GENERATING REAL WORLD DATA OF BREAST CANCER IN BRAZIL: VALIDATION OF AMAZONA III RESULTS WITH CANCER REGISTRIES

Introduction: AMAZONA III study (GBECAM 0115) is a prospective real world breast cancer (BC) cohort performed in Brazil that showed a high proportion of patients diagnosed with < 50 years and clinical stage II-III. Differences in terms of the stage at diagnosis were found between public and private health system. The national cancer registry has some limitations such as low population coverage, limited clinical-pathological information and unequal participation among public and private institutions. Objective: Validate data from AMAZONA III with the national cancer registry by an Oncology Structured Information System (OSIS). Methods: The AMAZONA III included women > 18 years with newly diagnosed between Jan 2016 and March 2018 in 22 sites. OSIS data was collected from Registro Hospitalar de Câncer (RHC) and DATASUS AQ (Chemotherapy Database) including women > 18 years with newly diagnosed BC from 77 sites in São Paulo State. We present a comparison of the patients baseline characteristics at diagnosis between AMAZONA III and OSIS. No formal statistical comparison was performed. Results: A total of 2950 patients from AMAZONA III and 10173 patients from OSIS were included in this analysis. Public health insurance coverage was 63.1% vs. 80.8% in AMAZONA III and OSIS, respectively. Mean age at diagnosis was 53.9 (SD 13.4) vs. 56.9 (SD 13.5) years old and age distribution was similar: < = 35 years (8.4% vs. 5.4%), 36-50 years (34.8% vs. 30.7%) and > 50 years (56.8% vs. 63.9%). No difference in age distribution by health insurance was found between the registries. Skin color was similar, white 58.6% vs. 53.6% and brown 34.4% vs. 39.2%. Histologically, 80.9% vs. 82.6% were invasive ductal carcinoma and 6.9% vs. 6.2% lobular, respectively. Clinical stage distribution was 26.3% vs. 25.2% stage I, 41.6% vs. 37% stage II, 27% vs. 27.5% stage III and 5.1% vs. 10.2% stage IV, respectively. In both databases, a high proportion of patients were diagnosed with stage III disease in public vs. private system. BC pathological characteristics were ER and/or PgR positive in 78% vs. 74.7%, HER2 positive in 23.4% vs. 11.6% and grade 2 tumors in 51% vs. 51.6%. Conclusion: Observational BC studies produce data similar to national cancer registries with the advantage of collecting more detailed information on BC treatment, relapse and survival endpoints.

 

MissingLinkText

No conflict of interest has been declared by the author(s).

Publication History

Article published online:
23 October 2019