Definition ane Classifications
Bronchogenic carcinoma of the lung—lung cancer—comprises a group of malignant neoplasms that arise from bronchial epithelium. The four major pathologic cell types of lung cancer are small cell carcinoma, adenocarcinoma, squamous cell carcinoma, and large cell carcinoma. Because they have overlapping clinical behavior and response to treatment, adenocarcinoma, squamous cell carcinoma, and large cell carcinoma are generally grouped together in the category of non-small cell lung cancer (NSCLC). NSCLC represents 75% to 80% of all cases of lung cancer. Classification systems for the four major types of lung cancer have been formulated by the World Health Organization, the Armed Forces Institute of Pathology, and the Working Party for Lung Cancer [see Table 1].
Epidemiology and Etiology
In the United States, lung cancer is the second most common cancer in both men and women, surpassed only by prostate cancer in men and breast cancer in women. For 2003, a total of 169,400 new cases were predicted [see Table 2]. Lung cancer was expected to constitute 14% of all cancer diagnoses in men and 12% of those in women. However, lung cancer is the leading cause of cancer deaths, accounting for 31% and 25% of all cancer-related deaths in men and women, respectively. For 2003, expected deaths from lung cancer were 154,900 [see Table 2].1
The epidemiology of lung cancer in the United States directly reflects patterns in cigarette smoking, albeit with a 10- to 15-year lag time.2 Over recent decades, the prevalence of cigarette smoking in men has decreased from nearly 50% to approximately 25%, and the incidence of lung cancer in men has declined somewhat. During that same period, the prevalence of cigarette smoking in women has declined only 11%, to approximately 22%, and the incidence of lung cancer in women is only now leveling off.
In men, the incidence of lung cancer peaked in 1984, at 86.5 per 100,000 population, and by 1996 had declined to 70 per 100,000 population. For women, the incidence in 1996 was 42.3 per 100,000 population. Since 1987, more women have died each year from lung cancer than from breast cancer, and the margin between the two diseases continues to widen. Estimates suggest that in 2003, over 50% more women died of lung cancer than of breast cancer.1
Unfortunately, cigarette smoking became increasingly popular in teenagers in the 1990s. In the United States, the prevalence of cigarette smoking in high-school students increased during the 1990s, peaking during 1996 to 1997, then began a gradual decline.3 The popularity of smoking varied by ethnicity and race. In a 1999 survey of high-school students, smoking rates were 15.8% in blacks, 25.8% in Hispanics, and 32.8% in whites.4
Smoking cessation and lung cancer
Cigarette smoking continues to contribute to the risk of lung cancer long after a person has stopped smoking. The American Cancer Society evaluated this relationship in a 6-year prospective study involving more than 900,000 persons.5 This study included persons who had never smoked, current smokers, and former smokers. As expected, the risk of dying of lung cancer was lower in patients who had quit smoking early in life than in those who quit later on, and the risk was significantly lower in those who quit than in those who did not. In a person who smoked 26 cigarettes a day starting at 17 years of age and stopped smoking between the ages of 30 and 49, the risk of death from lung cancer is slightly greater than that for persons who never smoked. For a person quitting smoking between the ages of 50 and 64, the risk of death from lung cancer plateaus at the risk level at the time of quitting and remains level until about the age of 75, when the risk appears to increase further. In this model, the annual lung cancer mortality for current smokers at age 75 is 1% for men and 0.5% for women, which is approximately 20 times higher than that of nonsmokers. Nonsmokers (i.e., persons with a lifetime exposure of less than 100 cigarettes) have a relative risk of lung cancer of 0.05 or less as compared with current smokers. For former smokers, the relative risk of lung cancer death depends on the age of smoking cessation. The risk was 0.45 for smokers who quit in their early 60s, 0.2 for those who stopped smoking in their early 50s, and 0.1 for those who stopped smoking in their 30s. All available data indicate that the lung cancer risk for former smokers is still consistently greater than for those who never smoked. Stopping smoking at any age can reduce lung cancer mortality, but the risk reduction is much greater for smokers who quit at a younger age.
Table 1 Major Classifications of Lung Cancer
|
System- |
Non-Small Cell Lung Cancer |
Small Cell Cancer |
||
|
Squamous Cell* Carcinoma |
Adenocarcinoma |
Large Cell Carcinoma |
||
|
World Health Organization, No. 2 (WHO-No. 2) |
Spindle cell variant |
Acinar |
Oat cell |
|
|
Papillary |
Giant cell |
Intermediate cell |
||
|
Bronchioloalveolar |
Clear cell |
Combined |
||
|
Solid carcinoma with mucin |
||||
|
Armed Forces Institute of Pathology (AFIP) |
11 J CC . . . i Well differentiated |
|||
|
Well differentiated |
Moderately differentiated |
Undifferentiated |
Lymphocyte-like (oat cell) Polygonal (intermediate) |
|
|
Moderately differentiated |
Poorly differentiated |
Giant cell |
Combined (usually |
|
|
Poorly differentiated |
Bronchioloalveolar |
Clear cell |
squamous) |
|
|
Working Party for Lung Cancer (WPLC) |
Well differentiated |
Well differentiated |
With mucin production |
|
|
Moderately differentiated |
Moderately differentiated |
With stratification |
Lymphocyte-like (oat cell) |
|
|
Poorly differentiated |
Poorly differentiated |
Giant cell |
Intermediate cells (fusiform, polygonal, others) |
|
|
Bronchioloalveolar/papillary |
Clear cell |
|||
Note: both the WHO-No. 2 and AFIP systems have a fifth category, adenosquamous cell carcinoma; benign lesions, dysplasia, carcinoma in situ, carcinoid tumors, soft tissue sarcomas, and other respiratory tract lesions, which account for only a few percent of all lung cancers, are not included in this table. *For the WPLC system, the classification is epidermoid. For the WPLC system, the classification is small cell anaplastic.
Table 2 Epidemiology of Lung Cancer in the United States, 20031
|
|
New Cases |
Deaths |
|
Men |
90,200 |
89,200 |
|
Women |
79,200 |
65,700 |
|
Total |
169,400 |
154,900 |
In addition to age effects, there is a dose-response relationship for smoking and lung cancer. The risk for lung cancer increases with the duration of smoking and the number of cigarettes smoked. Earlier age of starting to smoke, deeper inhalation, and use of cigarettes that are unfiltered or have a high tar and nicotine content also increase the risk of lung cancer. In the current United States population over the age of 50, 23% are current smokers and 35% are former smokers. Because both groups remain at elevated risk for lung cancer for their lifetimes, clinicians should take an accurate quantitative smoking history in all patients.
Genetic susceptibility and molecular mechanisms
The risk of lung cancer is affected by genetic susceptibility. Women smokers may be at higher risk for the development of lung cancer than men with a similar smoking history. Furthermore, lung cancer mortality appears to be higher in African Americans.
Mechanisms for genetic susceptibility to lung cancer include genes that govern smoking behavior, which affect dopamine reward mechanisms related to nicotine and nicotine metabolism, as well as gender6; individual capacity for carcinogen metabolism; germline mutations coding for dysfunctional genes; and capacity to repair DNA damage from carcinogens. Several genetic abnormalities have been associated with lung cancers [see Table 3].
Lung cancer in nonsmokers
Given the dominant role of cigarette smoking in the etiology of lung cancer, determining the risk posed by other substances is difficult.7 As many as 25% of cases of lung cancer in non-smokers may result from second-hand tobacco smoke. A small percentage of lung cancers result from occupational exposure to carcinogens, including asbestos, arsenic, cadmium, chromium, radiation, radon, and chemicals such as chromoethyl ether. Heavy residential exposure to radon may be synergistic with cigarette smoking in promoting lung cancer, but the risk from residential radon for nonsmokers remains unclear.
Pathophysiology and Pathogenesis
The prevalences of histologic subtypes of lung cancer in men and women have changed in ways that mirror the changes in smoking habits. In the early studies that established the association between smoking and lung cancer, cigarettes were unfiltered, most of the participants were men, and squamous cell carcinoma was the most common cell type. Now, with filtered cigarettes widely popular and larger numbers of women smoking, adeno-carcinoma is the most common type of lung cancer in both young men and women. This changing pattern of histology correlates temporally with the change from unfiltered to filtered cigarettes and with reductions in the tar and nicotine content of cigarettes. Those changes in cigarette manufacturing have led to deeper inhalation of smoke into the lungs, which exposes the distal airways more heavily to the carcinogenic influences of tobacco smoke. Other factors likely play a part as well. In nonsmokers, adenocarcinomas are the most common histologic type of lung cancer.
The initiation of carcinogenesis from cigarette smoke is related to a complex mixture of carcinogens and tumor promoters combined with the delivery vehicle of inhalation. Serial studies of bronchial epithelium in smokers demonstrate an evolution from dysplasia to metaplasia to neoplastic changes.9-12 Each stage has been associated with a number of genetic alterations, and the pivotal mechanisms are a topic of intense investigation. Factors associated with genetic susceptibility have yet to be identified and may emerge from studies of lung cancer in nonsmokers. Thus, the main clinical criterion for susceptibility remains a history of current or former smoking.
Table 3 Selected Molecular Genetic Abnormalities Associated with Lung Cancer
|
Abnormal Genes |
Mutation |
Frequency of Abnormal Expression (%) |
|
|
NSCLC |
SCLC |
||
|
Oncogenes |
30 |
Not reported |
|
|
myc family |
DNA amplification/overexpression |
10 |
10-40 |
|
HER-2/neu |
Increased expression of p185neu |
25 |
Not reported |
|
Tumor suppressor genes |
|||
|
p53 |
Deletion |
50 |
80 |
|
Point mutation |
|||
|
Overexpression |
|||
|
Rb |
Deletion |
15 |
> 90 |
|
3p |
Deletion |
50 |
90 |
NSCLC—non-small cell lung cancer
SCLC—small cell lung cancer
Prevention
Primary prevention
Given that 87% of cases of lung cancer occur in smokers and that the risk of lung cancer is lower by at least 20-fold in persons who have never smoked, the obvious strategy for primary prevention is to keep young persons from starting to smoke and to promote smoking cessation in smokers of all ages [see CE:III Reducing Risk of Injury and Disease]. Although public health measures that discourage smoking in public places and in the workplace, as well as the development of negative societal attitudes toward smoking, are helpful in reducing the prevalence of smoking in adults, progress against smoking has been slow and teenage smoking rates remain unacceptably high.
Secondary prevention
The use of nutritional supplements by smokers as a strategy to reduce lung cancer was suggested by an epidemiologic association of lower serum levels of p-carotene, vitamin E, and retinoids with a higher risk of lung cancer.13 Unfortunately, in clinical trials, these agents did not reduce lung cancer risk.
One of the best known trials, the CARET (p-Carotene and Retinol Efficacy Trial), comprised over 18,000 smokers of both sexes randomized to receive a retinoid drug, retinol palmitate, in combination with p-carotene or placebo.14 In this trial, patients who received p-carotene and retinol palmitate had a higher rate of development of lung cancer (relative risk = 1.36) and higher lung cancer mortality (relative risk = 1.59). In another placebo-controlled trial, from Finland, that studied the effects of vitamin E and p-carotene, smokers who received p-carotene were more likely to develop lung cancer (relative risk = 1.16). Vitamin E produced no effect. The risk of harm from p-carotene in this trial was more pronounced in heavy smokers. In a placebo-controlled trial in patients with resected stage I NSCLC, the use of 13-cis-retinoic acid increased the rate of lung cancer recurrence and mortality in patients who continued to smoke. Thus, at present, no evidence supports recommending vitamins to prevent lung cancer, and there is some evidence that p-carotene and retinoids may have harmful effects in smokers, as well as in persons with occupational exposure to asbestos.
Diagnosis
Screening
Most patients with lung cancer present with advanced inoperable disease. Screening for detection of lung cancer at an earlier stage is therefore an attractive idea, especially because persons at high risk for lung cancer can be readily identified by a smoking history.
Early studies of screening produced disappointing results. Randomized trials of screening, conducted in the United States and in the former Czechoslovakia, suggested that chest x-ray alone was not a satisfactory screening tool to detect early lung cancer tumors. 16 Curable tumors are often too small or indistinct to be detected on a standard chest x-ray.
Spiral CT scanning may be a more sensitive technique for lung cancer screening. With this technique, radiologists obtain a low-resolution image of the entire thorax in a single breath-hold, with low radiation exposure and relatively rapid throughput compared with standard CT scans. A number of studies have demonstrated the feasibility of spiral CT scanning in screening for lung cancer. In the Early Lung Cancer Action Project (ELCAP), 1,000 asymptomatic persons older than 60 years with a smoking history of 10 or more pack-years underwent both spiral CT and chest x-ray.17 CT detected malignant nodules in 2.7% of the patients, compared with 0.6% by chest x-ray. Benign nodules were detected at a rate of 20.6% by CT versus 6.1% by chest x-ray, so careful follow-up is critical for avoiding unnecessary biopsy. A Mayo Clinic study of spiral CT18 also demonstrated enhanced detection of malignant nodules, most of which were early-stage lung cancer, but an even higher yield of benign nodules (60%), which emphasizes the potential drawback of this technique.
At present, no data from randomized trials exist to allow an evidence-based recommendation either for or against lung cancer screening. Despite encouraging results from nonrandom-ized trials, several issues remain to be addressed, including lead-time bias, generalization to a broader population, application to younger patients at lower risk of lung cancer, and long-term benefit in terms of lower lung cancer mortality. Furthermore, a decision and cost-effectiveness analysis has suggested that the cost of implementing such a strategy would be substan-tial.19 Currently, spiral CT screening cannot be recommended except in the context of a clinical trial. Other new technologies that deserve consideration as potential screening methods include analysis of sputum cytology by molecular markers and localization of tumors by fluorescence bronchoscopy.20
The National Cancer Institute is currently enrolling patients in the National Lung Screening Trial (NLST), a randomized, controlled trial that will compare standard chest x-rays with spiral CT as a screening method for lung cancer.21 The NLST will enroll 50,000 current or former smokers between the ages of 55 and 74 years at clinical trial sites throughout the United States. Study participants will receive either a chest x-ray or a spiral CT once a year for 3 years and will then undergo monitoring until 2009. The researchers will be looking for a reduction in mortality of 20% or more with either modality. In addition to the screenings, some NLST centers will test for biologic markers that may have potential for screening.
Clinical manifestations and laboratory studies
The signs and symptoms of lung cancer vary with the anatomic location of the tumor, its extension into surrounding structures, metastatic spread, and the systemic effects of para-neoplastic syndromes. Unfortunately, only 6% of patients with lung cancer are asymptomatic at the time of diagnosis. The remainder of the patients present with symptoms resulting from regional spread of the tumor, mediastinal lymph node involvement, or distant metastases.
Pulmonary Manifestations
The most common manifestation of the primary tumor is cough, which results from endobronchial erosion and irritation. Others are, in decreasing order of frequency, dyspnea, chest pain, hemoptysis, and postobstructive pneumonia or pneumonitis [see Table 4]. Centrally located tumors also typically cause stridor, wheezing, hemoptysis, dyspnea, or chest pain, often central in location. Occlusion of the airway by a tumor can lead to a postobstructive infiltrate or pneumonia. Large tumors may cavitate and present as a lung abscess.
Manifestations of Intrathoracic Disease
Intrathoracic extension of the tumor or spread to mediastinal lymph nodes may produce a variety of symptoms [see Table 4]. Although individually these symptoms occur in fewer than 10% of patients with lung cancer, collectively they represent significant complications of locally advanced NSCLC, either at diagnosis or during the subsequent disease course. Hoarseness may result from invasion of the recurrent laryngeal nerve and resultant vocal cord paralysis. Dysphagia may be a sign of compression of the esophagus. Extensive tumor involvement of the right mediastinal lymph nodes often results in the superior vena cava syndrome, which is characterized by plethoric appearance; distention of the venous drainage of the arm and neck; and edema of the face, neck, and arms. Vena caval obstruction usually progresses gradually, allowing the development of collateral venous drainage that may be detected on physical examination.
Table 4 Common Signs and Symptoms of Lung Cancer at Diagnosis
|
Site of Tumor |
|
Percentage of |
|
Involvement |
Signs or Symptoms |
Patients Affected |
|
Pulmonary |
Cough |
50-75 |
|
Dyspnea |
30-40 |
|
|
Chest pain |
25-40 |
|
|
Hemoptysis |
15-30 |
|
|
Pneumonia/pneumonitis |
10-25 |
|
|
Intrathoracic |
Hoarseness |
< 10 |
|
Dysphagia |
< 10 |
|
|
Facial/arm swelling |
< 10 |
|
|
Shoulder/arm pain |
< 10 |
|
|
Pleural/chest wall pain |
< 10 |
|
|
Pleural/pericardial effusion |
< 10 |
|
|
Paraneoplastic syndromes |
< 10 |
|
|
Extrathoracic |
Anorexia/weight loss |
30-50 |
|
Generalized weakness |
20-40 |
|
|
Bone pain |
20-30 |
|
|
Liver abnormalities |
10-20 |
|
|
Headache/CNS abnormalities |
5-15 |
|
|
Flank pain |
< 10 |
|
|
Other (e.g., subcutaneous nod- |
< 10 |
|
|
ule, distant lymph nodes) |
Shoulder and arm pain from superior sulcus (Pancoast) tumor syndrome is a commonly misdiagnosed sign of lung cancer. The pain results from local extension of a tumor in the apex of the lung, with involvement of the eighth cervical and first thoracic nerves. Unfortunately, this condition is often mistaken for arthritis. In many cases, careful physical examination will identify ipsilateral Horner syndrome, which is characterized by ptosis, meiosis, and anhydrosis. The Horner syndrome is related to paravertebral extension and sympathetic nerve involvement of the tumors.
Pleuritic pain and chest wall pain occur most commonly in patients with primary tumors in the lung periphery that spread to the pleura and, in some cases, extend directly to the chest wall. Associated pleural effusion may occur in such cases; large effusions may cause dyspnea. Malignant pericardial effusions may also develop and can cause cardiac tamponade.
Paraneoplastic syndromes A minority of lung cancer patients present with paraneoplastic manifestations. The biology of these syndromes remains poorly characterized, but the syndromes appear to be cytokine-mediated responses to antigens from the intrathoracic lung tumor, rather than the result of distant spread of cancer.
The most common paraneoplastic feature associated with lung cancer is clubbing of the fingers from periosteal swelling of the distal phalanges, which may occur in 5% to 15% of patients. In a small percentage of patients, clubbing may be part of a symptomatic hypertrophic osteoarthropathy. These patients often complain of a distal symmetrical arthritis that most commonly involves the ankles or knees but can also involve the wrists, elbows, and other joints. Misdiagnosis of this condition as a strictly rheumatologic phenomenon often results in delayed recognition of the underlying neoplasm.
Although weight loss and fatigue are commonly an indication of distant metastasis, they can also represent a paraneoplastic phenomenon that occasionally occurs even with early-stage tumors. Especially in patients with small cell lung cancer (SCLC), paraneoplastic manifestations can also take the form of specific neurologic syndromes, such as the Lambert-Eaton syndrome. These patients present with muscle weakness, a variety of peripheral neuropathies, and central nervous system involvement such as subacute cerebellar degeneration or limbic encephalitis.
Another category of neoplastic syndromes relates to aberrant hormone or peptide production by lung cancer tumor cells. The most common of these is hyponatremia secondary to production of antidiuretic hormone (SIADH). Hypercalcemia can result from tumors that secrete parathyroid hormone; and Cushing syndrome, from tumors that secrete adrenocorti-cotropic hormone. In general, these hormonal syndromes are more common in SCLC than in NSCLC, because of the neu-roendocrine nature of SCLC. However, hypercalcemia can have a range of causes—including both remote effects and direct interactions between tumor and bone—and is much more common in NSCLC than in SCLC.
