| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Review |
1 Oncogene Science, Bayer HealthCare, Cambridge, MA 02142.
2 Tufts New England Medical School, Department of Pathology, Boston, MA 02111.
3 Medizinische Fakultaet der Universitaet zu Koeln, D-50931 Koeln, Germany, and Bayer Vital GmbH, D-51368 Leverkusen, Germany.
4 Section of Hematology-Oncology, Pennsylvania State University/Hershey Medical Center, Hershey, PA 17033.
5 Department of Medicine, Veterans Administration Medical Center, Lebanon, PA 17042.
6 Bayer HealthCare, Diagnostics Division, Stoke Court, Stoke Poges, Slough, Berkshire SL2 4LY, United Kingdom.
7 Clinical Biochemistry, University of Oxford, Oxford OX1 2JD, United Kingdom.
aAddress correspondence to this author at: Oncogene Science, Bayer HealthCare, 80 Rogers St., Cambridge, MA 02142. Fax 617-492-8438; e-mail walter.carney.b{at}oncogene.com.
 |
Abstract |
|---|
 Top Abstract IntroductionMethods of Determining her-2/neu...Methodology of Literature ReviewResultsDiscussionReferences |
|---|
Methods: We performed a systematic review of the peer-reviewed literature on circulating ECD with respect to prevalence, prognosis, prediction of response to therapy, and monitoring of breast cancer.
Results: The prevalence of increased ECD in patients with primary breast cancer varied between 0% and 38% (mean, 18.5%), whereas in metastatic disease the range was from 23% to 80% (mean, 43%). Some women with HER-2/neu-negative tumors by tissue testing develop increased ECD concentrations in metastatic disease. Increased ECD has been correlated with indicators of poor prognosis, e.g., overall survival and disease-free survival. Increased ECD predicts a poor response to hormone therapy and some chemotherapy regimens but can predict improved response to combinations of Herceptin and chemotherapy. Many studies support the value of monitoring ECD during breast cancer progression because serial increases precede the appearance of metastases and longitudinal ECD changes parallel the clinical course of disease.
Conclusions: The monitoring of circulating HER-2/neu ECD provides a tool for assessing prognosis, for predicting response to therapy, and for earlier detection of disease progression and timely intervention with appropriate therapy.
|
Introduction |
|---|
|
TopAbstractIntroductionMethods of Determining her-2/neu...Methodology of Literature ReviewResultsDiscussionReferences |
|---|
The HER-2/neu oncogene has been localized to chromosome 17q and encodes a transmembrane tyrosine kinase growth receptor that is expressed on cells of epithelial origin. The full-length glycoprotein has a molecular mass of 185 000 Daltons (p185) and is composed of the internal tyrosine kinase domain, a short transmembrane portion, and an extracellular domain (ECD) that is similar to the three other members of the HER family (2)(5). The ECD portion of the receptor protein is heavily glycosylated, has a molecular mass in the 97- to 115-kDa range, and has been shown to be shed into culture fluids of SKBR-3 cells (6) as well as plasma(7) and serum (8)(9) from healthy individuals and patients with breast cancer (7)(8)(9).
The mechanism of activation of the HER-2/neu pathway is not completely understood, but studies have shown that the ECDs of the HER family of receptor tyrosine kinases form homodimers and heterodimers and that receptor dimerization activates a cascade of events in the HER-2/neu signaling pathway (4). A ligand that binds to the HER-2/neu receptor has not been identified, but a family of peptide ligands named neu differentiation factors, or heregulins, has been identified. Heregulins are 45-kDa growth factors, with homology to EGF, that bind directly to HER-3 and HER-4 receptors and cause the formation of heterodimeric receptor complexes. The dimers subsequently induce transphosphorylation and activate the HER-2/neu receptor (Fig. 1
).
|
In the study by Aguilar and Salmon (10), the biological activity of the plasma membrane-anchored heregulin was evaluated in human breast cancer cells. It was reported that transmembrane heregulin binds to cells expressing HER-3 and induces HER-2/neu phosphorylation and increases DNA synthesis in cells overexpressing HER-2/neu. The biological response to heregulin appears to depend directly on the degree of HER-2/neu expression in breast cancer cells; therefore, the regulatory pathway for control of proliferation may depend on the concentrations of both the heregulins and HER-2/neu ECD. Perhaps measuring growth factors such as heregulins and EGF will have a separate clinical utility or clinical utility in conjunction with measuring HER-2/neu ECD. In addition, heterodimers can also form between the EGFR and HER-2/neu on binding of EGF or transforming growth factor-
to the EGFR (11). Dimerization of the various HER receptors leads to tyrosine phosphorylation and activation of the HER-2/neu kinase with subsequent activation of downstream transduction pathways and signaling through ras, c-Src, phosphatidylinositol 3-kinase, and phospholipase C
pathways. Formation of heterodimers increases the affinity of the partnering receptor for its ligand and leads to potentiation of the mitogenic signal (12).
In a report by Codony-Servat et al. (13), it was suggested that cleavage of the HER-2/neu ECD involves matrix metalloproteinase (MMP) activity and that the process of ECD cleavage was inhibited by the MMP inhibitor TIMP-1 but not by TIMP-2. The MMPs have been strongly implicated in multiple stages of cancer progression, including invasion and metastasis (14), but it is not yet known whether the HER-2/neu ECD functions in the process of invasion and metastasis. Codony-Servat et al. (13) showed that HER-2/neu ECD shedding was inhibited by broad-spectrum MMP inhibitors such as EDTA, TAPI-2, and Batimastat and confirmed the data from Christianson et al. (15) showing that HER-2/neu ECD cleavage leads to the release of a truncated phosphorylated p95 fragment. Molina et al. (16) showed that the p95 fragment could be detected in only 14 of the 24 of the human breast tumors analyzed and that p95 band expression was highly variable. Studies continue to determine both the biological role and the potential clinical value of the phosphorylated membrane fragment.
In their report, Molina et al. (16) also showed that HER-2/neu shedding was activated by 4-amino-phenylmercuric acetate (a well-known MMP activator) in HER-2/neu-overexpressing breast cancer cells and that it could be blocked by the MMP inhibitor, Batimastat. The increase in ECD shedding also enhanced the production of the p95 fragment. In the same report (16), they demonstrated that Trastuzumab® [commonly referred to as Herceptin and manufactured by Genentech (San Francisco, CA)], an anti-HER-2/neu therapy, had a direct inhibitory effect on the basal and activated processes involved in HER-2/neu cleavage from HER-2/neu-overexpressing breast cancer cells. The HER-2/neu ECD shedding that was activated by 4-amino-phenylmercuric acetate could be blocked with Herceptin, however, leading to a reduction in the release of the p95 fragment (16).
Trastuzumab is a humanized monoclonal antibody (Mab) developed to target the HER-2/neu receptor that is overexpressed in 25–30% of breast cancers cells. Herceptin binds with high affinity to the ECD of HER-2/neu and inhibits proliferation of tumor cells that overexpress the HER-2/neu oncoprotein. The results of a large multicenter phase 3 clinical trial demonstrated that Herceptin, when added to conventional chemotherapy, can provide benefit to patients with metastatic breast cancers (MBCs) that overexpress HER-2/neu. Compared with the best available standard chemotherapy, concurrent treatment with Herceptin and first-line chemotherapy was associated with significantly longer times to disease progression, a higher rate of response, longer duration of response, and improved overall survival (OS) (17).
Because Herceptin treatment [reviewed in Ref. (18)] has been in practice for only a few years, the mechanism of action by which Herceptin either alone or in combination with chemotherapy enhances survival of MBC patients still needs to be understood. According to a recent report (19), it is possible that Herceptin binds the ECD, causing internalization and degradation, and subsequently inhibits signal transduction pathways. Another possible mechanism by which Herceptin is effective is the induction of cytolytic activity through cytotoxic lymphocytes (19). However, many additional studies are necessary to understand the mechanism(s) by which Herceptin provides benefit to women with MBC.
In this report, we will review HER-2/neu circulating ECD concentrations in relation to prevalence, prognosis, prediction of response to therapy, monitoring in MBC, and monitoring for early detection of recurrence.
|
Methods of Determining her-2/neu Status |
|---|
|
TopAbstractIntroductionMethods of Determining her-2/neu...Methodology of Literature ReviewResultsDiscussionReferences |
|---|
Since the earliest studies in 1987 (22) and 1989 (23), numerous tumor studies using formalin-fixed, paraffin-embedded tumor tissue have shown that HER-2/neu DNA amplification and protein overexpression occur in 
25–30% of women with primary breast tumors (20)(21). However, a few reports have described a wider range (15–40%) of HER-2/neu tissue expression, which appears to be a function of the composition of the sample being evaluated by the pathologist. For example, Latta et al. (24) reported that IHC staining was seen in 34% of women with pure ductal carcinoma in situ, 17% of women with invasive carcinomas with associated ductal carcinoma in situ, and 12.5% of pure invasive carcinomas. Sixty percent of cases of high-grade ductal carcinoma in situ showed increased IHC overexpression vs the 27% of high-grade invasive carcinomas with associated ductal carcinoma in situ and 22% of high-grade pure invasive ductal carcinomas. In part the wide range of IHC staining of primary breast tumors could be attributable to a variety of factors. It could be the heterogeneity of the tumor specimen within the tissue section and/or variations in methodologies, antibodies, or operators. It could also be attributable to the time of diagnosis or the existence of the early detection programs at the various institutions.
The current availability of commercial IHC reagent sets for HER-2/neu detection has now led to greater standardization relative to past testing, when a variety of methodologies and antibodies were used. For example, the US Food and Drug Administration (FDA) granted clearance for the use of two commercially available IHC assays to identify women who are candidates for Herceptin-based therapies. These include Herceptest, a rabbit polyclonal (DAKOCytomation, Copenhagen, Denmark), and Pathway (Ventana, Medical Systems, Tucson, AZ), which uses the Mab CB11. The first step in the IHC scoring system used to establish HER-2/neu status is to determine whether >10% of the breast tumor cells are stained with a HER-2/neu-specific antibody. The second step is to ascertain whether the membrane staining is partial or complete, and the third step is to determine the intensity of staining (3+, 2+, or 1+).
The second method used to assess the HER-2/neu status of a primary breast cancer (PBC) is FISH analysis. The advantages of this technique include the stability of chromosomal DNA and the ability to determine gene amplification or the number of HER-2/neu copies within tumor cells. The FISH technique is very reliable, providing 95.5% sensitivity and 100% specificity for detection of the HER-2/neu gene amplification. Currently there are two commercially available FISH reagent sets cleared for use by the FDA. PathVysion (Vysis, Downers Grove, IL) detects the HER-2/neu gene with a directly labeled probe. The primary advantage of this test is that it includes an internal control probe for the assessment of polysomy. The control probe simultaneously detects the centromere of chromosome 17, the locus of the HER-2/neu gene. PathVysion was approved originally to select node-positive, stage II patients for anthracycline treatment and was recently expanded to include selection for Herceptin therapy. The second FISH test cleared by the FDA is called the Inform test (Ventana Medical Systems, Tucson, AZ) and uses a biotin-labeled oligonucleotide probe that is detected by an avidin-biotin FITC system for signal amplification (21). This test was cleared for use as a prognostic factor in node-negative breast cancer patients.
Overall, tissue testing with either IHC or FISH is performed on the primary tumor tissue, and these are the only methods cleared by the FDA to determine eligibility of women with MBC for Herceptin therapy (21). At present, the IHC and FISH methods are used in combination to offer enhanced reliability for determining the HER-2/neu status and eligibility of a woman for Herceptin-based therapies. Herceptin-based therapy is a term that is used to refer to any chemotherapy that is used in combination with Herceptin. Currently, a woman is considered to have a positive HER-2/neu tumor if she has >10% of the breast tumor cells staining with a 3+ intensity. In contrast, women with <10% positive cells are considered not to have a HER-2/neu-positive tumor and according to current guidelines would not be considered eligible for Herceptin therapy.
Although tissue testing is very important in determining the HER-2/neu status of the primary tumor, there are limitations to tissue testing. The most important is that they are one-time tests used to determine the HER-2/neu status of the PBC, which in turn determines the eligibility for Herceptin. Herceptin is given primarily in the metastatic setting, but the HER-2/neu status is determined from the original breast tumor, which may have been removed many years earlier.
The tissue tests for protein overexpression by IHC and gene amplification by FISH are both subject to technical problems. For IHC these include, but are not limited to, differences in the methodology between laboratories and between operators, variability in operator interpretation, and variation in reagents. Although the FISH technology appears to be reproducible, there are several limitations to FISH testing. The FISH instrumentation is expensive and not widely available in diagnostic pathology laboratories. In addition, the standardization of fixative solutions, fixation times, and digestion conditions continue to be technical issues, suggesting that problems still exist with FISH as well as IHC.
More recently, it has been proposed that chromogenic in situ hybridization (CISH) is an alternative to FISH. In contrast to FISH and fluorescence, CISH technology is based on a peroxidase reaction that can be visualized by light microscopy and eliminates the need for fluorescence equipment. CISH also provides a permanent record and is less expensive than FISH. Although CISH appears to be more practical and accurate, it had not been cleared by the FDA for use in the USA. Additional studies are also required to determine the exact concordance between FISH and CISH before CISH can be accepted as a method to select patients for Herceptin therapy (25)(26).
A review of the recent literature comparing FISH and IHC methods on the same tumor specimens showed a considerable degree of concordance between the methods. Many reports in the past few years have shown a >90% concordance between the methods, which approached 100% when comparing 3+ IHC readings with FISH (12)(21). However, cases with a 2+ score are not very reproducible and therefore are no longer used alone to determine eligibility of a woman for Herceptin. Patients with a 2+ IHC score must have a confirmatory FISH test to be eligible for Herceptin therapy.
In a retrospective analysis performed to explore the correlation between gene amplification by FISH and clinical outcomes of women treated with Herceptin-based therapies, it was concluded that FISH was the superior method for selecting patients likely to benefit from Herceptin therapy (27). Additional studies are warranted to substantiate this observation.
Some studies have also used tissue testing to compare the HER-2/neu status of the primary tumor and the metastatic lesion. For examples, in a report by Gancberg et al. (28), the primary breast tumors and at least one distant metastatic lesion from 107 patients were analyzed by IHC and FISH. Similar degrees of amplification were observed in the primary (25%) and metastatic lesions (24%). However, in the FISH analysis, performed on 68 paired primary and metastatic tumors from breast cancer patients, 5 of the 68 (7%) cases showed discordance; in 3 of the 5 samples there was amplification in the metastasis but not in the primary tumor. In the case of IHC, there was 6% (6 of 100) discordance, and in all 6 cases there was greater HER-2/neu overexpression in the metastatic tissue than in the primary tumor.
The third method used to determine the HER-2/neu status is the ELISA, and it can be used to quantify either the full-length p185 in tumor tissue or the circulating ECD in serum or plasma. Zabrecky et al. (6), using Mabs directed to the ECD(29), demonstrated that the ECD was shed into the culture supernatant of SK-BR-3 breast cancer cells. Studies using specific Mabs against the HER-2/neu protein combined with immunoprecipitation and Western blot techniques showed that the ECD was a glycoprotein with molecular mass between 97 and 115 kDa (6). Subsequent studies illustrated that the ECD could be detected in the plasma of healthy individuals and was increased in women with primary and MBC (7). These observations were later confirmed by Leitzel et al. (8) and Pupa et al.(9).
In the last few years many reports have described a variety of ELISA formats that have been used to quantify the ECD in serum or plasma of breast cancer patients and control groups. However, it has been difficult to compare results between publications because of a lack of standardization between the ELISAs. For example, three publications reporting ECD results with one particular commercial assay (Triton-Ciba Corning-Chiron) used at least three different cutoff values (3, 12, and 30 units/mL) to separate healthy and diseased populations (8)(30)(31). In some reports, antibody specificity or assay validation for HER-2/neu has not been demonstrated (32), nor have adequate references been provided to demonstrate that the antibodies in the ELISAs specifically detect the HER-2/neu ECD.
In summary, IHC and FISH testing can be used to determine the HER-2/neu status in primary tumor tissue but are not adequate for assessing the HER-2/neu status of a woman after the tumor is removed. In contrast, the ELISA method for measuring the circulating HER-2/neu ECD is the only way to obtain the real-time status of HER-2/neu and the only practical way to monitor changes in the HER-2/neu ECD concentrations post surgery.
|
Methodology of Literature Review |
|---|
|
TopAbstractIntroductionMethods of Determining her-2/neu...Methodology of Literature ReviewResultsDiscussionReferences |
|---|
All of the titles generated by the search, and the abstracts when available, were reviewed for relevance by at least three of the authors, and the full articles were obtained. Each of the articles was then read by two of the authors, and those for subsequent inclusion in the data extraction stage were identified by the following criteria: (a) main objective of assessing the clinical utility of the HER-2/neu ECD measurement; (b) patient population defined, including age and pathology; (c) number of patients and any exclusion criteria identified; (d) timing of collection of blood samples identified; (e) analytical methods defined; (f) cutoff values for HER-2/neu ECD value defined; (g) data for assessing prognosis, predictive value, and clinical course available; and (h) indication whether the operators performing assays were blinded to the data from the assessment of the clinical status of the patient. Any discrepancy in the findings of the two primary reviewers was then assessed by a third reviewer, and a decision was made to include or exclude the data, the latter whenever the discrepancy could not be resolved.
|
Results |
|---|
|
TopAbstractIntroductionMethods of Determining her-2/neu...Methodology of Literature ReviewResultsDiscussionReferences |
|---|
and represent circulating HER-2/neu ECD measurements in >6500 patients with breast cancer. A review of 24 references used to evaluate ECD concentrations in PBC showed that in studies involving 1923 patients, 18.5% had circulating HER-2/neu ECD concentrations that were above the control cutoff described in each publication.
|
In contrast, a review of 45 references and 4622 patients with MBC showed that 43% of the patients had circulating HER-2/neu ECD values above the cutoff for the control group presented in the publication. In 15 of the 45 publications, ECD concentrations were increased above the control group in >50% of the patients studied. In a publication by Fehm et al. (49), 62% of the MBC patients were reported to have increased ECD. Fehm et al. also reported that 31% of the patients with PBC had increased ECD concentrations and in that population of patients, 62% developed increased concentrations during the metastatic disease.
It is interesting to note that the data in Table 1
came from essentially six different assays. The automated Immuno-1 HER-2/neu test (manufactured by Bayer HealthCare, Tarrytown, NY) and the Oncogene Science manual microtiter plate HER-2/neu test (manufactured by Oncogene Science, Cambridge, MA) are essentially the same assays in that both use anti-HER-2/neu Mabs NB-3 and TA-1, (29), a soluble p97–115kDa calibrator, and a 15 µg/L cutoff (44)(74)(83). It should also be noted that there was a strong correlation between the automated and manual HER-2/neu assays (44)(74)(83). Both the automated and manual HER-2/neu assays have been cleared by the FDA for use in the management and monitoring of women with MBC. These are the only two assays currently cleared for the measurement of the circulating p97–115kDa HER-2/neu ECD. There were 20 publications in Table 1
that used the two methods described above. There were 21 references to the Triton-Ciba-Chiron assays, with 11 different cutoffs ranging from 5 to 30 units/mL as well as a 120 or 450 fmol/mL cutoff. These assays are essentially the same and are no longer commercially available for studies. There were five references to the Nicherei assay, three references to the Calbiochem or ORP assay, three references to the Dianova assay, and two references to the Bender assay. We could not find any references that described the antibody specificities or standard materials used in the Calbiochem, Dianova, or Bender assays, nor could we find any references that validated biochemically that the assays clearly measured the circulating HER-2/neu ECD. The Calbiochem, Dianova, and Bender assays are available for research use, which means that their performance characteristics have not been determined. In fact, the Bender assay claims to measure the circulating soluble p185, but there has been never been a scientific report of a circulating full-length p185, nor does the manufacturer present data to support the claim. Several other reports, however, have reproducibly demonstrated that the only HER-2/neu fragment found circulating is the truncated p97–115kDa (6)(7)(8)(9) ECD, so it is unclear what the Bender assay actually measures.
In their report, Andersen et al. (34) showed that increased HER-2/neu ECD concentrations were detected in the serum of 8% of preoperative breast cancer patients and in only 3% of postoperative sera from patients without recurrent breast cancer. In contrast, 59% (55 of 93) of patients with recurrent breast cancer developed increased HER-2/neu ECD concentrations. They also reported that increased ECD concentrations were detected significantly more often in patients with distant metastases than in patients with recurrent disease restricted to local metastasis (68% vs 19%). This observation was supported by Watanabe et al. (30), who concluded that the circulating HER-2/neu ECD concentration was closely related to tumor mass because the HER-2/neu ECD concentration in recurrent disease was found to be significantly higher than in nonrecurrent disease.
Regardless of the ELISA used, studies show that there is great variation in shed ECD in patients with increased concentrations. The concentrations can range from thousands of micrograms per liter to slightly above the 15 µg/L cutoff in MBC patients. It is unclear what mechanisms account for such a wide variation in shedding and what is the biological or clinical significance of the shedding. However, as mentioned earlier, MMPs are involved in proteolytic cleavage of ECD, and the increased shedding may be indicative of up-regulated MMP activity. MMPs have been associated with tumor invasion and metastasis (14), and therefore, enhanced ECD shedding may be indicative of the more aggressive cancers.
In their report, Andersen et al. (34) showed that 14 of 24 patients who had IHC-positive breast tumors also had increased HER-2/neu serum concentrations during the metastatic phase of the disease. In contrast, 28 of 82 (34%) patients who had IHC-negative primary breast tumors developed increased serum concentrations during the metastatic disease. Kandl et al. (57) also reported that some patients with negative HER-2/neu tumor staining developed extremely high concentrations of circulating HER-2/neu during MBC and that increasing concentrations correlated with progressive disease, which also correlated with extensive disease burden. Molina et al. (66) also reported that 23% of patients with recurrent breast cancer with no tissue overexpression had increased ECD concentrations, once again supporting the concept that there is a subpopulation of women with HER-2/neu-positive tumors that are not identified by tissue testing.
Several reports [e.g., Andersen et al. (34), Fehm et al. (48), Kandl et al.(57), Krainer et al. (61), and Molina et al.(66)] thus show that there is a population of women with PBC who have HER-2/neu-negative tumors by tissue testing but who develop increased concentrations of HER-2/neu ECD in the metastatic setting. There may be several possible explanations for this observation. It could be that different methods are used to assess the HER-2/neu status (IHC for tissue and ELISA for ECD), and another is that the HER-2/neu status is determined at different times during the disease and, in some cases, years apart. As mentioned above, IHC is used to determine the HER-2/neu status in PBC, whereas ELISA is currently used to track ECD changes in HER-2/neu status in the metastatic setting. In addition, the comparison of the HER-2/neu IHC results of the primary tumor with circulating HER-2/neu ECD concentrations in MBC represents two entirely different biological situations. The major reason for the observation, however, may be that an individual was determined to have <10% HER-2/neu-positive staining cells in the primary tumor and was designated as HER-2/neu negative. It is certainly possible therefore that the few percentage of the HER-2/neu-positive cells in the primary tumor are sufficient to produce metastatic lesions, which can subsequently produce the increased circulating concentrations of ECD. A major question, therefore, is whether a woman with <10% HER-2/neu-positive staining cells can really be considered to have a HER-2/neu-negative tumor and not be eligible for HER-2/neu-targeted therapies. Additional clinical studies need to be performed to investigate this point because a significant number of women may be missing important treatment options.
The majority of published studies show that by tissue testing, 25–30% of women diagnosed with PBC have gene amplification or protein overexpression, but as documented by our search of the literature, 18.5% of women with PBC have increased HER-2/neu ECD concentrations. It is possible that the increased ECD concentrations found at primary diagnosis may reflect the existence of occult metastases and greater tumor burden than is evident by physical examination (49). As mentioned above, some reports show that >50% of women with MBC can have an increased ECD concentration. A possible explanation could be that the variations in ECD concentrations seen at the presentation of the disease may be attributable to variation in the activity of the mechanism responsible for release of ECD into the circulation and/or its subsequent metabolism. Is this lack of concordance therefore attributable to biological variation or is it a reflection of a disease process that evolves with cancer progression? These observations may also reveal that different analytical methods will be necessary to determine HER-2/neu status at different time points in the transition from PBC to MBC.
serum HER-2/neu as a prognostic indicator
We analyzed 20 publications and outcomes data from 4430 breast cancer patients (3338 metastatic patients and 1092 PBC patients) to determine whether increased HER-2/neu ECD concentrations correlated with poor prognosis. This analysis took into account several indicators of prognosis, including time to progression (TTP), OS, and disease-free survival (DFS). The data are summarized in Table 2
; in some cases, where available, the hazard ratios had also been calculated, and these have been summarized in Fig. 2
.
|
|
Bewick et al. (35) examined the clinical significance of shed ECD plasma concentrations in MBC patients and showed that 46% of the patients receiving high-dose chemotherapy (HDCT) had increased ECD concentrations compared with controls. The results showed that patients with high ECD concentrations had a significantly poorer OS and shorter progression-free interval after high-dose therapy with paclitaxel and autologous stem cell transplantation than patients with ECD concentrations below the cutoff. For example, the median OS of patients with low ECD concentrations was 29.8 months, which was significantly longer than the 15.9 months seen in patients with high ECD concentrations. In addition, the progression-free survival period was significantly longer: 13 months for patients with low ECD concentrations compared with 8.6 months for patients with high ECD concentrations (36). In a similar study by Harris et al. (52), patients treated with HDCT and bone marrow transplant and who had increased ECD concentrations did worse than patients receiving the same treatment but who had low ECD concentrations.
When the prognostic value of this oncoprotein was evaluated by Molina et al. (66), the patients with abnormally high presurgical serum HER-2/neu concentrations had a worse prognosis than those patients with concentrations below the cutoff in both node-negative and node-positive patients. In addition, serum ECD concentrations in patients with advanced breast cancer were related to the site of recurrence, with significantly higher values in patients with metastases (45.4%) than in those with locoregional recurrence (9.2%). In 2002, a report by Ali et al. (33) demonstrated that the median OS for MBC patients with increased serum HER-2/neu was 17.1 months. In contrast, women with serum HER-2/neu ECD concentrations below the cutoff had a median OS of 29.0 months, indicating that increased HER-2/neu concentrations correlated with poor clinical outcome.
In summary, the collective evidence presented in Table 2
from >4000 breast cancer patients shows a strong correlation between increased HER-2/neu ECD concentrations and worse prognosis as demonstrated by a decrease in TTP, decreased OS, and decreased disease-free progression.
serum HER-2/neu as a predictive indicator
Hormone therapy.
In the early 1990s, Wright et al. (84), using IHC, reported that patients with MBC who had HER-2/neu overexpression and estrogen receptor (ER) positivity had a response rate of only 20% to first-line hormone therapy, whereas 48% of the MBC patients who were ER-positive and who had a normal IHC expression of HER-2/neu responded to the first-line hormone therapy.
Since that report, there have been several publications concerning HER-2/neu ECD concentrations and the response rate of MBC patients to first- and second-line hormone therapies. Table 3
lists several published studies representing 1778 MBC patients and 119 patients with PBC and their response rates to hormone therapy with respect to serum HER-2/neu ECD concentrations.
|
In the 2002 report by Lipton et al. (63), the authors showed that MBC patients (n = 711) who were treated with second-line hormonal therapy (either the progestin megestrol acetate or the aromatase inhibitor fadrozole) and who had pretreatment serum HER-2/neu concentrations above the reference cutoff of 15 µg/L were less likely to respond to the therapy. In fact, only 20.7% of those with concentrations >15 µg/L responded, whereas 40.9% of MBC patients who had serum HER-2/neu concentrations below the cutoff of 15 µg/L responded. The response rates reported here were nearly identical to those reported by the Wright et al. (84) for their IHC studies. The studies by Lipton et al. (63) also demonstrated that patients with increased pretreatment HER-2/neu ECD concentrations had a shorter duration of response, a shorter TTP (Fig. 3
), and a shorter OS than ER-positive patients who had pretreatment serum ECD concentrations below the 15 µg/L cutoff. In a recent report, Lipton et al. (64) also demonstrated that MBC patients (n = 562) treated with either an aromatase inhibitor (letrozole) or an antiestrogen (tamoxifen) and who had increased pretreatment serum HER-2/neu concentrations had a shorter TTP, a shorter time to treatment failure, a decreased objective response rate (complete response + partial response), and a decreased clinical benefit rate (complete response + partial response + stable disease >24 months) than similar patients who had serum HER-2/neu concentrations below the cutoff of 15 µg/L. Therefore, both forms of hormonal therapy were less effective if the pretreatment ECD concentrations were above the 15 µg/L cutoff value. This study also showed that MBC patients with serum HER-2/neu ECD concentrations below the cutoff responded better to letrozole than to tamoxifen. However, there was no significant difference in response rate to letrozole or tamoxifen if the serum HER-2/neu concentration was increased. Therefore, the superiority of letrozole was the greatest in those patients with serum HER-2/neu concentrations below the cutoff.
|
Overall, the studies by Lipton et al. (63)(64) have evaluated 1273 MBC patients and demonstrated that those with increased pretreatment circulating HER-2/neu concentrations were relatively resistant to both first- and second-line hormonal therapy. Similarly, Fehm et al. (49) identified a subpopulation of breast cancer patients who had increased concentrations of HER-2/neu ECD at the time of metastases who had a poor response rate to first-line therapy, regardless of whether it was hormonal or chemotherapy. These patients also had a shorter OS and a shorter survival after relapse.
Yamauchi et al. (82) studied circulating HER-2/neu ECD concentrations in MBC patients who received droloxifene as first-line hormonal therapy. They reported that increased pretreatment concentrations of HER-2/neu ECD correlated with a significantly lower response rate (9% vs 56%), a shorter TTP, and a shorter OS. However, in a more recent publication (53), this group reported that although HER-2/neu was increased in 35–40% of MBC patients and that increased ECD was associated with poorer prognosis, as reported previously, they did not observe, in this particular group of patients, a predictive role for the HER-2/neu ECD in MBC patients who received either endocrine therapy or anthracycline-based chemotherapy. However, as pointed out by Hait in the same issue (85), these data must be interpreted with caution because of the small sample size, the retrospective study design, and the unique characteristics of the participants.
Chemotherapy.
Early studies by Gusterson et al. (86) and Muss et al.(87) demonstrated that overexpression of HER-2/neu correlated with decreased responsiveness to combination therapy of cyclophosphamide–methotrexate–5-fluorouracil (CMF).
In 1995, Tsai et al. (88) demonstrated that increased expression of p185neu led to enhanced chemoresistance after transfection of HER-2/neu into non-small cell lung cancer cell lines established from untreated patients. In a follow-up 1996 report, the same investigators examined a panel of 20 non-small cell lung cancer cell lines for HER-2/neu ECD concentrations in the cell culture supernatants and then correlated the ECD concentrations with chemoresistance or chemosensitivity (89). A variety of cytotoxic drugs were used, such as doxorubicin, cisplatin, etoposide, and malphalan. The results showed that high HER-2/neu ECD concentrations correlated with chemoresistance and that cell lines expressing low ECD concentrations were relatively chemosensitive. In the report, multivariate analysis revealed that the concentration of p185 neu was the only predictor for chemoresistance to doxorubicin and etoposide. In a 1997 report by Fehm et al. (47), they showed that in node-positive breast cancer patients, those with increased HER-2/neu ECD concentrations had a worse outcome than patients with nonincreased HER-2/neu concentrations when treated with adjuvant CMF or cyclophosphamide-Novantrone-fluorouracil. In 1997, Pegram et al. (90) also reported that HER-2/neu overexpression altered chemotherapeutic drug sensitivity in a variety of human breast and ovarian cancer cells.
The next step, therefore, was to analyze the predictive role of HER-2/neu ECD concentrations and response to various regimens of chemotherapy. The data are summarized in Table 4
. We reviewed 12 publications, which contained data on 131 PBC patients and 1228 patients with MBC.
|
In their study of 58 MBC patients, Colomer et al. (43) showed that the probability of obtaining complete response to a paclitaxel–doxorubicin chemotherapy regimen was significantly lower in patients with increased HER-2/neu ECD concentrations compared with patients having nonincreased concentrations. In addition, the duration of clinical response was significantly shorter in patients with increased HER-2/neu ECD concentrations compared with the cases with nonincreased concentrations. For example, the duration of response was only 7.5 months for patients with increased ECD concentrations compared with 11 months for patients with ECD concentrations below the cutoff. Overall, increased concentrations of ECD correlated with reduced efficacy of a paclitaxel–doxorubicin chemotherapy combination.
In their report, Mehta et al. (91) concluded that breast cancer patients with more than three positive lymph nodes would benefit from determining the prechemotherapy concentrations of HER-2/neu ECD and that this could serve as an important tool to predict the response of breast cancer patients to chemotherapy. Postchemotherapy c-erbB-2 concentrations were also a prognostic indicator for DFS in patients who received chemotherapy. The reports by both Mehta et al. (91) and Fehm et al.(49) concluded that patients with increased HER-2/neu ECD concentrations had a lower response to first-line therapy than breast cancer patients with serum HER-2/neu ECD concentrations below the cutoff.
Nunes and Harris (12) and Kaptain et al.(11) reviewed the role of HER-2/neu ECD in relation to chemotherapy containing alkylating agents, anthracycline-containing agents, or taxanes. They concluded that HER-2/neu-positive tumors are relatively resistant to CMF-containing regimens but had an increased sensitivity to anthracycline-containing regimens. The clinical value of HER-2/neu ECD testing and response to various chemotherapeutic regimens continues to be an area of research investigation. Therefore, the availability of an FDA-cleared serum HER-2/neu ECD test will allow standardized studies to be conducted and the comparison of data from laboratory to laboratory.
Yu et al. (92) showed that cells that overexpress HER-2/neu protein are more resistant to taxanes such as paclitaxel. In addition, Yu et al. (92) reported that their experimental findings were supported by recent reports from a phase 3 clinical trial that the response rate to Taxol was significantly improved in breast cancer patients when HER-2/neu was down-regulated with Herceptin (93). This concept is also support by xenograft models in which Herceptin appeared to reverse the resistance to both paclitaxel and docetaxel (94). These observations are especially interesting because some of the best current response rates in patients are seen when Herceptin is combined with docetaxel or paclitaxel (95)(96).
Herceptin-based therapy.
We reviewed the four available publications representing 127 MBC patients treated with Herceptin-based therapies. The data are summarized in Table 5
.
|
In a report in 2002, Esteva et al. (95) analyzed the response rate of 30 MBC patients (all shown to be HER-2/neu positive by tissue testing) treated with docetaxel and Herceptin who were separated into responders with high or low ECD concentrations. The circulating HER-2/neu ECD concentrations were measured at baseline and at the time of response evaluation for all 30 patients. The median concentration at baseline was 41.9 µg/L (range, 7.1–666.5 µg/L). Twenty-one patients (70%) had increased HER-2/neu (>15 µg/L) concentrations at baseline. The patients with high baseline ECD concentrations (>15 µg/L) had the highest response rate to Herceptin-based therapy (76%), whereas only 33% of those with low ECD concentrations (15 µg/L) responded. In comparison, 67% of the FISH-positive patients responded. The overall response rate achieved was 63%; however, when patients with minor responses and stable disease were considered, 83% of the patients obtained some clinical benefit from the combination of weekly docetaxel and Herceptin. Esteva et al. (95) also reported that serial changes in serum HER-2/neu ECD concentrations correlated very well with clinical response to weekly docetaxel and Herceptin therapy. The authors concluded that additional research is warranted to determine the value of serum HER-2/neu ECD testing in selecting and monitoring patients undergoing Herceptin-based therapy.
In a retrospective study designed to determine whether HER-2/neu concentrations could predict outcomes of Herceptin-based therapy, Hoopmann et al. (96) examined plasma samples from 20 MBC patients. HER-2/neu ECD concentrations were measured in samples collected at the beginning of Herceptin-based therapy and at the time of first diagnostic imaging. HER-2/neu concentrations were evaluated in relation to the clinical course of each patient and using imaging results, and then the data were analyzed with respect to prediction of the patient response to therapy, defined as the TTP. The data presented showed that plasma HER-2/neu concentrations were predictive of clinical course in 16 of 20 patients (80%) and that all therapy responders had ECD concentrations below the cutoff (15 µg/L). The patients with permanently increased (15 µg/L) or increasing serum HER-2/neu concentrations displayed a poor clinical outcome in 6 of the 10 cases. On the basis of these results, the authors concluded that plasma HER-2/neu concentrations might be an early predictor of response to Herceptin-based therapies.
Finally, in a report by Dnistrian et al. (46) MBC patients were stratified according to pretreatment serum HER-2/neu concentrations to determine whether ECD concentrations could predict the response to Herceptin-based therapies. Of 18 patients treated with Herceptin and Taxol who had pretreatment ECD concentrations below the cutoff (>15 µg/L), 6 responded favorably to Herceptin-based therapy. In contrast, 30 of the 36 patients (83%) who had increased pretreatment serum HER-2/neu concentrations (mean, 344 µg/L) had a favorable response. Furthermore, in all patients with abnormally high pretreatment serum HER-2/neu concentrations, the ECD concentration decreased significantly with disease regression, and in most cases it returned to values below the cutoff.
The observation that increased serum HER-2/neu concentrations may be predictive of response to Herceptin-based therapy is interesting in light of the publications demonstrating that increased pretreatment HER-2/neu ECD concentrations predict poor response to hormone therapy and some regimens of chemotherapy. However, it is not entirely unexpected that high concentrations can predict response to Herceptin because the HER-2/neu receptor provides the binding site for Herceptin that is necessary for its action. The shed ECD may therefore provide an indication of the amount of receptor that is available for binding, as well as the activity of the malignancy. Additional studies will be needed to clarify this possibility.
monitoring HER-2/neu ecd concentrations in mbc patients treated with hormone or chemotherapy
We reviewed 16 references (14 related to MBC and 4 related to PBC; summarized in Table 6
) representing 1148 breast cancer patients. The 1148 patients were divided into 499 MBC patients and 649 PBC patients. In these studies, serial or longitudinal changes in serum HER-2/neu ECD values were compared with the clinical course of disease in women with MBC. The clinical course of a patient’s disease (defined as progression, response, or stable disease) was determined by monitoring results of clinical tests such as x-rays or computed tomography scans. Patients were then classified as to whether changes in serum HER-2/neu concentrations did or did not correspond to the clinical course of disease. For example, it was determined whether the serum value increased with progression or decreased with response to therapy to determine correspondence.
|
In the study by Cook et al. (44), longitudinal monitoring was performed on 103 stage IV MBC patients who were being treated with various regimens of hormonal or chemotherapy to determine whether serum HER-2/neu changes correlated with changes in the clinical course of disease. Thirty-eight (36.9%) of the 103 stage IV patients had an increased ECD (>15 µg/L), 33 of whom showed longitudinal HER-2/neu values that paralleled the clinical course of disease, which gave an overall sensitivity of 86.8%.
Schwartz et al. (74) reported serum HER-2/neu concentrations in patients with MBC who were receiving a variety of conventional therapeutic regimens. All patients had a pretherapy serum specimen, and four had posttherapy specimens. Longitudinal testing of the serum HER-2/neu concentrations clearly showed that women who expressed HER-2/neu in their tissue had increased serum concentrations (15 µg/L) and that changes in the ECD concentrations reflected the clinical course of the patient’s disease. This report also investigated whether there were ECD serial changes in healthy individuals over a several-month period. In the study, six specimens were drawn monthly from seven premenopausal and eight postmenopausal women and tested for serum HER-2/neu concentrations. These studies demonstrated that the serial HER-2/neu values were consistent for a given person over a several-month period. In one representative example, the serum HER-2/neu values over several months were very consistent and showed very little variation. The serum values were 8.2, 8.3, 7.7, 8.3, 8.1, and 8.1 µg/L. Similar data were obtained with the remaining healthy individuals and were a good basis for comparative monitoring studies in MBC patients.
Cheung et al. (41) examined serial changes in serum HER-2/neu concentrations in 30 MBC patients from two multicenter trials in which patients received either docetaxel-based therapy or doxorubicin-based therapy. The authors concluded that among the patients with positive tissue staining, sequential changes in serum completely paralleled the initial response to therapy.
In a study by Lueftner et al. (65), serum samples were taken weekly from 35 patients to monitor changes in the serum HER-2/neu ECD concentrations and to correlate those changes to the clinical course of disease. The MBC patients received dose-intense paclitaxel treatment. In this study, the overall response rate was 36%, but the response among the HER-2/neu-positive patients was 62%, showing a high sensitivity of the HER-2/neu-positive patients to dose-intense paclitaxel treatment. In all responders, the HER-2/neu concentration decreased below the detection limit either before the clinical diagnosis of response or by the end of the next cycle. However, in this cohort of patients, normalization of the HER-2/neu concentrations also occurred in patients who were stable or had progressive disease. The authors speculate that this could be explained by the fact that the chemotherapy is effective against the HER-2/neu-positive tumor cells but not against the HER-2/neu-negative tumor cells, which is the cell population responsible for the progressive disease. It was also pointed out by the authors that chemotherapy may alter the mechanisms by which the ECD is shed, but no data were presented to support the hypothesis.
In summary, numerous reports of patients receiving hormone or chemotherapy showed that longitudinal changes in serum HER-2/neu concentrations paralleled the clinical course of a patient’s disease. Overall, several studies showed that increases in serum HER-2/neu concentrations were reflective of progressive breast cancer, whereas decreasing serum HER-2/neu concentrations were reflective of response to therapy or a prolonged lack of disease progression.
monitoring mbc patients treated with herceptin-based therapies
Women with MBC being treated with Herceptin and various chemotherapies (n = 107) (46)(73)(95) were monitored for changes in HER-2/neu ECD concentrations before treatment and then serially thereafter. Previous studies reported by Payne et al. (83) showed that Herceptin does not interfere with measuring ECD concentrations in the HER-2/neu assay used in all three of the studies described below. In an initial report by Schwartz et al. (74) in 2000, it was suggested that changes in serum HER-2/neu during Herceptin-based therapy might parallel the clinical course of disease; however, there were too few patients in the report to make a valid conclusion.
In a 2002 study by Esteva et al. (95), 30 MBC patients treated with docetaxel (Taxotere) and Herceptin were monitored for changes in serum HER-2/neu to determine whether serial changes would reflect the clinical course of disease. Studies showed that ECD concentrations decreased in 14 of 16 (87%) responding patients. As seen in studies of patients receiving conventional therapies, serial changes in serum HER-2/neu concentrations did parallel the clinical course of disease after Herceptin treatment.
In their study, Schoendorf et al. (73) serially monitored 23 MBC patients treated with Herceptin and chemotherapy (median time of treatment, 13 months; range, 4–22 months) for changes in serum HER-2/neu ECD. The changes in HER-2/neu ECD concentrations were then evaluated in conjunction with response or lack of response to therapy. In the group of patients with increased concentrations (12 of 19), 35 events of either response or progression were documented. The serial changes in serum HER-2/neu correlated with remission or disease progression in 74% of the patients. The correlation between serial ECD changes and clinical changes was increased when the analysis focused on MBC patients with visceral metastasis. This group of clinical investigators concluded that serial changes in plasma HER-2/neu ECD concentrations did parallel changes in the clinical course of disease.
In a 2003 report (97), a group 57 MBC patients treated over a 2-year period with Herceptin and Taxol were monitored for changes in serum HER-2/neu, and the data were correlated with clinical changes. The study clearly showed that serum HER-2/neu ECD changes paralleled the clinical course of disease. The women with serially decreasing ECD concentrations responded to Herceptin-based therapy, whereas women with progressing breast cancer that did not respond to the combined therapy had serially increasing ECD concentrations. These data were consistent with the observations made in MBC patients who were treated with conventional hormone therapies and chemotherapies.
In summary, these four publications collectively studied a small number of women with MBC who received Herceptin-based therapies and who had their serum HER-2/neu concentrations monitored for up to 2 years. All four reports showed that serial changes in the ECD concentrations reflected the clinical course of disease. Women who had serially decreasing ECD concentrations responded to Herceptin-based therapy, whereas women with progressive breast cancer had serially increasing ECD concentrations. Several studies have recently been submitted for publication substantiating these data.
serum HER-2/neu concentrations and detection of early recurrence
In the 1996 report by Molina et al. (67), the authors evaluated the utility of measuring HER-2/neu, carcinoembryonic antigen (CEA), and CA 15-3 in the early diagnosis of recurrence. Serial serum measurements were performed in 200 PBC patients (no evidence of residual disease) who were followed for 1–4 years (median, 2.2 years). Of the 89 patients who developed metastases, 28% had HER-2/neu concentrations above the cutoff, 30% had a CEA that was above the reference cutoff, and 47% had increased concentrations of CA 15-3. The mean (SD) lead time before diagnosis was 4.5 (2.4) months for HER-2/neu, 4.8 (2.4) months for CA 15-3, and 4.9 (2.4) months for CEA. However, sensitivity was clearly related to the site of recurrence, with the lowest sensitivity found in patients with locoregional relapse and the highest sensitivity in patients with visceral metastasis. When patients with locoregional relapse were excluded, the sensitivity for HER-2/neu improved to 31%, and the overall sensitivity of early detection with all three markers combined was 76%. These data and others are summarized in Table 6
. The increase in sensitivity observed by Molina et al. (67) was also supported by studies by Watanabe et al. (30) and Schwartz et al.(74); however, Eskilinen et al. (98) reported only limited value in measuring serum CEA, CA 15-3, and HER-2/ neu in conjunction with other cancer tests.
In the study by Dnistrian et al. (46), HER-2/neu, CA 15-3, and CEA were all measured serially from baseline to investigate the changes in 54 MBC patients undergoing Herceptin and Taxol therapy. When the data for the individual markers were combined, the sensitivity for monitoring for response was 76%, similar to that reported by Molina et al. (67), but individually the sensitivities were 67% (only 31% in the report by Molina et al.) for HER-2/neu, 54% for CA 15-3, and 43% for CEA (46).
In the study by Ali et al. (33), CA 15-3 (a surrogate marker for disease burden) was measured along with serum HER-2/neu in 566 ER/progesterone receptor-positive MBC patients. These patients were treated with second-line hormone therapy (megestrol acetate) or an aromatase inhibitor (Fadrozole). Overall, 30% of the patients had increased HER-2/neu ECD concentrations (n = 168), and 60% had increased CA 15-3 concentrations; however, there was only a weak correlation between the two. Similar to a previous report, the clinical benefit (complete response + partial response + stable disease >24 weeks) of endocrine therapy was significantly lower in patients with increased HER-2/neu ECD concentrations. The investigators concluded that HER-2/neu was a significant independent predictive and prognostic factor in hormone receptor-positive MBC patients even when adjusted for tumor burden as measured by CA 15-3. The combination of increased HER-2/neu and CA15-3 predicted a worse prognosis for MBC patients than did increased CA 15-3 alone.
In their report, Colomer et al. (43) suggested that a panel of tumor tests such as CEA, CA 15-3, and HER-2/neu could be used to monitor patients postoperatively to increase the sensitivity of detecting early recurrence. The value of early detection will increase with the introduction of a variety of targeted therapies used as either monotherapy or in conjunction with either hormone therapy or chemotherapy.
|
Discussion |
|---|
|
TopAbstractIntroductionMethods of Determining her-2/neu...Methodology of Literature ReviewResultsDiscussionReferences |
|---|
More recently, the HER-2/neu receptor has become a major therapeutic target for the treatment of MBC. Similar to the ER status, which guides hormone therapy, HER-2/neu status now guides the administration of the anti-HER-2/neu therapy Herceptin. The importance of determining the HER-2/neu status was substantiated in a recent report in which the erbB2 status was shown to be superior to the ER as a prognostic factor (101). In addition, Slamon et al. (17) confirmed the importance of knowing the HER-2/neu status by reporting that patient survival was increased when Herceptin was combined with chemotherapy for treatment of women with MBC.
Studies have now shown that the HER-2/neu receptor is proteolytically cleaved from the cell surface, and a report by Molina et al. (16) indicates that it is a process involving MMP activity. This is interesting because reports have also indicated that HER family members can significantly up-regulate specific MMPs, which are key enzymes involved in angiogenesis and invasion (14). These results suggest a complex interaction between the HER family members, MMPs, and signal transduction (100). The circulating HER-2/neu ECD has been reproducibly shown to be a glycoprotein with a molecular mass of 97–115 kDa by Western blot studies and has been shown to be measurable in both plasma and serum specimens by immunoassay. Before the recent clearance by the FDA of a serum HER-2/neu assay with a cutoff of 15 µg/L, there had been numerous publications describing "research use only" assays or homebrew assays and a complete lack of assay standardization. This lack of standardization has led to considerable confusion in comparing concentrations of the circulating ECD from publication to publication. The same can be said for IHC and FISH assays. However, FDA clearance of IHC and FISH assays and their wide availability has gone a long way to enhance the correlation between IHC and FISH and to harmonize the results on primary tumor tissue. In contrast to ECD testing, which can be performed at any time during breast cancer, the tissue tests are limited to measuring either HER-2/neu p185 overexpression or DNA amplification in the primary tumor tissue. Once the tumor is removed, measuring the circulating HER-2/neu ECD concentrations are the only practical way of assessing HER-2/neu status.
The immunoassays measuring the circulating HER-2/neu (p97–115kDa) ECD have been used in numerous research studies to determine the ECD concentrations in all stages of breast cancer. To understand the collective contributions of the publications pertaining to the circulating HER-2/neu ECD and the potential clinical utilities of measuring the ECD concentrations, we preformed a systematic review of the literature. We focused on publications that investigated circulating HER-2/neu ECD with respect to clinical outcome data such as OS and TTP.
Before the decision by the FDA in 2000, which established a cutoff value of 15 µg/L for an immunoassay of the serum HER-2/neu ECD, many publications used a variety of assay formats with various antibody specificities, calibrators, and cutoff values to measure the circulating HER-2/neu. For example, publications using the Triton/Ciba/Chiron assays reported at least 11 different cutoff values to separate the control group from the cancer group. In their study, Pegram et al. (102), using a homebrew assay and Mab 4D5 as the captured reagent, reported that 259 of 443 (58%) patient sera did not have measurable ECD concentrations in the serum. To date, this is the only report indicating that ECD concentrations are not measurable in control individuals. In a report by Cobleigh et al. (103), using the same unvalidated assay and Mab 4D5, serum concentrations of baseline shed HER-2/neu were below the detectable concentration in 73 of 191 patients (38%). The investigators concluded that there was no significant correlation demonstrable between shed ECD concentrations and response status. However, in an abstract presented at the American Society for Clinical Oncology in 2000, Wong et al. (104) reported that ELISAs using Mab 4D5 could not detect all circulating HER-2/neu ECD, thus offering a possible explanation for the low results presented in the reports by Pegram et al. (102) and Cobleigh et al. (103). These publications demonstrate that use of homebrew or unvalidated assays could lead to inaccurate and incorrect conclusions about measuring HER-2/neu ECD concentrations.
To be objective in our analysis of the data, we compared the ECD concentrations reported in each publication for the cancer population with the control group from that particular publication, using the cutoff established in the particular publication. We also considered the data in conjunction with the clinical outcomes that were also presented in the particular publication. Although there are doubts about some of the assays in terms of absolute antibody specificities, assay standardization, and validation, there is a consistency, with the exceptions of the articles by Willsher et al. (32), Pegram et al. (102), and Cobleigh et al.(103), of the conclusions that can be drawn from the analysis of the data. Because there is now an immunoassay that is cleared for measuring the circulating ECD concentrations in women with MBC, there is an important need for larger multicenter prospective trials to confirm proposed clinical utilities that have been suggested from the numerous published research studies.
The vast majority of the reports published since 1991 have used the IHC test to measure p185 overexpression and FISH assays to measure DNA amplification to determine the HER-2/neu status of the primary tumor. However, the tissue result, which indicates HER-2/neu status of the primary tumor, may not necessarily reflect the HER-2/neu status of the patient with MBC. This is illustrated by several reports (34)(48)(57)(61)(66) that demonstrated that women determined to have a HER-2/neu-negative tumor by tissue testing had increased HER-2/neu ECD concentrations during the metastatic phase of the disease. This concept is also supported by the report by Gancberg et al. (28), which indicated that three of five breast cancer samples tested by FISH showed amplification in the metastatic lesion but not in the primary breast tumor. In addition, 6% of the cases studied by IHC showed greater HER-2/neu expression in the metastatic than in the primary tumor. In addition, Fehm et al. (49) reported that 31% of women with PBC had an increased ECD concentration, which in the metastatic phase of the disease progressed to 62%. In our analysis of the literature we found that 43% of the 4622 women with MBC from 45 publications had an increased HER-2/neu ECD concentration and that at least 15 publications reported values >50% and as high as 80%. This is in contrast to numerous reports over the past 20 years in which the authors state that only 25–30% of women with breast cancer have HER-2/neu-positive tumors. This has important therapeutic implications for women with MBC because only women with HER-2/neu status confirmed by tissue testing are eligible for treatment with the HER-2/neu-targeted Herceptin therapy. In addition, there is a subset of women with MBC for which a tissue block for tissue testing is not available; these women are therefore not eligible for Herceptin treatment. Studies need to be performed to determine whether serum testing can qualify such women for Herceptin therapy.
Our systematic review of the literature also revealed several interesting observations with respect to the circulating HER-2/neu ECD. With respect to the prevalence of increased ECD concentrations, an analysis of 24 references and 1923 PBC patients showed that 18.5% of the patients had an increased ECD concentration at the time of cancer diagnosis. Because several studies have shown that increased concentrations are associated with tumor mass and distant metastases, Fehm et al. (49) suggested that increased concentrations in PBC might be indicative of occult metastatic lesions. There have been studies of the immunohistochemical localization of HER-2/neu in the primary tumor with subsequent study of lymph node tissue at the time of progression in which the discordance of positivity between primary and metastatic tissue has varied form 7% to 25% (105). This has raised the question of wh
, DFS.