Abstract PD05-08: Genomic characterisation of invasive breast cancers with heterogeneous HER2 gene amplification

Aims: HER2 gene amplification is observed in up to 15% of breast carcinomas. In a rare subset of breast cancers classified as HER2-positive by immunohistochemistry and in situ hybridisation, HER2 overexpression and gene amplification are restricted to a subset of >30% but not all cancer cells. Here we sought to characterise the repertoire of gene copy number aberrations and somatic mutations in the HER2-positive and HER2-negative components of cases with heterogeneous HER2 overexpression and gene amplification.

Material and methods: Cases diagnosed as HER2 positive but with >30% but <100% of cells displaying HER2 overexpression were retrieved from the authors' institutions. HER2 heterogeneity status was re-assessed using immunohistochemistry and chromogenic and/ or fluorescence in situ hybridisation. For cases with confirmed HER2 gene amplification heterogeneity, HER2-positive and HER2-negative components were microdissected from tissue sections stained with the Herceptest antibody. DNA samples extracted from both components of each case were subjected to microarray-based comparative genomic hybridisation (aCGH), using a 32K BAC array platform with 50Kb resolution. The HER2-positive and HER2-negative components of cases with frozen material were also subjected to massively parallel targeted exome sequencing.

Results: Twelve cases yielded sufficient DNA for aCGH analysis. Tumours were preferentially ER positive (83%) and of histological grade 3 (67%). The HER2-positive and HER2-negative components of all cases shared most of the copy number aberrations. A pairwise comparison of the genomic profiles of the two components from each case revealed that in ten of the twelve cases, copy number aberrations in addition to 17q12 amplification encompassing the HER2 gene locus were restricted to one of the two components. Exome sequencing of two cases suggested that the HER2-positive and HER2-negative components from each case harboured >30 somatic mutations in common, including identical TP53 somatic mutations in both components of each case. The HER2-negative component of one of the cases displayed a somatic mutation in NRG2, an ERBB receptor ligand, and the HER2-negative component of the other case harboured a mutation in PTTG1IP, a proto-oncogene with putative oestrogen receptor elements.

Conclusions: Our results demonstrate that in HER2-positive breast cancers with heterogeneous HER2 gene amplification, the HER2-positive and HER2-negative components are clonally related. The distinct genomic profiles of HER2-positive and HER2-negative components, however, suggest that, at least in some of these cases, HER2 amplification may constitute a relatively late event in tumour evolution. Exome sequencing revealed mutations restricted to the HER2-negative components of HER2-positive tumours with heterogeneous HER2 overexpression/gene amplification, which may constitute potential drivers in the absence of HER2 overexpression/gene amplification.

Citation Information: Cancer Res 2012;72(24 Suppl):Abstract nr PD05-08.

Abstract P5-01-06: 18[F]-FDG Tumoral Uptake and Metabolic Response after 1 Cycle of Primary Chemotherapy According to Tumoral Phenotype of Breast Cancer

To compare the 18[F]-FDG-PET (PET) metabolic characteristics in triple-negative, luminal and positive HER2 breast cancer. Material and Methods: One hundred and forty-five patients with newly diagnosed locally advanced breast cancer were evaluated. Twenty one metastatic patients (discovered in baseline PET) were excluded and 9 patients were lost. One hundred and fifteen had undergone FDG PET before and after the first course of neoadjuvant chemotherapy. Breast cancer lesions were imaged at 80-90 minutes after administration of FDG. Maximum standardized uptake values (SUV) were measured at both time point (SUV1 and SUV 2). Metabolic response was measured by the relative decrease of SUV (ΔSUV). Using Immunohistochemistry as a surrogate for expression profiling, the tumours were classified as follows: triple negative defined by the lack of oestrogen receptors, progesterone receptors and human epidermal growth factor receptors 2 (HER2) expression, luminal (hormonal receptor positive, HER2 negative) and HER2 positive (overexpression of HER 2). Relationships between baseline [18F]-FDG uptake and clinical, histopathological and biological parameters were assessed by Mann-Whitney test. Relationships between SUV1, SUV2, ΔSUV and the tumoral phenotype were assessed by Kruskal-Wallis test.

Results: In the overall population the mean ±SD SUV1, SUV2 and ΔSUV values were 6.7± 5.9, 3.7± 3 and −36.7% ± 35.7%, respectively. Significant relationships were found between baseline FDG uptake and initial clinical and histopathological parameters : a high mitotic activity (p≥0.0001), a high nuclear pleomorphism (≥0.0002), a high tumour grading (p≥0.0001) and a negative oestrogen hormonal receptor status (p≥0.0001) were associated with a high baseline SUV. Patient age, lymph node involvement, architectural differentiation, progesterone hormonal receptor and HER2 status were not found to be related with SUV values.

Metabolic characteristics (mean ± SD) according to tumoral phenotype:

Conclusion: Our results show a significant positive relationship between the baseline FDG uptake and the proliferation markers, the tumour grading, the number of mitoses and the nuclear pleomorphism. A significant negative relationship was found between the baseline FDG uptake and the oestrogen hormonal receptor status. The triple negative phenotype was associated with higher baseline FDG uptake and higher residual FDG uptake after one course of chemotherapy commensurate with their aggressive biology. Luminal tumours showed a lower ΔSUV reflecting their lower chemosensibility.

Abstract P5-01-05: Prognostic Value at 4 Years of FDG PET, after the First Course of Neoadjuvant Chemotherapy in Breast Cancer

FDG-PET may be a powerful examination to evaluate the effectiveness of neoadjuvant chemotherapy: it appears to be possible to predict the preoperative histological response after the first cycle of chemotherapy. The aim of this prospective study was to investigate the prognostic value of the early decrease of tumoral glucidic metabolism, evaluated with FDG-PET, after the first cycle of neoadjuvant chemotherapy in breast cancer. Material and methods: FDG-PET was performed before and after the first cycle of neoadjuvant chemotherapy in 92 patients with large or locally advanced, non inflammatory, breast cancer. The change of 18F-FDG tumoral uptake was calculated from the maximum Standard Uptake Value corrected for body surface and glycemia (ΔSUVmax). Disease free survival (DFS) was assessed, and the prognostic value of ΔSUVmax was first studied among all patients. Using immunohistochemistry as a surrogate for expression profiling, patients were then classified according to tumoral phenotypes as follow: triple negative (defined by the lack of estrogen receptor (ER), progesterone receptor (PR), and human epidermal growth factor receptor 2 (HER2) expression), luminal (ER positive or PR positive, HER2 negative) and HER2 positive.

Results: The median follow-up was 46.3 months (range: 37–52 months). Twenty one patients had recurrent disease, 9 of whom died. Using univariate Cox regression analysis with all patients, DFS showed significant correlation with initial tumor size measured with ultrasound scan (p=0.03), and ΔSUVmax after the first cycle of neoadjuvant chemotherapy (cut-off = -45%, p=0.01). Using multivariate cox regression analysis, only ΔSUVmax remained an independent prognostic factor of DFS at four years (p=0.028). According to tumoral phenotypes, ΔSUVmax was a prognostic factor of DFS for the 50 patients with a luminal cancer : the risk of relapse was 8.7 times higher in women whose SUVmax decreased less than 24% after the first course of chemotherapy (p=0,04). No significant correlation between

ΔSUVmax and DFS was shown in the 22 triple negative phenotypes. For the 31 patients overexpressing HER2, a tumoral SUVmax lower than 1,8 after one cycle of chemotherapy tended to be a favorable prognostic factor with a lower risk of relapse (p = 0,058) but ΔSUVmax had no significant prognostic value (p=0.08).

Conclusion: Using immunohistochemistry, the change of tumoral FDG uptake after the first cycle of neoadjuvant chemotherapy in operable breast cancer patients appears to be an early surrogate marker of disease free survival for luminal tumors in this study. For HER2 positive tumors, FDG-PET seems to provide prognostic information but, as for the triple negative tumors, it was not statistically significant. FDG-PET might be useful to guide the early therapeutic choice for breast cancer treated with neoadjuvant chemotherapy.

Citation Information: Cancer Res 2010;70(24 Suppl):Abstract nr P5-01-05.

[Neoadjuvant chemotherapy FEC-HD in locally advanced breast cancer]

The tolerance and the clinical and histological efficacy of a neoadjuvant chemotherapy FEC-HD including hematopoietic growth factors have been studied in 40 patients with stade II or III breast cancer between February 1991 and February 1997. Four courses were given, every 21 days, with 5-fluorouracil (750 mg/m2/day D1 to D4 by continuous infusion), epirubicin (35 mg/m2/day D2 to D4) and cyclophosphamide (400 mg/m2/day D2 to D4) with G-CSF (5 mug/kg/day D6 to D15). The surgery was performed 3 or 4 weeks after the end of the chemotherapy. All patients had radiotherapy. The neoadjuvant chemotherapy induced 37.5% CR, 45% PR, and 15% SD. In 40% of the patients, the surgery was conservative. An histological CR was obtained in 15% with no axillary involvement one time out of two. There was intraductal carcinoma without invasive carcinoma in 7.5%. There was no differences between the response of inflammatory and non inflammatory tumors. One hundred and fifty-eight courses have been delivered. A grade 3 or 4 leuconeutropenia, anemia and thrombopenia have been observed in respectively 34.6%, 6.3% and 8.8% of the courses. A grade 3 or 4 mucositis has been noticed in 2.5% of the courses. A febrile granulocytopenia has occurred in 3.8% of the courses. The median survival without metastatic progression was 48 months and the median overall survival was not achieved. In stade II and III breast cancer, neoadjuvant chemotherapy with FEC-HD obtains an important histological response with an acceptable toxicity. The role of the dose-intensity increase on survival remains to be determined.