Central and East European Oncology Group

LA-REMA - Retrospective analysis of molecular factors predictive for clinical efficacy of lapatinib in HER2-positive advanced breast cancer patients.

Study team

Study Coordinator: Renata Duchnowska, MD, PhD,
Military Institute of Medicine, Warsaw, Poland

Study Sponsor:

Central and East European Oncology Group
Chairman: Prof. Jacek Jassem


Introduction of HER2-targeted therapies has led to a major breakthrough in the treatment of patients with HER2-positive breast cancer. However, despite their impressive clinical efficacy, there are still many patients who are resistant to HER2-targeting agents. Several molecular resistance mechanisms to HER2-targeting agents (mainly to trastuzumab) have been extensively studied.

Lapatinib has entered clinical practice later than trastuzumab and its mechanism of action is much different than that of the HER2-specific monoclonal antibody. Therefore, the molecular mechanisms of resistance to lapatinib cannot be simply extrapolated from known trastuzumab resistance mechanisms. Resistance to lapatinib may be caused by various mechanisms occurring at various levels within a cancer cell: outer/inner leaflet of plasma membrane, cytoplasm or nucleus.

Therefore, we will analyze particular molecular factors (surface receptors, kinases involved in signal transduction, transcription factors) and retrospectively correlate their status with clinical efficacy of lapatinib measured as a progression-free survival (time from the start of lapatinib therapy to progression or death from any cause).

Inadequate activity of various signal transduction elements has been demonstrated to be associated with resistance to HER2-targeted agents (mostly trastuzumab). Normally, activation of growth factor-associated signaling cascades is initiated at the plasma membrane in response to receptor activation (homo, or heterodimerization). Subsequently, the signal is transmitted downstream toward the nucleus via a signaling network which comprises multiple kinases. In cancer cells signal transduction pathways may become activated regardless of the receptor status. Spontaneously activated signal transduction elements may be responsible for resistance to receptor-targeted therapies, since crucial pathways remain active despite receptor blockade. Therefore, it is important to analyze the signal transduction molecules whose activity may correlate with resistance to lapatinib by correlation with PFS.

The aim of our study is to shed new light on the molecular mechanism involved in susceptibility/resistance of breast cancer to lapatinib. Since we will use advanced, validated molecular technologies to analyze tissue samples from a well-defined, large and tightly controlled patient population, we are convinced that our results will be reliable and highly reproducible and may help to optimize lapatinib based therapies in breast cancer patients.

Expected outcome: Defining reliable predictive factors for lapatinib therapy should considerably increase the therapeutic efficacy of this compound and its cost-effectiveness measures. This should provide an incentive to use lapatinib therapy rather than other therapeutic options in second-line therapy of advanced HER2-postive breast cancer and in perspective will move this agent to first-line treatment.


Determination of molecular factors predictive for treatment efficacy of lapatinib (measured by PFS) in HER-2 positive advanced breast cancer patients

Study design

A multicenter, retrospective translational study including HER2 positive advanced breast cancer patients who were previously treated with lapatinib in combination with capecitabine. The majority of these patients have participated in a GSK Expanded Access Program (EAP). Patients must have received up to 3 lines of previous chemotherapy for advanced disease and lapatinib must have been administered for at least of 6 weeks. The translational analyses will be carried out on formalin-fixed, paraffin-embedded (FFPE) tumor specimens collected from all participating patients.


The study population will include approximately 250 HER2 positive advanced breast cancer patients.

Inclusion criteria

  • HER-2 positive advanced breast cancer patients (3+ protein expression and/or FISH-positive)
  • age >18 years
  • no previous or concomitant malignant disease except for basal cell carcinoma of the skin
  • treated with a combination of lapatinib plus capecitabine at standard doses
  • received up to 3 lines of previous chemotherapy for advanced disease
  • evaluable for therapeutic response to lapatinib (i.e. having evaluable disease and administered lapatinib for at least of 6 weeks)
  • for whom FFPE tumor tissue specimens are available
  • alive patients will have to provide written informed consent prior to inclusion into the study according to regulations in particular countries

Study procedures

The study protocol will have to be accepted by ethics committees in particular institutions according to local regulations and by regulatory bodies, wherever necessary. Due to retrospective character of this study, it is expected that most patients will be deceased at the time of molecular analysis.

The tissue specimens (archival paraffin embedded tumor blocks) will be obtained from cancer centers from selected European countries and the study will be conducted by CEEOG (Central and East European Oncology Group).

All data will be fully coded to secure full protection of personal information.

We will analyze particular molecular factors (surface receptors, kinases involved in signal transduction, transcription factors) and retrospectively correlate their status with clinical efficacy of lapatinib measured as a progression-free survival (time from the start of lapatinib therapy to progression or death from any cause).

In addition, a Cox regression analysis will be conducted which includes known predictors of PFS in this group of patients. This will allow for assessing the predictive value of the markers.

Pathological and molecular analyses

Analyses will be performed at three institutions:

Medical University of Gdansk, Poland

Monogram Biosciences Lab San Francisco, Ca, USA

University of Medical Sciences at Greater Poland Cancer Center, Poznan, POLAND.

Medical University of Gdansk, Poland will perform central pathology review including tumor type, grade and HER2 expression.

Laboratories of Monogram Biosciences will analyze several molecular factors located at the level of cell plasma membrane. The company will use a unique VeraTag assay technology for assessing the majority of the molecular analytes. Monogram Biosciences has developed the VeraTag™ Assay technology which is capable of monitoring the quantitative expression levels, protein-protein interactions (i.e., dimerization), and post-translational modifications (e.g. phosphorylation) of selected protein targets involved in cancer cell signaling with high sensitivity and specificity. The VeraTag™ assay platform is based on the proximity and light-dependent release of antibody-conjugated fluorescent reported tags, each of which has a unique migration property detected by capillary electrophoresis (CE). Detailed methodology and validation summaries for the VeraTag technology platform and the VeraTag HERmark® Breast Cancer Assay, respectively, have been published.1, 2 In addition to using the VeraTag technology, PI3Kca mutation analyses will be done using a pyrosequencing method or the DxS PI3K Mutation Test Kit, and HER2 FISH testing will be performed at CMBP using the Vysis PathVysion kit.

Monogram’s South San Francisco, California and/or its affiliate CMBP will perform analyses using following assays.

VeraTag Assays:

  1. HER2 Total (H2T) expression (HERmark, validated)

  2. HER2:HER2 homodimers (H2D) (HERmark, validated)

  3. Total HER 3 (ErbB3) expression (pending validation)

  4. Total HER1 (EGFR, ErbB1) expression (pending validation)

  5. HER2:HER3 heterodimers (advanced development)

  6. HER1:HER2 heterodimers (advanced development)

  7. IGF-1R (R&D)

  8. cMET (advanced development)_

Non-VeraTag assays:

  1. PI3Kca oncogene mutation analysis (CMBP or Monogram)

  2. HER2 gene copy number (FISH), serving as a reference assay (CMBP)

Specimen requirements for VeraTag, FISH, and mutation assays:

1. Formalin-fixed (10% neutral buffered formalin), paraffin-embedded specimens. Non-formalin fixed samples are not acceptable for the VeraTag assay.

2. Tissue sections should be prepared as 5-um sections, one section/slide (unless tumor area is < 10 mm2, and then multiple sections per slide may be prepared), on positively-charged glass slides.

3. Total number of slides needed/patient for the entire study: 15 slides plus one H&E per patient.

4. Freshly cut sections will be sent to Monogram Biosciences within 4 weeks of preparation. Unstained slides will be stored and shipped at 4 degrees Celsius. Monogram will provide shipping kits with ice packs (if necessary during summer months) and prepaid airbills.

5. Monogram will perform macrodissection on all tissue sections prior to performance of VeraTag assays.

7. The acceptance criteria for FFPE specimens in VeraTag assays:

  • Invasive breast carcinoma (areas of DCIS or LCIS will be excluded by macrodissection prior to VeraTag assay)

  • Total tumor area on glass slide > 10 mm2

  • Total % of viable tumor (vs. normal tissue, in situ carcinoma, and/or necrosis) > 60% (samples with less than 60% invasive tumor will be macrosdissected prior to assay in order to enrich for invasive tumor content).

University of Medical Sciences at Greater Poland Cancer Center, Poznan, Poland will perform semi-quantitative analyses of phosphorylated signal transduction elements according to standard, previously described procedures3-4 .

Formalin-fixed paraffin-embedded tissue samples will be stained with antibodies against PTEN, pP70S6K, HIF-2, pERK1/2, or AMPK. Sections of 5 mm will be deparaffinized in xylol and alcohol. Slides will be pretreated in citrate buffer, in distilled water, blocked with normal goat serum and then incubated with the primary antibody, overnight. Binding of the monoclonal antibody will be detected with biotin-labeled goat anti-mouse or anti-rabbit IgG and horseradish peroxidase-labeled avidin – biotin complex. Immunohistochemical staining will be scored using a scoring system that incorporates staining intensity and percentage of positive tumor cells5.

Following antibodies will be used for analysis of particular intracellular molecules

  1. PTEN - (28H6) sc-56205 (mouse IgG)

  2. pERK 1/2 (Thr202) sc-101760 (rabbit IgG)

  3. pERK 1/2 (Tyr204) sc-101761 (rabbit IgG)

  4. pAMPK alpha1 (THR172) sc-101630 (rabbit IgG)

  5. p-P70S6K alpha (A-6) sc-8416 (mouse IgG)

  6. HIF2 alpha ab8365 (mouse IgG)

Statistical analysis

Where possible, independent review of disease assessments will be utilized in order to determine PFS.

Tissue samples from primary tumor will be submitted for analysis of 16 biomarkers. We hypothesize that a strong trend for improvement in PFS will be present for patients with the highest level for a given marker. For example, patients with high expression (staining intensity of 2+ or 3+) will have approximately 1.5 to 2 fold difference in median PFS compared to patients with low expression (staining intensity of 0 or 1+).

Testing approximately 228 samples will be sufficient to detect a difference (3.0 months median PFS vs. 5.0 months median PFS) with a one-sided level of significance of 0.05 and power of 80% (Addplan 5.0). Assuming that 10% of patients will be ineligible, sample size has to be increased to 250.

Continuous data will be summarized using descriptive statistics (mean, standard deviation, median, inter-quartile range, range).

Categorical data will be summarized using frequency counts and percentages.

All p-values will be calculated as 2-sided. Progression-free survival was chosen as the clinical endpoint; the Cox proportional hazards model will be used to test for the effects of tumor baseline characteristics for 16 individual molecular assays, and relevant clinical characteristics on clinical outcome.

Multivariate analysis: A multivariate Cox proportional hazards regression model will include a number of baseline patient characteristics. These covariates (depending on availability) will include visceral disease at baseline, number of sites of disease, stage of disease, bone metastasis, lung metastasis, liver metastasis, time from diagnosis of metastatic disease to randomization, hormone receptor status, and Eastern Cooperative Oncology Group performance status. Univariate proportional hazards models will be used to screen the covariates and those significant at P < .20 will be included in a backward elimination multivariate Cox regression model according to a selection-stay criterion of P < .05. Baseline tumor lesion size will be forced into the multivariate model. In addition, biomarkers of specific interest can be forced it into the model to determine if they are significant in the presence of the covariates. Effects will be considered significant if P < .05; all P values will be 2-sided. Covariates or biomarkers where 20% of patients have missing data will be excluded from the multivariate analysis in order to maximize the number of patients included in the analysis.

In addition to the multivariate Cox model described above, dimerization status of the ERBB receptors will be tested in a separate model in which only homodimers and heterodimers, testing for coexpression (correlation) among all pairs of appropriate factors, will be presented.

Univariate biomarker analysis: Significant biomarker results based on the Cox model may be used to form strata (expression/non-expression) and the logrank test will be used to compare PFS across strata.

Exploratory Biomarker Analyses

Additional exploratory analyses may be conducted on any of the biomarker data described above. This may include other efficacy measures (response rate, overall survival, clinical benefit proportion), pattern of relapse and others.


  1. Shi Y, Huang W, Tan Y, et al: A Novel Proximity Assay for the Detection of Proteins and Protein Complexes: Quantitation of HER1 and HER2 Total Protein Expression and Homodimerization in Formalin-fixed, Paraffin-Embedded Cell Lines and Breast Cancer Tissue. Diagn Mol Pathol 18:11-21, 2009

  2. Larson JS GL, Tan, Y, Defazio-Eli L, Paquet AC, Cook JW, Rivera A, Strommen K, Bose J, Chen L, Cheung J, Shi Y, Irwin S, Kiss LDB, Huang W, Utter S, Sherwood T, Bates M, Weidler J, Gordon Parry: Analytical Validation of a Highly Sensitive, Accurate, and Reproducible Assay (HERmark®) for the Measurement of HER2 Total Protein and HER2 Homodimers in FFPE Breast Cancer Tumor Specimens. Pathology Research International: Article ID 814176, 14 pages. doi:10.4061/2010/814176, 2010.

  3. Montero C. The antigen-antibody reaction in immunohistochemistry. J Histochem Cytochem51:1-4, 2003

  4. O'Malley F and Pinder S, Breast Pathology, 1st. Ed. Elsevier 2006. ISBN 978-0-443-06680-1

  5. Walker RA. Quantification of immunohistochemistry - issues concerning methods, utility and semiquantitative assessment I. Histopathology 49:406-10, 2006



Medical University of Gdańsk
Department of Oncology and Radiotherapy
7 Dębinki Street, 80-211 Gdańsk, Poland
phone: +48 58 584 4560
e-mail: This email address is being protected from spambots. You need JavaScript enabled to view it.