Tubal sterilization and risk of ovarian, endometrial and cervical cancer. A Danish population-based follow-up study of more than 65 000 sterilized women

Article (PDF Available)inInternational Journal of Epidemiology 33(3):596-602 · June 2004with24 Reads
DOI: 10.1093/ije/dyh046 · Source: PubMed
Abstract
On the basis of a population-based cohort, we assessed the cancer risk, focusing on gynaecological cancers and pre-malignant lesions, among women with a previous tubal sterilization. Using the Danish Hospital Discharge Register we identified 65 232 women who had a tubal sterilization (1977-1993). The cohort was followed for cancer occurrence, and compared with the expected number based on the national cancer incidence rates. The overall risk of ovarian cancer was decreased (standardized incidence ratio [SIR] = 0.82; 95% CI: 0.6, 1.0), and it was still decreased > or =10 years after the sterilization (SIR = 0.65; 95% CI: 0.4, 1.0). The rate of endometrial cancer was also decreased (SIR = 0.66; 95% CI: 0.5, 1.0), the risk continued being moderately reduced during follow-up, although it was not statistically significant. In this nationwide, population-based study we find that women with tubal sterilization have a decreased risk of subsequent development of ovarian cancer. As the protective effect is not decreasing with years of follow-up, our data do not support that 'screening' bias can explain the protective effect, but indicate that the sterilization itself may convey a reduction in risk. The same pattern is found for endometrial cancer, the association being less strong.
IJE vol.33 no.3 © International Epidemiological Association 2004; all rights reserved. International Journal of Epidemiology 2004;33:596–602
Advance Access publication 26 May 2004 DOI: 10.1093/ije/dyh046
Tubal sterilization and risk of ovarian,
endometrial and cervical cancer. A Danish
population-based follow-up study of more
than 65 000 sterilized women
Susanne K Kjaer,
1
Lene Mellemkjaer,
1
Louise A Brinton,
2
Christoffer Johansen,
1
Gloria Gridley
2
and Jørgen H Olsen
1
Accepted 6 October 2003
Background On the basis of a population-based cohort, we assessed the cancer risk, focusing
on gynaecological cancers and pre-malignant lesions, among women with a
previous tubal sterilization.
Methods Using the Danish Hospital Discharge Register we identified 65 232 women who
had a tubal sterilization (1977–1993). The cohort was followed for cancer
occurrence, and compared with the expected number based on the national
cancer incidence rates.
Results The overall risk of ovarian cancer was decreased (standardized incidence ratio
[SIR] = 0.82; 95% CI: 0.6, 1.0), and it was still decreased 10 years after the
sterilization (SIR = 0.65; 95% CI: 0.4, 1.0). The rate of endometrial cancer was
also decreased (SIR = 0.66; 95% CI: 0.5, 1.0), the risk continued being
moderately reduced during follow-up, although it was not statistically significant.
Conclusions In this nationwide, population-based study we find that women with tubal
sterilization have a decreased risk of subsequent development of ovarian cancer.
As the protective effect is not decreasing with years of follow-up, our data do not
support that ‘screening’ bias can explain the protective effect, but indicate that
the sterilization itself may convey a reduction in risk. The same pattern is found
for endometrial cancer, the association being less strong.
Keywords Tubal sterilization, cohort study, risk, gynaecological cancer
596
1
Danish Cancer Society, Institute of Cancer Epidemiology, Copenhagen,
Denmark.
2
Division of Cancer Epidemiology and Genetics, National Cancer Institute,
Bethesda, MD, USA.
Correspondence: Susanne K Kjaer, Danish Cancer Society, Institute of Cancer
Epidemiology, Strandboulevarden 49, DK-2100, Copenhagen Ø, Denmark.
E-mail: susanne@cancer.dk
A history of tubal sterilization has been associated, in some
studies, with a reduced risk of ovarian cancer.
1–7
In at least two
other studies, however, no such association was seen.
8,9
In
addition, there has been little consensus with regard to ovarian
cancer risk in relation to years of follow-up.
Only a few studies have reported on the risk of other types of
cancer. In one case-control study the risk of endometrial cancer
was moderately, but non-significantly, decreased,
10
while two
other case-control studies showed a non-significantly increased
risk for this cancer type.
11,12
In yet another study, a small, non-
significant reduction of invasive cervical cancer was reported,
particularly during the first 5–10 years following the tubal
sterilization.
13
In relation to both ovarian cancer and endometrial and breast
cancer, it has been hypothesized that the decreased risk, if
confirmed, could be related to alterations in the levels of
endogenous hormones (oestrogen, progesterone), which may
occur following tubal sterilization.
14–16
A possibly decreased risk
of cervical cancer has been related to a screening effect as tubal
ligation may provide an opportunity for Pap smear screening and
thereby secondary prevention of cervical cancer.
13
The purpose of the present study was to examine the risk of
gynaecological cancer in a cohort of more than 65 000 Danish
women who had a tubal sterilization performed during
1977–1993. In addition, we report on the risk of ovarian border-
line tumours and cervical intraepithelial neoplasia grade 3
by guest on June 6, 2013http://ije.oxfordjournals.org/Downloaded from
TUBAL STERILIZATION AND GYNAECOLOGICAL CANCERS 597
(CIN3), which to our knowledge has not previously been
evaluated in follow-up studies.
Methods
The population-based Danish Hospital Discharge Register
(HDR) keeps a record of virtually all (~99%) the somatic
hospitalizations in Denmark since 1977.
17
This computerized,
central register contains, for each discharge, information on the
personal identification number of the patient, dates of
admission/discharge, up to 20 diagnoses, and every surgical
procedure performed during the respective admission. The
surgical procedures are recorded and classified according to The
Danish Classification of Surgical Procedures and Therapies.
18
There were 65 366 women (45 years) who underwent tubal
sterilization (operation codes: 60800 60810 60820 60830
60840) in the time period 1977–1993. On these women we also
collected information on hysterectomy (operation codes: 61000
61020 61040) and bilateral oophorectomy (operation codes:
60120, 60320).
Since 1968, all Danish inhabitants have been assigned a
unique personal identification number, comprising information
on sex and date of birth, and these identification numbers are
registered in the computerized Central Population Register. The
number allows correct linkage of information between different
registers. The study cohort was linked to the Central Population
Register to verify the personal identification number and to
obtain information on date of death or date of migration, when-
ever appropriate. A total of 134 women (0.2%) were excluded
because of invalid personal identification number.
To ascertain the cancer occurrence in the cohort, it was linked
to the Danish Cancer Registry, which contains information on
all cases of cancer in Denmark since 1943, including bladder
papillomas and benign brain tumours. With regard to the
cervical precancerous lesions, carcinoma in situ and severe
dysplasia (CIN 3) may be close to complete registration, the
notification of moderate and mild dysplasia is incomplete.
19
Since 1987, it has been mandatory to report cancer cases as well
as precancerous lesions to the Danish Cancer Registry. The
cancers are classified according to a modified version of the
Seventh Revision of the International Classification of Diseases
(ICD).
19
The 65 232 women were followed for cancer occurrence
(including borderline ovarian tumours and CIN3) from the date
of the hospital discharge for tubal sterilization until the date of
emigration, date of death, or 31 December 1995, whichever
came first. Additional exit dates were applied when the risk of
certain cancers was examined. Thus, follow-up for endometrial
cancer was ended at the date of hysterectomy, follow-up for
cervical cancer/CIN3 was discontinued at the date of total
hysterectomy, and follow-up for ovarian cancer/borderline
tumour was ended at the date of bilateral oophorectomy.
The observed number of cancer cases was compared with the
number of expected cases based on the age- and calendar year-
specific cancer incidence rates for women from the Danish
Cancer Registry and accumulated person-years. Standardized
incidence ratios (SIR) (observed number of cases divided by the
expected number) were computed with corresponding 95% CI
for each type of cancer, assuming a Poisson distribution of the
observed cancer cases.
20
Results
The study cohort of 65 232 sterilized women accrued 643 761
person-years during follow-up. The mean follow-up time was
9.9 years (range: 0–19 years). Less than 2% of the women
were 24 years at the time of sterilization, while 15% were
25–29 years, 30% were 30–34 years, 35% 35–39 years, and
about 18% were 40–45 years when they had their tubal
sterilization. In the follow-up for ovarian cancer/borderline
tumours, endometrial cancer and invasive cervical cancer/CIN3,
respectively, 641 702, 604 248, and 605 631 person-years were
accrued. Thus, censoring the observation time subsequent to
bilateral oophorectomy, hysterectomy, and total hysterectomy
reduced the total number of person-years by 0.3%, 6.1%, and
5.9%, respectively. For 99.1% of the women, the type of tubal
sterilization was only registered as laparoscopic sterilization
without further specification.
A total of 1894 cancers were observed among the 65 232
women who previously had a tubal sterilization. This was
slightly less than expected (1981 cases) on the basis of the rates
from the general female population (SIR = 0.96; 95% CI: 0.9,
1.00) (Table 1). None of the non-gynaecological cancers
contributed in particular to this lower overall cancer rate,
except for breast cancer, which occurred less frequently than
expected (SIR = 0.9; 95% CI: 0.9, 1.0). In contrast, the group of
gynaecological cancers played a role as a total of 277 cases was
observed in this population against 328.0 expected (SIR = 0.84:
95% CI: 0.8, 1.0).
In Tables 2 and 3, the SIR are presented for specific types of
gynaecological cancers as well as for the ovarian borderline
tumours and CIN3. The overall risk of ovarian cancer was
decreased (SIR = 0.82), although the association only reached
borderline statistical significance (95% CI: 0.6, 1.0). However,
when considering the groups with long-term follow-up, there
was still a pattern of a decreased risk. Ten years or longer after
the sterilization the SIR was 0.65 (95% CI: 0.4, 1.0) (Table 2),
and even in women followed for 15 years after the tubal
sterilization, the risk tended to be reduced, although this was
not statistically significant (SIR = 0.65; 95% CI: 0.2, 1.4) (data
not shown). We also looked at the risk of ovarian cancer in
relation to age at sterilization. We observed the same decreased
ovarian cancer risk in women 35 years of age at time of
sterilization (SIR = 0.78; 95% CI: 0.5, 1.2) as in women who
were 35 years (SIR = 0.83; 95% CI: 0.6, 1.1) (data not
shown).
A total of 21 ovarian borderline tumours were observed in
this cohort, yielding a non-significantly decreased risk (SIR =
0.82; 95% CI: 0.5, 1.3). The risk estimates did not change much
with years of follow-up after the sterilization (Table 2).
The risk of ovarian cancer of different histological types
(serous, mucinous, clear cell/endometrioid, others) was also
assessed (Table 2). For both the group of serous invasive tumours
as well as for the group of ‘other’ ovarian cancers, the overall risk
was decreased; SIR = 0.72 (95% CI: 0.5, 1.0) and SIR = 0.60
(95% CI: 0.3, 1.0), respectively. In addition, the clear
cell/endometrioid group was non-significantly decreased (SIR =
0.86; 95% CI: 0.5, 1.44). In contrast, the mucinous ovarian
cancers were more common than expected, SIR = 1.49 (95% CI:
0.9, 2.3). For women with 5 years of follow-up, this risk of
mucinous ovarian carcinoma increased to 1.86 (95% CI: 1.1, 3.0)
by guest on June 6, 2013http://ije.oxfordjournals.org/Downloaded from
598 INTERNATIONAL JOURNAL OF EPIDEMIOLOGY
The rate of endometrial cancer was also decreased in women
who had a tubal sterilization (SIR = 0.70; 95% CI: 0.5, 1.0) and
except for the first year after the operation. The risk continued
being moderately reduced during follow-up, however the associa-
tions were not statistically significant (Table 3).
The observed number of cervical cancer cases was only
slightly lower than expected, both overall and during follow-up,
except for the first year where a non-significantly increased risk
was observed (Table 3). In contrast, CIN3 lesions were slightly in
excess (SIR = 1.06; 95% CI: 1.0, 1.1), the risk being significantly
Table 1 Standardized incidence ratios (SIR) for gynaecological cancer and specified groups of non-gynaecological cancer following tubal
sterilization among 65 232 Danish women
Cancer site Observed Expected SIR 95% CI
All malignant neoplasms 1894 1980.7 0.96 (0.9, 1.0)
Gynaecological cancers 277 328.0 0.84 (0.8, 1.0)
Non-gynaecological cancers 1617 1652.7 0.98 (0.9, 1.0)
Buccal cavity and pharynx 17 19.5 0.87 (0.5, 1.4)
Digestive organs 134 146.8 0.91 (0.8, 1.1)
Lung 89 99.0 0.90 (0.7, 1.1)
Breast 630 988.4 0.92 (0.9, 1.0)
Urinary system 45 41.2 1.09 (0.8, 1.5)
Melanoma of the skin 143 129.1 1.11 (0.9, 1.3)
Other skin 301 286.6 1.05 (0.9, 1.2)
Brain and nervous system 81 82.7 0.98 (0.8, 1.2)
Thyroid 26 21.5 1.21 (0.8, 1.8)
Lymphatic and haematopoietic 96 81.0 1.18 (1.0, 1.5)
Other specified 24 25.0 0.96 (0.8, 1.8)
Secondary and unspecified 30 31.1 0.97 (0.7, 1.4)
Table 2 Standardized incidence ratios (SIR) for ovarian cancer and borderline ovarian tumour by length of follow-up after tubal sterilization
Ovarian cancer Borderline ovarian tumour
Years since sterilization Obs. Exp. SIR 95% CI Obs. Exp. SIR 95% CI
All histological types
1 4.7 0.43 (0.1, 1.6) 1.5
1–4 17 24.0 0.71 (0.4, 1.1) 7.5 0.94 (0.4, 1.9)
5–9 37 34.1 1.08 (0.8, 1.5) 9.5 0.84 (0.4, 1.7)
10 19 29.1 0.65 (0.4, 1.0) 7.0 0.86 (0.3, 1.9)
Total 75 91.8 0.82 (0.6, 1.0) 25.5 0.82 (0.5, 1.3)
Serous type
4 9 11.3 0.80 (0.4, 1.5)
5 17 24.9 0.68 (0.4, 1.1)
Total 26 34.2 0.72 (0.5, 1.1)
Mucinous type
4 3 4.1 0.73 (0.2, 2.1)
5 16 8.6 1.86 (1.1, 3.0)
Total 19 12.8 1.49 (0.9, 2.3)
Clear cell/endometriod type
4 1 4.6 0.22 (0.0, 1.2)
5 13 11.7 1.11 (0.6, 1.9)
Total 14 16.3 0.86 (0.5, 1.4)
Other types
4 6 8.6 0.70 (0.3, 1.5)
5 10 18.0 0.56 (0.3, 1.0)
Total 16 26.6 0.60 (0.3, 1.0)
by guest on June 6, 2013http://ije.oxfordjournals.org/Downloaded from
TUBAL STERILIZATION AND GYNAECOLOGICAL CANCERS 599
increased within the first year after sterilization and also in the
group with the longest follow-up.
Discussion
In the present study following more than 65 000 women with a
tubal sterilization, we find that tubal sterilization is asso-
ciated with a reduced risk of ovarian cancer. This is in line with
both early studies (summarized by Hankinson et al.
2
and
Miracle-McMahill et al.
6
) and with those most recently
published
3–7
(Table 4). It has been suggested that screening of
the ovaries during the surgical procedure may be the main
reason for the protective effect of sterilization. If true, it should
be anticipated that the protective effect would decrease with
number of years since the operation. Our results disagree with
this hypothesis, as the risk remained decreased even after
10–15 years of follow-up. The results are in line with those of
case-control studies showing a protective effect that remained
Table 3 Standardized incidence ratios (SIR) for endometrial and cervical cancer and cervical intraepithelial neoplasia grade 3 (CIN3) by length of
follow-up after tubal sterilization
Endometrial cancer Cervical cancer CIN3
Years since sterilization Obs. Exp. SIR 95% CI Obs. Exp. SIR 95%CI Obs. Exp. SIR 95% CI
1 1.6 1.29 (0.1, 4.7) 15.7 1.21 (0.7, 1.9) 122.2 1.70 (1.5, 2.0)
1–4 4 9.0 0.44 (0.1, 1.1) 62 63.7 0.97 (0.8, 1.3) 362 417.3 0.87 (0.8, 1.0)
5–9 15.8 0.70 (0.4, 1.2) 48 56.9 0.84 (0.6, 1.1) 284 279.4 1.02 (0.9, 1.1)
10 13 16.3 0.80 (0.4, 1.4) 26 27.7 0.94 (0.6, 1.4) 124 100.7 1.23 (1.0, 1.6)
Total 30 42.7 0.70 (0.5, 1.0) 155 164.1 0.94 (0.8, 1.1) 919.5 1.06 (1.0, 1.1)
Table 4 Recent studies of the association between tubal sterilization and ovarian cancer
Size of study.
Study design Frequency of
Author Study period exposure Measures of association Comments
Rosenblatt et al., 1996 case-control 385 cases OR
a
adj = 0.71 (0.47, 1.08) Adjusted for: parity and OC
b
use,
WHO Collaborative Study (hospital-based) (104 borderline) No trend with time since Contraceptive use.
1979–1988 2563 controls tubal sterilization Estimates based on 1 exposed
34 cases (8.8%) and Clear cell: case for both clear cell tumours
426 controls (17.1%) OR = 0.33 (0.007, 2.68) and for Endometrioid tumours
with tubal Endometrioid:
sterilization OR = 0.21 (0.048, 1.49)
Cornelison et al., 1997 (USA) case-control 300 cases ORadj = 0.52 (0.31, 0.85) Adjusted for: socioeconomic
(hospital-based) 606 controls For 5–20 years after tubal level, marital status, parity,
1982–1988 26 cases (9%) and Sterilization, risk tended to age at first pregnancy,
93 controls (15.5%) be more decreased than Menarche, menopause,
with tubal sterilization for 21 years irregular menses, breast
feeding, body habitus,
and OC use.
Green et al., 1997 (Australia) case-control 824 cases ORadj = 0.61 (0.46, 0.85) Adjusted for: age, education,
(controls randomly 855 controls Risk was decreased even body mass index, parity,
selected from 104 cases (13%) and after 25 years duration of OC use, smoking
electoral roll) 194 controls (23%) and family history of
1990–1993 with tubal sterilization ovarian cancer.
Miracle- Cohort (ovarian cancer 396 114 women HR
c
= 0.64 (0.42, 0.96) Adjusted for: age at
McMahill et al., 1997 (USA) mortality) 799 ovarian cancer HRadj = 0.68 (0.45, 1.03) interview, race, body mass
1982–1991 deaths Risk was greater the first index, education, family,
24 women (3%) had 20 years than later history of ovarian/breast
a tubal sterilization cancer, no. of pregnancies,
marital status, menarche,
menopause, OC use,
HRT use, no. of
miscarriages, smoking
Krieger et al., 1997 (Canada) Cohort 251 907 women SIR
d
= 0.57, P 0.001 No adjustment for
(ovarian cancer having had tubal Risk was decreased even confounding factors
incidence) ligation after 10 years
1973–1993 108 ovarian cancers
a
Odds ratio.
b
Oral contraceptive.
c
Hazard ratio.
d
Standardized incidence ratio.
by guest on June 6, 2013http://ije.oxfordjournals.org/Downloaded from
600 INTERNATIONAL JOURNAL OF EPIDEMIOLOGY
up to 25 years after sterilization.
5
They are also in agreement with
the results of the only other follow-up study assessing ovarian
cancer incidence by years since tubal ligation.
7
Finally, a pro-
tective effect of the surgical procedure of sterilization itself (rather
than a screening effect) may be supported indirectly by the finding
of Krieger et al.
7
of no protective effect of unilateral ovariectomy,
where a similar screening effect could be anticipated.
Several hypotheses have been suggested to establish the
biological plausibility and explain a protective effect of tubal
sterilization on ovarian cancer risk. The majority of ovarian
cancers develop from the cells of the surface epithelium. During
each ovulation, these cells are involved in follicular rupture and
increased cell division in relation to the subsequent wound
repair. Fatallas ‘incessant ovulation’ hypothesis suggested that
these repeated minor traumas to the ovarian surface epithelium
increase the risk of ovarian cancer.
21
In line with this, the
number of ovulations during lifetime, as well as factors
associated with increased/decreased number of ovulations
(nulliparity, oral contraceptive [OC] use), have been identified
as important risk determinants for cancer of the ovaries.
22,23
Likewise, high levels of circulating gonadotrophins from the
pituitary gland have been suggested as playing an important
role in ovarian carcinogenesis (‘gonadotrophin hypo-
thesis’).
24,25
It has been suggested that following sterilization,
the ovarian circulation may be impaired
26
causing suppressed
ovarian hormone production followed by some degree of
anovulation, and thereby maybe a reduction in the risk of
ovarian cancer. Furthermore, the levels of circulating hormones
may be changed, also affecting the ovarian cancer risk. In
addition, it has been hypothesized that because of a reduction
in the utero-ovarian circulation, caused by the tubal
sterilization, reduced concentrations of uterine growth factors
reach the ovaries resulting in decreased ovarian cancer risk.
27
Similarly, the changes in the endogenous hormone level after
tubal sterilization may imply a different oestrogen/progesterone
ratio and this has been hypothesized to be related to the risk of
endometrial cancer.
28,29
However, results from the studies
forming the basis for these hypotheses are somewhat conflicting.
In some studies, a decreased oestrogen and maybe progesterone
level has been observed after tubal sterilization,
16,30,31
whereas
in others no changes were found.
32,33
Finally, another theory
has suggested that tubal ligation may prevent talc or other
carcinogens from ascending the tubes and coming in contact
with the ovaries. However, the current conclusion is that the
mechanism underlying the apparent protective effect of tubal
ligation is still not adequately elucidated.
Rosenblatt et al.
3
reported that the effect of tubal ligation was
seen only for clear cell and endometrioid ovarian carcinomas.
However, the results were based only on one exposed case in
each group. This finding could not be confirmed in the present
study. In our study, the overall decrease in ovarian cancer risk
seemed to affect the non-mucinous histological types only. It
has previously been suggested that mucinous ovarian cancers
have a different pattern of risk factors from the non-mucinous
types.
34
This relates in particular to use of OC and to parity.
However, the currently available data on this issue are
equivocal. Some studies have found use of OC and increasing
parity to decrease the risk of both mucinous and non-mucinous
ovarian cancer.
35–37
In contrast, others observed such a
decreased risk only for the non-mucinous tumours, whereas for
ovarian cancer of the mucinous type they found no or even an
increased risk associated with OC use and parity.
34,38–40
If the
development of mucinous ovarian tumours is less dependent on
hormonal/ovulatory factors as indicated by some studies,
41
it
might be anticipated that a protective effect of sterilization
would also be less pronounced for these tumours. Indeed, the
results of the present paper may support a different aetiological
mechanism for ovarian cancers of the mucinous type, however,
due to relatively small numbers when ovarian cancer is divided
into histological groups, we cannot exclude the possibility that
the findings may be due to chance.
Only a few case-control studies and no follow-up studies
have focused on the association between tubal ligation and the
risk of endometrial cancer.
10–12,42
In one study, a significantly
decreased crude risk was observed; however, after adjustment
for confounding factors like age and parity, only a moderately
(non-significant) decreased risk remained.
10
In the same study,
the reduction in risk tended to diminish with time since the
sterilization, maybe pointing to some degree of surveillance
effect. In another study, based on results from the World Health
Organization Collaborative Study of Neoplasia and Steroid
Contraceptives, it was concluded that the existing findings with
regard to the association between endometrial cancer and
previous sterilization could most likely be ascribed to random
variation.
11
This is supported by the most recent study in which
no significant association was observed between sterilization
and endometrial cancer after adjustment for parity was done.
12
In our study, we find an overall decreased risk of endometrial
cancer, and we observe no significant tendency that the effect
diminishes with length of follow-up. However, as we did not
have information on other risk factors, we were unable to make
any adjustments for potential confounding factors. Thus, even
though it is in theory biologically plausible that sterilization
affects endometrial cancer risk, it is still unclear whether a
causal association exists.
We find that the occurrence of uterine cervical cancer was
slightly lower and the occurrence of CIN3 higher than could be
expected from the rates in the general population. Our results
for invasive cervical cancer are almost identical to those recently
reported by Li et al.
13
The association between tubal sterilization
and CIN3 has not previously been addressed. The finding of a
decrease (although limited) in cancer of the cervix and simul-
taneously increased detection of CIN3, which is only rarely
symptomatic, most likely point to a screening effect i.e. in
connection with the surgical procedure, a high proportion of
women will also have a Pap smear.
A well-defined study population and a large cohort size
strengthen this nationwide, population-based study. Further-
more, information on outcome is nearly complete and there is
virtually no loss to follow-up, due to the accuracy and
completeness of the Danish Cancer Registry and the Central
Population Register. Finally, except for the study by Krieger
et al.,
7
this is the only study where information on exposure is
not dependent on patient recall. It is a limitation of the present
study, however, that we had no information on the type of
sterilization, and thus were unable to assess whether the cancer
risk varied with the different methods of tubal sterilization. In
addition, no information on potential confounding factors was
available. It is likely that women having a tubal sterilization
have more children than other women or at least a lower
by guest on June 6, 2013http://ije.oxfordjournals.org/Downloaded from
TUBAL STERILIZATION AND GYNAECOLOGICAL CANCERS 601
prevalence of infertility, and consequently an a priori lower risk
of ovarian/endometrial cancer. It is also clear that having a tubal
ligation, and the age at which this occurs, could influence the
total duration of OC use as well as the number of children. Both
of these variables constitute important risk determinants for
ovarian cancer, and thus our inability to adjust for these factors
may well have influenced the results of our study. Nevertheless,
it seems less likely that this can account entirely for our findings
with regard to ovarian cancer. In none of the recent studies
mentioned in Table 4, where information on such factors was
collected, neither parity alone nor in combination with other
risk determinants could explain the association between tubal
sterilization and ovarian cancer (Table 4). Furthermore, if the
association could be explained by a reduction in the length of
time when a woman could use OC or could have children, one
might expect a different magnitude of association in women
having their tubal ligation at a young age compared with
women sterilized later in life, and we did not observe such a
difference (data not shown). Finally, inclusion of women who
are actually not at risk (e.g. women with a hysterectomy or
bilateral oophorectomy) in the denominator of the standard
rate may imply underestimation of the expected number and
thus overestimation of the relative rate. As only about 1800
women undergo bilateral oophorectomy and 5500 women have
a hysterectomy each year (out of a population of approximately
1.8 million women aged 15–69 years),
43
the effect on the
ovarian cancer SIR estimate is considered limited.
In conclusion, in this nationwide, population-based study we
find that women with tubal sterilization have a decreased risk
of subsequent development of ovarian cancer, which seems to
last for at least 10–15 years, and likewise, a slight reduction in
the occurrence of ovarian borderline tumours is observed. Our
data do not support that screening bias can explain the protec-
tive effect against ovarian cancer, but indicate that the steriliza-
tion itself may convey a reduction in risk, although the biological
mechanism is not entirely clear. The same overall pattern is
found for endometrial cancer; however, the association is
weaker than for cancer of the ovaries. In contrast, the modera-
tely decreased cervical cancer risk and increased risk of CIN3
observed in our cohort indicate that the effect of tubal steriliza-
tion in this case can be explained by an increased cervical cancer
screening among sterilized women.
Acknowledgement
The study was supported by Master Agreement Order Contract
No. MAO NO1-CP-61111 from the National Cancer Institute,
Bethesda, Maryland, USA.
References
1
Whittemore AS, Harris R, Itnyre J et al. Characteristics relating to
ovarian cancer risk: collaborative analysis of 12 US case-control
studies. II. Invasive epithelial ovarian cancer in white women. Am J
Epidemiol 1992;136:1184–203.
2
Hankinson SE, Hunter DJ, Colditz GA et al. Tubal ligation,
hysterectomy and risk of ovarian cancer. JAMA 1993;270:2813–18.
3
Rosenblatt KA, Thomas DB and The World Health Organization
Collaborative Study of Neoplasia and Steroid Contraception (1996).
Reduced risk of ovarian cancer in women with a tubal ligation or
hysterectomy. Cancer Epidemiol Biomark Prev 1996;5:933–35.
4
Cornelison TL, Natarajan N, Piver MS, Mettlin CJ. Tubal ligation and
the risk of ovarian carcinoma. Cancer Detect Prev 1997;21:1–6.
5
Green A, Purdie D, Bain C et al. Tubal sterilization, hysterectomy and
decreased risk of ovarian cancer. Int J Cancer 1997;71:948–51.
6
Miracle-McMahill HL, Calle EE, Kosinski AS et al. Tubal ligation and
fatal ovarian cancer in a large prospective study. Am J Epidemiol
1997;145:349–57.
7
Krieger N, Sloan M, Cotterchio M, Parsons P. Surgical procedures
associated with risk of ovarian cancer. Int J Epidemiol 1997;26:
710–15.
8
McGowan L, Parent L, Lednar W et al. The woman at risk for
developing ovarian cancer. Gynecol Oncol 1979;7:325–44.
9
Koch M, Starreveld AA, Hill GB et al. The effect of tubal ligation on
the incidence of epithelial cancer of the ovary. Cancer Detect Prev
1984;7:241–45.
10
Castelsagué X, Thompson WD, Dubrow R. Tubal sterilization and the
risk of endometrial cancer. Int J Cancer 1996;65:607–12.
11
Rosenblatt K, Thomas D. Association between tubal ligation and
endometrial cancer. Int J Cancer 1997;71:129–30.
KEY MESSAGES
Women with a tubal sterilization have a decreased risk of developing ovarian cancer.
The effect on borderline tumours was less pronounced.
The protective effect does not seem to diminish significantly with time since the sterilization.
The decreased risk did not apply to the group of mucinous ovarian cancers, and thus our data may support the
previously suggested hypothesis of a different aetiological mechanism for this histological type.
A decreased risk of endometrial cancer following tubal sterilization was observed, although the association
seemed less strong than for ovarian cancer.
A slightly decreased risk of cervical cancer together with an increased risk of cervical intraepithelial neoplasia
grade 3 (CIN3) point to increased cervical cancer screening activity among sterilized women.
by guest on June 6, 2013http://ije.oxfordjournals.org/Downloaded from
602 INTERNATIONAL JOURNAL OF EPIDEMIOLOGY
12
Lacey JV, Brinton LA, Mortel R et al. Tubal sterilization and the risk of
cancer of the endometrium. Gynecol Oncol 2001;79:482–84.
13
Li H, Thomas DB. Tubal ligation and risk of cervical cancer.
Contraception 2000;61:323–28.
14
Sørensen T, Ladehoff P, Lindholm P, Quist K. Follicular stimulating
hormone, luteinizing hormone and estrogen levels before and after
female sterilization. Acta Obstet Gynecol Scand 1981;50:559–61.
15
Cattanach JF, Milne BJ. Post-tubal ovulatory problems correlated
with ovarian steroidogenesis. Contraception 1988;38:541–50.
16
Hakverdi AU, Taner CE, Esden AC, Satici O. Changes in ovarian
function after sterilization. Adv Contracept 1994;10:51–56.
17
Danish National Board of Health. Activity in the Hospital Care System
1979. Copenhagen: Danish National Board of Health, 1981.
18
Danish National Board of Health. Classification of Surgical Procedures and
Therapies. 1st, 2nd and 3rd Edns. Copenhagen: Danish National Board of
Health 1973, 1980, 1988.
19
Storm HH, Mandres T, Friis S, Bang S. Cancer Incidence in Denmark in
1988. Copenhagen: Danish Cancer Society, 1991.
20
Bailar JC, Ederer F. Significance factors for the ratio of a Poisson
variable to its expectation. Biometrics 1964;20:639–43.
21
Fatalla MF. Incessant ovulation—a factor in ovarian neoplasia? Lancet
1971;ii:163.
22
Purdie D, Green A, Bain C et al. Reproductive and other factors and
risk of epithelian ovarian cancer: an Australian case-control study. Int
J Cancer 1995;62:678–84.
23
Risch HA, Marrett LD, Howe GR. Parity, contraception, infertility and
the risk of epithelial ovarian cancer. Am J Epidemiol 1994;140:585–97.
24
Stadel BV. The etiology and prevention of ovarian cancer. Am J Obstet
Gynecol 1975;123:772–74.
25
Cramer DW, Welch WR. Determinants of ovarian cancer risk.
II. Inferences regarding pathogenesis. J Natl Cancer Inst 1983;71:717–21.
26
Rilman T, Persson I, Nilsson S. Hormonal aspects of epithelial ovarian
cancer: review of epidemilogic evidence. Clin Endocrinol 1998;49:
665–70.
27
Cramer DW, Xu H. Epidemiologic evidence for growth factors in the
pathogenesis of ovarian cancer. Ann Epidemiol 1995;5:310–14.
28
Henderson BE, Ross RK, Pike MC, Casagrande JT. Endogenous
hormones as a major factor in human cancer. Cancer Res 1982;42:
3232–39.
29
Henderson BE, Feigelson HS. Hormonal carcinogenesis. Carcinogenesis
2000;21:427–33.
30
Radwanska E, Headley SK, Dmowski P. Evaluation of ovarian
function after tubal sterilization. J Reprod Med 1982;27:376–84.
31
Cattanach JF. Oestrogen deficiency after tubal ligation. Lancet
1985;i:847–49.
32
Garza-Flores J, Vázquez-Estrada L, Reyes A et al. Assessment of
luteal function after surgical tubal sterilization. Adv Contracept
1991;7:371–77.
33
Erruo W, Bilian X, Weigian Y, Hui L, Beisheng I. Hormonal profile of
the menstrual cycle in Chinese women after tubal sterilization.
Contraception 1992;45:583–93.
34
Risch AR, Marret LD, Howe JM. Differences in risk factors for
epithelial ovarian cancer by histology. Am J Epidemiol
1996;144:363–72.
35
Wittenberg L, Cook LS, Rossing MA, Weiss NS. Reproductive risk
factors for mucinous and non-mucinous epithelial ovarian cancer.
Epidemiology 1999;10:761–63.
36
Weis NS, Lyon JL, Liff JM, Vollmer WM, Daling JR. Incidence of
ovarian cancer in relation to the use of oral contraceptives. Int J Cancer
1981;28:669–71.
37
Cramer DW, Hutchison GB, Welch WR, Scully RE, Ryan RE.
Determinants of ovarian cancer risk. I. Reproductive experience and
family history. J Natl Cancer Inst 1982;71:711–16.
38
Kvåle G, Heuch I, Nilsen S, Beral V. Reproductive factors and risk of
ovarian cancer: a prospective study. Int J Cancer 1988;42:246–51.
39
World Health Organization Collaborative Study of Neoplasia and
Steriod Contraceptives. Epithelial ovarian cancer and combined oral
contraceptives. Int J Epidemiol 1989;18:538–45.
40
Cramer DW, Hutchison GB, Welch WR, Beral V. Factors affecting the
association of oral contraceptives and ovarian cancer. N Engl J Med
1982;307:1047–51.
41
Purdie DM, Siskind V, Bain CJ, Webb PM, Green A. Reproduction-
related risk factors for mucinous and nonmucinous epithelial ovarian
cancer. Am J Epidemiol 2001;153:860–64.
42
Kelsey JL, Livolsi VA, Holford TR et al. A case-control study of the
endometrium. Am J Epidemiol 1982;116:333–42.
43
Danish National Board of Health. Surgical Activity in Danish Hospitals
1992. Copenhagen: Danish National Board of Health, 1992.
by guest on June 6, 2013http://ije.oxfordjournals.org/Downloaded from
    • "In a meta-regression of six of these studies, we did not observe a difference in the relative risk of ovarian cancer between women who had a tubal ligation less than 10 years ago (summary RR = 0.69, 95%CI: 0.59, 0.79) and those women who had a tubal ligation 10 or more years ago (summary RR = 0.68, 95%CI: 0.54, 0.87) (P-heterogeneity = 0.78) (Table 2). Of the other studies, a prospective cohort study of ovarian cancer mortality reported tubal ligation to be associated with a reduced risk for women who had the procedure within 20 years, with a smaller non-significant reduced risk for those who had the procedure 20 or more years ago.[19] However , a prospective cohort study based in China observed a non-significant increase in risk that was similar for both women who had a tubal ligation less than 33 years ago and women who had a tubal ligation 33 or more years ago [14].Figure 2 Forest plot for 30 studies of the association between tubal ligation and ovarian cancer risk. "
    [Show abstract] [Hide abstract] ABSTRACT: The purpose of this meta-analysis was to determine the strength of the association between gynecologic surgeries, tubal ligation and hysterectomy, and ovarian cancer. We searched the PubMed, Web of Science, and Embase databases for all English-language articles dated between 1969 through March 2011 using the keywords "ovarian cancer" and "tubal ligation" or "tubal sterilization" or "hysterectomy." We identified 30 studies on tubal ligation and 24 studies on hysterectomy that provided relative risks for ovarian cancer and a p-value or 95% confidence interval (CI) to include in the meta-analysis. Summary RRs and 95% CIs were calculated using a random-effects model. The summary RR for women with vs. without tubal ligation was 0.70 (95%CI: 0.64, 0.75). Similarly, the summary RR for women with vs. without hysterectomy was 0.74 (95%CI: 0.65, 0.84). Simple hysterectomy and hysterectomy with unilateral oophorectomy were associated with a similar decrease in risk (summery RR = 0.62, 95%CI: 0.49-0.79 and 0.60, 95%CI: 0.47-0.78, respectively). In secondary analyses, the association between tubal ligation and ovarian cancer risk was stronger for endometrioid tumors (summary RR = 0.45, 95%CI: 0.33, 0.61) compared to serous tumors. Observational epidemiologic evidence strongly supports that tubal ligation and hysterectomy are associated with a decrease in the risk of ovarian cancer, by approximately 26-30%. Additional research is needed to determine whether the association between tubal ligation and hysterectomy on ovarian cancer risk differs by individual, surgical, and tumor characteristics.
    Full-text · Article · May 2012
    • "Thus, simultaneous risk-reducing salpingectomies may become more widespread whenever a hysterectomy needs to be performed for benign indications and may be worth evaluating as isolated preventive procedures for women at a certain age group. It is well established that tubal ligation has a protective effect against the development of ovarian cancer434445 . Although the mechanisms remain unclear, with a tubal model of serous carcinogenesis, this protective risk may be explained in 3 ways: 1) For mid tubal ligation procedures, the mechanical barrier created by the ligation may prevent potentially mutagenic factors (inflammation, retro-grade menstruation material) from reaching the distal fallopian tube 2) For distal tubal ligation procedures, removal of the spatial relationship that as previously noted, facilitates the movement of tubal epithelium to the ovary, and 3) Removal of significant segments of tubal epithelium removes potential foci of malignant transformation . "
    Full-text · Article · Jan 2012
    • "Nevertheless, it is possible that the current generation of women are also subject to coercion: it is technically very easy to have a TO if the abdomen has to be opened anyway. In addition, the CS/TO procedure is not associated with any extra effort, pain, anxiety, risks or costs on the woman's part, and there are also, just like with some other methods, non-contraceptive benefits [37], [38].These factors can be enticing and should lead to extra careful counselling. However, not allowing women to choose whether they want a TO with their CS is also a form of coercion. "
    [Show abstract] [Hide abstract] ABSTRACT: In The Netherlands, caesarean sections (CSs) are rarely combined with tubal occlusion (TO), partly because discussing CS/TO near delivery is considered unethical and earlier hypothetical counselling--i.e. suppose you happen to need a CS--is rare. This results in more unintended pregnancies and is inconsistent with informed choice. We explored whether TO should indeed not be made routinely available to eligible women. A questionnaire was mailed to 515 Para ≥2 who underwent in the past ≥1 CS. 498 (96.7%) responded. They were on average 35.3 years old, had 2.5 children, had undergone 1.6 CSs, and 3.3 years had passed since their index delivery, either a CS (393) or vaginal birth (105) after a previous CS. 87% of the 498 believed that pregnant mothers with ≥1 children should be routinely counselled about CS/TO. Indeed, 58% and 85% respectively, thought women/couples expecting their second or third child should still be given the TO option days before delivery, if omitted earlier. Counselled women, 138/498 (27.8%), were far more often satisfied than those without CS/TO option. 33/393 had a CS/TO. None indicated regret in the questionnaire. Another 119 also would have elected a CS/TO if given that option. Therefore, 152 (38.7%) of 393 Para ≥2 had or would have liked a concurrent TO. 118/119 wrote they still regretted missing this opportunity. The exception's husband had had a vasectomy. 100/119 were good TO candidates: they were ≥28 years when they delivered an apparently healthy baby of ≥37 weeks. The current contraceptive use of these 100 suggests that this group will have at least 8 unintended pregnancies before age 50. The experiences and opinions of previous potential candidates for a CS/TO do not support the reluctance of Dutch obstetricians to counsel pregnant Para ≥1 about the TO option for a (potential) CS.
    Full-text · Article · Mar 2011
Show more