|
|
||||||||
Guest Access | Sign In via User Name/Password |
|||||||||
* From the University of California-Davis Medical Center.
Correspondence to: Susan Murin, MD, FCCP, 4150 V St, Suite 3400, Sacramento, CA 95817; e-mail: sxmurin{at}ucdavis.edu
| Abstract |
|---|
|
|
|---|
Design: A case-control study.
Setting: The University of California, Davis Medical Center.
Participants: Eighty-seven women patients with unilateral, invasive breast cancer and pulmonary metastatic disease were identified as cases, and each patient was matched with two control patients who did not have pulmonary metastatic disease. Case patients and control patients were matched for year of diagnosis, age at diagnosis, size of primary tumor, and nodal status.
Data analysis: Multivariate analysis using conditional logistic regression was used to determine the odds of smoking among women with pulmonary metastatic disease compared to matched control patients without pulmonary metastatic disease, after correction for potential confounding factors.
Results: Thirty-eight percent of the case patients vs 29% of the control patients were classified as ever-smokers; 24.1% of case patients were actively smoking at the time of breast cancer diagnosis vs 15.3% of the control patients. The unadjusted odds ratio for active smoking was 1.76 for women with pulmonary metastatic disease compared to women without pulmonary metastatic disease (p = 0.06). In the final multivariate model, the odds ratio for active smoking among women with pulmonary metastatic disease was 1.96 (p = 0.06).
Conclusions: There appears to be an association between cigarette smoking and the development of pulmonary metastatic disease among women with breast cancer. This may explain the previously noted higher breast cancer fatality rate among smokers. The relationship between smoking behavior and pulmonary metastasis from breast and other cancers warrants further investigation.
Key Words: breast cancer metastasis smoking
| Introduction |
|---|
|
|
|---|
Smokers have been found to have an increased rate of death from breast cancer in several epidemiologic studies,6 7 although they do not have an increased incidence of the disease.8 9 10 One potential explanation for this is that smoking, through its pulmonary or systemic effects, adversely affects the natural history of breast cancer. Since the lung is a common site of metastasis from breast cancer, and smoking is a cause of numerous changes in the lung that could affect the likelihood of metastatic spread to this organ, it is plausible that smoking might alter the course of breast cancer by increasing the frequency with which breast cancer metastasizes to the lung. Cigarette smoking and breast cancer are both sufficiently common that an effect of smoking on the natural history of breast cancer may have important implications for the health of women. We examined the relationship between cigarette smoking and pulmonary metastatic disease from unilateral, invasive breast cancer in a case-control study.
| Materials and Methods |
|---|
|
|
|---|
= 0.05; ß = 0.20),
assuming 0.2 as the proportion of smokers in the control group. The
study was approved by the University of California-Davis Human Subjects
Committee, and patient consent for medical records review was not
required.
Data Collection
The medical records of case patients and control patients were
reviewed, and the data were abstracted by a single trained abstractor.
Information on date of diagnosis, age at diagnosis, race, menopausal
status, tumor size and histology, tumor hormone receptor status, number
of positive lymph nodes, body mass, breast cancer treatment, and
smoking status was recorded on a standardized data collection
instrument.
Definition of Metastatic Disease
Patients were considered to have pulmonary metastatic disease if
their conditions met the following criteria: (1) pleural effusion with
cytologic evidence of malignancy; (2) pleural effusion, exudative in
character, without alternative explanation and attributed, in the
medical record, to metastatic disease; (3) multiple pulmonary nodules
on chest radiograph or CT scan, which were interpreted by the clinician
and radiologist to be metastatic in etiology, whether or not a biopsy
was performed; and (4) a radiographic pattern on plain film or CT scan
that was interpreted by the radiologist to be consistent with
lymphangitic carcinomatosis. Patients with single-mass lesions
of the lung were excluded because of the potential for
misidentification of a lung cancer lesion as a metastatic lesion. Other
sites of metastatic involvement were identified by a review of all
radiologic and pathologic reports in the chart and in the computerized
databases of the University of California-Davis Radiology and Pathology
Departments. A bone scan, head CT scan, MRI scan, or other scan that
was read as consistent with metastatic disease was considered to be
evidence of metastasis, even in the absence of pathologic confirmation.
Smoking Status
Information on smoking history was obtained from a review of all
potential sources contained in the medical record, including hospital
admission histories, clinic notes, and consultations. Patients were
classified as ever-smokers or never smokers. Ever-smokers were further
classified according to smoking status at the time of diagnosis as
active smokers or former smokers.
Statistical Analysis
Univariate analyses were performed using t tests for
continuous variables and
2 or Fishers Exact
Tests for the comparison of proportions. Age was calculated in years.
Body mass index was recorded in kilograms per square meter and was
regarded as a continuous variable. Tumor size was coded as a categoric
variable, with categories of < 2 cm, 2 to 5 cm, and > 5 cm.
Positive lymph nodes were considered to be a categoric variable, with
categories of 0, 1 to 3, 4 to 10, and > 10 positive nodes. Race was
coded as a categoric variable, with categories of white, black, and
other. Family history of breast cancer, menopausal status, radiation
therapy, chemotherapy, and hormonal therapy (ie, tamoxifen
therapy or oophorectomy) were coded as dichotomous variables. The
number of pack-years was coded as a categoric variable by dividing the
data into quartiles (nonsmokers, 1 to 19 pack-years, 20 to 39
pack-years, and
40 pack-years). Multivariate analysis was performed
using conditional logistic regression and a forward selection variable
selection algorithm (SAS statistical software, version 6; SAS
Institute; Cary, NC).
| Results |
|---|
|
|
|---|
|
|
|
| Discussion |
|---|
|
|
|---|
The susceptibility of an organ to invasion by malignant cells is a dynamic variable. Smoking causes a host of changes in the lung, including increased permeability12 13 and altered local immune function,14 15 16 17 18 and is associated with changes in the incidence and/or natural history of a broad variety of lung diseases.19 Smoking causes mild lung injury, and lung injury from a variety of causes, such as hyperoxia, radiation, and bleomycin exposure, has been shown to increase the likelihood of developing pulmonary metastatic disease in animal models of cancer.20 21 22 23 24 In our study, the association between pulmonary metastatic disease and cigarette smoking was strongest for active smoking. This is not unexpected given that many of the biological effects of smoking that might plausibly affect the susceptibility of the lung to pulmonary metastatic disease, such as changes in permeability or immune function, are transient and reversible.
A weakness of this study was our inability to control for other lifestyle factors that might confound the relationship between smoking and metastatic disease. For example, smokers are known to have less healthful diets and to be less physically active than their nonsmoking counterparts.25 It is possible that smoking is not itself the cause of a change in the natural history of breast cancer but, instead, is associated with an unidentified confounding factor. With the matching of patients and control subjects for tumor size and nodal status, our study design has accounted for any possible difference between smokers and nonsmokers in disease stage at diagnosis, a factor that has been suggested as a possible explanation for the higher rate of fatal breast cancer among smokers.6
In our study population, women who had pulmonary metastatic disease were significantly less likely to have been treated with hormonal therapy (ie, tamoxifen therapy or oophorectomy), which is of demonstrated efficacy in improving breast cancer survival,26 than were women without pulmonary metastatic disease. The association between the failure to receive hormonal therapy and the development of metastatic disease is not surprising. The reasons that some women did not receive such therapy are not clear. It is possible that such therapy was not offered to, or was refused by, women who smoked because of the increased risk of thrombosis associated with both smoking and the use of tamoxifen, but this is purely speculative. The relationship between pulmonary metastatic disease and smoking status persisted after we corrected for hormonal therapy in our multivariable model.
The possibility of ascertainment bias must be considered as a potential contributing factor to the results of this study. Smokers might be more likely to have chest radiographs performed, with an increased chance of detecting pulmonary metastatic disease. While this might lead to an earlier diagnosis of pulmonary metastatic disease, one would not expect it to significantly alter the overall rate of detection of that disease. Chest radiographs are performed frequently in the care of patients with cancer, and pulmonary metastatic disease is frequently symptomatic. One would not expect there to be a significant number of patients for whom there was a sustained failure to diagnose metastatic disease of the lung.
Our study, as well as the prior work of Scanlon et al,11 focused specifically on the association between smoking and pulmonary metastatic disease. It is possible, however, that an effect of smoking on metastatic propensity is not limited to the lung. Smoking has systemic effects that could affect tumor defense mechanisms external to the lung. For example, the number of circulating natural killer cells is reversibly decreased in active smokers.27 28 Smoking also affects platelet function and coagulability,29 30 factors that are believed to play a role in the bodys defense against tumor cells lodged within capillary beds. In addition, oxidant constituents of cigarette smoke affect signal transduction mechanisms involved in the metastatic process.31 Cigarette smoking is associated with an antiestrogenic effect, and smokers undergo menopause at an earlier age,32 which could affect the biological behavior or therapeutic response of hormone-sensitive tumors such as those in breast cancer. The effect of smoking on metastasis to organs other than the lung has not, to our knowledge, been examined. Our patients were significantly more likely than control subjects also to have metastatic disease at sites other than the lung, which is not surprising in view of the natural history of metastatic cancer. Our study design does not allow for conclusions about the temporal pattern of the metastases or about the assessment of the effect of smoking on other organ metastases.
The phenomenon of an adverse impact of cigarette smoking on the course and outcome of cancer may not be limited to breast cancer, in that smoking has been suggested as a predictor of a more lethal course for other cancers as well.33 34 35 36 The relationship between cigarette smoking and cancer outcome deserves further study. A prospective study, with the ability to control for other potentially confounding lifestyle factors, should be performed. If the relationship between smoking and an adverse effect on the natural history of this and/or other cancers is confirmed, we will need to explore whether the effect is remediable by smoking cessation at the time of cancer diagnosis.
| Conclusion |
|---|
|
|
|---|
| Footnotes |
|---|
Received for publication July 17, 2000. Accepted for publication January 5, 2001.
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
L. Xu and X. Deng Protein Kinase C{iota} Promotes Nicotine-induced Migration and Invasion of Cancer Cells via Phosphorylation of {micro}- and m-Calpains J. Biol. Chem., February 17, 2006; 281(7): 4457 - 4466. [Abstract] [Full Text] [PDF] |
||||
![]() |
R. T. Chlebowski Obesity and Early-Stage Breast Cancer J. Clin. Oncol., March 1, 2005; 23(7): 1345 - 1347. [Full Text] [PDF] |
||||
![]() |
L. Xu and X. Deng Tobacco-specific Nitrosamine 4-(Methylnitrosamino)-1-(3-pyridyl)-1-butanone Induces Phosphorylation of {micro}- and m-Calpain in Association with Increased Secretion, Cell Migration, and Invasion J. Biol. Chem., December 17, 2004; 279(51): 53683 - 53690. [Abstract] [Full Text] [PDF] |
||||
![]() |
S. Murin, K. E. Pinkerton, N. E. Hubbard, and K. Erickson The Effect of Cigarette Smoke Exposure on Pulmonary Metastatic Disease in a Murine Model of Metastatic Breast Cancer Chest, April 1, 2004; 125(4): 1467 - 1471. [Abstract] [Full Text] [PDF] |
||||
![]() |
H. W. Daniell and S. Murin Breast Cancer in Smokers Chest, May 1, 2003; 123(5): 1771 - 1772. [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |