I. Introduction

Lung cancer is the leading cause of death from cancer in the USA, accounting for more deaths per year than cancers of the breast, colon, prostate, and cervix combined(Bach et al, 1999; Parkin et al, 1999). The American Cancer Society estimates there will be 169,500 new cases of lung cancer diagnosed in 2002 and 157,400 deaths (American Cancer Society, 2003). Lung cancer usually presents as an advanced, unresectable tumor and is usually fatal disease. It has been a challenge during the past decades to develop efficient screening tools and treatment for such an aggressive neoplasm. The overall survival at five years measured by the National Cancer Institute's Surveillance, Epidemiology, and End Results (SEER) program in the United States is 14%. For European countries, the average ~~five year~~five-year survival is 8%, the same as for developing countries (Parkin et al, 1999; Black, 2000). However, for resectable patients with pathologic stage I non-small-cell lung cancer, 5-year survival rates range from 65 to 80% (Flehinger et al, 1992; Inoue et al, 1998). Since stage I disease has the potential for curative treatment, it has been hoped that detection of early stage cancer in asymptomatic individuals can result in meaningful benefit (Mulshine et al, 1989).

II. What population is under high risk for the development of lung cancer?

The risks and benefits of screening methods make them applicable predominantly to high-risk populations, since costs and harms that they might generate can only be justified by a reduction in mortality for that specific cause screened. Factors that identify those at risk for lung cancer have been recognized through epidemiologic study over the past 30 years (Tockman et al, 1987).

There are two categories of evidence that indicate smoking to be the major cause of human lung cancer. Without exception, epidemiological studies have demonstrated a consistent association between smoking and lung cancer in both genders. Also, chemical analyses of cigarette smoke reveal a multitude of known mutagens and carcinogens that are absorbed and metabolized, and cause demonstrable genetic changes in smokers (Loeb et al, 1984; McLemore et al, 1990). Across studies, risk increases with: number of cigarettes smoked (the relative risk is 3.0 for 1 pack-per-day smokers), years smoked, earlier age at onset of cigarette smoking, degree of inhalation, tar and nicotine content of cigarettes smoked and use of nonfiltered compared with filtered cigarettes (Loeb et al, 1984). Lung cancer rates decrease when smoking is stopped and approaches those of people who have never smoked at 10 years after cessation (Garfinckel and Silverberg, 1991).A study from the United Kingdom evaluating the relationship between smoking, smoking cessation and lung cancer points to the fact that even people who stop smoking at 50 or 60 years of age avoid most of their subsequent risk of developing lung cancer, and those who stop at 30 years of age avoid more than 90% of the risk attributable to tobacco of those who continue to smoke (Peto et al, 2000).

Advanced age is also considered a risk factor for lung cancer development. Tockman and colleagues (1987) reported a 2.8 relative risk for lung cancer for an age of greater than 59 years. Most screening programs to date have incorporated a minimum age cut-off of 45 years for recruitment of subjects to be screened.

Among cigarette smokers, the presence of airways obstruction is a greater risk for the subsequent development of lung cancer than is age alone or the level of smoking. A measure of the forced expiratory volume in the first second (FEV₁) of less than 60% of the predicted value is associated with a 6.4 times greater risk for lung cancer compared with the risk associated with the absence of ventilatory impairment (Tockman et al, 1987). One possible explanation could be that chronic obstructive pulmonary disease may interfere with the ability of the airways to eliminate inhaled carcinogens effectively (Cohen, 1980). In the presence of damaged bronchial mucosa and prolonged exposure to carcinogens, a carcinogenic effect would be enhanced.

Occupational exposure is estimated to account for approximately 5% of lung cancer in the United States (Beckett, 2000) and the majority of these cancers are caused by asbestos, followed by radon, silica, chromium, cadmium, nickel, arsenic and beryllium. Among these, asbestos has received the most study (Weiss, 1999).Cigarette smoke and asbestos interact strongly in causing bronchogenic carcinomas and the evidence indicates that the presence of asbestosis is a much better predictor of excess lung cancer risk than measures of exposure to asbestos.

Patients with resected stage I NSCLC (T1N0M0 and T2N0M0) have an annual incidence of a second primary lung cancer of 3-5% and this is an order of magnitude greater (10-fold greater) than the incidence experienced by the middle-aged, heavy smokers in the NCI Early Lung Cancer trial almost two decades ago (Mountain, 1997). Second primary lung cancer (SPLC) has been defined as a lung cancer, either of a different histologic cell type which appears within 2 years following resection of the index cancer, or as a tumor of the same cell type, that has characteristics of a primary lung cancer and arises in a different lobe if it appears after 2 years following resection (Tockman et al, 1988). Since cured NSCLC patients exhibit such high rates of second lung cancer they might constitute an appropriate population for early detection and chemoprevention trials.

Similarly, patients successfully treated for head and neck cancers and small cell lung cancer (SCLC) are at high risk for second primary lung cancer. Association of a squamous cell carcinoma of the head and neck and a pulmonary cancer is frequent: prevalence is about 5 to 10% (De Mones et al, 1999). Also, cured SCLC patients have one of the highest rates of second primary tumor (SPT) development, being SPTs the most common cause of cancer death 4 or more years after definitive primary therapy. SPTs occur most commonly in tobacco-exposed sites (i.e., lungs, esophagus and head and neck) (Lippman et al, 1994).

Currently no available diagnostic procedure has yet been proven to be of significant value for general use in the early detection of lung cancer in these high-risk populations.

III. Early Lung Cancer Detection - early trials

Much research has been conducted to evaluate the effectiveness of various combinations of screening tests on lung cancer mortality. In the 1950s and 1960s, several uncontrolled and nonrandomized controlled studies evaluated various combinations of roentgenograms and cytology given at various time intervals, ranging up to once every 6 months. These studies failed to show a benefit from lung cancer screening(Eddy, 1989). The lack of success of those previous sputum cytologic screening programs has been attributed in part to the technical inability to localize the source of expectorated cancer cells in patients with negative chest roentgenograms (Berlin et al, 1984).With the development of the fiberoptic bronchoscope, localization of occult endobronchial neoplasms became possible and in the early 1970s it became feasible to design studies to determine whether a reduction in lung cancer mortality might be achieved by screening cytology and chest roentgenography(Berlin et al, 1984; Carbone et al, 1970).

From 1971 to 1982, three National Cancer Institute sponsored studies on Screening for Early Lung Cancer were conducted at the Mayo Clinic, Johns Hopkins and Memorial Sloan-Kettering Cancer Center. Those studies enrolled a total of 31,360 men 45 years of age and older who smoked at least 1 pack of cigarettes per day or had smoked this amount within the year prior to enrollment.

In the Mayo Clinic trial,all participants received chest roentgenograms and sputum cytologic examinations prior to random allocation into screened and control groups (Figure 1) (Fontana et al, 1984; Fontana et al, 1986; Sanderson, 1986). Those free of cancer were randomized to a study group (chest roentgenograms and sputum cytologic tests every 4 months), or a control group (recommendations for annual chest roentgenograms and sputum cytology, without efforts to assure compliance). The study design unfortunately failed to establish that the control group was truly not screened. More than half of the control population underwent periodic screening, including recommendations described above.

The Johns Hopkins and Memorial Sloan-Kettering studies randomly allocated their volunteers into “dual screen” and “X-ray only screen” groups (Figure 2) (Melamed et al, 1984; Tockman, 1986). The dual-screen group received an annual chest roentgenogram examination plus annual sputum cytologic testing followed by a 3-day collection of sputum for cytologic examination every 4 months. The “X-ray only” group received only annual chest roentgenograms. Results of the initial screen (prevalence) of the three trials showed that 2,815 patients were found to have indeterminate abnormalities on the screening roentgenograms or x-rays considered “suspicious for cancer”, indicating the need for further evaluation. From those patients, 120 (4.3%) were found to have lung cancer. Only 79 (0.4%) of the 21,127 men who received the dual screen had sputum examinations reported as showing either marked atypia or carcinoma cells. From those patients, 55 had lung cancer and 12 had cancer of the upper respiratory tract (predictive value of 85%). Of the 160 cases detected in the dual- screen group, 67 (41%) would have been detected by cytologic examination alone and 123 (77%) by X-ray alone. Half of the lung cancers detected on dual screening (81 patients) were stage I. Those that were stage I and were resected had a 5-yr survival that was almost 80%. Of the 81 stage I lung cancers, approximately 60% were visible roentgenographically and nearly 40% were roentgenographically occult and detected by cytology alone. The early stage cancers detected by cytology alone were central squamous cell carcinomas.

The most frequently encountered early-stage cancers detected by roentgenograms alone were peripheral adenocarcinomas. Only 9% of stage I cancers were detected by both techniques. It is clear that a higher proportion of early stage cancers were detected and that patients staged as I or II had a better 5-year survival than those in stage III. However, the three trials failed to demonstrate improved resectability or survival rates among the study groups compared to the controls. These trials did not produce lowered overall lung cancer mortality, which is considered the ultimate indicator of the effectiveness of a screening test.

The Mayo Lung Project reported that the 5-year survival for patients who developed lung cancer was better in the screened group (40% versus 15%). However, with extended follow-up the overall lung cancer mortality for the entire population screened was not shown to be improved (Marcus et al, 2000). An analysis (Strauss, 2002) performed to determine which endpoint of that study (survival versus mortality) provides an unbiased measure of effectiveness shows that among resected patients, survival at 7 years was 50% (96% CI, 39% to 61%). The survival plateau demonstrated in the Kaplan-Meier survival curves argues in favor that overdiagnosis is not a confounder factor. In contrast, among those not undergoing resection, survival at 5 years was only 2% (95% CI, 1% to 8%). The author concluded that survival was superior in the screened population, and that this advantage was not attributable to lead-time bias, length bias, or overdiagnosis bias. Mortality though was biased, because incidence differences (30% higher incidence in the screened group) confounded the ability of mortality to reflect the true effect of screening.

IV. Possible Causes of Screening Biases

In the evaluation of a screening program, several potential sources of bias must be considered: lead-time bias, length bias and overdiagnosis (MacLean, 1996).

The first of these is lead-time bias: patients who are diagnosed at an early stage in their illness may appear to survive longer than those diagnosed at a later stage only because the diagnosis is established for a longer period of time.

Length bias is related to tendency of slow-growing tumors to be discovered during screening while fast-growing tumors are more likely to become clinically evident between screening intervals. Length bias is expected to improve survival rates for those patients who had a better prognosis from the start. However, if any of these biases accounted for the favorable survival in the screened groups of the Mayo, Sloan-Kettering and Hopkins studies reported above, one would expect that the same stage I patients would have comparable survival even if they remained untreated. Fortunately, this data is available. Although the majority of patients in those studies diagnosed with a stage I non-small-cell lung cancer were treated by surgical resection, 5 to 21% of patients with stage I NSCLC failed to be treated surgically either because patients refused surgery or because there were medical contraindications to surgery. Flehinger (1992) described five-year survivals (with lung cancer death as an endpoint) for those who were operated on ranged from 52 to 62%, while for those who did not undergo surgery survival ranged from 0 to 8%.

The third type of potential bias,and perhaps the only screening bias that can account for improvements in stage distribution, resectability, survival, higher incidence, and equal mortality as end-point in a screened compared to a control population is overdiagnosis (Strauss et al, 1993; Strauss, 1997; Strauss et al, 1997). This term refers to the detection, by screening, of lesions that are not clinically significant and would not adversely affect the lifespan of a patient. It means that some of the slow growing cancers never would have been diagnosed in the absence of screening and those individuals would have died of another disease without their subclinical cancer being recognized (Parkin and Moss, 2000). This concept is especially applicable for slow-growing neoplasms such as prostate cancer. It is well known that the clinical incidence of this disease does not match the prevalence noted at autopsy, where more than 40% of men over 50 years of age are found to have carcinoma of the prostate, while 9.5% of 50 year-old men will actually develop clinically apparent disease over their lifetime (Fried et al, 1997). In a series of over 20,000 autopsies carried out over a 25-year period, 700 cancers (11%) were found in whom the diagnosis of cancer had not been considered relevant clinically (Karwinski et al, 1990). In over half of them the unrecognized tumor was considered an incidental finding, being kidney and prostate the main organs involved in those cases. In contrast, the unrecognized cancers that caused death were almost often from pancreas or lung. These patients tended to be older than those with clinically recognized disease. In conclusion, it seems that because of its biological behavior, lung cancer is unlikely to be an overdiagnosed neoplasm in screening programs.

V. Innovative Screening Technologies

The specimens obtained during the NCI-sponsored trials provided a potential means for identifying other markers of early lung cancer (Mulshine, 1999). To develop a more sensitive detection technique, Tockman and co-workers used immunostaining of sputum cells with lung cancer-associated monoclonal antibodies (MoAbs) to determine if early preneoplastic cytologic changes of lung cancer could be more reliably detected (Mulshine et al, 1991). Using the two MoAbs 703D4 and 624H12, developed against non-small cell and small-cell lung cancer cells, specimens from the Johns Hopkins study were reexamined. They found that in subjects with moderate atypical metaplasia, these antibodies could predict the later development of lung cancer at least 2 years prior to clinical recognition, with a sensitivity of 91% and a specificity of 88% (Tockman et al, 1988). Of the 22 known positive cases of lung cancer, 20 were detected by dual antibody analysis. Nineteen of these cases were recognized by the 703D4 MoAb alone, whereas 624H12 was less sensitive, detecting eight cases of lung cancer alone. Because of the higher sensitivity of the 703D4 antibody, much of the biochemical research has been focused on this marker 12 that recognizes heterogeneous nuclear ribonucleoprotein (hnRNP) A2/B1 involved in cellular proliferation and the carcinogenesis cascade.

A prospective validation trial of sputum immunocytochemical early detection of the lung cancer is under way (Tockman et al, 1994; Mulshine et al, 2000) In this study, eleven centers collaborated in the accrual of 1,000 patients with stage I NSCLC who had undergone complete resection. Any patient currently in follow-up 6 months or more after surgical resection, chemotherapy or radiotherapy performed an annual sputum induction. The MoAbs examined in this study are 624H12 and 703D4, the same ones shown to be of value in retrospective studies. Preliminary results from this study indicate that when the predictive value of MoAb markers is compared to routine morphologic study, the positive predictive value of hnRNP MoAb is 67% (Mulshine, 1999).

Another promising technology for screening is the detection of growth factors associated with chronic lung injury and transformation of this epithelium from normal, to dysplastic, to frankly invasive cancer. Gastrin-releasing peptide (GRP), Insulin-like Growth Factor I (IGF-I) and Transferrin (TF) are examples of growth factors associated with carcinogenesis. The presence of elevated levels of these growth factors in the bronchial lavage of subjects at high risk for lung cancer may provide complementary information to the sputum immunostaining regarding the early detection of lung cancer (Mulshine et al, 1991). Also, the knowledge of the existence of such molecules has lead to development of monoclonal antibodies, which might be of value as therapy in high-risk individuals.

More recently, new early lung cancer screening programs have been designed using low-dose spiral computed tomography (CT) to improve the likelihood of detection of small non-calcified nodules, and thus of lung cancer at an earlier and potentially more curable stage. Once exposure to radiological examinations is associated with the risk of induction of malignancy, this has to be balanced against the benefits of early detection of a malignant tumor. Effective dose equivalent ranges from 0.06 to 0.25 millisieverts (mSv) with chest radiography in 2 views, 3-27 mSv with CT using conventional examination parameters, and 0.3-0.55 mSv using low dose CT settings. Based on considerations by the International Commission on Radiological Protection, it can be expected that radiation exposure with an effective dose equivalent of 1 mSv would lead to 5 additional malignancies in 100,000 individuals exposed (International Commission on Radiological Protection, 1991).

The results of a Japanese study conducted by Kaneko et al (1996) brought to focus the idea of screening high-risk populations for lung cancer using a new screening method: spiral CT. A total of 1,369 members from the for-profit organization called Anti-Lung Cancer Association (ALCA) – mostly men, average age 60 years and smokers (> 20 pack-year) were submitted to dual screening using spiral CT and chest radiography. Chest radiography was able to detect only 4 of the 15 cases of lung cancer that were diagnosed with low-dose spiral CT. Although the yield was low, these results confirmed the superiority of that new technique in the detection of early stage peripheral lung cancers.

Sone and colleagues (1998) developed a mass-screening program also in Japan to evaluate the usefulness of annual screening for lung cancer by low-dose CT and the characteristics of identified lung cancers. The prevalence data obtained in 1996 showed that among 5,483 individuals aged between 40 and 74 years, including smokers and non-smokers, suspicious nodules were detected in 279 (5.1%) subjects and 22 (8% of the 279) were confirmed surgically to have lung cancer. Corresponding figures in 1997 were 173 suspicious nodules of 4425 individuals, being 25 of them (14%) confirmed to be lung cancers. In 1998, 136 of 3878 (3.5%) individuals were diagnosed with suspicious nodules and of those only 9 (7%) of them were confirmed malignant. The mean size of lesions was 17mm and CT identified almost ten times as many cancers (0.48%) than a standard mass screening done 1 year before in the same area using roentgenograms and sputum cytology (0.03-0.05%). The sensitivity and specificity of detecting surgically confirmed lung cancers were 55% and 95%, respectively, in 1996. The sensitivity and specificity were 83% and 97% in the 1997 screening. Of the 60 cases of lung cancers identified in the screening 88% were surgically confirmed as stage IA. Such a small number of neoplasms detected in this study can be attributed to the fact that the screening was done in a rural area of Japan where the rate of lung cancer has been reported to be low. That constituted a low-risk population, once non-smokers were included in the study.

The Early Lung Cancer Action Project (ELCAP)was designed to evaluate baseline and annual repeat screening by low-dose CT in people at high risk of lung cancer and is under way at Cornell University-New York Presbyterian Hospital. The project’s overall design and findings from baseline screening were recently published (Henschke et al, 1999). ELCAP has enrolled 1,000 symptom-free volunteers, aged 60 years or older, with at least 10 pack-years of cigarette smoking, who were medically fit to undergo thoracic surgery. Non-calcified nodules were detected three times more commonly than by chest radiography, malignant tumors four times more commonly, and stage I tumors six times more commonly. Of the 27 CT-detected lung cancers, 26 (96%) were resectable, compared to only 30 (51%) of the 59 tumors detected on baseline chest radiography done under Mayo Lung Project. Critics to this study include mainly lack of randomization of subjects to a non-screened arm and a number of enrolled individuals that is not sufficient to provide statistical power to evaluate impact in survival and mortality in long-term follow-ups.

Another prospective cohort study conducted at Mayo Clinic enrolled 1,520 individuals aged 50 years or older who had smoked at least 20 pack-years to undergo a baseline screening with spiral computed tomography and sputum cytology followed by a three-year annual screening (Swensen et al, 2002). One or more lung nodules were identified in 1,000 (66%) of the 1,520 participants, although primary lung cancers were documented in only 25 of the 1,464 individuals (1.7%) who returned for the first of their annual incidence scans and sputum examinations. CT alone detected 23 cases and sputum cytology alone detected 2 cases. Twelve (57%) of the 21 non-small cell cancers detected by computed tomography were stage IA, although a stage shift can not be confirmed. Of significant concern was the extremely high rate of false-positive results lesions identified by the spiral CT.

The study conducted by Diederich et al (2002) at the University of Münster, Germany, screened with low-dose spiral CT a total of 817 asymptomatic smokers (minimum age of 40 years and minimum tobacco consumption of 20 pack-years). The authors were able to demonstrate a prevalence of lung cancer of 1.3% in this high-risk population. Only seven out the twelve lesions (58%) diagnosed as being malignant were actually stage I.

Nawa et al (2002) developedanother screening program in Japan, targeting health care professionals. From April 1998 to August 2000, spiral CT screening was performed as part of annual health examinations on a total of 7,956 individuals. The majority of patients were men, with ages ranging from 50 to 59 years. 62.1% of the subjects were current or former smokers. After baseline screening, a total of 36 cases of primary lung cancer were histologicaly confirmed (0.44% prevalence). 28 patients (77.7%) were classified as stage IA. The most common histology was adenocarcinoma (35 of 37) and the mean diameter was 17 mm.

The retrospective study by Patz et al (2000) evaluated 510 patients with surgically resected, pathologic stage IA NSCLC. The purpose of this study was to determine the relationship between size and survival in patients with stage IA NSCLC (lesions < 3 cm). No correlation between decrease in tumor dimension and improvement in survival was found, and the authors used this result to caution against the use of low-dose spiral CT as an effective tool for early diagnosis of lung cancer. One explanation why the expected relationship between tumor size and survival was not observed was the fact that the authors reported survival based on deaths from all causes, rather than deaths related only to lung cancer (Black, 2000). Nearly half of deaths in those patients were actually due to other causes and the inclusion of these deaths may have reduced the power of this study to detect differences in lung cancer survival. The size of the studied population was also a concern regarding the statistical power of that study when it comes to evaluation of cancer survival.

In summary, several studies have evaluated the utility of CT screening in the detection of early lung cancer. Depending largely upon the population studied, it appears that the detection rates range from 0.4% to 1.7%. The tumors detected tend to be small, peripheral, and are resectable in 50-70% of cases. None of these studies included a prospective control group, therefore the true effect of CT screening on cancer related mortality is unknown.

Positron emission tomography (PET) has gained attention as a new modality that appears useful in differentiating benign from malignant lesions when investigating a solitary pulmonary nodule detected previously by chest radiographs or CT scans. The technique is based on the uptake of a radioactive glucose, fluorodeoxyglucose, in metabolically active cells. In a recent study, Pitman and colleagues (2001)evaluated the accuracy of blinded reading of PET scans in 40 of 60 consecutive patients referred for evaluation of an indeterminate lung nodule or mass. The results showed that PET yielded 23 true positives, 13 true negatives, 3 false positives (2 tuberculosis, 1 sarcoidosis) and 1 false negative (an adenocarcinoma), giving a sensitivity of 96%, a specificity of 81%, a negative predictive value of 93% and a positive predictive value of 88% for malignancy.

When it comes to the discussion whether it is cost effective to screen individuals using expensive techniques, Mahadevia and colleagues (2003) evaluated the potential clinical and economic implications of an annual lung cancer-screening program based on helical CT. Using a computer-simulated model they compared annual helical CT screening to no screening for hypothetical cohorts of 100,000 current, quitting, and former heavy smokers. Over a 20-year period, assuming a 50% stage shift, the current heavy smoker cohort had 13% lung cancer-specific mortality reduction, with cost-effectiveness measured at $ 116,300 per quality-adjusted life-year (QALY) gained. The QALY is a measure of the quantity of life gained from a treatment, weighted by the quality of that life. It is generally agreed that any intervention costing more than $100,000 per QALY is not considered to be cost-effective (Earle et al, 2000). Therefore, spiral CT was not considered a cost-effective tool for lung cancer screening by the results of that study.

Several additional studies are currently under way. The National Cancer Institute and the American College of Radiology Imaging Network is enrolling 25,000 persons at high risk for lung cancer for a multicenter, randomized-controlled trial(National Cancer Institute and American College of Radiology Imaging Network, 2003). The study has the objective to compare lung cancer-specific mortality in high-risk subjects who undergo low-dose spiral CT scan of the chest versus chest radiography. In addition, the National Lung Screening Trial (NLST) – sponsored by the National Cancer Institute- has been enrolling participants for what will become the largest randomized controlled screening trial for lung cancer (Figure 3) (National Institutes of Health, National Cancer Institute, 2002). A total of 50,000 participants (25,000 per arm) will be accrued for this study within 2 years. The study aims to compare spiral computed tomography and chest radiography as screening tools for lung cancer. The study as it is designed has sufficient power to determine if there is 20% or greater reduction in lung cancer mortality by either technique.

VI. Conclusions

Screening for lung cancer has been the target of extensive research and controversy over the past decades. A high risk population can be identified, including smokers, those submitted to occupational exposures and survivors of prior lung cancer and head and neck cancer. Previous screening trials showed limited value on early detection and mortality. Those studies, however, did not employ modern technology and some suffered from flaws in their design.

An enthusiastic era of prospective trials using promising technologies such as low-dose spiral CT and sputum immunocytochemistry is facing again the challenge of proving that early detection of lung cancer can actually save lives. Until positive results of such studies are available assuring the practice of mass screening for lung cancer in high-risk populations, the decision to screen asymptomatic patients remains an individual decision. Physicians must consider the current value of available screening techniques, their patients’ level of risk and whether they would be suitable for surgical intervention before making individualized decisions for screening. Until prospective studies of lung cancer screening show a significant improvement in mortality, general population-based screening cannot be routinely recommended.

Until the results of these ongoing prospective studies are known, we cannot recommend population screening. However, we can identify individuals at extremely high risk for lung cancer. These individuals would include those with smoking history of greater than 20 pack years who would be surgical candidates plus (1) a history of resected stage I lung cancer, (2) early stage head and neck cancer that appears cured, and/or (3) abnormal pulmonary function tests (FEV1 < 60% of predicted).

However, there is strong evidence that the most effective approach to lung cancer is primary prevention– cessation of cigarette smoking (Wolpaw et al, 1996). Attention to this area needs to increase despite their difficulties and frustration. Although physicians agree that it is appropriate for doctors to counsel patients to stop smoking, they are inconsistent in providing such advice. The most commonly reported barriers to counseling are the belief that most patients are uninterested, perceived lack of skills and time constrains (Frank et al, 1991; Jaen et al, 1994). Studies have found though, that brief, directive smoking interventions delivered during routine care are cost-effective and have the potential for significant public health benefit. The assessment of smoking status and smoking cessation interventions should be definitely integrated into standard office practice (Robinson et al, 1995).

References

American Cancer Society: Cancer Facts and Figures 2003. Atlanta, GA: American Cancer Society. Last accessed March 24, 2003.

Bach PB, Cramer LD, Warren JL, Begg CB (1999) Racial differences in the treatment of early-stage lung cancer. N Engl J Med 34, 1198-1205.

Beckett WS (2000) Occupational respiratory diseases. N Engl J Med 342, 406-413.

Berlin NI, Buncher R, Fontana RS et al (1984) The National Cancer Institute Cooperative Early Lung Cancer Detection Program – Results of the initial screen (Prevalence). Am Rev Respir Dis 130, 545-549.

Black, WC (2000) Unexpected Observations on Tumor Size and Survival in Stage IA Non-small Cell Lung Cancer. Chest 117, 1532-1534.

Carbone PP, Frost JK, Feinstein AR, et al (1970) Lung cancer: early detection, diagnosis and localization. In: Lung cancer: perspectives and prospects. Ann Intern Med 73, 1003-1024.

Cohen BH (1980) Chronic obstructive pulmonary disease: a challenge in genetic Epidemiology. Am J epidemiol 112, 247-288.

De Mones E, Hans S, Le Pimpec-Barthes F, Riquet M, Laccourreye O, Brasnu D (1999) Squamous cell carcinoma of the head and neck and pulmonary cancer. Report of 50 cases. Ann Otolaryngol Chir Cervicofac 116, 341-350.

Diederich S, Wormanns D, Thomas M, Lenzen H, Roos N, Heindel W (2002) Screening for early lung cancer with low-dose spiral CT: prevalence in 817 asymptomatic smokers. Radiology 222, 773-781.

Earle CC, Chapman RH, Baker CS, Bell CM, Stone PW, Sandberg EA, Neumann PJ (2000) Systematic Overview of Cost-Utility Assessments in Oncology. J Clin Oncol 18, 3302-3317.

Eddy DM (1989) Screening for lung cancer. Ann Intern Med 111, 232-237.

Flehinger BJ, Kimmel M, Melamed MR (1992) The effect of surgical treatment on survival from early lung cancer. Chest 101, 1013-1018.

Fontana RS, Sanderson DR, Taylor WF, et al (1984) Early Lung cancer detection: results of the initial (prevalence) radiologic and cytologic screening in the Mayo Clinic study. Am Rev Respir Dis 130, 561-565.

Fontana RS, Sanderson DR, Woolner LB, et al (1986) Lung cancer screening: the Mayo program. J Occup Med 28, 746-750.

Frank E, Winkleby MA, AltmanDG, et al (1991) Predictors of physicians’ smoking cessation advice. JAMA 266, 3139-3144.

Fried RM et al (1997) Prostate cancer screening and management. Med Clin North Am 81, 801.

Garfinckel L, Silverberg E (1991) Lung cancer and smoking trends in the United Sates over the past 25 years. CA 41, 137.

Henschke CI, McCauley DI, Yankelevitz DF, et al (1999) Early Lung Cancer Action Project: overall design and findings from baseline screening. The Lancet 354, 99-105.

Inoue K, Sato M, Fujimura S, Sakurada A et al (1998) Prognostic assessment of 1310 patients with non-small-cell lung cancer who underwent complete resection from 1980 to 1993. J Thorac Cardiovasc Surg 116, 407-411.

International Commission on Radiological Protection (1991) 1990 Recommendations, Publication 60. Oxford, England: Pergamon Press.

Jaen CR, Stange KC, Nutting PA (1994) Competing demands of primary care: a model for the delivery of clinical preventive services. J Fam Pract 38, 166-171.

Kaneko M, Eguchi K, Ohmatsu H, Kakinuma R, Naruke T, Suemasu K, Moriyama N (1996) Peripheral lung cancer: screening and detection with low-dose spiral CT versus radiography. Radiology 201, 778-802.

Karwinski B, Svendsen E, Hartveit F (1990) Clinically undiagnosed malignant tumors found at autopsy. APMIS 98, 496-500.

Lippman SM, Benner SE, Hong WK (1994) Cancer Chemoprevention. J Clin Oncol 12, 851-873.

Loeb LA, Ernster VL, Warner KE et al (1984) Smoking and lung cancer: an overview. Cancer Res 44, 5940-5958.

MacLean CD (1996) Principles of Cancer Screening. Med Clin North Am 80:1-14.

Mahadevia PJ, Fleisher LA, Frick KD, Eng J, Goodman SN, Powe NR (2003) Lung Cancer Screening with Helical Computed Tomography in Older Adult Smokers: A Decision and Cost-effectiveness Analysis. JAMA 289, 313-322.

Marcus PM, Bergstralh EJ, Fagerstrom RM (2000) Lung Cancer Mortality in the Mayo Lung Project: Impact of Extended Follow-up. Journal of the NCI 92: 1308-1316.

McLemore TL, Adelberg S, Liu MC et al (1990) Expression of CYP1A1 gene in patients with lung cancer: evidence for cigarette smoke-induced gene expression in normal lung tissue and for altered gene regulation in primary pulmonary carcinomas. J Natl Cancer Inst 82, 1333-1339.

Melamed MR, Flehinger BJ, Zaman MB, et al (1984) Screening for early lung cancer – Results of the Memorial Sloan-Kettering Study in New York. Chest 86, 44-53.

Mountain CF (1997) Revisions in the international system for staging lung cancer. Chest 111, 1710-1717.

Mulshine JL (1999) Reducing Lung Cancer Risk – Early detection. Chest 116, 493S-496S.

Mulshine JL, De Luca LM, Dedrick RL, et al (2000) Considerations in Developing Succseeful, Population-Based Molecular Screening and Prevention of Lung Cancer. Cancer 89, 2465S-2467S.

Mulshine JL, Tockman MS, Smart CR (1989) Considerations in the development of lung cancer screening tools. J Natl Cancer Inst 81, 900-905.

Mulshine JL, Treston AM, Scott FM, et al (1991) Lung cancer: rational strategies for early detection and intervention. Oncology 5, 25-32.

clinicaltrials. Date last modified: December 23, 2003.

National Institutes of Health, National Cancer Institute. Fifty thousand current and former smokers needed for National Lung Screening Trial. September 18, 2002. Available at http://cancer.gov/clinicaltrials. Accessibility verified March 24, 2003.

Nawa T, Nakagawa T, Kusano S, Kawasaki Y, Sugawara Y, Nakata H (2002) Lung cancer Screening Using Low-Dose Spiral CT: Results of Baseline and 1-Year Follow-up Studies. Chest 122, 15-20.

Parkin DM, Pisani P, Ferlay J (1999) Global cancer statistics. CA Cancer J Clin 49, 33-64.

Parkin MD, Moss SM (2000) Lung Cancer Screening: Improved Survival But No Reduction in Deaths – The Role of “Overdiagnosis”. Cancer 89, 2369S-2369S.

Patz EF Jr, Rossi S, Harpole DH Jr, Herndon JE, Goodman PC (2000) Correlation of tumor size and survival in patients with Stage IA non-small cell lung cancer. Chest 117, 1568-1571.

Peto R, Darby S, Deo H, Silcocks P, et al (2000) Smoking, smoking cessation, and lung cancer in the UK since 1950: combination of national statistics with two case-control studies. BMJ 321, 323-329.

Pitman AG, Hicks RJ, Kalff V, Binns DS, Ware RE, McKenzie AF, Ball DL, MacManus MP (2001) Positron emission tomography in pulmonary masses where tissue diagnosis is unhelpful or not possible. Medical Journal of Australia 175, 303-7.

Robinson MD, Laurent SL, Little JM Jr (1995) Including smoking status as a new vital sign: it works. J Fam Pract 40, 556-563.

Sanderson DR (1986) Lung Cancer Screening – The Mayo Study. In: IV World Conference on Lung Cancer. Chest 84, 324S.

Sone S, Takashima S, Li F, et al (1998) Mass screening for lung cancer with mobile spiral computed tomography scanner. The Lancet 351, 1242-1245.

Strauss GM (1997) Measuring Effectiveness of Lung Cancer Screening: From Consensus to Controversy and Back. Chest 112, 216S-228S.

Strauss GM (2002) The Mayo Lung Cohort: a Regression Analysis Focusing on Lung Cancer Incidence and Mortality. J Clin Oncol 20, 1973-1983.

Strauss GM, Gleason RE, Sugarbaker DJ (1993) Screening for Lung Cancer Re-examined: a Reinterpretation of the Mayo Lung Project Randomized Trial on Lung Cancer Screening. Chest 103, 337S-341S.

Strauss GM, Gleason RE, Sugarbaker DJ (1997) Screening for Lung Cancer: Another Look; A Different View. Chest 111, 754-768.

Swensen SJ, Jett JR, Sloan JA, Midthun DE, Hartman TE, Sykes AM, Aughenbaugh GL, Zink FE, Hillman SL, Noetzel GR, Marks RS, Clayton AC, Pairolero PC (2002) Screening for lung cancer with low-dose spiral computed tomography. Am J Respir Crit Care Med 165, 508-513.

Tockman MS (1986) Survival and Mortality from Lung Cancer in a Screened Population – The Johns Hopkins Study. Chest 84, 324S-325S.

Tockman MS, Anthonisen NR, Wright EC, Donithan MG (1987) The Intermittent Positive Pressure Trial Group and the Johns Hopkins Lung Project for the Early Detection of Lung Cancer. Ann Intern Med 106, 512-518.

Tockman MS, Erozan YS, Gupta P, et al (1994) The early detection of second primary lung cancers by sputum immunostaining. Chest 106, 385s-390s.

Tockman MS, Gupta PK, Myers JD, et al (1988) Sensitive and specific monoclonal antibody recognition of human lung cancer antigen on preserved sputum cells: a new approach to early lung cancer detection. J Clin Oncol 6, 1685-1693.

Weiss W (1999) Asbestosis: a marker for the increased risk pf lung cancer among workers exposed to asbestos. Chest 115, 536-549.

Wolpaw DR (1996) Early Detection in Lung Cancer: Case Finding and Screening. Med Clin North Am 80, 63-82.

Screening for lung cancer: a review and current status

Received: 05 May 2003; Accepted: 08 May 2003; electronically published: May 2003

I. Introduction