SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. Disclosure Statement
  8. Abbreviations
  9. References

The role of tobacco smoking in the multistage carcinogenesis at the cervix is not fully understood because of a paucity of prospective data. To assess the relationship between smoking and spontaneous regression of cervical precursor lesions, a total of 516 women with low-grade squamous intraepithelial lesion (LSIL) were monitored by cytology and colposcopy every 4 months. Probability of LSIL regression within 2 years was analyzed in relation to smoking behaviors, with regression defined as at least two consecutive negative Pap smears and normal colposcopy. Women’s age, initial biopsy results, and human papillomavirus (HPV) genotypes were included in the multivariate models for adjustments. Our study subjects included 258 never-smokers and 258 smokers (179 current and 79 former smokers). During a mean follow-up time of 39.8 months, 320 lesions regressed to normal cytology. Probability of regression within 2 years was significantly lower in smokers than in never-smokers (55.0%vs 68.8%, P = 0.004). The risk of LSIL persistence increased with smoking intensity and duration and with younger age at starting smoking (P = 0.003, P < 0.001, and P = 0.03, respectively). Smokers had twice as high a risk of persistent HPV infection compared to never-smokers (odds ratio, 2.50; 95% confidence interval, 1.30–4.81; P = 0.006). In young women, passive smoking since childhood reduced probability of regression within 2 years (56.7%vs 85.9%, P < 0.001). Further adjustments for a wide range of cervical cancer risk factors did not change the findings. In conclusion, tobacco smoking may interfere with regression of cervical precursor lesions. Childhood exposure to second-hand smoke may increase a risk of persistent cervical abnormalities among young women. (Cancer Sci 2010)

Tobacco smoking is the leading preventable causes of death. There is convincing evidence that tobacco smoking causes cardiovascular disease, chronic lung disease, and several epithelial cancers including lung cancer.(1,2) Tobacco use is estimated to cause 5.4 million deaths annually worldwide.(3) This number is projected to rise to more than 8 million by 2030.(4) In the USA, prospective data from 104 519 women demonstrated that 64% of deaths among current smokers and 28% of deaths among former smokers were attributable to tobacco smoking.(5) Nevertheless, several countries are experiencing an increase in the prevalence of smoking, particularly among younger women.(2)

Cervical cancer remains the second most common cancer in women worldwide, with nearly 500 000 women developing the disease every year.(6) Infections by oncogenic human papilloma viruses (HPVs) are established as a major risk factor for cervical cancer.(7) In addition, a few prospective studies(8,9) and many case-control studies(10,11) have consistently demonstrated that women who smoke are at an increased risk of developing cervical cancer and its immediate premalignant lesions (CIN 3, cervical intraepithelial neoplasia grade 3). However, the role of tobacco smoking in multistage carcinogenesis is not fully understood because of a paucity of prospective data. Moreover, much less is known about a causal relationship between passive smoke exposure and cervical cancer.(12,13)

Most low-grade cervical lesions are known to regress spontaneously, whereas only a small fraction progress to cervical cancer.(14–18) Using follow-up data of 516 women with mildly abnormal cervical cytology (LSIL, low-grade squamous intraepithelial lesion) found not to have CIN 3 or worse by initial colposcopy-directed biopsy, we examined the relationship between tobacco smoking and natural history of low-grade cervical abnormalities.

Materials and Methods

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. Disclosure Statement
  8. Abbreviations
  9. References

Study design.  This analysis was based on data from a prospective non-intervention cohort study conducted by the JHACC (Japan HPV And Cervical Cancer) Study Group for identifying determinants of LSIL/CIN regression and progression. Between April 1998 and August 2004, 905 women with low-grade cervical abnormalities were recruited from nine hospitals that performed conventional Pap smears, colposcopy, and cervical biopsies. They entered the study only after giving signed informed consent voluntarily. Cervical smears were classified according to the 2001 Bethesda System as normal, atypical, LSIL, or high-grade squamous intraepithelial lesion (HSIL).(19) Only at the entry, two small specimens were taken by colposcopy-directed punch biopsy and histological diagnosis was made using hematoxylin–eosin (H&E)-stained sections according to the World Health Organization (WHO) classification. The inclusion criteria of this analysis were evident LSIL cytology, biopsy-positive CIN 1 or CIN 2 lesion, age 18 to 54 years, and first detection of cervical abnormalities. Two cytopathologists (Y.H. and Masafumi Tsuzuku) and two pathologists (R.F. and T.K.) reviewed all cytological and histological specimens at entry. After the central cytology and pathology review, 225 women were excluded from the study: 34 women with HSIL cytology or CIN 3 histology, 90 women with LSIL cytology alone, 60 women with histological CIN 1–2 findings alone, and 41 women without LSIL or CIN 1–2 findings. We excluded women with cytology-positive and biopsy-negative lesions from the analysis because they were more likely to have regression than women with both positive lesions. Also, we excluded women with cytology-negative and biopsy-positive lesions from the analysis because we could not define cytological regression for them. At enrollment, the study subjects were tested for cervical HPV-DNA and serum IgG antibodies to Chlamydia trachomatis and herpes simplex virus type 2 (HSV2). Information about smoking, alcohol drinking, contraceptive and reproductive history, and sexual behavior was obtained from a self-administrated questionnaire. Questions on smoking habits included status (never, former, or current smoker), intensity (number of cigarettes smoked per day), duration of the habit (age at starting and, for former smokers, age at cessation), and household exposure to second-hand smoke (status and duration). Former smokers were defined as those women who had abstained from tobacco smoking for at least 12 months prior to the entry. To assess a dose-dependent relationship between active smoking and risk of LSIL persistence and progression, smoking behaviors were classified as smoking intensity (never, using 1–9 cigarettes per day, and using ≥10 cigarettes per day) and smoking duration (never, using 1–9 years, and using ≥10 years). The Brinkman Index (BI) was calculated by multiplying the average number of cigarettes smoked per day by the smoking years.

A total of 680 eligible subjects were routinely followed at 3- to 4-month intervals and received cytology and colposcopic examinations at each visit. They were re-tested for cervical HPV-DNA at 24 months. To avoid interference by the biopsy procedure on natural course of the disease, a cervical biopsy was performed only when women had HSIL smears and major colposcopic changes (dense aceto-white change, iodine negativity, and coarse punctation/wide irregular mosaics of differing size) that were suggestive of progression to CIN 3 or worse. For women who were regarded as progressing based on cytology and histology in the participating hospitals, two cytopathologists and two pathologists reviewed all cytological and histological specimens collected for diagnosis of disease progression. In this analysis, progression was defined as histological CIN 3 lesions or worse diagnosed upon rigorous pathology review. Occasionally a few difficult cases were adjudicated by joint review with consideration of cytology as well as histology. We defined regression as at least two consecutive negative smears and normal colposcopy. Women were regarded as having persistent lesions when they did not have either regression or progression over the period of follow-up.

In this study, 110 women had an insufficient number (<2) of follow-up visits to be included in the analysis and 54 women gave no answer to a self-administrated questionnaire on smoking. Therefore, we analyzed the follow-up data from 516 women who reported their smoking habits completely in entry questionnaires.

Institutional ethical and research review boards of the participating institutions approved the study protocol.

Human papillomavirus (HPV) testing.  We detected HPV-DNA in cervical samples by polymerase chain reaction (PCR)-based methodology described previously.(20) In brief, exfoliated cells from the ectocervix and endocervix were collected in a tube containing 1 mL of (phosphate-buffered saline) PBS and stored at −30°C until DNA extraction. Total cellular DNA was extracted from cervical samples by a standard sodium dodecyl sulfate (SDS)-proteinase K procedure. Human papillomavirus (HPV)-DNA was amplified by PCR using consensus-primers (L1C1/L1C2 + L1C2M) for the HPV L1 region. A reaction mixture without template DNA was included in every set of PCR runs as a negative control. Also, primers for a fragment of the β-actin gene were used as a control to rule out false-negative results for samples in which HPV-DNA was not detected. Human papillomavirus (HPV) types were identified by restriction fragment-length polymorphism (RFLP) that has been shown to identify at least 26 types of genital HPVs.(21) Based on HPV type prevalence data from the International Agency for Research on Cancer (IARC)-pooled analyses(22,23) and meta-analyses,(24,25) 13 oncogenic HPV types detected by the Hybrid Capture 2 (Digene, Gaithersburg, MD, USA) and Amplicor HPV tests (Roche Molecular Systems, Branchburg, NJ, USA) were separated into two groups in the present study. Group 1 high-risk (G1HR) types included HPV16, 18, 31, 33, 35, 45, 52, and 58, the eight most common oncogenic types associated with invasive cervical cancer worldwide. Group 2 high-risk (G2HR) types consisted of HPV39, 51, 56, 59, and 68. All other HPV types were classified as low-risk types (non-oncogenic).

Serology.  The level of IgG antibodies to Chlamydia trachomatis and HSV2 was determined by using commercially available enzyme-linked immunosorbent assay (ELISA) kits: HITAZYME (Hitachi Chemical, Tokyo, Japan) for Chlamydia trachomatis and HerpeSelect 2 ELISA IgG (Focus Diagnostics, Cypress, CA, USA) for HSV2. The serological assay for Chlamydia trachomatis utilized purified EB outer-membrane proteins of the Chlamydia trachomatis L2 strain as antigens and did not detect antibody to Chlamydia pneumoniae.(26) These serological assays were performed at a clinical testing laboratory (SRL, Tokyo, Japan).

Statistical analysis.  All time-to-event analyses were based on the actual date of the visits because analyses estimating date of event occurrence as the mid-point between two visits did not change the findings. For regression or progression, time to event was measured from the date of the index visit (i.e. the first instance of an abnormal cytology result) to the date of the visit at which cytological transition to normal or CIN 3 was first detected. Women whose lesions persisted or who dropped out of the study were censored at their last recorded return visit dates. Subjects who had only one negative colposcop/cytology result before loss to follow-up were censored at the last date of positive Pap tests. Subjects who were biopsied were censored at the time of their biopsy, regardless of the biopsy results, to reduce the potential for interference by the biopsy procedure on estimates of time of regression. Cumulative probability of disease regression or progression was estimated using the Kaplan–Meier method and compared with the log-rank test, and the Cox proportional-hazards model was used for statistical adjustments. Women’s age (20–29, 30–39, or 40–54 years), initial biopsy results (CIN 1 or CIN 2), and HPV risk category (group 1 high-risk, group 2 high-risk, low-risk HPV types, or HPV negative) were included as confounding factors in the multivariate models for statistical adjustments. Since additional adjustments for other cervical cancer risk factors (lifetime numbers of sexual partners, age at first sexual intercourse, parity, oral contraceptives use, and IgG antibodies to Chlamydia trachomatis and HSV2) did not change the findings, we mainly presented the results adjusted for the basic confounding factors of women’s age, initial biopsy results, and HPV risk category (basic adjusted models). For comparison of characteristics of our study subjects, Student’s t-test and χ2-test were used. All analyses were carried out using the STATA 9 (StataCorp, College station, TX, USA) statistical packages. Two-sided P-values were calculated throughout and considered to be significant at <0.05.

Results

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. Disclosure Statement
  8. Abbreviations
  9. References

We analyzed the follow-up data from 516 women who reported their smoking habits completely in entry questionnaires. The mean age of the study subjects was 35.9 years (median, 36.0; range, 20–54). They completed 4493 visits and the mean follow-up time was 39.8 months (range, 6.8–86.4). During the period of follow-up, 320 lesions spontaneously regressed to normal cytology, while 42 progressed to CIN 3.

Table 1 shows the basic demographics of the 516 women included in this analysis. At enrollment, 258 women had never smoked while 258 women were ever-smokers, including 179 current smokers and 79 former smokers. Ever-smokers were significantly younger than never-smokers (mean age, 34.3 years vs 37.6 years; Student’s t-test: P < 0.001). Smokers were more likely than never-smokers to have CIN 2 lesions (rather than CIN 1), although the differences were not statistically significant (18.2%vs 13.9%, χ2-test: P = 0.19). The detection rate and distribution of oncogenic HPV types did not appreciably differ between smokers and never smokers (χ2-test: P = 0.15). However, the differences in lifetime numbers of sexual partners, age at first sexual intercourse, oral contraceptive use, and IgG antibody to Chlamydia trachomatis were statistically significant between smokers and never-smokers (P < 0.05), probably reflecting higher sexual activity in smokers.

Table 1.   Distributions of baseline characteristics between smokers and never-smokers
 Never-smokersSmokersP-value*
AllCurrentFormer
  1. *Differences between smokers and never-smokers were analyzed by χ2-test. CIN, cervical intraepithelial neoplasia grade 3; HPV, human papillomavirus; HSV2, herpes simplex virus type 2. Bold letters denote statistical significance.

Number of women25825817979 
Age (years)
 Mean (SD)37.6 (7.1)34.3 (7.4)33.5 (7.4)36.0 (7.2)<0.001
 20–2935 (13%)64 (25%)49 (27%)15 (19%)
 30–39115 (45%)130 (50%)95 (53%)35 (44%)
 ≥40108 (42%)64 (25%)35 (20%)29 (37%)
Histology at entry
 CIN grade 1222 (86%)211 (82%)143 (80%)68 (86%)0.19
 CIN grade 236 (14%)47 (18%)36 (20%)11 (14%)
HPV genotypes
 16, 18, 31, 33, 35, 45, 52, 58118 (47%)139 (56%)94 (54%)45 (58%)0.15
 39, 51, 56, 59, 6870 (28%)56 (22%)43 (25%)13 (17%)
 Low-risk types or HPV negative61 (25%)55 (22%)36 (21%)19 (25%)
Number of lifetime sexual partners
 0–170 (28%)14 (5%)4 (2%)10 (13%)<0.001
 2–383 (33%)50 (20%)34 (20%)16 (21%)
 ≥499 (39%)188 (75%)137 (78%)51 (66%)
Age at first intercourse (years)
 ≤1846 (18%)135 (54%)101 (58%)34 (44%)<0.001
 19–21120 (47%)90 (36%)62 (35%)28 (36%)
 ≥2288 (35%)27 (11%)12 (7%)15 (19%)
Parity
 080 (31%)100 (39%)74 (41%)26 (33%)0.09
 1–2135 (52%)128 (50%)88 (49%)40 (51%)
 ≥343 (17%)30 (12%)17 (10%)13 (16%)
Use of oral contraceptives
 Users16 (6%)31 (12%)23 (13%)8 (10%)0.02
 Non-users238 (94%)218 (88%)149 (87%)69 (90%)
IgG antibody to Chlamydia trachomatis
 Low90 (36%)76 (30%)50 (29%)26 (35%)0.04
 Medium88 (36%)77 (31%)51 (29%)26 (35%)
 High69 (28%)97 (39%)74 (42%)23 (30%)
IgG antibody to HSV2
 Low85 (34%)79 (32%)56 (32%)23 (31%)0.62
 Medium78 (32%)89 (36%)58 (33%)31 (41%)
 High84 (34%)82 (33%)61 (35%)21 (28%)

Smoking and risk of LSIL persistence and progression.  Low-grade squamous intraepithelial lesion (LSIL) was significantly less likely to regress among ever-smokers than among never-smokers (probability of regression within 2 years: 55.0%vs 68.8%, log-rank test: P = 0.005, Fig. 1a). Even after adjusting for age, histological grades, and HPV risk category, the difference was still statistically significant (P = 0.004). In addition, the risk of LSIL persistence was significant for both current and former smokers (both P = 0.02, Table 2), indicating that the excess risk of LSIL persistence due to smoking was not eliminated rapidly upon cessation. The risk of LSIL persistence approached a never-smoker’s risk more than 10 years after quitting smoking (P for trend = 0.007, Table 2).

image

Figure 1.  Risk of low-grade squamous intraepithelial lesion (LSIL) persistence and progression in relation to smoking behaviors. Kaplan–Meier plot was used to estimate cumulative probabilities of LSIL persistence and progression in relation to smoking behaviors. Low-grade squamous intraepithelial lesion (LSIL) was more likely to persist long among smokers than among never-smokers (log-rank test P = 0.005, a), while active smoking did not correlate with risk of progression within 5 years (P = 0.73, b).

Download figure to PowerPoint

Table 2.   Active smoking and cumulative probability of LSIL regression within 2 years
Smoking behaviorsnEventsProbabilitiy of regression within 2 yearsMedian time to regression mo. (95% CI)Hazard ratio for regression (95% CI)
UnadjustedP-valueAdjusted*P-value*
  1. *Adjusted for women’s age (20–29, 30–39, or 40–54 years), initial biopsy results (CIN 1 or CIN 2), and HPV risk category (HPV16/18/31/33/35/45/52/58, other high-risk types, low-risk types, or HPV negative). †For the analysis of age at smoking initiation, smoking duration was also added to the multivariate models. CI, confidence interval; CIN 1, cervical intraepithelial neoplasia grade 1; HPV, human papillomavirus; LSIL, low-grade squamous intraepithelial lesion; NA, not available.; Bold letters denote statistical significance.

Current/former status
 Never-smokers25816568.6 (62.5–74.6)11.1 (7.9–13.8)1 1 
 Smokers25812655.0 (48.4–61.9)19.3 (13.5–25.5)0.72 (0.57–0.90)0.0050.70 (0.55–0.89)0.004
  Current1799057.2 (49.2–65.5)16.6 (12.9–25.5)0.75 (0.58–0.97)0.030.73 (0.56–0.95)0.02
  Former793650.1 (39.0–62.5)22.6 (11.0–41.4)0.64 (0.45–0.93)0.020.64 (0.44–0.93)0.02
Time since quitting smoking
 Never-smokers25816568.6 (62.5–74.6)11.1 (7.9–13.8)1 1 
 ≥10 years since quitting smoking311554.3 (36.7–73.9)22.6 (8.2–NA)0.69 (0.41–1.17)0.170.89 (0.52–1.53)0.68
 <10 years since quitting smoking482147.6 (33.8–63.6)NA0.61 (0.39–0.97)0.040.52 (0.32–0.85)0.01
 Current smokers1799057.2 (49.2–65.5)16.6 (12.9–NA)0.75 (0.58–0.97)0.030.72 (0.55–0.95)0.02
             P for trend 0.02          P for trend 0.007
Cigarettes per day
  025816568.6 (62.5–74.6)11.1 (7.9–13.8)1 1 
 1–9753858.2 (51.7–64.7)18.1 (12.0–26.3)0.75 (0.53–1.07)0.110.76 (0.53–1.09)0.13
 ≥101828753.5 (45.8–61.8)20.9 (13.3–31.0)0.70 (0.54–0.91)0.0070.67 (0.51–0.88)0.004
             P for trend 0.006          P for trend 0.003
Smoking duration (years)
  025816568.6 (62.5–74.6)11.1 (7.9–13.8)1 1 
 1–9865267.9 (56.6–78.6)9.1 (6.1–15.8)1.09 (0.80–1.49)0.580.95 (0.68–1.33)0.78
 ≥101727448.8 (40.9–57.4)25.5 (18.5–48.4)0.58 (0.44–0.76)<0.0010.60 (0.45–0.79)<0.001
            P for trend <0.001       P for trend  <0.001
Brinkman index
  025816568.6 (62.5–74.6)11.1 (7.9–13.8)1 1 
 1–99965261.3 (50.5–72.3)13.5 (8.2–22.6)0.87 (0.64–1.19)0.390.86 (0.62–1.19)0.36
 ≥1001617351.1 (42.8–60.0)21.0 (15.8–41.5)0.63 (0.48–0.83)0.0010.61 (0.46–0.82)0.001
             P for trend 0.001          P for trend 0.001
Age of smoking initiation (years)
 ≥181909557.0 (49.2–65.0)18.5 (13.3–24.9)1 1 
 <18673048.5 (36.5–62.0)25.5 (9.8–NA)0.85 (0.57–1.27)0.430.61 (0.39–0.95)†0.03†

The probability of progression within 5 years was similar between smokers and never-smokers (10.7%vs 13.0%, log-rank test: P = 0.73, Fig. 1b). Statistical adjustments did not change the results (P = 0.56 in the basic adjusted model).

Intensity, duration, and age at initiation of smoking.  Current and former smokers were combined in all subsequent analyses because similar results were obtained from current and former smokers. Even after adjusting for women’s age, initial biopsy results, and HPV risk category, both smoking intensity and duration strongly correlated with the risk of persisting LSIL cytology (P for trend = 0.003 and <0.001, respectively, Table 2). Accordingly, the risk of LSIL persistence also increased steadily with increasing BI (P for trend = 0.001, Fig. 2a). The median time to regression was twice longer among women with BI ≥100 than among never-smokers (21.0 months vs 11.1 months, Table 2). However, active smoking did not increase the 5-year risk of progression to CIN 3 even among the women with greater smoking intensity and duration (P for trend = 0.56, Fig. 2b).

image

Figure 2.  Dose-dependent relationship between tobacco smoking and low-grade squamous intraepithelial lesion (LSIL) persistence. The Brinkman Index (BI) was calculated by multiplying the average number of cigarettes smoked per day by the smoking years. Kaplan–Meier plot revealed a dose–response relationship between smoking and risk of LSIL persistence (log-rank test: P = 0.001, a). However, active smoking did not increase the 5-year risk of progression to CIN 3 even among the women with greater smoking intensity and duration (P = 0.56, b).

Download figure to PowerPoint

Furthermore, LSIL was less likely to regress among women who started smoking at the age of <18 years compared to ≥18 years (probability of regression within 2 years: 48.5%vs 57.0%). The difference was statistically significant after adjustment for women’s age, initial biopsy results, HPV genotypes, and smoking duration (hazard ratio for regression, 0.61; 95% confidence interval [CI], 0.39–0.95; P = 0.03, Table 2).

Persistent HPV infection.  Human papillomavirus (HPV)-DNA data at entry and 24 months were analyzed to examine whether tobacco smoking was associated with persistent HPV infections among women with low-grade cervical abnormalities. The results of second HPV testing at 24 months were available for 284 women. Inclusion criteria of this sub-analysis were as follows: women had to be HPV-DNA positive at entry and individual HPV genotypes had to be explicitly determined at both entry and 24 months. Overall, the sub-analysis consisted of 192 women including 88 smokers and 103 never-smokers. In this analysis, a woman was considered to have a persistent infection when at least one of the genotypes continued to be detected at 24 months. Genotype-specific persistence for any HPV genotypes was more common in smokers than in never-smokers (46.6%vs 27.2%, χ2-test: P = 0.007). Smokers had at least twice as high the risk of persistent infection compared with never smokers, after adjusting for women’s age, initial biopsy results and baseline HPV genotypes (odds ratio, 2.50; 95% CI, 1.30–4.81; P = 0.006).

Second-hand smoke exposure.  We analyzed an association between passive smoking and risk of LSIL persistence (Fig. 3). In young women aged <30 years, long-term passive smoke exposure (≥20 years) was significantly associated with reduced probability of regression within 2 years (56.7%vs 85.9%, log-rank test: P = 0.002). After adjusting for initial biopsy results, HPV genotypes and active smoking status, hazard ratio for LSIL regression was 0.29 (95% CI, 0.16–0.54) among young women exposed to second-hand smoke for ≥20 years compared with those never exposed or exposed for <20 years (P < 0.001). The risk of persistent LSIL related to second-hand smoke exposure was statistically significant for both young women who had smoked and who had never smoked (P = 0.002 and 0.03, respectively, Table 3). For women aged 30 years or more, however, long-term passive smoking did not correlate with the risk of LSIL persistence (Table 3). Even when the analysis was confined to the women who had been exposed to passive smoke for a longer period of time since childhood, we could not find any association between passive smoking and LSIL persistence in this age group (data not shown).

image

Figure 3.  Passive smoking and risk of low-grade squamous intraepithelial lesion (LSIL) persistence among young women. In young women aged <30 years, long-term secondhand smoke exposure (≥20 years) significantly reduced probability of regression within 2 years (56.7%vs 85.9%, log-rank test: P = 0.002). After adjusting for HPV genotypes, histology and active smoking status, hazard ratio for LSIL regression was 0.29 (95% confidence interval [CI], 0.16–0.54) among young women exposed to secondhand smoke for ≥20 years compared with those never exposed or exposed for <20 years (P < 0.001).

Download figure to PowerPoint

Table 3.   Passive smoking and cumulative probability of LSIL regression within 2 years
Passive smoke exposurenEventsProbabilitiy of regression within 2 yearsMedian time to regression mo. (95% CI)Hazard ratio for regression (95% CI)
UnadjustedP-valueAdjusted*P-value*
  1. *Adjusted for initial biopsy results (CIN 1 or CIN 2) and HPV risk category (HPV16/18/31/33/35/45/52/58, other high-risk types, low-risk types, or HPV negative). †The women were analyzed separately according to their own smoking status. CI, confidence interval; CIN 1, cervical intraepithelial neoplasia grade 1; HPV, human papillomavirus; LSIL, low-grade squamous intraepithelial lesion; NA, not available. Bold letters denote statistical significance.

Women aged <30 years
 Never exposed252086.6 (67.7–97.2)4.8 (3.8–7.6)1 1 
 Exposed704264.4 (52.1–76.3)10.6 (6.0–16.3)0.51 (0.30–0.87)0.010.47 (0.26–0.87)0.02
 Duration
  <20 years261885.8 (58.8–98.7)6.6 (4.2–12.0)0.75 (0.40–1.42)0.381.01 (0.49–2.10)0.97
  ≥20 years422254.4 (39.8–70.4)15.8 (7.2–NA)0.38 (0.21–0.70)0.0020.30 (0.15–0.59)0.001
          P for trend  0.001     P for trend  <0.001
 Never-smokers†
  Never or exposed <20 years211785.9 (65.9–97.2)5.7 (4.2–9.1)1 1 
  Exposed ≥20 years13756.7 (32.1–83.7)16.6 (5.3–NA)0.43 (0.18–1.04)0.060.30 (0.10–0.89)0.03
 Smokers†
  Never or exposed <20 years302177.0 (58.7–91.3)5.4 (4.0–7.7)1 1 
  Exposed ≥20 years281555.4 (37.8–74.7)12.4 (5.3– NA)0.48 (0.24–0.94)0.030.28 (0.12–0.62)0.002
Women aged ≥30 years
 Never exposed563058.3 (45.0–72.2)13.7 (7.9–NA)1 1 
 Exposed35819560.0 (54.5–65.6)16.6 (12.9–19.9)1.00 (0.68–1.47)0.991.24 (0.82–1.87)0.32
 Duration
  <20 years1688657.1 (52.9–61.3)18.1 (12.2–NA)0.93 (0.62–1.42)0.751.17 (0.75–1.84)0.48
  ≥20 years17710162.5 (58.6–66.5)16.3 (11.0–19.6)1.05 (0.70–1.58)0.801.29 (0.83–1.99)0.26
          P for trend  0.60     P for trend  0.25
 Never-smokers†
  Never or exposed <20 years1217363.7 (54.7–72.6)12.1 (7.9–17.3)1 1 
  Exposed ≥20 years926274.4 (64.1–83.6)10.7 (7.2–18.4)1.16 (0.82–1.62)0.401.15 (0.81–1.63)0.44
 Smokers†
  Never or exposed <20 years1004047.9 (42.1–53.7)NA1 1 
  Exposed ≥20 years823749.1 (43.2–55.0)NA1.13 (0.72–1.76)0.601.17 (0.75–1.84)0.49

Confounding by other risk factors.  One may speculate that the results presented herein may still reflect confounding effects by other factors, even after controlling for the strong effects of HPV infections, women’s age, and histological grade at entry. To examine residual confounding effects by other cervical cancer risk factors, we performed additional adjustments for possible confounding factors, such as lifetime numbers of sexual partners,(27) age at first sexual intercourse,(27) parity,(28) oral contraceptive use,(29) and infections by Chlamydia trachomatis(30) and HSV2.(31) Although several associations were attenuated by multiple adjustments, these analyses did not change the conclusions (Table 4). This suggests that the associations may be independent of these cervical cancer risk factors.

Table 4.   Cox proportional-hazards models illustrating hazard ratios (95% confidence intervals) for LSIL regression in relation to smoking behaviors
Tobacco smoke exposureHazard ratio for regression (95% confidence interval)
Basic adjusted model*P-valueFully adjusted model†P-valueFully adjusted model plus STD infections‡P-value
  1. *Adjusted for women’s age (20–29, 30–39, or 40–54 years), initial biopsy results (CIN 1 or CIN 2), and HPV risk category (HPV16/18/31/33/35/45/52/58, other high-risk types,low-risk types, or HPV negative). This basic model was used in Tables 1 and 2. †Adjusted for women’s age, initial biopsy results, HPV risk category, lifetime numbers of sexual partners (0–1, 2–3, or ≥4), age at fist sexual intercourse (≤18, 19–21, or ≥22), parity (0, 1–2, or ≥3) and oral contraceptives use (users or non-users). ‡Adjusted for women’s age, initial biopsy results, HPV risk category, lifetime numbers of sexual partners, age at fist sexual intercourse, parity, oral contraceptives use and IgG antibodies to Chlamydia trachomatis (low, medium, or high,) and herpes simplex virus type 2 (low, medium or high). §For the analysis of age at smoking initiation, smoking duration was also added to the multivariate models. ¶The women were analyzed separately according to their own smoking status. CIN 1, cervical intraepithelial neoplasia grade 1; HPV, human papillomavirus; LSIL, low-grade squamous intraepithelial lesion; STD, sexually transmitted disease. Bold letters denote statistical significance.

Active smoking
 Cigarettes per day
 1 1 1 
  1–90.76 (0.53–1.09)0.130.82 (0.57–1.18)0.290.86 (0.59–1.24)0.42
  ≥100.67 (0.51–0.88)0.0040.74 (0.54–1.00)0.050.75 (0.55–1.03)0.07
            P for trend 0.003         P for trend 0.04         P for trend 0.07
 Smoking duration (years)
 1 1 1 
  1–90.95 (0.68–1.33)0.781.00 (0.70–1.41)0.991.08 (0.75–1.54)0.68
  ≥100.60 (0.45–0.79)<0.0010.65 (0.48–0.89)0.0080.66 (0.48–0.91)0.01
           P for trend <0.001          P for trend 0.01          P for trend 0.01
 Brinkman Index
 1 1 1 
  1–90.86 (0.62–1.19)0.360.90 (0.64–1.26)0.540.94 (0.66–1.33)0.72
  ≥100.61 (0.46–0.82)0.0010.68 (0.49–0.94)0.020.69 (0.50–0.96)0.03
            P for trend 0.001         P for trend 0.02         P for trend 0.03
 Age at smoking initiation
  ≥181 1 1 
  <180.61 (0.39–0.95) §0.03§0.59 (0.35–0.99) §0.04§0.61 (0.36–1.02) §0.06 §
Passive smoking among women aged <30 years
 Never exposed1 1 1 
 Exposed0.50 (0.28–0.89)0.020.35 (0.17–0.71)0.0030.37 (0.18–0.76)0.007
 Duration
  <20 years1.06 (0.53–2.14)0.870.66 (0.28–1.55)0.340.80 (0.32–1.99)0.63
  ≥20 years0.32 (0.16–0.61)0.0010.23 (0.11–0.50)0.0010.24 (0.11–0.55)<0.001
           P for trend <0.001         P for trend <0.001         P for trend <0.001
 Never smokers¶
  Never or exposed <20 years1 1 1 
  Exposed ≥20 years0.30 (0.10–0.89)0.030.23 (0.06–0.79)0.020.03 (0.00–0.32)0.003
 Smokers¶
  Never or exposed <20 years1 1 1 
  Exposed ≥20 years0.28 (0.12–0.62)0.0020.23 (0.08–0.61)0.0030.18 (0.07–0.49)0.001

Discussion

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. Disclosure Statement
  8. Abbreviations
  9. References

Most low-grade cervical lesions are known to regress spontaneously, whereas only a small fraction progress to cervical cancer.(14–18) Several experimental studies have demonstrated direct oncogenic effects of chemical tobacco-related carcinogens,(32,33) while others have reported the modulation of host immunity by tobacco smoking.(34–36) Given natural history of low-grade cervical lesions, experimental evidence suggests that tobacco smoking may promote progression to cancer as a direct chemical carcinogen and/or interfere with lesion regression as an immune modulator suppressing T-cell responses against HPV infections.(37) Additionally, a recent in vitro study reported that exposure of cervical cells to benzo(a)pyrene, a major carcinogen found in cigarette smoke, induced high levels of HPV synthesis.(38) High viral load observed among current smokers may favor persistence of HPV-induced lesions.(39,40)

In our study, tobacco smoking significantly increased the risk of persistent LSIL smears, but did not correlate with progression to CIN 3 during the course of the study. In addition, data from second HPV testing at 24 months showed that a risk of persistent HPV infections was significantly increased among smokers, in keeping with several prospective studies of HPV infections.(41–43) These observations suggest that tobacco smoking may contribute to cervical carcinogenesis by interfering with regression of HPV-induced lesions rather than by promoting progression as a chemical carcinogen. Thus, the mechanism by which tobacco smoking contributes to carcinogenesis may differ between cervical cancer and other tobacco-related cancers.

Our data demonstrated a strong dose–response relationship between smoking and risk of persisting LSIL smears, implying for the beneficial effects of smoking reduction and cessation. In addition, younger age (<18 years) at smoking initiation significantly increased the risk of persistent LSIL even after adjustment for smoking duration, supporting school smoking prevention programs. These findings suggest that smoking intensity and duration and age at smoking initiation may be critical for determining the risk of lesion persistence among smokers. A large pooled analysis of 23 epidemiological studies also reported that the risk of cervical squamous cell carcinoma significantly increases with the number of cigarettes smoked per day and with younger age at starting smoking.(11)

The question is still open as to how quickly smoking cessation can reduce the excess risk of cervical cancer among smokers. In this study, the risk of LSIL persistence was also significant for former smokers, although their excess risk approached a never-smoker’s risk more than 10 years after quitting smoking. However, the collaborative reanalysis of case-control studies could not find a significant trend toward reduced risk of cervical cancer or CIN 3 with increasing time since stopping smoking.(11) A prospective study reported that smoking cessation reduced the size of low-grade cervical lesions, but probability of regression was not significantly different between the women who continued to smoke and who stopped smoking completely for at least 6 months.(44) Taken together, the risk of LSIL persistence due to smoking appears to persist at least in a short-term follow-up even after the women stop smoking. Among tobacco-related cancers, the beneficial effect of smoking cessation was first observed for lung cancer.(45) Accumulated evidence has shown that smoking cessation clearly decreases the excess risk of lung cancer, although the lung cancer risk in former smokers remains higher than in never-smokers, even after a long period (>20 years) of complete abstinence from smoking.(2,45,46)

Among young women under 30 years of age, those women exposed to second-hand smoke at home for 20 years or more, probably from fathers and mothers since very early stages of life, were found to be at an increased risk of persistent cervical abnormalities, independently of their own smoking status. In young women, the effect of long-term passive smoking was stronger than that of current smoking. Since the magnitude of exposure to toxic chemicals from second-hand smoke is far less than the exposure that occurs with active smoking, infants and children might be particularly susceptible to the adverse effects of second-hand smoke.

Interestingly, exposure to passive smoke for ≥20 years or during childhood did not increase risk of persistent LSIL cytology among women older than 30 years. Although it is difficult to offer a biologic explanation for the age-specific relationship between passive smoking and LSIL persistence, a possible explanation could be that passive smoke exposure may be a greater risk factor in women who have not experienced other major competing risk factors.

In this study, the questionnaires after enrollment revealed the high smoking rate in women with LSIL cytology. The number of former and current smokers (n = 218) was accidentally equal to that of never-smokers (n = 218). The rate of current smokers in the study subjects (35.0%) was much higher than the smoking rate in Japanese women aged 20–29 (14.3%), 30–39 (18.0%), or 40–49 (13.4%) in 2008.(47) However, this is not surprising. Previous case-control studies have consistently demonstrated higher smoking rates in women with low-grade cervical lesions compared with the control women: 33.8%vs 15.0% in Japan,(48) 34.7%vs 20.4% in the USA,(49) and 40.2%vs 21.7% in Sweden.(50)

The present study has several limitations that may have affected the conclusion. First, this cohort study investigated the natural history of prevalent, not incident, cervical abnormalities. Therefore, the differences in time from LSIL/CIN1-2 development to study enrollment could be a bias. Since tobacco use is a risk factor for persistent HPV infections,(41–43) women with long-term persistent lesions among smokers and women with relatively new lesions among never-smokers may have been preferentially collected as the study subjects. This may have affected the time to regression and the number of women regressing. Second, information regarding smoking was recruited only at enrollment, thus not accounting for the possible changes in smoking behaviors during the course of the study. Third, 110 women lacking a sufficient number of follow-up visits and 54 women missing data on smoking were excluded from the analysis. Although the age distributions, initial biopsy results and baseline HPV genotypes in the excluded and included women were similar (data not shown), this rate of loss could have influenced the results. Finally, the periods of follow-up were not very long (mean, 39.8 months; range, 6.8 to 86.4). In addition, current smokers were more like to be lost to follow-up compared with never-smokers (mean follow-up periods, 36.1 months vs 41.9 months). This may have resulted in an underestimation of the impact of smoking on progression. Although tobacco smoking was not associated with LSIL progression in a short-term follow-up, tobacco smoking should eventually increase the risk of disease progression in a long-term observation because women with persistent LSIL smears have more chances of progression to CIN 3. Therefore, the women with persistent lesions are still followed up by cytology and colposcopy to assess a long-term tobacco-related risk of progression to CIN 3 or worse. A longitudinal study of women that were positive for high-risk HPV types with normal cytology has demonstrated that the risk of CIN 3 diagnosed within 10 years is significantly elevated among smokers.(8)

In conclusion, tobacco smoking may interfere with spontaneous regression of precursor lesions. The results from this study support previous research that has shown that women who smoke are at an increased risk of developing cervical cancer and precancer.(8–11,37) Equally important, exposure to second-hand smoke since childhood may increase a risk of persistent cervical abnormalities among young women. Although further study is warranted to validate our results, protecting girls from second-hand smoke and stopping tobacco use in any age may reduce the future burden from cervical cancer.

Acknowledgments

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. Disclosure Statement
  8. Abbreviations
  9. References

The authors thank Dr Tadahito Kanda (Center for Pathogen Genomics, National Institute of Infectious Diseases, Tokyo, Japan) for his comments on the study design and the manuscript; Mr Masafumi Tsuzuku (Department of Cytopathology, Cancer Institute Hospital, Japanese Foundation of Cancer Research, Japan) for his cytological review; many others who facilitated this study; and all the women who participated in the study. This work was supported by a grant from the Ministry of Education, Science, Sports and Culture of Japan (grant number 12218102) and in part by a grant from the Smoking Research Foundation.

Abbreviations

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. Disclosure Statement
  8. Abbreviations
  9. References
CI

confidence interval

CIN

cervical intraepithelial neoplasia grade 3

HPV

human papillomavirus

HSV2

herpes simplex virus type 2

LSIL

low-grade squamous intraepithelial lesion

NA

not available

STD

sexually transmitted disease

References

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. Disclosure Statement
  8. Abbreviations
  9. References
  • 1
    Centers for Disease Control and Prevention (CDC). Smoking-attributable mortality, years of potential life lost, and productivity losses – United States, 2000-2004. MMWR Morb Mortal Wkly Rep 2008; 57: 12268.
  • 2
    World Health Organaization. WHO Report on the Global Tobacco Epidemic, 2009: Implementing smoke-free environments. Available at http://www.who.int/tobacco/mpower/2009/gtcr_download/en/index.html Accessed 10 April 2010.
  • 3
    Ezzati M, Lopez AD. Estimates of global mortality attributable to smoking in 2000. Lancet 2003; 362: 84752.
  • 4
    Mathers CD, Loncar D. Projections of global mortality and burden of disease from 2002 to 2030. PLoS Med 2006; 3: e442.
  • 5
    Kenfield SA, Stampfer MJ, Rosner BA, Colditz GA. Smoking and smoking cessation in relation to mortality in women. JAMA 2008; 299: 203747.
  • 6
    Parkin DM, Bray F, Ferlay J, Pisani P. Global cancer statistics, 2002. CA Cancer J Clin 2005; 55: 74108.
  • 7
    Schiffman M, Castle PE, Jeronimo J, Rodriguez AC, Wacholder S. Human papillomavirus and cervical cancer. Lancet 2007; 370: 890907.
  • 8
    Castle PE, Wacholder S, Lorincz AT et al. A prospective study of high-grade cervical neoplasia risk among human papillomavirus-infected women. J Natl Cancer Inst 2002; 94: 140614.
  • 9
    McIntyre-Seltman K, Castle PE, Guido R et al. ALTS Group. Smoking is a risk factor for cervical intraepithelial neoplasia grade 3 among oncogenic human papillomavirus DNA-positive women with equivocal or mildly abnormal cytology. Cancer Epidemiol Biomarkers Prev 2005; 14: 116570.
  • 10
    Plummer M, Herrero R, Franceschi S et al. IARC Multi-centre Cervical Cancer Study Group. Smoking and cervical cancer: pooled analysis of the IARC multi-centric case-control study. Cancer Causes Control 2003; 14: 80514.
  • 11
    International Collaboration of Epidemiological Studies of Cervical Cancer, Appleby P, Beral V, Berrington de Gonzalez A et al. Carcinoma of the cervix and tobacco smoking: collaborative reanalysis of individual data on 13,541 women with carcinoma of the cervix and 23,017 women without carcinoma of the cervix from 23 epidemiological studies. Int J Cancer 2006; 118: 148195.
  • 12
    Slattery ML, Robison LM, Schuman KL et al. Cigarette smoking and exposure to passive smoke are risk factors for cervical cancer. JAMA 1989; 261: 15938.
  • 13
    Trimble CL, Genkinger JM, Burke AE et al. Active and passive cigarette smoking and the risk of cervical neoplasia. Obstet Gynecol 2005; 105: 17481.
  • 14
    Östör AG. Natural history of cervical intraepithelial neoplasia: a critical review. Int J Gynecol Pathol 1993; 12: 18692.
  • 15
    Melnikow J, Nuovo J, Willan AR, Chan BK, Howell LP. Natural history of cervical squamous intraepithelial lesions: a meta-analysis. Obstet Gynecol 1998; 92: 72735.
  • 16
    Holowaty P, Miller AB, Rohan T, To T. Natural history of dysplasia of the uterine cervix. J Natl Cancer Inst 1999; 91: 2528.
  • 17
    Schlecht NF, Platt RW, Duarte-Franco E et al. Human papillomavirus infection and time to progression and regression of cervical intraepithelial neoplasia. J Natl Cancer Inst 2003; 95: 133643.
  • 18
    Castle PE, Schiffman M, Wheeler CM, Solomon D. Evidence for frequent regression of cervical intraepithelial neoplasia-grade 2. Obstet Gynecol 2009; 113: 1825.
  • 19
    Solomon D, Davey D, Kurman R et al. The Bethesda System 2001: terminology for reporting the results of cervical cytology. JAMA 2002; 287: 21149.
  • 20
    Yoshikawa H, Kawana T, Kitagawa K, Mizuno M, Yoshikura H, Iwamoto A. Detection and typing of multiple genital human papillomaviruses by DNA amplification with consensus primers. Jpn J Cancer Res 1991; 82: 52431.
  • 21
    Nagano H, Yoshikawa H, Kawana T et al. Association of multiple human papillomavirus types with vulvar neoplasias. J Obstet Gynaecol Res 1996; 22: 18.
  • 22
    Muñoz N, Bosch FX, De Sanjosé S et al. 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.
  • 23
    Muñoz N, Bosch FX, Castellsagué X et al. Against which human papillomavirus types shall we vaccinate and screen? The international perspective. Int J Cancer 2004; 111: 27885.
  • 24
    Clifford GM, Smith JS, Plummer M, Muñoz N, Franceschi S. Human papillomavirus types in invasive cervical cancer worldwide: a meta-analysis. Br J Cancer 2003; 88: 6373.
  • 25
    Smith JS, Lindsay L, Hoots B et al. Human papillomavirus type distribution in invasive cervical cancer and high-grade cervical lesions: a meta-analysis update. Int J Cancer 2007; 121: 62132.
  • 26
    Yamamoto T, Moji K, Kusano Y, Kurokawa K, Kawagoe K, Katamine S. Trend in Chlamydia trachomatis infection among pregnant women in the past ten years in Japan: significance of Chlamydia trachomatis seroprevalence. Sex Transm Dis 1998; 25: 51621.
  • 27
    International Collaboration of Epidemiological Studies of Cervical Cancer. Cervical carcinoma and reproductive factors: collaborative reanalysis of individual data on 16,563 women with cervical carcinoma and 33,542 women without cervical carcinoma from 25 epidemiological studies. Int J Cancer 2006; 119: 110824.
  • 28
    Muñoz N, Franceschi S, Bosetti C et al. International Agency for Research on Cancer. Multicentric Cervical Cancer Study Group. Role of parity and human papillomavirus in cervical cancer: the IARC multicentric case-control study. Lancet 2002; 359(9312): 1093101.
  • 29
    Moreno V, Bosch FX, Muñoz N et al. International Agency for Research on Cancer. Multicentric Cervical Cancer Study Group. Effect of oral contraceptives on risk of cervical cancer in women with human papillomavirus infection: the IARC multicentric case-control study. Lancet 2002; 359(9312): 108592.
  • 30
    Anttila T, Saikku P, Koskela P et al. Serotypes of Chlamydia trachomatis and risk for development of cervical squamous cell carcinoma. JAMA 2001; 285: 4751.
  • 31
    Smith JS, Herrero R, Bosetti C et al. International Agency for Research on Cancer (IARC) Multicentric Cervical Cancer Study Group. Herpes simplex virus-2 as a human papillomavirus cofactor in the etiology of invasive cervical cancer. J Natl Cancer Inst 2002; 94: 160413.
  • 32
    Prokopczyk B, Cox JE, Hoffmann D, Waggoner SE. Identification of tobacco-specific carcinogen in the cervical mucus of smokers and nonsmokers. J Natl Cancer Inst 1997; 89: 86873.
  • 33
    Melikian AA, Sun P, Prokopczyk B et al. Identification of benzo[a]pyrene metabolites in cervical mucus and DNA adducts in cervical tissues in humans by gas chromatography-mass spectrometry. Cancer Lett 1999; 146: 12734.
  • 34
    Stämpfli MR, Anderson GP. How cigarette smoke skews immune responses to promote infection, lung disease and cancer. Nat Rev Immunol 2009; 9: 37784.
  • 35
    Barton SE, Maddox PH, Jenkins D, Edwards R, Cuzick J, Singer A. Effect of cigarette smoking on cervical epithelial immunity: a mechanism for neoplastic change? Lancet 1988; 2: 6524.
  • 36
    Poppe WA, Ide PS, Drijkoningen MP, Lauweryns JM, Van Assche FA. Tobacco smoking impairs the local immunosurveillance in the uterine cervix An immunohistochemical study. Gynecol Obstet Invest 1995; 39: 348.
  • 37
    Castellsagué X, Muñoz N. Chapter 3: cofactors in human papillomavirus carcinogenesis – role of parity, oral contraceptives, and tobacco smoking. J Natl Cancer Inst Monogr 2003; 31: 208.
  • 38
    Alam S, Conway MJ, Chen HS, Meyers C. The cigarette smoke carcinogen benzo[a]pyrene enhances human papillomavirus synthesis. J Virol 2008; 82: 10538.
  • 39
    Xi LF, Koutsky LA, Castle PE et al. Relationship between cigarette smoking and human papilloma virus types 16 and 18 DNA load. Cancer Epidemiol Biomarkers Prev 2009; 18: 34906.
  • 40
    Xi LF, Hughes JP, Edelstein ZR et al. Human Papillomavirus (HPV) type 16 and type 18 DNA Loads at Baseline and Persistence of Type-Specific Infection during a 2-year follow-up. J Infect Dis 2009; 200: 178997.
  • 41
    Giuliano AR, Sedjo RL, Roe DJ et al. Clearance of oncogenic human papillomavirus (HPV) infection: effect of smoking (United States). Cancer Causes Control 2002; 13: 83946.
  • 42
    Kjaer SK, Munk C, Winther JF, Jørgensen HO, Meijer CJ, Van Den Brule AJ. Acquisition and persistence of human papillomavirus infection in younger men: a prospective follow-up study among Danish soldiers. Cancer Epidemiol Biomarkers Prev 2005; 14: 152833.
  • 43
    Maucort-Boulch D, Plummer M, Castle PE et al. Predictors of human papillomavirus persistence among women with equivocal or mildly abnormal cytology. Int J Cancer 2010; 126: 68491.
  • 44
    Szarewski A, Jarvis MJ, Sasieni P et al. Effect of smoking cessation on cervical lesion size. Lancet 1996; 347(9006): 9413.
  • 45
    US Department of Health and Human Services (USDHHS). The Health Benefits of Smoking Cessation: A Report of the SURGEON General. Atlanta, GA: US Department of Health and Human Services, Centers for Disease Control, Center for Chronic Disease Prevention and Health Promotion, Office on Smoking and Health, 1990.
  • 46
    Peto R, Darby S, Deo H, Silcocks P, Whitley E, Doll R. Smoking, smoking cessation, and lung cancer in the UK since 1950: combination of national statistics with two case-control studies. BMJ 2000; 321: 3239.
  • 47
    Military of Health, Labour and Welfare. A recent survey of tobacco use in Japan. Available at http://www.health-net.or.jp/tobacco/product/ pd100000.html, Accessed on 14 May 2010.
  • 48
    Yoshikawa H, Nagata C, Noda K et al. Human papillomavirus infection and other risk factors for cervical intraepithelial neoplasia in Japan. Br J Cancer 1999; 80: 6214.
  • 49
    Schiffman MH, Bauer HM, Hoover RN et al. Epidemiologic evidence showing that human papillomavirus infection causes most cervical intraepithelial neoplasia. J Natl Cancer Inst 1993; 85: 95864.
  • 50
    Kjellberg L, Hallmans G, Ahren AM et al. Smoking, diet, pregnancy and oral contraceptive use as risk factors for cervical intra-epithelial neoplasia in relation to human papillomavirus infection. Br J Cancer 2000; 82: 13328.