The HER2 phenotype of circulating tumor cells in HER2-positive early breast cancer: A translational research project of a prospective randomized phase III trial

Background

HER2 is one of the predominant therapeutic targets in breast cancer. The metastatic selection process may lead to discrepancies between the HER2 status of the primary tumor and circulating tumor cells (CTCs). This study analyzed the HER2 status of CTCs in patients with HER2-positive primary breast cancer at the time of diagnosis. Aim of the study was to assess potential discordance of HER2 status between primary tumor and CTCs, as this may have important implications for the use of HER2-targeted therapy.

Methods

The number and HER2 status of CTCs out of 30ml peripheral blood were assessed in 642 patients using the CellSearch System (Janssen Diagnostics, USA). The cutoff for CTC positivity was the presence of at least 1 CTC, and the cutoff for HER2 positivity of CTCs was the presence of at least 1 CTC with a strong HER2 staining.

Results

258 (40.2%) of the 642 patients were positive for CTCs (median 2; range 1–1,689). 149 (57.8%) of these 258 patients had at least 1 CTC with strong HER2 staining. The presence of HER2-positive CTCs was not associated with tumor size (p = 0.335), histopathological grading (p = 0.976), hormone receptor status (ER: p = 0.626, PR: p = 0.263) or axillary lymph node involvement (p = 0.430). Overall, 83 (32.2%) of the CTC-positive patients exclusively had CTCs with strong HER2 staining, whereas 31 (12.0%) had only CTCs with negative HER2 staining. Within-sample variation in the HER2 status of CTCs was found in 86 (57.8%) of the 149 patients with more than 1 CTC.

Conclusion

This study demonstrated that discordance between the HER2 expression of CTCs and that of the primary tumor frequently occurs in early breast cancer. Future follow-up evaluation will assess whether this discrepancy may contribute to trastuzumab resistance.

Abstract P1-04-06: Discordance of the ER- and HER2-status on disseminated tumor cells compared to the primary tumor in patients with early breast cancer

Background: Differences in ER- and HER2-expression on metastases compared to the primary tumor (PT) are a known phenomenon and may have clinical implications in respect of targeted systemic treatment approaches. The aim of this study was to evaluate both ER- and HER2-status on disseminated tumor cells (DTCs) in the bone marrow (BM) of patients (pts) with early breast cancer (EBC) and to compare these with the corresponding PT.

Methods: BM aspirates were obtained at the time of first surgery. After Ficoll enrichment for mononuclear cells two cytospins with 106 BM cells were evaluated for ER-, HER2- and cytokeratin (CK) -expressions simultaneously by immunocytochemistry using a triple fluorescence staining method with antibodies directed against human ER (secondly labeled with Cy3, red), HER2 (Coumarin-AMCA, blue) and CK (DyLight488, green). The manual analysis was conducted using a computerized fluorescence microscope (Axioskop, Zeiss, Germany). Criteria for CK- and HER2-positivity were the ring-like appearance of the respective membrane stainings and for ER-expression a nuclear staining. Only pts with the detection of CK positive cells (DTC+) and known ER- and HER2-status of the PT (n = 54) were selected for this analysis.

Results: The median number of DTCs was 13 (range 1-95; total number of DTCs detected: 1082). 40 (74%) of the pts had at least one ER-positive (pos) DTC, 24 (44%) at least one HER2-pos DTC, 14 (26%) at least one ER-pos/HER2-pos DTC, and 50 (93%) at least one ER-negative/HER2-negative (neg) DTC, while 10 (19%) pts had only ER-neg/HER2-neg DTCs.

The concordance rate between ER-status on DTCs and PT was 74%. Pts with an ER-pos PT were significantly more likely to have at least one ER-pos DTC (34 out of 42) than pts with an ER-neg PT (6 out of 12; Chi-square test, χ2 = 4.66, p = 0.031). 39 (93%) of the 42 pts with ER-pos PT had at least 1 ER-neg DTC.

The concordance rate between HER2-status on DTCs and PT was 52%. The probability of having at least one HER2-pos DTC was not related to the HER2-status of the PT (Chi-square test, χ2 = 0.34, p = 0.56). 22 (46%) of the 48 pts with a HER2-neg PT had at least one HER2-pos DTC. All of the 6 pts with a HER2-pos PT had at least one HER2-neg DTC.

7 out of 10 pts with a triple-neg PT had at least one DTC pos for ER, HER2 or both. Further the heterogeneity of the ER- and HER2-expression on DTCs compared to the PT for different DTC counts was evaluated. We detected all possible combinations of ER- and HER2-experssion on DTCs regardless of the respective status of the PT.

Conclusions: Our study confirms that the ER- and/or HER2-status on DTCs may differ compared to the PT. This discordance could be especially important for pts with a triple-neg PT and ER-pos or HER2-pos DTCs, since they might respond favorably to an endocrine or HER2-targeted therapy. On the other hand, the presence of ER-neg or HER2-neg DTCs in pts with ER-pos or HER2-pos PT might explain some of the failures of adjuvant endocrine or HER2 targeted therapy.

Citation Information: Cancer Res 2013;73(24 Suppl): Abstract nr P1-04-06.

Abstract P2-01-02: Circulating Tumor Cells (CTC) may Express HER2/neu in Patients With Early HER2/neu Negative Breast Cancer – Results of the German SUCCESS C Trial

Background: There is growing evidence that the HER2/neu-status of distant metastases or minimal residual disease in blood and bone marrow may differ from the primary tumor in patients with breast cancer. The HER2/neu-status of CTCs was prospectively evaluated in patients with HER2/neu negative primary breast cancer randomized into the German multicenter SUCCESS C study.

Methods: The SUCCESS C trial is a randomized Phase III study comparing FEC-Docetaxel (FEC-Doc) vs. Docetaxel-Cyclophosphamid (DC) as well as 2 years of a lifestyle-intervention in patients with early, HER2/neu negative, node positive or high-risk node negative primary breast cancer.

As part of the translational research program, 23ml peripheral blood were drawn after adjuvant chemotherapy. In 505 samples, the prevalence of CTCs and their HER2/neu-status were assessed using the CellSearch System (Veridex, USA). After immunomagnetic enrichment with an anti-Epcam-antibody, cells were labelled with anti-CK8/18/19 and anti-CD45 antibodies. A fluorescein conjugate antibody with anti-CK-Fluorescein Isothiocyanate (FITC) was used for HER2/neu phenotyping. The cut-off for CTC-positivity was ≥ 1 CTC and for HER2/neu ≥ 1 CTC with strong HER2/neu-staining (+++).

Results: 26,9% of pts (n = 136) were positive for CTCs (mean 1.78; range 1–7; median = 1). The number of detected CTC was distributed as follows: 1 CTC (n = 76; 55.9%), 2 CTCs (n = 35; 25.7%), 3 CTCs (n = 13; 9.6%), 4 CTCs (n = 7; 5.2%) and ≥ 5 CTCs (n = 5; 3.7%). HER2/neu staning of CTCs was not detectable or weak in 26.5% (n = 36) and 4.4% (n = 6) of CTC positive patients respectively and therefore categorized as HER2/neu negative. In 32.4% of the CTC-positive patients (n = 44), we detected moderate and in 36.8% (n = 50) strong HER2/neu-staining of ≥ 1 CTC per sample. No association was found between CTCs or the HER2/neu-status of CTCs with tumor size, histopathological grading, hormone receptor status or axillary lymph node involvement.

Conclusions: The data of this trial confirm previous findings that patients with HER2/neu negative primary breast cancer can show HER2/neu positive minimal residual disease. These results underline the importance of frequent HER2/neu determination during follow up and disease progression. Survival data within the Success C trial will give further insight into the tumor biology of HER2/neu negative disease.

Citation Information: Cancer Res 2012;72(24 Suppl):Abstract nr P2-01-02.

Abstract P4-13-11: Prognostic factors in young breast cancer patients over time – a 40 year longitudinal analysis

Background: In younger patients (pts) breast cancer (BC) is associated with a worse prognosis compared to older pts. Only few data are available presenting the development of prognostic factors regarding a period of several decades. This 40 year (y) longitudinal comparison of the main prognostic factors was performed to investigate differences in younger women with primary BC in contrast to older women over time.

Patients and Methods: In this retrospective analysis a consecutive pts cohort of 4010 pts was analyzed. Pts were documented and treated for primary invasive breast cancer between 1963 and 2003 at two University Hospitals in Germany. To be eligible, pts were required to have identified tumor characteristics, including TNM-status. Pts with carcinoma in situ or distant metastases were excluded. The cohort was divided in two age groups, ≤40y and >40y. Furthermore to reveal trends and changes over the duration of 41 years the period of analysis was split into 3 time frames: 1963–1976, 1977–1989 and 1990–2003. We analyzed the main prognostic factors for BC including tumor size, grading, nodal status and HR-status in longitudinal comparison regarding the three time frames, respectively. During 1963–77 HR-status was determined in just 12.6% of pts. Thus, this time frame was excluded in the analysis of HR-status.

Results: In 41 yrs, 747 (18.6%) pts were treated between 1963–76, 1722 pts (42.9%) in 1977–89 and 1541 pts (38.4%) in 1990–2003. Overall 358 pts were ≤40y and 3652 pts were over the age of 40.

A significant reduction of tumor size (metric assessment) at primary diagnosis was observed for both age groups (pts≤40y: p = 0.012; pts>40y: p < 0.0001) with no difference between the time frames, respectively (1963–76: p = 0.289; 1977–89: p = 0.647; 1990–2003: p = 0.937).

The number of node-positive pts significantly decreased in pts >40 y (p = 0.001) whereas no difference could be seen in pts aged ≤40 y (p = 0.991).

In both age groups the number of G2/3 tumors increased over the yrs (pts≤40y: p = 0.001; pts>40y: p < 0.0001). Between 1963–76 more G1-tumors were diagnosed in younger pts (p = 0.041) whereas in the two following periods significantly more G2/3 tumors (p = 0.001; p = 0.002) were observed in this group.

In young pts, 54.2% were HR-negative in 1977–89 and 47.6% in 1990–2003 (p = 0.323). In pts >40y 38.4% and 21.7% were HR-negative, respectively (p < 0.0001). Comparing the age groups, significantly more pts were HR-negative in the young patient group (p = 0.001 (1977–89) and p < 0.0001 (1990–2003)).

Conclusions: Concerning the large period of 40yrs, technical improvement and increasing awareness for BC are reasons for decreasing tumor size at the time of primary diagnosis. However, the rate of node-positive pts in the young patient group remained stable. Furthermore, the high percentage of HR-negative pts and the increasing number of pts with unfavorable tumor grade demonstrate more aggressive tumor types in younger pts. These data confirm the need to improve screening tools to early identify young women with the risk to develop breast cancer, and for personalized treatment approaches in these patients.

Citation Information: Cancer Res 2012;72(24 Suppl):Abstract nr P4-13-11.

Abstract OT1-1-10: DETECT III - A multicenter, randomized, phase III study to compare standard therapy alone versus standard therapy plus lapatinib in patients with initially HER2-negative metastatic breast cancer but with HER2-positive circulating tumorcells

Background: Despite a HER2-negative primary tumor approximately 20–30% of patients develop HER2-positive metastases (Zidan et al. 2005; Tewes et al. 2009). As previously described in the DETECT I trial (Fehm et al. 2010) determination of HER2 status on circulating tumor cells (CTCs) is one option for re-evaluating HER2-status in the metastatic setting. Currently it is unclear if HER2-targeted therapy based on the assessment of HER2-status of CTCs reveals a clinical benefit.

Trial design: DETECT III is a randomized, open-label, two arm phase III study comparing standard treatment alone vs. standard treatment plus HER2-targeted therapy with lapatinib in HER2-negative metastatic breast cancer patients with HER2-positive CTCs. Choices of chemotherapy and endocrine therapy include: docetaxel, paclitaxel, capecitabine, vinorelbine, non pegylated liposomal doxorubicin, letrozole, exemestane and anastrozole.

Main eligibility criteria: 1. metastatic breast cancer with HER2-negative primary tumor tissue and/or biopsies from metastatic sites or locoregional recurrences2. evidence of ≥ 1 HER2-positive CTC3. ≥ 1 evaluable metastatic lesion according to RECIST4. Tumor evaluation within 6 weeks before randomization

Specific aims: Objective: The objective of the trial is to prove the clinical efficacy of lapatinib in patients with metastasizing breast cancer who exhibit HER2-positive circulating tumor cells (CTC) although the primary tumor tissue and/or biopsies from metastatic sites were investigated for HER2 status and showed HER2-negativity.

Primary endpoint: Progression free survival

Secondary endpoints: Overall response rateClinical benefit rateOverall survivalDynamic of CTCQuality of lifeSafety and tolerability of lapatinib

Statistical methods: The primary endpoint will be analyzed by Kaplan-Meier method using the logrank test in order to compare the PFS distributions of the two arms. Efficacy, toxicity and other event rates are calculated, providing confidence intervals. In case of comparison between patient groups, these rates will be analyzed by Fisher's exact test or test. The Kaplan Meier analysis for all event related data will be carried out overall for the whole patient population. Furthermore a Cox regression analysis will be done using the following covariates Hormone receptor status (positive/negative)Number of prior chemotherapy lines for metastatic diseasePrior endocrine therapy for metastatic diseaseEndocrine treatment vs. cytotoxic treatmentOne metastatic site vs. multiple metastatic sitesBone metastases vs. no bone involvementPerformance status ECOG Score (0/> 0)

Present accrual and target accrual: As only half of the patients with HER2-negative metastatic breast cancer show CTC-positivity and of those approximately 32% will exhibit HER2-positive CTC (Fehm et al. 2010), screening of about 1420 patients is required to enroll 228 patients. First patient was screened in February 2012. As of July 27th 2012 117 patients were screened and 21 were found to have HER2-positive CTCs

Citation Information: Cancer Res 2012;72(24 Suppl):Abstract nr OT1-1-10.

P4-07-06: Correlation of Two Analytical Methods for Circulating Tumor Cells in Peripheral Blood of Patients with Primary Breast Cancer

Background

While the evidence for circulating tumor cells (CTCs) as a prognostic marker in metastatic breast cancer has been well established, there is still a lack of data in primary disease. In the SUCCESS A trial two different techniques for the detection of CTCs in early breast cancer were prospectively evaluated.

Material and Methods: SUCCESS A compared FEC-Docetaxel vs. FEC-Docetaxel-Gemcitabine and 5 vs. 2 years of treatment with zoledronic acid in primary breast cancer patients and node positive or high-risk node negative disease. Two different techniques to detect CTCs were prospectively evaluated in two consecutive, but comparable subgroups of the whole study population.

In 3515 samples the CellSearch® System (Veridex, Warren, USA) was used for CTC detection. Immunomagnetic enrichment with an EPCAM-antibody was followed by labeling with monoclonal antibodies specific for cytokeratin (8, 18, 19) and leukocytes (CD45). 2165 samples were evaluated with a manual immunocytochemistry (MICC) protocol. Cytospins were prepared after mononuclear cell enrichment based on Oncoquick® centrifugation (greiner bio-one, Frickenhausen, Germany). Staining was performed with the monoclonal pancytokeratin antibody A45-B/B3 (Micromet, Munich, Germany) and the APAAP technique. Conventional light field microscopy (Axiophot; Zeiss, Oberkochen, Germany) was used for the detection of stained cells.

For both methods, the cut-off value for positivity was ≥ 1 CTC. All events were evaluated by two independent observers.

Results: CTCs were examined in a total number of 3243 patients before and after chemotherapy (CHT). The two subgroups evaluated with one or the other method were well-balanced regarding clinical parameters as tumor size, grading, lymph node-status, hormone receptors and Her2. Furthermore there was no significant correlation between the CTC positivity and one of these clinical parameters using CellSearch or the MICC, respectively (p > 0,05 using the chi square test each time).

Before adjuvant CHT 21. 3% (424 out of 1994) and 21.1 % (264 out of 1249) of the patients were found positive for CTCs using CellSearch® or the MICC respectively, with a mean CTC level of 5.9 (range: 1 to 827) and 3.1 (range: 1 to 256).

Immediately after CHT 21.9% (333 out of 1521) and 16.5% (151 out of 916) of the patients were positive for CTCs using CellSearch® or the MICC. The mean CTC level decreased to 3.0 (range: 1 to 124) and 2.1 (range: 1 to 23) in both analytical methods.

Using CellSearch® there was a significant correlation between the presence of CTCs before CHT and disease progression (p = 0.0044), as well as survival (p = 0.0001), whereas the MICC did not predict any of these (p = 0.3143 and p = 0.0801 respectively; the chi-square test was used each time).

Conclusion: We found comparable prevalence of CTCs before and after adjuvant chemotherapy both with the CellSearch® System or the MICC. However, prognostic relevance could only be shown for CTCs detected with the CellSearch® System. This may be attributed to the high standardization and reproducibility of the automated system, as well as the additional CD45 counterstaining. According to our findings, the FDA approved CellSearch® System should be used as gold standard for CTC detection in future clinical trials.

Citation Information: Cancer Res 2011;71(24 Suppl):Abstract nr P4-07-06.