|
|
||||||||
Guest Access | Sign In via User Name/Password |
|||||||||
* From the Department of Medicine (Drs. Khurana, Bejjanki, and Owens), Overton Brooks VA Medical Center; and Department of Biometry (Dr. Caldito), Louisiana State University Health Sciences Center, Shreveport, LA.
Correspondence to: Vikas Khurana, MD, Overton Brooks VAMC, East Stoner Ave, 151 (Research), Shreveport, LA 71101; e-mail: khuranav{at}gmail.com
Abstract
Background: Statins are commonly used cholesterol-lowering agents that are noted to suppress tumor cell growth in several in vitro and animal models.
Methods: We studied the association of lung cancer and the use of statins in patients enrolled in the Veterans Affairs (VA) Health Care System. A retrospective case-control study nested in a cohort study was conducted using prospectively collected data from the Veterans Integrated Service Networks 16 VA database from 1998 to 2004. We analyzed data on 483,733 patients from eight states located in south central United States. The primary variables of interest were lung cancer and the use of statins prior to the diagnosis of lung cancer. Multiple logistic regression analysis was done to adjust for covariates including age, sex, body mass index, smoking, diabetes, and race. Statistical software was used for statistical computing.
Results: Of the 483,733 patients in the study, 163,662 patients (33.8%) were receiving statins and 7,280 patients (1.5%) had a primary diagnosis of lung cancer. Statin use > 6 months was associated with a risk reduction of lung cancer of 55% (adjusted odds ratio, 0.45; 95% confidence interval, 0.42 to 0.48; p < 0.01). Furthermore, the protective effect of statin was seen across different age and racial groups and was irrespective of the presence of diabetes, smoking, or alcohol use.
Conclusions: Statins appear to be protective against the development of lung cancer, and further studies need to be done to define the clinical utility of statins as chemo protective agents.
Key Words: lung cancer pharmacology preventive medicine
Lung cancer is currently the most common cause of cancer mortality in the United States and throughout the world. The American Cancer Society estimates that lung cancer will be responsible for approximately 160,000 deaths in the United States during 2005, in comparison with 125,000 deaths from the combined mortality of colorectal, breast, and prostate cancers.1 Worldwide, almost 600,000 deaths were due to lung cancer in 1995, and it is projected that this number will continue to rise over the next decades.
Lung cancer is different from most other cancers because of the major modifiable risk factor of exposure to tobacco smoke. Exposure to asbestos, arsenic, halo ethers, nickel, and polycyclic aromatic hydrocarbons are other identifiable risk factors for lung cancer. Genetic factors, dietary factors, and the presences of underlying benign forms of parenchymal lung diseases such as pulmonary fibrosis are other potential risk factors. To date, there is no effective chemopreventive agent identified as a means to reduce the incidence of lung cancer.
3-Hydroxy-3-methylglutaryl coenzyme A reductase inhibitors (statins) are widely used drugs for the treatment of lipid disorders, especially hypercholesterolemia. Their role in reduction of mortality and morbidity in both primary23 and secondary prevention456 of coronary artery disease is well proven. They also reduce long-term cerebrovascular events, particularly after an initial coronary event.78910 Statins are currently one of the most widely prescribed drugs and are generally well tolerated and have a safe side effect profile, with the most concerning adverse effects being hepatotoxicty and myotoxicity.11 Several studies1213 have shown beneficial effects of statins on the incidence of cancers. We conducted this study to determine if statins have a protective effect against lung cancer in a US veteran population.
Materials and Methods
Data Source
The Veterans Health Administration (VHA) is organized into 21 administrative regions called Veterans Integrated Service Networks (VISNs). VISN 16, or the South Central Veterans Affairs (VA) Health Care Network, provides health-care treatment to veterans in an eight-state region that includes Florida, Alabama, Mississippi, Louisiana, Arkansas, Missouri, Oklahoma, and Texas. The network, an integrated health-care system, includes 10 medical centers, 33 community-based outpatient clinics, 7 nursing homes, and 2 domiciliaries. It is one of the largest of the 21 VISNs of the VHA. Geographically, the network spans approximately 170,000 square miles and includes > 1.9 million veterans. The Veterans Health Information Systems and Technology Architecture (Vista)14 is the integrated electronic medical record system for the VHA that contains, among other information, records of inpatient stays and outpatient visits, diagnostic codes, current procedural terminology codes, pharmacy records, and laboratory test results.15 All this information, except the content of narrative progress notes, can readily be tabulated by means of automated queries. Prior to 1995 the various facilities maintained their Vista databases separately. At that time, VISN 16 initiated a VISN-wide information system consolidation by standardizing the data from all the facilities in VISN 16. It contained the data since October 1, 1996. The VISN 16 data warehouse is an enriched administrative database and contains prospectively collected data of the patients treated in VISN 16. The computerized records contain both clinical and administrative factors including all patient encounters, pharmacy, laboratory, vital signs, financial, and patient demographic data. In addition, several health-care factors such as smoking and alcohol intake are also recorded. All the diagnoses are coded according to International Classification of Diseases (ninth version, clinical modification). There are 750 million rows of data arranged in relational tables that can be accessed by the Microsoft SQL server (Microsoft; Redmond, WA). The data from each facility are extracted monthly from the Vista, and tables are populated after standardization of the data from all the facilities in the VISN. Access to this data is monitored and controlled. The study protocol, including access and use of the computerized patient clinical information, was authorized by the VISN 16 multisite institutional review board. The informed consent requirement was waived for the study.
Study Design
A retrospective case-control study nested in a cohort was conducted using as subjects those included in the VISN 16 network with visits to VA Health Care System between October 1, 1998, and June 1, 2004. A Microsoft SQL server was used to identify these patients.
The VA administration defines a frequent VA Health Care System user as any person who was seen in a primary care clinic or was hospitalized in the VA Health Care System during the last 2 years. We used this definition to identify recent users of the system during the period between October 1, 1998, and June 1, 2004, for inclusion in our study. It excludes patients who utilize VA for prescriptions only. Thus, of 1.4 million patients in the VISN 16 network, a total of 483,733 recent users of the VA Health Care System were included in the study. The case population consisted of the patients with lung cancer, and patients without lung cancer made up the control population.
Multivariate analysis using a multiple unconditional logistic regression was used to determine significant effect of statin use on the risk for lung cancer adjusting for the effects of age, sex, race, body mass index (BMI), smoking, alcohol use, and diabetes. The data were limited to those from 18 to 100 years of age. BMI was defined as the first reported weight (in kilograms) for the patient divided by the height squared (in meters). VA administrative database-defined criteria for race was used. Smoking and alcohol use were reported as health factors by the clinical provider for those patients with smoking and alcohol use. Detailed smoking data were retrieved with > 40 different fields for the recoding of the tobacco-related health factors as entered by the primary care physicians. The fields were consolidated into two groups, one of which indicated they smoked and the other indicated they never smoked. Patients with missing smoking data were deleted from the multiple logistic regression analysis so the adjusted odds ratios (ORs) comparing odds for cancer between the statin and nonstatin users (Tables 123 ) refer only to those with smoking information. Diabetes was recorded as present or absent, and date of diagnosis of diabetes was not factored into the analysis. The duration of statin use was defined as the time of usage of statins prior to the diagnosis of lung cancer or the time of usage of the statin until the data collection completion date. Assumption was made that the patient had not received statins prior to receiving them in the VA. Patients receiving statins after the diagnosis of lung cancer were excluded from the statin use group. The dose and type of statin use were not factored in the analysis. Statistical software (SAS version 9.1.2; Cary, NC) was used for statistical computing. Multiple unconditional logistic regression analysis was used to determine significant association between statin use and lung cancer adjusted for the significant effects of covariates. The OR and its 95% confidence interval (CI) were used universally in the data analysis.
|
|
|
The demographics of the case and control population are as shown in Table 1. Men comprised 97.9% of those with lung cancer (n = 7,124). Of the 483,733 patients in the study population, 7,280 patients (1.5%) had lung cancer and 163,662 patients (33.8%) were receiving statins. Use of statins was seen in 1,994 of the 7,280 patients (27.4%) with lung cancer and 161,668 of 476,453 patients (33.9%) without lung cancer. Those who received statins after the diagnosis of lung cancer were included in the nonstatin group in the analysis to determine significant association of statin with lung cancer. In addition to statin use, other factors significantly associated with lung cancer were race, age, sex, BMI, diabetes, alcohol use, and smoking. There was no significant interaction between statin use and any of these other significant factors for lung cancer. The multiple logistic regression with just the main effects of statin use and these other significant covariates as predictors for lung cancer was a good fit to the data using the deviance and Pearson goodness of fit test. Statin use
6 months was associated with a significant risk reduction of lung cancer by 55%% (adjusted OR, 0.45; 95% CI, 0.42 to 0.48; p < 0.01). Smoking information was missing for 173,439 patients (35.8%). There were 3,623 smokers (72.6%) and 180,896 smokers (59.2%) smokers among lung cancer cases and control subjects, respectively; and the adjusted OR for smoking was significant at 2.13 (95% CI, 1.98 to 2.30).
Table 2 lists the effects of duration of statin therapy on lung cancer. The protective effect of statin against lung cancer increased with statin duration. There was a 77% reduction in odds for lung cancer among those who used statin for
4 years (adjusted OR, 0.23; 95% CI, 0.20 to 0.26; p value < 0.01).
Data for statin use
6 months were analyzed with respect to age, race, BMI, smoking, alcohol use, and diabetes to determine if the protective effect of statins was restricted to particular risk factor groups. Results are shown in Table 3. The protective effect of statin was seen across all age, race, BMI, smoking, alcohol, and diabetes groups. In this analysis, statin use > 6 months was associated with a risk reduction of 53% (adjusted OR, 0.47; 95% CI, 0.43 to 0.51; p < 0.01) even in subjects who smoked, which is the major risk factor for lung cancer.
Discussion
To date, there is no effective chemopreventive agent identified for lung cancer. Several epidemiologic surveys16 have shown that high levels of beta-carotene in the diet or in the blood are associated with lower risk of cancer in general and lung cancer in particular. Some studies1718 have suggested that low serum concentrations of certain antioxidants vitamins, especially derivatives of vitamins A and E, are associated with the development of lung cancer. Many studies192021 now suggest that an increased consumption of green and yellow vegetables, fruits, and some micronutrients is associated with a substantial lower risk of lung cancer, both among cigarette smokers and nonsmokers. While some studies2223 have suggested that beta-carotene may actually increase the risk of lung cancer, others showed a beneficial24 or no effect at all.25 Despite these data, evidence from randomized trials22232425 aimed at primary prevention with antioxidants has been conflicting.
Initially, there was concern that statins have intrinsic carcinogenic properties based on studies in animal models and from epidemiologic data in humans. Newman and Hulley26 reviewed animal studies with lipid-lowering therapy and concluded that statins and fibrates might cause cancers in rodents. Epidemiologic cohort studies2728 in humans also demonstrated that low cholesterol levels were associated with an increase in cancer death. However, confounding variables such as the effect of preexisting cancer and the retrospective nature of the study were major limitation of these studies.29 In contrast to these studies are the findings from a meta-analysis of randomized-controlled trials and case control studies, which revealed either no association,30313233 or even a decreased cancer incidence34353637 among patients using statins. Graaf et al35 conducted a case-control study with 3,129 cancer cases and 16,976 control subjects, which showed a 20% reduction in overall cancer risk in statin users. A nested case-control study36 found a 28% lower risk of all cancers in statins users compared with those receiving bile acid binders. Collectively, these studies seem to suggest that statins do not augment cancer incidence and may even provide protective effects. In vitro and cross-sectional studies1213383940414243444546 have shown anticancer activity of statins on several kinds of cancers including colon, prostate, lung, hepatocellular carcinoma, breast cancer, astrocytomas, glioblastoma multiform, mesothelioma, leukemia, myeloma, medulloblastoma, and renal cell carcinoma. The safety of statins as an anticancer agent has been investigated in various clinical trials.474849
Our data indicate that statin use
6 months in duration was associated with a significant risk reduction in the incidence of lung cancer across all age groups, race, and BMI. The study found an increased OR for lung cancer for the group in whom cancer developed within 6 months of the first recorded prescription of statins. There is no available data for the condition that brought the patients into the VA system. We do not believe that the statins actually caused an increase in lung cancer. We postulate that this represents a skewed group due to recording of the data at the time of entry of the patient into the database and may represent an old diagnosis or recording issue at the time of entry into the VA Health Care System. The data obtained after 6 months of statin use (Table 2) clearly shows a decreasing risk for lung cancer with increasing duration of statin use.
This protective effect was even seen in smokers. These data are also consistent with the results of small, nested case-control studies3536 from Netherlands and Canada that showed a significant reduction in the occurrence of all cancers among statin users as compared with persons who did not use statins. In vitro studies5051 have shown the beneficial effects of statins on human and animal lung tumor models; these effects were attributed in part to the antiproliferative, proapoptotic, and anti-invasive properties of statins. The present study is the first, large population-based study of the effect of statins on the incidence of lung cancer
While thought provoking, the results of this study should be evaluated with caution. This study is population based, in which the population consists of VA users during a certain period in eight US southern states. Our study population was predominantly male. The study was an observational retrospective case-control study nested in a cohort study; hence, we cannot rule out unknown biases or confounders. Although the observed significant protective effect of statins against lung cancer for our study population was adjusted for smoking, age, sex, BMI, alcohol use, and diabetes, we did not adjust for other possible risk factors of lung cancer such as exposure to asbestos, halo ethers, polycyclic aromatic hydrocarbons, nickel, arsenic, passive smoke, and radon. Other potential risk factors, including dietary factors, genetic factors, and the presence of underlying benign forms of parenchymal lung disease, were not included in our analysis. Our analysis did not include patients who obtained statin prescriptions from outside the VA system. However, most veterans would prefer to obtain prescription medication from VA system to avoid additional medical costs. Also, dose, duration, and type of statin use were not factored into our analysis. We used the administrative definition of a frequent VA system user that includes any person who was seen by a primary care clinic or was hospitalized in the VA system during the last 2 years. The definition would have excluded several patients; however, the exclusion was not selective for cases and control subjects. We did not use other possible patient selection definitions, which could be a limitation of our study. One of the strengths of our study is the use of a computerized database, with prospective data gathering from patients, allowing for inclusion of 483,733 patients. Use of a computerized pharmacy database minimizes the exposure misclassification and recall bias. Furthermore, established risk factors for lung cancer that were not studied do not play a role in prescription of statins. Nevertheless, the possibility that the findings of our retrospective study may have been confounded by the indication for statin use could be a limitation of our study.
Conclusion
Statin use > 6 months was associated with a 55% reduction in the odds for lung cancer. Due to the high prevalence of statin use and grave prognosis of lung cancer, even a modest risk reduction means a considerable effect on public health. Our study suggests that statins have a potential role in primary chemoprevention for lung cancer. Well-designed randomized prospective double-blinded placebo controlled clinical trials are necessary to validate the value of statins in lung cancer prevention and treatment.
Footnotes
Abbreviations: BMI = body mass index; CI = confidence interval; OR = odds ratio; VA = Veterans Affairs; VHA = Veterans Health Administration; VISN = Veterans Integrated Service Network; Vista = Veterans Health Information Systems and Technology Architecture
The authors have no conflicts of interest to disclose.
Received for publication April 6, 2006. Accepted for publication December 15, 2006.
References
-tocopherol and subsequent risk of lung cancer among male smokers. J Natl Cancer Inst 1999;91,1738-1743Related Editorial
This article has been cited by other articles:
![]() |
J. Gray, W. M. Alberts, and G. Bepler Statins and Lung Cancer Risk Chest, May 1, 2007; 131(5): 1274 - 1275. [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |