Medscape   Home   Site Map   Marketplace   My Medscape   CME Center   Feedback   Help Desk

American Society of Clinical Oncology 36th Annual Meeting
Day 4 - May 23, 2000

HER2 Testing and HER2-Based Therapy

Harold J. Burstein, MD, PhD

One of the most exciting areas in solid tumor oncology has been the emergence of more rational therapeutics based on biologic properties of tumors. Nowhere has this been more fully exploited to date than in breast cancer. In the span of less than 2 years, the human epidermal growth-factor receptor 2 (HER2) has moved from being a laboratory-based prognostic factor to a widely studied predictive marker, as well as the target for a selective, active therapy, the monoclonal antibody trastuzumab, which binds to the HER2 protein.

It is now reasonable to test all new cases of breast cancer -- both early and late stage -- for HER2 overexpression. However, the rapid growth of clinical interest in HER2 testing and in the appropriate use of trastuzumab has created new clinical problems: what is the optimal testing strategy, and who really benefits from trastuzumab treatment? A variety of presentations at ASCO 2000 tackled aspects of HER2 and its implications for treatment in breast cancer.

Strengths and Weaknesses of HER2 Testing

HER2 overexpression occurs in roughly 25% to 35% of all cases of breast cancer. Overexpression most often arises from genetic amplification of the HER2 gene, resulting in increased gene number, DNA content, mRNA expression, and surface protein expression. There are many possible ways to test for HER2 overexpression in tumor cells (Table 1).

Table 1. Strategies for HER2 Testing

Method Tests Ease of Testing Laboratory Readout Laboratory Comment
Southern blot DNA content Not standardized for laboratory evaluation    
Northern blot Northern blot RNA content Not standardized for laboratory evaluation nbsp; nbsp;
Immunohistochemistry Cell surface protein expression Easily reproduced in laboratory; inexpensive Usually scored 0 (negative) to +3 (strongly positive); each point roughly corresponds to a 5-fold increase in cell surface expression Variety of antibody tests available; can be done on fixed, preserved tissue
Fluorescence in situ hybridization (FISH) Gene copy number per cell Easily reproduced in laboratory; expensive Usually scored positive (amplified gene) or negative Two commercial tests available; not widely practiced; can be done on fixed, preserved tissue
Serum HER2 Soluble, circulating extracellular domain of HER2 protein Not routinely available; correlations with cellular HER2 expression not well characterized Standards in evolution  
Immunohistochemistry is widely practiced by pathologists around the world and is fast, cheap, and widely available. However, it requires well-preserved tissue, and may be subject to variable interpretations. By contrast, testing for nucleic acid quantity by Southern or Northern blot analysis is not routinely done in clinical laboratories, and is therefore more labor intensive and less widely accessible. For the clinically used methods -- immunohistochemistry and FISH -- a variety of particular tests are available. Each test is likely to have its own operator characteristics and unique sensitivity/specificity issues.

In general, there is good agreement between the testing strategies. However, there are cases of immunohistochemical "positive" tests with no evidence of gene amplification (FISH negative; so-called "single copy overexpressors"). There are conversely cases of genetic amplification without increased surface expression. It is not known what accounts for the discordance in this latter instance -- whether these are truly false-negative tests or whether there are other biological processes that limit surface expression of HER2 despite genetic amplification.

What Does a Positive Test Mean for Therapy?

With the development of anti-HER2 therapy (trastuzumab), the crucial question for the clinical laboratory has become: which test most reliably predicts response (or lack of response) to trastuzumab? Such information is routinely needed in deciding whether to offer trastuzumab therapy. Preliminary clues are becoming available. In a trial of weekly paclitaxel and trastuzumab, patients whose tumors were either HER2-positive or HER2-negative were treated with the combination regimen. The response rate in the HER2-negative subgroup of patients was considered equivalent to the historic response rate seen with weekly paclitaxel by itself. This result has been widely interpreted to mean that patients whose tumors are HER2-negative by immunohistochemistry do not benefit from trastuzumab. A second important clue emerged from the trial of single agent trastuzumab as first-line therapy for advanced breast cancer. In that study, the response rate among women whose tumors were HER2 +3 positive was substantially higher (31% response rate) than was seen in women whose tumors were HER2 +2 positive (0% to 5% response rate).

A number of important presentations at ASCO serve to clarify how HER2 tests relate to clinical outcomes for women with advanced breast cancer. Dr. Andrew Seidman,[1] from Memorial Sloan-Kettering in New York, NY, reported on the final outcomes from the clinical trial examining weekly paclitaxel (90 mg/m2) with trastuzumab (4mg/kg loading dose, 2 mg/kg weekly thereafter). The response rate to the combination regimen was dramatically different in subsets of patients whose HER2 overexpression was tested using different antibodies in immunohistochemical assays, or by FISH analysis. As seen in Table 2, response rates vary by up to 20%, depending on how "positivity" was defined.

Table 2. Response Rates and Methods of HER2 Testing

Antibody/Method for HER2 Testing % of Patients on Study With Positive HER2 Overexpression % Response Rate to Trastuzumab and Paclitaxel in HER2-Positive Patients
Dako HercepTest (IHC) 55 69
P-AB 1 (IHC) 66 67
TAB250 (IHC) 38 83
CB11 (IHC) 43 76
FISH 46 75
In general, this study suggests that the less sensitive the HER2 test is (and hence, the lower the number of positive cases), the greater the likelihood of benefit from trastuzumab-based therapy.

Go FISH?

There is considerable interest in correlating FISH results and immunohistochemistry results. Dr. Seidman reported the following correlations (Table 3):

Table 3. Concordance Between HercepTest and FISH Analysis: Memorial Sloan-Kettering/M.D. Anderson

IHC Level (HercepTest) No. of Cases % FISH Positive
0-1 37 7%
2 18 39%
3 33 75%
Similar concordance patterns were reported by Genentech scientists, updating data from the early trastuzumab trials (Table 4).[2] In this study, the immunohistochemical assay was not the HercepTest but the clinical trials assay, another standardized measure of HER2 expression.

Table 4. Concordance Between IHC and FISH: Genentech

IHC Level No. of Cases % FISH Positive
0 214 3
+1 30 7
+2 88 24
+3 197 89
Finally, a German pathology group has done a single-center concordance analysis, which again showed that at either end of the spectrum, there is reasonable concordance between IHC and FISH (see Table 5).[3] They evaluated 142 cases of breast cancer and analyzed IHC and FISH results.

Table 5. Concordance Between IHC and FISH: German Pathology Group

IHC Level No. of Cases No. FISH Positive % FISH Positive
0 or +1 44 0 0
+2 72 22 (10 moderate, 12 high) 31
+3 26 26 100
In general, IHC has decent concordance with FISH. There are small percentages of patients who are negative by IHC who appear to have FISH overexpression, on account of either false-negative tests or other intrinsic biologic factors that limit surface expression. Conversely, most but not all +3 overexpressing patients (75% to 100%) have genetic amplification. In all 3 series, a substantial fraction (61% to 76%) of tumors that are +2 by IHC lack FISH overexpression.

Who Benefits From Trastuzumab?

The important question is, Do these IHC-positive patients get clinical benefit from trastuzumab, or is benefit restricted to those patients (the minority of +2 positive patients) who have genetic amplification (ie, are FISH positive). Data presented at ASCO now clarify that question to encompass use of trastuzumab as a single agent and in combination with chemotherapy.

Dr. Charles Vogel,[3,4] from the Aventura Comprehensive Cancer Center, Aventura, FL, presented the update of the study analyzing single-agent use of trastuzumab in women with metastatic breast cancer who had not previously received chemotherapy for advanced disease. It is now clear that all overt clinical responses in this study of more than 100 women occurred in the 76% of patients with HER2 +3 positive tumors; there were no responses to single-agent trastuzumab in the 24% of patients who were HER2 +2 positive. The analysis of response rates based on FISH phenotyping, done on a subset of patients for whom adequate tissue was available, is presented in Table 6.

Table 6. Utility of FISH at Predicting Response to First-Line Single-Agent Trastuzumab Therapy

  FISH + FISH -
No. of Patients 41 21
Response Rate 41% 5%
A similar analysis was done for response to single-agent trastuzumab offered as second- or third-line treatment for advanced breast cancer.[3,4] As seen in the first-line trial, responses were observed only among patients whose tumors overexpress HER2 as measured by FISH (Table 7).

Table 7. Response Rates to Single-Agent Trastuzumab Therapy as Second- or Third-Line Therapy

IHC Level FISH + FISH -
+2 20% 0%
+3 20% 0%
Thus, the "lion's share" of response to single-agent trastuzumab is found in that subset of women with either +3 overexpression by immunohistochemistry or gene amplification as measured by FISH. For women whose tumors lack either of these features, the response rate is clearly less than 5% and may be nonexistent.

Dr. Seidman[1] provided response rate information in the trial of trastuzumab with weekly paclitaxel. Among patients who were FISH negative, the response rate to trastuzumab and paclitaxel was 48%, consistent with historic controls for weekly paclitaxel alone. By contrast, the response rate among FISH-positive patients was 75%.

Dr. Robert Mass,[4] from Genentech, Inc., San Francisco, California, reported on the utility of FISH testing in the large randomized study that compared trastuzumab and chemotherapy (either paclitaxel or doxorubicin/cyclophosphamide) with chemotherapy alone. The overall results of that study, presented last year at ASCO, showed that adding trastuzumab to chemotherapy led to significant improvement in response rates, time to progression, and overall survival for women with HER2-positive metastatic breast cancer. The response rates have now been analyzed as a function of FISH expression and are shown in Table 8.

Table 8. Response Rates to Chemotherapy +/- Trastuzumab as a Function of FISH Analysis

Treatment FISH + FISH -
Chemotherapy 27% 39%
Chemotherapy & trastuzumab 54% 41%
There are 3 dramatic conclusions to be drawn from this analysis. First, chemotherapy works differently in metastatic breast cancer patients defined by HER2 status. Higher response rates to chemotherapy are observed in women whose tumors are HER2 negative. Because of this finding, the vast literature on chemotherapy response rates for advanced breast cancer is more or less obsolete, as hardly any studies have reported on outcomes as a function of HER2 expression. From now on, characterizing HER2 status is essential for understanding the true activity of any treatment trial.

Second, adding trastuzumab to chemotherapy has essentially no effect for women whose tumors lack HER2 gene amplification as measured by FISH. The response rates were, in essence, identical in the FISH-negative subset.

Finally, adding trastuzumab dramatically affects outcome for HER2-positive patients. In that group of women, on average, adding trastuzumab to chemotherapy doubles the response rate. This study strongly suggests that advanced breast cancer needs to be thought of as 2 diseases: an HER2-negative variant, for whom standard chemotherapy is more effective and trastuzumab is of little benefit, and an HER2-positive variant, for whom standard chemotherapy is less effective, but in whom adding trastuzumab may overcome the relative resistance to chemotherapeutic agents.

Other researchers at Genentech have been analyzing serum HER2 as a marker for predicting response to trastuzumab-based therapy.[5] Serum HER2 can be measured by ELISA. It was found in the pivotal Genentech trials that 33% of HER2 +2 patients had serum HER2, and that 74% of HER2 +3 patients had serum HER2. In the randomized trial comparing chemotherapy alone to chemotherapy + trastuzumab, the patients who responded to therapy were as likely to have elevated serum HER2 as those who did not respond to treatment. Thus, for the moment, it does not appear that serum HER2 is a good predictor of response to trastuzumab-based therapy, and that tissue-based methods are more reliable.

Putting HER2 Testing Together: Which Tumors to Test With What?

To date, trastuzumab has been shown to be effective only in women with advanced breast cancer whose tumors had greater surface expression of HER2 as measured by IHC. IHC has the virtue of being readily available at most centers and being relatively inexpensive in comparison to FISH. Retrospective analysis suggests that among these women, real clinical activity in response to single-agent trastuzumab or additional benefit from trastuzumab with chemotherapy was restricted to those women with genetic amplification of the HER2 gene in their tumors. There is reasonably good concordance between IHC +3 overexpression and FISH positivity. In the metastatic setting, using either of these measures still seems a reasonable way to select candidates for trastuzumab therapy. In cases where IHC testing is equivocal or intermediate, FISH testing will further define an appropriate patient population and should be offered for IHC +2 patients. IHC +2 cases that are FISH negative should not receive trastuzumab.

A small percentage of women who are IHC 0 to +1 will have genetic amplification of HER2. However, because most cases of breast cancer are HER2 IHC negative, the actual number of these IHC neg/FISH pos cases is considerable. It is not known whether trastuzumab has any clinical activity in such women, and clinical trials are urgently needed to determine whether gene amplification (ie, FISH positive) is adequate for creating sensitivity to trastuzumab, even in the absence of IHC-detected HER2. Until the results of such a trial are available, trastuzumab probably should not be routinely offered to such patients, and most cases of breast cancer should be initially evaluated with IHC before consideration of FISH testing. However, it is also apparent that not all IHC testing is equivalent. Refinements in selection of antibodies for IHC testing and actual IHC methods will enhance the predictive value of such testing. For patients in whom other histologic factors or the clinical course of the disease suggest that IHC results might be falsely negative, evaluation with FISH is also a reasonable alternative.

Adjuvant trials of trastuzumab are currently in development. Given these recent data, it seems reasonable to restrict accrual in these trials to the patient population most likely to get potential benefit from trastuzumab: those women with unequivocal IHC results or FISH-positive expression. If trastuzumab is shown to have a role in the adjuvant setting, it will subsequently become important to determine which method of tumor testing should be offered up-front for all women with breast cancer diagnoses.

It is worth noting that patient selection for trastuzumab trials based on different diagnostic methods may profoundly affect the clinical results. Studies of patients carefully selected by IHC and FISH overexpression are likely to have higher response rates than studies including patients selected by less stringent methods. The field of HER2 testing is likely to continue to evolve rapidly. In the near term, IHC still has a role, but as there is a rapid growth in experience with FISH testing, this strategy may become more familiar, cheaper, more available, and more commonly used in the clinic.

A New Trastuzumab and Chemotherapy Regimen

So far, there are reports of trastuzumab used in combination with doxorubicin/cyclophosphamide, cisplatin, and paclitaxel, but a variety of other combinations are currently being evaluated. Use of trastuzumab with anthracyclines has been limited on account of the unacceptable incidence of cardiac toxicity when the drugs are used together. There is a need for defining new and effective combination regimens for trastuzumab and chemotherapy.

Investigators from Dana-Farber Cancer Institute in Boston, Massachusetts, reported on a phase II study of trastuzumab with vinorelbine.[6] This 40-person study included women with HER2-positive tumors (either +2 or +3) who could have received 0, 1, or 2 prior chemotherapy regimens for advanced breast cancer. Women were treated with trastuzumab on the standard 4 mg/kg loading dose, with 2 mg/kg weekly doses thereafter, and with vinorelbine at 25 mg/mg2/week. Vinorelbine dose was adjusted for weekly neutrophil counts. On 80% of the treatment days, patients received both trastuzumab and vinorelbine.

The regimen proved to be very well tolerated. Neutropenia was the only side effect, with treatment-related grade 3 or 4 toxicity, and no G-CSF was used in the study. There were minimal GI symptoms, and women reported minimal alopecia. There were no cases of symptomatic heart failure; 8% of patients had drops in their ejection fraction. All of these patients had received prior doxorubicin and had borderline ejection fractions at baseline. The overall response rate was 75%. Responses were seen despite extensive prior therapy (Table 9).

Table 9. Response Rates to Trastuzumab & Vinorelbine

  No. of Patients Response Rate
No. of Prior Regimens for Metastatic Breast Cancer
   0 19 84
   1 14 64
   2 7 71
Types of Prior Regimens (Adj or Metastatic)
   None 7 71
   Nonanthracyline, nontaxane 4 100
   Anthracycline-based 8 88
   Taxane-based 6 50
   Both anthracycline- and taxane-based 15 73
HER2 Level of Expression (IHC)
   +3 30 80
   +2 9 55
   Other 1 100
As has been seen in other trastuzumab-based trials, response rates were highest among patients with +3 HER2 expression.

The encouraging results of this study and the well-tolerated treatment regimen suggest that trastuzumab and vinorelbine merit further evaluation as a regimen for widespread use among women with HER2-positive breast cancer.

HER2 Testing: Beyond Breast Cancer

Most clinical studies of HER2 and trastuzumab have focused on breast cancer patients. There are data suggesting that HER2 overexpression occurs in other tumors as well, including ovarian, lung, and other malignancies, but the clinical significance of such findings is not clear. The true incidence of HER2 overexpression in other malignancies is not known.

One study[7] reported on the incidence of HER2 overexpression among patients with androgen-independent prostate cancer. The investigators used the more specific TAB250 antibody to detect HER2 expression and analyzed 22 cases. HER2 positivity was seen in 27% of cases overall, though in only 3 cases was expression considered either +2 or +3 positive. Based on such studies, investigators at Memorial Sloan-Kettering[8] have initiated a phase II trial of trastuzumab alone or with paclitaxel among patients with prostate cancer. Ninety-seven patients have been evaluated for HER2 expression; the incidence of HER2-positive tumors is 12%. To date, there have been no responses to single-agent trastuzumab therapy in men with prostate cancer.

Lung cancers are also being tested for HER2 overexpression. At the University of Colorado, researchers are characterizing the expression of HER2 in non-small-cell lung tumors.[9] Adenocarcinomas and large-cell lung cancers appear to have frequent surface expression of HER2 (65% with the HercepTest), and in 2 of the 36 analyzed were found to have modest overamplification of HER2 genes on FISH analysis.

References

  1. Seidman AD, Fornier M, Esteva F, et al. Final report: weekly (W) Herceptin (H) and taxol (T) for metastatic breast cancer (MBC): analysis of efficacy by HER2 immunophenotype (immunohistochemistry [IHC]) and gene amplification [fluorescent in situ hybridization]. Program and abstracts of the American Society of Clinical Oncology 36th Annual Meeting; May 20-23, 2000; New Orleans, Louisiana. Abstract 319.
  2. Buehler H, Bangemann N, Evers K, et al. Effective HER-2/neu diagnosis in breast cancer by a combination of immunohistochemistry and FISH. Program and abstracts of the American Society of Clinical Oncology 36th Annual Meeting; May 20-23, 2000; New Orleans, Louisiana. Abstract 294.
  3. Vogel C, Cobleigh M, Tripathy D, et al. First-line non-hormonal treatment of women with HER2 positive metastatic breast cancer with Herceptin (trastuzumab, humanized anti-HER2 antibody). Program and abstracts of the American Society of Clinical Oncology 36th Annual Meeting; May 20-23, 2000; New Orleans, Louisiana. Abstract 275.
  4. Mass RD, Sanders C, Charlene K, et al. The concordance between the clinical trials assay (CTA) and fluorescence in situ hybridization (FISH) in the Herceptin pivotal trials. Program and abstracts of the American Society of Clinical Oncology 36th Annual Meeting; May 20-23, 2000; New Orleans, Louisiana. Abstract 291.
  5. Wong WL, Bajamonde A, Nelson B, et al. Baseline serum HER2 (sHER2) levels in the pivotal Herceptin breast cancer trials: a comparison of 2 ELISA methods. Program and abstracts of the American Society of Clinical Oncology 36th Annual Meeting; May 20-23, 2000; New Orleans, Louisiana. Abstract 297.
  6. Burstein HJ, Kuter I, Richardson P, et al. Herceptin and vinorelbine for HER2-positive metastatic breast cancer: a phase II study. Program and abstracts of the American Society of Clinical Oncology 36th Annual Meeting; May 20-23, 2000; New Orleans, Louisiana. Abstract 392.
  7. Reese DM, Small EJ, Waldman FM, et al. HER2 expression in androgen-independent prostate cancer. Program and abstracts of the American Society of Clinical Oncology 36th Annual Meeting; May 20-23, 2000; New Orleans, Louisiana. Abstract 1365.
  8. Morris MJ, Reuter VE, Kelly WK, et al. A phase II study of Herceptin alone and with taxol for the treatment of prostate cancer. Program and abstracts of the American Society of Clinical Oncology 36th Annual Meeting; May 20-23, 2000; New Orleans, Louisiana. Abstract 1298.
  9. Hirsch F, Veve R, Varella-Garcia M, et al. Evaluation of HER2/neu expression in lung tumors by immunohistochemistry and fluorescence in situ hybridization (FISH). Program and abstracts of the American Society of Clinical Oncology 36th Annual Meeting; May 20-23, 2000; New Orleans, Louisiana. Abstract 1900.

Return To Day 4 Summaries

  Home   Site Map   Marketplace   My Medscape   CME Center   Feedback   Help Desk

Medscape Search Options
Select a database to search, enter a search term, then click “go.”    Advanced Search Forms

All material on this website is protected by copyright. Copyright © 1994-2000 by Medscape Inc. All rights reserved. This website also contains material copyrighted by 3rd parties. Medscape requires 3.x browsers or better from Netscape or Microsoft.