Tetsuro Yahata, Chiaki Banzai, Kenichi Tanaka and Niigata Gynecological Cancer Registry

645..650 doi:10.1111/j.1447-0756.2011.01755.x J. Obstet. Gynaecol. Res. Vol. 38, No. 4: 645 650, April 2012 Histology-specific long-term trends in the incidence of ovarian cancer and borderline tumor in Japanese females: A population-based study from 1983 to 2007 in Niigata Tetsuro Yahata, Chiaki Banzai, Kenichi Tanaka and Niigata Gynecological Cancer Registry Division of Obstetrics and Gynecology, Graduate School of Medical and Dental Sciences, Niigata University, Niigata, Japanjog_1755 Abstract Aim: The histology-specific long-term trends in the incidence of ovarian cancer and borderline tumors in Japanese women were examined, based on data from the population-based cancer registry in Niigata, Japan. Material and Methods: Data were obtained from the Niigata Gynecological Cancer Registry, which covered the entire female population in Niigata prefecture, Japan, during the period from 1983 to 2007. Results: A total of 3134 females with epithelial ovarian cancer, including borderline tumor cases, were diagnosed between 1983 and 2007. The age-standardized rates (ASRs) of both ovarian cancer and borderline tumors have steadily increased, with significant changes in ovarian cancer in all age groups, and borderline ovarian tumors in subjects aged <50. The ASRs of endometrioid adenocarcinoma showed a steady increasing trend, and those of clear cell and mucinous adenocarcinomas showed significant increasing trends in the total population. The ASRs of clear cell, mucinous, and endometrioid adenocarcinomas in the 50+ age group were significantly increased, especially the incidence of clear cell adenocarcinoma, which strikingly increased by approximately threefold from 1.2 (1983 1989) to 3.5 (2000 2007) per 100 000 females. Conclusion: This prefecture-wide study showed the practical trends in ovarian cancer and borderline tumors in Japanese females. The incidence of ovarian cancer has steadily increased, with significant increases in the incidence of clear cell and mucinous adenocarcinomas in the total population during the past two decades. Because of the poor response rate of these histological subtypes to platinum-based regimens, novel treatment approaches should be adopted to improve the prognostic outcome in patients with ovarian cancer in Japan. Key words: borderline ovarian tumor, clear cell adenocarcinoma, incidence, mucinous adenocarcinoma, ovarian cancer. Introduction Ovarian cancer is the sixth most common cancer in Japanese females, with an estimated 8304 cases in 2005, 1 and is the leading cause of death among female cancer patients worldwide as the majority of patients diagnosed with ovarian cancer present with advanced stage disease. Despite the significant advances in surgery and chemotherapy achieved over the past two decades, the resulting overall 5-year survival rate in patients with advanced-stage disease remains approximately 40%. The incidence of ovarian cancer in Japan is relatively low compared with other developed countries, such as those in Northwestern Europe and North America. These countries have showed decreasing trends in Received: April 2 2011. Accepted: August 3 2011. Reprint request to: Dr Tetsuro Yahata, Division of Obstetrics and Gynecology, Graduate School of Medical and Dental Sciences, Niigata University, 1-757 Asahimachi-dori, Chuo Ward, Niigata 951-8510, Japan. Email: yahatat@med.niigata-u.ac.jp 2012 The Authors 645

T. Yahata et al. ovarian cancer incidence, whereas it has been increasing in Japan since the 1970s, with steady increases in the older age group. 2 The Research Group for Population-based Cancer Registration in Japan also has reported an increasing incidence of ovarian cancer over the past two decades based on the data from 5 12 population-based cancer registries; 3 however, no information regarding the incidence according to the histological subtype could be obtained. The histological subtype has been recognized to be one of the important prognostic factors in ovarian cancer. In particular, clear cell and mucinous adenocarcinomas have been identified as adverse prognostic features because of their poorer response to platinum-based chemotherapy compared to serous or endometrioid adenocarcinoma. The primary objective of this study was to use data from the prefecture-wide gynecological cancer registry in Niigata (Niigata Gynecologic Cancer Registry) to describe the females diagnosed with ovarian cancer and borderline tumors within the prefecture. This analysis evaluated 24 years (1983 2007) of population-based datasets in order to investigate the trends in the incidence of ovarian cancer and borderline tumors in Japanese females according to the histological subtype. Material and Methods These analyses were based on malignant ovarian tumor cases reported to the Niigata Gynecologic Cancer Registry between 1983 and 2007. This registry was established in 1983 and it is a population-based registry which covers the entire female population in Niigata prefecture of around 1.2 million female inhabitants. All information on newly diagnosed gynecological malignant tumors, as well as pre-malignant tumors in females aged 10 years and over, has been collected. The information sources include all hospitals with gynecological departments, where patients with malignant gynecological tumors are diagnosed and/or treated, in Niigata prefecture. The inclusion criteria for the analysis were primary ovarian cancer cases and primary borderline ovarian tumor cases with microscopic confirmation of disease occurring between 1 January 1983 and 31 December 2007. The histological subtypes were classified into five categories: (i) serous adenocarcinoma, (ii) mucinous adenocarcinoma, (iii) endometrioid adenocarcinoma, (iv) clear cell adenocarcinoma, and (v) others, which included undifferentiated or unspecified carcinomas. Non-epithelial ovarian malignancies were excluded from the analysis. The trends in the ovarian cancer incidence in Niigata prefecture from 1983 to 2007 were investigated. Agestandardized incidence rates (ASRs) were calculated by the direct method with 95% confidence intervals (CI) for each disease, with the world population proposed by Segi et al. 4 as the standard. The 95% CI was calculated by the formula 95% CI = ASR 1.96x [Standard Error of ASR]. Patients were divided into two groups based on age: <50 years and 50+ years, for both ovarian cancer and borderline tumors. The changes in the ASRs in three chronological periods, 1983 1989, 1990 1999, and 2000 2007, were examined according to the histological subtype. Differences in the ASRs between the first period (1983 1989) and the last period (2000 2007) were evaluated using rate ratios (RR) and the corresponding 95% CI. The SPSS statistical software program (SPSS, IBM, Somers, NY, USA) was used for the statistical analyses. Results The trends in the ASRs are illustrated in Figure 1. A total of 3134 females with ovarian cancer, including borderline tumor cases, were diagnosed between 1983 and 2007. The ASRs of both ovarian cancer and borderline tumors steadily increased from 3.9 (in 1983) to 8.0 (in 2007) per 100 000 females, and from 0.9 (in 1983) to 3.4 (in 2007) per 100 000 females, respectively. The trends in the ASRs of ovarian cancer and borderline tumors by age group and the period of diagnosis are shown in Table 1. The incidence rates of ovarian cancer and borderline tumors significantly increased during the last period (2000 2007) in comparison to the Figure 1 Trends in age-standardized incidence rates of ovarian cancer and borderline tumors. Standard population: world standard population. 646 2012 The Authors

Ovarian cancer in Japan Table 1 Trends in the age-standardized incidence rates of ovarian cancer and borderline tumor by age group and period of diagnosis Time period Age group <50 50+ All ages Ovarian cancer 1983 1989 3.2 (2.8 3.6) 10.0 (8.9 11.2) 4.0 (3.5 4.5) 1990 1999 4.1 (3.5 4.5) 14.7 (13.6 15.9) 6.2 (5.6 6.8) 2000 2007 4.2 (3.5 4.9) 17.7 (15.4 19.9) 6.7 (6.0 7.8) Rate ratio (95%CI) 1.3 (1.1 1.6) 1.7 (1.5 2.0) 1.5 (1.3 1.7) Borderline tumor 1983 1989 0.7 (0.4 1.0) 1.5 (0.9 2.1) 0.8 (0.5 1.0) 1990 1999 1.2 (0.9 1.6) 1.2 (0.8 1.6) 1.2 (1.0 1.5) 2000 2007 1.5 (0.8 2.1) 1.8 (1.2 2.4) 1.6 (0.9 2.2) Rate ratio (95%CI) 2.4 (1.2 3.6) 2.2 (0.1 4.3) 1.8 (1.2 2.5) Rate ratio: 2000 2007 versus 1983 1989. Bold items show statistically significant differences compared with 1983 1989. Values are incidence rate (95% confidence interval). Incident rates are per 100 000 and age-adjusted to the world standard population. Table 2 Trends in the age-standardized incidence rates of ovarian cancer according to histological type by age and period of diagnosis Time period Age group <50 50+ All ages Serous adenocarcinoma 1983 1989 1.1 (0.9 1.3) 5.0 (4.2 5.9) 1.9 (1.4 2.4) 1990 1999 1.2 (1.0 1.5) 5.9 (5.4 6.4) 2.2 (1.7 2.7) 2000 2007 1.2 (1.0 1.3) 6.1 (5.0 7.2) 2.1 (1.7 2.6) Rate ratio (95%CI) 1.1 (0.9 1.3) 1.2 (0.9 1.5) 1.2 (0.9 1.4) Mucinous adenocarcinoma 1983 1989 0.7 (0.4 1.1) 1.6 (1.4 1.8) 0.9 (0.7 1.1) 1990 1999 1.0 (0.7 1.4) 2.8 (2.3 3.3) 1.4 (1.1 1.6) 2000 2007 0.9 (0.7 1.1) 2.5 (1.9 3.1) 1.2 (1.0 1.5) Rate ratio (95%CI) 1.9 (0.6 3.2) 1.6 (1.3 1.9) 1.5 (1.1 1.8) Endometrioid adenocarcinoma 1983 1989 0.4 (0.2 0.6) 0.8 (0.4 1.2) 0.5 (0.3 0.7) 1990 1999 0.7 (0.5 0.8) 1.8 (1.4 2.1) 0.9 (0.6 1.2) 2000 2007 0.7 (0.4 1.1) 2.0 (1.6 2.5) 1.0 (0.6 1.4) Rate ratio (95%CI) 1.5 (0.7 2.3) 2.1 (1.1 3.1) 2.8 (0.7 4.8) Clear cell adenocarcinoma 1983 1989 0.3 (0.2 0.5) 1.2 (0.7 1.8) 0.5 (0.3 0.7) 1990 1999 0.4 (0.3 0.6) 2.3 (1.8 2.7) 0.8 (0.5 1.1) 2000 2007 0.7 (0.5 1.0) 3.5 (2.4 4.7) 1.3 (0.8 1.8) Rate ratio (95%CI) 2.5 (1.3 3.6) 3.1 (1.8 4.5) 2.4 (1.7 3.1) Rate ratio: 2001 2007 versus 1983 1989. Bold items show statistically significant differences compared with 1983 1989. Values are incidence rate (95% confidence interval). Incident rates are per 100 000 and age-adjusted to the world standard population. first period (1983 1989) in the total population. Greater changes were observed in the older age group than in the younger age group in ovarian cancer cases, whereas a significant change was observed only in the younger age group in the borderline tumor cases. The trends in the ASRs of ovarian cancer according to the histological subtype by age and period of diagnosis are shown in Table 2 and Figure 2. The ASRs of endometrioid adenocarcinoma showed a steady increasing trend, and those of clear cell and mucinous adenocarcinomas showed significant increasing trends in the total population. In the 50+ age group, the steady increasing trends were observed for all histological subtypes, with significant increasing trends in clear cell, mucinous, and endometrioid adenocarcinomas. The incidence of clear cell adenocarcinoma was strikingly increased by approximately threefold from 1.2 (1983 1989) to 3.5 (2000 2007) per 100 000 females 2012 The Authors 647

T. Yahata et al. in the 50+ age group. Serous adenocarcinoma is still the most common histological subtype; however, the proportion of serous adenocarcinomas decreased from 59% (in 1983) to 29% (in 2007), whereas the proportion of clear cell adenocarcinoma increased from 13% (in 1983) to 24% (in 2007). For borderline tumors, we examined the time-trends based on the two main histological subtypes: serous and mucinous tumors. The mucinous type showed two- to fourfold higher rates than the serous type, and the ASRs for both showed a significant increasing trend in the total population during the study period (Table 3). Rate per 10 000 population women 14 12 10 8 6 4 2 0 <50 years old 1983 1989 1990 1999 2000 2007 50+ years old 1983 1989 1990 1999 2000 2007 serous mucinous endometrioid clear cell Figure 2 Trends in age-standardized incidence rates of ovarian cancer according to histological type by age and period of diagnosis. Other histological types, which included undifferentiated or unspecified carcinomas, are not included in this figure. Standard population: world standard population. Discussion This study investigated the evolution of trends in the incidence of ovarian cancer and borderline tumors in Japanese females during the last 24 years using data from the prefecture-wide gynecological cancer registry in Niigata. There was a significant increase in the incidence of both ovarian cancer and borderline tumors during the study period. It was noteworthy that an analysis of the trends in the incidence of ovarian cancer according to the histological subtype showed a significant increase in clear cell adenocarcinoma (3.1-fold) among women in the 50+ age group during past two decades. Ovarian cancer represents the sixth most commonly diagnosed cancer among women in Japan, as well as in the world, and causes more deaths per year than any other cancer of the female reproductive system. 5 Large geographical variations in the incidence of ovarian cancer have been reported. Higher incidence rates are reported from Northwestern Europe and North America, where the incidence rates have remained almost constant or have shown decreasing trends over the past few decades. 6 8 Ovarian cancer is relatively uncommon in Japan, but an increase in the incidence has been observed in recent years. The ASR of ovarian cancer was reported to have increased by approximately 1.3-fold (from 4.0 to 5.4 per 100 000 women) in Osaka, Japan from 1975 to 1998. 9 The national cancer incidence in Japan has been estimated by the Research Group for Population-based Cancer Registration in Japan since 1975, based on the data from 5 12 population-based cancer registries. 3 They also have reported an increasing ASR of ovarian cancer from Table 3 Trends in the age-standardized incidence rates of borderline tumor according to histological type by age and period of diagnosis Time period Age group <50 50+ All ages Serous borderline 1983 1989 0.1 (0.1 0.2) 0.1 (0.0 0.2) 0.1 (0.0 0.2) 1990 1999 0.3 (0.2 0.5) 0.3 (0.1 0.4) 0.3 (0.2 0.4) 2000 2007 0.4 (0.2 0.6) 0.4 (0.1 0.7) 0.4 (0.2 0.6) Rate ratio (95%CI) 2.6 (1.3 3.9) 1.7 (0.2 3.2) 2.6 (1.2 4.0) Mucinous borderline 1983 1989 0.2 (0.0 0.4) 1.0 (0.6 1.3) 0.4 (0.2 0.5) 1990 1999 0.6 (0.4 0.7) 0.8 (0.5 1.2) 0.6 (0.5 0.8) 2000 2007 1.0 (0.6 1.3) 1.0 (0.7 1.4) 1.0 (0.7 1.3) Rate ratio (95%CI) 1.9 (0.5 3.3) 1.1 (0.1 2.1) 2.6 (1.8 3.4) Rate ratio: 2000 2007 versus 1983 1990. Bold items show statistically significant differences compared with 1983 1990. Values are incidence rate (95% confidence interval). Incident rates are per 100 000 and age-adjusted to the world standard population. 648 2012 The Authors

Ovarian cancer in Japan 4.7 (1984) to 7.4 (2005). 1,3 The trend in the incidence rate in this prefecture-wide study was similar to the estimated trends in ovarian cancer in Japanese females determined by the Research Group. Borderline ovarian tumors are a type of epithelial ovarian cancer of low malignant potential, which account for approximately 15% of all epithelial ovarian cancers. It tends to afflict a young population, and the majority of patients are diagnosed with stage I tumors with a favorable prognosis; however, recurrence and death may develop in a small number of patients. 10 The information regarding the trend in the incidence of borderline ovarian tumors is limited because there have been only a few epidemiological series. Increases in the incidence between the 1970s and 1990s have been reported in Northwestern Europe, reaching ASRs of 2.7 4.8 per 100 000 females. 11 However, a decrease in the incidence in Northern Europe has been observed during the past decades, possibly reflecting a favorable impact of the use of combined oral contraceptives. 12 The present series showed an increased incidence in borderline ovarian tumors in Japanese females during the past two decades, especially in younger patients. Several risk and protective factors have been reported to be associated with epithelial ovarian cancer. The most well-studied and consistent findings have been reproductive and hormonal factors, such as the number of live births, menopausal status, and use of oral contraceptives. Among them, parity has been reported to be an even stronger protective factor in low-incidence populations. Since 1950, the average parity per woman of childbearing age in Japan has rapidly decreased, to a nadir of 1.26 in 2005. Thus, the trend of declining parity may be an explanation for the increasing incidence of both epithelial ovarian cancer and borderline tumors in Japan. The histological subtypes of ovarian cancer are markedly different from what has been reported in Caucasian populations, especially the high incidences of clear cell adenocarcinoma in Japanese patients. The proportion of clear cell adenocarcinoma has been increasing, and reached 24% in 2007 in this study. However, the frequency of clear cell adenocarcinoma in United States was only 4.1% according to the Surveillance, Epidemiology and End Results database for 1992 1999. 13 Although the reason for these differences in the breakdown of the histological subtype is unknown, they may be due to the differences in genetic background, as it has been reported that the incidence of ovarian clear cell adenocarcinoma is higher in women of Asian descent. In very few studies has it been possible to both evaluate and draw conclusions about time trends in the incidence of the different histological subtypes of ovarian cancer because of the difference in genetic and environmental backgrounds between studies. Cramer et al. reported an increase in the incidence of endometrioid and clear cell adenocarcinomas of the ovary in the USA during the 1970s, while no change occurred in the overall incidence of ovarian cancer. 14 A recent study in Switzerland showed the histological distribution of ovarian cancer according to the calendar period 1974 1988. 15 The proportion of serous and endometrioid adenocarcinomas tended to increase with time in Swiss patients. The present study showed steady increasing trends in all histological subtypes, with significant changes in the incidence of clear cell, endometrioid, and mucinous adenocarcinomas in the 50+ age group. The incidence rates for clear cell and endometrioid adenocarcinomas were especially increased, by two- to threefold, during the past two decades. Many epidemiological studies have shown that endometriosis increases the risk of the development of ovarian cancer. Recently, the incidence of endometriosis has increased in Japan, possibly because of delayed marriage, decrease in parity, and unavailability of oral contraceptives. The increasing trends in the incidence of endometrioid and clear cell adenocarcinomas may be, at least partly, due to the increased rate of endometriosis, as both histological subtypes were suggested to be associated with endometriosis by epidemiological and clinicopathological studies. 16 Significant increases in clear cell and mucinous adenocarcinomas in the total population were shown in the present study. Both histological subtypes have been identified as having adverse prognostic features because they have a poorer response to platinum-based chemotherapy than serous or endometrioid adenocarcinomas. Despite their aggressive clinical course, these cancers are still treated similarly to the other epithelial ovarian cancers at the present time, because the rarity of the histological subtype has prevented randomized studies from being performed. The development of more active agents than those that are currently available is needed to improve the prognostic outcome in such patients. An international randomized control study of TC (paclitaxel/carboplatin) versus CPT-P (CPT11/cisplatin) chemotherapy for ovarian clear cell adenocarcinoma by the Japanese Group of Gynecologic Oncology has been ongoing since 2006. Patient enrollment was completed in February 2011, with a 2012 The Authors 649

T. Yahata et al. total of 662 patients, and this trial might lead to a new treatment strategy for ovarian clear cell adenocarcinoma in the near future. The treatment of ovarian mucinous adenocarcinoma remains a challenge. Currently, optimal cytoreduction represents the most important factor for better survival, and should be sought aggressively by these patients. Additional translational studies are clearly warranted to identify the molecular pathways unique to these cell types, to design novel biological agents to overcome the drug resistance of these aggressive cancers. The present population-based study, like others, was restricted by a lack of central pathological review. Although a pathological diagnosis was made by more than one pathologist in each cancer treatment hospital, our conclusion may potentially be misleading if there were any significant disagreements among the pathologists in various hospitals contributing to the Niigata Gynecologic Cancer Registry. Additional concerns about the interpretation of the data include possible advances in the diagnosis of ovarian cancer. Various imaging modalities have emerged as important contributors to the diagnosis of ovarian cancer, such as MRI, transvaginal ultrasonography, and the introduction of color flow Doppler imaging during the study period. As this study included only pathologically-confirmed ovarian cancer and borderline tumor cases, advances in these diagnostic modalities in daily practice may have increased the number of operable cases, and thus the incidence estimates may not reflect the true number of patients. In summary, the present prefecture-wide study showed the practical trends in ovarian cancer and borderline tumors by histological type in Japanese females. The incidence of both ovarian cancer and borderline tumors has steadily increased, with significant increases in the incidence of clear cell and mucinous adenocarcinomas in the total population over the past two decades. Because of the poor response rate of these histological subtypes to platinum-based regimens, novel and more active treatment approaches should be adopted in order to improve the prognostic outcome in such patients. Acknowledgments We thank the members of the Gynecologic Cancer Registry of Niigata for data collection. Disclosure None declared. References 1. Matsuda T, Marugame T, Kamo K, Katanoda K, Ajiki W, Sobue T. Cancer incidence and incidence rates in Japan in 2005: based on data from 12 population-based cancer registries in the Monitoring of Cancer Incidence in Japan (MCIJ) project. Jpn J Clin Oncol 2011; 41: 139 147. 2. Tamakoshi K, Kondo T, Yatsuya H, Hori Y, Kikkawa F, Toyoshima H. Trends in the mortality (1950 1997) and incidence (1975 1993) of malignant ovarian neoplasm among Japanese women: analyses by age, time, and birth cohort. Gynecol Oncol 2001; 83: 64 71. 3. Research Group for Population-based Cancer Registration in Japan. Cancer incidence and incidence rates in Japan in 1984 estimates based on data from eight cancer registries Research Group for Population-based Cancer Registration in Japan. Jpn J Clin Oncol 1989; 19: 82 85. 4. Segi M, Fujisaku S, Kurihara M, Narai Y, Sasajima K. The age-adjusted death rates for malignant neoplasms in some selected sites in 23 countries in 1954 1955 and their geographical correlation. Tohoku J Exp Med 1960; 72: 91 103. 5. Permuth-Wey J, Sellers TA. Epidemiology of ovarian cancer. Methods Mol Biol 2009; 472: 413 437. 6. Goodman MT, Shvetsov YB. Incidence of ovarian, peritoneal, and fallopian tube carcinomas in the United States, 1995 2004. Cancer Epidemiol Biomarkers Prev 2009; 18: 132 139. 7. Adami HO, Bergstrom R, Persson I, Sparen P. The incidence of ovarian cancer in Sweden, 1960 1984. Am J Epidemiol 1990; 132: 446 452. 8. Bray F, Loos AH, Tognazzo S, La Vecchia C. Ovarian cancer in Europe: cross-sectional trends in incidence and mortality in 28 countries, 1953-2000. Int J Cancer 2005; 113: 977 990. 9. Ioka A, Tsukuma H, Ajiki W, Oshima A. Ovarian cancer incidence and survival by histologic type in Osaka, Japan. Cancer Sci 2003; 94: 292 296. 10. Silva EG, Gershenson DM, Malpica A, Deavers M. The recurrence and the overall survival rates of ovarian serous borderline neoplasms with noninvasive implants is time dependent. Am J Surg Pathol 2006; 30: 1367 1371. 11. Levi F, La Vecchia C, Randimbison L, Te VC. Borderline ovarian tumours in Vaud, Switzerland: incidence, survival and second neoplasms. Br J Cancer 1999; 79: 4 6. 12. La Vecchia C, Tavani A, Franceschi S, Parazzini F. Oral contraceptives and cancer. A review of the evidence. Drug Saf 1996; 14: 260 272. 13. Quirk JT, Natarajan N. Ovarian cancer incidence in the United States, 1992 1999. Gynecol Oncol 2005; 97: 519 523. 14. Cramer DW, Devesa SS, Welch WR. Trends in the incidence of endometrioid and clear cell cancers of the ovary in the United States. Am J Epidemiol 1981; 114: 201 208. 15. Levi F, Franceschi S, La Vecchia C, Ruzicka J, Gloor E, Randimbison L. Epidemiologic pathology of ovarian cancer from the Vaud Cancer Registry, Switzerland. Ann Oncol 1993; 4: 289 294. 16. Van Gorp T, Amant F, Neven P, Vergote I, Moerman P. Endometriosis and the development of malignant tumours of the pelvis. A review of literature. Best Pract Res Clin Obstet Gynaecol 2004; 18: 349 371. 650 2012 The Authors