Cigarette smoking and risk of histological subtypes of epithelial ovarian cancer in the EPIC cohort study.

Abstract

New data regarding a positive association between smoking and risk of epithelial ovarian cancer (EOC), especially the mucinous tumor type, has started to emerge. The purpose of this study was to examine the association between different measures of smoking exposures and subtypes of EOC in a large cohort of women from 10 European countries. The European Prospective Investigation into Cancer and Nutrition (EPIC) cohort is a multicenter prospective study initiated in 1992. The questionnaires included data about dietary, lifestyle, and health factors. Information about cigarette smoking was collected from individuals in all participating countries. We used Cox proportional hazard regression models to estimate hazard ratio (HR) of EOC overall and serous, mucinous, and endometroid histological subtypes, with 95% confidence intervals (CIs) associated with different measures of smoking exposures adjusting for confounding variables. Altogether 836 incident EOC cases were identified among 326,831 women. The tumors were classified as 400 serous, 83 mucinous, 80 endometroid, 35 clear cell, and 238 unspecified. Compared with never smokers, current smokers had a significantly increased risk for mucinous tumors [HR = 1.85 (95% CI 1.08-3.16)] and those smoking more than 10 cigarettes per day had a doubling in risk [HR = 2.25(95% CI 1.26-4.03)] as did those who had smoked less than 15 pack-years of cigarettes [HR = 2.18 (95% CI 1.07-4.43)]. The results from the EPIC study add further evidence that smoking increases risk of mucinous ovarian cancer and support the notion that the effect of smoking varies according to histological subtype.

Full text

Page 1

Cigarette smoking and risk of histological subtypes
of epithelial ovarian cancer in the EPIC cohort study
Inger T. Gram1,2, Annekatrin Lukanova3, Ilene Brill4, Tonje Braaten1, Eiliv Lund1, Eva Lundin5, Kim Overvad6,
Anne Tjønneland7, Francoise Clavel-Chapelon8,9, Nathalie Chabbert-Buffet10, Christina Bamia11, Antonia Trichopoulou11,12,
Dimosthenis Zylis11,12, Giovanna Masala13, Franco Berrino14, Rocco Galasso15, Rosario Tumino16, Carlotta Sacerdote17,18,
Oxana Gavrilyuk1, Steinar Kristiansen1, Laudina Rodrı ´guez19, Catalina Bonet20, Jose ´ Marı ´a Huerta21,22,
Aurelio Barricarte23,24, Maria-Jose ´ Sa ´nchez22,25, Miren Dorronsoro26, Karin Jirstro ¨m27, Martin Almquist28, Annika Idahl29,30,
H. Bas Bueno-de-Mesquita31,32, Marie Braem33, Charlotte Onland-Moret33, Konstantinos K. Tsilidis34,
Naomi E. Allen34, Veronika Fedirko35, E. Riboli36and Rudolf Kaaks3
1Institute of Community Medicine, Faculty of Health Sciences, University of Tromsø, Tromsø, Norway
2Norwegian Centre for Integrated Care and Telemedicine, University Hospital of North Norway, Tromsø, Norway
3Division of Cancer Epidemiology, German Cancer Research Center, Heidelberg, Germany
4Department of Epidemiology, School of Public Health, University of Alabama at Birmingham, Birmingham, AL
5Department of Medical Biosciences, Pathology, Umea ˚ University, Umea ˚, Sweden
6Department of Epidemiology, School of Public Health, Aarhus University, DK-8000 Aarhus C, Denmark
7Institute of Cancer Epidemiology, Danish Cancer Society, Copenhagen, Denmark
8Inserm, Centre for Research in Epidemiology and Population Health, U1018, Institut Gustave Roussy, F-94805, Villejuif, France
9Paris South University, UMRS 1018, F-94805, Villejuif, France
10Ob-Gyn Department, APHP Hospital Tenon and UMR S_938 University P and M Curie, Paris, France
11WHO Collaborating Center for Food and Nutrition Policies, Department of Hygiene, Epidemiology and Medical Statistics, University of Athens Medical
School, Athens, Greece
12Hellenic Health Foundation, Athens, Greece
13Molecular and Nutritional Epidemiology Unit, Cancer Research and Prevention Institute (ISPO), Ponte Nuovo, Via delle Oblate n.2, 50139 Florence, Italy
14Fondazione IRCCS Istituto Nazionale Tumori, Milan, Italy
15Istituto di Ricovero e Cura a Carattere Scientifico - Centro di Riferimento Oncologico di Basilicata, Unit of Clinical Epidemiolgy, Biostatistics and Cancer
Registry, Rionero in Vulture, Italy
16Cancer Registry and Histopathology Unit, ‘‘Civile – M.P.Arezzo’’, Hospital ASP 7, Ragusa, Italy
17Center for Cancer Prevention (CPO-Piemonte), Torino, Italy
18Human Genetic Foundation, (HuGeF), Torino, Italy
19Public Health and Participation Directorate, Health and Health Care Services Council, Asturias, Spain
20Unit of Nutrition, Environment and Cancer, Catalan Institute of Oncology, Barcelona, Spain
21Department of Epidemiology, Murcia Regional Health Authority, Murcia, Spain
22CIBER Epidemiologı ´a y Salud Pu ´blica (CIBERESP), Spain
23Navarra Public Health Institute, Pamplona, Spain
24Consortium for Biomedical Research in Epidemiology and Public Health, Spain
25Andalusian School of Public Health, Granada, Spain
26Epidemiology and Health Information, Public Health Division of Gipuzkoa, Basque Regional Health Department, Avda. San Sebastian, Spain
27Department of Clinical Sciences, Pathology, Lund University, Ska ˚ne University Hospital, Lund, Sweden
28Department of Surgery, University Hospital Lund, Lund, Sweden
29Department of Public Health and Clinical Medicine, Nutritional Research, Umea ˚ University, Umea ˚, Sweden
30Department of Clinical Sciences, Obstetris and Gynecology, Umea ˚ University, Umea ˚, Sweden
31National Institute for Public Health and the Environment (RIVM), Bilthoven, The Netherlands
32Department of Gastroenterology and Hepatology, University Medical Centre Utrecht (UMCU), Utrecht, The Netherlands
33Julius Center for Health Sciences and Primary Care, University Medical Centre Utrecht (UMCU), Utrecht, The Netherlands
34Cancer Epidemiology Unit, Nuffield Department of Clinical Medicine, University of Oxford, Oxford, United Kingdom
35International Agency for Research on Cancer, Lyon, France
36Department of Cancer Epidemiology and Prevention, School of Public Health, Imperial College, London, United Kingdom
Key words: ovarian cancer, smoking, mucinous, serous, EPIC cohort
Abbreviations: BMI: body mass index; CI: confidence interval; EOC: epithelial ovarian cancer; EPIC: European Prospective Investigation into
Cancer and Nutrition; HR: hazard ratio; HT: hormone therapy; IARC: International Agency for Research on Cancer; OC: oral contraceptive
DOI: 10.1002/ijc.26235
History: Received 13 Apr 2011; Accepted 26 May 2011; Online 15 Jun 2011
Correspondence to: Inger Torhild Gram, Institute of Community Medicine, University of Tromsø, N-9037 Tromsø, Norway,
Tel.: þ[4777644816], Faxþ[4777644831], E-mail: inger.gram@uit.no
Short Report
Int. J. Cancer: 130, 2204–2210 (2012) V
C 2011 UICC
International Journal of Cancer
IJC

Page 2

New data regarding a positive association between smoking and risk of epithelial ovarian cancer (EOC), especially the
mucinous tumor type, has started to emerge. The purpose of this study was to examine the association between different
measures of smoking exposures and subtypes of EOC in a large cohort of women from 10 European countries. The European
Prospective Investigation into Cancer and Nutrition (EPIC) cohort is a multicenter prospective study initiated in 1992. The
questionnaires included data about dietary, lifestyle, and health factors. Information about cigarette smoking was collected
from individuals in all participating countries. We used Cox proportional hazard regression models to estimate hazard ratio
(HR) of EOC overall and serous, mucinous, and endometroid histological subtypes, with 95% confidence intervals (CIs)
associated with different measures of smoking exposures adjusting for confounding variables. Altogether 836 incident EOC
cases were identified among 326,831 women. The tumors were classified as 400 serous, 83 mucinous, 80 endometroid, 35
clear cell, and 238 unspecified. Compared with never smokers, current smokers had a significantly increased risk for
mucinous tumors [HR 5 1.85 (95% CI 1.08–3.16)] and those smoking more than 10 cigarettes per day had a doubling in risk
[HR 5 2.25(95% CI 1.26–4.03)] as did those who had smoked less than 15 pack-years of cigarettes [HR 5 2.18 (95% CI
1.07–4.43)]. The results from the EPIC study add further evidence that smoking increases risk of mucinous ovarian cancer and
support the notion that the effect of smoking varies according to histological subtype.
Ovarian cancer is the fifth most common cancer among
women in Europe with an estimated number of 67,000
(4.5%) new cases, and the sixth most common cause of can-
cer death with an estimated number of 42,000 (5.5 %) deaths
in 2008.1Except for the well-established protection by parity
and oral contraceptive use, relatively little is known about
other factors that may influence the risk of developing the
disease.2,3In 2004, the Monograph on ‘‘Tobacco smoke and
involuntary smoking’’ from the International Agency for
Research on Cancer (IARC) reported that there was insuffi-
cient evidence to draw conclusions regarding the possible
effect of smoking on ovarian cancer risk.4However, new data
regarding a positive association between smoking and risk of
epithelial ovarian cancer (EOC), especially with the mucinous
tumor type, have started to emerge. A systematic review
from 2006,5including a meta-analysis on the association of
smoking and ovarian cancer, with data from one cohort
study6and 18 case–control studies, showed a significant two-
fold increased risk of mucinous tumors among current com-
pared with never smokers. Since then, three7–9cohort and
two case control10,11studies have examined the relationship
between smoking and histological subtypes of EOC, all sup-
porting the notion that the histologic subtypes of EOC may
represent different entities. In the most recent summary
statement from IARC in 2009, both colorectal cancer and
mucinous EOC are reported as new tumor sites for which
there is sufficient evidence to list tobacco smoke as a cause of
carcinogenicity in humans.12
The purpose of this study was to examine the association
between different measures of smoking exposures and sub-
types of EOC in a large cohort of women from 10 European
countries.
Material and Methods
The European Prospective Investigation into Cancer and
Nutrition (EPIC) cohort population and the procedures of
data collection have been described elsewhere.13In brief,
approximately 370,000 women and 150,000 men were
recruited between 1992 and 2000 from 23 centers in 10 Euro-
pean countries (France, Italy, Spain, United Kingdom (UK),
The Netherlands, Greece, Germany, Sweden, Denmark, and
Norway). Approval for the study was obtained from the local
ethics committees in the participating countries and the in-
ternal review board of the International Agency for Research
on Cancer.
The following women were excluded: 19,707 with preva-
lent cancer other than nonmelanoma skin cancer, 2,209 with
incomplete follow-up data, 10,500 with bilateral oophorec-
tomy, 28 with a nonepithelial ovarian cancer, 509 lacking
baseline questionnaire, and 8,209 with missing information
on smoking status, leaving 326,831 women in the analytical
cohort.
Incident cancer cases were identified through linkage to
population cancer registries in Denmark, Italy, The Nether-
lands, Spain, Sweden, UK, and Norway, or using a combina-
tion of methods including linkage to health insurance
records, cancer and pathology registries, and active follow-up
of study participants or their next of kin in France, Germany,
and Greece. Invasive and borderline surface epithelial-stromal
ovarian tumors were referred to as EOC. They were further
categorized according to histology as serous, mucinous, endo-
metroid, clear cell, and others, including those with unspeci-
fied/missing information.
The baseline questionnaires elicited the following informa-
tion on cigarette smoking: smoking status at baseline (cur-
rent, former, or never), age started smoking, age quitting
(former smokers), number of cigarettes per day currently
smoked (current smokers), and number of cigarettes smoked
per day during different periods of life. Based on this infor-
mation, the following variables were calculated; duration of
smoking in years, defined as the current age (or age at quit-
ting for former smokers) minus age at start, intensity of
smoking, as the lifetime average number of cigarettes smoked
and pack-years of smoking as the lifetime average number of
cigarettes smoked multiplied by the number of years smoked
Short Report
Gram et al.
2205
Int. J. Cancer: 130, 2204–2210 (2012) V
C 2011 UICC

Page 3

divided by 20 (1 pack ¼ 20 cigarettes). For former smokers,
the number of years since quitting was calculated as the dif-
ference between current age and the age at which they quit
smoking. The questionnaires also asked about lifestyle and
reproductive factors. Information on body height and body
weight was collected allowing us to calculate body mass index
(BMI) as weight in kilograms divided by height in meters
squared.
Statistical analysis
We used T-Test and v2test for differences in distribution of
selected characteristics between cases/non-cases and between
current and never smokers and current and former smokers.
We used the Cox proportional hazards model with age as the
underlying time scale to estimate multivariate adjusted haz-
ard ratios (HR) with 95% confidence intervals associated with
different measures of smoking exposure for EOC overall, se-
rous, mucinous, and endometroid tumors, with never smok-
ers as the reference group. Entry time was defined as age at
recruitment and exit time was age at diagnosis of EOC, the
date of any incident cancer (except basal cell carcinoma) di-
agnosis, death, or the end of follow-up, whichever occurred
first. All models were stratified by recruitment center to con-
trol for center effects such as follow-up procedures and ques-
tionnaire design and age at recruitment (in 1-year interval)
to account for duration of follow-up.
Each of the following factors was evaluated as a potential
confounder of the relation between cigarette smoking and
EOC: four levels of education (none or primary school, tech-
nical or professional, secondary school, university degree),
postmenopausal (yes/no), age at menopause (premenopausal/
perimenopausal, postmenopausal: ?45, 46–50, 51–52, >52),
parous (yes/no), number of full-term pregnancies (0, 1, 2, 3,
4þ), ever oral contraceptive (OC) use(yes/no), duration of
OC use (never; ever: ?4, 5–9, ?10 years), ever postmeno-
pausal hormonal therapy (HT) use (yes/no), duration of HT
use (never, ever: ?4, 5þ), nondrinkers (yes/no), life-time
alcohol consumption in grams per day (0,<5, 5þ), simple
hysterectomy (yes, no), unilateral ovariectomy (yes/no), fertil-
ity problems (yes/no), and BMI (in quartiles). For each cova-
riate missing values were assigned to separate categories. We
included number of full-term pregnancies and duration of
OC use as established risk factors for EOC in the multivariate
model. After including these covariates, none of the factors
listed above changed the HR estimates materially and was
therefore not included in the final models. Tests for linear
trend were obtained by creating an ordinal exposure variable
with equally spaced scores and including it in the model. We
tested for heterogeneity between different histological tumor
types with Wald chi-square statistics.
Two-sided p-values <0.05 were considered statistically sig-
nificant. All analyses were conducted using the SAS version
9.2 (SAS Institute, Cary, North Carolina, USA).
Results
Altogether 836 incident EOC (87% invasive, 8% borderline,
and 5% unclassified) cases were identified during a median
follow-up time of 8.8 years. The tumors were classified as
400 (47.8%) serous, 83 (9.9%) mucinous, 80 (9.6%) endomet-
roid and 35 (4.2%) clear cell, while 238 (28.5%) cases were
undefined, not otherwise specified or had missing data on
histology.
EOC cases were older, had less years of education, were
more likely to be postmenopausal, and if so older at meno-
pause. Cases were younger at age of last birth, were less likely
to have a history of ever OC use, and duration of OC use
was shorter among ever users compared with noncases. Cases
were more likely to be ever HT users, duration of HT use
was longer among ever users, they reported to be heavier,
and to consume less alcohol compared with noncases (all p’s
<0.05; data not shown). Women with mucinous tumors were
younger at enrolment and at diagnosis, fewer were postmeno-
pausal and fewer were ever HT users compared with those
with serous tumors (all p’s <0.05; data not shown).
Forty-three percent of the women reported being ever
smokers. Current smokers were younger at enrolment and at
diagnosis of ovarian cancer and compared with never smokers
more were postmenopausal and menopause was reached at an
earlier age, on average. Compared with never smokers, current
smokers also had fewer children and a lower BMI and were
more likely to be ever users of OC and HT, and to be long-
term users for both medications. Fewer current smokers were
non-drinkers, and current smokers had higher average alcohol
consumption compared with never smokers. Compared with
former smokers, current smokers reported to have smoked on
average more cigarettes per day, on a lifetime basis, and for
more years (all p-values <0.001; data not shown).
Table 1 shows that current smokers had an 85% increased
risk of mucinous [HR ¼ 1.85 (95% CI 1.08–3.16)], while
both former [HR ¼ 0.83 (95% CI 0.48–1.45)] and current
[HR ¼ 0.58 (95% CI 0.29–1.15)] smokers had a nonsignifi-
cantly decreased risk of endometroid tumors compared with
never smokers. Compared with never smokers, current smok-
ers, and those smoking more than 10 cigarettes per day had
a doubling in risk [HR ¼ 2.25 (95% CI 1.26–4.03)] of muci-
nous tumors as did those who had smoked less than 15
pack-years of cigarettes [HR ¼ 2.18 (95% CI 1.07–4.43)]. No
consistent dose response relationship was found between the
different measures of smoking exposure and the histological
subtypes (Table 1).
The multivariate adjusted HR’s for current versus never
smokers did not differ significantly between mucinous and
serous (p for heterogeneity ¼ 0.06) while it did for mucinous
and endometroid tumors (p for heterogeneity ¼ 0.009).
Discussion
Our study finds that current smokers have an increased risk
of mucinous EOC compared with never smokers. We
Short Report
Int. J. Cancer: 130, 2204–2210 (2012) V
C 2011 UICC
2206
Smoking and subtypes of ovarian cancer

Page 4

Table 1. Multivariate1adjusted hazard ratio (HR) estimates of epithelial ovarian cancer with 95% confidence intervals (CI) overall and
histological subtype among smokers compared with never smokers: The EPIC Cohort Study
Smoking status
All tumors (N 5 836)Serous (N 5 400) Mucinous (N 5 83) Endometroid (N 5 80)
Cases/NoncasesHR CICases HRCI CasesHRCI CasesHR CI
Never 479/184,6721.00(ref) 2361.00 (ref)421.00 (ref)51 1.00(ref)
Former 193/75,5531.02 0.85–1.21850.91 0.71–1.18 171.00 0.56–1.79 180.83 0.48–1.45
Current164/65,770 1.120.93–1.35791.040.79–1.3624 1.851.08–3.16 110.58 0.29–1.15
Trend test2p0.270.940.040.11
Former
Age started smoking
20þ
<20
570.900.68–1.20260.91 0.60–1.3830.640.19–2.094 0.540.19–1.53
127 1.100.90–1.35570.980.72–1.32121.060.54–2.06141.020.55–1.90
Trend test2p0.480.810.990.86
Smoking duration (years)
0–19870.880.70–1.12380.780.55–1.1170.700.31–1.5980.680.32–1.48
20þ
891.200.95–1.52431.210.86–1.6981.430.65–3.1370.790.35–1.78
Trend test2p0.330.630.690.39
Life time number of cigarettes per day3
0–950 0.860.63–1.17170.590.35–0.9840.640.22–1.8350.740.28–1.93
10þ
641.080.81–1.42341.180.80–1.7360.990.40–2.4381.070.48–2.38
Trend test2p0.860.890.770.98
Pack-years of smoking
0–14760.850.66–1.11340.760.51–1.1160.600.25–1.4770.660.28–1.53
15þ
381.320.94–1.87171.290.77–2.1541.660.57–4.8661.630.66–4.02
Trend test2p0.590.890.970.67
Time since quitting (years)
0–9661.130.87–1.48321.110.76–1.6350.950.37–2.4550.650.26–1.66
10þ
1170.940.76–1.16510.850.62–1.16121.060.55–2.06100.720.36–1.45
Trend test2p0.621.001.000.25
Current
Age started smoking
20þ
<20
611.050.79–1.38301.060.71–1.5781.960.89–4.3550.680.26–1.74
1001.170.93–1.48481.040.74–1.45161.800.94–3.4360.580.23–1.41
Trend test2p0.200.79
<0.050.18
Smoking duration (years)
0–1920 1.240.77–2.0070.950.44–2.0952.200.77–6.3310.490.06–3.80
20þ
1411.100.90–1.35711.060.79–1.41191.780.98–3.24100.640.31–1.30
Trend test2p0.320.720.040.20
Life time number of cigarettes per day3
0–9521.190.88–1.62301.290.86–1.9481.600.71–3.6040.740.25–2.15
10þ
590.940.70–1.27331.010.68–1.50101.450.67–3.1250.650.24–1.75
Trend test2p0.920.760.270.35
Current number of cigarettes per day
0–9451.250.91–1.70221.210.77–1.9020.600.14–2.5320.430.10––1.79
10þ
1131.070.86–1.34551.000.73–1.37212.251.26–4.0390.700.33–1.49
Trend test2p0.410.90
<0.010.27
Short Report
Int. J. Cancer: 130, 2204–2210 (2012) V
C 2011 UICC
Gram et al.
2207

Page 5

observed a doubling in risk among those smoking more than
10 cigarettes daily as well as among those who had smoked
less than 15 pack-years of cigarettes. However, we did not
reveal any consistent dose response relationship between the
different measures of smoking exposure and mucinous tumors.
The present study has the largest numbers of cases among
all prospective cohort studies so far that focused on the rela-
tionship between smoking and EOC. The prospective design,
the large sample size, the classification according to histologi-
cal subtype, and the relatively large proportion of women
being ever smokers at enrolment are strengths of our study.
Data on daily smoking were standardized across study cen-
ters, and data on a wide range of potential confounders was
also collected in a similar way. The fact that known risk fac-
tors for ovarian cancer differed between cases and non-cases
in the expected way clearly indicates high internal validity of
our cohort study.
A remaining limitation of this study, in spite of the very
large cohort size overall, is the small number of ovarian can-
cer cases by histological subtype. Other limitations are that
borderline tumors were clearly under-reported and that for a
rather large proportion of cases histological subtype data was
missing. We did not have information on changes in smok-
ing habits during follow-up, but we find it likely that any
misclassification of smoking exposure and tumor histology
will have attenuated the reported associations. Although we
adjust for potential confounders, we cannot rule out the pres-
ence of residual confounding.
The four cohort studies summarized in Table 2 also had
fewer than 100 mucinous cases.6–9The Canadian Study6
reported a significant doubling in risk of mucinous tumors
among current smokers, while the Nurses Health Study I8
found an association of this magnitude among both former
and current compared with never smokers. In a subsequent
report,9which included the Nurses Health Study I & II, these
associations were no longer significant and were of similar
magnitude as those reported from Scandinavia by Gram et
al.7(Table 2). The results from the present study are in
agreement with those of the other cohort studies and contrib-
ute to strengthening the evidence.
The previously mentioned meta analysis, including a total
of 910 women with mucinous and 5,564 with non-mucinous
ovarian cancers, found that the risk of mucinous tumors
among current smokers increased with increasing amount
smoked but returned to that of never smokers within 20–30
years of quitting smoking, while current smokers had a risk
estimate close to unity for serous and a nonsignificantly
decreased risk of endometroid tumors.5The two more recent
case–control studies10,11also found that smokers had a signif-
icantly increased risk of mucinous and a nonsignificant
decrease in risk of endometroid tumors.
Emerging data on molecular features do support the hy-
pothesis that mucinous EOC is different from other subtypes.
Differences in genetic alterations that have been found as
mucinous have a significantly higher prevalence of KRAS
mutation and a lower frequency of BRCA and p53 abnormal-
ities compared with serous tumors. These subtypes have also
been found to differ in gene expression analyses and in im-
munohistochemical studies.14
Gene expression studies have shown that mucinous and
endometorid histotypes of EOC correlated with changes in
normal colonic mucosa and endometrium, respectively.15We
have previously found that smoking increased the risk of
colorectal cancer in both the Norwegian Women and Cancer
Study16and EPIC cohorts.17Furthermore, we have observed
a reduced risk of endometrial cancer with smoking among
postmenopausal and an increased risk among premenopausal
women, in another report from the EPIC study.18Our find-
ing suggesting that smoking may influence mucinous and
endometroid EOC in opposite direction is supported by the
significant test for heterogeneity, although this may still be
due to chance, given the relatively small numbers of ovarian
cancer cases by histological sub-type.
The available data obtained from past decades show that
DNA adduct formation is a key step in tobacco carcinogene-
sis.19Benzo(a) pyrene [B(a)p] is a potent carcinogen present
in cigarettes that acts locally.20B(a)p adducts have been
found in ovarian follicular cells among women exposed to
cigarette smoke. Presence of these adducts may increase the
risk for DNA damage through a direct carcinogenic effect.21
This supports the biological plausibility of a positive associa-
tion between smoking and any of the subtypes of EOC.
The results of the current study lend further support to
the heterogeneity of epithelial ovarian cancer and that
Table 1. Multivariate1adjusted hazard ratio (HR) estimates of epithelial ovarian cancer with 95% confidence intervals (CI) overall and
histological subtype among smokers compared with never smokers: The EPIC Cohort Study (Continued)
Smoking status
All tumors (N 5 836)Serous (N 5 400)Mucinous (N 5 83)Endometroid (N 5 80)
Cases/NoncasesHRCICasesHRCICasesHRCICasesHRCI
Pack–years of smoking
0–14521.210.89–1.66281.280.84–1.96132.181.07–4.4330.490.14–1.67
15þ
Trend test2p
590.930.70–1.25351.020.69–1.5250.850.32–2.2860.860.34–2.13
0.870.770.700.56
1Stratified by center and age, adjusted for number of full-term pregnancies and duration of oral contraceptives use, all at enrolment.2Includes
never smokers.3Women from Sweden and France excluded due to missing data on this variable.
Short Report
2208
Smoking and subtypes of ovarian cancer
Int. J. Cancer: 130, 2204–2210 (2012) V
C 2011 UICC

Page 6

smoking may influence the subtypes differently. It will take
many years before the respective cohorts have enough cases to
explore these associations in greater detail. To better under-
stand the risks revealed in the meta-analyses of case–controls
studies,5we suggest a pooling of the available data from the
cohort studies together with new updates of EOC cases.
Acknowledgements
The coordination of EPIC is financially supported by the European
Commission (DG-SANCO) and the International Agency for Research
on Cancer. The national cohorts are supported by Danish Cancer Soci-
ety (Denmark); Ligue contre le Cancer, Mutuelle Ge ´ne ´rale de l’Education
Nationale, Institut National de la Sante ´ et de la Recherche Medicale
(France); Deutsche Krebshilfe, Deutsches Krebsforschungszentrum and
Federal Ministry of Education and Research (Germany); Ministry of
Health and Social Solidarity, Stavros Niarchos Foundation and Hellenic
Health Foundation (Greece); Sicilian Government, AIRE-ONLUS Ragusa,
AVIS-Ragusa, Italian Association for Research on Cancer (AIRC) and
National Research Council (Italy); Dutch Ministry of Public Health,
Welfare and Sports (VWS), Netherlands Cancer Registry (NKR), LK
Research Funds, Dutch Prevention Funds, Dutch ZON (Zorg Onderzoek
Nederland), World Cancer Research Fund (WCRF); Statistics Nether-
lands (The Netherlands); Norwegian Cancer Society (Norway); Health
Research Fund (FIS), Regional Governments of Andalucı ´a, Asturias, Bas-
que Country, Murcia and Navarra, ISCIII RETIC (RD06/0020) (Spain);
Swedish Cancer Society, Swedish Scientific Council and Regional Gov-
ernment of Skåne and Va ¨sterbotten (Sweden); Cancer Research UK,
Medical Research Council UK (United Kingdom).
Table 2. Previous cohort studies of cigarette smoking and histological type of epithelial ovarian cancer (EOC)
First author,
study year,
(Ref.)
country
All tumors, N,
Smoking status
RR1
(95%CI)Serous
RR1
(95%CI)Mucinous
RR1
(95%CI)Endometroid RR1(95%CI)
Terry,N ¼ 4542,
Never
N ¼ 1842
N ¼ 322
N ¼ 67
2003, 1.001.00 1.001.00
(6)Ref Ref RefRef
CanadaFormer1.07 0.981.19 0.67
(0.86–1.34) (0.69–1.40)(0.48–2.93(0.37–1.22)
Current 1.181.04 2.290.81
(0.93–1.50)(0.71–1.53)(1.00–5.28)(0.43–1.54)
Gram,N ¼ 3433
Never
N ¼ 1943
N ¼ 553
NA
20081.01.001.00
(7)RefRefRef
NorwayFormer1.31.31.4
/Sweden(1.0–1.7)(0.9–1.8)(0.7–2.6)
Current1.41.31.5
(1.0–1.8)(0.9–1.9)(0.7–2.9)
Tworoger ,N ¼ 7373
Never
NA693
NA
20081.001.00
(8), NHS IRefRef
USAFormer1.052.02
(0.86–1.28)(1.15–3.55)
Current < 15/d1.08
(0.92–1.27)2.224
Current ? 15/d1.06(1.16–4.24)
(0.74–1.51)
Gates,N ¼ 8763
Never
N ¼ 4682
N ¼ 843
N ¼ 134
20101.001.001.001.00
(9 )RefRefRefRef
NHS I & IIFormer1.051.091.540.59
USA(0.91–1.22)(0.89–1.34)(0.85–2.74)(0.39–0.90)
Current1.11 1.141.42 0.93
(0.92–1.35)(0.88–1.49)(0.76–2.63)(0.59–1.47)
1Multivariate adjusted.2Invasive tumors.3Include invasive and borderline tumors.4Includes all current smokers.
Short Report
Gram et al.
2209
Int. J. Cancer: 130, 2204–2210 (2012) V
C 2011 UICC

Page 7

References
1. Ferlay J, Parkin DM, Steliarova-Foucher E.
Estimates of cancer incidence and
mortality in Europe in 2008. Eur J Cancer
2010;46:765–81.
Gertig DM, Hunter D. Ovarian cancer. In:
Adami HO, Hunter D, Trichopoulos D, eds.
Textbook in cancer epidemiology. New
York: Oxford University Press, 2002.378–94.
Lukanova A, Kaaks R. Endogenous
hormones and ovarian cancer:
epidemiology and current hypotheses.
Cancer Epidemiol Biomarkers Prev 2005;14:
98–107.
IARC. IARC Monographs on the
evaluation of carcinogenic risks to humans.
Tobacco smoke and involuntary smoking.
Lyon, France: IARC Press, 2004.
Jordan SJ, Whiteman DC, Purdie DM,
Green AC, Webb PM. Does smoking
increase risk of ovarian cancer? a
systematic review. Gynecol Oncol 2006;103:
1122–9.
Terry PD, Miller AB, Jones JG, Rohan TE.
Cigarette smoking and the risk of invasive
epithelial ovarian cancer in a prospective
cohort study. Eur J Cancer 2003;39:
1157–64.
Gram IT, Braaten T, Adami HO, Lund E,
Weiderpass E. Cigarette smoking and risk
of borderline and invasive epithelial
ovarian cancer. Int J Cancer 2008;122:
647–52.
Tworoger SS, Gertig DM, Gates MA,
Hecht JL, Hankinson SE. Caffeine, alcohol,
smoking, and the risk of incident epithelial
ovarian cancer. Cancer 2008;112:1169–77.
Gates MA, Rosner BA, Hecht JL, Tworoger
SS. Risk factors for epithelial ovarian
2.
3.
4.
5.
6.
7.
8.
9.
cancer by histologic subtype. Am J
Epidemiol 2010;171:45–53.
10. Soegaard M, Jensen A, Hogdall E,
Christensen L, Hogdall C, Blaakaer J, Kjaer
SK. Different risk factor profiles for
mucinous and nonmucinous ovarian
cancer: results from the Danish MALOVA
study. Cancer Epidemiol Biomarkers Prev
2007;16:1160–6.
11. Rossing MA, Cushing-Haugen KL,
Wicklund KG, Weiss NS. Cigarette
smoking and risk of epithelial ovarian
cancer. Cancer Causes Control 2008;19:
413–20.
12. Secretan B, Straif K, Baan R, Grosse Y, El
GF, Bouvard V, Benbrahim-Tallaa L, Guha
N, Freeman C, Galichet L, Cogliano V. A
review of human carcinogens. Part E:
tobacco, areca nut, alcohol, coal smoke,
and salted fish. Lancet Oncol 2009;10:
1033–4.
13. Riboli E, Hunt KJ, Slimani N, Ferrari P,
Norat T, Fahey M, Charrondiere UR,
Hemon B, Casagrande C, Vignat J,
Overvad K, Tjonneland A, et al. European
Prospective Investigation into Cancer and
Nutrition (EPIC): study populations and
data collection. Public Health Nutr 2002;5:
1113–24.
14. Frumovitz M, Schmeler KM, Malpica A,
Sood AK, Gershenson DM. Unmasking the
complexities of mucinous ovarian
carcinoma. Gynecol Oncol 2010;117:491–6.
15. Marquez RT, Baggerly KA, Patterson AP,
Liu J, Broaddus R, Frumovitz M, Atkinson
EN, Smith DI, Hartmann L, Fishman D,
Berchuck A, Whitaker R, et al. Patterns of
gene expression in different histotypes of
epithelial ovarian cancer correlate with
those in normal fallopian tube,
endometrium, and colon. Clin Cancer Res
2005;11:6116–26.
16. Gram IT, Braaten T, Lund E, Le ML,
Weiderpass E. Cigarette smoking and risk
of colorectal cancer among Norwegian
women. Cancer Causes Control 2009;20:
895–903.
17. Leufkens AM, van Duijnhoven FJ, Siersema
PD, Boshuizen HC, Vrieling A, Agudo A,
Gram IT, Weiderpass E, Dahm C, Overvad
K, Tjonneland A, Olsen A, et al. Cigarette
smoking and colorectal cancer risk in the
European prospective investigation into
cancer and nutrition study. Clin
Gastroenterol Hepatol 2011;9:
137–44.
18. Al-Zoughool M, Dossus L, Kaaks R,
Clavel-Chapelon F, Tjonneland A, Olsen A,
Overvad K, Boutron-Ruault MC, Gauthier
E, Linseisen J, Chang-Claude J, Boeing H,
et al. Risk of endometrial cancer in
relationship to cigarette smoking: results
from the EPIC study. Int J Cancer 2007;
121:2741–7.
19. Lodovici M, Bigagli E. Biomarkers of
induced active and passive smoking
damage. Int J Environ Res Public Health
2009;6:874–88.
20. Hecht SS. Cigarette smoking: cancer risks,
carcinogens, and mechanisms. Langenbecks
Arch Surg 2006;391:603–13.
21. Zenzes MT, Puy LA, Bielecki R.
Immunodetection of benzo[a]pyrene
adducts in ovarian cells of women exposed
to cigarette smoke. Mol Hum Reprod 1998;
4:159–65.
Short Report
2210
Smoking and subtypes of ovarian cancer
Int. J. Cancer: 130, 2204–2210 (2012) V
C 2011 UICC