NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

Humphrey LL, Johnson M, Teutsch S. Lung Cancer Screening: An Update for the U.S. Preventive Services Task Force [Internet]. Rockville (MD): Agency for Healthcare Research and Quality (US); 2004 May. (Systematic Evidence Reviews, No. 31.)

  • This publication is provided for historical reference only and the information may be out of date.

This publication is provided for historical reference only and the information may be out of date.

Cover of Lung Cancer Screening

Lung Cancer Screening: An Update for the U.S. Preventive Services Task Force [Internet].

Show details

4Results

In our searches, we identified 809 citations and reviewed their abstracts. One hundred and forty-nine full text papers were reviewed. From them, we identified 5 randomized controlled trials28, 29, 4346; 4 non-randomized, uncontrolled trials4751; 6 case-control studies5257; and 1 non-randomized, controlled trial58 of chest x-ray and/or sputum cytology screening for lung cancer. A randomized trial of chest x-ray in conjunction with a multiphasic screening program also was reviewed.59, 60 In addition, we reviewed 6 uncontrolled cohort studies of lung cancer screening with computerized tomography (CT).6169

Lung Cancer Screening with Chest X-Ray +/- Sputum Cytology

Randomized Controlled Trials

Tables 2 and 3 summarize the controlled trials of lung cancer screening and their methods and quality. Figure 2 shows the relative risks and confidence intervals of the randomized trials. In the 1960s, a controlled nonrandomized trial of chest x-ray screening every 6 months involving approximately 55,000 men older than 40 years was conducted by the Northwest London Mass Radiography Service.45, 46 In this cluster-randomized trial, 29,723 male factory workers from 75 randomly identified firms were offered chest x-ray every 6 months and compared to 25,300 controls from other factories offered screening at baseline and at 3 years. After 3 years, the annual mortality rate from lung cancer in the intervention group was 0.7/1,000; the rate was 0.8/1,000 in the control population, not different statistically.

Table 2. Controlled trials of lung cancer screening with chest x-ray with or without sputum cytology.

Table 2

Controlled trials of lung cancer screening with chest x-ray with or without sputum cytology.

Table 3. Methods and quality of randomized controlled trials of lung cancer screening.

Table 3

Methods and quality of randomized controlled trials of lung cancer screening.

Figure 2. Mortality in randomized controlled trials of lung cancer screening with chest x-ray with or without sputum cytology.

Figure

Figure 2. Mortality in randomized controlled trials of lung cancer screening with chest x-ray with or without sputum cytology.

The Kaiser-Permanente Multiphasic Evaluation Study was a randomized trial designed to determine whether or not encouraging middle-aged people to have annual multiphasic health checkups (MHC) would reduce mortality “from a group of diseases hypothesized in advance to have a fatal outcome preventable or postponable through periodic MHCs.”60 The MHC consisted of several screening tests including chest x-ray. More than 10,700 members of Kaiser's large health maintenance organization aged 35–54 of average risk (17% smokers) were randomized by medical record number into a study group (n=5,156) who were encouraged to undergo the MHC annually, and a control group (n=5,557) who were not advised regarding the MHC but who could receive one if they requested it. Approximately 60% of the study group and 20% of the control group underwent the MHC annually. There were 44 lung cancer deaths in the study group and 42 in the control group (death rates per 1,000 for the 16-year follow-up period of 8.6 and 7.6 respectively)—not a statistically significant difference.

Three National Cancer Institute-sponsored randomized controlled trials of lung cancer screening in male smokers were conducted in the United States in the 1970s28, 44, 7075 and a fourth in Czechoslovakia.43, 76 Those conducted at the Memorial Sloan Kettering (MSK) Cancer Center28, 71, 72, 7779 and at Johns Hopkins (JH) University44, 70, 8083 were identical in design and were conducted to evaluate the incremental benefit of adding sputum cytology to annual chest x-ray. Each was rated of fair quality based on USPSTF criteria. The reasons for each quality score are shown in Table 1. Of the 20,400 male smokers (at least 20 pack years of smoking) age 45 and above who volunteered for these 2 studies, 10,194 were randomized into a “dual screen” group that was offered screening with annual chest x-ray and sputum cytology every 4 months for 5 years, and 10,233 to a chest x-ray group that was offered annual chest x-ray screening for 5 years. Each group was followed for 5–8 years.

In the MSK study, the baseline (prevalence) screen identified 30 (6.0/1,000) lung malignancies in the “dual screen” group and 23 (4.6/1,000) in the chest x-ray group.71 Average 5-year survival of the prevalence cases among the dual screen patients was 48% and among those in the chest x-ray group, 37%. Following the prevalence screen, 114 subsequent (incident) lung cancers were identified in the dual screen group and 121 in the annual x-ray group, with 33 and 32 cases, respectively, diagnosed in the 2 years following screening. Combining the incidence and prevalence tumors, 144 lung cancers were detected in each group28, 72, 79; 40% of all lung cancers detected were stage I. Survival of both groups was approximately 35% at 5 years. This compares with an average 5-year survival for lung cancer in the general population at that time of 10%. The mortality rate was 2.7/1,000 person-years in both the chest x-ray and dual screen groups.

In the JH study, the prevalence screen identified 39 malignancies in the dual screen and 40 in the chest x-ray group.44, 82 Prevalence varied significantly with age and reached 2.2% in individuals above age 65. Fifty-six percent of the cancers in the prevalence cases were resected and 5-year survival was 33% and 59% for the chest x-ray and dual screen groups, respectively.44 After 8 years of follow-up, 194 incidence cancers were identified in the dual screen and 202 in the chest x-ray group. Survival at 8 years was approximately 20% for both groups, with a mortality rate of 3.4/1000 person-years in the dual screen group and 3.8/1,000 person-years in the chest x-ray group, not statistically significant differences. These rates were similar to community lung cancer mortality rates at the time.82

The authors of these studies concluded that adding sputum cytology to annual chest x-ray screening had no benefit in reducing lung cancer mortality. It is notable that the minimum screening intensity in these studies involved annual chest x-ray over a period of 6 years, and that the observed 5-year survival of 35% in both groups was higher than the average lung cancer survival at the time the study was conducted, and continues to be higher than current usual 5-year survival. However, the absence of a mortality benefit among trial participants in the Johns Hopkins study when compared to the population at the time argues against an important screening benefit and suggests survival was prolonged relative to the community because of screening biases, particularly lead-time and length bias. In a recent Cochrane review, the results of the Memorial Sloan-Kettering and Johns Hopkins studies were pooled using a random effects model and showed a trend toward reduced lung cancer mortality in the intervention group (RR 0.88; 95% CI, 0.74–1.03).41

The first trial to evaluate the value of intense screening with chest x-ray was the Mayo Clinic Lung Project (MLP) involving 10,933 male smokers age 45 and above.29, 74, 75, 8491 All participants underwent a prevalence screen with sputum cytology and chest x-ray; 91 cancers were identified (prevalence 0.8%) with resectability rates of 54% and 5-year survival of 40%, which was more than twice the survival of an age-similar Mayo Clinic comparison group with lung cancer. Almost half the prevalence cases were stage I or II and among these cases, 5-year survival was 70%.29, 73, 75

After the prevalence screen, 9,211 men were randomized to either a study group (n = 4,618) and screened with chest x-ray and pooled 3-day sputum cytology every 4 months for 6 years, or to a control group (n=4,593) and advised to have annual chest x-ray and sputum cytology. During the study period, 206 incidence cases of lung cancer were identified in the experimental group and 160 incidence cases identified in the control group. Resectability of the cases was 46% in the experimental group and 32% in the control group, with 5-year survival of 33% in the experimental group and 15% in the control group. After 6 years of follow-up, there were 115 lung cancer deaths in the control group and 122 in the study group, with death rates of 3.0/1,000 patient-years in the control group and 3.2/1,000 patient-years in the study group, not significantly different. After 20.5 years of follow-up the lung cancer death rates were 4.4 (95% CI, 3.9–4.9) and 3.9 (95% CI, 3.5–4.4) per 1,000 person-years in the intervention and control groups, respectively.88

The Mayo Lung Project was the first individually randomized controlled trial to specifically evaluate the role of chest x-ray in lung cancer screening and was the most influential in determining current public health policy. Although it is rated of fair quality by USPSTF criteria, there are several limitations of the study. 1. A prevalence screen detected 91 cases (0.8%). Thus, there was no completely unscreened control group. Also, these cases were followed separately and not evaluated in the randomized comparison. Thus, any effect of these cases on mortality could not be determined. 2. Nearly half of the control subjects obtained annual chest x-rays during the course of the study, with one-third of the malignancies in the control group discovered by screening chest x-ray. Seventy-three percent of the controls received chest x-rays during the study's last 2 years. 3. Compliance of the intervention group was 75%, reducing the study's power.73 4. Assuming full compliance of the study group, the study was underpowered from the beginning with a 48% chance of detecting a 20% reduction in mortality.73 Based on screening of the control group, the reduced compliance of the intervention group, the initial study power, and the relatively short duration of follow-up, 1 analysis of the first study results reflecting 6 years of follow-up determined that it would have less than a 20% probability of showing a benefit from screening.74 However, in the most recent MLP follow-up, Marcus et al. effectively demonstrate that with a greater length of follow-up, a major reduction in lung cancer mortality was not missed due to low power.88, 92

The incidence of lung cancer in the experimental group in the Mayo Clinic Study was approximately 22% higher than in the control group. Even 3 years after the end of the trial, there were 46 more malignancies in the intervention group than in the control group.73 It is unlikely that the increase in incidence was a consequence of radiation from the screening x-rays, based on radiation exposure literature.93 Strauss et al. has suggested that this may be due to non-random distribution of important lung cancer risk factors between the control and intervention group.94 This issue was evaluated by Marcus89 and the distribution of several potential lung cancer risk factors was not found to vary significantly between the intervention and control groups. Although little detailed information is provided, there is evidence on review of the MLP publications that not all patients were asymptomatic,29, 73 which could alter the findings of the screening study if patients with symptoms were disproportionately enrolled in the intervention arm of the study. However, there is no evidence to support this. Another possibility is that the higher incidence of lung cancer in the screened population may represent the diagnosis of insignificant lung cancers, e.g., overdiagnosis. This issue is complex and also relevant to the CT studies and is discussed below.34, 50, 94, 95

The most recent randomized controlled trial was conducted in Czechoslovakia43, 76, 96 where 6,364 male smokers ages 40–64 received chest x-ray and sputum cytology as a prevalence screen; 19 lung cancer cases were identified (3.0/1,000). After exclusion of the prevalence cases, 3,172 men were randomized to the study group and 3,174 to the control group. The study group received chest x-ray and sputum cytology every 6 months for 3 years and the control group received a single chest x-ray and sputum cytology at the end of the screening period. In the next 3 years of follow-up, chest x-rays were administered annually to subjects in both groups. In the first 3 years of the study, 36 lung cancers were identified in the experimental group and 19 in the control group. After 6 years, there were 108 malignancies identified in the experimental group and 82 in the control group, with 85 lung cancer deaths in the study group and 67 in the control group, rates that were not statistically different. A problem in randomization is suggested by the finding of significantly higher total all-cause, cancer, and smoking-related mortality in the intervention group (341 versus 291 deaths), over the entire 6-year study period.76 This study is of poor quality, based on USPSTF criteria, and we consider the results likely to be invalid.

All cause mortality was calculated in the recent Cochrane review and data were available from the Erfurt County, Czechoslavakian, MSK, MLP, and Kaiser studies. Pooled analysis comparing frequent chest x-ray screening with less frequent screening, excluding the Czech trial, identified a relative risk of 0.97 (95% CI, 0.94–1.01) using a fixed effects model.41

In summary, 2 fair-quality randomized trials evaluating intense screening among high risk males with sputum cytology have shown no benefit of adding cytology to annual chest x-ray. Two fair-quality trials among high risk men comparing intense chest x-rays with less intense chest x-rays showed no benefit. Finally, a fair-poor randomized trial of multiple procedures in a low lung cancer risk population showed no benefit of annual chest x-ray.

Case-Control Studies

Six case-control studies were identified in our search, 5 conducted in Japan and 1 in Berlin. The five Japanese studies are all of fair quality and the German study of poor quality, based on USPSTF criteria; the studies' methods and findings are summarized in Table 4. In the 1950s a case-control study based on a tuberculosis screening program in which chest x-ray was offered every 1–2 years to all adults in the former German Democratic Republic showed no association between periodic chest x-ray and reduced lung cancer mortality.52 This study's quality is rated poor, based on poor smoking assessment among controls, no report of response quality rates and use of hospital controls.

Table 4. Case control studies of lung cancer screening.

Table 4

Case control studies of lung cancer screening.

A 1992 case-control study from Japan35 reported data from 15 municipal lung cancer screening programs where participants were screened using chest x-ray, and in some areas, sputum cytologic tests. In general, the population offered screening consisted of high-risk men (smoking exposure) and low-risk women. The cases were comprised of 273 patients (208 men, 65 women) with fatal lung cancer. Cases' screening histories were compared with those of 1,269 control subjects matched by sex, age, smoking status, and type of health insurance. The odds ratio for dying from lung cancer was lower at 0.72 (95% CI, 0.50–1.03) for those screened with chest x-ray +/- sputum within 12 months of diagnosis. For those screened between 12 and 24 months prior to the diagnosis, the odds ratio was 0.83 (95% CI, 0.56–1.23). The odds ratio for women screened within 12 months of diagnosis was 0.42 (95% CI, 0.20–0.87) (Table 4).

In a recent case-control study performed in Japan,54 the study group was comprised of 193 persons dying of lung cancer (158 men and 35 women), aged 40–74, and holding national health insurance. Three controls for each case were selected randomly from living national health insurance holders matched by residence, gender, and year of birth (n=579). The screening histories of the cases and controls were reviewed dating from the time of diagnosis of the case. After adjusting for smoking history, the odds of dying from lung cancer for participants screened within 12 months of diagnosis compared with non-screened individuals was 0.54 (95% CI, 0.34–0.85). The odds ratio for screening in the 12–24 months prior to diagnosis was 0.64 (95% CI, 0.30–0.97); no significant benefit was observed for screening occurring over 24 months prior to the diagnosis of lung cancer (OR 0.59; 95% CI, 0.30–1.15) (Table 4).

A case-control study conducted in the Miyagi Prefecture in Japan,55 where mass screening with chest x-ray and sputum cytology had been conducted since 1982, was of similar design to the above Japanese studies and involved 328 cases of fatal lung cancer diagnosed after 1990 among individuals aged 40–79 (70 low-risk women, 258 mixed-risk men). Each case was matched by gender, age, municipality, and smoking history, to 6 controls. The odds ratio for screening within 12 months was 0.54 (95% CI, 0.41–0.73) and 0.62 (95% CI, 0.42–0.92) for screening within 24 months. The odds ratio was 0.64 (95% CI, 0.36–1.14) for screening within 36 months. Screening 36–48 months prior to diagnosis showed no benefit. Among women (all “low-risk”), the odds ratio for screening within 12 months was 0.57 (95% CI, 0.30–1.11).

Another case-control study conducted among holders of National Health Insurance in the Nigata Prefecture of Japan,56 with similar methods among 25 low-risk women and 149 high-risk men with fatal lung cancer, identified reduced risk of fatal lung cancer among individuals screened within 12 months (OR 0.40; 95% CI, 0.27–0.59). Unlike 2 of the other Japanese case-control studies, the benefit did not extend past 12 months. Among women (all “low risk”), the odds ratio was 0.61 (95% CI, 0.23–1.68).

Finally, a case-control study was conducted among Japanese individuals in the Okayama Prefecture where population-based lung cancer screening had been conducted. Four hundred and twelve individuals with fatal lung cancer aged 40–79 were matched with 2–10 controls by gender, age, and district. After adjustment for smoking exposure, the odds ratio for screening within 12 months was 0.59 (95% CI, 0.46–0.74). Among women of mixed smoking status, the odds ratio associated with screening within 12 months was 0.39 (95% CI, 0.24–0.64), lower than among men (RR 0.67; 95% CI, 0.51–0.87).

In summary, 5 fair quality case-control studies among high to average risk men and low risk women, suggest a screening benefit. All are limited by lack of control for occupational exposures and family history and, as discussed below, the findings must be interpreted cautiously due to lack of randomization and possible healthy screenee bias.

Controlled, Non-randomized Studies

A study conducted in Germany between 1972 and 1977 compared chest x-ray screening every 6 months in 41,532 males in 4 districts with chest x-ray screening every 18 months in 102,348 men in 10 districts. After 10 years, lung cancer mortality rates were similar at 0.6/1,000 and 0.8/1,000 in the intervention and control groups, respectively, not statistically different.58 The details of this study are described in Table 5.

Table 5. Nonrandomized controlled studies of lung cancer screening.

Table 5

Nonrandomized controlled studies of lung cancer screening.

Uncontrolled Studies

Four non-randomized, uncontrolled studies of lung cancer screening were conducted in the 1950s and 1960s. Two showing no survival benefit with screening when compared to usual population survival were the Philadelphia Pulmonary Neoplasm Research Project47, 48 and the Veterans Administration Trial.49 The Tokyo Metropolitan Government Study50 and the South London Cancer Study51 evaluated periodic chest x-ray screening and showed improved survival from lung cancer when compared to population rates at the time. Details of these studies are shown in Table 6. Because they do not have control groups they receive little emphasis in this review and were not rated in quality.

Table 6. Nonrandomized uncontrolled studies of lung cancer screening.

Table 6

Nonrandomized uncontrolled studies of lung cancer screening.

Lung Cancer Screening with CT

Several recent cohort studies, all without control groups, have evaluated screening for lung cancer with low dose CT. The details of these studies are shown in Table 7. In 1999, a study by Henschke et al. reported findings from the baseline screen of the Early Lung Cancer Action Project (ELCAP).65 The goals of this uncontrolled study were to compare low-dose CT and chest x-ray (in the same subjects) and to determine how nodule size affects survival. The study involved 1,000 symptom-free volunteers age 60 and above, with at least 10 pack years of smoking, and no prior malignancy, who were evaluated as medically fit for surgery and who each underwent chest x-ray and low-dose CT. Each CT and chest x-ray were read independently by chest radiologists; all nodules were radiographically characterized and evaluated according to standardized protocols, including high-resolution CT (HRCT), biopsy, or short-term follow-up. The study results are summarized in Table 7. The study details are provided here also given the current intense public interest in screening CT, based largely on this trial and because the study contributes important information about the test characteristics of current chest x-ray screening.

Table 7. Low dose computerized tomography lung cancer screening outcomes.

Table 7

Low dose computerized tomography lung cancer screening outcomes.

The population was 46% female, had a median age of 67, and median pack years of 45. Approximately 14% of the entire population had been exposed to asbestos. Baseline (prevalence screen) chest x-ray identified 68 individuals with non-calcified nodules (NCN), of which 33 were confirmed by low-dose CT scan; 7 were malignant, and all were resectable. Baseline (prevalence screen) low-dose CT identified 233 individuals with 1 to 6 NCN. High-resolution CT scan was recommended to 233 and 183 complied. Among them, 63 were negative and 16 had biopsies recommended, with 13 malignancies identified. One hundred and four other individuals with abnormal high-resolution CT were followed with repeat CT and from these, 14 more biopsies were recommended; all revealed malignancy. Of the 30 recommended biopsies, which included 9 video assisted thorascopic surgeries (VATS), 27 malignancies were identified, of which 26 were resectable and 23 were stage I disease. Four other lung cancers were also diagnosed based on non-nodule CT abnormatlities. Thus, the prevalence of lung cancer following a baseline low-dose CT and follow-up of abnormalities was 3.1%. These findings are shown in Table 7.

Six to eighteen months after the baseline exam, 841 underwent a first repeat screen and 461 underwent a first repeat (incidence) exam. Based on the timing of the incidence exam the investigators estimate there were 1184 “annual” examinations. Among the original 1,000 alive and without cancer, 2 developed symptom-detected lung cancer (interval cancer). Thirty of the 1,184 (2.5%) individuals undergoing annual screening required further evaluation. Two of these 30 died prior to workup and of the remaining 28, 12 nodules resolved over 1 month, 8 were followed with HRCT, and 9 underwent biopsy resulting in 7 lung cancer diagnoses (6 stage IA).97

The significant findings from this study were that the prevalence of lung cancer was relatively high, and that low-dose CT has the ability to detect malignant tumors 4 times more commonly than chest x-ray, and stage I tumors 6 times more frequently than chest x-ray, suggesting that the sensitivity of chest x-ray compared to low-dose CT is 25% for detecting malignancy. There are no mortality data available yet on the ELCAP cohort; however, based on data from lung cancer survival in the usual practice setting, the expected 5-year survival of the 85% of patients with stage I non-small cell lung cancer is estimated at 65–70%10, 98 and possibly higher based on the size distribution within the stage I carcinomas detected on the initial screen. The ELCAP study is also helpful in describing the screening test accuracy of chest x-ray since CT is usually the definitive test for evaluating an abnormal chest x-ray. The test characteristics of chest x-ray based on this study and using CT as a gold standard are shown in Table 8.

Table 8. Test characteristics of chest x-ray based on ELCAP CT study.

Table 8

Test characteristics of chest x-ray based on ELCAP CT study.

Three LDCT studies have been conducted in Japan involving large numbers of both high- and low-risk men and women. Each study used different protocols but each included chest x-ray and sputum cytology. The details of these studies are described in Table 7. One study involved 5,483 prevalence and 8,303 incidence screens and detected 59 primary lung cancers among men and women aged 40–74 with LDCT; 1 lung cancer was detected with sputum cytology only. Fifty-five out of 60 were stage I at diagnosis; no survival or mortality data have been published.61, 68 Interestingly, the prevalence (baseline) screen identified more malignancies in non-smokers than smokers (0.44% versus 0.40%, respectively). Incident screens revealed more tumors in smokers than non-smokers, suggesting different biological properties of the tumors in smokers and non-smokers.

Another study included 742 participants who had been screened biennially with chest x-ray and sputum cytology for a number of years and 940 new participants (n=1,611) who underwent LDCT screening exams beginning in 1993. These participants were current smokers, age 40 and above, and recruited from the general population (1,415 male, 196 female). Among 1,611 baseline screens, 11.5% had abnormal CTs, 3.4% abnormal chest x-rays, and 0.8% abnormal sputum cytology from which 13 lung cancers (11 Stage I) were diagnosed based on CT and one lung cancer was diagnosed with sputum cytology only. 7,891 repeat screens resulted in 721 (9.1%) abnormal nodules, 719 HRCT and 35 biopsies, with 19 lung cancer diagnoses (18 Stage I) based on CT findings; 3 lung cancers were detected with sputum cytology only.62, 99 Among the 56 biopsies performed because of abnormalities on LDCT, 20 VATS procedures were performed resulting in 14 lung cancer diagnoses. Finally, another Japanese cohort study involved 7,956 Hitachi employees (1,637 women, 6,319 men) aged 50–69 where chest CT was conducted as part of annual health examinations.67 In this study, the prevalence screen identified 2,099 individuals with abnormal nodules from which 541 underwent HRCT, and 64 were further evaluated; 24 malignancies were identified in men and 12 in women. The annual incidence screen among 5,568 individuals identified 4 new malignancies. Thoracotomy, (including VATS) was performed in 57 patients resulting in 40 diagnoses of lung cancer.

A LDCT study conducted at the Mayo Clinic involved 1,520 men and women aged 50 or older with 20 or more pack years of smoking.66, 69, 100 Each participant underwent a baseline (prevalence) screen and from this screen, 782 (51%) individuals had 1 or more non-calcified nodules and 26 (1.7%) were diagnosed with primary lung cancer; one individual was diagnosed with lung cancer with sputum cytology only. Among 1,464 individuals from this cohort who underwent an annual incidence screen,(2,916 screens), 191 (13%) individuals were found to have new nodules and 10 new diagnoses of lung cancer were made (6.7/1000). There were also 2 interval and 2 cancers diagnosed with sputum cytology. Of the 50 individuals with malignancies identified, 36 were non-small cell lung cancer (NSCLC) of which 31 (86%) were resected for cure; 8 patients underwent surgery for benign disease in this study; there were also 2 interval cancers and 2 cancers diagnosed with sputum cytology only (1 at baseline and 1 at incidence screen). It is possible that the high number of nodules in this study is because of the Mayo Clinic's Midwest location and higher rates of histoplasmosis in this region.

A German study101 involving 817 asymptomatic volunteers age 40 and above with at least 20 pack years of smoking was conducted between November 1995 and July 1999. The median age was 53, and 229 of the 817 were women. All underwent LDCT and each CT was read by 1 of 2 radiologists. All non-calcified nodules greater than 10 mm were considered potentially malignant and evaluated with HRCT or follow-up in 3, 6, 12, and 24 months. Non-calcified nodules less than 10 mm were followed with repeat low dose CT. The prevalence findings included 350 individuals with non-calcified nodules and HRCT was recommended. Of the 350, 81 were re-evaluated at 3 months and 269 underwent HRCT, which identified 32 nodules in 29 individuals. Seventeen individuals (18 nodules) had morphology suggesting benign causes and were followed; 1 of these grew over 24 months and was diagnosed as stage I adenocarcinoma. Twelve underwent biopsy (including 8 VATS procedures) and malignancy was diagnosed in 10 with one additional interval lung cancer; 6 were stage IA. After an average of 2.7 years of follow-up, 6 are alive without evidence of recurrence. In total, 13 individuals underwent biopsy and 11 malignancies were diagnosed.

In summary, 6 cohort studies evaluating screening with low dose CT have shown that CT is significantly more sensitive than chest x-ray for identifying non-calcified nodules, that the false positive rates are high but significantly reduced with non-invasive follow-up (CT or time) and that the majority of patients undergoing biopsy are diagnosed with lung cancer, usually at an early stage. Importantly, no conclusions about the overall impact of this procedure on reducing lung cancer mortality can be made in the absence of randomization and a control group with mortality as an outcome.

Lung Cancer Screening Among Women

Lung cancer is the leading cause of cancer-related death among women in the United States.102 Most lung cancer in women is attributed to smoking.102 However, women have substantial exposure to passive smoking and a significant proportion of lung cancer in non-smoking women is attributed to passive smoking.12 In addition, although it is controversial, some studies suggest that for any level of smoking, women are at higher risk of developing cancer than men.3, 14, 15 For unknown reasons, women also tend to develop adenocarcinoma of the lung disproportionately to men,10, 15, 23 and adenocarcinoma is also found more commonly among non-smokers.23 This cell type tends to occur peripherally10, 103 and may be more apt to be detected with chest x-ray and/or CT than other cell types. As a consequence, radiologic imaging and screening for lung cancer may perform differently and may actually be better among women. Unfortunately, no randomized trials of specific lung cancer screening have included women. The only data evaluating screening among women and including control populations come from 4 Japanese case-control studies evaluating screening among primarily non-smoking women (passive smoking not assessed). These studies are summarized in Table 9 and show odds ratios for screening conducted within 12 months of lung cancer diagnosis 0.39–0.61, 2 studies statistically significant; however, interpretation is limited by the screening biases discussed in this review. Five studies of LDCT have included women; mortality data are not yet available. In addition, randomized trials of lung cancer screening with chest x-ray and/or CT involving women are currently underway.

Table 9. Lung cancer screening studies including women.

Table 9

Lung cancer screening studies including women.

Views

  • PubReader
  • Print View
  • Cite this Page

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...