Japanese Journal of Clinical Oncology Advance Access originally published online on July 8, 2005
Japanese Journal of Clinical Oncology 2005 35(8):464-469; doi:10.1093/jjco/hyi125
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
© 2005 Foundation for Promotion of Cancer Research
Influence of Age on Cervical Cancer Survival in Japan
Department of Cancer Control and Statistics, Osaka Medical Center for Cancer and Cardiovascular Diseases, Osaka, Japan
For reprints and all correspondence: Akiko Ioka, Department of Cancer Control and Statistics, Osaka Medical Center for Cancer and Cardiovascular Diseases, 1-3-3 Nakamichi, Higashinari-ku, Osaka 537-8511, Japan. E-mail: akiko3{at}gol.com
Received April 2, 2005; accepted May 26, 2005
 |
Abstract |
|---|
 TOP Abstract INTRODUCTIONPATIENTS AND METHODSRESULTSDISCUSSIONReferences |
|---|
Background: Relative 5-year survival for cervical cancer has been reported to be lower in older women in Japan. A population-based study was carried out to clarify why increased age is associated with decreased survival in spite of a nationwide cervical cancer screening program having been carried out since 1982 in Japan.
Methods: The Osaka Cancer Registry's data were used to investigate associations between age groups and survival for cervical cancer patients. Survival analysis was restricted to the reported 8966 cases diagnosed in 1975–1996 who lived in Osaka Prefecture (except for Osaka City), or resided in Osaka City in 1993–1996, since active follow-up data on vital status 5 years after the diagnosis were available.
Results: Relative 5-year survival for cervical cancer cases was lower in older age groups (88.6% in <30 years, 78.1% in 30–54 years, 67.7% in 55–64 years and 54.4% in 65+ years), as was the proportion of the detection by screening (6.3, 9.8, 9.2 and 6.0%), the proportion of the localized stage (83.0, 67.3, 51.0 and 42.7%) and the proportion of women who underwent surgery (79.2, 83.2, 65.6 and 35.2%). Among localized cases detected with screening, the survival in those 
55 years old was >92% and almost comparable with that in 30 to 54 year olds, but significantly lower among those detected without screening.
Conclusion: Lower survival among older women was caused mainly by the presence of more advanced disease at diagnosis. Further extension of the nationwide cervical cancer screening program should result in improved diagnosis of earlier stage disease, which might improve differences of cervical cancer survival among these age groups.
Key Words: survival • cervical cancer • age • cancer stage • cervical cancer screening
|
INTRODUCTION |
|---|
|
TOPAbstractINTRODUCTIONPATIENTS AND METHODSRESULTSDISCUSSIONReferences |
|---|
Cervical cancer is the second most common cancer among women worldwide (GLOBOCAN 2002, http://www.depdb.iarc.fr/globocan/GLOBOframe.htm). In Japan, both the mortality from and incidence of uterine cancer have decreased steadily during the last few decades (1,2). The relative 5-year survival for uterine cancer has increased slightly, but less so with older women in Osaka, Japan (3). Similar findings were also obtained in studies of relative 5-year survival for cervical cancer (4,5).
Widespread use of cervical screening has been associated with the substantial reduction in the incidence rate of and mortality from cervical cancer (6–8). Cervical cancer screening has been recommended to be carried out every year for those aged 30 years and over by the Japanese Ministry of Health, Labor and Welfare since 1982, and revised clinical guidelines addressing who (women aged 20 years and over) and how often (every 2 years) to screen for cervical cancer were introduced in 2004. The observed increase in survival for cervical cancer must at least partly be due to the nationwide cervical cancer screening program. However, survival has been lower for older women. Are lower survival rates associated with poorer uptake of the program among older women? In this study, using the Osaka Cancer Registry's data, we try to clarify why increased age is associated with decreased survival, and discuss what improvements could be made to the program.
|
PATIENTS AND METHODS |
|---|
|
TOPAbstractINTRODUCTIONPATIENTS AND METHODSRESULTSDISCUSSIONReferences |
|---|
DATA SOURCES
Data on 8966 reported cervical cancer cases (ICD Tenth Revision, C53) who were newly diagnosed in 1975–1996 were retrieved from the Osaka Cancer Registry's database. The Osaka Cancer Registry (OCR) has been operating since December 1962, covering Osaka Prefecture with its population of 8.8 million (2000 census). Cancer incidence data in Osaka have been reported in ‘Cancer Incidence in Five Continents’, volumes III in 1976 to VIII in 2002 (9). The quality of these data, therefore, can be assumed to have met the standards set up by the International Association of Cancer Registries during the last three decades. In 1993–1997, the proportion of death certificate only (DCO) cases was 4%, and the mortality to incidence (M/I) ratio was 0.31 for cervix uteri (9).
To obtain information on the vital status of registered cases, the OCR has used the following three procedures: (i) collation between the master file and the cancer death file; (ii) collation between living cases from the master file and the file of all death certificates of Osaka residents; and (iii) confirmation of the cases' living status by referring to registers in local municipality offices of inhabitants 5 years after the diagnosis. This final step was, however, not conducted for those residing in Osaka City in 1975–1992. Therefore, survival analysis was restricted to cases who lived in Osaka Prefecture (except for Osaka City) in 1975–1996 or resided in Osaka City in 1993–1996, since they had active follow-up information. In addition, they met the following criteria: (i) in the case of multiple tumors, only the first was included; and (ii) cases diagnosed as carcinoma in situ were excluded.
In total, 426 cases (4.8%) were lost to follow-up and treated as censored cases at the latest date when they were confirmed as alive.
Cervical cancer screening was recommended to be carried out every year for those aged 30 years by the Japanese Ministry of Health, Labor and Welfare in 1982–2003, and almost all Japanese women have had their menopause by the age of 55 years. There is an evidence-based report which recommends that women with a history of regular normal Pap smears should discontinue screening at age 65 as the risk of cervical cancer and other abnormalities falls with age; fewer than 1 in 1000 women aged 60 years with a history of normal Pap smears will develop cervical intraepithelial neoplasia (CIN 3) or cancer, and the longer the history of prior normal Pap tests the greater is the reduction in risk (10,11). Age at first diagnosis was classified into four categories: <30 years, from 30 to 54 years, from 55 to 64 years and 65 years and over. The cancer stage at diagnosis was classified into the following three groups in OCR: localized (cancer is confined to the original organ), regional (cancer spreads to regional lymph nodes and/or spreads to immediately adjacent organs/tissues) or distant (cancer metastasizes to distant organs/tissues). These stages were defined as follows: localized, FIGO (International Federation of Gynecology and Obstetrics) stage I (excluding extension to corpus uteri); regional, FIGO stages I (with extension to corpus uteri), II, III and IV A inclusive; distant, FIGO stage IV B.
The definition of the localized stage is the same between OCR and US-SEER (Surveillance, Epidemiology and End Results), but not for regional/distant stages: in US-SEER, regional stage is defined as FIGO stages I (with extension to corpus uteri), II and III inclusive, and distant stage as FIGO stage IV (12).
STATISTICAL ANALYSIS
The distributions of patients' characteristics were assessed with 
2 tests for categorical variables. Cumulative observed survival was estimated using the Kaplan–Meier method according to age group. Survival time was computed from the date of first diagnosis to the end-point, defined as death from any cause. Closing date was defined as the date 5 years after the first diagnosis. Relative 5-year survival was calculated adjusting for differences in the probability of dying of causes other than cancer among subjects. Relative survival was calculated as the ratio of observed survival to expected survival, the latter being estimated using the survival probability in the general population of Japan of similar subjects with respect to sex, age and calendar year at diagnosis. The Ederer II method was employed (13,14). Differences were considered as statistically significant if P-values were <0.05 by two-sided test. Statistical package software, SPSS (version 11.0), was used for statistical analysis.
|
RESULTS |
|---|
|
TOPAbstractINTRODUCTIONPATIENTS AND METHODSRESULTSDISCUSSIONReferences |
|---|
Table 1 presents all patients' characteristics and relative 5-year survival. In all cases, the average age was 54.6 years (SE 0.1), the proportion of the detection with screening was only 8.6%, and relative 5-year survival was 70.2% (SE 0.5) during 1975–1996. The survival decreased gradually with greater age (88.6, 78.1, 67.7 and 54.4%), as did the proportion of detection by screening (6.3, 9.8, 9.2 and 6.0%), the proportion of the localized stage (83.0, 67.3, 51.0 and 42.7%) and the proportion of women who underwent surgery (79.2, 83.2, 65.6 and 35.2%). Among cases detected with screening/localized cases/cases who underwent surgery, the survival also decreased with increased age: the survival in 30 to 54 year olds was obviously higher than that in those aged 55–64 and
65 years. In those <30 years old, the survival among cases who underwent surgery was somewhat lower than that of those who did not undergo surgery, although this difference was not statistically significant. This unexpected finding was caused by a difference in stage distribution between the two groups (data not shown).
|
Table 2 shows patients' characteristics among cases detected by screening or not, where unknown cases for screening were excluded. The proportion of the localized stage decreased markedly with older age groups among cases detected both with screening (73.3, 56.3 and 47.4%, P < 0.01) and without screening (67.4, 51.3 and 44.0%, P < 0.01), and also the proportion of women who underwent surgery among both cases detected with screening (85.5, 73.9 and 43.0%, P < 0.01) and without screening (83.4, 65.7 and 35.6%, P < 0.01).
|
Table 3 indicates relative 5-year survival among cases detected by screening or not. The survival decreased with older age among cases detected both with screening (84.3, 75.4 and 64.2%) and without screening (77.6, 67.1 and 55.0%). Interesting findings were observed regarding survival if considering the cancer stage in addition to detection with screening or not: for localized cases, the survival did not decrease appreciably with older age among cases detected by screening (94.0, 92.3 and 92.1%), while, among those detected without screening, the survival in 55–64 and
65 year olds was significantly lower than that in 30 to 54 year olds (85.7 versus 91.8%, P < 0.01) (75.8 versus 91.8%, P < 0.01). For cases who underwent surgery, the survival did not decrease remarkably with greater age among cases detected by screening (89.0, 83.1 and 85.1%), while, among those detected without screening, the survival in 55–64 and 65 year olds was obviously lower than that in 30 to 54 year olds (76.3 versus 83.3%, P < 0.01) (73.4 versus 83.3%, P < 0.01). As >85% of all cervical cancer cases were cases detected without screening in 1975–1996, the survival when the study subjects were restricted to these cases was similar to that for all cases.
|
|
DISCUSSION |
|---|
|
TOPAbstractINTRODUCTIONPATIENTS AND METHODSRESULTSDISCUSSIONReferences |
|---|
The proportion of cervical cancer was >70% in all uterine cancer (ICD 10, C53-55) cases (>80% in all excluding uterine cancer NOS) in 1975–1996, and the proportion of the localized stages among these cases was lower with greater age, even though the nationwide cervical cancer screening program for those aged 30 years and over started in 1982 in Japan. Primarily as a result of the presence of more advanced disease at diagnosis, the survival of cervical cancer cases was lower in older women.
Relative 5-year survival for cervical cancer for 55–64 year olds (67.7%) was lower than that from 30–54 year olds (78.1%), as was that in 65 year olds (54.4%). The proportion of the localized stage was lower with increased age despite the nationwide screening program. Moreover, comparing the survival for localized cases by age groups, the survival also decreased with older age group (91.9, 86.5 and 76.4%). A marked decline in relative survival among women aged from 55 to 64 years is likely to be due to the presence of more advanced cancer stages at diagnosis, while among women aged 65 years it is likely to be due to the increased difficulty in applying complete therapy protocols as general health declines in addition to the presence of more advanced cancer stages at diagnosis. Cervical cancer screening procedures have been shown to reduce morbidity and mortality in cervical cancer (6–8), and Raffle et al. (15) have reported that in the UK National Health Service cervical screening program, 
1000 women need to be screened for 35 years to prevent one death. However, many women are not using these services in Osaka: the coverage of this nationwide screening program was only 20.3% among women aged 30 years and over in 2001 (Comprehensive Survey of the Living Conditions of People on Health and Welfare, http://www.dbtk.mhlw.go.jp/toukei/cgi/j_kensaku), and it was lower in older age groups (29.9% in those aged 40–49 years, 26.3% in those aged 50–59 years and 18.1% in those aged 60–69 years). Our population-based study suggests that there are increasing demands for improvement in the nationwide cervical cancer screening program in order to increase coverage, especially in those aged from 55 to 64 years, because the proportion of localized cases showing relatively good survival is very low.
Relative 5-year survival for localized cases was different depending on whether a cancer was identified through cervical cancer screening or not: the survival did not decrease appreciably with older age among cases detected with screening, but decreased among those detected without screening. Should women aged 65 years and over receive routine screening in Japan? In those aged 65 years old and over, relative 5-year survival among cases who were born in 1910–1939 was significantly higher than that of those who were born before 1909 (59.1 versus 43.9%, P < 0.01), therefore, the survival was higher among cases who had more chances to undergo cervical cancer screening in their younger aged period in our study. On the other hand, the US Preventive Services Task Force is against continuing routine screening of women after age 65 if they have had adequate recent screening with normal results and are not otherwise at increased risk for cervical cancer (16), and also in the UK National Health Service cervical screening program. Moreover relative 5-year survival among women aged 65 years and over in Osaka was a little higher than that in the USA in 1975–1996 (54.4 versus 51.0%), as was the distribution of cancer stages (42.7 versus 33.1%) (17). Although many older American women were reported as having cervical cancer screening more frequently than recommended (almost half of women aged 65–74 with no history of abnormal smears) (18), the proportion of localized cases and the survival were lower in older age groups in the USA. We suggest that it is better to continue cervical cancer screening of women aged 65 years and over in Japan, although more investigation will be required to clarify whether it should be recommended to continue routine screening or not.
Survival for cervical cancer patients would be expected to increase by extending cervical cancer screening in all regions of Japan, as there have been no significant improvements in outcomes of treatment for cervical cancer over the past 20 years. Although average ages for cervical cancer remain almost unchanged in 1975–1996 in our study (53.5 years in 1975–1979, 53.7 years in 1980–1984, 55.4 years in 1985–1989 and 55.8 years in 1990–1996), relative 5-year survival for cervical cancer was similar in each diagnosed year group (65.7% in 1975–1979, 72.5% in 1980–1984, 71.7% in 1985–1989 and 70.3% in 1990–1996). In England, the national call and recall system established in 1988, which invites women who are registered with a general practitioner, keeps track of any follow-up investigation, and, if all is well, recalls the women for screening in 3 or 5 years time. Financial incentives were first introduced with general practitioner contracts in 1990. They have increased coverage to 85% (from 42% in 1988 to 85% in 1994), and studies based in the regional cancer registries indicated that since 1988 the stage distribution has shifted towards earlier stages in England (19). This has resulted in falls in the incidence and mortality (6), with almost stable survival for cervical cancer (from 62.6% in 1985–1989 to 63.8% in 1990–1994) (20). In Japan, the number of smears taken in the screening program must increase rapidly, so first we need to implement the organized nationwide cervical cancer screening program like the National Health Service call and recall system. After that, we should use outreach intervention (21,22) increasing the number of screening procedures for non-adherent women who lack opportunities to receive cervical cancer screening.
There are several limitations in this study that should be considered before accepting any of our conclusions. First, there might be misclassification of the information regarding detection by screening as the proportion of cases detected through screening was very low in our study. Cases detected with screening might be registered as cases detected without screening. Secondly, we need to take into consideration the completeness of reporting to the cancer registry by hospitals, because the proportion of DCO cases was 4% for cervix uteri in OCR in 1993–1997, which was higher than that in North American or northern/western European registries. On the other hand, as the survival for cervical cancer is relatively better than that for other cancers, hospitals might tend to be more lax about having poor completeness of notification about patients who are still alive, thus underestimating survival in our study. We need to reanalyze the association between cervical cancer survival and age after having achieved satisfactory completeness of notification. Thirdly, the present study suggested that older age groups were associated with lower survival for cervical cancer in Osaka by analyzing OCR's database only. It is necessary to ascertain from national population-based cancer registry data whether or not similar problems exist in other locations in Japan.
Despite some limitations inherent in this study, survival for cervical cancer among women aged 55 years and over was much lower than that among those aged from 30 to 54 years in relation to the localized stage proportion at diagnosis. The authors consider that, among women aged 55 years and over, further extending cervical cancer screening with resulting increased detection of earlier stage disease at diagnosis would be desirable for improving the survival of cervical cancer patients in Japan, and that it should be feasible to extend the nationwide screening program including the provision of effective outreach interventions.
|
Acknowledgments |
|---|
This work was supported in part by a Grant-in-Aid for Cancer Research (14-2) from the Ministry of Health, Labour and Welfare.
|
References |
|---|
|
TOPAbstractINTRODUCTIONPATIENTS AND METHODSRESULTSDISCUSSIONReferences |
|---|
1 Hirose K, Takezaki T, Tominaga S, Tajima K. Cancer mortality in Japan (1950–2000). In: Tajima K, Kuroishi T, Oshima A, editors. Cancer Mortality and Morbidity Statistics: Japan and the World—2004, Gann Monograph on Cancer Research No. 51. Tokyo: Japan Sci Soc Press 2004;1–93.
2 Ioka A, Tsukuma H, Ajiki W, Oshima A. Trends in uterine cancer incidence in Japan 1975–98. Jpn J Clin Oncol 2003;33:645–6.
3 Ajiki W, Tsukuma H, Oshima A. Trends in cancer incidence and survival in Osaka. In: Tajima K, Kuroishi T, Oshima A, editors. Cancer Mortality and Morbidity Statistics: Japan and the World-2004, Gann Monograph on Cancer Research No. 51. Tokyo: Japan Sci Soc Press 2004;137–63.
4 Osaka Prefectural Department of Environment and Public Health, Osaka Medical Association, Osaka Medical Center for Cancer and Cardiovascular Diseases: Survival of cancer patients in Osaka 1975–89. Osaka Foundation for Prevention of Cancer and Circulatory Disease;998.
5 Kinoshita N, Ajiki W, Koyama Y, Hasuo S, Tanaka H, Tsukuma H, et al. Trends in relative 5-year survival rates of uterus cancer patients. Introduction of age standardization. Jpn J Cancer Clin 1998;44:1101–6. (in Japanese with English abstract)
6 Quinn M, Babb P, Jones J, Allen E. Effect of screening on incidence of and mortality from cancer of cervix in England: evaluation based on routinely collected statistics. Br Med J 1999;318:904–8.
7 Macgregor JE, Campbell MK, Mann EM, Swanson KY. Screening for cervical intraepithelial neoplasia in north east Scotland shows fall in incidence and mortality from invasive cancer with concomitant rise in preinvasive disease. Br Med J 1994;308:1407–11.
8 Devesa SS, Silverman DT, Young JL, Jr, Pollack ES, Brown CC, Horm JW, et al. Cancer incidence and mortality trends among whites in the United States, 1947–84. J Natl Cancer Inst 1987;79:701–70.[Web of Science][Medline]
9 Parkin DM, Whelan SL, Ferlay J, Teppo L, Thomas DB, editors. Cancer Incidence in Five Continents Vol. VIII. IARC Scientific Publication No.155. Lyon, France: International Agency for Research on Cancer;2002.
10 Curran DR, Stigleman S, Neher JO. Should we discontinue Pap smear screening in women aged >65 years? J Fam Pract 2004;53:308–10.[Medline]
11 Hartmann KE, Hall SA, Nanda K, Boggess JF, Zolnoun D. Screening for Cervical Cancer. Systematic Evidence Review number 25. Rockville, MD: Agency for Health Research and Quality;2002.
12 Young JL, Jr., Roffers SD, Ries LAG, Fritz AG, Hurlbut AA, editors. SEER Summary Staging Manual-2000: Codes and Coding Instructions, National Cancer Institute, NIH Publication No. 01-4969. Bethesda, MD;2001.
13 Estev J, Benhamou E, Raymond L. Statistical Methods in Cancer Research Vol. IV: Descriptive Epidemiology. IARC Scientific Publication No. 128. Lyon, France: International Agency for Research on Cancer;1994:231–45.
14 Ajiki W, Matsuda T, Sato Y, Fujita M, Yamazaki S, Murakami R, et al. Standard method of calculating relative survival rates in population-based cancer registries—an investigation using stomach cancer patients. Jpn J Cancer Clin 1997;43:1005–14. (in Japanese with English abstract)
15 Raffle AE, Alden B, Quinn M, Babb PJ, Brett MT. Outcomes of screening to prevent cancer: analysis of cumulative incidence of cervical abnormality and modelling of cases and deaths prevented. Br Med J 2003;326:901–4.
16 US Preventive Services Task Force. Guide to Clinical Preventive Services, 2nd edn. Baltimore: Williams & Wilkins;1996:105–17.
17 US Department of Health and Human Services: SEER*Stat 5.0, SEER Cancer Incidence Public-Use Database, 1973–2000. Bethesda: NIH;2003.
18 Sirovich BE, Welch HG. The frequency of Pap smear screening in the United States. J Gen Intern Med 2004;19:243–50.[Medline]
19 Herbert A, Breen C, Bryant TN, Hitchcock A, Macdonald H, Millward-Sadler GH, et al. Invasive cervical cancer in Southampton and South West Hampshire: effect of introducing a comprehensive screening programme. J Med Screen 1996;3:23–8.[Medline]
20 Coleman MP, Gatta G, Verdecchia A, Esteve J, Sant M, Storm H, et al. EUROCARE Working Group. EUROCARE-3 summary: cancer survival in Europe at the end of the 20th century. Ann Oncol 2003;14:128–49.
21 Marcus AC, Crane LA. A review of cervical cancer screening intervention research: implications for public health programs and future research. Prev Med 1998;27:13–31.[CrossRef][Web of Science][Medline]
22 Skinner CS, Strecher VJ, Hospers H. Physicians' recommendation for mammography: do tailored messages make a difference? Am J Public Health 1994;84:43–9.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  K. Aklimunnessa, M. Mori, M. M. H. Khan, F. Sakauchi, T. Kubo, Y. Fujino, S. Suzuki, S. Tokudome, A. Tamakoshi, and for the JACC Study Group Effectiveness of Cervical Cancer Screening Over Cervical Cancer Mortality Among Japanese Women Jpn. J. Clin. Oncol., August 1, 2006; 36(8): 511 - 518. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||