Linda F. Barr
Lung cancer is the leading cause of visceral cancer and cancer-related death in the United States. In 1999 there were an estimated 171,600 new cases of lung cancer and 158,900 deaths from the disease. For each lung cancer patient, an average of 14.7 years of life is lost prematurely (1). The turn of the millennium has seen a small decline in lung cancer mortality for men but a large increase in mortality for women (2). Because of its close association with tobacco smoke, lung cancer is generally a preventable tumor. Unfortunately, the diagnosis usually occurs late in the course of the disease after metastasis has occurred and determined the outcome. New radiographic methods enhance lung cancer detection but remain unproven for benefiting survival. Advances in the management of lung cancer have resulted from both the improved delivery of older chemotherapeutic agents and the addition of new, tumor-specific drugs.
Epidemiology
Tobacco
From 80% to 90% of lung cancers are caused by tobacco smoke, most importantly from cigarettes, but also from
P.926
pipes and cigars. The increasing death rate from lung cancer in the past 50 years lags 20 years behind a parallel rise in cigarette smoking. Ominously, although the prevalence of current cigarette use declined over >30 years, it significantly increased among high-school students in the United States from 27.5% in 1991 to 36.4% in 1997. Furthermore, cigar smoking has increased by 50% among all age groups (3). When cigarette use has declined, so also has lung cancer incidence. During its first decade, the California tobacco control program led to significant declines in smoking, in association with declines in lung cancer and heart disease mortality (4).
The lifetime lung cancer mortality for the general population is approximately 10% for moderate smokers, 20% for heavy smokers, and 1% for nonsmokers (5), a compelling statistic because 47 million Americans smoke. For smokers, the most important determinant of lung cancer risk is the duration of cigarette smoking, and the number of cigarettes smoked per day has a multiplicative effect (6). The risk of lung cancer increases approximately with the fourth power of the number of years of smoking and the square of the number of cigarettes smoked daily (7). Because of the duration effect, individuals who start smoking before age 15 years are four times more likely to develop lung cancer than those who begin after age 25 years (6). Furthermore, exposure to smoke from other persons’ cigarettes (“passive smoking”) leads to an increased risk for lung cancer (8, 9, 10). The risk of lung cancer declines after 5 years from smoking cessation and continues to decrease with duration of time from quitting; however, some risk remains for former heavy smokers. Indeed, almost half of lung cancers occur in former smokers.
Although those who smoke nonfiltered cigarettes have the highest risk for lung cancer, the amount of tar in each cigarette does not correlate with cancer risk (11). This is because smokers of low-tar and low-nicotine cigarettes typically increase the depth and length of their cigarette inhalation to get their required dose of nicotine (12). Research demonstrates that carcinogen uptake is only modestly and transiently reduced when the number of daily cigarettes is decreased (13). Although the risk of lung cancer is strongly related to the number of cigarettes smoked daily, these results suggest that the only way to decrease lung cancer risk from cigarettes is to stop smoking completely.
Occupational Exposure
Table 61.1 lists other exposures that increase the risk for lung cancer and examples of relevant occupations (14). It is especially important to identify people with asbestos exposure. Not only do asbestos-exposed individuals have an increased incidence of mesothelioma, they also have a sixfold greater risk for developing lung cancer than the general population. Those exposed people who smoke are 60 times more likely to develop bronchogenic carcinoma than are nonsmoking nonexposed people (15).
|
TABLE 61.1 Occupational Agents Associated with Lung Cancer |
||||||||||||||||||||||||||||||||||||||
|
||||||||||||||||||||||||||||||||||||||
Radon
Radon is a pulmonary carcinogen that is a naturally produced radioactive gas found universally in the soil and air. The contribution of radon exposure to excess lung cancer in uranium and other underground miners (including those who mine iron, zinc, tin, and fluorspar) is well established. Of unproven but theoretical concern is the lung cancer risk from exposure to radon in contaminated soil beneath some homes. By extrapolation from miners’ data, it is estimated that such exposure may be responsible for a relative risk of 1.14 (95% confidence interval = 1.0-1.3) at 150 Bq/m3 (the standard measure of radiation
P.927
exposure) and may account for 6,000 to 36,000 lung cancer deaths each year in the United States. There may be a greater than additive risk for cigarette smoking and home radon exposure (16). Because of uncertainty in the risk estimates, it has been suggested that homes should be tested for radon using commercially available tests and corrective measures taken when the exposure rate approaches 150 Bq/m3. Whether such measures are effective in reducing the risk of lung cancer is unknown.
Other Risk Factors
It is important to consider other groups with increased risk for bronchogenic carcinoma. First, those with previous lung cancers or with other tobacco-associated cancers are at markedly increased risk for developing a second cancer of the respiratory or upper digestive tract (discussed later in Follow-up of Patients Surviving Lung Cancer section). Second, some studies suggest that genetic predisposition influences the risk of lung cancer. Stratified case control studies controlled for cigarette smoking have determined that lung cancer patients have an odds ratio of 1.7 to 5.3 for having a first-order relative with lung cancer (17,18). This increased risk may be the result of an inadequacy of DNA repair capacity. Studies of the repair of bleomycin- and benzo[a]pyrene diol epoxide–induced chromatid damage (the latter chemical is a carcinogenic derivative of tobacco smoke) show significant differences in the DNA repair ability between lymphocytes derived from lung cancer patients and those derived from age- and ethnicity-matched control subjects (19). Alternatively, genetic differences in lung cancer risk may reflect differences in the functioning of enzymes in the metabolic pathways of carcinogens.
Third, the presence of chronic lung disease may increase lung cancer risk. This has been described for both chronic obstructive pulmonary disease (20) and interstitial disease (21). Fourth, human immunodeficiency virus infection is associated with a relative risk of lung cancer of 6.5 compared with the general population. As with other acquired immunodeficiency syndrome (AIDS)-associated malignancies, lung cancer is more aggressive and manifests a worse prognosis in the AIDS patient (22).
Screening
The diagnosis of lung cancer is generally made in the last quarter of the tumor's life cycle, usually after metastases have occurred and essentially have determined the outcome (23). This observation underlies the dismal survival of patients with the disease. For all stages combined, 5-year survival for patients with non–small cell lung cancer (NSCLC) is 14% and for those with small cell lung cancer (SCLC) is 6%, with only a modest gain over the past 2 decades. However, more than half of patients with disease discovered at a localized stage survive for at least 5 years (24). Thus, there is much interest in detecting lung cancers at an earlier curable stage. As with other adult solid tumors, lung cancers have a long preclinical phase characterized by accumulating genetic changes over a decade or more. This may reflect the relationship between tumor size and doubling time. A 1-mm tumor has already undergone 20 doublings, and a 1-cm tumor has already undergone 30 doublings. Over the extrapolated range of tumor doubling times, this implies a lifespan of 5 to 10 years with accumulating genetic damage, angiogenesis, and metastasis occurring along the way (25).
The presence of a large pool of undiagnosed lung cancer in current and former smokers is supported by the discovery of a 6.4% incidence of unsuspected malignant nodules in the lung specimens of patients undergoing lung volume reduction surgery (26). Earlier large randomized studies at multiple centers showed that screening cigarette smokers with yearly chest x-ray films and sputum cytology improved the detection of cancer and the survival (time from detection to death) but did not lead to a significant reduction in mortality (27). Whether this paradox results from lead time bias (the time of diagnosis moved forward but the date of death unaffected) or overdiagnosis (length) bias (bias toward detection of less aggressive or indolent tumors in a periodically screened sample) (28), or whether the tumors were not detected at a sufficiently early point in their development to have altered outcome is controversial.
The same debates surround the development of low-energy chest computed tomography (CT) as a screening tool for lung cancer (29). This technique is more sensitive than chest radiography and detects nodules <1 cm; however, it has relatively poor specificity. This is exemplified by the findings from a prospective study of annual screening using low-dose helical chest CT at the Mayo Clinic (30). Two years after baseline CT scanning, noncalcified nodules were detected in 69% of 1,520 current and former smokers, with the majority of tumors almost certainly benign. Given the high rate of benign nodule detection, the researchers cautioned that if the annual rate of new indeterminate pulmonary nodules seen in the study remained at 9% to 13%, then “almost all patients will have at least one false-positive CT examination result after several more years of screening.” At the 3-year point, 55 participants (3.6%) underwent 60 thoracic operations: benign disease was found in 10 patients (18%) and lung cancer was found in 45 (82%). A 5-year followup study of the Mayo Clinic cohort showed that annual screening chest CT did not lead to a difference in the percentage of patients diagnosed with stage I disease. The authors note, “CT screening for lung cancer offers the possibility of reducing mortality from lung cancer. Our preliminary results do not support this possibility and may raise concerns that false-positive results and overdiagnosis could actually result in more harm than
P.928
good.” In addition to the considerable economic burden of screening and then followup of all these patients, the interventional evaluation carried a significant morbidity: surgical complications occurred in 27% of patients, and operative mortality was 1.7% (31). Larger trials to measure the benefits of screening chest CT in at-risk patients are ongoing.
In addition to improving radiologic methods for lung cancer detection, research has focused on developing lung cancer biomarkers. Candidate abnormalities found in cigarette smokers include mutations in ras and p53 genes and DNA sequence loss at three loci: 3p14 (location of the FHIT tumor suppressor gene), 9p21 (location of the p16 tumor-suppressor gene), and 17p13 (location of the p53 tumor-suppressor gene). Differences between lung cancer patients and control subjects have also been found for the functional status of members of the family of enzymes responsible for carcinogen activation and degradation and the ability of patients’ lymphocytes to repair the genetic damage induced by the cigarette carcinogen benzo[a]pyrene (19,32). Analysis of the expression of large numbers of genes and proteins in individual tumors by DNA microarrays and advances in proteomics highlight patterns that may later prove useful for tumor detection and for tumor characterization both by type and by chemotherapeutic responsiveness. Measurement of these markers is not yet clinically useful.
The U.S. Preventive Services Task Force has extensively reviewed the current data and concluded that there is no support for lung cancer screening by any method, although the data also are insufficient to conclude that screening does not work (33).
Histology
Lung cancer is classified broadly into two major groups—NSCLC and SCLC—both of which are associated with cigarette smoking. The distribution of these cancers is 80% and 20%, respectively. In the NSCLC category, approximately 36% of patients have squamous cell carcinomas,45% have adenocarcinomas, 9% have large cell carcinomas, 2% to 4% have bronchioloalveolar carcinomas, and 1% to 2% havecarcinoids. There is significant heterogeneity in the histology: 45% of patients with NSCLC have mixed NSCLC phenotypes, 10% to 20% of NSCLC tumors have SCLC-like neuroendocrine features, and 9% of SCLC have regions of NSCLC tumor cells (34). This heterogeneity supports the hypothesis that all pulmonary cancers arise from a single pluripotent stem cell that has the capacity to differentiate into the major bronchioloalveolar mucosal cell types, including neuroendocrine, glandular, and epithelial.
The past 15 to 20 years have seen a significant shift in NSCLC subtypes in North America, but not in Europe, from a majority squamous cell to a majority that is adenocarcinoma. This has been attributed to the use of “low-tar” cigarettes in the United States and Canada, which lead to a deeper inhalation pattern that allows greater carcinogen exposure to the peripheral lung (where adenocarcinomas generally arise). Another contributor may be the increased proportion of women with lung cancer, a population that is more prone to adenocarcinomas (35,36).
The incidence of bronchioloalveolar carcinoma has doubled over the past 20 years; in one study, this tumor constituted 15% of all lung cancers (37). The reason for this increase is unclear. Patients with this cancer tend to be younger, are more likely to be female, and are less likely to be smokers than are those with other lung cancers. The treatment of bronchioloalveolar carcinoma is similar to that of other non–small cell carcinomas. However, this tumor has a greater propensity to spread locally (or to arise in multiple locations, consistent with a “field” defect) compared with other NSCLC phenotypes, and repeated surgical excisions may be necessary (38).
Malignant mesothelioma is a pleural tumor strongly associated with asbestos exposure. There is a time lag of 25 to 40 years between asbestos exposure and cancer presentation. A proposed relationship between the development of this tumor and infection with the DNA virus SV40 (simian virus) is controversial. Mesothelioma usually is heralded by a pleural effusion. The diagnostic yield of thoracentesis for mesothelioma is <40%, improved by only 10% by repeat thoracentesis and pleural needle biopsy. Thus, thoracoscopic-guided biopsy or open lung biopsy may be required for this diagnosis. A common problem with procedures used to diagnose this tumor is malignant seeding along needle and incisional tracts. A positron emission tomography (PET) scan is useful in the evaluation of tumor extent and prognosis. The histologic differentiation of malignant mesothelioma from adenocarcinoma is an occasional problem and may necessitate electron microscopy or staining for immunohistochemical markers. The treatment of mesothelioma entails debulking surgery, with either intracavitary or external radiation. Platinum-based chemotherapy in conjunction with gemcitabine and other agents may be useful. A promising new treatment for mesothelioma is pemetrexed, which targets folate metabolic enzymes to disrupt nucleotide synthesis. Still, the prognosis is poor, and the median survival time is 12 months (39,40).
History
The symptoms of lung cancer can be categorized according to those caused by the mass effect of the tumor in the airway, those caused by impingement of the tumor on extrapulmonary mediastinal structures, those of paraneoplastic syndromes, and those of distant metastases.
P.929
Pulmonary Symptoms
Ninety percent of patients with lung cancer have symptoms at the time of diagnosis, and most of these symptoms are respiratory. The respiratory manifestations include cough, dyspnea, chest pain, hemoptysis, and symptoms related to postobstructive pneumonia. Cough occurs in almost all patients during the course of lung cancer. Cough and sputum production are nonspecific symptoms in smokers, but the appearance of a chronic cough, with or without expectoration, or a change in cough pattern in an older smoker should raise a suspicion of lung cancer.
Hemoptysis (generally blood-streaked sputum) is the initial manifestation of cancer in many patients. Potential extrapulmonary sources of hemoptysis include the mouth, nasopharynx, and gastrointestinal tract. Evaluation includes history and physical examination, including oral pharyngeal and nasal examination, and a chest CT scan, preferably with contrast. Referral for bronchoscopy should be considered for patients who are older than 40 years or who have an abnormal radiograph, a history of hemoptysis for >1 week, a history of tobacco smoking, a chronic cough, anemia, or weight loss. Lung cancer is found at bronchoscopy in one third of patients with hemoptysis and one or more of these risk factors (41). Whether bronchoscopy is indicated in the absence of these risk factors is unclear. With rare exceptions, a negative bronchoscopy is reliable for excluding lung cancer in patients with hemoptysis and negative radiologic studies (42). Because a bronchoscopy is not as good for diagnosing upper airway disease as a nasopharyngoscopy and mirror examination, patients with a negative pulmonary evaluation will require examination by an otolaryngologist and possibly also by a gastroenterologist.
Chest pain is reported by approximately half of patients with lung cancer on presentation, usually as an intermittent dull ache on the side of the tumor. The cause of this pain is unclear. If the tumor appears otherwise resectable, this symptom should not preclude surgery. More severe pain may indicate metastatic disease to the parietal pleura or to bone and may necessitate a bone scan or PET scan. Shoulder pain, sometimes mistaken for arthritis, may be caused by a Pancoast (superior sulcus) tumor (see Extrapulmonary Thoracic Symptoms section) or may be caused by a tumor involving the diaphragm.
Dyspnea in a patient with lung cancer may be caused directly by the primary tumor or may result from factors indirectly related to the tumor, from treatment of the cancer, and from other concurrent medical disease. Primary tumor-related etiologies include obstruction of the airway with postobstructive atelectasis or pneumonia, pleural effusion, lymphangitic carcinomatosis, or pericardial metastasis. The tumor can impinge on nerves and lead to recurrent laryngeal or phrenic nerve palsy. Dyspnea indirectly related to the cancer may result from pulmonary embolism, an electrolyte or hormonal disorder from a paraneoplastic syndrome, or inanition. Radiation may result in pneumonitis. Chemotherapeutic agents may cause anemia or other toxicities, such as cardiomyopathy or pneumonitis, which may contribute to this symptom. Finally, because cigarettes are a common risk factor for other medical conditions, patients with lung cancer may have pre-existing lung disease, usually chronic obstructive lung disease or cardiac disease that may produce dyspnea.
Pneumonia resulting from airway obstruction by tumor may be the presenting manifestation of lung cancer. There are no formal recommendations on whether to obtain a followup chest x-ray film in a patient with pneumonia. Whether to obtain a chest x-ray film to document resolution of pneumonia several weeks after the acute illness should therefore be based on (a) whether there were findings suggestive of lung cancer on the patient's initial film; (b) the patient's risk factors for lung cancer, including age and history of cigarette smoking or other exposure; (c) the presence of pre-existing lung disease or other medical comorbidities that increase the risk of pneumonia complications; and (d) the patient's clinical course. A pneumonia that fails to resolve should prompt a consultation with a pulmonologist.
Extrapulmonary Thoracic Symptoms
Intrathoracic extrapulmonary extension may present in a variety of ways, including the superior vena cava syndrome, Pancoast tumor syndrome, Horner syndrome, dysphagia, hoarseness, and symptoms of pericardial disease. Patients with these symptoms should be rapidly evaluated for definitive diagnosis and treatment.
Thesuperior vena cava syndrome consists of edema and rubor of the upper trunk and face, sometimes with syncope. It is caused by obstruction of the superior vena cava by involvement of the mediastinum with tumor. The chest x-ray film usually shows a right upper lobe mass and a widened mediastinum. In 25% of cases, a right pleural effusion is noted.
ThePancoast syndrome is caused by a tumor that is located at the pulmonary apex and involves adjacent structures such as the chest wall, lymphatics, ribs, vertebrae, vessels, and nerves. The tumor may cause shoulder pain, arm pain, and paresthesia, usually in the ulnar distribution, indicating encroachment of the brachial plexus.
Horner syndrome results from tumor invasion of the lower cervical or upper thoracic sympathetic trunk and consists of miosis, ptosis, enophthalmos, facial flushing, and anhidrosis on the affected side.
Persistent hoarseness may be caused by involvement of the trachea with tumor or by vocal cord paralysis caused by entrapment of the recurrent laryngeal nerve, usually on the left, by a mediastinal mass. This symptom should be
P.930
evaluated with a flow–volume loop (performed in a pulmonary function laboratory) or by laryngoscopy. Additionally, contrast CT should be performed, starting at the sinuses and down through the chest, to evaluate the upper airway, mediastinum, and pulmonary parenchyma.
Wheezing usually is caused by small airway obstruction but also may be produced by upper airway or bronchial obstruction by an intrinsic or extrinsic mass. Therefore, upper airway involvement should be evaluated in all wheezing patients by auscultation over the neck. The presence of stridor or of monophonic wheezing that localizes to the neck should be emergently evaluated with flow–volume loops and/or laryngoscopy and chest x-ray and neck films.
Mediastinal tumors or enlarged lymph nodes can cause dysphagia by extrinsically impinging on the esophagus. Finally, lung cancers can directly invade, or metastasize to, the pericardium and produce tamponade.
Extrathoracic Symptoms
Anorexia, cachexia, weight loss, and fever are common in patients with lung cancer, particularly as the disease advances. Extrathoracic metastatic disease is found at autopsy in most patients with NSCLC and in almost all patients with SCLC. Metastases may be seen in any organ. Common metastatic sites for lung cancer include the pleura, bone, brain, liver, and adrenal glands. In addition, SCLC frequently invades the bone marrow, leading to hematologic abnormalities.
Paraneoplastic Syndromes
Paraneoplastic syndromes are common in patients with lung cancer. It is important to recognize that the symptoms of paraneoplastic syndromes may precede the other symptoms of lung cancer. Manifestations of these syndromes may mimic metastatic disease and mislead treatment decisions. These symptoms generally improve with successful treatment of the underlying malignancy.
Endocrine syndromes may be seen in 10% to 15% of patients. The most common is hypercalcemia secondary to a parathyroid hormone-related protein, with an N-terminus homologous with that of normal parathyroid hormone so that it binds to the parathyroid hormone receptor. This syndrome is usually caused by an NSCLC. The syndromes of inappropriate antidiuretic hormone release and of ectopic adrenocorticotropic hormone secretion are less common and are more likely to be seen in patients with SCLC. Increased levels of these ectopic hormones can be found in the blood of many more patients with SCLC than in the few percent that have clinical manifestations of the syndromes. The syndrome of inappropriate antidiuretic hormone release is associated with hyponatremia (see Chapter 81). The ectopic adrenocorticotropic hormone syndrome is characterized by mild hypertension, hyperglycemia, hypokalemia, alkalosis, and occasional hyperpigmentation but generally does not have the other manifestations of Cushing syndrome.
Clubbing andhypertrophic pulmonary osteoarthropathy are common in, although not limited to, patients with lung cancer. Clubbing is seen in one third of patients with lung cancer. Hypertrophic pulmonary osteoarthropathy with pain, swelling, tenderness, and positive bone scan may be seen in as many as 10% of these patients. Lung cancer is the underlying cause in >80% of adult cases of this osteoarthropathy. The symptoms usually respond to aspirin or other nonsteroidal anti-inflammatory medicines.
Neurologic paraneoplastic processes are rare but dramatic and may be the presenting symptoms of SCLC and occasionally of NSCLC (43). PET scanning may be useful for diagnosis of occult malignancy, whereas magnetic resonance imaging (MRI) seems to have limited value (44). In the Eaton-Lambert syndrome, proximal muscle weakness and paresthesias may mimic myasthenia gravis, from which it can be distinguished by electromyography. The diagnosis of other neurologic paraneoplastic syndromes, including peripheral neuropathies, subacute cerebellar degeneration, cortical degeneration, polymyositis, and intestinal dysmotility, may be facilitated by measurement of type I antineuronal nuclear antibody (ANNA-1/Hu-1) (43). Treatment of the tumor may result in improvement of these symptoms.
Hematologic abnormalities may occur in patients with lung cancer and with mesotheliomas. An increased tendency to clot may be manifest as either typical or atypical (Trousseau syndrome) venous or arterial thromboembolic disease (see Chapter 57). Anemia and thrombocytosis are common.
Physical Examination
The purpose of a careful physical examination of patients with lung cancer is to determine the presence of metastatic disease, to evaluate the patient's candidacy for therapy, and to assess the patient's functional capacity, which is a major contributor to prognosis. Constitutional findings are important; weight loss may be evidence of advanced disease. Attention must be paid to areas of pain or tenderness, which may indicate bony metastasis. The examination should include an evaluation of lymph nodes in the anterior and posterior cervical, supraclavicular, and axillary regions. Suspicious lymph nodes are those >1 cm in diameter, hard in consistency, or fixed. Hoarseness or signs of superior vena cava syndrome or Horner syndrome should be noted. The pulmonary examination may reveal evidence of chronic lung disease and local wheezing or bronchial breath sounds from bronchial obstruction or basilar dullness from a pleural effusion. D’Éspene sign, the inappropriate transmission of whispered pectoriloquy
P.931
below the level of the T2 vertebral body, suggests mediastinal involvement with tumor. The cardiac examination, including evaluation of the jugular veins, is important in preoperative assessment and when considering the possibility of a malignant pericardial effusion. Hepatomegaly (liver span >13 cm) may suggest hepatic metastases. The extremities may demonstrate clubbing, edema, or cyanosis. A careful neurologic examination is important to look for evidence of brain metastases.
Diagnostic Procedures
Chest X-Ray Film
A review of old chest x-ray films can demonstrate the rapidity of disease onset and may aid in prognostication. A mass with a doubling time of <2 weeks or >450 days is unlikely to be malignant. (For a more extensive discussion of radiologic characteristics helpful in differentiating benign from malignant pulmonary lesions, see Solitary Pulmonary Nodule.)
Certain chest x-ray patterns are suggestive of particular types of lung cancer. Squamous cell and small cell carcinomas tend to be centrally located. Adenocarcinomas and large cell carcinomas tend to arise peripherally. Centrally located tumors are more likely to present with symptoms of obstruction, such as atelectasis, pneumonia, and dyspnea. SCLC may demonstrate only hilar adenopathy with no visible primary tumor on the initial chest x-ray film. Squamous cell carcinoma is the most likely bronchogenic carcinoma to cavitate, although even this tumor cavitates uncommonly.
Bronchioloalveolar cell carcinoma usually presents peripherally and tends to be multifocal. This cancer has a variety of radiologic manifestations and may appear as a single mass or as multiple nodules, or it may imitate a pneumonia. Malignant mesothelioma is associated with a pleural effusion, and the underlying pleural surface is thickened and lumpy.
Computed Tomography and Magnetic Resonance Imaging of the Chest
A contrast CT of the chest is routine in the evaluation of patients with suspected lung cancer to determine the anatomic extent of the tumor and to guide diagnostic and therapeutic procedures (reviewed in ref. 45). CT may demonstrate lesions that are unseen on chest x-ray film. Furthermore, CT may aid in characterizing the size, shape, and composition of these lesions. For example, CT is better than chest x-ray film in determining calcification of a solitary pulmonary nodule (SPN); a densely calcified nodule is more likely to be benign. CT may demonstrate fat in a nodule, highly suggestive of a hamartoma. In addition, the presence and size of mediastinal or hilar lymph nodes can be evaluated by CT. In patients with lung cancer, nodes >1 cm in diameter have a 60% chance of representing metastatic disease (46). If a surgical cure is contemplated, sampling by transbronchial needle aspiration (TBNA), mediastinoscopy, or mediastinotomy is required (see these specific subsections to follow). Finally, extending CT cuts down through the upper abdomen may demonstrate hepatic or adrenal metastases, although these radiographic findings will require biopsy if therapy would be influenced by the histology of these lesions. CT is neither sensitive nor specific in detecting the invasion of chest wall or mediastinal structures.
MRI is superior to CT for evaluation of possible chest wall or vascular invasion by tumor and for examination of superior sulcus tumors. MRI is more sensitive than contrast CT for diagnosing brain metastases. MRI is not as informative as CT for evaluating pulmonary parenchymal disease, nor is it helpful for evaluating mediastinal lymph nodes.
Positron Emission Tomography
PET, using the glucose analogue fluorodeoxyglucose (FDG), aids in the evaluation of some patients with SPN and in the staging of lung cancer (see Staging). FDG is taken up more avidly by more metabolically active cells and is not easily eliminated. Malignant tissue is enhanced by this method, whereas normal tissue is not. For focal pulmonary densities of all sizes, the sensitivity of PET is 97% and the specificity is 78% (47). Infections, other inflammatory lesions, and granulomatous disease such as sarcoidosis can demonstrate high FDG uptake, leading to false-positive results. Some tumors are falsely identified as benign by this study, especially bronchioloalveolar carcinomas and some adenocarcinomas (see Solitary Pulmonary Nodule). Finally, the PET scan is insensitive for densities <1 cm and in the presence of a high serum glucose level.
Sputum Cytology
The frequency of diagnosis of lung cancer from cytology submitted from spontaneously expectorated sputum depends on the cell type and the tumor location (48). Squamous cell tumors shed cells into the airways, and cytologies are positive approximately 80% of the time. However, patients with a peripheral adenocarcinoma have positive cytologies <5% of the time. Patients should be instructed to produce a forceful cough. Three early morning specimens should be collected in a tightly fitting container. If the specimens cannot be submitted to the laboratory within 2 to 3 hours of collection, a fixative should be added, in which case the specimens can be pooled and submitted together. Saccomanno solution, which contains alcohol and Carbowax, is one of the best fixatives. Submitting more
P.932
than three samples does not increase the likelihood of a positive diagnosis of cancer. If the patient is not producing sputum, induction by inhalation of an aerosolized solution of saline can be helpful. The patient inhales normal saline or Hanks balanced salt solution that is aerosolized by an ultrasonic nebulizer (DeVilbiss 3583). In general, sputum induction is not a routine office procedure and should be done by experienced personnel. In addition to the sputum collected immediately after induction, good material is produced the following morning. The addition of chest vibration or percussion does not improve the yield from sputum cytology. In cases with obstruction of the bronchus as evidenced by the physical examination or chest x-ray film, it is reasonable to start collecting sputum after therapy, which includes antibiotics and perhaps bronchodilators and may reestablish airway patency and allow sputum production.
Bronchoscopy
Bronchoscopy is used both to diagnose and to stage bronchogenic carcinoma (49). Bronchoscopy is important to search for a second synchronous malignancy and to evaluate vocal cord function. A fixed vocal cord suggests involvement of the recurrent laryngeal nerve in the mediastinum. Furthermore, the bronchoscope visualizes the proximal extent of the tumor, which is helpful in planning surgery (tracheal “sleeve” resections allow resection of tumors close to the carina). Transbronchial biopsy using forceps, brushing for cytology, and bronchoalveolar lavage all can be performed through the bronchoscope. In addition, transbronchial needle aspiration (TBNA) allows sampling of hilar, subcarinal, and other mediastinal lymph nodes. The diagnostic yield for sampling tumors outside of the airway is improved with endoscopic guidance, which is not available in all centers. Fiberoptic bronchoscopy is an outpatient procedure that causes minimal discomfort with use of mild sedation (e.g., midazolam or propofol), a short-acting narcotic (fentanyl), sometimes a drying and vagolytic agent (usually atropine or glycopyrrolate), and topical lidocaine. Occasional circumstances dictate preoperative endotracheal intubation or use of a rigid bronchoscope; these procedures are done with general anesthesia.
The diagnostic yield of bronchoscopy is 60% to 80% for tumors >2 cm in diameter and 20% to 40% for those <2 cm, and it is higher for visualized masses in the central airways. Necrotic tumors, submucosal carcinomas, and large tumors that displace feeding bronchi may be more of a diagnostic challenge and may require more than one bronchoscopic procedure to obtain diagnostic material. Many peripheral tumors are accessible by bronchoscopy using fluoroscopically guided needles, brushes, and biopsy forceps. However, masses <2 cm or in the outer third of the lung usually are better approached from the outside by percutaneous transthoracic needle aspiration (PTNA).
Although bronchoscopy is generally a safe and well-tolerated procedure, potential complications include pulmonary hemorrhage, pneumothorax, laryngospasm or bronchospasm, hypoxemia, and transient cardiac arrhythmias. Patients with primary or secondary coagulation disorders, thrombocytopenia, uremia, pulmonary hypertension, or the superior vena cava syndrome are at increased risk for bleeding. Patients receive supplemental oxygen during the procedure and shortly thereafter, and they are monitored with continuous oximetry. Pneumothorax complicates approximately 5% of transbronchial forceps biopsies and <1% of TBNA procedures, although only half of patients who develop a pneumothorax require a chest tube. Transient atrial and ventricular arrhythmias occur in approximately 5% of elderly patients and may be related to transient hypoxemia or induction of the vagal response by passage of the bronchoscope through the upper airway. However, death is rare and almost always is associated with a transbronchial biopsy that caused significant bleeding (50).
Patients may have a low-grade fever after bronchoscopy, but the occurrence of pneumonia is low. Although the incidence of bacteremia during bronchoscopy is <2%, some advocate prophylactic antibiotics for patients at risk for bacterial endocarditis (50).
The interval between a myocardial infarction and safe performance of fiberoptic bronchoscopy is unknown and must be established on an individual basis, balancing the goals of the procedure and the therapeutic implication of the results.
Patient Experience
Bronchoscopy can be performed on an ambulatory basis by a pulmonologist or a thoracic surgeon. The patient is told to fast (including liquids) for at least 8 hours before bronchoscopy (although generally medicines can be taken with a sip of water) and to not eat for approximately 1 to 2 hours after completion of the procedure until the effects of topical anesthesia wear off. The only discomfort the patient will experience is coughing caused by irritation of the trachea and main bronchi, which is treated with topical 1% to 2% lidocaine. There usually is no pain associated with this procedure. A low-grade fever (<100.5°F oral) or blood-streaked hemoptysis may be present during the first 24 hours after bronchoscopy; however, the patient should call the pulmonologist or go to the emergency room for higher fever, hemoptysis that is more than blood streaking, chest pain, severe shoulder pain (which may indicate a pneumothorax), or exacerbation of shortness of breath.
In the unusual situation where sputum cytology demonstrates malignancy but the chest x-ray film and chest CT scan do not localize a suspicious area (“occult carcinoma”), a long bronchoscopic procedure is undertaken under general anesthesia. In the procedure, a meticulous upper airway evaluation is followed by sequential
P.933
sampling of each pulmonary segment. Positive findings mandate a confirmatory procedure in which repeat sampling is done of subsegments corresponding to the positive material. The yield of this procedure is 13% to 35%, and up to 15% of these patients have multicentric carcinomas. New metachronous primary lung cancer develops in patients with “occult carcinoma” at the rate of 5% per year. Almost all these patients are heavy smokers and have increased surgical morbidity and mortality; thus, lung-sparing surgery or photodynamic therapies may be indicated (51).
Percutaneous Transthoracic Needle Aspiration
CT-directed PTNA of the lung is most useful when a mass or a nodule is located peripherally near the pleura or in the apex of the lung. It is especially helpful in the diagnosis of metastatic carcinoma because these tumors arise outside of the airway and are difficult to diagnose by bronchoscopy, particularly when they are <2 cm in size. It also may be preferred when the differential diagnosis includes infection because interpretation of the significance of an infectious agent obtained by bronchoscopy is complicated by passage of the scope through the nonsterile nasopharynx or oropharynx. PTNA is performed by either a radiologist or a pulmonologist with fluoroscopic or CT guidance. The technique and experience of the operator and the cytopathologist are of paramount importance in the success of the procedure, and the yield is increased with a repeat procedure. However, the false-negative rate for any fluoroscopically guided procedure is significant. Thus, nondiagnostic or nonspecific findings require followup, the nature of which is determined by the adequacy of the sample and the clinical scenario. PTNA is usually an outpatient procedure, done with local anesthetic, and requires a cooperative patient (52).
Contraindications to the procedure include large blebs or blood vessels that are in the direct path of the needle; a patient with an uncontrollable cough, in which case general anesthesia may be necessary; and contralateral pneumonectomy, in which case the production of a pneumothorax would be devastating. In addition, patients with pulmonary hypertension have an increased risk for bleeding.
Patient Experience
Patients will receive intravenous sedation and local anesthetic. The patient will feel a mild pressure sensation with introduction of the needle. Otherwise, there is no significant pain. Potential complications include a 5% risk of bleeding (usually of minimal amount) and a 10% to 15% risk of pneumothorax (half of these cases require chest tube insertion). Most patients can resume normal activity within 24 hours.
Mediastinoscopy, Mediastinotomy, and Thoracoscopy
Mediastinoscopy and mediastinotomy are indicated when enlarged mediastinal lymph nodes or masses cannot be adequately sampled by less invasive techniques (TBNA or PTNA; see specific sections above). Lymph nodes accessible to cervical mediastinoscopy are those located in the pretracheal area from the thoracic inlet to 1 cm beyond the carina bilaterally. Those anterior to the aortic arch and in the aortic–pulmonary window are not accessible to the cervical mediastinoscope and require anterior mediastinotomy. Potential complications include wound infection, injury to major vascular structures, and recurrent laryngeal nerve paralysis (53). These procedures are done by thoracic surgeons and require general anesthesia and hospitalization.
A biopsy done through a thoracoscope may be better tolerated than one done during a limited thoracotomy. Thoracoscopy may also be helpful, in selected cases, in the removal of peripheral lung masses (54) and in diagnosing the cause of pleural effusions. A medical thoracoscopy can be done under general anesthesia or in the patient sedated with local anesthetic. The visceral and parietal pleuras are examined with either a rigid or fiberoptic endoscope inserted through a chest tube.
A video-assisted thoracic surgical procedure usually requires general anesthesia. The procedure involves inducing a controlled pneumothorax and single lung ventilation, followed by insertion of the instruments through multiple small incisions. The patient generally has 2 to 3 days of hospital observation with a chest tube in place after thoracoscopy. The incidence of complications is similar to that of the closed procedures (55).
Staging
The staging of the patient with lung cancer has two goals: to determine the anatomic extent of the tumor and to determine the physiologic capacity of the patient to undergo therapy.
Non–Small Cell Lung Cancer
For NSCLC, the anatomic extent of the tumor is defined by the TNM classification system, where T is tumor size, N is nodal involvement, and M is distant metastasis. The TNM categories then are grouped into four stages with therapeutic and prognostic implications (Tables 61.2,61.3, and 61.4).
Mediastinal nodal involvement can be assessed by chest CT, with sampling of lymph nodes >1 cm in transverse diameter, by bronchoscopy (TBNA), or by mediastinoscopy or surgery. PET is complementary to CT for this examination. For detection of mediastinal node metastases,
P.934
P.935
the sensitivity and specificity of PET scans are 79% and 91%, respectively, and of CT scans are 60% and 77%, respectively (56,57). However, unlike CT, PET scans cannot adequately distinguish between intrapulmonary lymph nodes and those in the mediastinum, nor can PET differentiate hyperplastic from neoplastic disease, which may require histologic sampling of positive areas.
|
TABLE 61.2 Staging of Non–Small Cell Lung Cancer: TNM Definitions |
||||||||||||||||||||||||||||||||||||||||||||||
|
||||||||||||||||||||||||||||||||||||||||||||||
|
TABLE 61.3 Staging of Non–Small Cell Lung Cancer: New International Revised Stage Grouping |
||||||||||||||||||||||||||||||||||||||||||||||
|
||||||||||||||||||||||||||||||||||||||||||||||
|
TABLE 61.4 Five-Year Survival of Non–Small Cell Lung Cancer by Stage at Presentation |
||||||||||||||||||||||||||||||||||||||||||||||||
|
||||||||||||||||||||||||||||||||||||||||||||||||
The major sites of NSCLC metastases are brain, bones, liver, and adrenal glands. The extent of routine pretreatment evaluation of these sites has been controversial. However, two large meta-analyses have demonstrated that the negative predictive value of a good clinical evaluation for metastatic disease is >97% (58,59). This finding led the American Thoracic Society/European Respiratory Society (60) to recommend the NSCLC staging evaluation (Table 61.5). These recommendations may require modification for use in the individual patient. For example, constitutional signs or symptoms of advanced disease or evidence of borderline operability by physiologic criteria may necessitate a more aggressive staging evaluation.
Whole-body PET helps in the preoperative staging of patients with potentially resectable NSCLC. PET has good sensitivity and specificity for detection of distant metastases (56,57). In one prospective study of 102 patients, PET resulted in a change in tumor stage in 62 patients (56). However, because of a 17% false-positive rate for distant metastases, these areas need to be sampled. In addition, PET cannot visualize tumors in the brain. A combined PET–CT scan improves the diagnostic accuracy for diagnosing and staging malignancies (61).
MRI is useful for specific indications, such as examination of paravertebral tumors, or when imaging of vascular and neural structures in the mediastinum or thoracic inlet is required. Gadolinium-enhanced brain MRI scans are superior to CT for detection of brain metastases and should be considered for patients with stage III or IV NSCLC.
Both patients with NSCLC and those with SCLC should undergo evaluation of their performance status. The performance status determines the ability to undergo therapy and is a major contributor to the prognosis (62). Common scales for this measurement are the Eastern Cooperative Oncology Group (ECOG) and Karnofsky performance scales, which measure the patient's physiologic status as a result of the cancer and concurrent medical problems (Table 61.6).
Small Cell Lung Cancer
SCLC is classified by a two-stage system. In limited-stage disease, the cancer is confined to the hemithorax and regional lymph nodes (including mediastinal, ipsilateral hilar, and supraclavicular nodes), which essentially delineate the extent of disease that can be contained within a tolerable radiation port. In extensive-stage disease, the cancer lies outside these boundaries. Rarely SCLC presents as a localized nodule, which is the only time curative surgery is possible. SCLC is a more aggressive tumor than NSCLC and metastasizes early and widely. Therefore, the staging on presentation is more extensive than with NSCLC. The staging of SCLC patients generally entails the same recommendations as in Table 61.5, part A, as well as additional studies. Because one third of these patients have bone involvement at presentation, a bone scan or PET scan is indicated. One fourth of patients have liver metastases. Because liver enzymes are not sensitive for metastases, a CT through the liver is necessary. The central nervous system is involved at presentation in one fourth, and eventually in half, of the patients with SCLC. Symptoms are absent in 10% of those with central nervous system
P.936
involvement. A cranial MRI with gadolinium is superior to head CT for demonstration of metastases. A bone marrow aspirate and biopsy are no longer routine. Performance status must be evaluated, as for NSCLC (Table 61.6).
|
TABLE 61.5 Pretreatment Evaluation of Patients with Lung Cancer |
||||||||||||||||||||||||||||||||||||||||
|
||||||||||||||||||||||||||||||||||||||||
|
TABLE 61.6 Performance Status Scales |
||||||||||||||||||||||||||||||||||||||||
|
||||||||||||||||||||||||||||||||||||||||
Treatment
Lung cancer therapy has been evolving. Treatment options may be complicated and often require the input of a pulmonologist, oncologist, radiotherapist, and thoracic surgeon. Current therapeutic trends include (a) optimizing the current therapies for benefit and risks; (b) introducing new chemotherapy agents, notably those resulting from improved understanding of cancer biology; (c) targeting therapy to the tumor to spare normal tissue and to be patient specific; and (d) improving supportive care.
The indication for chemotherapy has expanded to include all stages of lung cancer. Alternative radiation strategies are being investigated. Multimodality therapy, including concurrent radiation and chemotherapy, is increasingly being used. Adjuvant therapy is the addition of chemotherapy or radiotherapy to surgical treatment. Studies suggest that adjuvant therapy may have a role for some patients with complete surgical resection of NSCLC. Neoadjuvant therapy is the use of chemotherapy or radiotherapy prior to surgery, to shrink the tumor and allow for more complete resection.
The current standard treatments for NSCLC and SCLC are reviewed by Spira and Ettinger (63) and outlined in Table 61.7.
Non–Small Cell Lung Cancer
The most important prognostic criterion for NSCLC is the TNM classification because it defines resectable, and thus most of the potentially curable, disease. Surgery is the standard treatment for patients without clinical evidence of mediastinal or metastatic disease and is an option for selected patients with tumor involvement of the mediastinum or chest wall. Chemotherapy and radiotherapy improve survival of most stages (compare the 5-year survival outlined in Table 61.4 with that in Table 61.7).
Stages I, II, and IIIA
Surgical resection results in a 5-year survival of approximately 70% and 50% for patients with clinical stage I and stage II NSCLC, respectively, and is appropriate for selected patients with stage IIIA disease for whom 5-year survival is 15% by presurgical staging (although better for those defined by surgical staging). These results indicate significant mortality even in patients with disease that appears to be local at presentation. Postoperative radiation therapy improves local control and may have a small effect on survival for stage II and IIIA patients. Adjuvant or neoadjuvant platinum-based chemotherapy (e.g., cisplatin or carboplatin), generally in combination with other chemotherapeutic agents and external radiation, improves survival of patients with stage IIIA disease. Cisplatin-based chemotherapy has a small impact on survival of patients with stage I and II disease but is associated with considerable toxicity (64). Recommendations may be changed by newer therapies currently being evaluated, such as the well-tolerated oral agent uracil–tegafur (65,66).
|
TABLE 61.7 General Approach to the Treatment of Lung Cancer According to Stagea |
||||||||||||||||||||||||||||||||||||||||||||||||||||
|
||||||||||||||||||||||||||||||||||||||||||||||||||||
P.937
Stage IIIB and IV
The 5-year survival rate of patients with stage IIIB disease is approximately 5% and with stage IV disease is negligible. Median survival times with best available therapies are 8 to 10 months. For these patients, performance status is the key to determining further therapy. For stage IIIB patients, chemotherapy with a platinum-based regimen and definitive local radiation therapy is indicated for patients with good performance status. Treatment should be started soon after diagnosis because a delay in therapy may negate any survival benefit. For stage IV patients, chemotherapy is appropriate for patients with good performance status; local radiotherapy should be used for palliation (67). However, the subgroup of patients with solitary brain metastases and otherwise resectable lung cancer (stage I or II) are treated with combined thoracotomy and craniotomy, possibly in conjunction with chemotherapy (68). Approximately 20% of patients with advanced NSCLC benefit from treatment with the oral epidermal growth factor receptor tyrosine kinase inhibitors gefitinib and erlotinib. Responders are more likely to be female, nonsmokers, and patients with adenocarcinoma (69). For Pancoast tumor, an aggressive, combined modality approach results in a 5-year survival of 41% (70). Use of this targeted therapy is discussed later in this chapter.
The most common complications of chemotherapy are nausea and vomiting, myelosuppression, nephrotoxicity, electrolyte disturbances, ototoxicity, and neuropathy (see Chapter 10). The epidermal growth factor receptor inhibitors may cause a rash.
Surgical Considerations
Patients with NSCLC tumors that appear anatomically localized after staging evaluation are considered surgical candidates if the risks of the thoracotomy and of removal of functional lung tissue are considered medically reasonable. The prediction of residual lung function after resection is based on pulmonary function studies. If radiologic studies show an irregular distribution of underlying lung disease, a nuclear perfusion study may be needed to determine the contribution of the proposed lung resection to the total lung function. The predicted postoperative forced
P.938
expiratory volume in 1 second (FEV1) is calculated from the results of the nuclear study. All patients should be evaluated as potential candidates for both pneumonectomy and lobectomy if operative findings dictate that the former will be necessary for cure. A pneumonectomy is feasible if the preoperative FEV1 is ≥2 L or 80% of predicted, if the maximum voluntary ventilation is ≥50% of predicted, or if the maximal oxygen consumption on exercise testing is >20 mL/kg/min. Alternatively, a lobectomy or pneumonectomy is feasible if the predicted postoperative FEV1 is ≥0.8 L or 40% of predicted. For borderline cases or when the degree of dyspnea or disability is greater than the FEV1 would predict, a carbon monoxide diffusion capacity (DCo) ≥40% of predicted or a change of oxygen saturation of ≤2% with exercise suggests that a patient is operable. Many surgeons also rely on the patient's exertional tolerance. Good tolerance as demonstrated by the ability to walk a flight of stairs or by excellent baseline activity level predicts adequate functional reserve to tolerate pulmonary resection. Patients who do not meet these criteria or who have hypoxemia or hypercarbia on baseline arterial blood gas have a high risk for respiratory complications after thoracotomy. Additional considerations, such as cardiovascular status, other comorbid disease, and the informed patient's wishes, must be considered in the surgical decision (71). Video-assisted thoracoscopic lung resection is an alternative to open thoracotomy for selected patients.
Possible postoperative pulmonary complications of patients undergoing thoracotomy include the need for prolonged mechanical ventilatory support or the development of bronchospasm, atelectasis, bronchitis, pneumonia, or prolonged leakage of air from the chest tube. Several preoperative measures decrease the risk for these complications in patients with chronic obstructive lung disease. These measures include sustained cessation of smoking, use of inhaled bronchodilators and inhaled steroids (for asthmatics), treatment of pulmonary infection with antibiotics, and treatment of increased secretions with measures such as chest physiotherapy, postural drainage, and/or use of a flutter valve or Acapella device. Finally, all patients should be taught how to use the incentive spirometer.
Thoracotomy results in a temporary decrease in vital capacity of 30% (in part due to pain and atelectasis) and an increased upper airway bacterial load. Thus, important postoperative measures include using adequate, but not excessive, pain control and encouraging the patient to cough frequently and to use the incentive spirometer. Both cancer and chronic obstructive lung disease are independent risk factors for deep venous thrombosis and pulmonary embolism. Therefore, as with all surgical patients, early ambulation, low-dose unfractionated or low-molecular-weight heparin, and lower-extremity sequential compression devices are essential preventive measures. Normal activities can be resumed within several weeks of surgery, depending on the pulmonary reserve.
Radiotherapy
Definitive radiotherapy is not as effective as surgery, but it offers a potential for cure or for prolonged survival when surgery is not an option. Although compromised lung function may limit the ability to deliver radiation, definitive external radiotherapy may cure 15% to 20% of patients with stage I or II disease who have contraindications to surgery (72). As discussed earlier, radiotherapy has a role in stage IIIA, IIIB, and IV disease and, possibly as part of a multimodal approach, in stage I and II disease. Alternative radiation schemes, such as increasing total radiation dose and hyperfractionation (twice-daily therapy), are being investigated in an attempt to improve tumor response and decrease complications.
The complications of external radiation include pneumonitis, pulmonary fibrosis, and dysphagia from transient esophagitis. The rate of these complications depends on the time and dosage of radiation delivery and the volume of normal tissue exposed. However, complication rates have fallen in recent years as techniques have improved. Minimizing pulmonary risk in a patient with underlying lung disease requires individualization of radiation protocols by the radiation oncologist, taking into account the extent of the pulmonary disease and the anatomy of the tumor.
Palliative Therapies
For patients who are not candidates for other therapies, local symptom palliation may be accomplished by external-beam radiation therapy. Palliative radiotherapy is delivered over shorter intervals and in lower dosages than those used for definitive therapy. It relieves hemoptysis, superior vena cava syndrome, and dyspnea in 60% to 80% of patients. However, vocal cord paralysis rarely responds. Although <20% of NSCLC patients with atelectasis respond to radiotherapy, collapsed lobes or lungs re-expand in 57% of those with SCLC (see Small Cell Lung Cancer subsection) (73).
Bronchoscopic mechanical debulking, laser therapy, brachytherapy, and bronchoscopic placement of prosthetic stents are other options for symptomatic relief of major airway obstruction (51). These procedures may require hospitalization and must be performed by physicians with experience and expertise in their use. The neodymium: yttrium–aluminum–garnet laser is useful for debulking obstructing endobronchial lesions. Symptomatic relief is immediate and dramatic but may last only 1 to 3 months in two thirds of patients because of tumor regrowth. Thus, repeated procedures may be required, with increasing procedure-related morbidity and mortality.
Endoscopically placed brachytherapy with 192Iridium is used to reduce either endobronchial or extraluminal compressing masses. This therapy relieves symptoms
P.939
more rapidly than external radiation, although not as rapidly as laser therapy, and has a longer duration of response with fewer complications compared with laser. Brachytherapy does not cause the regional pneumonitis that limits the use of external radiation. The risks of endobronchial brachytherapy and laser therapy are bronchial perforation and hemorrhage with asphyxiation. Local edema with critical airway occlusion can complicate laser therapy, and fistulas may form after brachytherapy.
Brachytherapy can be delivered intraoperatively by permanent implantation of 125Iodine or 103Palladium seeds. Chest wall lesions may be treated with temporary 192Iridium seeds. This type of brachytherapy is particularly helpful for superior sulcus tumors. The roles of laser or brachytherapy as definitive therapies for bronchogenic carcinoma remain to be defined (74).
Role of Consultants
Given the therapeutic considerations, including the increased role of chemotherapy at all stages, all patients with lung cancer should be referred to a medical oncologist. Stages I, II, and IIIA tumors are potentially resectable, and patients with tumors in these categories should be referred for surgical evaluation, depending on comorbid medical conditions and pulmonary function, as outlined above. Patients in these stages who are not surgical candidates should also be referred to a radiation oncologist for possible external radiation therapy.
Patients with bone pain, pulmonary symptoms resulting from airway obstruction, or brain metastases should be referred to a radiation oncologist for palliative therapy. As noted above, some patients with solitary brain metastases and stage I or II NSCLC may be candidates for neurosurgical intervention (75).
Hospice care (see Chapter 13) is a consideration for patients with stage IV tumors and significant comorbid conditions, those with poor performance status, or those who have not responded to repeated therapies.
Small Cell Lung Cancer
Almost all patients with SCLC have either clinically evident or subclinical metastases at presentation. Age and performance status are prognostic factors and influence treatment selection. Except for the rare isolated peripheral SCLC, surgery generally is not an option for curative treatment. With only supportive care, the median survival of patients with limited disease is 3 months and with extensive disease is 1.5 months. A single chemotherapeutic agent doubles the median survival; combination chemotherapy, and, depending on stage and symptoms, radiotherapy improve survival further (Table 61.7). Therefore, all patients with SCLC should be referred to a medical oncologist; this consultant usually involves a radiation oncologist as well.
For limited disease, chemotherapy is combined with concurrent or early radiotherapy. Prophylactic cranial irradiation decreases brain metastases, may increase survival, and is a consideration for those who achieve complete remission (63). The current favored chemotherapy is a platinum compound and etoposide, although gemcitabine and vinorelbine may be better tolerated by the elderly and debilitated. Treatment is given every 3 to 4 weeks for four to six cycles. The response rate for patients with limited disease is 80% to 90%, with a complete response rate of 50% to 60% and a median survival of 12 to 16 months. The major toxicities of the chemotherapeutic agents used for SCLC are nausea and vomiting, myelosuppression, neurotoxicity, mucositis, and diarrhea (see Chapter 10 for other common adverse effects of chemotherapy).
Among patients with extensive disease, there are essentially no 5-year survivors, but chemotherapy can improve the quality and length of life. Chemotherapeutic agents are generally used in combination and can be modified for patients with poor functional status. Irinotecan, a topoisomerase inhibitor, is an agent that improves survival of patients with extensive SCLC, when used in combination with cisplatin (76). Overall response of patients with extensive SCLC to chemotherapy is 60% to 80%, with a complete response of only 20% to 25% (77). Routine chest and cranial radiotherapy has not been shown to improve survival for those with extensive disease, but radiation may offer palliation when delivered to symptomatic sites.
Most patients with SCLC relapse within 1 year of chemotherapy. The probability of response to a second course of therapy is less than with the first course and correlates with the duration and completeness of the first remission.
Adjuvant Therapies for Cancer-Associated Symptoms
Treatment with recombinant erythropoietin (10,000 U subcutaneous injection three times per week or 40,000 U weekly) or darbepoetin (200 µg subcutaneous injection every 2 weeks) to maintain a hemoglobin ≥12.0 g/dL improves fatigue and quality of life and may improve lung cancer survival (78). For those with dyspnea, treatment consists of confronting the underlying cause(s), and administration of narcotics. Weight loss is addressed with dietary supplementation. Megestrol acetate, 800 mg elixir daily, may be useful for treatment of anorexia. Pain, depression, nausea, and constipation are common symptoms and amenable to medical management. Treatment of infections in the setting of neutropenia may be aided by administration of a recombinant human granulocyte colony-stimulating factor (filgrastim or long-acting pegylated filgrastim), usually by the patient's oncologist.
P.940
New Chemotherapeutic Agents
New chemotherapies have derived from laboratory findings that continue to elucidate lung cancer biology. Lung cancer results from the dysregulation of nuclear and signaling processes that control cell cycling and death and allow unrestricted cell growth, invasion, metastasis, and development of treatment resistance. Dysregulated nuclear proteins that contribute to cancer cell mitosis and abrogate the appropriate apoptotic (programmed cell death) pathway include Myc and p53 (in both NSCLC and SCLC), Rb (in SCLC), and other proteins in these pathways. Bcl-2/bax is an inhibitor of apoptosis that is overexpressed in many SCLC and NSCLC tumors. Dysfunctional signaling pathways and proteins that promote carcinogenesis include Ras (in NSCLC), protein kinase C (in SCLC), and two tyrosine kinase growth factor pathways: (a) epidermal growth factor receptor (EGFR) and its ligands EGF and transforming growth factor (TGF)-α (in NSCLC), and (b) c-Kit and its ligand stem cell factor (in SCLC). The attraction of blood vessels is necessary for tumors to grow beyond a small size, and vascular endothelial growth factor (VEGF) receptor and its ligands and other vascular-specific integrins mediate this step. Many trials for chemotherapies that act on these targets are ongoing.
New-generation chemotherapies are increasingly able to decrease toxicity by improving tumor cell specificity. This effort is dependent on the new field of pharmacogenetics. For example, gefitinib and erlotinib interfere with the epithelial-specific, epidermal growth factor receptor (EGFR/erbB)-tyrosine kinase signal, and improve survival and symptoms for patients with NSCLC (79). The response to these drugs will be predictable for each patient, as it is determined by the genetic profile of the EGFR kinase domain in the cells of each tumor (80). Research showed that the eventual loss of responsiveness to gefitinib and relapse are caused by a point mutation in the EGFR that interferes with binding of gefitinib, but which may be overcome by using EGFR-inhibiting drugs with different targets (81).
Pharmacogenetics may provide information on use of the chemotherapeutic agent irinotecan, a topoisomerase I inhibitor that improves outcome for patients with extensive SCLC. Irinotecan is a prodrug whose activation and elimination are affected by polymorphisms of hepatic enzymes that affect drug efficacy and toxicity. Thus, selecting patients and/or modifying drug protocol based on determination of these polymorphisms in each patient may improve the benefit–toxicity ratio (82). Finally, profiling each patient's tumor genetics with DNA microarrays may eventually help personalize lung cancer treatment approaches.
Besides improving the outcome for chemoresponsive tumors, newer agents are extending the survival of patients with tumors that previously were less responsive to chemotherapy. Pemetrexed may improve the outcome for patients with both advanced NSCLC and mesothelioma (83,40). Gefitinib and erlotinib are efficacious for bronchoalveolar carcinoma.
Followup of Patients Surviving Lung Cancer
Lung cancer is most likely to recur within the first 5 years, although 9% of patients with resected stage I disease have recurrences >5 years after treatment. One third of first reappearances are local. Although lung cancers can metastasize anywhere, the most common sites are regional lymph nodes, brain, lung, liver, and bone (84).
Patients who have had a lung cancer have significant risk for a second primary cancer. In one large study of patients with resected stage I NSCLC, one third had second synchronous or metachronous cancers discovered during 10 years of followup; one third were second primary lung cancers (85). Chemotherapy increases the incidence of leukemia and myelodysplastic syndromes. Cranial irradiation may produce central nervous system toxicity, and symptoms may not appear for several years.
Followup at least every 3 to 6 months for the first few years after resection of a lung cancer, and at least yearly thereafter, is advised to detect metastases or local recurrence and second primary malignancies. Patients with lung cancers may be a group for which serial low-energy CT and/or PET may be useful screening maneuvers, although this has not been studied. Close attention to standard recommendations for early detection of extrapulmonary cancers, such as yearly history and physical examination, monitoring for occult fecal blood, and routine mammography, are advisable. Symptoms of bone pain may be assessed with bone scans and x-ray films. Neurologic signs or symptoms may be evaluated by head CT or gadolinium-enhanced MRI. Chest symptoms can be evaluated by x-ray film, CT, and/or bronchoscopy. Abdominal CT is more sensitive than liver enzymes for evaluation of patients with signs or symptoms of hepatic metastases.
Chemoprophylaxis
The goal of chemoprophylaxis is to both inhibit and reverse pulmonary carcinogenesis. Pulmonary carcinogenesis is characterized by progressive accumulation of genetic lesions and pathologic dysplasia. A “field defect” involving the interaction of genetic predisposition with environmental triggers likely is involved.
Primary chemoprevention means preventing cancers in patients at risk (i.e., smokers). Secondary chemoprevention means preventing cancers in smokers who already have histologic changes (squamous metaplasia or dysplasia or
P.941
sputum atypia) with the goal of preventing the progression to malignancy. Tertiary chemoprevention means preventing second primary tumors in patients who have already had a first cancer. Cigarette smoking cessation is the only maneuver that definitely meets all three preventive goals (4,10,85).
Results of studies of other preventive strategies have been mixed. Results of two large primary prevention trials of vitamins have been discouraging. Both the Carotene and Retinol Efficacy Trial (CARET), which tested the combination of β-carotene and retinyl palmitate against placebo for 18,314 men and women (including 14,254 cigarette smokers), and the Alpha-Tocopherol Beta-Carotene (ATBC) trial, which examined the effects of β-carotene, with or without α-tocopherol, in 29,133 cigarette-smoking men found that β-carotene significantly increased the risk for lung cancer (reviewed in ref. 36). In addition, a prospective investigation of the relationship between diet and cancer in 478,021 individuals from 10 European countries found an inverse association between fruit consumption and lung cancer risk but no association with vegetables (86). The ongoing Specialized Program of Research Excellence (SPORE) trials of primary and secondary prevention of NSCLC are designed to examine the efficacy of two molecularly targeted agents for chemoprevention: (a) gefitinib, the previously described EGFR inhibitor that treats NSCLC, and (b) tipifarnib, an inhibitor of farnesylation, which prevents Ras from localizing to the membrane and becoming active (87). Both of these trials, and those of cyclo-oxygenase (COX)-2 inhibitors, are on hold pending evaluation of safety issues for chemoprevention studies in healthy individuals (88). However, new information describing the relationship between specific EGFR mutations and sensitivity to gefitinib may allow targeted primary or secondary chemoprevention (89).
Finally, the European Study on Chemoprevention with Vitamin A and N-Acetylcysteine (EUROSCAN), a tertiary prevention study by the European Organization for Research and Treatment of Cancer, found no tumor-free or survival advantage of supplemental retinyl palmitate and/or N-acetylcysteine for 2,592 patients with head and neck cancer or lung cancer (90).
Pleural Effusions
Patients with bronchogenic carcinoma and pleural effusion usually have dyspnea and cough, although one fourth of patients are asymptomatic. In contrast, most patients with malignant mesothelioma and effusion have chest pain that usually is dull and nonpleuritic. The effusions are bilateral in one third of patients with carcinomatous pleurisy. Three fourths of patients have moderate to large pleural effusions. Indeed, the absence of a mediastinal shift in the setting of a large effusion is highly suggestive of a tumor that is fixing the mediastinum, obstructing a main bronchus, or simulating an effusion in the pleural space.
The two categories of pleural effusion in the setting of lung cancer are carcinomatous, with involvement of the pleura (malignant effusion), and paramalignant effusion, in which the pleura is not involved with tumor. Malignant effusions usually are exudative, are serosanguineous or bloody, and may have lymphocytosis, low pH, and low glucose concentration (see Chapter 59). Elevated amylase (salivary isotype) level in a patient without an esophageal rupture may be seen in pleural effusions with adenocarcinoma, and elevated hyaluronic acid may be seen in those associated with malignant mesothelioma. The paramalignant effusion may be either exudative or transudative. These effusions may result from mediastinal or peripheral carcinoma obstructing lymph drainage, central venous obstruction by the tumor, or postobstructive pneumonia or atelectasis. Because patients who smoke and are older are at increased risk for coronary artery disease, congestive heart failure must be considered in the differential diagnosis of a transudative effusion in this group of patients.
Patients with a malignant effusion are not surgically curable, and most patients die within 6 months of diagnosis. However, those with paramalignant effusions may be candidates for surgery, generally as part of a combined modality therapy. Thus, the cause of pleural effusion must be established in patients with otherwise curable bronchogenic carcinoma, as well as for those with nonresectable carcinoma to help direct palliative therapy.
Thoracentesis, with cytologic evaluation of fluid, provides a diagnosis in more than two thirds of cases of malignant effusions, and the yield is improved by repeating the procedure. Potential complications include pneumothorax (5%–10%), bleeding (approximately 1%), and parietal pleural seeding (1%–2%). Unfortunately, for 10% to 20% of patients with malignant effusions, thoracentesis is nondiagnostic. Both thoracentesis and pleural biopsy are routinely performed in an ambulatory setting, the latter only by pulmonologists or thoracic surgeons.
Blind pleural biopsy has only a modest diagnostic yield for malignancy. Thus, nondiagnostic thoracenteses usually are followed by thoracoscopy. Thoracoscopically directed pleural biopsy correctly identifies malignancy in 95% of affected patients (91). More studies are needed to validate the promise of PET scanning for differentiating malignant from benign pleural effusions (92).
Patient Experience
For thoracentesis and pleural biopsy, patients are administered a local anesthetic before the needle is inserted into the pleural space. The patient feels the injection of lidocaine, which stings for a few seconds until the onset of anesthesia. Lung re-expansion usually causes some coughing and may be associated with a chest sensation that some patients find uncomfortable or even painful. Pneumothorax, which
P.942
may result from either needle trauma to the lung or the sucking in of air into an emptied pleural space in the setting of a noncompliant lung, may be associated with increased shortness of breath or with chest or shoulder pain on the affected side. However, pneumothorax may be asymptomatic, so a chest x-ray film after the thoracentesis is mandatory for all patients. The pneumothorax may be followed in an outpatient with serial chest x-ray films or may require chest tube placement, depending on the clinical scenario.
For diagnostic medical thoracoscopy, a short-acting sedative (e.g., midazolam or propofol) and a short-acting opiate (fentanyl) are injected intravenously. Then a small incision is made with the patient under local, anesthesia and an endoscope is passed into the pleural space. Patients feel pressure during the procedure. For video-assisted thoracic surgery, patients require general anesthesia.
Both thoracentesis and thoracoscopic evaluation of the pleura take approximately 30 minutes to perform.
Treatment of pleural effusions in the setting of lung cancer depends on the cause. Systemic chemotherapy is the main option for patients with malignant effusions resulting from SCLC and NSCLC. Various palliative procedures can be attempted.
Patients with effusions secondary to mediastinal lymphatic obstruction may be candidates for external radiotherapy. Effusions secondary to postobstructive processes may be amenable to external radiotherapy or to laser therapy, brachytherapy, or placement of a prosthetic stent, all procedures directed at reestablishing bronchial patency.
Large malignant effusions, which usually recur after they are drained, can be treated with repeated thoracentesis. However, the most common approach to the malignant effusion is pleurodesis. In general, the optimal candidate for pleurodesis is a patient who has already demonstrated symptom relief with thoracentesis and has an anticipated survival of more than a few months. Patients with bulky pleural disease or central bronchial obstruction are less likely to respond to pleurodesis because the lung must re-expand and the pleural surfaces must appose in order to seal. Pleurodesis is helpful in three fourths of selected patients. Surgical pleural abrasion or pleurectomy has higher morbidity and mortality than closed pleurodesis but may be useful for certain patients (91).
Patient Experience
For pleurodesis, the pleural space is drained as completely as possible with a chest tube. The patient usually requires narcotics for analgesia. When drainage has decreased, a sclerosing agent (talc, doxycycline, or bleomycin) is instilled into the pleural space. The chest tube is clamped, and the patient is told to move sequentially in different positions over the subsequent several hours. Doxycycline pleurodesis may be painful, and simultaneous lidocaine instillation into the pleural space and/or narcotics systemically are given. Talc is the most popular agent for pleurodesis and may be the most successful, but its use is controversial because of potential complications. Talc is delivered through the chest tube or sprayed through a thoracoscope. Talc pleurodesis may produce fever, with an onset at 4 to 12 hours and duration up to 72 hours. Talc has been associated with the development of a reversible arterial desaturation syndrome that rarely requires mechanical ventilation and generally is treated with high-dose corticosteroids (93). The sclerosing agent may require reinstillation after a few days if the first treatment is ineffective. Pleurodesis generally requires hospitalization, but for some patients with nonloculated effusions, a pigtail catheter can be inserted on an ambulatory basis for drainage and for delivery of the sclerosing agent.
Solitary Pulmonary Nodule
An SPN is defined as a single focal spherical density in the pulmonary parenchyma that is not associated with any other parenchymal process or with adenopathy. Its diameter is variously designated in the medical literature as up to 3 to 4 cm. In a non-screened population, one fourth to half of these lesions are malignant, either primary pulmonary cancers or metastases from an extrapulmonary source (compare with statistics of a screened population in Screening section). Surgical resection is possible in 80% of patients with an SPN, and the 5-year survival rate of those with a malignant SPN is 40%. These figures are significantly better than are the figures for patients with lung cancer as a group. The nonmalignant lesions result from a diverse group of processes, including the residua of prior granulomatous diseases (e.g., tuberculosis or fungal infection), hamartomas, bronchial adenomas, organizing pneumonia, pulmonary infarcts, and arteriovenous malformations (94). Management of the SPN for an individual patient is based on risk–benefit analysis, including the probability that the nodule is malignant, the risks of the contemplated diagnostic and therapeutic procedures, the accuracy of biopsy techniques, the risk that a delay in therapy will affect outcome, and, finally, the informed patient's preference.
There are four possible management strategies for a patient with an SPN. The first is to do no further followup. The second is to refer the patient for an immediate invasive study for diagnosis. The third is to refer the patient for immediate surgical excision. The fourth is to observe the SPN carefully over time with serial chest x-ray films or unenhanced CT scans (“watchful waiting”) for any signs that would move the patient to one of the other management strategies. The course of action for any individual patient is determined by the answers to four questions and may be aided by PET scanning.
Question 1:Is the visualized lesion really an SPN? It has been estimated that 10% to 20% of lesions interpreted as
P.943
possible SPNs on the initial chest x-ray film are not actual SPNs (95). In these cases, what appeared to be a pulmonary nodule may actually be an extraparenchymal process (e.g., bone, vascular, or chest wall lesion). Alternatively, the pulmonary nodule may not be a solitary process (i.e., additional disease is present). Other radiologic studies, including a review of old films and CT scans, may be helpful.
Question 2:What is the age and stability of the nodule? Comparing the current studies with prior x-ray films is needed to answer this question. The doubling time for a pulmonary malignancy ranges from 20 to 400 days. Thus, the lack of growth of an SPN for several years is sufficient evidence of benign disease.
Question 3:What are the characteristics of the patient that increase the risk of malignancy? The establishment of smoking status (or other toxic exposure as outlined in Table 61.1) is one cornerstone of this investigation. The second is the patient's age. The prevalence of malignancy in SPNs is 1% in patients younger than 35 years but increases rapidly in those older than 35 years (96). Furthermore, the history of a prior malignancy mandates a more aggressive approach because the SPN will prove to be either a primary or metastatic malignancy in 80% of these patients (97). It is important that an SPN in a patient with a known extrapulmonary primary tumor be biopsied because these nodules are more likely to be a new primary tumor of the lung than a metastasis (98).
Question 4:What are the risk characteristics of the nodule? The most important risk characteristic of an SPN is its growth rate, as previously discussed. Second, although calcification may be seen in malignant disease, certain patterns of calcification favor a benign diagnosis. A dense central nidus and a diffuse or laminated pattern of calcification are reliable signs of a healed granuloma. A popcorn pattern of calcification suggests a hamartoma. However, an eccentric nidus of calcification is uninformative and may result from a cancer engulfing a prior calcified lesion. CT densitometry is more sensitive than chest x-ray film for demonstrating calcification (99).
Other characteristics are also important to examine. A nodule >3 cm in diameter likely is malignant. An irregular lesion with poorly defined or spiculated borders has a high likelihood of malignancy.
To answer these questions, workup of the SPN includes first obtaining the necessary films to verify the presence of an SPN, examining all prior x-ray films, and, finally, obtaining a chest CT scan to confirm the presence of disease, demonstrate associated disease, examine the characteristics of the nodule (including size, shape, and presence of calcification, fat, fluid, and vessels), and guide the diagnostic approach.
Two of these management strategy groups should fall out immediately based upon the initial appraisal. For one group of patients, the benign nature of the process may be assured after some or all of these steps; these patients will not require further evaluation. Patients who have a high risk for malignancy based on the criteria should undergo a staging evaluation (as per above discussion) and should be referred for surgery if appropriate. For the patient who does not fall into one of these two groups, the decision on whether to refer the patient for an immediate biopsy or to adopt a watchful waiting approach may be aided by the findings of a PET scan. Prospective studies show that the sensitivity of PET for detection of malignant nodules is 90% to 92%, and the specificity is 83% to 90%. The sensitivity of PET decreases to 80% for nodules less <1.5 cm and is significantly lower for nodules <1.0 cm. Carcinoids, bronchioloalveolar carcinomas, and a few adenocarcinomas may be missed (100,101). Thus, followup of negative scans (the watchful waiting strategy) is mandatory. A positive PET scan would favor obtaining an immediate biopsy.
The diagnostic techniques for obtaining a biopsy of an SPN include fluoroscopically directed PTNA, fluoroscopically directed transbronchial biopsy or TBNA performed through a fiberoptic bronchoscope, thoracoscopic biopsy, or thoracotomy. These procedures were discussed in Diagnostic Procedures section. The diagnostic yield of PTNA is 70% to 90% and of transbronchial biopsy is 30% to 70% (94,102). Nonneoplastic causes of peripheral nodules (especially certain infections) may be diagnosed by transbronchial biopsy or PTNA. However, the specter of false-negative results mandates that nonspecific findings be followed either by another procedure or by watchful waiting, as described below. The choice of the initial diagnostic technique and of follow-up management is best determined by a pulmonologist in consultation with the patient.
Based on this discussion, the watchful waiting strategy would be appropriate management for the following examples: a patient younger than 35 years old who has no risk factors and has a small well-defined nodule that is negative on PET, a patient with a nodule <2.5 cm that has been unchanged in size for 2 years by prior chest x-ray films, a patient who has few risks and negative diagnostic studies, and a patient for whom the risk of the diagnostic and therapeutic procedures outweighs the risks of watchful waiting. Given the range of tumor doubling times, the delay inherent in this strategy likely would not result in significantly decreased survival of the patient with a pulmonary nodule that is malignant (94). A scheme for monitoring serial unenhanced chest CT scans might be to obtain repeat films every 3 to 4 months for 1 year, then every 6 months for 1 year, and then to repeat at yearly intervals thereafter. The total number of years to follow a nodule depends on the suspicion of malignancy based on the criteria. For patients for whom the level of suspicion is low, serial studies for 2 years are adequate. For patients for whom the level of suspicion for malignancy is high, 5-year followup is necessary to encompass the unusually slow-growing carcinoma (usually an adenocarcinoma).
P.944
Specific References*
For annotated General References and resources related to this chapter, visit http://www.hopkinsbayview.org/PAMreferences.
P.945