SEARCH

SEARCH BY CITATION

Keywords:

  • cervical cancer;
  • HPV;
  • adenocarcinoma;
  • squamous cell carcinoma

Abstract

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

Squamous cell carcinomas account for about 80% of cancers of the uterine cervix, and the majority of the remainder are adenocarcinomas. There is limited evidence on the extent to which these histological types share a common etiology. The International Collaboration of Epidemiological Studies of Cervical Cancer has brought together and combined individual data on 8,097 women with invasive squamous cell carcinoma, 1,374 women with invasive adenocarcinoma and 26,445 women without cervical cancer (controls) from 12 epidemiological studies. Compared to controls, the relative risk of each histological type of invasive cervical cancer was increased with increasing number of sexual partners, younger age at first intercourse, increasing parity, younger age at first full-term pregnancy and increasing duration of oral contraceptive use. Current smoking was associated with a significantly increased risk of squamous cell carcinoma (RR = 1.50, 95% CI: 1.35–1.66) but not of adenocarcinoma (RR = 0.86 (0.70–1.05)), and the difference between the two histological types was statistically significant (case-case comparison p < 0.001). A history of screening (assessed as having had at least one previous nondiagnostic cervical smear) was associated with a reduced risk of both histological types, but the reduction was significantly greater for squamous cell carcinoma than for adenocarcinoma (RR = 0.46 (0.42–0.50) and 0.68 (0.56–0.82), respectively; case–case comparison, p = 0.002). A positive test for cervical high-risk HPV-DNA was a strong risk factor for each histological type, with 74% of squamous cell carcinomas and 78% of adenocarcinomas testing positive for HPV types 16 or 18. Squamous cell and adenocarcinoma of the cervix share most risk factors, with the exception of smoking. © 2006 Wiley-Liss, Inc.

Squamous cell carcinomas account for about 80% of cancers of the uterine cervix, and the majority of the remainder are adenocarcinomas.1 While infection with high-risk types of the human papillomavirus (HPV) is the major cause of both of these types of cervical cancer,2, 3 there is limited evidence about the extent to which other risk factors are shared.3, 4 The International Collaboration of Epidemiological Studies of Cervical Cancer has brought together worldwide evidence on the effects of hormonal contraceptives and of other factors5, 6 on the risk of cervical cancer. Here we compare risk factors for the two main histological types of invasive cervical cancer in a reanalysis of individual patient data from 12 epidemiological studies of cervical cancer, which had classified women according to the histological type of their cancer.

Material and methods

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

Identification of studies and collection of data

Full details of the methods used in the collaboration have been published previously.6 For this report, epidemiological studies of invasive cervical cancer that included both women with squamous cell carcinoma and women with adenocarcinoma and that had information on the histological type of the tumor were eligible for inclusion. All studies included in the collaboration were required to have information on the use of hormonal contraceptives. Cohort (prospective) studies were eligible if they included at least 30 cases of invasive cervical cancer and case–control studies were eligible if they had at least 100 cases of invasive cervical cancer. Individual subject information was requested from principal investigators of each study and was checked and analyzed centrally so that variables could be defined as consistently as possible across studies. The one eligible cohort study was analyzed as a nested case–control study with up to 4 controls (women without cervical cancer) selected randomly and matched per case according to age in single years. Data on socio-economic factors, smoking history, reproductive factors, sexual behavior, hormonal contraceptive use, Pap smear history and, if available, on cervical HPV infection were coded from information provided by principal investigators in a standard format. To maximize comparability between studies, analyses of data on HPV type were restricted to studies that measured cervical HPV-DNA by a PCR-based method. HPV types were classified as “high risk” according to the definition of Muñoz et al.7 Although the range of HPV types included in the tests varied between the studies, all studies included tests for HPV 16 and 18.

Histological classification of cancers was based on that described by the World Health Organization (WHO) for tumors of the uterine cervix.8 Microinvasive tumors were classified as invasive. For the main analyses adenosquamous carcinomas were included in the category of adenocarcinoma. Women with carcinomas of unknown histological type (n = 282) and of specified histological types other than squamous cell or adenocarcinoma (n = 108) were excluded.

As before, analyses were restricted to women who were aged between 16 and 89 years, who had not had a hysterectomy and who had reported one or more sexual partners.6 For the analysis of screening history a woman was considered likely to have been screened if she had had at least one previous screening cervical (Papanicolau, or Pap) smear. In the few studies where it was not certain that diagnostic smears had been excluded, smears taken within the 12 months prior to diagnosis/pseudodiagnosis were excluded if possible. Adequate data on the number and timing of all previous Pap smears were not available for most of the studies, hence more detailed analyses of screening history were not feasible.

Statistical methods and presentation of data

Conditional logistic regression was used to calculate relative risks (RR) and their corresponding 95% confidence intervals (CI). When only two groups are compared, the relative risk of cervical cancer and the associated CI are presented. However, when more than two groups are compared, variances are estimated by treating the relative risks as floating absolute risks.9 Use of floating methods does not alter the estimates of relative risk, but yields floated standard errors and floated confidence intervals (FCIs) that enable valid comparisons to be made between any two exposure groups, even if neither is the baseline group. The unfloated (“conventional”) CIs are given whenever results for only two groups are compared (e.g. in the text). Tests for trend were carried out using the median value within a given category as the level for that category. All tests of statistical significance were two-sided.

In all analyses women were stratified by study (and within study by centre), by single year of age, and where appropriate, by age at first intercourse (<18, 18–20 and 21+ years), duration of oral contraceptive use (never, <10 and 10+ years' use), number of full-term pregnancies (0, 1–2, 3–4 and 5+), smoking status, previous Pap smear (yes/no) and by lifetime number of sexual partners (1, 2–5 and 6+). Where studies or individual women were missing information on any of these factors they were included by creating a category for “missing” for the relevant variable, and sensitivity analyses performed to assess the effect of this assumption.

Separate analyses of risk factors were conducted first for women with squamous cell carcinoma and for women with adenocarcinoma, taking women without cervical cancer as the comparison group. However, many of the controls in the included studies were the same for analyses of both squamous cell and adenocarcinoma cases and so RR estimates for each histological type are not independent and cannot be compared directly. A direct comparison between risk factors for squamous cell and adenocarcinomas was therefore made using a case–case analysis, directly comparing women with squamous cell and adenocarcinoma. Because the analyses were conducted using conditional logistic regression, the RR estimated from the case–case analysis is not necessarily equal to the ratio of the RRs from the two corresponding case–control analyses.

Results

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

Individual data from 12 studies10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29 were analyzed, including a total of 8,097 women with invasive squamous cell carcinoma of the cervix, 1,374 women with invasive adenocarcinoma and 26,445 women without carcinoma of the cervix (Table I). Median age at diagnosis was similar for both histological types (squamous cell carcinoma, 45 years and adenocarcinoma, 43 years).

Table I. Characteristics of the 12 Studies Included in this Analysis
Study nameCountryAdenocarcinoma casesSquamous cell cancer casesControls
nMedian year of diagnosisMedian age at diagnosisnMedian year of diagnosisMedian age at diagnosisnMedian age
Cohort studies
Million Women Study10UK572000551171999572,80056
Case–control studies, population controls
Brinton USA11USA6319834241719834279143
UK cervical cancer12UK18119863639719873492835
Daling Seattle13, 14USA1191991405581992401,42241
USA adeno15USA961994408319943830037
Latvia16Latvia1619996120219995423750
Case–control studies, hospital controls
WHO—developed countries17Australia51980373219804164735
Israel11198242651982371,92939
WHO—developing countries17Colombia61982462119824419444
Chile161982481161982431,05136
Kenya1019833810219823968133
Mexico361984432361983411,46738
Nigeria21980322519804014937
Philippines2819814112019814273342
Thailand (Chiang Mai)1111985436801985462,51446
Thailand (Chulalongkorn)881984434951984442,35740
Thailand (Siriraj)1101984446441984442,61341
Brinton Latin America18Colombia1119865419819864740845
Mexico2219864512319864529145
Panama1619864516219864831047
Costa Rica1119873916319864436644
Milan19Italy9619885146919855287854
IARCParaguay20719895010919894810144
Brazil211819914518119905122551
Thailand224119924434519915035450
Philippines233419924935319924738647
Morocco241619935219819934920340
Peru252519964917319964719649
Algeria261519985518319985320252
India271219984519319984721347
Bangkok28Thailand4619924324019924476140
Johannesburg29South Africa4919975569719975173853
Total 1,3741987438,09719874526,44543

Compared to women without cervical cancer, the RRs of both squamous cell carcinoma and adenocarcinoma of the cervix increased with increase in number of sexual partners and with decrease in age at first intercourse (Table II). When the cases were compared directly there was no significant difference between the histological types for number of sexual partners or age at first intercourse (p-trend for squamous cell compared to adenocarcinoma = 0.4 and 0.2, respectively).

Table II. Relative Risk (RR)1 of Invasive Cervical Cancer in Relation to Sexual Behavior, Reproductive Factors, Oral Contraceptive Use, Body Mass Index and Smoking
 nRR (and 95% FCI/CI2)
Squamous cell carcinomaAdenocarcinomaControlsSquamous cell vs controlsAdenocarcinoma vs controlsSquamous cell vs adenocarcinoma
  • 1

    RR stratified by study, age at diagnosis and, where appropriate, number of sexual partners, age at first intercourse, number of full-term pregnancies, smoking status, Pap smears and duration of oral contraceptive use; 95% floated confidence intervals (FCIs).

  • 2

    95% confidence intervals are given for duration of oral contraceptive use (per year use).

  • 3

    Trend in parous women only.

(a) Number of sexual partners
 13,99268415,8071.00 (0.94–1.06)1.00 (0.88–1.14)1.00 (0.84–1.19)
 2–53,1454495,8792.00 (1.91–2.09)1.63 (1.47–1.80)1.16 (1.05–1.29)
 6+7791641,2892.98 (2.62–3.40)2.64 (2.07–3.36)1.22 (0.92–1.62)
    p-trend < 0.001p-trend < 0.001p-trend = 0.4
(b) Age at first intercourse (years)
 21+1,5463107,9431.00 (0.93–1.07)1.00 (0.86–1.16)1.00 (0.83–1.20)
 18–202,4234387,5201.60 (1.51–1.68)1.50 (1.35–1.67)1.10 (0.96–1.24)
 <183,2685166,8862.24 (2.11–2.38)2.06 (1.83–2.33)1.15 (0.99–1.34)
    p-trend < 0.001p-trend < 0.001p-trend = 0.2
(c) Number of full-term pregnancies
 Nulliparous4211332,9300.69 (0.60–0.78)0.94 (0.74–1.18)0.84 (0.63–1.12)
 1–22,1344569,8541.00 (0.94–1.07)1.00 (0.89–1.13)1.00 (0.86–1.16)
 3–42,3624127,3881.50 (1.43–1.59)1.36 (1.22–1.52)0.99 (0.87–1.12)
 5+3,1583726,2462.08 (1.95–2.23)1.61 (1.37–1.90)1.31 (1.10–1.55)
    p-trend3< 0.001p-trend3< 0.001p-trend3= 0.01
(d) Age at first birth (years)
 Nulliparous4211332,9300.85 (0.75–0.97)1.14 (0.91–1.44)0.93 (0.70–1.23)
 25+1,2502556,4771.00 (0.93–1.08)1.00 (0.86–1.16)1.00 (0.84–1.20)
 20–242,9995099,9421.66 (1.57–1.74)1.44 (1.30–1.60)1.23 (1.09–1.40)
 17–192,3933725,4762.16 (2.03–2.30)2.10 (1.84–2.40)1.09 (0.94–1.26)
 <17982981,4642.72 (2.42–3.05)2.01 (1.53–2.65)1.49 (1.12–1.99)
    p-trend3< 0.001p-trend3< 0.001p-trend3= 0.1
(e) Duration of oral contraceptive use (per year) by recency2
 Current users508951,6171.08 (1.06–1.09)1.07 (1.04–1.11)1.00 (0.97–1.04)
 2–9 years since stopping542991,7581.03 (1.02–1.05)1.03 (1.00–1.06)1.00 (0.97–1.03)
 10+ years since stopping5311271,4520.98 (0.96–1.00)1.02 (0.97–1.07)0.95 (0.91–0.99)
    p-trend < 0.001p-trend < 0.001p-trend = 0.1
(f) Body mass index (kg/m2)
 20–251,2653273,9981.00 (0.93–1.08)1.00 (0.88–1.14)1.00 (0.86–1.17)
 25–305641692,1301.05 (0.94–1.18)1.24 (1.02–1.50)0.99 (0.78–1.25)
 30+200658040.94 (0.78–1.14)1.13 (0.83–1.55)0.85 (0.59–1.25)
 <20380828921.14 (0.99–1.32)0.97 (0.74–1.27)1.11 (0.81–1.51)
    p-trend = 0.3p-trend = 0.1p-trend = 0.3
(g) Smoking status
 Never4,39076011,9581.00 (0.94–1.06)1.00 (0.89–1.13)1.00 (0.86–1.17)
 Past6361361,9681.01 (0.91–1.14)0.74 (0.60–0.92)1.35 (1.06–1.74)*
 Current1,3962272,6541.50 (1.38–1.63)0.86 (0.73–1.01)1.66 (1.38–1.99)**
      p = 0.04*, < 0.001**

Compared to women without cervical cancer, the RRs for both histological types increased with increase in number of full-term pregnancies (Table II), and the case–case comparison showed a significantly greater trend for squamous cell compared to adenocarcinoma (p = 0.01). Younger age at first full-term pregnancy was associated with an increased risk of cervical cancer that did not differ significantly between squamous cell and adenocarcinoma (p trend = 0.1) There were no significant differences between squamous cell and adenocarcinomas for use of combined oral contraceptives or for body mass index.

There was, however, a highly significant difference between the two histological types with respect to the risks associated with tobacco smoking (Table II). Current smoking was associated with a significantly increased risk of squamous cell carcinoma (RR = 1.50 (95% CI: 1.35–1.66)), but not of adenocarcinoma (RR = 0.86 (0.70–1.05)), and on case–case analysis the difference between the histological types was highly statistically significant (p < 0.001).

Data on cervical HPV-DNA from a PCR-based test were available for 2,084 women with squamous cell carcinoma, 248 with adenocarcinoma and 2,656 controls from 4 studies.14, 16, 20, 21, 22, 23, 24, 25, 26, 27, 28 Taking women without cervical cancer as the comparison group, a positive test for cervical HPV-DNA was a highly significant risk factor for both histological types, but the magnitude of the risk associated with testing positive for a “high-risk” HPV type was greater for squamous cell (RR = 189 (95% CI: 133–267)) than for adenocarcinoma (RR = 110 (55–220)); when the two histological types were compared directly this difference was statistically significant (p = 0.007) (Table III). HPV 16 was the most common type in cases of both histological types (Table IV). A greater proportion of squamous cell than of adenocarcinoma cases tested positive for HPV 16 (56% versus 40%, p-difference < 0.001); but a greater proportion of adenocarcinoma than of squamous cell carcinoma cases tested positive for HPV 18 (35% versus 16%, p-difference < 0.001). A significantly higher proportion of squamous cell than of adenocarcinoma cases tested positive only for high-risk types other than HPV 16 and 18 (15% verses 4%, p-difference < 0.001).

Table III. Relative Risk (RR)1 of Invasive Cervical Cancer in Relation to Testing Positive for HPV-DNA
 nRR (and 95% CI)
Squamous cell carcinomaAdenocarcinomaControlsSquamous cell vs controlsAdenocarcinoma vs controlsSquamous cell vs adenocarcinoma
  • HR, high risk.

  • 1

    RR stratified by study, age at diagnosis; 95% confidence intervals (CI).

HPV negative130362,233111.00
HPV positive (any type)1,954212423114 (85–154)65 (36–119)2.10 (1.23–2.56)
HPV positive (HR types)1,856204259189 (133–267)110 (55–220)2.10 (1.23–2.59)
Table IV. Numbers and Proportions of Women Who Tested Positive for High Risk HPV-DNA According to Positivity for Types 16 and 18
 Squamous cell carcinomaAdenocarcinomaControls
n%n%n%
HPV 16 +ve1,1595698401235
HPV 18 +ve333168835442
HPV 16 and 18 +ve4327311<1
Positive for other high-risk types only32115114813
Positive for low-risk types only985831646
HPV −ve130636152,23384
Total (with test results)2,0841002481002,656100

Compared to women without cervical cancer, having had a previous Pap smear was associated with a marked reduction in risk of squamous cell carcinoma (RR = 0.46 (0.42–0.50)) and with a significant but smaller reduction in risk of adenocarcinoma (0.68 (0.56–0.82)); the difference between types was statistically significant (p = 0.002) (Table V). Exclusion of 4 studies in which diagnostic smears could not reliably be excluded12, 13, 14, 19, 26 did not materially alter the findings with respect to Pap smear history.

Table V. Relative Risk (RR)1 of Invasive Cervical Cancer in Relation to Previous Pap Smears
 nRR (and 95% CI)
Squamous cell carcinomaAdenocarcinomaControlsSquamous cell vs controlsAdenocarcinoma vs controlsSquamous cell vs adenocarcinoma
  • 1

    RR Stratified by study, age at diagnosis, number of sexual partners, age at first intercourse, number of full-term pregnancies, smoking status and duration of oral contraceptive use; 95% confidence intervals (CI).

No previous Pap smear4,14654911,0251.001.001.00
Previous Pap smear2,2055899,1640.46 (0.42–0.50)0.68 (0.56–0.82)0.70 (0.56–0.88)

Seven studies in this collaborative reanalysis had used independent histological review of cases.13, 14, 15, 16, 17, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28 When analyses were restricted to these studies the results were not materially altered (Appendix Table AI; all of the studies with PCR-based HPV information had reviewed case histology). Similarly, sensitivity analyses in which women with missing information on one or more confounding variables were omitted produced results very similar to those of the main analyses.

In total 180 women from 5 studies11, 15, 17, 18, 28 were known to have been diagnosed with adenosquamous carcinoma (26% of all “adenocarcinoma” cases in these studies). In the remaining studies it was not possible to distinguish between adenocarcinoma and adenosquamous carcinoma. Analyses of risk factors for adenosquamous carcinoma were conducted, but generally lacked statistical power due to the relatively small number of cases (Appendix Table AII). With respect to cervical HPV infection, test results were available for only 21 cases of adenosquamous cancer of which 13 tested positive for high-risk HPV types (six for HPV type 16 and seven for HPV type 18).

Discussion

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

This collaborative reanalysis of individual data from 12 epidemiological studies is the largest and most comprehensive direct comparison to date of risk factors for squamous cell and adenocarcinoma of the cervix. It demonstrates that there are no substantial differences between the two most common histological types of invasive cervical cancer with respect to the role of number of sexual partners, age at first intercourse, age at first birth, body mass index and use of oral contraceptives. The exception is current tobacco smoking, which is associated with an increased risk of squamous cell but not of adenocarcinoma of the cervix. There is some evidence that parity may be a stronger risk factor for squamous cell than for adenocarcinoma, which others have also suggested.3

Although a positive test for high-risk HPV was a highly significant risk factor for both histological types, the magnitude of the risk was greater for squamous cell than for adenocarcinoma. The lower prevalence of HPV positivity in adenocarcinomas may partly reflect differences in the extent of HPV typing in different studies. Where individual studies have used exhaustive typing there appears to be no substantial difference in prevalence of HPV positivity between squamous cell and adenocarcinomas of the cervix.30

Screening by cervical smears appears to be effective in reducing the risk of both histological types of cervical cancer. The results of this analysis are to some extent limited by lack of detailed data on women's pattern of cervical screening but potential biases are minimized by using a case–case comparison and the findings are consistent with the view that cervical screening may reduce the risk of squamous cell carcinoma more than that of adenocarcinoma.31

References

  1. Top of page
  2. Abstract
  3. Material and methods
  4. Results
  5. Discussion
  6. References
  7. Appendix
  • 1
    ParkinDM, WhelanSL, FerlayJ, TeppoL, ThomasDB, eds. Cancer incidence in five continents, vol. VIII. Lyon: IARC Press, 2002. IARC Scientific Publications No. 155.
  • 2
    Clifford GM, Smith JS, Plummer M, Munoz N, Franceschi S. Human papillomavirusvirus types in invasive cervical cancer worldwide: a meta-analysis. Br J Cancer 2003; 88: 6373.
  • 3
    Castellsague X, Diaz M, de Sanjose S, Munoz N, Herrero R, Franceschi S, Peeling RW, Ashley R, Smith JS, Snijders PJ, Meijer CJ, Bosch FX. Worldwide human papillomavirus etiology of cervical adenocarcinoma and its cofactors: implications for screening and prevention. J Natl Cancer Inst 2006; 98: 30315.
  • 4
    Berrington de González A, Sweetland S, Green J. Comparison of risk factors for squamous cell and adenocarcinomas of the cervix: a meta-analysis. Br J Cancer 2004; 90: 178791.
  • 5
    International Collaboration of Epidemiological Studies of Cervical Cancer. Cervical carcinoma and reproductive factors: collaborative reanalysis of individual data on 16,563 women with cervical cancer and 33,542 women without cervical cancer from 25 epidemiological studies. Int J Cancer 2006; 119: 110824.
  • 6
    International Collaboration of Epidemiological Studies of Cervical Cancer. Carcinoma of the cervix and tobacco smoking: collaborative re-analysis of individual data on 15,756 women with carcinoma of the cervix and 29,043 women without carcinoma of the cervix from 23 epidemiological studies. Int J Cancer 2006; 118: 148195.
  • 7
    Munoz N, Bosch FX, de Sanjose S, Herrero R, Castellsague X, Shah KV, Snijders PJ, Meijer CJ. International Agency for Research on Cancer Multicenter Cervical Cancer Study Group. Epidemiologic classification of human papillomavirus types associated with cervical cancer. N Engl J Med 2003; 348: 51827.
  • 8
    WHO. In: TavassoliFA, StrattonMR, eds. WHO classification of tumours: pathology and genetics of tumours of the breast and female genital organs. Lyon: IARC Press, 2002. Chapter 5: Tumours of the Uterine Cervix, pp 25989.
  • 9
    Plummer M. Improved estimates of floating absolute risk. Stat Med 2004; 23: 93104.
  • 10
    Beral V. Breast cancer and hormone-replacement therapy in the Million Women Study. Lancet 2003; 362: 41927.
  • 11
    Brinton LA, Huggins GR, Lehman HF, Mallin K, Savitz DA, Trapido E, Rosenthal J, Hoover R. Long-term use of oral contraceptives and risk of invasive cervical cancer. Int J Cancer 1986; 38: 33944.
  • 12
    Green J, Berrington de Gonzalez A, Sweetland S, Beral V, Chilvers C, Crossley B, Deacon J, Hermon C, Jha P, Mant D, Peto J, Pike M, Vessey MP. Risk factors for adenocarcinoma and squamous cell carcinoma of the cervix in women aged 20–44 years: the UK National case-control study of cervical cancer. Br J Cancer 2003; 89: 207886.
  • 13
    Daling JR, Madeleine MM, McKnight B, Carter JJ, Wipf GC, Ashley R, Schwartz SM, Beckmann AM, Hagensee ME, Mandelson MT, Galloway DA. The relationship of human papillomavirus-related cervical tumors to cigarette smoking, oral contraceptive use, and prior herpes simplex virus type 2 infection. Cancer Epidemiol Biomarkers Prev 1996; 5: 5418.
  • 14
    Madeleine MM, Daling JR, Schwartz SM, Shera K, McKnight B, Carter JJ, Wipf GC, Critchlow CW, McDougall JK, Porter P, Galloway DA. Human papillomavirus and long-term oral contraceptive use increase the risk of adenocarcinoma in situ of the cervix. Cancer Epidemiol Biomarkers Prev 2001; 10: 1717.
  • 15
    Lacey JVJ, Brinton LA, Abbas FM, Barnes WA, Gravitt PE, Greenberg MD, Greene SM, Hadjmichael OC, McGowan L, Mortel R, Schwartz PE, Silverberg SG et al. Oral contraceptives as risk factors for cervical adenocarcinomas and squamous cell carcinomas. Cancer Epidemiol Biomarkers Prev 1999; 8: 107985.
  • 16
    Silins I, Wang X, Tadesse A, Jansen KU, Schiller JT, Avall-Lundqvist E, Frankendal B, Dillner J. A population-based study of cervical carcinoma and HPV infection in Latvia. Gynecol Oncol 2004; 93: 484492.
  • 17
    WHO. Invasive squamous-cell cervical carcinoma and combined oral contraceptives: results from a multinational study. WHO collaborative study of neoplasia and steroid contraceptives. Int J Cancer 1993; 55: 22836.
  • 18
    Herrero R, Brinton L, Reeves WC, Brenes MM, Tenorio F, de Britton R, Gaitan E, Montalvan P, Garcia M, Rawls WE. Risk factors for invasive carcinoma of the uterine cervix in Latin America. Bull Pan Am Health Organ 1990; 24: 26383.
  • 19
    Parazzini F, La Vecchia C, Negri E, Maggi R. Oral contraceptive use and invasive cervical cancer. Int J Epidemiol 1990; 19: 25963.
  • 20
    Rolon PA, Smith JS, Munoz N, Klug SJ, Herrero R, Bosch X, Llamosas F, Meijer CJ, Walboomers JM. Human papillomavirus infection and invasive cervical cancer in Paraguay. Int J Cancer 2000; 85: 48691.
  • 21
    Eluf-Neto J, Booth M, Munoz N, Bosch FX, Meijer CJ, Walboomers JM. Human papillomavirus and invasive cervical cancer in Brazil. Br J Cancer 1994; 69: 11419.
  • 22
    Chichareon S, Herrero R, Munoz N, Bosch FX, Jacobs MV, Deacon J, Santamaria M, Chongsuvivatwong V, Meijer CJ, Walboomers JM. Risk factors for cervical cancer in Thailand: a case-control study. J Natl Cancer Inst 1998; 90: 507.
  • 23
    Ngelangel C, Munoz N, Bosch FX, Limson GM, Festin MR, Deacon J, Jacobs MV, Santamaria M, Meijer CJ, Walboomers JM. Causes of cervical cancer in the Philippines: a case-control study. J Natl Cancer Inst 1998; 90: 439.
  • 24
    Chaouki N, Bosch FX, Munoz N, Meijer CJ, El Gueddari B, El Ghazi A, Deacon J, Castellsague X, Walboomers JM. The viral origin of cervical cancer in Rabat, Morocco. Int J Cancer 1998; 75: 54654.
  • 25
    Santos C, Munoz N, Klug SJ, Almonte M, Guerrero E, Alvarez M, Velarde C, Galdos O, Walboomers JM, Meijer C, Caceres E. HPV types and cofactors causing cervical cancer in Peru. Br J Cancer 2001; 85: 96671.
  • 26
    Hammouda D, Munoz N, Herrero R, Arlsan A, Bouhadef A, Oublil M, Djedeat B, Fontaniere B, Snijders P, Meijer C, Franceschi S. Cervical carcinoma in Algiers, Algeria: human papillomavirus and lifestyle risk factors. Int J Cancer 2004; 113: 4839.
  • 27
    Rajkumar T, Franceschi S, Vaccarella S, Gajalakshmi V, Sharmila A, Snijders P, Munoz N, Herrero R. Role of paan chewing and dietary habits in cervical carcinoma in Chennai, India. Br J Cancer 2003; 88: 138893.
  • 28
    Thomas DB, Qin Q, Kuypers J, Kiviat N, Ashley R, Koetsawang A, Ray RM, Koetsawang S. Human papillomaviruses and cervical cancer in Bangkok. II. Risk factors for in situ and invasive squamous cell cervical carcinomas. Am J Epidemiol 2001; 153: 7329.
  • 29
    Sitas F, Pacella-Norman R, Carrara H, Patel M, Ruff P, Sur R, Jentsch U, Hale M, Rowji P, Saffer D, Connor M, Bull D, Newton R, Beral V. The spectrum of HIV-1 related cancers in South Africa. Int J Cancer 2000; 88: 48992.
  • 30
    Zielinski GD, Snijders PJ, Rozendaal L, Daalmeijer NF, Risse EK, Voorhorst FJ, Jiwa NM, van der Linden HC, de Schipper FA, Runsink AP, Meijer CJ. The presence of high-risk HPV combined with specific p53 and p16INK4a expression patterns points to high-risk HPV as the main causative agent for adenocarcinoma in situ and adenocarcinoma of the cervix. J Pathol 2003; 201: 53543.
  • 31
    Sasieni P, Adams J. Changing rates of adenocarcinoma and adenosquamous carcinoma of the cervix in England. Lancet 2001; 357: 14903.

Appendix

  1. Top of page
  2. Abstract
  3. Material and methods
  4. Results
  5. Discussion
  6. References
  7. Appendix
Table AI. Relative Risk (RR)1 of Invasive Cervical Cancer in Relation to Sexual Behavior, Reproductive Factors, Oral Contraceptive Use, Body Mass Index, Smoking and in Relation to Previous Pap Smears (7 Studies with Independent Histology Review Only)
 nRR (and 95% FCI/CI2)
Squamous cell carcinomaAdenocarcinomaControlsSquamous cell vs controlsAdenocarcinoma vs controlsSquamous cell vs adenocarcinoma
  • 1

    RR stratified by study, age at diagnosis and, where appropriate, number of sexual partners, age at first intercourse, number of full-term pregnancies, smoking status, Pap smears and duration of oral contraceptive use; 95% floated confidence intervals (FCIs).

  • 2

    95% confidence interval presented for duration of oral contraceptive use (RR per year use) or previous Pap smear.

  • 3

    Trend in parous women only.

(a) Number of sexual partners
 13,49457314,1711.00 (0.93–1.08)1.00 (0.86–1.17)1.00 (0.82–1.22)
 2–51,8312634,0851.97 (1.87–2.08)1.65 (1.46–1.87)1.05 (0.94–1.19)
 6+4551099243.00 (2.50–3.60)2.95 (2.11–4.13)1.04 (0.68–1.58)
    p-trend < 0.001p-trend < 0.001p-trend = 0.9
(b) Age at first intercourse (years)
 21+1,3702717,1471.00 (0.93–1.08)1.00 (0.86–1.17)1.00 (0.83–1.21)
 18–202,0063276,4611.57 (1.48–1.67)1.41 (1.24–1.60)1.10 (0.95–1.27)
 <182,4093625,6562.09 (1.95–2.24)1.91 (1.66–2.20)1.10 (0.92–1.31)
    p-trend < 0.001p-trend < 0.001p-trend = 0.5
(c) Number of full-term pregnancies
 Nulliparous278952,2260.64 (0.54–0.75)1.03 (0.78–1.38)0.65 (0.45–0.93)
 1–21,4812756,8671.00 (0.93–1.08)1.00 (0.86–1.17)1.00 (0.83–1.20)
 3–41,7053055,5521.59 (1.50–1.69)1.61 (1.42–1.83)0.88 (0.76–1.02)
 5+2,3532895,1562.12 (1.97–2.29)1.76 (1.48–2.11)1.19 (0.98–1.44)
    p-trend3 < 0.001p-trend3 < 0.001p-trend3 = 0.05
(d) Age at first birth (years)
 Nulliparous278952,2260.79 (0.68–0.93)1.14 (0.86–1.51)0.76 (0.53–1.10)
 25+8931804,6411.00 (0.92–1.09)1.00 (0.84–1.19)1.00 (0.81–1.24)
 20–242,2553637,4281.78 (1.68–1.88)1.46 (1.29–1.65)1.26 (1.09–1.46)
 17–191,7382554,3252.27 (2.11–2.44)1.94 (1.65–2.27)1.12 (0.94–1.33)
 <17640681,1302.77 (2.42–3.19)1.85 (1.34–2.56)1.57 (1.11–2.21)
    p-trend3 < 0.001p-trend3 < 0.001p-trend3 = 0.1
(e) Duration of oral contraceptive use (per year) by recency2
 Current users7471581,9651.08 (1.06–1.09)1.07 (1.04–1.11)1.00 (0.97–1.04)
 2–9 years since stopping8231892,4441.04 (1.02–1.06)1.04 (1.00–1.08)1.00 (0.96–1.05)
 10+ years since stopping7231883,1180.98 (0.96–1.01)1.02 (0.97–1.07)0.93 (0.87–0.98)
    p-trend < 0.001p-trend < 0.001p-trend = 0.1
(f) Body mass index (kg/m2)
 20–256881631,5581.00 (0.90–1.11)1.00 (0.82–1.22)1.00 (0.79–1.27)
 25–30300886471.11 (0.95–1.30)1.27 (0.97–1.65)1.09 (0.79–1.51)
 30+117412331.10 (0.85–1.42)1.37 (0.89–2.09)0.83 (0.51–1.35)
 <20185293671.17 (0.94–1.44)0.81 (0.51–1.28)1.17 (0.73–1.92)
    p-trend = 0.9p-trend = 0.04p-trend = 0.5
(g) Smoking status
 Never3,2495618,2221.00 (0.92–1.08)1.00 (0.85–1.17)1.00 (0.80–1.25)
 Past323666991.02 (0.87–1.20)0.55 (0.38–0.78)1.88 (1.26–2.80)
 Current7811031,2721.33 (1.19–1.49)0.65 (0.50–0.84)1.75 (1.35–2.27)
(h) Pap smear2
 No previous Pap smear3,76850810,4901.001.001.00
 Previous Pap smear1,7144048,1450.42 (0.38–0.47)0.59 (0.48–0.73)0.67 (0.53–0.85)
Table AII. Relative Risk (RR)1 of Invasive Adenosquamous Cervical Cancer in Relation to Sexual Behavior, Reproductive Factors, Oral Contraceptive Use, Body Mass Index, Smoking Status and in Relation to Previous Pap Smears
 nRR (and 95% FCI/CI2)
Adenosquamous carcinomaControlsAdenosquamous vs controls
  • 1

    RR stratified by study, age at diagnosis and, where appropriate, number of sexual partners, age at first intercourse, number of full-term pregnancies, smoking status and duration of oral contraceptive use; 95% floated confidence intervals (FCIs).

  • 2

    95% confidence intervals for duration of oral contraceptive use (per year use) or previous Pap smear.

  • 3

    Trend in parous women only.

(a) Number of sexual partners
 19912,5531.00 (0.71–1.42)
 2–5663,9222.06 (1.65–2.57)
 6+94542.37 (0.94–5.95)
   p-trend < 0.001
(b) Age at first intercourse (years)
 21+396,0721.00 (0.67–1.48)
 18–20515,7321.37 (1.00–1.87)
 <18895,2642.73 (2.06–3.63)
   p-trend < 0.001
(c) Number of full-term pregnancies
 Nulliparous141,7820.91 (0.43–1.90)
 1–2447151.00 (0.70–1.43)
 3–4575,0501.23 (0.92–1.63)
 5+655,0141.12 (0.78–1.61)
   p-trend3 = 0.7
(d) Age at first birth (years)
 Nulliparous141,7821.15 (0.55–2.37)
 25+273,7421.00 (0.64–1.56)
 20–24586,6731.18 (0.89–1.58)
 <20784,1451.93 (1.42–2.65)
   p-trend3 = 0.05
(e) Duration of oral contraceptive use (per year) by recency2
 Current users181,4621.04 (0.96–1.12)
 2–9 years since stopping171,6361.00 (0.98–1.01)
 10+ years since stopping121,0501.06 (0.92–1.22)
   p-trend = 0.5
(f) Body mass index (kg/m2)
 20–25188651.00 (0.60–1.68)
 25–30134391.30 (0.69–2.52)
 30+41530.70 (0.16–2.97)
 <20102821.50 (0.68–3.32)
   p-trend = 0.5
(g) Smoking status
 Never846,3901.00 (0.70–1.42)
 Past104750.57 (0.20–1.65)
 Current151,0740.67 (0.36–1.25)
(h) Previous Pap smears2
 No previous Pap smear1019,5321.00
 Previous Pap smear445,7380.65 (0.42–1.00)