Attendee Confcal 1 Welcome Box

Browsing Over 172 Presentations

Sort By Upcoming First

Biomarkers and translational research and precision medicine (ID 666)

5P - First clinical validation of the physical biomarker based on the nanomechanical tissue profiling in breast cancer for prediction of treatment outcome in a neoadjuvant therapy setting (ID 590)

Abstract

Background

Neoadjuvant therapy (NAT) is evolving as a standard of care in breast cancer (BC) patients, but only 50% of the patients achieve complete response. Early prediction of therapy response, essential for treatment individualization, remains challenging. Subcellular mechanical alterations are emerging as biomarkers of cancer aggressiveness that can potentially aid treatment optimization. We explore if nanomechanical phenotyping of a tissue biopsy prior to first treatment cycle can identify patients that will respond to therapy in a neoadjuvant setting.

Methods

The Swiss-based monocentric, prospective, double blinded study was conducted on 545 patients to evaluate the potential of nanomechanical tissue profiling measured by atomic force microscopy based ARTIDIS (Automated and Reliable Tissue Diagnostics) technology as a physical biomarker for rapid BC diagnosis, prognosis and prediction of treatment outcome. This is a follow up report focused on the predictive value of nanomechanical signature for NAT response. Out of 31 patients assigned to NAT, 27 received chemotherapeutic regimens and 4 patients received endocrine therapy. ARTIDIS measurements were performed on the naïve biopsy and the tumor regression was determined by radiological imaging before and after NAT. The relative size and distribution of stiffness is integrated into a custom developed data analytics tool that systematically analyzes nanomechanical signatures from each patient in correlation to therapy response based on imaging and clinical follow up parameters.

Expected Results

The distribution of patients according to therapy response and the predictive accuracy (ROC) of the nanomechanical signature will be provided.

Expected Conclusions

Our exploratory study based on the follow up data from the Swiss Nano study will provide evidence if nanomechanical profiling of tumor tissue at diagnosis can be utilized for treatment outcome prediction and consequently therapy optimization. Validation studies are planned to further investigate the potential of the nanomechanical signature to provide guidance in neoadjuvant treatment of BC patients.

Clinical trial identification

BASEC 2015-00170

Collapse

6P - 25-year survival and benefit from tamoxifen therapy by the clinically used breast cancer markers in lymph node-negative and ER-positive/ HER2-negative breast cancer (ID 450)

Abstract

Background

The clinically used breast cancer markers are known to predict short-term survival, but whether these markers predict long-term (25-year) survival is unclear. We therefore aimed to determine whether the clinically used markers are long-term prognosticators and predictors of tamoxifen therapy benefit in patients with lymph node-negative and ER-positive/ HER2-negative breast cancer.

Methods

Secondary analysis of the Stockholm Tamoxifen (STO-3) trial conducted from 1976 to 1990, randomizing postmenopausal lymph node-negative breast cancer patients to receive adjuvant tamoxifen therapy versus no adjuvant therapy.

Distant recurrence-free interval (DRFI) by the clinically used markers was assessed by Kaplan-Meier and multivariable Cox proportional hazards analysis. Recursive partitioning analysis was performed to evaluate which markers best predict long-term survival.

Results

A statistically significant difference in 25-year DRFI was seen by tumor size (Log-rank, P<0.001) and tumor grade (Log-rank, P=0.019), but not by PR and Ki-67 status.

Patients with smaller tumor size (T1a/b Hazard ratio [HR], 0.31; 95% CI, 0.17-0.55; T1c HR, 0.58; 95% CI, 0.38-0.88), and tumor grade 1 had a significant reduced long-term risk (Grade 1 HR, 0.48; 95% CI, 0.24-0.95), compared to patients with larger (T2) tumor size and grade 3 tumors, respectively.

A significant tamoxifen therapy benefit was suggested for patients with larger tumor size (T1c HR, 0.53; 95% CI, 0.32-0.89; T2 HR, 0.34; 95% CI, 0.16-0.73), lower tumor grade (Grade 1 HR, 0.24; 95% CI, 0.07-0.82; Grade 2 HR, 0.50; 95% CI, 0.31-0.80), and PR-positivity (HR=0.38, 95% CI, 0.24-0.62).

Recursive partitioning analysis selected tumor size as the most important characteristic to predict survival, followed by trial arm for patients with larger tumors.

Conclusions

Our findings suggest that tumor size followed by grade are significant 25-year prognosticators of DRFI outcome. Further, a significant 25-year benefit from tamoxifen therapy was seen in patients with larger tumor size, lower tumor grade and PR-positive tumors.

Clinical trial identification

The trial center for the STO-3 trial was the Regional Cancer Center Stockholm-Gotland, in Stockholm Sweden. However, the start of the Stockholm Tamoxifen trial (STO-3) in 1976 was well before trial registration started in Sweden, therefore information on trial number is not available.

Collapse

7P - An optimized Breast Cancer Index node-positive (BCIN+) prognostic model for late distant recurrence in patients with hormone receptor-positive (HR+) node-positive breast cancer (ID 435)

Abstract

Background

BCI is a gene expression-based assay that reports a prognostic BCI score that significantly predicts risk of overall (10y), early (0-5y), and late (≥5y) distant recurrence (DR) in HR+, node-negative (N0) and node-positive (N1) breast cancer. The BCIN+ prognostic model was trained in the Trans-ATAC cohort, which evaluated primary adjuvant anastrozole versus tamoxifen. The current study optimized the BCIN+ prognostic model for late DR in N+ patients from the arm treated with 7.5 years of endocrine therapy in the translational cohort of the IDEAL trial.

Methods

Patients with 1 to 3 positive nodes (N1) in the 7.5-year endocrine treatment arm of the translational IDEAL cohort were used to examine cut-points for BCIN+ model across the range of 1 to 9 to classify patients into Low- and High-risk groups. Kaplan-Meier analysis was used to calculate the 15-year (post-diagnosis, 10-year post-randomization) late DR free interval (DRFI) as the primary endpoint. The cut-point was selected based on the classification of a low-risk group with <5% 15-year late DRFI. Initial validation of the optimized prognostic model was performed in a single institution retrospective cohort using Cox proportional hazards regression.

Results

241 N1 IDEAL patients (85% ≥ 55 y, 54% T1, 52% G2) were included. Evaluation of the BCIN+ prognostic model led to an adjusted cut-point, which classified 44 patients as BCIN+ Low Risk with a 15-year late DRFI of 3.2%, and 197 patients as BCIN+ High Risk with a DRFI of 19.0% (HR: 7.11, 95% CI: 0.97-52.17; p=0.024). Independent validation in a retrospective cohort of 349 patients (64% ≥55 y, 66% T1, 58% G2) showed that the optimized BCIN+ model was significantly prognostic for late DR (HR: 9.25, 95% CI: 1.27-67.45; p=0.007), and classified 66 and 283 patients as BCI Low- and High-risk with 1.6% and 15.2% 15-year late DRFI, respectively.

Conclusions

An optimized BCIN+ prognostic model was determined from patients in the randomized IDEAL trial, which was significantly prognostic for late DR in HR+ N1 patients. Additional studies in randomized N+ cohorts are required for further validation of this BCIN+ model optimized for late DR.

Clinical trial identification

BOOG 2006-05

Collapse

8P - Mutational analysis of circulating tumor DNA (ctDNA) in patients with ER+/HER2- advanced breast cancer (ABC) receiving palbociclib (P): results from the TREnd trial. (ID 286)

Abstract

Background

Cyclin-dependent kinase 4 and 6 inhibitors like palbociclib (P) are a mainstay of treatment for ER+/HER2- ABC; however intrinsic or acquired resistance is a major clinical issue. We performed a mutational analysis on ctDNA samples from patients (pts) included in the c-TREnd study, the translational cohort of the TREnd trial (NCT02549430) which randomized pts to receive P alone or in combination with the endocrine treatment (ET) to which they had progressed in the previous line of ET.

Methods

Forty-six pts were enrolled in c-TREnd. Plasma was collected before treatment (T0), after the first cycle of therapy (T1) and at the time of progression (T2). Hybridization and capture were performed using the TruSight Tumor 170 Kit (Illumina). Single nucleotide variants (SNVs) were detected and annotated using LoFreq and Oncotator, and further refined by ABEMUS. Tumor Mutational burden (TMB) was the sum of silent and non-silent mutations. Tumor fraction (TF) was based on the dispersion of Copy Number Alteration (CNA) genomic profiles. Progression free survival (PFS) was estimated using the Kaplan–Meier method and compared with the log-rank test.

Results

Thirty-two pts (87 samples), 14 from the P arm and 18 from the P+ET arm, were included in the final analysis. The most frequently mutated genes at T0 were ESR1 (23%), PIK3CA (17%), AR , FGFR2 and TP53 (10%). At T0, mutations in ESR1, but not in PIK3CA, were significantly prognostic (median PFS – mPFS 3.7 mo vs 11 mo in ESR1 mut vs WT, p=0.015). A significantly worse mPFS was observed when a broader analysis of PI3K pathway (adding AKT1, PTEN, TSC1/2 and BRAF) was performed (mPFS 5.2 m in mut vs 10.8 m in WT, p=0.04). Mutations in AR tended to confer a better, although not statistically significant, mPFS (14.2 mo vs 5.5 in WT, p=0.29). At T2 we observed the emergence of 9 new mutations in 7 genes (ESR1, AKT1, ARID1A, BRIP1, CCNE1, MTOR and TP53). TMB and TF at T0 or TF change between T1 and T0 were not associated with PFS.

Conclusions

Mutations in ESR1 and in PI3K pathway genes were associated with worse prognosis in pts treated with P, while TMB and TF were not. Larger studies are needed to validate these observations.

Collapse

9P - BRCA1 promoter methylation confers a more favorable prognosis to systemically untreated young triple-negative breast cancer patients than tumor BRCA1 mutation (ID 361)

Abstract

Background

The prognoses of systemically treated, triple-negative breast cancer (TNBC) patients with a pathogenic tumor BRCA1 mutation (tBRCA1m) or BRCA1 promoter methylation (BRCA1 PM) have been widely studied. However, the prognosis for systemically untreated women remains unknown. This study investigates the prognosis of systemically untreated young N0 TNBC patients according to tumor BRCA1 status.

Methods

Dutch women aged < 40 years, diagnosed with TanyN0M0 TNBC between 1989 and 2000 were selected from the Netherlands Cancer Registry. In that era, N0 patients were considered low risk and not given (neo)adjuvant systemic therapy. We analyzed tBRCA1m and BRCA1 PM using DNA from formalin-fixed paraffin-embedded tumor tissues. We built Cox and competing risk regression models, for invasive disease-free survival (iDFS), and distant recurrence-free survival (DRFS) with secondary primary tumors (SPTs) as competing events. Both models were adjusted for tumor characteristics and locoregional treatment.

Results

For 373 patients, tBRCA1m and BRCA1 PM status were available. Of these, 28% had pathogenic tBRCA1m, 36% had BRCA1 PM and the rest were classified as BRCA1 dual-negative. Compared to patients with BRCA1 dual-negative tumors, patients with BRCA1 PM had a favorable iDFS (adjusted hazard ratio [aHR] = 0.66, 95% confidence interval [CI] = 0.45 – 0.98) but similar DRFS (subdistribution hazard ratio [sHR] = 0.86, 95% CI = 0.50 – 1.46), while patients with tBRCA1m had poorer iDFS (aHR = 1.86, 95% CI = 1.30 – 2.65) and also similar DRFS (sHR = 1.04, 95% CI = 0.61 – 1.76). Furthermore, patients with BRCA1 PM had lower risk for SPTs (sHR = 0.38, 95% CI = 0.17 – 0.83), while patients with tBRCA1m had higher risk (sHR = 3.12, 95% CI = 1.79 – 5.45).

Conclusions

Although tBRCA1m and BRCA1 PM both cause BRCA1 gene inactivation, the iDFS differed significantly between systemically untreated young TNBC patients with these tumor types. This can be mainly attributed to substantially different risk for SPTs. Interventions to prevent SPTs, such as contralateral prophylactic mastectomy or secondary chemoprevention, should be considered for young tBRCA1m TNBC patients.

Collapse

10P - The clinical actionability of PTEN protein and gene expression analysis in HR- and HER2+ breast cancers (ID 593)

Abstract

Background

Phosphatase and tensin homologue (PTEN) loss is associated with tumorigenesis, tumor progression, and therapy resistance in breast cancer. However, the clinical value of PTEN as a biomarker requires further studies. We seek to identify clinically relevant subtypes of breast cancer based on PTEN status and other clinicopathologic features.

Methods

A cohort of 608 breast cancer patients previously profiled for PTEN protein expression was included. Based on the expression on the neoplastic cells compared to the normal internal controls by IHC, cases were classified as PTEN-low (PTEN-L) or PTEN-retained (PTEN-WT). The former constituted the study group, while the latter the control group. Analysis of gene expression was performed on 3,929 patients from the METABRIC and MSK cohorts retrieved from cBioPortal. The Shapiro-Wilk test was used to analyze the normal distributions of continuous variables. Relationships between PTEN status and the clinicopathologic and molecular features of the patient population were assessed using Fisher’s exact test or Chi-squared/Wilcoxon rank-sum test. Survival curves were built according to the Kaplan‐Meier method.

Results

Reduced expression of PTEN was significantly different at protein and gene levels, where the former was observed in 46.1% (n=280/608), while, not surprisingly, the latter in only 7.8% (n=308/3,929) cases. PTEN-L tumors were significantly enriched for in both HER2+ (n=63, 22.5%) and triple-negative (TN) (n=41, 14.6%) subtypes compared to PTEN-WT tumors (n=34, 10.4% and n=18, 5.5%, respectively; p<0.0001). When the relative expression of PTEN was decreased, both hormone receptor (HR)- and HER2 overexpression/amplification were significantly related to worse overall survival (OS) compared to the HR+/HER2- status (p<.0001). Of note, this condition was not statistically significant in the presence of a PTEN-WT expression. Moreover, PTEN-L protein expression but not gene expression was related to worse OS in HR+/HER2+ tumors compared to HR+/HER2- (p=0.002).

Conclusions

The combined analysis of PTEN protein and gene expression may provide additional data to perform a tailored risk assessment while evaluating patients with HR- and HER2+ breast cancers.

Collapse

11P - A Breast Cancer Index (BCI) prognostic model for N0 HR+ breast cancer optimized for late distant recurrence (ID 214)

Abstract

Background

BCI is a gene expression-based assay that reports a prognostic score for risk of overall (10y) and late (≥5y) distant recurrence (DR) in HR+ early-stage breast cancer. The current study optimized the BCI prognostic model for late DR utilizing N0 patients from the translational aTTom (Trans-aTTom) study.

Methods

N0 patients in the 5y tamoxifen arm of the Trans-aTTom cohort were used to examine BCI assay cut-points based on classification of a Low-risk group with >95% 15y (10y post-randomization) late DR free survival (DRFS) as estimated by Kaplan-Meier analysis. Validation was performed in an independent multi-institutional cohort using Cox proportional hazards regression.

Results

697 N0 patients (81% ≥55y, 71% T1, 44% G2) were included. A cut-point was determined with a 15y DRFS of 95.7% and 88.7% in the BCI Low- and -High Risk groups, respectively. At 10y and 20y post-diagnosis, patients were stratified as BCI-Low Risk (45%) with a DRFS of 98.0% and 92.9%, and as BCI-High Risk (55%) with a DRFS of 94.3% and 86.4% (HR: 2.16, [1.23-3.80]; p=0.006), respectively. Independent validation in 312 patients (47% <55y, 67% T1, 62% grade 2) showed that the optimized BCI model was significantly prognostic for late DR (HR: 2.16, p=0.006), stratifying BCI Low- (45%) and High-risk (55%) patients with 96.7% and 87.9% 10-year late DRFS, respectively.

Cohorts BCI Risk Groups 10y late DRFS (95% CI) 15y late DRFS (95% CI) 20y late DRFS (95% CI)
Trans-aTTom (n=697) BCI-Low 98.0% (96.4-99.6%) 95.7% (93.3-98.1%) 92.9% (89.3-96.8%)
BCI-High 94.3% (91.9-96.7%) 88.7% (85.3-92.2%) 86.4% (82.6-90.4%)
Multi-institutional (n=312) BCI-Low 96.7% (93.6-99.9%)
BCI-High 87.9% (82.8-93.3%)

Conclusions

In the current study, an optimized BCI prognostic model developed in the Trans-aTTom cohort was significantly prognostic for late DR in HR+ N0 breast cancer. Additional studies in cohorts with extended follow-up are required for further validation of this BCI late DR model.

Clinical trial identification

ISRCTN17222211; NCT00003678

Collapse

12P - The RODILIA pilot study for molecular screening of patients with metaplastic breast cancer (ID 575)

Abstract

Background

Metaplastic breast cancer (MPBC) is a rare disease characterized by aggressive features and dismal prognosis after standard therapy. Herein, we report the molecular screening of the Milan National Cancer Institute case series for the evaluation of potentially druggable alterations.

Methods

A total of 49 MPBC cases treated with curative intent were identified. Primary tumors were profiled using Oncomine Comprehensive Assay Plus panel (Thermo Fisher Scientific) for copy number alteration (CNA), mutation, tumor mutational burden (TMB), and microsatellite instability (MSI) analyses, according to the manufacturer instructions.

Results

Sequencing results are presented for the first 34 pathological reviewed cases. Quality control metrics were met in 33 cases. In total, 94 unique genes harbored at least one mutation representing 24% of the panel. The median number of mutations indexed per patient was 3.5 (range 0-29). Notably, 25 cases showed actionable mutated genes, including PTEN, mTOR, FGFR3, FGFR4. Most of the cases showed low TMB, the median value being 4.5 (range 0-28). MSI status was high only in 2 cases. Common CNAs included 13q (10%), 5q (9%) and 17p (6%). Eight out of ten canonical cancer pathways (cell cycle, Hippo, MYC, NOTCH, PI3K, RTK-RAS, TGFβ and β-catenin/WNT) were altered by both mutational and CNA events occurring at different proportions, with mutational events involving up to 33%, and CNA involving up to 42% of genes of the altered pathway. In the Hippo and RTK-RAS pathways the two types of alterations were instead equally represented whereas the remaining pathways (β-catenin/WNT, TGFβ, PI3K, NOTCH, MYC and Cell cycle) were more affected by CNA than mutations. NRF2 and TP53 signaling pathways were instead activated by mutational events only.

Conclusions

Mutational and copy number alterations conveyed complementary information in MPBC cooperating in activation of cancer pathways.These findings suggest to further study the value of CNAs in MPBC biological processes, especially immunogenicity, which cannot be explained by the low TMB and MSI found. Extended data with matched immune profile will be presented at the meeting.

Collapse

13P - Comparison study of different programmed death-ligand 1 (PD-L1) assays, readers and scoring methods in triple-negative breast cancer (TNBC) (ID 365)

Abstract

Background

Different immunohistochemical programmed death-ligand 1 (PD-L1) assays and scoring methods in triple-negative breast cancer (TNBC) have been reported to yield variable results. We compared the analytical concordance and interobserver variability of four clinically developed PD-L1 assays assessing immune cell (IC) score and combined positive score (CPS) in TNBC.

Methods

Archival primary TNBC resection specimens (n = 99) were stained for PD-L1 using VENTANA SP142, VENTANA SP263, DAKO 22C3 and DAKO 28-8. PD-L1 expression was scored by four trained readers according to guidelines for IC-score and CPS on whole slide images by virtual microscopy.

Results

The mean PD-L1 positivity ranged between 53%-74% for IC-score ≥1% and CPS ≥1 with highest levels for SP263. The concordance between IC-score ≥1% and CPS ≥1 across all readers for each assay was >93%. Intra-class correlation coefficients (ICCs) revealed poor-to-good inter-reader agreement for each assay for IC-score (0.489-0.793, highest value for SP142) and moderate-to-good agreement for CPS (0.653-0.794, highest value for 22C3). ICCs for each reader (ranging from 0.226-0.595) uncovered poor-to-moderate inter-assay agreements on PD-L1-positivity for both scores. The concordance of all readers across the four assays was 86.8% for CPS ≥1 and 88.4% for IC-score ≥1%. Kappa scores for inter-reader agreement for each assay at IC-score ≥1% were 0.728-0.777 and for CPS ≥1 0.680-0.735. Evaluation of inter-assay agreement for each reader revealed kappa scores 0.446-0.596 at IC-score ≥1% and 0.492-0.587 at CPS ≥1.

Conclusions

We demonstrate a certain degree of concordance between IC-score and CPS for the assessment of PD-L1-positivity across different assays and readers. However, the four PD-L1 assays are analytically not fully concordant.

Collapse

14P - Immunomodulatory effect of denosumab in early breast cancer: preliminary results of a randomized window-opportunity clinical trial D-Biomark (NCT03691311). (ID 384)

Abstract

Background

Most breast cancers (BC) exhibit low immune infiltration and are unresponsive to immunotherapy. Hence, the urgency to find new mechanisms of immune activation, postulating receptor activator of nuclear factor kappa-Β ligand (RANKL) and its receptor RANK as potential immunomodulator. Our previous data demonstrated that RANK pathway inhibitors, such as denosumab, used for the treatment of bone metastasis, could also prevent and/or treat BC and regulate the tumour immune crosstalk. However, the population of breast cancer patients who may benefit from denosumab remains to be identified.

Methods

Patients with early-stage HER2-negative BC, candidates to tumour excision as first therapeutic approach are included. Patients are randomized 2:1 to denosumab: control (no treatment); experimental arm received 2 doses of 120 mg subcutaneous denosumab (once per week) before surgery (2-4 weeks later). Putative changes in tumour cell proliferation by Ki67 immunohistochemistry (IHC), cell survival by cleaved caspase 3 IHC (primary endpoints) and stromal tumour infiltrating lymphocytes (TILs) quantified in the haematoxylin and eosin by between baseline (biopsy sample) and surgery are evaluated. Specific antibodies will be used to characterize infiltrating immune populations. Denosumab driven gene expression changes in tumour samples will be analysed and tools such as CIBERSORT will be used to characterize the immune infiltrate.

Results

We present results from the first 36 patients enrolled out of 60. Clinical and tumour characteristics were well balanced between the groups. No relevant toxicities were reported. No clinically significant differences in Ki67 and cleaved caspase-3 were observed after denosumab treatment. Interestingly, a statistically significant increase in TILs was observed in the denosumab treated group (p=0.03, Paired t test) but not in the control group (p=0.80). A 33% of patients treated denosumab showed a ≥10% increase in TILs vs 0% in the control group (p =0.05).

Conclusions

Short term neoadjuvant denosumab increases TILs in early BC.

Clinical trial identification

NCT03691311

Collapse

15P - The combined influence of receptor subtype, grade and TN status on breast cancer survival in recently treated women with non-metastatic disease in Norway (ID 432)

Abstract

Background

Receptor subtypes of breast cancer, as defined by immunohistochemistry, are surrogates for molecular subtypes. Despite new emerging molecular tumour classifications, these receptor subtypes are well-known independent predictors of breast cancer death and remain widely used in clinics. Few studies have disentangled the combined influence of receptor subtypes, grade, size (T) and nodal status (N) on breast cancer survival in large groups of recently treated patients.

Methods

From the Cancer Registry of Norway, we obtained detailed clinical information on 24137 women with invasive breast cancer aged 20 to 74 between 2005 and 2015. Of these, 17204 had non-metastatic breast cancer with known receptor status. Receptor subtype was defined by estrogen receptor (ER), progesterone receptor (PR) and human epidermal growth factor receptor 2 (HER2) status. Hazard ratios for breast cancer death up to 13 years after diagnosis were estimated using flexible parametric survival models and comparing combinations of receptor subtype, grade and TN status with adjustments for age, year and surgery type. Additional adjustment for adjuvant treatment was done on a subset of women with known treatment information.

Results

Receptor subtype, grade and TN status were strong predictors for breast cancer death, both independently and in combination. The combined effect of all factors was a 20- to 40-fold higher breast cancer mortality rate when comparing to women with the best outcome (ER+PR+HER2-, low grade, T1N0). In women with ER+HER2- subtypes and no nodal spread, a larger tumour size was associated with a significantly higher risk of breast cancer death. Women with ER+HER2- tumours also had an increased late (>5 years) mortality which was largely explained by intermediate/high grade and nodal spread. Women with ER+PR-HER2-, high grade, N+ tumours had particularly high mortality similar to triple negative breast cancer.

Conclusions

These results highlight the importance of thoroughly combining well-known tumour factors to describe the wide range of risks of dying from breast cancer, also among small, node negative tumours.

Collapse