Incidence, Mortality and Survival Analysis of Epithelial Ovarian Cancer in Brunei Darussalam

Background: This study provides population-based study of cancer incidence, mortality and survival rates for women diagnosed with epithelial ovarian cancer (EOC), and evaluate the prognostic factors of EOC patients survival in Brunei Darussalam. Methods: This is a retrospective study of patients diagnosed with EOC between 1st January 2007 and 31st December 2017 in Brunei Darussalam. Crude, age-specific, age-standardized incidence and mortality rates per 100,000 women were calculated. Kaplan-Meier method was used to determine the overall 5-years survival rate. Log-rank test was used to examine the differences in survival between groups. The multivariable Cox Proportional Hazard regression models were used to estimate the hazard ratio for overall survival and to identify the prognostic factors. Results: A total of 207 patients were included in the study. The crude incidence and mortality rates were 9.7 and 3.6 per 100,000 respectively while the age-standardized incidence and mortality rates were 11.3 (95% CI: 9.7,12.9) and 4.5 (95% CI: 3.4,5.6) per 100,000 respectively in the period 2007-2017. The overall mean age at diagnosis was 48.4 (standard deviation=15.3) years. The overall survival rates at 1, 3, and 5 years for EOC patients were 79.7%, 69.7%, and 61.4% respectively. Age at diagnosis, cancer stage, and histology were significant prognostic factors for patients’ survival. Older age at diagnosis (≥70 years vs <40 years), regional or advanced stage (vs localized stage) and having undifferentiated or other epithelial ovarian (vs serous carcinoma) were associated with having higher hazard of death. Conclusion: Early detection of disease should be emphasized through public education and raising awareness to improve survival rates of patients with EOC.


Introduction
Ovarian cancer (OC) is the eighth most commonly diagnosed cancer in female and the eighteenth most common cancer worldwide (Sung et al., 2021;WHO, 2020a). It is one of the most common gynecologic cancers that ranks third after cervical and uterine cancer (Sung et al., 2021;Bray et al., 2018) and is associated with having the worst prognosis and the highest mortality rate (Coburn et al., 2017). Epithelial ovarian cancer (EOC) is the most lethal gynecologic cancer affecting women globally (Ferlay et al., 2015;Kurosaki et al., 2016) and accounts for 90% of all ovarian malignancies (Torre et al., 2018). 2020 (54.4%) while Europe came second with 21.3% (WHO, 2020a). In Brunei Darussalam, OC was the fifth most common cancer among females (DPP, 2018). Like many cancers, the incidence of OC varies across the world. GLOBOCAN 2020 estimates of age-standardized (world) incidence rates and mortality rates of OC in Brunei Darussalam were 17.4 per 100,000 and 7.4 per 100,000 respectively (WHO, 2020b).

Incidence, Mortality and Survival Analysis of Epithelial Ovarian Cancer in Brunei Darussalam
Brunei Darussalam has an estimated population of 453,600 in 2020. The population consists of Malay (65.8%), Chinese (10.2%) and other ethnicities (24.0%). There are four districts namely Brunei Muara (69.7%), Tutong (11.4%), Belait (16.5%) and Temburong (2.5%) (DEPS, 2019). Health services are highly accessible and provided free to citizens and permanent residents of the country, including provision for overseas specialized medical care not available in the country. The country implemented an electronic medical record system since September 2012. It provides an efficient, effective healthcare information and management system, and enhance the accuracy, timeliness, and completeness of the country's national cancer registry .
This study aims to provide population-based cancer incidence, mortality and survival rates for women diagnosed with EOC in 2007-2017, and evaluate the prognostic factors of EOC patients survival in Brunei Darussalam.

Materials and Methods
This is a retrospective population-based study using de-identified data extracted from the Brunei Darussalam Cancer Registry (BDCR) of Ministry of Health, Brunei Darussalam. Female citizens and permanent residents who were diagnosed with EOC between 1 January 2007 and 31 December 2017 were included in the study. All cases were followed until 31 December 2018. Age at diagnosis was classified as <40, 40-49, 50-59, 60-69, ≥ 70 years old, and ethnicity was stratified into Malay, Chinese and Others. District was classified as Brunei Muara, Tutong, Belait and Temburong. Cancer staging information at time of diagnosis reported in BDCR was based on the Surveilance, Epidemiology, and End Results (SEER) summary staging, which categorizes OC spread from its origin into localized (SEER stage 1 equivalent to TNM Stage I and IIa: T1-T3/N0/M0), regional (SEER Stage 2-5 equivalent to TNM Stage IIa and IIb and III: T3-T4/AnyN/ AnyT/N1,2/M0) and distant (SEER stage 7 equivalent to TNM Stage IV: AnyT/AnyM/M1) (Young et al., 2001).
Crude rate and age-specific rate were calculated using the number of new cases/deaths divided by Brunei female population during the same period (DOS, 2020). The annual age-standardized incidence rates (ASIR) and mortality rates (ASMR) were standardized by the direct method using the World Health Organization (WHO) world standard population distribution (Ahmad et al., 2000). Patients were categorized into two time-periods, 2007-2011 and 2012-2017, to compare the change in the crude, age-specific and age-standardized incidence and mortality rates over time.
Descriptive statistics analyses were conducted to evaluate the distribution of each variable, followed by the Fisher's Exact test to analyze associations between cancer stages and other variables. Kaplan-Meier survival analysis method was used to determine the overall 5-years survival rate and the median survival time of EOC patients. Overall survival was defined as the period of time from diagnosis to death or end of follow-up, due to any cause. Patients who were still alive or lost to follow-up at the end of the study period were right-censored. Log-rank test was used to examine the differences in survival between groups. Multivariable Cox Proportional Hazard (PH) regression models were used to estimate the hazard ratio (HRs) for overall survival and to identify the prognostic factor of EOC patients. We evaluated the PH assumption over time for these models using Schoenfeld residuals. Adjusted HRs and 95% confidence intervals (CI) were generated. All statistical analyses were performed using the R statistical software (Version 4.0.2). Ethical

Results
A total of 207 patients diagnosed with EOC were included in this study. The highest number of cases were recorded in year 2013 and 2014 (n=24 cases, 11.6%) while the lowest was in year 2007 (n=9 cases, 4.3%), with an average of 18 cases per year. There were 77 deaths in total during the study period and a total of 73 cases had died within 5 years of diagnosis.
The crude incidence rate of EOC was 9.7 per 100,000 per year while crude mortality rate was 3.6 per 100,000 in the period 2007 to 2017 (Table 1). Crude incidence rate was 8.7 per 100,000 in the first period (2007)(2008)(2009)(2010)(2011) and increased to 10.5 per 100,000 in the second period (2012-2017). However, crude mortality rate was more than double in the second period (4.9 per 100,000) as compared to the first period (1.8 per 100,000). By age group, the age-specific incidence rate was lowest in patients below 40 years old (3.8 per 100,000) and highest at 34.4 per 100,000 for 70 years and above age group ( Figure 1). The same trend can be seen for the age-specific mortality rates where patients below 40 years old recorded the lowest rate (0.3 per 100,000), followed by 40-49 age group (5.2 per 100,000), 60-69 years (13.1 per 100,000), 50-59 years (17.5 per 100,000) and 70 years and above (19.9 per 100,000).
Patients' demographic and clinical characteristics are presented in Table 2. The overall mean age at diagnosis was 48.4 (standard deviation=15.3) years, with the majority of the patients in the 50-59 age group (29.0%), followed by less than 40 years (27.5%), 40-49 years old (22.2%), 60-69 years old (12.1%) and 70 years old and above (9.2%). Majority of the patients were of Malay stage (38.9%), followed by distant (33.7%), and regional stages (27.4%). The most common histological type was mucinous carcinoma (30.9%), serous carcinoma (28.0%), undifferentiated or other epithelial (25.1%), endometrioid carcinoma (10.6%) and clear cell carcinoma (5.3%). ethnicity (83.6%) while 10.6% and 5.8% were of Chinese and other ethnicities respectively. Most EOC patients resided in Brunei Muara district (68.2%), followed by Tutong (16.2%), Belait (13.6%) and Temburong (2.0%). Majority of the patients were diagnosed at localized    Table 3 shows the number of cases for EOC patients according to cancer stage. No significant differences in the proportion of districts (p=0.350) and ethnicities (p=0.443) were found between stages. More than half of the women aged 70 and above were diagnosed at distant stage (52.9%) while only 11.8% were diagnosed at localized stage. Approximately 62.5% of women aged 60-69 years were diagnosed at distant stage. However, most of the women aged 40 and below were diagnosed at localized stage (69.8%), followed by regional (17.0%)   Figure 2). However, the median survival time was not available as the estimated survival were above 0.5.    (Figure 3).
Multivariable Cox PH regression analyses found that age at diagnosis, cancer stage and histology were the significant prognostic factors for EOC patients' survival (p<0.05), shown in Table 4. No evidence of significant violations of proportional hazard assumptions were observed.
As compared to a patient below 40 years old, the expected hazard of death was highest for EOC patient in the oldest age group (≥70 years) [Adj.HR=7.21; p=0.001], followed by 50-59 years [Adj.HR=5.29; p=0.001] whereas the expected hazard of patients in the 40-49 and 60-69 age groups were found to be comparable to those below 40 years old (p>0.05). Patients with distant stage have significantly higher hazard [Adj.HR=11.99; p<0.001] compared to patients with localized stage. This trend is also observed in patients with regional stage [Adj. HR=6.12; p<0.001]. Patients with undifferentiated or other epithelial have significantly higher hazard of death [Adj.HR=3.12; p=0.001] compared to patients with serous carcinoma.

Discussion
Global ASIR of OC in year 2012, 2018 and 2020 were 6.1, 6.6 and 6.6 per 100,000 women respectively (Sung et al., 2021;Bray et al., 2018;Ferlay et al., 2015). Brunei Darussalam recorded the third highest standardized incidence rate of OC (8.8 per 100,000) among Asian countries in year 2012 (Razi et al., 2016). Adjusting crude EOC incidence rate (9.7 per 100,000) for WHO world standard population in this study revealed an ASIR of 11.3 per 100,000. Our rates were considerably higher compared to the world's average ASIR of all ovarian cancers. Past study has found a significant positive correlation between the Human Development Index (HDI) and the standardized incidence rate of OC, which means that incidence was higher among high HDI countries (Razi et al., 2016). Brunei Darussalam's HDI was 0.838 in year 2019 which put the country in the very high human development country (UNDP, 2020). OC incidence pattern may be explained by changes in risk factors related to lifestyle, population growth, tube ligation, decreased pregnancy, lactation, obesity, and diet (Razi et al., 2016;Rohani-Rasaf et al., 2013). Increasing spread of western diet and lifestyle in this country might also be the cause of the rapid increase in the burden of the disease.
The world's average ASMR of OC in 2012 was 3.8 (per 100,000) and the rate increased to 3.9 (per 100,000) in 2018 and 4.2 (per 100,000) in 2020 (Sung et al., 2021;Bray et al., 2018;Ferlay et al., 2012). In this study, our crude EOC mortality was 3.6 (per 100,000) but ASMR was 4.5 (per 100,000), when standardized using WHO world standard population. Our study showed a decrease in ASIR trend of EOC between the two time periods (11.6 vs 11.1 per 100,000). Comparison of the age-specific incidence rates between the two time periods showed incidence in women aged 70 years and above was much higher in the first period (57.0 vs 20.4 per 100,000), which might have contributed to the drop in ASIR in the second period. However, the ASMR was higher in the second period (5.6 vs 3.0 per 100,000). The high mortality rate observed in the second period was due to the higher agespecific mortality rates for all the age-groups, except for women aged 60 to 69 years, as compared to the first period. This is the first study looking at the overall survival rates and prognostic factors of EOC in Brunei Darussalam. The overall survival rates at 1-, 3-and 5-year for EOC patients in Brunei Darussalam were 79.7%, 69.7%, and 61.4% respectively. Studies from other Asian countries reported 5-year overall survival rates among OC patients of 51.2% in Malaysia (NCR, 2018), 54.8% in Indonesia (Aziz, 2009) and42.7% in Singapore (NRDO, 2021).
It is commonly reported that the risk of EOC is strongly related to age, highest in older females. Younger patients had better 5-year survival than older patients, consistent with other large population-based studies (Chan et al., 2006;Cabanes et al., 2009). 5-year survival in this study was 92.6% for women aged below 40 years at diagnosis, while survival ranged from 40.4% to 43.9% for women aged 50 and above at diagnosis. Better survival in younger women may be related to the fact that older women are more likely to have other health problems. In addition, OC may be diagnosed earlier in pre-menopausal women than in post-menopausal women, because one major symptom used to identify OC is a change in menstruation (Akhtar-Danesh et al., 2011).
Cancer stage, a prognostic factor recognized by previous studies (Lokman et al., 2017;Lee et al., 2018), was found to be significantly associated with survival of EOC patients in this study. About 33.7% of the total cases were diagnosed at distant stage. This study found patients with distant stage have significantly higher hazard compared to patients with localized and regional stages, consistent with previous studies (Lee et al., 2018;Liu et al., 2020;Chang et al., 2018). More than half of the women aged 70 and above (52.9%), and approximately 62.5% of women aged between 60 to 69 were diagnosed at distant stage. In most cases, OC is not diagnosed until it has progressed to an advanced stage and the cancer has spread beyond the ovaries which explains the lower survival from OC (Permuth-Wey and Sellers, 2009). OC symptoms in the early stage tend to be not noticeable (Verheijen and Zweemer, 2016).
Histological type is another prognostic factor for EOC patients in this study. While serous carcinoma is reported to be the most common histological type (Ishioka et al., 2004;Kurman et al., 2014), our study found that 30.9% were diagnosed with mucinous carcinoma, followed by serous carcinoma (28.0%). Our data showed the 5-year survival of patients with mucinuous carcinoma or clear cell carcinoma was better than those of serous carcinoma, endometroid carcinoma, and undifferentiated or other epithelial. Although the current managements for epithelial ovarian carcinoma were similar regardless different histological types (Chiang et al., 2013), the modification of the treatments based on the histological types is necessary especially for patients with undifferentiated carcinoma or other epithelial because of their poorer outcome.
EOC, although less common compared to the other gynecological cancers, is the most deadly. We found that the 5-year survival rate of OC patients is 61.4%, lower as compared to the other gynecologic cancer in Brunei Darussalam such as cervical cancer with 5-year survival rate of 68.6% . Treatments for OC have improved over time, such as the development of adjuvant therapy with platinum-based drugs, tumor cytoreductive surgery, taxane-based chemotherapy, and intraperitoneal delivery of chemotherapy (Cristea et al., 2010;Raja et al., 2012).
There are several limitations in our study. First, data were collected retrospectively and were based on the inherent accuracy of patient records. Therefore, our study may have biases especially with incomplete data collection. Second, cancer treatment, which is an important prognostic factor for ovarian tumours (Lee et al., 2018;Chang et al., 2018), is not available for this study. Other potential prognostic factors that have been reported in other studies such as BMI at diagnosis, CA125 levels, first-degree family history of cancer, grade of cancer, gene expression and marital status are also not available (Liu et al., 2020;Goode et al., 2020). Future studies should include these potential prognostic factors which would be useful for examining the impact of early diagnosis of survival and could contribute to understanding some of the variation observed between countries.
Despite these limitations, the findings of the present study provide policy makers, and health professionals with key data to call for increasing awareness among women about early symptoms of OC. Early detection of disease should be emphasized through public education and raising awareness to improve survival rates of patients with EOC.