ABSTRACT Objective: To evaluate the knowledge and practices of gynecologists, obstetricians, and family and community physicians in Brazil regarding breast cancer screening, mammographic findings defined by the BIRADS™ system, and their approach to suspected clinical lesions. Methods: This was an observational, cross-sectional, descriptive study conducted using an online research questionnaire distributed via email to 9,000 gynecologists and obstetricians and 5,600 family and community and preventive medicine doctors actively practicing in Brazil. Results: Among gynecologists and obstetricians, 42.8% follow the 40-74 years screening, 33.5% follow the 50-69 years screening, and 23.6% do not follow any specific protocol. Among the family and community physicians, 76.6% follow the 50-69 years screening protocol, and 23.4% do not follow any specific protocol. When we evaluated the responses regarding the behaviors of each BIRADS™ classification, 46.3% of responses were wrong among gynecologists and obstetricians, and 77.9% were wrong among community and preventive medicine doctors, exhibiting a significant difference. The role of breast ultrasound in screening was evaluated; 39.1% of gynecologists and obstetricians and 20.3% of community and preventive medicine doctors consider it as a screening method. Among gynecologists and obstetricians who do not follow any screening protocol, 94.7% consider ultrasound as a screening method. Among community and preventive medicine doctors, only 26.5% of physicians who follow the 50-69 years screening method consider it as a screening method. Conclusion: Inadequate training results in gynecologists and obstetricians, and family and community physicians performing inadequate screening and not following the recommended practices outlined in the BIRADS™ system. Objective BIRADS system lesions Methods observational crosssectional, crosssectional cross sectional, sectional cross-sectional 9000 9 000 9,00 5600 5 600 5,60 Results 428 42 8 42.8 4074 40 74 40-7 335 33 33.5 5069 50 69 50-6 236 23 6 23.6 protocol 766 76 76.6 234 4 23.4 classification 463 46 3 46.3 779 77 77.9 difference 391 39 1 39.1 203 20 20.3 947 94 7 94.7 265 26 26.5 Conclusion 900 00 9,0 560 60 5,6 42. 407 40- 33. 506 50- 2 23. 76. 46. 77. 39. 20. 94. 26. 90 0 9, 56 5,

OBJETIVOS: analisar a correlação entre o polimorfismo PROGINS e o câncer de mama. MÉTODOS: estudo caso-controle desenvolvido entre abril e outubro de 2004 com o pareamento de 50 mulheres com diagnóstico histopatológico de carcinoma de mama e 49 mulheres saudáveis. A inserção Alu de 306 pares de base no intron G do gene do receptor da progesterona denominada PROGINS foi detectada por meio de reação em cadeia da polimerase e analisada em gel de agarose 2% corado com brometo de etídio. Os grupos controle e experimental foram comparados, por meio de programa estatístico Epi-Info 6.0, quanto aos genótipos e às freqüências alélicas, utilizando-se o teste do chi2. RESULTADOS: em relação ao PROGINS encontramos uma prevalência na população estudada de 62 (62,6%) indivíduos homozigotos selvagens, 35 (35,3%) de heterozigotos e dois (2,1%) casos com a presença da mutação. Não foi evidenciada diferença significante em relação ao polimorfismo PROGINS, quando comparados os casos e controles, seja com relação à homozigose (62 vs 65,3%), heterozigose (36 vs 34,6%) ou à presença de mutação (2,0 vs 2,1%), com p de 0,920 (OR=1,01), 0,891 (OR=1,06) e 0,988 (OR=1,10), respectivamente. CONCLUSÕES: os resultados mostraram que o polimorfismo PROGINS não conferiu risco substancial de câncer de mama em seus portadores.

PURPOSE: to analyze the correlation between PROGINS polymorphism and breast cancer. METHODS: a case-control study was carried out from April to October 2004. The genotypes of 50 women with breast cancer and 49 healthy women were analyzed. The 306-base pair Alu insertion polymorphism in the G intron of progesterone receptor gene was detected by polymerase chain reaction and analyzed on 2% agarose gel stained with ethidium bromide. The control and experimental groups were compared regarding genotypes using the statistical Epi-Info 6.0 program and for frequencies the exact Fisher test or chi2 test were used. p value smaller p than 5% was considered to be significant. RESULTS: in relation to PROGINS we found in the studied population a prevalence of 62 (62.6%) wild homozygous, 35 (35.3%) heterozygous individuals and two (2.1%) cases with the presence of the mutation. Regarding PROGINS polymorphism, significance was not evidenced when cases and controls were compared, as related to homozygosis (62 vs 65.3%), heterozygosis (36 vs 34,6%) or the mutation (2.0 vs 2.1%), with p=0.920 (OR=1.01), 0.891 (OR=1.06), and 0.988 (OR=1.10), respectively. CONCLUSIONS: the results show that single-gene PROGINS polymorphism does not confer a substantial risk of breast cancer to its carriers.

Due to the conflicting results regarding the association between breast cancerand the GSTM1 null mutation, our aim was to research this associationin a Brazilian population and correlations withsmoking, reproductive history and several clinical pathologies. A case-control study was performed on 105 women with breast cancer and 278 controls. Extraction of DNA was accomplished according to the protocol of the GFX® kit and polymorphism analysis by the PCR technique. The control and experimental groups were compared and statistical analysis assessed by X² or Fisher's exact test. The deletion in the GSTM1 gene in the breast cancer group had a prevalence of 32 (30.4%) individuals with the presence of null mutation. In the control group, the null mutation was present in 104 (37.4%) women. Upon comparison of the two groups, no statistically significant difference of the GSTM1 gene was observed, with an odds ratio (OR) of 0.74, 95%, confidence interval (CI) 0.45 - 1.20, p = 0.277. The results conclusively show that singlegene GSTM1 polymorphisms do not confer a substantial risk of breastcancer to its carriers. Furthermore, in this study no correlation was found between GSTs andsmoking, reproductive history and several clinical pathologies with respect to cancer risk.