Ploidy analysis by in situ hybridization of interphase cell nuclei in fine-needle aspirates from breast carcinomas: correlation with cytologic grading

A four gene signature predicts benefit from anthracyclines: evidence from the BR9601 and MA.5 clinical trials

Chromosome instability (CIN) in solid tumours results in multiple numerical and structural chromosomal aberrations and is associated with poor prognosis in multiple tumour types. Recent evidence demonstrated CEP17 duplication, a CIN marker, is a predictive marker of anthracycline benefit. An analysis of the BR9601 and MA.5 clinical trials was performed to test the role of existing CIN gene expression signatures as predictive markers of anthracycline sensitivity in breast cancer. Univariate analysis demonstrated, high CIN25 expression score was associated with improved distant relapse free survival (DRFS) (HR: 0.74, 95% CI 0.54-0.99, p = 0.046). High tumour CIN70 and CIN25 scores were associated with aggressive clinicopathological phenotype and increased sensitivity to anthracycline therapy compared to low CIN scores. However, in a prospectively planned multivariate analysis only pathological grade, nodal status and tumour size were significant predictors of outcome for CIN25/CIN70. A limited gene signature was generated, patients with low tumour CIN4 scores benefited from anthracycline treatment significantly more than those with high CIN4 scores (HR 0.37, 95% CI 0.20-0.56, p = 0.001). In multivariate analyses the treatment by marker interaction for CIN4/anthracyclines demonstrated hazard ratio of 0.35 (95% CI 0.15-0.80, p = 0.012) for DRFS. This data shows CIN4 is independent predictor of anthracycline benefit for DRFS in breast cancer.

Chromosome 7 aneusomy in metaplastic breast carcinomas with chondroid, squamous, and spindle-cell differentiation

Metaplastic breast carcinomas (MBCs) are basal-like tumors that frequently express epidermal growth factor receptor (EGFR) via an unknown underlying genetic mechanism. In this study, the EGFR/CEP7 gene copy number in 17 MBCs with chondroid, squamous, and spindle-cell differentiation showing EGFR expression by immunohistochemistry was analyzed using fluorescence in situ hybridization. All cases had a balanced EGFR/CEP7 ratio. EGFR gene amplification was not observed in any case. Monosomy was found in 25% and polysomy in 12.5% of carcinomas with chondroid differentiation. All spindle-cell carcinomas and 50% of squamous carcinomas showed trisomy. Comparison with CEP7 copy number revealed aneusomy of chromosome 7, as opposed to increases or decreases specific to the EGFR gene or 7p. Although no direct correlation between EGFR expression by immunohistochemistry and aneusomy was observed, cases with a score of 3+ showed a higher frequency of EGFR gene copy gain. In the absence of EGFR amplification, chromosome 7 aneusomy might be a useful criterion for the determination of potential candidates for EGFR tyrosine kinase inhibitor clinical trials.

EGFR gene copy number heterogeneity in fine-needle aspiration cytology from breast carcinomas determined by chromogenic in situ hybridization

Most studies have shown epidermal growth factor receptor (EGFR) overexpression to be associated with poor prognostic factors in breast carcinomas. The relationship to EGFR gene copy number is unclear. The aim of our study was to investigate the heterogeneity of the EGFR gene copy number in breast carcinomas. The material consisted of air-dried smears from 29 breast carcinomas and 3 breast cancer cell lines (MCF-7, SKBR3, and T47D). Chromogenic in situ hybridization (CISH) was done using chromogenic detection. The mean signal numbers for EGFR gene and chromosome 7 as well as the EGFR gene/chromosome 7 centromere probe (CEP7) ratio were recorded. Immunohistochemical (IHC) staining was done on the corresponding paraffin sections. The copy number of the EGFR gene in each tumor/cell line ranged from 1.2 to 5.6. The EGFR gene/CEP7 ratio showed a biological continuum ranging from 0.59 to 1.94 with a mean of 1.04. EGFR gene copy loss was found in 16.6% of cases whereas copy gain was demonstrated in 19.4%. There was no relationship between IHC protein expression of EGFR and EGFR gene copy number or EGFR gene/CEP7 ratio.In conclusion, most breast carcinomas had a balanced EGFR gene/CEP7 copy number with a mean ratio of 1.04. Almost equal subpopulations revealed limited copy gain and copy loss. EGFR high dosage amplification, like in HER-2, was not demonstrated. Demonstration of EGFR gene copy loss might have a potential as a surrogate marker for EGFR gene mutation and/or deletion.

Patterns of aneusomy for three chromosomes in individual cells from breast cancer tumors

Multi-color fluorescence in situ hybridization (FISH) can determine the changes in the copy numbers of several chromosomes simultaneously and can therefore be used to identify aneusomic patterns in individual cells. Aneusomic patterns may be useful for determining the malignant nature of rare epithelial cells in the blood of cancer patients. Touch preparations from 74 primary breast tumors were evaluated for aneusomy of chromosomes 1, 8 and 17 by tri-color-FISH. In the first part of the analysis, percentages of aneusomy for individual chromosomes and their combinations were determined. In the second part of the analysis, aneusomic patterns for these three chromosomes were analyzed in individual tumor cells and compared to aneusomic patterns observed in leukocytes and in individual cells from benign and normal breast tissue to determine aneusomic patterns indicative of malignancy. Ninety-two percentage of the primary breast carcinomas showed aneusomy for one or more enumerator probes. Comparison with benign breast tissue identified six aneusomic patterns in individual carcinoma cells indicative for malignancy by statistical analysis and not observed in leukocytes. Hence, certain patterns of aneusomy in individual cells involving chromosomes 1, 8 and 17 are indicative of malignancy in individual breast tumor cells and may be useful for determining malignancy of rare epithelial cells in the blood of breast cancer patients.

Detection and quantitation by fluorescence in situ hybridization (FISH) and image analysis of HER-2/neu gene amplification in breast cancer fine-needle samples

BACKGROUND

Fine-needle sampling, although a practical and noninvasive method of tissue acquisition, has rarely been used for HER-2/neu fluorescent in situ hybridization (FISH). To assess HER-2/neu gene amplification in mammary carcinoma, FISH signals on cytology and corresponding tissue biopsies were detected visually and measured by image analysis. The results were correlated with patient and tumor characteristics.

METHODS

In situ HER-2/neu DNA probe hybridization was performed on 61 cytology specimens and on 47 corresponding frozen sections of breast carcinomas. Tumors were classified by visual evaluation as unamplified, moderately amplified, or highly amplified. Multiparametric image analysis was performed using the Discovery automated image analyzer (Becton Dickinson, Leiden, Netherlands). The integrated fluorescence ratio (IFR) was calculated for each sample as the integrated FISH fluorescence of the tumor cells divided by the integrated FISH fluorescence of internal control cells containing two spots. The percentage positive nuclear area (PPN), calculated as the area of FISH fluorescence divided by the area of nuclear DNA fluorescence, and the PPR, ratio of the PPN of the tumor cells divided by the control cells, were also calculated for each sample.

RESULTS

Visual analysis yielded 46 unamplified and 15 (24.6%) amplified (seven moderately amplified and eight highly-amplified) tumors. Strong (P < 0.001) correlation between results on cytological and histological materials was obtained. The FISH spots on the cytological preparations were more easily visualized and scored than those on the corresponding tissue sections. Visual HER-2/neu signal scoring was strongly correlated with IFR (P = 0.0001) and PPR (P = 0.0001). Within the tumors classified as highly amplified by visual examination, quantitation of the degree of amplification fluorescence signal was possible using image analysis.

CONCLUSIONS

Cytologic specimens were a suitable and representative source of materials for detection and quantitation of HER-2/neu gene amplification by FISH and image analysis. Cancer (Cancer Cytopathol)

Estimating loss of the wild-type p53 gene by in situ hybridization of fine-needle aspirates from breast carcinomas

TP53 mutations have been found in 16-64% of breast carcinomas. The aim of our study was to investigate loss of the wild-type TP53 gene by in situ hybridization (ISH) of fine-needle aspirates (FNAC) from breast carcinomas. The material consisted of FNAC from 33 breast carcinomas, with histologic specimens from 19 of the cases. Routine diagnostic smears were used for cytologic grading. ISH of the wild-type TP53 gene and chromosome 17 was performed on air-dried smears. Hybridization signals were counted in at least 100 nuclei, and the percentage for each signal number was calculated. FNAC from four fibroadenomas as well as cell preparations from five lymphocyte cultures were used as normal/benign controls. Cutoff for defining loss of p53 gene signals was set at 20% of cells with zero and one gene signal only. Concomitant p53 protein expression was determined on 20 histologic sections and eight additionally available air-dried smears. Loss of wild-type p53 gene was found in 20 carcinomas (60.6%). The rate of signal loss varied from 0.4% to 75.3% of the cells. All tumors with aneusomy of chromosome 17 revealed loss of p53 gene signals, as did 42% of the disome cases. Loss of wild-type p53 gene was present in 10 of 16 grade 1 cancers (62.5%), eight of 13 grade 2 tumors (61.5%), and two of four grade 3 cases. Signal loss did not correlate with p53 protein expression. In conclusion, subpopulations with loss of the wild-type p53 gene are a common finding in breast carcinomas; they are detected in more than 60% of the tumors, including grade 1 cancers.

Numerical abnormalities of chromosome 7 in interphase cell nuclei of breast carcinoma have no impact on immunohistochemically determined EGFR status

AIM OF STUDY

To investigate the relationship between epidermal growth factor receptor (EGFR) status and numerical aberrations of chromosome 7 in breast carcinomas.

DESIGN

In situ hybridization (ISH) of interphase cell nuclei on air-dried fine-needle aspirates (FNAC) from 33 breast carcinomas was evaluated for numerical abnormalities in chromosome 6, 7, 12 and 17. Immunohistochemical staining of EGFR was performed on corresponding histological specimens.

RESULTS

78% of the tumours were aneuploid by ISH. Aneusomy of chromosome 7 was found in 18 cases (60%). EGFR overexpression was observed in 30% of the carcinomas, and seven of nine were aneuploid by ISH. The same percentage of chromosome 7 aneusomy was found in both EGFR-positive and -negative cases. Five of seven EGFR-positive tumours revealed aneusomy of chromosome 7.

CONCLUSION

Numerical gain of chromosome 7 is a common finding, occurring in about 60% of breast carcinomas. Most EGFR-positive tumours are aneuploid and show numerical gain of chromosome 7, but abnormal numbers of chromosome 7 have no impact on the EGFR status.

Numerical aberrations of chromosome 17 in interphase cell nuclei of breast carcinoma cells: lack of correlation with abnormal expression of p53, neu and nm23 protein

The genes for p53, neu (c-erbB-2) and nm23 are all located on chromosome 17. Abnormal expression of their protein products is an important prognostic parameter. The aim of this study was to investigate if numerical aberrations of chromosome 17 are reflected in the expression of these markers. The immunohistochemical expression was analysed on histological specimens from 33 breast carcinomas. In situ hybridization (ISH) was performed on interphase cell nuclei in air-dried fine-needle aspirates from the same cases using a digoxigenin-labelled alpha-satellite probe for chromosome 17. ISH for chromosome 6, 7 and 12 was used additionally to give an estimate of ploidy. Of the carcinomas 76% were aneuploid, and numerical abnormalities of chromosome 17 were found in 34%. Abnormal p53 protein was expressed in 15% (five cases). All of these were aneuploid, but only one of them revealed aneusomy of chromosome 17. Neu overexpression was found in 18% of the tumours (six cases). Five of these were aneuploid, whereas two were aneusome for chromosome 17. Four cancers showed full (normal) expression of nm23 protein, whereas 29 had reduced expression. Reduced expression was found in 23 of 25 aneuploid tumours. Numerical aberrations of chromosome 17 were found equally in carcinomas with reduced and full nm23 protein expression. Abnormal numbers of chromosome 17 seem only to have a minor impact on these markers and are not reflected significantly in their expression.