SEARCH

SEARCH BY CITATION

Keywords:

  • cervical cancer;
  • screening programme;
  • incidence;
  • mortality

Abstract

  1. Top of page
  2. Abstract
  3. Material and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

With respect to cervical cancer management, Finland and the Netherlands are comparable in relevant characteristics, e.g., fertility rate, age-of-mother at first birth and a national screening programme for several years. The aim of this study is to compare trends in incidence of and mortality from cervical cancer in Finland and the Netherlands in relation to the introduction and intensity of the screening programmes. Therefore, incidence and mortality rates were calculated using the Cancer Registries of Finland and the Netherlands. Data on screening intensity were obtained from the Finnish Cancer Registry and the Dutch evaluation centre at ErasmusMC-Rotterdam. Women aged 30–60 have been screened every 5 years, in Finland since 1992 and in the Netherlands since 1996. Screening protocols for smear taking and referral to the gynaecologist are comparable. Incidence and mortality rates have declined more in Finland. In 2003, age-adjusted incidence and mortality in Finland were 4.0 and 0.9 and in the Netherlands 4.9 and 1.4 per 100,000 woman-years, respectively. Excess smear use in the Netherlands was estimated to be 24 per 1,000 women during a 5-year interval compared to 121 in Finland. The decline in mortality in Finland seems to be almost completely related to the screening programme whereas in the Netherlands it was initially considered to be a natural decline. Differences in risk factors might also play a role: the Netherlands has higher population density and higher percentages of immigrants and (female) smokers. The greater excess smear use in Finland might also have affected incidence. © 2007 Wiley-Liss, Inc.

Mass screening for cervical cancer has been performed in several countries with varying success, depending on the coverage and intensity of screening such as intervals between screens, age groups covered, attendance rate, quality of follow-up after a positive smear, coordination of organized and opportunistic screening (screening outside the screening programme) and other characteristics.1-3 Actual proof of the effectiveness of cervical cancer screening was never obtained from randomized cervical cancer screening trials; instead, the evidence of the efficacy and effectiveness is based on cohort follow-up studies and also on geographical correlation studies.4, 5 Consequently, debate has arisen on the contribution of screening to the decrease in cervical cancer mortality that was found in some areas as well as to the prevention of an increase in other areas where cervical cancer mortality did not change significantly despite extensive screening.1, 6, 7 The objective of cervical cancer screening is to prevent the occurrence of invasive cancer and thus death by detecting and treating high-grade intraepithelial lesions, being precursors of invasive cancer. The most widely used screening approach to detect lesions has been cervical cytology, followed by investigation of “positive” women with colposcopy and directed biopsy.4, 8 Addition of an HPV test to the screening programme or substitution of the cytological test by the HPV-DNA test and even the inclusion of HPV vaccination in the national vaccination programmes are now under consideration in several countries with mass screening programmes.9-14

It is now well-established that HPV infection is the central causal factor in cervical cancer. HPV is a common sexually transmitted infection and both women and men are usually exposed to the virus after the onset of sexual intercourse. The risk of infection with HPV and also the risk of cervical cancer is increased by the number of sexual partners, age at first intercourse and sexual behavior of the woman's male partners.15, 16 Additional risk indicators for cervical cancer are the number of live births, long-term use of oral contraceptives, cigarette smoking and immunosuppression.16 The incidence of cervical cancer varies across the world depending on the presence of the above-mentioned risk factors and the availability of a screening programme.17

The aim of the present study was to compare trends in the incidence of cervical cancer in Finland and the Netherlands and to relate the trends to the extent and intensity of the screening programmes. Finland and the Netherlands are comparable in most other relevant characteristics, e.g. gross domestic product, fertility rate and age at first birth.18, 19

Material and methods

  1. Top of page
  2. Abstract
  3. Material and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

Study population

Age-specific and age-adjusted world standardized rates (WSR) for cervical cancer incidence and mortality were calculated from the Finnish Cancer Registry (FCR) and the Netherlands Cancer Registry (NCR).

The FCR is a population-based nationwide cancer registry established in 1952, registration began in 1953. The FCR receives data on cancer cases from hospitals, health centres, medical practitioners and pathological and cytological laboratories. It also receives information about all death certificates which mention cancer. The file of all deaths occurring in Finland is checked annually against the files of the FCR.

Nine regional cancer registries submit their data to the NCR, which has been a population-based nationwide cancer registry since 1989. Registration in the Netherlands began in the region of the Eindhoven Cancer Registry in 1955 and was followed by the other regions during the 1980s. Notification is obtained from the Pathology Departments, the Dutch Network and National database for Pathology (PALGA), and Hematology Departments. Other sources are the Radiotherapy Departments of the hospitals, as well as the National Registry of Hospital Discharge Diagnosis, which accounts for up to 8% of new cases apparently without pathological notification.20 Death certificates are not available in an identifiable form to the regional cancer registries and the NCR. Data on deaths from cervical cancer were therefore derived from Statistics Netherlands.21

In the NCR, carcinoma in situ of the cervix is not registered; accordingly, also for Finland only cases of invasive cervical cancer were included in this study.

Mapping

Maps were made based on truncated age-adjusted (World Standard) incidence rates per 100,000 women. For cities with >100,000 inhabitants the rates were shown as such as circles on the maps. The radius of the circle indicates the size of the population and the color the WSR. The remaining rates were smoothed to prevent disturbing chance variation.22 Smoothing was done by calculating a weighted average of the age-adjusted incidence rates of the neighboring areas for each grid (size 2 km by 2 km) to define the color of that grid. The weights were inversely associated with the distance and reached 50% at a distance of 25 km and zero at 150 km. Every step between the categories represents a 1.13-fold increase in the rate.

History of screening programmes

In 1963 an organized cervical screening programme with 5-year intervals was introduced as a pilot project in 3 municipalities in Finland, extending to most parts of the country within a few years. By 1970, the coverage of the invitational programme already exceeded 80% of women in the target age group of 30–50 years. Later, in 1992 30–60 years became the national target age. Some municipalities also invited women 25–30 and/or 60–65 years old. Furthermore, from the early 1970s onwards, registered coverage has become almost complete. According to a bylaw drawn up in 1992 the municipalities had to offer cervical cancer screening to 30- to 60-year-old women with a 5-year screening interval.23

In the Netherlands, cytological screening has been available to women in some regions of the Netherlands since the mid 1970s within a combination of opportunistic screening and local and regional invitational programmes, with 3-year intervals.24 In 1988, a nationwide screening programme was initiated aimed at women aged 35–54 years screened at 3-year intervals.25 In 1996 screening activities were restructured for a new national programme. From then on women between 30 and 60 years old were screened at 5-year intervals, leaving the number of 7 invitations during a lifetime unchanged.26

Screening practices

In the Finnish screening programme, smears are taken out by trained nurses (midwives) in local health care centres and the smears are screened by cytotechnicians. Smear quality is under continuous control and assessed by the cytology laboratories. The cytologist checks every abnormal smear and a proportion of normal smears.23 Referral to the gynaecologist for colposcopy and biopsy takes place after a clear finding of dysplasia (Pap III-V) or after several borderline findings (Pap II), based on the recommendation of the cytologist (Table I).

Table I. Comparison of Pap Smear Classification and Bethesda System
WHO terminologyPap score27Bethesda28
Atypical cellsPap IIASCUS/AGUS
Mild/moderate dysplasiaPap IIIALSIL
Severe dysplasiaPap IIIBHSIL
Carcinoma-in-situPap IVHSIL/AIS
Squamous cell carcinoma/AdenocarcinomaPap VSquamous cell carcinoma/adenocarcinoma

In the Netherlands, most of the programme smears are taken in general practice, by general practitioners or their practice assistants. Investigation of the smears is performed by specially trained cytotechnicians. The nonnegative cases are evaluated by a head-cytotechnician. A cytopathologist has the final supervision and writes the final report on the nonnegative cases. Screening results are filed at the laboratories and in PALGA. Comparable to Finland, referral to the gynaecologist takes place after repeated borderline findings (Pap II or IIIA) or after finding positive cytology (Pap IIIB-V).

Participation

In Finland the participation rate exceeded 70% (72% in 2004) but in addition to the organized screening programme, opportunistic smears are also more common. It has been estimated that the coverage of smears during a 5-year period was about 90% (i.e. at least 1 smear made per female) and the coverage of women with a Pap-smear at least once in their lifetime was estimated to be 98%.23 Opportunistic screening is estimated to be more than 100% of the total screening activity (i.e. the use of smears is 2 times higher than recommended by the programme) and the excess use of Pap smears (all smears taken in a certain period that do not contribute to the observed coverage of the target population) was 121 per 1,000 women in a 5-year period.2, 23

In the Netherlands the participation rate was about 65%.29 At the end of 1996, together with opportunistic screening, the percentage of women 30–60 years old with at least 1 smear in the previous 5 years was ∼80% and coverage of women with a Pap-smear at least once in their lifetime was 90%.26 However, participation in each following round of the population-based screening programme became lower; furthermore and with increasing age, attendance declined, being 10–15% lower for the age group 50–59 than the age group 30–49.30, 31 The excess use of Pap smears was very low: 24 per 1,000 women in a 5-year period with opportunistic screening being <2% of the total screening activity in 2003.2, 29

Results

  1. Top of page
  2. Abstract
  3. Material and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

During the period 1955–1964, the incidence of invasive cervical cancer in Finland was 15 per 100,000 woman-years, age-adjusted to the world standard population, with a slightly increasing trend within that period (Fig. 1). From 1965 to 1990 incidence rapidly decreased to 2.8 per 100,000 woman-years in 1991, which gives an overall decrease of ∼70–80%. After 1991, the incidence increased to 4.0 per 100,000 woman-years in 2003. The mean age-specific incidence rate for the age group 25–39 increased from 2.1 in 1989–1994 to 4.6 in 1998–2003; it was higher in urbanized areas (Fig. 2a). Mortality from cervical cancer decreased continuously from 6.8 per 100,000 woman-years during 1958–1962 to 0.9 per 100,000 woman-years in 2003 (Fig. 3).

thumbnail image

Figure 1. Age-adjusted (world standard) incidence rates per 100,000 women in Finland and the Netherlands, according to year of diagnosis.

Download figure to PowerPoint

thumbnail image

Figure 2. (a) Truncated age-adjusted (world standard) incidence rates per 100,000 women in Finland, age group 25–39 years, in the period 1989–1994 (left) and the period 1998–2003 (b) Truncated age-adjusted (world standard) incidence rates per 100,000 women in the Netherlands, age group 25-39 years, in the period 1989–1994 (left) and the period 1998–2003.

Download figure to PowerPoint

thumbnail image

Figure 3. Age-adjusted (world standard) mortality rates per 100,000 women in Finland and the Netherlands, according to year of death.

Download figure to PowerPoint

In the Netherlands, the incidence of cervical cancer varied from 18 to 12 per 100,000 woman-years, age-adjusted to the world standard population, in the period 1960–1970 (Eindhoven registry) (Fig. 1). In 1970 incidence started to decrease from 12 per 100,000 woman-years (Eindhoven registry) to 3.6 per 100,000 woman-years in 2003 (national rates: 4.9 per 100,000). The mean age-specific incidence rate in the Netherlands for the age group 25–39 decreased from 11.6 in the period 1989–1994 to 9.8 in the period 1998–2003 with higher incidences in the cities and more urbanized areas (Fig. 2b). These differences were bigger than in Finland. Mortality from cervical cancer in the Netherlands decreased from 5.4 per 100,000 woman-years in 1970 to the lowest rate ever, 1.1 per 100,000 woman-years, in 2002 (Fig. 3).

Discussion

  1. Top of page
  2. Abstract
  3. Material and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

There is worldwide evidence of a considerable decline in incidence and mortality from cervical cancer in areas with active mass screening programmes. However, the effectiveness of these programmes varies, and initially the trends in the disease were hardly affected by screening in some countries such as the United Kingdom and Scotland.1 This study has shown similarities, but also differences in the trends in incidence of and mortality from cervical cancer between Finland and the Netherlands. In Finland, the 80% decrease in the incidence and mortality rates has mainly resulted from the national screening programme, even though wide-spread opportunistic screening could also have affected these rates but in a clearly less degree.5, 23, 32, 33 Changes in sexual behavior probably increased rather than decreased the background risk.34 Mortality in the Netherlands started to decrease in 1970, earlier than an effect of mass screening could be expected, but definitely continued to drop after the introduction of screening, being currently 40% lower than around 1960. A drop in the incidence of cervical cancer and a shift towards diagnosis of invasive cervical cancer in a less advanced stage seems the best explanation for the “natural” decline.35, 36 In Finland, the death records have been continuously linked with the incidence records on an individual level. In the Netherlands, unfortunately there is no linkage of the cancer registry with the cause-of-death registry. However after 1970, there were very few uterus NOS cases of cancer in the Netherlands and therefore the trends in mortality were affected negligibly.21 Differences in mortality rates may also partly be explained by a difference in background incidence between Finland and the Netherlands, the latter having more urbanized areas and more migration from abroad than Finland. These migrant women originate mainly from countries with higher cervical cancer incidence rates, the WSR per 100,000 women in 2002 being 12 for Morocco, 16 for Indonesia, 13 for women from the Caribbean islands to even 27 for Suriname37, 38; 4% of all cervical cancer cases in the Netherlands are diagnozed in immigrant women.39

Part of the difference in the decrease of incidence of cervical cancer could be explained by risk factors for cervical cancer. In contrast to Finland, the Netherlands does have a very high population density in most of the country and a higher percentage of (female) smokers, both of which are risk factors for cervical cancer.21 Furthermore, the fertility rate was higher, especially in catholic parts of the Netherlands up until 1970.21

In Finland, a recent increase was observed in the incidence of cervical cancer in young women, possibly related to changes in sexual behavior during the last few decades, suggesting an increasing role of some potentially oncogenic sexually transmitted infections, such as HPV.21, 34 Also among young Finnish women smoking increased during the 1980s. In the Netherlands, an increase in the incidence of cervical cancer could not yet be observed.40 However, HIV and other sexually transmitted infections have been increasing, according to the latest surveillance data.41

In both countries, discussion has started on a Quadrivalent Human Papillomavirus (Types 6, 11, 16, 18) Recombinant vaccine, which has been approved in the EU for prescription to women aged 9 years and older, to include it in the vaccination programmes to reduce further (and definitely) the incidence of cervical and other HPV-related cancers.9, 12, 42 However, follow-up studies of the vaccines on the effect on cancer risk, i.e. the problems and limitations of the vaccine, are not available. Furthermore, vaccination may create a false feeling of complete protection. The screening programme therefore may need to be continued in addition to the vaccination for several decades more.

The percentages of hysterectomy found for both countries might also explain the difference in incidence rates. Hysterectomy has been under discussion since the 1970s because of regional and international variations in frequency, indications and surgical methods.43 In some countries, the increasing frequency of hysterectomy has led to an underestimation of the actual risk for cervical cancer. A study from Finland calculated 11% higher hysterectomy-corrected rates than the uncorrected rates.43 The prevalence of hysterectomy in Finland in 1987–1989 was approximately the same as in the Netherlands. However, from 1991 to 1999 the annual number of hysterectomies in Finland increased by 16% and decreased by 24% in the Netherlands in almost the same period.43, 44 Therefore, the adjustment for age-specific fractions of women with a hysterectomy may have had some influence on the cervical cancer rates in both countries: the corrected incidence rates for Finland being somewhat higher and for the Netherlands somewhat lower.

To conclude, incidence and mortality rates of cervical cancer rates became very low in both Finland and the Netherlands, in fact the lowest in Europe due to a large extent their mass screening programmes. In the Netherlands, there still seems to be some room for improvement, whereas Finland might need to pay more attention to young women and the high rate of excess smear use. As long as cervical cancer occurs in women who are screened, most attention must be directed toward minimizing false-negative smears. Furthermore, anxiety might be caused in women by repeated testing of low-grade cervical abnormalities and colposcopic evaluation of high-grade abnormalities.4, 45 Unfortunately, more intensive screening greatly increases the need for more interventions for lesions which would never have developed into tumors.46 Attention to quality of life and potential adverse aspects should therefore be part of the evaluation of screening programmes.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Material and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

We thank Mr. Toni Patama from Kuopio University, Finland, for making the maps.

References

  1. Top of page
  2. Abstract
  3. Material and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References