Background: We aim to review different factors associated with the relapse and the mortality in breast cancer patients in restricted-resource settings over a five-year period. Method: A retrospective cohort study including 133 women underwent breast cancer surgery was performed at the Surgical Oncology Unit of the Conakry University Hospital in Guinea. Socio-demographical characteristics, clinical information, treatment and data on relapse and survival were retrieved from medical files. Predictor factors of relapse were analyzed by using univariate and multivariate logistic regression. Factors associated with mortality were analyzed by Kaplan Meier survival and Cox models. Results: Breast cancer surgery was conservative in 13 cases (9.8%) and radical in 120 cases (90.2%). Five-year relapse was 33.8% ± 8.0% and the predictor factors of this relapse were age and stage III (p = 0.005). The overall 5-year mortality was 42.1% and the risk factors independently associated with death were body mass index (p = 0.01), origin of patients (p = 0.02), radiotherapy (p = 0.01) and cancer relapse (p = 0.001). Conclusion: The proportions of relapse and death were higher in breast cancer patients. The age and the stage of cancer at the time of surgery were the predictor factors of relapse. Body mass index, origin of patients, radiotherapy and cancer relapse were associated with death.

Bilateral gigantomastia is a rare condition, often associated with pregnancy that is characterized by a diffuse enlargement of both breasts. Here we present a case of a late 20s woman in her seven months pregnancy with a bilateral gestational gigantomastia associated with multiple breast lumps. Histological analysis revealed a fibroadenoma. Her prolactin level after caesarean delivery was found particularly high. A significant decrease in breast size was achieved with bromocriptine treatment in conjunction with a bilateral lumpectomy. This case report highlights the diversity of gigantomastia and emphasizes the importance of a tailored, multidisciplinary approach to the diagnosis and treatment of this condition.

A 30-year-old HIV positive woman presented with a multifocal mass tumour associated with axillary and lateral-cervical lymphadenopathy in the right breast. Laboratory examination of the biopsy confirmed a case of mammary Burkitt's lymphoma with a nodular infiltration of the breast. Antiretroviral treatment and chemotherapy were effective to control the tumour. Although Burkitt's lymphoma rarely involves the breasts, it should be considered during routine breast examination of African woman.

393 Background: The clinical use of molecular expression profiles could result in more accurate and objective diagnoses of cancers as well as prognoses of disease or response to the treatment. Unsupervised hierarchical clustering (UHC) analysis is a common method to profile the molecular expression of tissue microarray data. Methods: TM4: a free, open-source system for microarray data management and analysis was used in order to identify expression patterns of interest in our cohort of renal cell carcinomas (RCC) (n=80). We investigated 5 pathological predictors: (a) the histological type, (b) Fuhrman grade, (c) depth of infiltration, (d) metastasis in lymph node, (e) TNM stages, and 27 immunohistochemical molecular predictors involved in different pathways of tumor development and progression including: p53 and VHL tumor suppressor proteins; Bcl-2 and Survivin antiapoptotic proteins; Hif1-alpha and Notch proteins as a transcription factors; EGFR, PDGFR-α and VEGF proteins involved in tumor growth and proliferation; Glut proteins involved in tumor cell metabolism. Results: Unfavorable prognosis was significantly correlated with pathological predictors. Clear RCC histological subtype had a worse prognosis than the other ones studied to be accompanied by increased expression of Hif1-alpha and CAIX (p<0.001 in both cases) in patients with a number of metastatic nodes weighed (p<0.001). We found in Fuhrman Grade III samples a higher expression of VEGF (p=0.029). UHC analysis was done using average method with Spearman rank test to group the data with different predictor profiles. A preliminary analysis by hierarchical clustering of RCC produced three separate clusters groups. Conclusions: UHC based on a extended immunoprofile might be a useful, promising and powerful tool for further translational studies and should lead us to define a diagnostic and prognostic signature for RCC. Our laboratory is currently involved in this issue. No significant financial relationships to disclose.

No abstract