[Back to Breast Cancer Index]

The goal of the HER-2 evaluation of a tumor is to predict Herceptin responsiveness or non-responsiveness ² as to Herceptin-based chemotherapy. "Positivity" predicts (1) responsiveness to Herceptin & indicates (2) likely resistance to endocrine therapies & (3) likely less benefit from non-anthracycline, non-taxane-containing chemo but better response to anthracycline therapy & paclitaxel⁸.

METHODS: Product expression is assessed by IHC. Gene amplification is asessed by in-situ gene or oncogene hybridization with either a non-permanent fluorescent tag (FISH) or permanenmt marker tags of chromogenic type (CISH) or silver deposition type (SISH). Or, there is the reverse transcriptase quantitative PCR (RTq-PCR) test. These are all done on properly formalin fixed & paraffin embedded tumor tissue (FFPE).

Bullet Points :

• IHC tumor tissue test: We perform this @ LML, & it is our first-line check of the invasive tumor...true negativity means there is nothing for Herceptin to go after.
  • 10% neutral buffered formalin-fixed tumor tissue (from cores, lumpectomy, or mastectomy) [Dr. Michael Lagios² has always used alcoholic formalin & says IHC works fine with it; Allen Gown²says FISH works with it, too...but alcohol damages nuclear markers such as ER & TTF-1 if nuclei not yet formalin fixed]. Adequate…8 hours or more…fixation is vital^{2, 5}! ...at least 6 and no more than 24 hours fixation⁸...though 2008 finds some authnorities defending 48 hours in order to account for fixation over a weekend.
  • similar results are (should be) seen whether primary tumor or a metastasis⁴.
  • the tissue block needs to contain normal breast & tumor so that the interpreting pathologist can “normalize⁷” (mentally subtract) the degree of staining in that patient’s benign breast tissue against the tumor staining (tumor IHC minus normal gland IHC = reported result) in order to grade the degree of over-expression & obtain IHC/FISH concordance⁷...this happens in some kits and must not be confused with overall too-heavy marking.
  • IHC is directed at extracellular domain of the membrane receptor protein coded by the HER-2/neu oncogene.
  • degrees of IHC staining ...only membrane staining counts (we use current CAP-ASCO criteria)… & FISH correlation is as follows:
  • 0 =...no staining at all or up to <10% of cells have partial membrane staining when normalized (correlates with normal FISH negative for gene amplification).
  • 1 + =...>10% of tumor cells with faint/barely perceptible membrane-pattern of partially circumferential staining. Probably FISH non-amplified. If you can only appreciate the membrane staining at 40x...most likely 1+ and not 2+. 
  • 2 + =... weak to moderate circumferentially complete membrane staining greater than the internal normal tissue in <30% of tumor cells. If you can see the staining at 10x, could be 2-3+. Most likely (1) diploid chromosomes with only 3-6 FISH-positive dots...normal is 2...which would give a ratio between 1.5 & 3...gene amplification being a ratio greater than 2; or (2) aneuploid chromosome 17 which turns out not to have a ratio above 2.
  • 3 +...>30%⁸) with moderate to strong..."intense"...complete membrane staining. If see distinct pattern at 4x, most likely 3+...slide should seem brownish to the naked eye. FISH amplified in 95% if true 3+. FISH positive if >6 copies per nucleus⁸).
  • concordance/discordance: IHC & FISH/CISH are trying to predict/assure a Herceptin response if Herceptin is used (Dr. Press² notes a $50,000+ cost of just the drug for a 52-week course; risk of Tx-induced cardiomyopathy increased over 400% from 1% up to 4-6%⁶) & normalized IHC 3+ responds and a 2+ with gene amplification (especially if high-level amp.) responds like a 3+. By current convention you expect true product over-expression to be associated with gene amplification.
    • archetype: really overexpressed IHC, gene-amplified by FISH/CISH, grade III, high Ki67, node positivity, and relative chemoresistance to some agents, and half of such cases are low-degree ER/PR positive².
    • really overexpressed IHC without remarkable Ki67 elevation is odd
      ( discordant ?).
    • 0 negative & 1+ are strongly predictive of negativity for gene amplification; 2+ is indeterminate; 3+ is strongly predictive of positivity for gene amplification and strongly predictive of Herceptin responsiveness.
    • 3+ IHC with FISH negative (or 3+ & ratio only 2.1 L07-4483) for gene amplification may be an example of “single copy over-expression” of product [LMC-00-6135; LMC-05-6300], an over-interpretation, or an insufficiently controlled technical problem such as too-intense antigen retrieval or wrong antibody used such as E-cadherin [L09-801].
    • 2+ IHC, especially if the lab does not “normalize” and mentally subtract out the degree of normal epithelial staining on the slide, may not be true over-expression; a true normalized 2+ likely corresponds to low amplification of gene copies (3-6) or to a non-amplified gene ratio in an aneuploid-increase of chromosome 17.
    • IHC not 2-3+ but gene is amplified according to FISH or CISH or SISH results (possibly as many as 1-3% of cases⁹):
    1. indicates absence of full translation of the genetic information.
    2. amplification is reported, but it is really pseudoamplification of the gene in aneuploid nuclei & was not detected because lab did not use a ratio with another gene on same chromosome such as CEP 17.
    3. amplified gene is producing a defective or aberrant oncoprotein which the IHC antibody does not adequately recognize.
    4. testing on a post-chemotherapy tumor can show non-over-expressed IHC while the gene status is actually amplified (chemo has damaged oncoprotein production?) [LMC-01-4957] .
    5. a true status of product over-expression relates to increased proliferation and is thought to happen rarely in nuclear grade I (IDCs, classic lobular, or tubular ca.)²...warrants careful review.
• FISH tumor tissue test: We have this done at PhenoPath or Clarient labs & tends to be requested in 2+ indeterminate cases.
• reasonable to FISH, CISH, or SISH test cases IHC <3 +, at least when tumor ER/PR neg., and high proliferation by such as Ki67².
• Is a surrogate line of evidence to help define whether membrane product is over-expressed…amplification implies product over-expression.
• results expressed as a ratio of average number of detected HER-2/neu “genes” [oncogenes] stained per tumor nucleus to the number of chromosome 17 genes stained per nuclei (this adjusts for variations in nuclear “ploidy”).
• RTq-PCR: this is a quantitative test done from FFPE tissue and capable of correlating concordantly in p7% of cases with ISH¹⁰.
• serum test: useful in patients who have an initial test value = to or > 15 ng/ml;serum sample is forwarded to a reference lab. Indications for serum test:
• if original tumor block is missing or recurrence is impractically located for biopsy
• to complement the results of the IHC test
• to monitor response to hormonal therapy (cases with ER positivity & low-volume, indolent metastatic disease)
• to monitor response to Herceptin/chemo combination (ER neg. & symptomatic multi-organ disease)

Discussion :

Human epidermal growth factor receptor 2 gene ERBB2 (commonly referred to as HER-2...sometimes as HER-2/neu) amplification and/or membrane product over-expression correlates with aggressive traits likely for that breast cancer², with HER-2 gene amplification in about 18-20% of all breast cancers. Produced in the cytoplasm, protein goes to cell surface membrane and normally aggregates to become or induce clusters of about a thousand receptors. If protein is neoplastically upregulated, those clusters of receptors enlarge and coalesce & lead to heavy IHC staining ⁸. The receptor is a 185-kd protein cell-surface membrane receptor (in the family of human epidermal growth factor receptors) with tyrosinase activity in the intra-cellular domain (and the molecule having a transmembrane domain and an extracellular domain). It is found normally in a number of epithelial cell types. Through binding of specific growth factors, the receptor regulates some aspects of cell growth and division. Increased receptors are thought to increase signalling for cell growth. This proto-oncogene is located on the long arm of chromosome 17.

Proto-oncogenes encode important proteins utilized normally in normal cell growth & differentiation, and they are genes with oncogenic potential.  When proto-oncogenes are altered by point mutation, translocation or gene amplification, they thereby become oncogenes and produce growth signals that may lead to cellular transformation and the development of cancer. Herceptin attaches to the HER-2/neu protein (neu) product (membrane receptors) and somehow, sometimes, inhibits or cancels the dysfunctional effect of excess product or amplified gene (oncogene).

The mutated, amplified, or altered form of a proto-oncogene is called an oncogene…an amplified HER-2 gene is an oncogenic HER-2 genetic tumor status.

Breast cancer is diagnosed  in about 180,000 American women each year. The risk of breast cancer increases with age. For more than a decade, circulating breast cancer antigens have been used in breast cancer for monitoring therapy and evaluating possible recurrence. It is conventionally considered that a breast cancer never changes its HER-2 status. Since about 2003, we have seen in our practice the use of the IHC tissue stain in a search for over-expressed  HER-2/neu oncogene-product protein as a prognostic indicator and a marker related to choice of therapy (15-30% of breast cancers over-express HER-2 product). This latter point has become particularly useful since the introduction of Herceptin® as a chemotherapy agent which targets the HER-2/neu receptor. And, in ER positive tumors,HER-2 positivity predicts responsiveness to anthracyclin based chemo and to tamoxifen⁷. With immunohistochemistry (IHC) techniques, it has been found that breast cancer patients who over-express HER-2/neu product in their cancers have a poor prognosis with shorter disease-free and overall survival than patients who do not over-express HER-2/neu.

Some suggest that elevations of shedded extracellular domain HER-2/neu in serum¹ may be used as evidence of early recurrence (the serum material is part of the membrane product but not sure yet of any correlations with intensity of IHC staining of tumor). In addition, initially elevated levels of serum HER-2/neu correlate with the presence of metastatic disease and suggest poor prognosis. These levels may be valuable in predicting response to various forms of therapy. Combining HER-2/neu serum measurements with other tumor markers may improve the sensitivity of detection of recurrence.

A reference in 2002 indicated that Herceptin was only effective in cases of IHC product over-expression; one in 2003 indicated effectiveness only if gene status is amplified. Dr. Gown says that Herceptin response has to do with product over-expression²… as does NEJMed⁴…over-expression is the target for treatment⁴.

Synonyms: neu, c-erbB2, HER-2

References:

1. from an online Bayer report of 4 serum test reports at 12/01 San Antonio breast conf., 31 Dec. 2001 [not online as of 7 July 2003].
2. verbal or e-mail info from experts (such as Dr. Allen Gown, Dr. Press).
3. Genentech website & then drop HER-2 into site search engine.
4. articles & editorials, NEJMed 353(16):1652-1738, 20 October 2005.
5. Neal Goldstein, et. al. AJCP 120:86-92, July 2003.
6. 12/2005 San Antonio Breast Conf.
7. Gown AM, et. al., "Extremely High Concordance Between Immunohistochemistry and FISH Testing for HER2/neu STATUS in Breast Cancer: a Study of 6104 Cases, presented at 12/2006 San Antonio Breast Conference [showing the value of "normalization" in obtaining IHC & FISH concordance.
8. CAP/ASCO Guidelines ("New Guidelines Recommended Determining HER2 Status for All Invasive Breast Cancer") released 11-18 Dec 2006 & in article in CAP Today Dec. 2006.
9. Dr. Jeff Ross, Albany Med. Ctr., Albany, NY @ Genetech sponsored lecture, Columbia, S. C. 5/13/2009.
10. Baehner FI, et. al., JCO 28:4300-6, 2010 & reported by Allen Gown in spring 2011 vol. 14(2) issue of Phenomena by PhenoPath Lab.

(posted 2004; latest addition 6 June 2011)