The T and B cells repertoire diversity represent one of the immune defence level which controls the integrity of the organism and determines its ability to recognize and control infectious attacks and development of tumours. The study of the lymphocytes TCR and BCR diversity could permit to better understand how lymphopenia act on overall survival and to improve detection of high risk patients who could benefit of adapted therapies for better care.

The therapeutic management recommendations of patients with metastatic cancer offer standard treatment in specific situations. But, there is always a subgroup of patients who do not benefit from treatment and which has a very low survival. The risk of death for patients is very variable depending on the initial cancer site, tumor aggressiveness and chemosensitivity of tumours.

It's therefore important to have relevant prognostic tools to predict such an excess of relative toxicity or drug resistance. Simple prognostic factors for survival as the performance status (PS) have already been highlighted in several studies. Thus, the possibility to identify a group of patients with a higher risk of mortality could be of major interest for clinicians. In fact such stratification will allow:

• To limit this risk by adjusting the therapy and/or associated treatments (antibiotic prophylaxis, dose reduction ...),
• To develop protocols for testing innovative strategies specific to this high risk population.

The objectives of these innovative protocols would be designed to correct lymphodivpenia.

The main objective is to show that T divpenia (low TCR combinatorial diversity <30%) is a risk factor for early death after chemotherapy (early death: any death occurring within 3 months (lung cancer) or within 6 months (breast cancer) after the start of chemotherapy).

The secondary objectives are:

• To establish that the divpenia factor is independent of clinical and biological prognostic factors (PS, LDH, haemoglobin, lymphopenia or ANC) to predict a early death,
• To establish that the prognostic score NDL which will combine in a two-dimensional graph the CD4 count or total lymphocytes count and TCR repertoire diversity will allow a better stratification of lympho-divpenic patients who will benefit from more appropriate treatments,
• To characterize other circulating markers this could improve the identification of the early death risk (phenotypic markers, cytokines ...) in combination with the previous settings.

Prognostic models have been established in many tumor types at the initial stage or at time of the relapse (breast and lung cancers ...). It seems necessary to highlight other clinical and biological prognostic factors that would estimate a lower survival at 6 months, whatever the nature or the original site of the tumour.

The strong prognostic value of lymphopenia in the early death after chemotherapy or hematologic toxicity of chemotherapy has recently been established in several tumor models. 25% of patients with metastatic solid cancers have a systemic immune dysfunction.

Further analysis incorporating prior any treatment (D0) a systematic phenotypic analysis of lymphocyte subpopulations in the patients' blood showed that a significant reduction in the absolute number of peripheral CD4+ T cells (<450/μl) was an independent factor risk of early death and febrile neutropenia in patients with solid cancers of different origin (lymphoma, myeloma, sarcoma, breast cancer) treated with chemotherapy.

However, early death remains rare events and a more recent study has established that a lymphopenia (<1000 lymphocytes/µl) was an independent prognostic factor for overall survival in patients with solid tumours.

Recent studies have shown the importance of combinatorial diversity of T and B lymphocytes repertoire (TCR/BCR) in the efficiency of the immune response against infection.

A preliminary analysis of TCR repertoire diversity, in a retrospective cohort of patients with metastatic breast cancer, demonstrated the independent predictive value of divpenia for overall survival in these patients. This research has demonstrated that patients with cancer had a very large disparity in immune TCR diversity and that it was not always correlated with lymphocyte count.

These preliminary data show that the discriminating power of TCR diversity is greater than the measurement of cell count. The combined analysis of these data in a NDL® graph may help to better discriminate patients at risk for which it is necessary to develop new therapeutic strategies.

Inclusion Criteria:

• Age ≥ 18 years old,
• Patients with an histologically proven, inoperable breast or lung tumour,
• Metastatic disease before the start of any chemotherapy,
• Signed written informed consent form,
• Covered by a medical insurance,
• Patient accepting the conservation of biological samples,
• Locally advanced incurable disease (only breast tumour).

Exclusion Criteria:

• Hematological tumour,
• Auto-immune disease (including HIV-positive - AIDS stage) or patients with immunosuppressive therapy,
• Metastatic disease that had progressed after a first line chemotherapy,
• Pregnant or lactating female or female of child-bearing potential not employing adequate contraception,
• Patient deprived of liberty by a judicial or administrative,
• Adult protected by law.

2 cohorts of patients (breast and lung cancer), with locally advanced or metastatic cancer