an expression signature of 10 genes as a putative biomarker to predict the PI3K status and response to AIs.

Citation Information: Cancer Res 2012;72(24 Suppl):Abstract nr P3-06-01.

High miR-26a and low CDC2 levels associate with decreased EZH2 expression and with favorable outcome on tamoxifen in metastatic breast cancer

For patients with metastatic breast cancer, we previously described that increased EZH2 expression levels were associated with an adverse outcome to tamoxifen therapy. Main objective of the present study is to investigate miR-26a and miR-101 levels, which both target EZH2, for their association with molecular pathways and with efficacy of tamoxifen as first-line monotherapy for metastatic breast cancer. Expression levels were measured using quantitative Real-Time Polymerase Chain Reaction (qRT-PCR) in primary breast cancer specimens of 235 estrogen receptor-α (ER)-positive patients. Pathway analysis was performed on microarray data available for 65 of these tumors. Logistic regression and Cox uni- and multivariate analysis were performed to relate expression levels with clinical benefit and time to progression (TTP). Increasing levels of miR-26a were significantly (P < 0.005) associated with both clinical benefit and prolonged TTP, whereas miR-101 was not. Cell cycle regulation and CCNE1 and CDC2 were the only significant overlapping pathway and genes differentially expressed between tumors with high and low levels of miR-26a and EZH2, respectively. In addition, increasing mRNA levels of CCNE1 (P < 0.05) and CDC2 (P < 0.001) were related to poor outcome. Multivariate analysis revealed miR-26a and CDC2 as an optimal set of markers associated with outcome on tamoxifen therapy, independently of traditional predictive factors. To summarize, only miR-26a levels are related with treatment outcome. Cell cycle regulation is the only overlapping pathway linked to miR-26a and EZH2 levels. Low mRNA levels of EZH2, CCNE1, and CDC2, and high levels of miR-26a are associated with favorable outcome on tamoxifen.

Decreased expression of EZH2 is associated with upregulation of ER and favorable outcome to tamoxifen in advanced breast cancer

The purpose of this study is to investigate EZH2 in a large series of breast cancer patients for its prognostic and predictive value, and to evaluate its functional role in treatment response in vitro. EZH2 levels were measured using quantitative Real-Time Polymerase Chain Reaction (qRT-PCR) in primary breast cancer specimens and related to clinicopathologic factors and disease outcome. EZH2 expression was downregulated with siRNAs in MCF7, to assess expression alterations of putative EZH2 downstream genes and to determine cell numbers after treatment with the anti-estrogen ICI 164384. In 688 lymph node-negative patients who did not receive adjuvant systemic therapy, EZH2 was not significantly correlated with metastasis-free survival (MFS). In 278 patients with advanced disease treated with first-line tamoxifen monotherapy, the tertile with highest EZH2 levels was associated with the lowest clinical benefit (OR = 0.48; P = 0.02) and with a shorter progression-free survival (PFS) in both univariate (HR = 1.80; P < 0.001) and multivariate analysis, including traditional factors (HR = 1.61; P = 0.004). In vitro, EZH2 silencing in MCF7 caused a 38% decrease in cell numbers (P < 0.001) whereas ICI 164384 treatment resulted in a 25% decrease (P < 0.001) compared to controls. Combining EZH2 silencing with ICI treatment reduced cell numbers with 67% (P < 0.001) compared to control conditions. EZH2 downregulation was associated with an almost two-fold upregulation of the estrogen receptor alpha (ER) (P = 0.001). In conclusion, EZH2 has no prognostic value in breast cancer. High levels of EZH2 are associated with poor outcome to tamoxifen therapy in advanced breast cancer. Downregulated EZH2 leads to upregulation of the ER and better response to anti-estrogens.

Serum proteomic patterns for ovarian cancer monitoring

We set out to discover ovarian cancer biomarkers useful for monitoring progression during and after chemotherapy and possibly for diagnosis. Surface-enhanced laser desorption/ionization time-of-flight mass spectrometry was used to create serum protein profiles of ovarian cancer patients before chemotherapy or at progression (n = 51) (trial initiated by the Gynecological Cancer Cooperative Group of the European Organization for Research and Treatment of Cancer trial) that were compared with those of healthy individuals (n = 31). In addition, sera profiles from ovarian cancer patients after chemotherapy (n = 12) were compared with those of ovarian cancer patients at progression (n = 24). One of the discovered biomarkers was identified and subsequently confirmed and validated using enzyme-linked immunosorbent assay (ELISA). Eight primary (sens = 94%, spec = 97%, P < 0.0001) and seven progression tumor biomarkers (sens = 91%, spec = 97%, P < 0.0001) were discovered. In addition, we discovered eight potential progression monitoring biomarkers (sens = 75%, spec = 83%, P = 0.0008) of which one, a biomarker of 11.7 kd, was further identified as serum amyloid A1. Independent validation (ELISA) showed an elevated expression of this protein at relapse in four of the seven ovarian cancer patients tested. Combining the eight newly discovered progression monitoring biomarkers with CA125 resulted in a clear increase of the sensitivity (91-100%). These biomarkers, in combination with for instance CA125, should be validated in large ovarian cancer and control groups. The resulting multimarker assay could be suitable for disease monitoring during and after therapy and might also be useful for ovarian cancer screening.