BACKGROUND: Breast cancer is the commonest cancer amongst Malaysian women but local survival data are scarce. The present study was therefore conducted to assess overall survival and prognostic factors in Malaysian breast cancer patients. METHODS: The research sample was a prospective cohort of 413 patients diagnosed with breast cancer in the University of Malaya Medical Centre between 1993 to 1997. Survival data were obtained from the National Registry of Birth and Deaths in December 2000. The clinico-pathological variables studied were age, ethnic group, stage, tumour size, lymph node status, oestrogen receptor status and grade. The data were analysed utilizing Splus statistical software. The important prognostic factors were identified by fitting the Cox's proportional hazard model to the data set. Survival probabilities were estimated using the Kaplan-Meier method and differences were compared by the log-rank test. RESULTS: The overall 5-year survival was 59.1%. The Cox's proportional hazard model identified stage, lymph node status, size and grade as factors that correlated with prognosis. Age was not a significant prognostic factor. The Cox regression model by stepwise selection showed stage, nodal status and grade of tumour to be independent prognostic factors, whereas ethnicity, age and ER status were not. INTERPRETATION: The overall survival in our centre was low. Recognizing factors that affect prognosis of breast cancer patients in Malaysia may improve delivery of health care to at-risk groups by strategizing interventions as survival depends on early detection and effective treatment.
BACKGROUND:Breast cancer is the commonest cancer amongst Malaysian women but local survival data are scarce. The present study was therefore conducted to assess overall survival and prognostic factors in Malaysian breast cancerpatients. METHODS: The research sample was a prospective cohort of 413 patients diagnosed with breast cancer in the University of Malaya Medical Centre between 1993 to 1997. Survival data were obtained from the National Registry of Birth and Deaths in December 2000. The clinico-pathological variables studied were age, ethnic group, stage, tumour size, lymph node status, oestrogen receptor status and grade. The data were analysed utilizing Splus statistical software. The important prognostic factors were identified by fitting the Cox's proportional hazard model to the data set. Survival probabilities were estimated using the Kaplan-Meier method and differences were compared by the log-rank test. RESULTS: The overall 5-year survival was 59.1%. The Cox's proportional hazard model identified stage, lymph node status, size and grade as factors that correlated with prognosis. Age was not a significant prognostic factor. The Cox regression model by stepwise selection showed stage, nodal status and grade of tumour to be independent prognostic factors, whereas ethnicity, age and ER status were not. INTERPRETATION: The overall survival in our centre was low. Recognizing factors that affect prognosis of breast cancerpatients in Malaysia may improve delivery of health care to at-risk groups by strategizing interventions as survival depends on early detection and effective treatment.
Authors: Nirmala Bhoo-Pathy; Mikael Hartman; Cheng-Har Yip; Nakul Saxena; Nur Aishah Taib; Siew-Eng Lim; Philip Iau; Hans-Olov Adami; Awang M Bulgiba; Soo-Chin Lee; Helena M Verkooijen Journal: PLoS One Date: 2012-02-21 Impact factor: 3.240