  |
|
|
||||||||||||||
|
||||||||||||||||
| Journal of the New Zealand Medical Association,
11-February-2005, Vol 118 No 1209
|
||||||||||||||||
|
The impact of breast cancer screening on breast cancer
registrations in New Zealand
Ann Richardson, Brian Cox, Thelma Brown, Paul Smale
BreastScreen Aotearoa started in December 1998. Up to June
2004, when the age-range was increased to 45–69 years,
BreastScreen Aotearoa has offered 2-yearly two-view mammography to New
Zealand women aged 50–64 years, with the aim of reducing the mortality
from breast cancer in this population. Results from randomised controlled trials
of breast cancer screening have shown that well-organised programmes can reduce
deaths from breast cancer by about one-third among women aged 50 years and over
who are offered screening.1,2
Among women who accepted the offer of screening in these
trials, breast cancer deaths were reduced by as much as
40%.3 The effect of screening in the routine
healthcare setting is likely to be less, but the early results of at least one
organised screening programme (albeit with very high participation of 85%)
suggest a 24% reduction in breast cancer mortality is
attainable.4 The benefits of screening for
women aged under 50 years are
uncertain.5–7
Breast cancer screening should advance the time of diagnosis
by detecting breast cancer earlier than it would otherwise be diagnosed. An
indicator that a screening programme detects breast cancer early is an increase
in breast cancer incidence in the first (prevalence) screening round. In the
prevalence round of a new screening programme, most women have not previously
had mammograms. If screening detects breast cancer early, it will identify
tumours that may otherwise not have been diagnosed until some time later.
The early detection of these tumours causes an initial rise
in the detection rate of breast cancers in the age group offered screening, to
2.5 to 3.0 times the pre-screening incidence. Some over-diagnosis of invasive
breast cancer may occur. That is, some cancers may be detected which would not
have presented as symptomatic disease because of their extremely slow growth.
This may be about 6% of invasive cancers detected in a breast cancer screening
programme.7
Another important indicator is the stage at diagnosis. If
screening detects cancer early, a shift in the stage distribution towards
earlier stage tumours would be expected, compared with the stage distribution of
tumours diagnosed before the advent of widespread breast cancer screening in New
Zealand.
If BreastScreen Aotearoa detected breast cancer early, there
should have been an increase in the incidence of breast cancer in women aged
50–64 years during the prevalence screening round. Pilot breast cancer
screening programmes had started earlier (during 1991 and 1992), but the pilot
programmes screened only about 12% of New Zealand women aged 50–64
years.
MethodsThe degree of spread of breast
cancers at diagnosis was available from the New Zealand Cancer Registry. We
have used the term ‘degree of spread’ rather than
‘stage’, to avoid confusion—because the information on stage
of disease reported by the Cancer Registry is not the TNM stage. The stage
reported by the Cancer Registry is closely related to the clinical stage of
disease at diagnosis, and has been used to assess the effects of early detection
in the breast screening programme.
Age-specific breast cancer incidence rates for New
Zealand women aged 50–64 years were compared for the years before and
after the establishment of
BreastScreen Aotearoa.8 The degree of
spread of breast cancer diagnosed in women screened in BreastScreen Aotearoa was
compared with the degree of spread of breast cancers registered during
1979–1988 before the introduction of the pilot programmes or the
establishment of BreastScreen Aotearoa.
ResultsIn the year (1999) following the
establishment of BreastScreen Aotearoa, there was a marked increase in the
age-specific incidence of breast cancer in New Zealand women aged 50–64
years (Figure 1).
Figure 2 shows a shift towards less spread of disease at
diagnosis for screened women, compared with women in the same age range
(50–64 years) before the introduction of population screening for breast
cancer in New Zealand.
DiscussionThe increase in the age-specific
incidence of breast cancer in New Zealand women aged 50–64 years in the
first year of screening suggests that BreastScreen Aotearoa detected breast
cancer early in some women.
There was a shift towards earlier diagnosis of invasive
cancer for screened women compared with women in the same age range (50–64
years) before the introduction of population screening for breast cancer in New
Zealand. BreastScreen Aotearoa detected a higher proportion of localised
invasive breast cancers compared with 1979–1988. Only a small proportion
of this is likely to be due to over-diagnosis of extremely slow-growing invasive
breast cancer.7 This comparison is between
screened women and an unscreened population, and there may be selection effects,
whereby women who choose to be screened are those who would be more likely to
present early in the absence of screening. Selection effects have been detected
in randomised controlled trials of
screening.3
Figure 1. Age-specific breast cancer incidence in New
Zealand before and during the prevalence screening round of BreastScreen
Aotearoa
 Figure 2. Stage distribution of invasive breast cancer
diagnosed in BreastScreen Aotearoa during 1999-2001, compared with the
stage distribution of invasive breast cancer diagnosed in New Zealand women aged
50–64 years during 1979-1988 (before the establishment of
screening)
 Figure 2.
Continued
χ2=271.1;
df =3; p<0.0001; BSA= BreastScreen Aotearoa.
Unfortunately, adjustments for selection cannot be made for
BreastScreen Aotearoa; there is no population register of women eligible for
screening, and considerations of privacy have meant that no data are recorded
about women who have declined an offer of screening.
During 1999–2001, 56% of eligible women were screened
in BreastScreen Aotearoa. Only a major selection effect could completely explain
the shift seen in Figure 2, if screening had no impact. BreastScreen Aotearoa
also detected a similar proportion of node-negative invasive tumours to the
Swedish Two-County Trial of breast cancer screening, with 75% of invasive
tumours being node-negative.9 The Swedish
Two-County Trial resulted in a significant reduction in breast cancer mortality
among women offered screening, and the results from this trial have been used to
develop international targets for breast cancer screening. The similarity in the
nodal status of invasive tumours detected by BreastScreen Aotearoa and the
Swedish Two-County Trial is encouraging.
Other indicators of the impact of BreastScreen Aotearoa were
being monitored for women screened up to June 2002, and have been published in
the reports of the BreastScreen Aotearoa Independent Monitoring Group until
early 2003.10
It is impossible to
prove that any national screening
programme reduces breast cancer mortality, since there is no control group. This
is why internationally, screening programmes measure their progress against
targets derived from randomised controlled trials. Because any impact of breast
cancer screening on breast cancer mortality is not immediate, it is important to
measure early indicators such as changes in breast cancer registration. The
changes in patterns of breast cancer registration following the introduction of
BreastScreen Aotearoa reported here, do not prove that the programme will reduce
breast cancer mortality, but they provide an encouraging early indication. A
breast cancer screening programme that did not show these effects would be
unlikely to achieve a reduction in breast cancer mortality in the longer
term.
Author information:
Ann Richardson, Senior Lecturer in Epidemiology, Department of Public
Health and General Practice, Christchurch School of Medicine and Health
Sciences, University of Otago, Christchurch; Brian Cox, Director; Thelma Brown,
Research Fellow; Paul Smale, Research Fellow, Hugh Adam Cancer Epidemiology
Unit, Dunedin School of Medicine, University of Otago, Dunedin
Correspondence: Dr
Ann Richardson, Department of Public Health and General Practice, Christchurch
School of Medicine and Health Sciences, PO Box 4345, Christchurch. Fax: (03) 364
3614; email: ann.richardson@chmeds.ac.nz
References:
|
|
||||||||||||||
| | Current
issue | Search journal |
Archived issues | Classifieds
| Hotline (free ads) Subscribe | Contribute | Advertise | Contact Us | Copyright | Other Journals |
|