Objective
The purpose of this study was to assess cigarette use and environmental smoke exposure in women with cervical cancer.
Study Design
Smoking behavior was recorded prospectively in a clinical trial of women with locally advanced cervical carcinoma.
Results
Of 315 participants, 133 women (42%) were current smokers; 72 women (23%) were former smokers, and 110 women (35%) were never smokers. Current smokers began smoking earlier (16 vs 18 years; P = .009), for more years (29 vs 24 years; P = .005), and in greater amounts (20 vs 11 cigarettes/d; P < .001) than former smokers. Active smokers lived more often with another smoker (63.3%), compared with former smokers (35.0%; P < .001) or never-smokers (28.7%; P < .001). Agreement between self-report and urine cotinine level was high (kappa = 0.872; P < .001). A significant decrease in cotinine level during treatment occurred in 5.2% of current smokers.
Conclusion
Prevalence of smoking and tobacco consumption was twice that of the North American female population. Few smokers quit or decreased consumption during treatment.
Cigarette smoking is an important risk factor for the development of squamous cell carcinoma of the cervix. The estimated prevalence of cigarette smoking in North American women who are diagnosed with cervical cancer is approximately 50%, or more than twice the prevalence among adult women in the United States. Smoking also has an adverse effect on survival for women with locally advanced cervical cancer. In a recent Gynecologic Oncology Group (GOG) treatment study (GOG 165), unadjusted Cox survival analyses revealed that current smoking was associated with a 45% increase in the risk of disease progression (95% confidence interval [CI], 1.05–1.99; P = .024) and a 58% increase in the risk of death (95% CI, 1.11–2.23; P = .011). After adjustment for patient age, race, performance status, Federation of International Gynecologic Oncology stage, tumor grade, histologic subtype, pelvic node status, and treatment, current smoking remained a significant predictor of worse survival (hazard ratio, 1.51; 95% CI, 1.01–2.27; P = .045). Analyses performed with cotinine-derived smoking status (current smoker [>50 ng/mL] vs non-smoker [<50 ng/mL]) produced similar results in regards to progression-free and overall survival as did self-reported smoking status. The present study was conducted to gain insight into the extent of tobacco addiction and exposure to environmental smoke in subjects with locally advanced cervical cancer. These factors directly impact the success of smoking cessation efforts and may help explain the lack of observed smoking cessation in the study group during treatment.
It is hoped that smoking cessation before or during treatment for cervical cancer may abrogate the poorer prognosis that has been identified among smokers. Smoking is associated with impaired oxygen delivery to tissues because of small vessel constriction and/or increased levels of carboxyhemoglobin. Tumor hypoxia has been associated in clinical studies with more aggressive cancer behavior. Radiation therapy is the primary treatment for women with locally advanced cervical cancer, and improvements in tumor oxygenation after smoking cessation may be reflected in a better response to treatment. Success at smoking cessation has been linked to many factors that include desire to quit, daily cigarette consumption, and exposure to household or environmental cigarette smoke. These variables have not been well studied in women with cervical cancer. As part of GOG 165, a standardized questionnaire that focused on smoking behavior was completed for study subjects before they began chemoradiation therapy. Participants also provided a urine sample for measurement of cotinine before and during treatment to validate reported smoking behavior and to provide an estimate of changes in cigarette use during therapy. Results from the questionnaire and cotinine analysis allowed comparisons between never-smokers, former smokers, and current smokers.
Materials and Methods
Patients
GOG 165 was designed to detect a 33% decrease in the hazard ratio that is associated with disease progression when patients with locally advanced cervical cancer were treated with radiation plus 5-fluorouracil (5-FU), compared with radiation plus cisplatin. The period of enrollment was from October 1997 through August 2000. The study was closed prematurely when interim analysis of the 328 enrolled subjects indicated that the 5-FU treatment arm would not result in an improvement in progression-free survival, compared with weekly cisplatin-based chemoradiation. Patients who were enrolled in GOG 165 had primary, previously untreated, histologically confirmed (by central review of the GOG Pathology Committee) invasive squamous cell carcinoma, adenocarcinoma, or adenosquamous carcinoma of the uterine cervix (stage II-B, III-B, IV-A). Ineligibility criteria included patients with an initial performance score of 4, a positive pregnancy test, and/or positive aortic nodes as determined by lymphangiogram, lymphadenectomy, computed tomography, or magnetic resonance imaging.
Informed consent
The subject consent form for GOG 165 focused mainly on the issues that pertained to the administration of the 2 possible chemoradiation regimens. Local institutional review board approval was obtained. With regard to the smoking component of the trial, the following information was included in the consent: “I understand that current therapies for cervical cancer which has spread beyond the cervix do not cure all patients. For some types of cancer it appears that cigarette smoking may influence the effectiveness of treatment. I understand that another aim of this study is to determine if smoking or being exposed to other’s cigarette smoke influences the effectiveness of the type of therapy I will be receiving for my cervical cancer. I understand that my participation in the study is requested even if I have never smoked or have quit smoking.” Given the lack of any consistent link between smoking and prognosis in this cohort, study sites were not required to provide further advice or recommendations regarding smoking cessation to the study subjects.
Smoking questionnaire
All patients who participated in GOG 165 were required to complete a questionnaire regarding their current and previous smoking history that included an estimate of environmental tobacco smoke exposure before the onset of chemoradiation therapy. The questionnaire was adapted from the National Health Interview Survey and was administered by personnel interview at each GOG site. Smoking status was assigned based on reported smoking behavior in the questionnaire. Subjects were considered to be current smokers if they reported having previously smoked at least 100 cigarettes during their lifetime and were still smoking at least 1 cigarette per week. Nonsmokers included never-smokers (subjects who had smoked <100 cigarettes during their lifetime) and former smokers. Former smokers reported smoking at least 100 cigarettes ever and currently smoking <1 cigarette per week.
Cotinine assessment in a pretreatment urine specimen
Cotinine is the major metabolite of nicotine, is chemically stable, and has a relatively long half-life (19-40 hours), which permits the estimation of long-term tobacco exposure. Analysis of urine cotinine levels was performed by the Institute for Cancer Prevention (Valhalla, NY) with a modification of the enzyme immunoassay as developed by Orasure Technologies, Inc. (OTI, Bethlehem, PA). The lower limit of detection with the enzyme immunoassay is 5 ng cotinine/mL urine. Levels >50 ng/mL are consistent with active smoking. Detectable levels <50 ng/mL are consistent with exposure to environmental smoke or infrequent, light, personal smoking.
Statistical analysis
Statistical analyses were performed with Statistical Package for the Social Sciences (version 10.1; SPSS Inc, Chicago, IL) and Statistical Analysis System (version 8.2; SAS Institute Inc., Cary, NC). Patient characteristics and smoking behaviors were compared between never, former, and current smokers with the Pearson Chi-square method (for categoric variables ) or nonparametric Kruskal-Wallis test (for continuous variables ). Spearman correlation coefficient was generated as a nonparametric measure of the correlation between 2 ordinal variables. All tests were 2-sided. Cohen’s kappa was used as a measure of the agreement between self-reported current smoking status and cotinine-derived current smoking status. A kappa of 1 indicated perfect agreement, whereas a value of 0 indicates that agreement is no better than chance.
Results
There were 328 patients enrolled in GOG 165 who were assigned randomly to receive radiation plus cisplatin vs radiation plus prolonged venous infusion 5-FU. Twelve patients were excluded for various reasons that included incorrect primary, wrong stage, or improper surgical staging. Among the 316 eligible patients, 315 patients completed the questionnaire and were evaluable for the smoking component of this study. The median patient age at protocol entry was 48 years. Sixty-two percent of the women were white; 20% of the women were African American; 13% of the women were Hispanic; 4% of the women were Asian/Pacific Islander, and 1% of the women were Native American. The following percentages were the distribution by Federation of International Gynecologic Oncology stage: 65% stage IIB, 31% stage IIIB, and 4% stage IV-A. Results from the smoking questionnaire categorized 133 women (42%) as current smokers and 182 women (58%) as non-smokers. Of the nonsmokers, 72 women were former smokers, and 110 women were never-smokers. There is evidence that current smokers were significantly younger and more often white, compared with former or never-smokers ( Table 1 ).
| Characteristic | Total (n = 315) | Smoking status | |||
|---|---|---|---|---|---|
| Never (n = 110) | Former (n = 72) | Current (n = 133) | P value | ||
| Age, y a | 48 (27–84) | 49 (29–84) | 52 (27-84) | 46 (28–74) | .003 |
| Race, n (%) | < .001 | ||||
| White | 196 (62.2) | 51 (46.4) | 48 (66.7) | 97 (72.9) | |
| African American | 63 (20.0) | 21 (19.1) | 13 (18.1) | 29 (21.8) | |
| Other b | 56 (17.8) | 38 (34.6) | 11 (15.3) | 7 (5.3) | |
a Data given as median (range);
b Includes Asian/Pacific Islander, 12 women; Hispanic, 42 women, and Native American, 2 women.
Details regarding subjects’ smoking histories and exposure to environmental or second-hand smoke are summarized in Table 2 . Compared with former smokers, current smokers started smoking cigarettes at an earlier age (median, 16 vs 18 years; P = .009) and smoked more years (median, 29 vs 24 years; P = .005). When asked to recall the average number of cigarettes smoked per day since starting smoking, current smokers smoked more cigarettes per day than former smokers (median, 20 vs 11 cigarettes per day; P < .001). Current smokers also reported smoking a similar number of cigarettes per day in the previous week, compared with the average number of cigarettes smoked per day since starting smoking, which indicated that it was unlikely that these subjects reduced their tobacco intake since being informed of the diagnosis of cervical cancer.
| Self-reported status | Smoking status a | P value | ||
|---|---|---|---|---|
| Never (n = 110) | Former (n = 72) | Current (n = 133) | ||
| Age smoked first cigarette, y | .009 | |||
| Median (range) | Not applicable | 18 (8–50) | 16 (5–45) | |
| Interquartile range | 15–20 | 14–18 | ||
| Calculated no. of years | .005 | |||
| Median (range) | Not applicable | 24 (1–58) | 29 (7–55) | |
| Interquartile range | 16–35 | 24–36 | ||
| Average no. of cigarettes smoked/day since starting smoking | < .001 | |||
| Median (range) | Not applicable | 11 (1–60) | 20 (1–80) | |
| Interquartile range | 5–20 | 10–20 | ||
| Average no. of cigarettes smoked/d in the previous week | ||||
| Median (range) | Not applicable | Not applicable | 20 (0–140) | |
| Interquartile range | 7–22 | |||
| Currently lives with a smoker, n (%) | < .001 | |||
| No | 75 (70.8) | 39 (65) | 44 (36.7) | |
| Yes | 31 (29.2) | 21 (35) | 76 (63.3) | |
| No answer provided | 4 | 12 | 13 | |
| No. of cigarettes smoked by housemate(s)/d while subject is at home | n = 30 | n = 29 | n = 82 | .369 |
| Median (range) | 14 (0–60) | 6 (0–60) | 20 (0–60) | |
| Interquartile range | 9–30 | 1–35 | 7–26 | |
| Exposed to smoke outside the home | n = 110 | n = 72 | n = 113 | < .001 |
| No, n (%) | 69 (62.7) | 31 (43.1) | 29 (22) | |
| Yes, n (%) | 41 (37.3) | 41 (56.9) | 103 (78) | |
| No answer provided | 0 | 0 | 1 | |
| No. of hr/wk exposed to smoke outside the home | n = 44 | n = 33 | n = 99 | < .001 |
| Median (range) | 2 (0–56) | 3 (0–80) | 8 (0–100) | |
| Interquartile range | 1–5 | 1–6 | 3–20 | |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
