Introduction

Asia is the world’s largest and most populous continent, comprising over 60 % of the world’s population. Except for a few countries (Singapore, Taiwan, Hong Kong, Japan, South Korea, Israel, Saudi Arabia, and Macau) that are classified as high income countries, the rest of Asia includes low income and middle income countries [1, 2]. Over recent decades, South East Asia has seen large differences in socio-economic growth, leading to sharp contrasts in health systems developments between countries [2].

Compared to Western countries, where breast cancer incidence rates have stabilized or even decreased over the last two decades [3–5], most Asian countries have seen a rapid rise in breast cancer incidence [6–10]. With the Westernization of Asian countries, changes in dietary pattern, and increased exposure to environmental and reproductive risk factors among Asian women, it is quite likely that in the near future, the majority of breast cancer patients will be of Asian descent.

Singapore is a newly industrialized Asian country where approximately 75 % of the population is Chinese, 14 % is Malay, and 9 % is Indian [11]. Classified as a high income Asian country, Singapore sees a 95 % literacy rate and a life expectancy at birth of 81 years [12]. Rapid economic growth and low unemployment rates [12] have converted Singapore from a developing to a developed country within three decades [13], with rising standards of living and advanced healthcare facilities. Healthcare systems in Singapore have undergone major reforms from the early 1960s (when decentralization took place) to the early 1980s, where the National Health Plan outlined a 20 year plan to modernize healthcare facilities and raise medical standards [14]. Current healthcare provision in Singapore is considered at par with that from other developed countries [14].

Like Singapore, Malaysia also comprises three major ethnic groups—i.e., Malay (~54 %), Chinese (~26 %), and Indians (~8 %) [15], with a life expectancy at birth of 74 years [15]. An upper middle income country [16], Malaysia has seen sustained economic growth over the past few years, with an increasing proportion of people falling into the middle class category [17]. Although healthcare systems in Malaysia have undergone significant improvements over the last three decades, there are still gaps in terms of resource allocation, funding, and infrastructure that need to be filled before Malaysian healthcare can be considered at par with that from other developed countries [18].

Several studies in the West have shown that breast cancer occurs more frequently in developed countries and among women with a high socioeconomic status (SES) [19–23]. Incidence rates of breast cancer in Singapore (developed country) and Malaysia (less developed country) are 60.0 and 46.2 per 100,000 populations, respectively [24, 25]. Survival after breast cancer, on the other hand, is generally lower in low income countries and in women with a low SES or educational level [20, 26].

This study compares breast cancer presentation, treatment, and outcome of patients from these two neighboring countries in South East Asia with different levels of development.

Materials and methods

Data for this study were obtained from the Singapore Malaysia Breast Cancer Hospital Based Registry [27]. This registry combines data from two hospital-based registries, i.e., the National University Hospital (NUH) breast cancer registry (Singapore, high income country) and the University of Malaya Medical Center (UMMC) hospital-based registry (Kuala Lumpur, Malaysia, middle income country).

The NUH breast cancer registry was started in 1995 and contains information on 2,449 consecutive breast cancer patients diagnosed between 1990 and 2007. From the NUH registry we selected 2,141 patients diagnosed between 1993 and 2007. The UMMC breast cancer registry, started in 1993, contains information on 3,320 patients diagnosed between 1993 and 2007. Details on both these registries are described elsewhere [27, 28]. In both centers, patients were monitored through follow-up in the specialist outpatient clinics. Data on mortality were obtained from the hospitals’ medical records, as well as active follow-up through the patients’ next-of-kin. Follow-up for each patient was calculated from the date of diagnosis to the date of death or end of follow-up (July 2010 for NUH patients and November 2010 for UMMC patients).

For individual patients, the registry provides information on age at diagnosis, ethnicity (Chinese, Malay, Indian, and other), estrogen receptor (ER) and progesterone receptor (PR) status (if ≥10 % of epithelial tumor cells expressing receptors, negative, and unknown), stage (in situ, I, II, III, IV, and unknown), differentiation (good, moderate, poor, unknown), tumor size, nodal status (pN0, pN1, pN2, pN3, and unknown), regional nodes examined (0, 1–3, 4–9, and 10 or more). Treatment variables included type of surgery (mastectomy, breast-conserving surgery [BCS], and no surgery), radiotherapy (yes/no), chemotherapy (yes/no), hormone therapy (yes/no), and neoadjuvant chemotherapy (yes, no, and unknown).

Statistical analysis

Demographics, tumor characteristics, and treatment received by patients at the National University Hospital (Singapore) (n = 2,141) or the University of Malaya Medical Center (Kuala Lumpur, Malaysia) (n = 3,320) were compared by logistic regression analysis. Age at diagnosis and tumor size (as continuous variables) were presented as a median and compared with the Mann-Whitney U-test.

The proportion of patients receiving adequate (standard) treatment (defined as surgery for patients with stage in situ, I, II or III, chemotherapy for patients with ER negative lymph node positive invasive tumors, hormone therapy for patients with ER positive tumors, and radiotherapy for patients treated with breast-conserving surgery) were compared between the two institutions with the chi square test.

Kaplan-Meier analysis and the log rank test were used to compare overall survival between countries, and Cox regression analysis was used to estimate the adjusted relative risk of all cause mortality for patients treated in Malaysia as compared to those treated in Singapore. To gain insight into the factors contributing to survival disparities, we entered all variables univariately associated with survival into a multivariate Cox model in a stepwise manner. The first model consisted of crude hazard ratios (HR) representing the relative risk of death of patients from Malaysia as compared to those from Singapore. The second model presented HR adjusted for age at diagnosis, year of diagnosis, and ethnicity. The next model was additionally adjusted for tumor characteristics (i.e., tumor size, grade, nodal status, and ER status) and the final model was additionally adjusted for type of surgery type, radiotherapy, chemotherapy, and hormone therapy.

All analyses were performed with SPSS version 16, and a p value <0.05 was considered significant.

Results

The median follow-up for the Malaysian and Singaporean patients was 5.1 and 6.1 years, respectively. Malaysian and Singaporean patients presented at similar ages (median age 50 years for both countries). Malaysian patients were less likely to be diagnosed with in situ breast cancer than patients from Singapore (adjusted odds ratio (ORadj) 0.2; 95 % CI 0.1–0.3) and more likely to be diagnosed with advanced disease [(22.3 versus 14.4 %, respectively, for stage III; ORadj 1.6; 95 % CI 1.3–2.0); 10.8 versus 7.9 %, respectively, for stage IV; ORadj 1.2; 95 % CI 1.1–1.4)] as compared to Singaporean patients (Table 1). The tumor size of Malaysian patients was larger than that of Singaporean patients (median tumor size 30 mm compared to 22 mm; p <0.001). Malaysian patients were more likely not to undergo surgery for stage I–III disease (9.0 versus 0.6 %, respectively; p value <0.001) (Table 2). Malaysian patients with invasive, non metastatic disease were less likely to receive radiotherapy (RT) following BCS as compared to the Singaporean patients (78.0 versus 89.8 %, respectively; p value <0.001) (Fig. 1). Malaysian women were just as likely to receive chemotherapy for estrogen receptor (ER) negative lymph node (LN) positive disease (87.6 % compared to 90.1 %; p value >0.05) and hormone therapy for ER positive disease (91.2 % compared to 89.1 %; p value >0.05) as the Singaporean patients.

Table 1 Patient and tumor characteristics by place of diagnosis and the likelihood of these characteristics being associated with being diagnosed in Malaysia as determined by logistic regression
Full size table
Table 2 Treatment administered to nonmetastatic invasive breast cancer patients from Malaysia and Singapore, and the likelihood of treatment being associated with being diagnosed in Malaysia as determined by logistic regression
Full size table
Fig. 1
figure 1

Country stratified differences in proportion of stage in situ, I, II, and III patients receiving surgery; ER negative LN positive patients receiving chemotherapy; ER positive patients receiving hormone therapy; and patients receiving BCS followed by radiotherapy (excluding metastatic cases and cases with unknown stage)

Full size image

A total of 209 (10.8 %) Singaporean patients and 610 (18.9 %) Malaysian patients with invasive breast cancer received incomplete locoregional treatment defined as no surgery or BCS without RT or ER negative LN positive without chemotherapy, or ER positive without hormone therapy.

The 5 year overall survival for Malaysian patients was substantially lower than that of Singaporean patients (69.0 % compared to 80.0 %; log rank test p <0.001) (Fig. 2). Overall survival estimates for both countries improved with calendar time, with the improvement in survival being stronger for Malaysia (5 year survival estimates for Malaysians diagnosed between 1993–2000 and 2001–2007 were 62.0 and 73.0 %, respectively, while for the Singaporeans, estimates were 79.0 and 81.0 %, respectively; Table 3).

Fig. 2
figure 2

Life table estimates for Malaysia and Singapore (excluding in situ patients)

Full size image
Table 3 Five-year overall survival estimates for Malaysia and Singapore (excluding in situ patients)
Full size table

Univariate Cox regression analysis showed that besides country of diagnosis (i.e., Singapore or Malaysia), age at diagnosis, period of diagnosis, ethnicity, ER status, PR status, type of surgery, radiotherapy, chemotherapy, hormone therapy, regional nodes examined, nodal status, cell differentiation (grade), tumor size, and receipt of neoadjuvant chemotherapy were significantly associated with risk of all cause mortality. Multivariate Cox regression analysis showed that country of diagnosis remained independently and significantly associated with survival, even after adjusting for tumor characteristics and treatment in a stepwise manner (Table 4), with patients diagnosed and treated in Malaysia having a 67 % higher mortality risk than patients diagnosed in Singapore (adjusted hazard ratio [HRadj] 1.67, 95 % CI 1.44–1.92) (Table 5). Patients diagnosed in both countries receiving incomplete locoregional treatment or no surgery for invasive disease had a similar risk of death, whereas Malaysian patients receiving chemotherapy or presenting with node negative disease had a significantly higher risk of death than their Singaporean counterparts (Table 6).

Table 4 Stepwise modeling for Cox regression analysis
Full size table
Table 5 Cox regression models for all-cause mortality for invasive breast cancer patients
Full size table
Table 6 Subgroup analysis—multivariate Cox regression models for all cause mortality (excluding in situ patients)
Full size table

Discussion

This study highlights important differences in survival between breast cancer patients from tertiary hospitals in Singapore (high income country) and Malaysia (middle income country). Despite only small differences in way of presentation and access to treatment, Malaysian patients were more than 60 % more likely to die within the first 5 years after diagnosis than the Singaporean patients. This increased risk was not explained by more advanced staging and less optimal treatment.

Breast cancer survival disparities between countries have been well documented, and studies have shown that patients from countries with enhanced diagnostic facilities and up-to-date treatment options have better survival rates [29–34]. Although incidence rates of breast cancer are lower in middle income countries than in high income countries, 55 % of breast cancer deaths occur in low income countries, a finding that can be attributed to two major determinants: namely, late stage at presentation and inadequate treatment [35, 36]. In a comparative study of 12 countries in Africa, Asia, and Central America, differences in cancer outcome correlated with level of development of health services [37].

Differences in presentation between Singaporean and Malaysian patients could be a result of a higher level of health systems development in Singapore, where screening is more commonplace and diagnostic and healthcare facilities are advanced as compared to Malaysia. However, social and cultural factors are likely to play a role as well. In Malaysia, factors like lower awareness about the disease and or reluctance to approach physicians due to cultural taboos are more prevalent [25, 38].

As for presentation, we found only small differences in treatment patterns between patients from the two countries. Singaporean patients were more likely to receive standard treatment (radiotherapy in cases of treatment with BCS, and surgery for nonmetastatic disease) as compared to the Malaysians. However, the differences were small.

Large differences in breast cancer survival rates between Singaporean and Malaysian patients highlight a scope for improvement in the management of breast cancer in Malaysia. Several factors, such as differences in population structure (life expectancy) [39, 40], low access to screening [41, 42], lower socio-economic status [26, 43], low access to high-quality healthcare [25], poor treatment compliance [44–46], poor lifestyle after diagnosis [47], and lack of knowledge of the disease and its outcome [48] among Malaysians could explain the disparities in survival between patients from Malaysia and Singapore. Several studies have shown that type of treatment received is associated with breast cancer survival [49–52], but from our study, it is unlikely that differences in treatment would explain the overall survival differences as the stepwise adjusted hazard ratios (adjusting for demographic characteristics followed by adding tumor characteristics and finally treatment to the Cox model) did not differ significantly from the unadjusted HR.

We acknowledge that our study suffers from several shortcomings, including a relatively short follow-up time for patients from both countries. In addition, we assessed all cause mortality as our endpoint, as no data on cause of death were available. Third, because this is a hospital-based study rather than a population-based study, extrapolating these findings to the general population of the respective countries might not be feasible. However, the catchment area of NUH, Singapore, which treats an estimated 10 % of breast cancer cases in Singapore, sees patients with demographics that are not different from other areas of the country [28]. The UMMC in Kuala Lumpur, Malaysia, serves a predominantly middle income urban population, and hence our findings may not necessarily reflect the overall situation of breast cancer in Malaysia [27]. The presentation of breast cancer in the rural Malaysian settings for instance, may be more advanced than in our study [27]. Another limitation of our study is that some prognostic factors, such as co-morbidity, body mass index, HER2/neu status, and local/systemic recurrence were largely missing and hence their impact on our results will be difficult to gauge.

Conclusions

Differences in way of presentation (except stage and tumor size) and treatment of breast cancer patients from Singapore and Malaysia are small. Patients from Malaysia present slightly more often with advanced stage disease and unfavorable characteristics. The overall survival of breast cancer patients from Malaysia is much lower than that of Singaporean patients. Poorer compliance with treatment, unfavorable life style factors, and competing risks could explain the higher mortality risk of Malaysian breast cancer patients.