HER2/neu Gene Amplification Heterogeneity: The Significance of Cells With a 3:1 HER/CEP17 Ratio

Lester J. Layfield, MD and Robert L. Schmidt, MD, PhD, MBA From the Department of Pathology, University of Utah and ARUP Laboratories, Salt Lake City, UT

Abstract
Criteria for Diagnosis of Heterogeneity College of American Pathologists published guidelines for assessment of HER2/neu genetic heterogeneity. When twenty cells are counted for evaluation of HER2/neu amplification, a single 3:1 HER2/CEP17 ratio cell characterizes the sample as heterogeneous. Heterogeneity for HER2/neu amplification may indicate biologically important characteristics including likelihood of amplification in metastases. We performed fluorescence in situ hybridization on 1546 cases. For each case, twenty cells of invasive carcinoma were analyzed for HER2/CEP17 ratio. Cases were assessed as non-amplified (ratio < 1.8), borderline amplified (1.8 < ratio < 2.2) or amplified (ratio > 2.20). Heterogeneity was present when the percentage of cell with ratios greater than 2.20 was 5% but < 50%. Individual cells were typed by probe ratios and distribution of cell types determined. The distribution of HER2/CEP17 ratio was determined with the number of 3:1 HER/CEP17 cells plotted against the number of amplified cells. 3:1 HER2/CEP17 ratio cells occur with low frequency (2.2%) but are the determining factor for heterogeneity in 46% of heterogeneous cases. 35% of heterogeneous cases were due to a single 3:1 cell. Single 3:1 cells are a poor predictor for additional amplified cells. Inclusion of cells with a 3:1 HER2/CEP17 ratio in the definition of heterogeneity may be too broad as these cells are the determining factor in approximately one-third of diagnoses of heterogeneity but are not strongly associated with other measures of amplification. Moreover, 3:1 HER2/CEP17 ratio of cells are a poor predictor for the presence of additional amplified cells in a sample.
Abnormal Cells (Excluding 3:1) 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 Total Number of 3:1 HER2/CEP17 cells per case 0 911 75 17 12 2 2 2 2 1 1 2 0 1 2 1 3 4 7 14 25 49 1,13 3 1 16 6 36 23 7 7 1 3 2 1 0 1 2 3 2 0 0 2 3 3 5 0 26 7 2 3 6 1 3 1 1 5 0 0 4 0 2 1 2 1 1 1 0 2 1 0 1 0 0 8 1 3 1 0 5 1 2 2 1 1 3 2 2 1 1 1 0 0 0 1 1 0 0 0 3 4 4 2 2 2 2 1 1 1 0 0 0 1 0 0 0 0 0 0 0 0 0 0 1 2 5 6 7 8 4 0 0 1 3 0 0 0 1 2 2 0 1 0 0 0 0 0 1 0 0 0 0 0 0 0 0 1 4 0 0 0 1 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 1 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Total Cases 1,130 Heterogeneous by 3:1 cells n = 219

Results
In total, 198 of 1,546 cases (12.8%) were either amplified or borderline amplified. Four hundred seventy four of the 1,546 (31%) cases were heterogeneous using the CAP definition. A total of 30,920 individual cells were counted with 683 of these cells (2.2%) disclosing a 3:1 HER2/CEP17 ratio. Thus, 3:1 ratio of HER2/CEP17 cells occurred at a relatively low frequency (2.2%). Abnormal cells defined by a ratio of HER2/CEP17 > 2.2 accounted for 12.5% of all cells if 3:1 HER2/CEP17 ratio cells are counted as abnormal and 10.3% of all cells if 3:1 HER2/CEP17 ratio cells are excluded from the abnormal category. The distribution of HER2/CEP17 ratios is shown in Figure 1. 3:1 HER2/CEP17 ratio cells determined the classification of heterogeneous in 46% of diagnoses of heterogeneity (Table I). Many of these heterogeneous or amplified cases (35.0%) were characterized as heterogeneous due to a single 3:1 cell. 3:1 HER2/CEP17 ratio cells were a poor predictor of heterogeneity as determined by other types of abnormal ratio cells (Table II). Table II demonstrates the relationship between the probability of heterogeneity (excluding 3:1 cells) as a function of the number of 3:1 HER2/CEP17 cells present in the sample. The graph shows that 3:1 ratio of cells are associated with other abnormal cells (HER2/CEP17 > 2.2) cells. This relationship is statistically significant (OR= 2.2, p < 0.001). Figure 2 shows the relationship between 3:1 cells per case and the HER/CEP17 ratio exclusive of 3:1 cells. The absence of 3:1 HER2/CEP17 ratio cells is a fairly specific indicator of non-heterogeneity. If no 3:1 ratio of cells are present, the probability is approximately 90% that no abnormal cells (amplified ratio >2.20) will be found (Table II). The presence of 3:1 ratio of cells is not a sensitive predictor of other types of abnormal cells. There was a statistically significant relationship between the number of 3:1 ratio cells and other types of abnormal cells (OR = 2.2, p<0.001) in that the probability of finding additional abnormal cells increases with the number of 3:1 cells. This pattern was consistent across non-amplified, borderline and amplified cases (data not shown). In borderline cases, 3:1 cells are positively associated (p< 0.002) with heterogeneity but in amplified cases, 3:1 HER2/CEP17 ratio cells show no association with heterogeneity (p<0.62).
80 Frequency 100

Introduction
Tumor heterogeneity for HER2/neu amplification has been recognized as an issue for analysis by the College of American Pathologists (CAP). Lewis et al1 documented significant intratumoral heterogeneity for HER2/neu in both IHC and FISH preparations. Heterogeneous breast cancers may contain clones of amplified cells responsible for some immunohistochemically 2+ but FISH nonamplified cases and giving rise to recurrences and metastases with an amplified status, different biological behavior and a different response to Herceptin. In response to this possibility, the CAP developed criteria for defining genetic heterogeneity in HER2/neu testing of breast carcinomas.2 The CAP defined heterogeneous amplification of HER2/neu as the presence of between 5% and 50% of tumor cells with a HER2/CEP17 ratio > 2.20. Using this definition, the presence of a single cell with a HER/CEP17 ratio of > 2.20 (3:1 HER2/ CEP17), when counting 20 cells, leads to the diagnosis of heterogeneous amplification.4 Authors have questioned the validity of the upper percentage cut point for defining heterogeneity,3 while others have questioned the validity of the cell ratio (> 2.20) for assessment of a cell as amplified.4 The clinical implications of heterogeneity are currently unclear with authors reporting both a negative prognostic impact associated with genetic heterogeneity5 and lack of correlation of heterogeneous HER2 amplification with outcome.6 Technical aspects of FISH determination of gene amplification are complex and may artifactually impact the determination of HER2/neu amplifications status particularly in low ratio amplified cells and in heterogeneous tumors.7,8

55 31 14 7 12 7 6 4 7 5 7 5 1 5 8 11 18 30 49 1,546

Heterogeneous n= 474

Amplified by 3:1 cells n=15

20

40

60

134

Not Amplified

Amplified

-1

1 ln (HER2/CEP17 ratio)

Amplified by 50% cells > 2.2 n=161

Figure 1: Histogram showing frequency of cases by HER2/CEP17 ratio. The horizontal axis is presented on a log scale. The vertical bars correspond to cutoff points for borderline cases (ln(1.8) = 0.59) and amplified cases (ln(2.2) = 0.79). The vertical axis shows the number of cases corresponding to each HER2/CEP17 ratio.

2 2 1

Conclusions

8 Borderline

Table I: Each entry is the number of cases with the corresponding number of non-3:1 abnormal cells (i.e. cells with a HER2/CEP17 ratio > 2.2, excluding 3:1 HER2/CEP17 cells).

Materials and Methods

1,546 consecutive cases undergoing HER2/neu FISH were obtained from the records of ARUP Laboratories, Salt Lake City, Utah. Fluorescence in situ hybridization was performed using the PathVysion (Abbot Laboratories, Chicago, IL). For each of the cases, 20 cells of invasive carcinoma were analyzed for HER2/CEP17 ratio. Cases were assessed using current CAP recommendations for analysis of HER2/neu amplification in breast carcinomas.3 Cases were assessed as non-amplified (ratio < 1.8), borderline amplified (1.8 ratio 2.2) or amplified (ratio > 2.20). Heterogeneity was present when the percentage of cells with ratios above 2.20 was 5% but < 50%. Individual cells were typed by probe ratios and the distribution of cell types determined. The distribution of HER2/CEP17 ratios was determined with the number of 3:1 HER2/CEP17 cells plotted against the number of amplified cells. The probability of a heterogeneous population being present was plotted against the number of 3:1 HER2/CEP17 ratio cells in the sample. Scatter plots of heterogeneity versus amplification were developed as well as graphically investigating the relationship between the distribution of 3:1 HER2/CEP17 cells and amplification status.

0 1 2 3 4 5 6 7 8 Total

No 1019 187 46 17 3 5 1 0 1 1279

Yes 114 80 35 17 9 9 1 2 0 267

Total 1133 267 81 34 12 14 2 2 1 1546

Prob(H et) 10.1 30.0 43.2 50.0 75.0 64.3

95% CI Lowe r Upper 8.3 11.8 24.4 35.5 32.1 54.2 32.2 67.7 46.3 100 35.6 93.0

3:1 Cells

3:1 Cell Count

Heterogeneous (exclusive of 3:1 cells)

Not Amplified 4

Amplified

Our data indicate that cut points of heterogeneity are placed in a region that is sensitive to counting and technical error. A small error in counting may result in a diagnosis of heterogeneity. The 3:1 ratio of cells may be particularly sensitive to technical errors. Modifications in cut points as suggested by Schmidt et al5 may improve the reproducibility and predictive value of findings of genetic heterogeneity and amplification. Determination of genetic heterogeneity within a breast cancer can have several important clinical implications including differences in HER2/neu amplification between original primary, recurrences and metastases. Elimination of 3:1 HER2/CEP17 ratio of cells from the definition of heterogeneity appears to reduce cases designated as heterogeneous and improves the relationship between single abnormal cells and other measures of amplification. While the biologic/pathologic significance of 3:1 HER2/CEP17 ratio cells is as yet unknown, they influence many diagnoses of heterogeneity using the CAP guidelines. Appropriate guidelines for assessment of heterogeneity may be therapeutically and prognostically important.

-1

.79 1 ln (HER2/CEP17 ratio)

References
1. Lewis JT, Ketterling RP, Halling KC, Reynolds C, Jenkins RB, Visscher DW. Analysis of intratumoral heterogeneity and amplification status in breast carcinomas with equivocal (2+) HER-2 immunostaining. Am J Clin Pathol. 2005 Aug;124(2):273-281. 2. Vance GH, Barry TS, Bloom KJ, Fitzgibbons PL, Hicks DG, Jenkins RB, Persons DL, Tubbs RR, Hammond ME; College of American Pathologists. Genetic heterogeneity in HER2 testing in breast cancer: panel summary and guidelines. Arch Pathol Lab Med. 2009 Apr;133(4):611-612. 3. Hsu CY, Li AF, Yang CF, Ho DM. Proposal of modification for the definition of genetic heterogeneity in HER2 testing in breast cancer. Arch Pathol Lab Med. 2010 Feb;134(2):162; author reply 163. 4. Schmidt RA, Dintzis SM, Allison KH. High-ICR:a unifying data-driven reporting system for HER2 FISH analysis based on intra-tumor heterogeneity. Mod Pathol 2010; (62A):251. 5. Shafi H, Schmidt M, Bose S. Her2/neu genetic heterogeneity (GH) in breast cancer as defined by CAP guidelines correlates with negative prognostic factors. Mod Pathol 2010; (63A):252. 6. Bartlett AI, Starcyznski J, Robson T, Maclellan A, Campbell FM, van de Velde CJ, Hasenburg A, Markopoulos C, Seynaeve C, Rea D, Bartlett JM. Heterogeneous HER2 gene amplification: impact on patient outcome and a clinically relevant definition. Am J Clin Pathol. 2011 Aug;136(2):266-274. 7. Vanden Bempt I, Drijkoningen M, De Wolf-Peeters C. The complexity of genotypic alterations underlying HER2-positive breast cancer: an explanation for its clinical heterogeneity. Curr Opin Oncol. 2007 Nov;19(6):552-557. 8. Farabegoli F, Santini D, Ceccarelli C, Taffurelli M, Marrano D, Baldini N. Clone heterogeneity in diploid and aneuploid breast carcinomas as detected by FISH. Cytometry. 2001 Feb 15;46(1):50-56.

Table II : Heterogeneity is determined exclusive of 3:1 cells. Prob(Het) is the probability of heterogeneity without counting 3:1 cells. For example, there were 267 cases with one 3:1 cell. Eighty of these had a sufficient number of amplified cells (exclusive of 3:1 cells) to be classified as heterogeneous. Thus, if one 3:1 cell is present the probability of additional amplified cells being present is 30%.

Figure 2: The figure shows the relationship between the number of 3:1 cells and the HER2/CEP17 ratio per case. For this graph, the HER2/CEP17 ratio was calculated without including 3:1 HER2/CEP17 cells. The vertical bars indicate the cut points for borderline and amplified cases. The horizontal axis is presented on a log scale to facilitate presentation of data points at the low end of the scale.