Long-Term Breast Cancer Survival in Vietnamese Women

Abstract

To the editor:The breast cancer incidence in Vietnam was consid-erably lower than that in the USA (1). Data on sur-vival of breast cancer in Vietnam are scarce. Thecurrent study examined the correlation of survivalwith prognostic factors and treatment of Vietnamesebreast cancer patients in Hanoi, Vietnam.Study participants (248) with primary breast cancer,were recruited at the National Cancer Hospital, Hanoi,Vietnam between June 2002 and October 2003.Patients were treated with modified radically mastec-tomy or conservative surgery and axillary node sam-pling (median: 10 lymph nodes excised). Histology wasclassified following WHO criteria. Tumor was gradedfollowing Scarff-Bloom-Richardson (SBR) and Elston-Ellis methods (2). Informed consent was obtained fromall patients before treatment. Patients with modifiedradical mastectomy were treated with adjuvant radio-therapy (tumors ≥3 cm: 50 Gy to the chest wall; andnode positive: 50 Gy to the axillary area). Patients withpartial mastectomy were given 50 Gy to the entirebreast and a booster (60–65 Gy) to the tumor bed.Patients with lymph node metastasis were receivedadjuvant chemotherapy as anthracycline or taxane, 4–6cycles.Hormone receptor content of tumors were definedas positive if immunohistochemistry results >1%stained nuclei. Patients with positive tumors werereceived endocrine therapy. The tumors were also ana-lyzed for HER2 expression and rate of cell prolifera-tion using an automated immunostaining platformand SISH procedure (3). The tumors were divided intofour categories: (a) Luminal A [ER(+) or/and PgR(+)and HER2( )]; (b) luminal B [ER(+) or/and PgR(+)and HER2(+)]; (c) HER2-enriched tumors [ER( ),PgR( ) and HER2(+)]; and (d) triple negative [ER( ),PgR( )andHER2( )].Comparison of clinicopathological characteristicsbetween premenopausal and postmenopausal patientswas performed using chi-squared test and odds ratio(OR). Disease-free survival (DFS) was calculated fromoperation date to the date of first detection of metasta-sis or contralateral breast cancer. Cancer-specific sur-vival (CSS) was calculated from operation date to thedate of death caused by cancer. Overall survival (OS)was calculated from operation date to date of death ofany cause or the last day of follow-up (July 31, 2011).Survival rates were estimated using Kaplan—Meiermethod. Log-rank test was used to compare DFS andOS between groups. We censored those who were stillalive after the last day of follow-up. Cox proportionalhazards models were used to determine the relation-ship between breast cancer deaths and prognosticfactors and treatment.Median follow-up time was 99 months. Of 248patients, 159 patients were premenopausal, 85 post-menopausal, and four unknown menopausal statuses.During follow-up, 48 patients died from breast cancer.Table 1. Relationship between Clinical Character-istics and Survival in Univariate and MultivariateModels