The European Organization for Research and Treatment of Cancer (EORTC) Breast Cancer Group is currently running eight studies: four randomised phase III trials, three phase II trials and one survey. In 2003, the Breast Cancer Group entered 1003 patients in these studies via 87 active participating centres. Table 1 presents a summary of the current open studies. The two most important studies will be reviewed here.

Another important trial that is in preparation and will be open in the near future is the MIND-ACT trial: a prospective, randomised study in node-negative breast cancer patients with two aims. The first is to test the performance of the gene prognosis signature (70-gene classifier) in comparison with the common clinic–pathologic criteria as a new prognostic tool that will reduce the risk of overtreatment with the correct identification of patients who need adjuvant chemotherapy. The second aim is to evaluate a docetaxel + capecitabine (nonanthracycline-based) regimen as a superior alternative to an anthracycline-based regimen, for high-risk node-negative breast cancer.

Until a few years ago an important part in staging breast cancer patients was the axillary lymph node dissection (ALND). Evidence has emerged that the sentinel node biopsy procedure by lymphatic mapping is an equivalent staging procedure to ALND 123. The introduction of a new surgical technique requires a process of quality-controlled implementation (preferably by treating patients in prospective studies) in order not to jeopardise, but to improve the level of care gained by the ALND, as has been achieved by the implementation of breast-conserving surgery 4. Only in this way can the questions regarding the importance of different technical aspects and the questions on the clinical implications be answered.

The EORTC 10981-22023 AMAROS trial is a phase III study comparing a complete ALND with radiotherapy to the axilla in sentinel node-positive patients, whereas sentinel node-negative patients are followed for the endpoints of the study as well (Fig. 1). Patients enrolled on the trial must have an operable invasive breast cancer greater than 5 mm and less than 3 cm, without clinically suspect regional lymph nodes. This intergroup study, which first enrolled patients in February 2001, is coordinated by the EORTC Breast Cancer Group in collaboration with the EORTC Radiotherapy Group, the BOOG (Breast Cancer Research Group) from The Netherlands and, in the near future, with former Axillary Lymphatic Mapping Against Nodular Axillary Clearance centres from the UK.

The main objective of the trial is to prove equivalent local/regional control for patients with proven axillary lymph node metastasis by sentinel node biopsy if treated with axillary radiotherapy instead of axillary lymph node dissection, with reduced morbidity. A second objective is to investigate whether adequate axillary control can be obtained by not subjecting patients with a negative sentinel lymph node to ALND.

As of 1 May, 18 institutes have included 1173 patients. It is assumed that the axillary recurrence-free rate in the ALND treatment group at 5 years equals 98%, and the aim of this study is to show that the axillary recurrence-free rate in the radiotherapy group at 5 years is not less than 96%. Given the ratio of 40 positive patients/60 negative patients, a total of 3485 patients are required to complete this study. The ratio to date of 35/63 (taking into account a 2% nonidentification rate) leads to a slight adjustment of the total number of patients needed. The study will at least be continued up to 2006.

In order to ensure that the trial results are not compromised by the use of improper techniques, or by a large variation in the quality of the treatment techniques used, the protocol provides guidelines for the sentinel node biopsy procedure, the pathology, the surgery and the radiotherapy. Quality assurance procedures were initiated to guarantee a uniform quality of treatment between institutions. The surgical quality control consists of quality control before starting to include patients in the trial, regular site visits and axillary failure analysis by the Independent Data Monitoring Committee.

Before a centre is allowed to include patients in the AMAROS trial, the quality of the sentinel node procedure is checked. This requirement is based on the existence of a learning curve 567. The surgeon and nuclear physician conducting the sentinel node procedure should have followed a training course. The trial protocol demands a so-called combined technique, using lymphoscintigraphy, preoperative blue dye injection and peroperative use of a gamma probe. At least 30 sentinel node procedures followed by an ALND of at least level I and level II have to be performed, with a minimum of 27 patients with accurate sentinel node identification. Not more than one false negative should be encountered. Besides inspection of the learning curve by an onsite review of 30 original patient files, the surgical quality control from the start consists of witnessing a lymphoscintigraphy and sentinel node biopsy procedure. So far 98% of the sentinel nodes have been identified, suggesting that the quality control is very effective. In addition, 63% of the sentinel node biopsies resulted in a negative sentinel node, which corroborates the superfluity of following axillary staging operations in the past.

The EORTC Radiotherapy Group has a strong tradition of institution-bound quality assurance 8910. Approval of the dummy run by the Radiotherapy Quality Assurance Committee is one of the prerequisites in the AMAROS trial. A recent study of the compliance of participating institutes to the radiotherapy guidelines of the AMAROS dummy run protocol resulted in a number of potential protocol deviations found at first assessment 11. Since recommendations led to a large number of adaptations by the participants, a considerable improvement in protocol compliance and inter-institutional consistency was achieved.

The evaluation of disseminated breast tumour cells in sentinel lymph nodes by quantitative real-time PCR and the comparison of the sensitivity of this assay with the routine histological analysis is part of the translational research 12. The quantitative real-time PCR assay with multiple mRNA markers for the detection of disseminated breast cancer cells in sentinel nodes resulted in an upstaging of sentinel nodes containing metastastic disease of 10% compared with the routine histological analysis. The application of this technique may be of clinical relevance, as it is suggested that micrometastatic disease in sentinel nodes is associated with further nodal nonsentinel node metastases in breast cancer.

More detailed information is available online http://www.amaros.nl.

The p53 study is the first prospective intergroup translational research trial assessing the potential predictive value of p53 using a functional assay in yeast in patients with locally advanced/inflammatory or large operable breast cancer prospectively randomised to a taxane regimen versus a nontaxane regimen. The EORTC Breast Cancer Group, the sponsor of the trial, collaborates in this trial with the Anglo-Celtic group, the SAKK (Swiss Group for Clinical Research) group and the Swedish group. Free drug supply (docetaxel) is offered by the pharmaceutical industry partner Aventis. Furthermore, the trial is supported by two grants from Aventis and Pfizer. Figure 2 depicts the scheme of the p53 study.

Factors predictive of the response to chemotherapy are virtually nonexistent. Experimental and clinical studies have shown that anticancer agents achieve their cytotoxic effect through apoptosis 13 and that p53 is a key gene in the apoptotic pathway. In vivo and in vitro data suggest that p53-mutated tumours are resistant to anthracyclines but remain sensitive to taxanes 14151617181920. Limited clinical data suggest that this finding may be true in patients with breast cancer 21. Unfortunately the immunohistochemistry method is not the best method to assess p53. There is a risk of false-negative and false-positive results when p53 is assessed by the immunohistochemistry method as compared with the sequencing method 22. A functional assay for p53 in yeast gives a direct information regarding p53 function (detects biologically important mutations) and is more sensitive than sequencing (it is insensitive to contamination of samples with normal tissue) 23.

In this study we are comparing prospectively a nontaxane anthracycline-based regimen versus a taxane regimen. We are using the p53 functional test in yeast. We are performing this trial in the neoadjuvant setting, which allows one to assess clinical and pathological responses after chemotherapy. Patients with either large operable or locally advanced cancers or inflammatory cancers are eligible.

The primary endpoints consist of comparison of the progression-free survival (PFS) of patients who received neoadjuvant chemotherapy without taxanes (treatment A) with the PFS of patients who received a chemotherapy with taxanes (treatment B), in both the normal p53 subgroup and in the p53-mutated subgroup. In addition we will compare the PFS of patients who received treatment A with the PFS of patients who received treatment B independent of the p53 status. The secondary endpoints are distant metastasis-free survival, overall survival and toxicity. In total, 1440 patients will be included in this study (including 10% of ineligible or nonassessable cases). We have already included 960 patients.

An intermediate analysis is planned when all the patients will be included in the study (number of expected events, 158). This analysis will provide some information concerning the p53-mutated subgroup (80% power to see an increase in the PFS of 30% in arm A to 50% in arm B; hazard ratio, 1.76; alpha, 0.008), but it will be too soon to obtain on the normal p53 subgroup.

If the recruitment in this study is performed over a period of 3 years, three additional years will be necessary before the final analysis. There are no early stopping rules in this trial.

The primary endpoint of this trial is a translational research question: do p53-mutated tumours respond better to taxanes than p53 wild-type tumours? A frozen tumour sample is mandatory before entering a patient in this trial. We will therefore have ideal conditions to perform translational research studies. For example, we are trying to identify a cDNA microarray profile that can predict for a complete pathological response after treatment A (anthracycline based) or treatment B (taxane based).

More detailed information is available online http://www.eortc.be.

None declared.

AMAROS = After Mapping of the Axilla: Radiotherapy or Surgery; ANLD = axillary lymph node dissection; EORTC = European Organization for Research and Treatment of Cancer; MIND-ACT = Microarray in Node Negative Disease and Anthracycline Chemotherapy or Taxane; PCR = poly-merase chain reaction; PFS = progression-free survival.