Non-Small-Cell Lung Cancer

, , , ,

This management guide for non–small-cell lung cancer (NSCLC) covers the symptoms, screening, diagnosis, staging, and treatment of adenocarcinoma, squamous cell carcinoma, and large-cell carcinoma of the lung.

Overview

Lung cancer has been the leading cause of cancer death among men in the United States for years and since 1988, it has become the number-one cause of cancer death among women. An estimated 224,390 new cases of lung cancer and 158,080 deaths due to this disease will occur in 2016, representing roughly 27% of all cancer deaths. This exceeds the combined number of deaths from cancer of the next leading sites (breast, prostate, and colon). Lung cancer accounts for 6% of all deaths in the United States.

Lung cancer develops from pulmonary parenchymal or bronchial supportive tissues. Although multiple cell types are often found within a single lung tumor, one type usually predominates. Based on the therapeutic approach, there are two major subdivisions of lung cancer: small-cell lung cancer (SCLC) and non–small-cell lung cancer (NSCLC). Non–small-cell tumors account for approximately 85% of all lung cancers. The three major tumor types included under this category are adenocarcinoma, squamous cell carcinoma, and large-cell carcinoma.

This chapter provides basic information on the epidemiology, etiology, screening, prevention, and signs and symptoms of lung cancer in general and then focuses specifically on the diagnosis, staging, pathology, and treatment of NSCLC, as well as on the pulmonary evaluation of lung cancer patients and follow-up of long-term survivors.

A separate chapter provides information on the staging, pathology and pathophysiology, and treatment of the far less common SCLC and concludes with brief discussions of mesothelioma and thymoma.

Epidemiology

Gender

In the United States, the estimated number of new lung cancer cases for 2016 is 117,920 for men and 106,470 for women. Although the incidence of lung cancer had been rising in women, the figures have stabilized and begun to decline recently. The incidence is decreasing in men.

Age

The age of patients at which lung cancer is diagnosed varies widely, but the median age at diagnosis is approximately 70 years. The likelihood of lung cancer in patients younger than 40 years of age is low, with less than 1% of all lung cancers occurring in patients under age 40.

Race

In the United States, the highest incidence of lung cancer in men and women is found in African Americans (93.4/100,000 for men and 51.4/100,000 for women), followed by Caucasians (79.3/100,000 for men and 58.7/100,000 for women).

Geography

There are geographic variations in the incidence of lung cancer, with the highest rates worldwide observed in North America and Eastern Europe; in the United States, the highest rates are found in northern urban areas and along the southern coast from Texas to Florida. The state with the highest incidence of lung cancer is Kentucky, with an incidence of 120.4/100,000 in men and 80.7/100,000 in women.

Survival

The overall 5-year survival rate for lung cancer is 17%; this has improved from a rate of approximately 12% in the 1970s.

Etiology and Risk Factors

Cigarette Smoking

Approximately 87% of all cases of lung cancer in the United States are related to cigarette smoking. Tobacco smoke contains more than 40 known carcinogens. There is a relatively strong dose-response relationship between cigarette smoking and the development of this cancer. The greater the number of cigarettes smoked on a daily basis and the greater the number of years of smoking, the greater is the risk of lung cancer. An individual who smokes one pack of cigarettes daily has a 20-fold increased risk of lung cancer compared with a nonsmoker.

The overall incidence of cigarette smoking decreased from 1974 through 1992. Smoking cessation decreases the risk of lung cancer, but a significant decrease in risk does not occur until approximately 5 years after discontinuation, and the risk remains higher in former smokers than in nonsmokers for at least 25 years. The benefit of smoking cessation is greater if it occurs at a younger age.

Smoking cessation is difficult. Recent data have suggested that a variety of hereditary factors increase the risk of addiction to nicotine among some individuals. Nevertheless, millions of former smokers have quit successfully. Smoking cessation programs that address both physical withdrawal from nicotine and psychological dependence appear to be more effective than either of these approaches alone. In addition, continued efforts are needed to prevent adolescents and preadolescents from beginning to smoke and to encourage them to quit after a brief period of experimentation.

Several cancer centers have recently reported that more than half of their patients with newly diagnosed lung cancer are former smokers, having quit more than 1 year before diagnosis. Healthy ex-smokers represent a large group of individuals who may benefit from effective tools for early detection and/or chemoprevention of lung cancer.

Secondhand Smoke

Not only is smoking risky for those who smoke, it also poses a hazard to nonsmokers who either live or work with smokers. It is estimated that approximately 3,000 lung cancer deaths per year in the United States are due to secondhand smoke. Individuals who live in a household with a smoker have a 30% increase in the incidence of lung cancer compared with nonsmokers who do not live in such an environment.

Asbestos Exposure

Exposure to asbestos is another risk factor for lung cancer. Cigarette smokers who are exposed to asbestos develop lung cancer at an extremely high rate. There is a 90-fold increase compared with unexposed individuals. Exposure to asbestos is also a major risk factor for the development of mesothelioma (see discussion of this cancer in the chapter on small cell lung cancer).

Radioactive Dust and Radon Exposure

Uranium miners who have been exposed to radioactive dust and radon gas also have an increased incidence of lung cancer. Although there has been some controversy about the risk posed by exposure to residential radon gas, a study conducted in Sweden showed an increased incidence of lung cancer in individuals who were exposed to a high level of radon in their homes. Ten percent of lung cancer cases appear to be associated with radon exposure.

Infection

Human papilloma virus (HPV) has been implicated as a cause of NSCLC; nearly 25% of cases in nonsmokers may be associated with the virus. Recurrent respiratory papillomatosis (RRP) is a disease typically found in young patients with laryngo-tracheobronchial HPV polyps. RRP may cause cough and obstruction, and may eventually degenerate into malignancy. HPV 16/18 infection is associated with the p53 mutation. Human immunodeficiency virus (HIV) infection increases the risk of lung cancer by 2- to 11-fold. It is the most common non–AIDS-associated cause of death, accounting for 16% of deaths. Treatment with antiretrovirals has not changed the risk or virulence of lung cancer. Stage-for-stage, NSCLC is far worse for the HIV patient than for patients without AIDS.

Inheritance

Family history has been considered a risk factor, and often a smoking-related history is uncovered. More recently, inheritable mutations have been discovered. One in particular is T790M, a germline mutation of EGFR gene that is associated with adenocarcinoma.

Air Pollution

Particulate matter in air pollution has been shown to increase the risk of lung cancer, especially adenocarcinoma. The risk is associated with the size and volume of the particulate matter. Outdoor pollution may account for 1% to 2% of lung cancer cases.

Chronic Lung Inflammation and General States of Inflammation

Lung cancer has been associated with idiopathic pulmonary fibrosis and many other inflammatory lung diseases; chronic obstructive pulmonary diseases; tuberculosis; and systemic states of inflammation such as rheumatoid arthritis, Crohn’s disease, and other chronic inflammatory conditions.

Screening and Prevention

Screening

Prior to 2011, screening for lung cancer among asymptomatic individuals at elevated risk because of smoking history or occupational exposures was not recommended. An unfortunate result of this policy was that most patients present at an advanced stage, and cure rates have improved little over the past 30 years. Only 7% of NSCLC patients received a diagnosis at stage IA.

Three randomized screening trials conducted in the United States in the 1970s failed to show a reduction in lung cancer mortality among the smokers who were screened by sputum cytology and chest radiography for lung cancer. Despite the fact that these American trials were not designed to evaluate chest radiography as a screening tool, the results led most experts to conclude that screening for lung cancer was not worthwhile. In addition, most investigators recommended that research efforts and resources be allocated to the prevention of lung cancer. A recent, randomized, prospective trial from Czechoslovakia showed that screening with chest radiography increased the diagnosis of early-stage lung cancer but failed to reduce the mortality from lung cancer.

The potential to screen for lung cancer has received renewed interest because of the superior performance of low-dose helical computed tomography (CT) (radiation dose 1.0–1.5 mSv compared with 8 mSv for regular chest CT) compared with chest radiography (radiation dose 0.06 mSv) in detecting small lesions.

Numerous studies are ongoing to evaluate chest CT scan for lung cancer screening. Several recent reports from Japan, Germany, and the United States have documented the ability of low-dose spiral CT scans to detect lung cancer at an early stage. In some recent trials, more than 80% of lung cancers detected by screening were diagnosed in stage I.

Kaneko (Radiology 1996) screened male smokers older than 50 years. Of the 15 cancers detected by CT scan, only 4 were seen on chest radiography; 14 of the 15 cancers were stage I, with an average diameter of 1.6 cm. Ohmatsu (ASCO 1999) found 35 lung cancers (0.37% detection rate) with 9,452 CT scans. Of these cancers, 27 were stage IA. These patients had a 3-year survival rate of 83%.

The I-ELCAP (International Early Lung Cancer Action Program) is a single-arm prospective study that has accrued more than 70,000 study subjects from 30 sites and documented that a high percentage of lung cancers are detected in stage I, a stage in which long-term survival can reasonably be anticipated in more than 60% of patients. These studies provide early evidence to suggest that CT screening in populations at high risk for lung cancer has the potential to reduce lung cancer mortality in the near future.

In a controversial article from I-ELCAP, Henschke et al presented the results of their low-dose CT screening trial in 31,567 high-risk participants. CT was performed every 7 to 18 months. Lung cancer was found in 484 (1.5%), of whom 412 (85%) had stage I cancer, with an 88% 10-year survival rate estimated in that cohort. Noncalcified pulmonary nodules were detected in 233 participants (23%; 95% CI, 21–26) by low-dose CT at baseline, compared with 68 (7%; 95% CI, 5–9) by chest radiography. Lung cancer was detected by CT in 27 patients (2.7%; 95% CI, 1.8–3.8) and by chest radiography in 7 patients (0.7%; 95% CI, 0.3–1.3).

Of the 27 CT-detected cancers, 26 were resectable. Stage I cancers were diagnosed in 23 of 27 patients (85%) by CT and 4 of 7 patients (57%) by chest radiography. In addition, low-dose CT detected four more nonparenchymal cases of lung cancer: two with endobronchial lesions and two in the mediastinum. These cases show an added benefit of low-dose CT over chest radiography, although the data were not included in the analysis. (The study primarily focused on malignant disease in noncalcified pulmonary nodules detected by low-dose CT or radiography.) It remains to be seen, however, whether lung cancer screening with low-dose spiral CT will reduce the lung cancer mortality of the study population or only improve the 5-year survival rate of the patients with diagnosed lung cancer. In response to the I-ELCAP results, Bach et al reviewed the collective results of three prospective, concurrently run, randomized screening trials in 3,246 high-risk patients. They found no reduction in lung cancer deaths or advanced cancer in the screened groups.

Based on growing evidence that spiral CT may truly provide for a successful early detection strategy, the National Cancer Institute (NCI) launched the National Lung Screening Trial (NLST) in September 2002. NLST has accrued 53,500 current and former smokers (aged 55 to 74) into a prospective trial, randomizing participants to receive annual spiral CT or annual chest radiography. In October 2010, the data safety monitoring board announced that there was a 20.3% reduction in lung cancer mortality and a 7% reduction in all-cause mortality in those screened with low-dose CT and closed the study. Concerns have been raised about the fact that more than 20% of patients are found to have abnormalities and the cost of screening this patient population could be as much as $38,000 per year of life gained. With the longer history of population-based lung cancer CT screening, perhaps the more realistic cost per year of life saved is 1 million yen (~$8,400) (Tabata 2014). The number of people needed to be screened with low-dose CT to prevent one lung cancer–related death was 320 in the NLST. The NLST group and others (Henschke 2011, Foy 2011, Pyenson 2014) demonstrated a 20% to 50% reduction in lung cancer mortality, especially increasing the percentage of stage I patients who do not require costly adjuvant therapy. Concerns about overdiagnosis and the resultant increased expensive and risky noninvasive and invasive testing are mitigated by using the algorithms provided by the National Comprehensive Cancer Network (NCCN), the American College of Chest Physicians (ACCP), the Fleischner Society, and I-ELCAP.

The increased use of CT scanning in the United States, currently more than 62 million scans per year, has significantly increased the exposure to radiation in the population and may be a future health issue. The exposure to radiation from two or three standard CT scans in an adult is similar to that experienced by many survivors of atomic bombs dropped on Japan in 1945. The risks associated with exposure to radiation are highest in children. Because of potential harm from overdiagnosis and the risk of radiation-induced cancers, this important study is the first step toward a targeted cost-effective public policy of encouraging the use of CT screening in populations in which the risk-reward ratio is most advantageous. Although models based upon the questionable theory of nonthreshold linear radiation carcinogenesis predict CT-caused cancers, there is no empirical evidence to suggest that any cancers result from low-dose CT scans in adults.

The efficacy of lung cancer screening is also being evaluated as part of the Prostate, Lung, Colorectal and Ovarian (PLCO) Cancer Screening Trial. Men and women were randomized to receive annual chest radiography or usual care. Eligibility was not based on risk of lung cancer, because given the large size of the study (> 100,000 participants), it was expected that there would be appreciable numbers of current and former smokers among the participants.

Biomarkers, proteomic evaluations, and circulating endothelial cells are currently under intensive investigation within and outside the Early Detection Research Network.

Updated guidelines in 2013–2015 from the NCCN in the United States recommend annual low-dose CT screening for those at high risk; they recommend no routine screening for moderate- or low-risk individuals. High risk was defined by the NCCN as age 55 to 74 years with a 30 pack-year history of smoking and, if no longer smoking, smoking cessation within the past 15 years; or a 20 pack-year history of smoking with one additional risk factor (other than secondhand smoke exposure). Similar guidelines were also issued in 2012 by ASCO, the American Association for Thoracic Surgery, and the American College of Chest Physicians.

Expert groups are in agreement that patient counseling of the risks and benefits of screening; the development of a registry to collect data on follow-up testing, smoking behavior, radiation exposure, and patient experience; the development of quality metrics for CT interpretation, similar to quality control for mammography; and emphasis on the importance of smoking cessation are all integral components of a screening strategy.

The lack of demonstrated benefit for the earlier radiographic screening approaches should not be misinterpreted as nihilism about the early detection of lung cancer. Individuals at risk (current and former smokers) who present with symptoms consistent with lung cancer deserve appropriate evaluation. A lack of resolution of abnormalities on a chest radiograph obtained after the completion of empiric antibiotic therapy for pneumonia should prompt further evaluation for possible lung cancer.

Chemoprevention

The concept of field carcinogenesis was originally developed for the aerodigestive tract in the early 1950s. Reducing the exposure of the epithelial mucosa to carcinogens, predominately cigarette smoke, has the greatest impact on reducing the incidence of cancer in high-risk individuals.

The Finnish Alpha-Tocopherol Beta-Carotene Study was conducted to determine whether certain vitamin supplements would prevent cancer. It evaluated 29,133 male smokers over 5 to 8 years and found there was an 18% increase in the incidence of lung cancer in the group taking beta-carotene. Other chemopreventive agents studied in a phase III fashion include retinol, aspirin, retinyl palmitate, etretinate, isotretinoin, α-tocopherol, 4-hydroxyphenyl retinamide, anethole dithiolethione, and N-acetylcysteine. None of them showed any preventive benefit.

Second primary lung tumors develop at a rate of 1% to 3% annually for the first 5 years following resection of stage I NSCLC. The Intergroup randomized trial that assessed the ability of 13-cis-retinoic acid to prevent the occurrence of a second primary cancer in patients with completely resected stage I NSCLC showed no impact of treatment on the incidence of second primary tumors. Furthermore, patients who continued to smoke and who received isotretinoin had a higher risk of recurrence of the index cancer. Also, there was no reduction in second primary tumors in the 13-cis-retinoic acid–treated group. Trials using cyclooxygenase 2 inhibition in former and current smokers are yet to be reported. Tyrosine kinase inhibition is currently being studied to reverse bronchial premalignant lesions and Ki-67 levels in the Lung Cancer Biomarkers Chemoprevention Consortium trial.

Selenium as L-selenomethionine has been shown to inhibit cell growth, induce apoptosis in vitro, and retard carcinogenesis at higher-dose levels in animal models. Epidemiologic data suggest an inverse relationship between selenium intake and lung cancer.

A study (Clark et al: JAMA 1996) designed to determine the effects of selenium on the incidence of basal or squamous cell carcinomas showed that nutritional supplementation with this agent had no consequences on the incidence of skin cancer; however, secondary analyses revealed that it was associated with significantly fewer cases of lung cancer.

A phase III Intergroup selenium prevention trial was designed to follow the lung cancer isotretinoin prevention trial. To reduce the incidence of second primary tumors, this double-blind design randomized patients by a 2:1 ratio to receive either selenomethionine (200 μg/d) or placebo daily for 48 months. Patients were monitored for safety, development of second primary tumors, and recurrence. At a planned interim analysis, there was a trend toward a lower incidence of second primary cancers in patients receiving placebo compared with selenium. Second primary tumor (lung/overall) incidence was 1.36–3.66 per 100 person years for placebo vs 1.91–4.11 for selenium (P = .150). The study was discontinued because of futility.

Educational programs

Given that results from the Intergroup study and other chemoprevention studies to date have been disappointing, it is important to continue educational efforts to prevent adolescents from smoking cigarettes and to advocate smoking cessation in active smokers. Some experts believe that educational programs must begin during childhood, probably between the ages of 6 and 10 years. Targeting children and young adults is a significant priority of any lung cancer reduction program.

Smoking cessation is more likely to be successful when a combination of behavioral and pharmacologic interventions is offered. A simple five-step algorithm called the 5 As that includes the elements of brief counseling for office practice is effective for increased abstinence rates. This system encourages clinicians to ask patients about their smoking status, advise smokers to quit, assess their readiness to quit, assist them with their smoking cessation effort, and arrange for follow-up visits or contact. First-line drug therapy for smokers includes nicotine replacement therapy, bupropion, or varenicline. A meta-analysis of 36 randomized trials of bupropion monotherapy found that bupropion nearly doubles the likelihood of smoking cessation, compared with placebo. The most common side effects of bupropion include insomnia, agitation, dry mouth, and headache, while seizures are a rare (0.1% of patients) but serious side effect.

Another meta-analysis found that varenicline increased the odds of quitting threefold compared with placebo and produced a quit rate of 33% at 6-month follow-up. Three randomized trials have also suggested that varenicline is superior to bupropion for rate of short-term smoking cessation and may be superior for long-term cessation; however, long-term cessation data are more equivocal. Varenicline may increase the risk of neuropsychiatric and cardiovascular side effects, and patients should be monitored closely while taking it.

Signs and Symptoms

The clinical manifestations of lung cancer depend on the location and extent of the tumor. In patients who have localized disease, the most common symptoms are related to obstruction of major airways; infiltration of lung parenchyma; and invasion of surrounding structures, including the chest wall, major blood vessels, and viscera.

Cough

Cough is a major manifestation of lung cancer and is present in nearly 80% of patients with symptomatic lung cancer. It is important to remember, however, that most lung cancer patients are current or former smokers, and they may have a cough related to chronic irritation of the upper and/or lower airways from cigarette smoke. Therefore, smokers should be asked whether there has been a change in their cough, such as an increase in frequency or severity.

Dyspnea and Hemoptysis

Increasing dyspnea and hemoptysis may be signs of lung cancer, although in the case of hemoptysis, 70% of patients bleed from nonmalignant causes, mostly infection, and more frequently, bronchitis. In patients who present with hemoptysis, are older than 40 years, and have a history of smoking and chronic obstructive pulmonary disease without an abnormality on chest radiographs, lung cancer should be considered in the differential diagnosis.

Pneumonia

Postobstructive pneumonia secondary to partial or complete bronchial obstruction occurs relatively frequently in association with lung cancer. It is important to obtain repeated chest radiographs in adults who have been treated for pneumonia to be certain that the radiographic abnormalities have cleared completely. Patients with recurring bronchitis and pneumonia should undergo further clinical investigation.

Pleural Effusion

Lung cancer may spread to the pleural surface or may obstruct segmental or lobar lymphatics, resulting in pleural effusion and increased dyspnea.

Other symptoms include fatigue, weight loss, and loss of appetite.

Chest Pain

Approximately 5% of lung tumors invade the chest wall. The resultant pain is a better predictor of chest wall invasion than are chest CT findings. An individual who complains of persistent chest pain should have chest radiography to exclude the presence of peripheral lung cancer that has invaded the chest wall.

Shoulder and Arm Pain

Apical tumors that infiltrate surrounding structures (also called Pancoast or superior sulcus tumors) produce shoulder and/or arm pain as a result of brachial plexus compression. Tumors in the apical lung segments that involve the superior sulcus of the chest may be difficult to detect on a routine chest radiograph; therefore, a person who complains of persistent shoulder pain, particularly with signs of neurologic involvement, should have a CT scan of the chest to look for an apical tumor. An MRI scan of the chest apex may be beneficial to assess depth, the spinal column, and mediastinal and vascular invasion.

Horner's Syndrome

Invasion of the sympathetic ganglion by an apical lung tumor causes Horner's syndrome (ipsilateral ptosis, myosis, and anhidrosis).

Hoarseness

Hoarseness secondary to vocal cord paresis or paralysis occurs when tumors and lymph node metastases compress, cause dysfunction in, or invade the recurrent laryngeal nerve. This situation is more common on the left side, where the recurrent laryngeal nerve passes under the aortic arch, but it may also occur with high lesions on the right side of the mediastinum.

Other Symptoms of Tumor Compression

Lung tumors may also cause dysphagia by compression or invasion of the esophagus or may cause superior vena cava syndrome by compression or invasion.

Some tumors may result in wheezing or stridor secondary to compression or invasion of the trachea and may also cause signs of cardiac tamponade secondary to involvement of the pericardial surface and subsequent accumulation of pericardial fluid. Spinal cord impingement may result in loss of function and/or sensation. Central nervous system involvement may result in headache of different levels of severity, seizure, dizziness, and other mental function changes.

Signs and Symptoms of Metastatic Disease

Lung cancer can metastasize to multiple sites, most commonly to bone, liver, brain, lungs (contralateral or ipsilateral), and adrenal glands.

A lung cancer patient who has brain metastases may complain of headaches or specific neurologic symptoms, or family members may notice a decrease in the patient's mental acuity. Also, metastatic lung cancer may cause spinal cord compression, resulting in a characteristic sequence of symptoms: pain, followed by motor dysfunction, and then sensory symptoms. The patient may have any or all of these symptoms.

Patients who complain of band-like pain encircling one or both sides of the trunk may have spinal cord compression. In addition, coughing and sneezing may cause significant exacerbation of pain from spinal cord compression.

A bone radiograph and/or a bone scan are warranted in lung cancer patients who complain of persistent pain in the trunk or extremities. If performed in the evaluation of lung cancer, 18F-fluorodeoxyglucose (FDG)-positron emission tomography (PET) supplants the need for a bone scan in most patients. PET appears to be more sensitive with similar specificity for bone metastases when compared with bone scan. If plain films are normal or equivocal for metastases, CT and/or magnetic resonance imaging (MRI) may be helpful to evaluate suspicious areas. MRI of the spine is the most effective way to evaluate suspected spinal cord compression.

Systemic Paraneoplastic Symptoms

Lung cancer is commonly associated with systemic manifestations, including weight loss (with or without anorexia). In addition, patients frequently complain of fatigue and generalized weakness. SCLC is associated with hormone production, which causes endocrine syndromes in a subset of patients, such as SIADH (syndrome of inappropriate antidiuretic hormone secretion) and via secretion of ACTH (adrenocorticotropic hormone) hypercortisolism.

Specific Neurologic Syndromes

Syndromes such as Lambert-Eaton syndrome (see the “Oncologic Emergencies and Paraneoplastic Syndromes” chapter), cortical cerebellar degeneration, and peripheral neuropathy may occur in lung cancer patients but are relatively rare.

Clubbing

Although clubbing may occur in a variety of conditions, it is important for the clinician to evaluate a patient's hands. If clubbing is noted, obtaining a chest radiograph may result in the early diagnosis of lung cancer.

Hypertrophic Pulmonary Osteoarthropathy

A relatively small percentage of patients with lung cancer may present with symptomatic hypertrophic osteoarthropathy. In this syndrome, periosteal inflammation results in pain in affected areas, most commonly the ankles and knees. It is frequently associated with clubbing.

Carcinoid Syndrome

This syndrome is extremely uncommon in patients who have a bronchial or pulmonary carcinoid tumor. Most of these patients are asymptomatic (tumors are found by radiography), and a few have cough from an endobronchial lesion.

Staging and Prognosis

Staging

Changes to TNM

The staging of lung cancer must be conducted in a methodical and detailed manner. The TNM staging system (Table 1) applies equally well to all histologies of NSCLC. SCLC is traditionally staged as limited (disease potentially amenable to definitive chemoradiation therapy, usually within one radiation port) or extensive. However, TNM staging is also recommended for epidemiologic purposes, particularly with the renewed recent debate about the role of surgery for the small numbers of SCLC cases that present while still localized. Most patients, however, have advanced disease at the time of presentation.

Stage is commonly reported as either clinical or pathologic, designated as “c” or “p,” respectively. Clinical stage is based on noninvasive (or minimally invasive) tests, whereas pathologic stage is based on tissue obtained during surgery (see section on “Diagnosis and staging evaluation”).

The International Association for the Study of Lung Cancer (IASLC) has developed changes in the TNM system (TNM7) based on an international collection and review of 100,809 patients.

TABLE 1: TNM7 staging of lung cancer

1. Moving tumors > 7 cm from T2 to T3

2. Reassigning the category given to additional pulmonary nodules

a. Additional nodules in the same lobe, T4 to T3

b. Additional nodules in a different ipsilateral lobe, M1 to T4

3. Reclassifying pleural effusion as an M descriptor, M1a.

In addition:

1. T2b, N0, M0 cases can be moved from stage IB to stage IIA

2. T2a, N1, M0 cases can be moved from stage IIB to stage IIA

3. T4, N0-1, M0 cases can be moved from stage IIIB to stage IIIA.

Prognostic Factors

Stage

The most important prognostic factor in lung cancer is the stage of disease.

Performance status and weight loss

Within a given disease stage, the next most important prognostic factors are performance status and recent, unexplained weight loss. The two scales used to define performance status are the Eastern Cooperative Oncology Group (ECOG) performance status or the Zubrod system and the Karnofsky system (see Appendix 1). Simply, patients who are ambulatory have a significantly longer survival than those who are nonambulatory. Similarly, patients who have lost more than 5% of body weight during the preceding 3 to 6 months have a worse prognosis than patients who have not lost a significant amount of weight.

Molecular prognostic factors

Several studies published over the past decade have indicated that mutations of ras proto-oncogenes, particularly KRAS, portend a poor prognosis in NSCLC, and in the case of adenocarcinomas, are more frequently found in smokers. Accordingly, selective research has focused on developing molecularly targeted therapeutic approaches to the ras proto-oncogenes, in particular, the farnesyl transferase inhibitors as well as EGFR inhibitors and antiangiogenic approaches (see section on “Novel and promising agents”).

Of equal relevance was the completion of large studies by Pastorino et al and Kwiatowski et al evaluating the prognostic importance of immunocytochemical and molecular pathologic markers in stage I NSCLC.

In the study by Pastorino et al, gene-expression profiles were performed in 198 tumor samples to develop a predictive model for recurrence, the lung metagene model. The model was then tested in three cohorts of stage IA patients and was found to be highly accurate in predicting the likelihood of recurrence. By identifying patients who are likely to experience recurrence, adjuvant therapy is more likely to be effective. In the study by Kwiatowski et al, gene-expression profiles were performed in 185 frozen samples, and a 5-gene signature that correlated with relapse-free and overall survival was identified.

The findings of these two studies suggest that pathologic stage and completeness of surgical resection may yield the most critical prognostic information, but mutation of the KRAS oncogene and absence of expression of the HRAS p21 proto-oncogene may augment the pathologic information. Further data indicate that a molecular prognostic model differentiating patients with lung cancer may be possible using proteomics or other molecular biology techniques.

Molecular profiling has been described as a means to molecularly characterize tumors and to identify their metastatic potential and regional growth character. Other molecular markers that portend a poor prognosis include reduced expression of RASSF1A and FHIT (fragile histidine triad); reduced expression of catenins and E-cadherin; reduced ERCC1 and RRM1; reduction in tumor suppressor genes p53, retinoblastoma gene, p16INK4A, and p15INK4B; increased telomerase activity; overexpression of EGFR (c-erbB-1), matrix metalloproteinase, HER2/neu (c-erb-2), and vascular endothelial growth factor (VEGF); c-Met overexpression; overexpression of mutated p53; urokinase plasminogen activator overexpression; promoters of angiogenesis; reduced expression of the nm23 gene; and low expression of KAI1/CD82.

Besides advanced tumor stage, risk factors that have been independently associated with recurrence of NSCLC include visceral invasion, lymphovascular invasion, number and percentage of lymph nodes involved, degree of lymph node involvement, number of lymph node stations involved, stain positive for TTF-1, and micropapillary histology (especially when there is more than 5% involvement).

At the 2014 Chicago Multidisciplinary Symposium on Thoracic Oncology, Kumar and Movsas et al presented prognostic biomarkers in patients whose NSCLC was treated with radiation therapy. Overall, they found that CD68 was predictive of increased risk of death, whereas MMP-9 was associated with decreased risk of death. Associations between racial disparities and survival were dependent on median household income. Of note, high expression of CD44 and low expression of MFG-E8 were predictive of local tumor control.

The ongoing Cancer and Leukemia Group B (CALGB) 140202 and the Southwest Oncology Group (SWOG) 0720 are two examples of prospective trials using molecular information to guide therapy. Patients less likely to benefit from adjuvant therapy are monitored for recurrence, and those exhibiting certain molecular features receive treatment.“Personalized medicine” is expected to have a greater role in preventing recurrence in early-stage lung cancer.

Sidebar: Kratz et al found that, in stage I NSCLC, a quantitative polymerase chain reaction 14-gene expression assay on formalin-fixed paraffin-embedded tissues was found capable of identifying low-, intermediate-, and high-risk patients for recurrence (P = .0003).

Quality of life

The Radiation Therapy Oncology Group (RTOG) reanalyzed the results of RTOG 9801 to assess the added value of quality of life (QOL) as a prognostic factor for overall survival. On multivariate analysis, they found that the global QOL score superseded the classic predictors of survival (eg, Karnofsky performance status, stage, etc) in patients with locally advanced NSCLC. Patients with a global QOL score less than the median (66.7 of 100) had a 70% higher rate of death than patients with a QOL of 66.7 or higher (P = .002). Of note, patients who were married or had a partner had higher QOL scores than those who were not (P = .004). A clinically meaningful increase in the QOL score (of 10 points) corresponded to a decrease in the hazard of death by 10% (P = .002). This study highlights the “added value” of patient-reported outcomes and the need to incorporate QOL measures into clinical oncology trials.

Diagnosis and Staging Evaluation

History and Physical Examination

The diagnosis and preoperative staging of lung cancer begin with a good history and physical examination. When obtaining the history, the clinician should keep in mind the tendency for lung cancer to involve major airways and other central structures. Similarly, the patterns of metastatic dissemination and systemic manifestations must be considered when conducting the physical examination.

Patients should be questioned specifically about the presence of palpable masses, dysphagia, bone pain, headache, or changes in vision. Careful auscultation and percussion may suggest the presence of atelectasis or pleural effusion. Auscultation of the chest also may provide evidence of large-airway obstruction and pulmonary consolidation. Rhonchi and wheezing may provide some helpful treatment planning information. An enlarged liver may indicate hepatic metastases.

Palpation of the Neck and Supraclavicular Fossa

Discovery of adenopathy of the neck and supraclavicular fossa may allow both diagnosis and staging, by needle or open biopsy.

Imaging Studies

Chest radiographs

These provide initial helpful information in patients with new respiratory symptoms. Posteroanterior and lateral chest radiographs are fundamental in assessing the local extent of the primary tumor and also may provide valuable information regarding metastatic disease.

The chest radiograph should be inspected for the presence of a pleural effusion or synchronous pulmonary nodules, and the bones should be examined for evidence of osseous metastases. A widened mediastinum usually indicates metastatic disease within the mediastinal lymph nodes. Comparison with previous radiographs is helpful and well worth the effort expended in their retrieval.

Chest CT from the lower neck to the mid level of the kidneys

Testing of this area, including the entire liver and adrenal glands with 5- to 10-mm slices, is performed routinely to further define the primary tumor and to identify lymphatic or parenchymal metastases. In a review of 20 studies that assessed the value of CT to determine mediastinal lymph node involvement, with an average prevalence of 28%, CT had a pooled sensitivity of 57%, a specificity of 82%, and a negative predictive value (NPV) of 83%. Benign enlargement of mediastinal nodes is more common in patients with postobstructive infection. Histologic confirmation of the presence or absence of tumor within the mediastinal lymph nodes is necessary whenever this information will change treatment recommendations. In patients who are considered surgical candidates, metastatic tumor is found in approximately 15% to 20% of mediastinal lymph nodes less than 1 cm in greatest diameter.

In mediastinal nodes larger than 1 cm in the minor axis (ie, in a given 2-dimensional image, the smaller of the two dimensions) by CT that are negative by FDG-PET, the false-negative rate is 13% to 25%, and the false-positive rate is approximately 56%.

It is important to remember that patients with persistent symptoms, such as cough and dyspnea, who have a normal chest radiograph may be harboring a central lesion that is not obvious on chest radiography but can be detected by chest CT. Also, as previously mentioned, apical tumors (Pancoast tumors) may be difficult to detect on a chest radiograph but are usually readily apparent on a CT scan.

PET

For lung masses (lung lesions > 3 cm in greatest dimension), FDG accumulation on PET implies a significant likelihood of malignancy.

The maximum standardized uptake value (SUV) threshold of 2.5 within the region of interest optimizes the sensitivity and specificity of PET in assessing suspicious lung lesions larger than 1 cm. An FDG-avid lesion on PET should never be assumed to be malignant, and a PET-negative lesion is not absolutely benign. Adenocarcinoma in situ and carcinoid are two cell types that do not readily accumulate FDG. Furthermore, higher SUV lesions do not imply a greater likelihood of cancer; the highest SUVs have been found in inflammatory lesions, such as granulomas and infections.

A review of 18 studies of the utility of PET to assess the mediastinal lymph nodes demonstrated a pooled sensitivity of 84% and a specificity of 89%, with a positive predictive value (PPV) of 79% and an NPV of 93%. Combining the results of CT and PET, the PPV and NPV were 83% to 93% and 88% to 95%, respectively. Thus, FDG-PET is superior to CT scanning in staging the mediastinal lymph nodes. An estimated 15% to 20% of patients with a known or suspected diagnosis benefit from a preoperative FDG-PET, because previously unrecognized metastatic disease will be discovered.

The false-negative rate can be as high as 25% to 30% for CT, PET, and PET-CT when the prevalence of mediastinal involvement is higher, such as in cases of lesions being large and/or more centrally located and/or when there is hilar lymph node involvement.

Several trials have evaluated the prognostic significance of FDG uptake on PET scan in NSCLC. Using multivariate Cox analysis, these studies noted that SUVmax (maximum SUV within the region of interest) of the primary lesion, particularly when greater than 7 to 10, was an independent prognostic factor.

PET scanning may also prove a valuable tool for staging and evaluating patients undergoing radiation therapy and patients with NSCLC treated with chemoradiotherapy or irradiation. PET response after treatment has been found to be a powerful predictor of survival.

Adrenal gland

The adrenal gland may be the sole site of metastatic disease in up to 10% of patients with NSCLC, although patients with adrenal masses should not be assumed to have metastatic disease and should not be denied a potentially curative surgery on the basis of a scan alone. Less than 1% of adrenal masses of at least 1 cm that are negative on FDG-PET are malignant in lung cancer patients. Contrast-enhanced MRI-weighted images may assist in achieving a diagnosis. Suspicious adrenal masses should be either biopsied or resected in potentially operable patients to confirm the stage of disease.

Obtaining a Tissue Diagnosis

The next step is to obtain a histologic or cytologic diagnosis of the radiologically revealed lesion, although preoperative histologic diagnosis need not be obtained in a highly suspicious lung mass without evidence of distant or locoregional metastases (see below).

Central lesions

Collecting daily sputum cytologies for 3 consecutive days provides a cytologic diagnosis for central lesions 71% of the time and for peripheral lesions 49% of the time. In clinically suspicious lung nodules/masses, a negative sputum cytology result warrants further clinical investigation. Flexible bronchoscopy is commonly required to achieve a diagnosis. For central lesions that are exophytic, at least three direct forceps biopsies should be performed to achieve a 74% sensitivity. Washings and brushings add to the sensitivity but by themselves have a sensitivity of 48% and 59%, respectively. Further improved sensitivity is obtained with bronchoscopic fluoroscopically directed transbronchial needle aspiration biopsies. For central lesions, the overall sensitivity for flexible bronchoscopy is, in experienced hands, 88%.

In addition, bronchoscopy may provide important staging information, such as whether the tumor involves the distal trachea or carina, and may help surgeons plan the appropriate operation (lobectomy or sleeve resection vs pneumonectomy). Determining the degree of bronchial involvement assists surgical planning. Bronchoscopy-directed biopsies should be performed to assist in determining the intended line of resection, especially when evaluating for submucosal involvement. One percent to 2% of patients with lung cancer will have an endobronchial synchronous primary or metastatic lesion.

Peripheral lesions

Bronchoscopy is less likely to yield a diagnosis in patients with peripherally located lesions. Bronchoscopic ability to make a diagnosis of malignancy in peripheral lesions is dependent on size; for those less than 2 cm, the sensitivity is 33%, and for lesions greater than 2 cm, it is 62%. If a bronchiole is seen traversing or extending to the lesion on CT, the sensitivity is reported to be twice as high, nearly 60%. Newer image-guided bronchoscopic technologies are being investigated to improve diagnostic capabilities.

Electromagnetic-directed navigational bronchoscopy uses CT-derived information and corresponding anatomic features to accurately direct bronchoscopic biopsies. For patients with solitary pulmonary nodules that are 3 cm or less in maximum diameter, there is an intermediary risk of malignancy, and diagnosis can be a challenge. Electromagnetic navigation bronchoscopy may offer an additional diagnostic means of achieving brushings, washings, and directed biopsies.

A CT-guided needle biopsy may diagnose up to 90% of peripheral lung cancers but is dependent on the quality of the CT scan and the experience of the radiologist performing the procedure. The false-negative rate is 20% to 30%. Needle biopsy is usually reserved for patients who are not candidates for an operation because of distant metastatic disease or poor health or performance status. If the patient is a candidate for surgery, resection is generally recommended for any suspicious mass, whether the result of needle biopsy is positive or nondiagnostic. Therefore, for patients with a suspicious peripheral lesion that is not associated with pleural effusion, mediastinal adenopathy, or other evidence of metastatic disease, it is reasonable to proceed directly to surgery.

Mediastinoscopy

Mediastinoscopy is a time-tested technique whereby the middle (cervical mediastinoscopy) and the anterior mediastinum (Chamberlain mediastinoscopy) may be assessed for direct or metastatic lymph node involvement. In the hands of a specialist, the risk of biopsy trauma to local structures (great vessels, trachea, or esophagus), bleeding, recurrent nerve paresis, infection, or death is minimal. Whole-node biopsies may be taken, achieving a great deal of information about the location and number of nodal stations involved, number of metastatic lymph nodes within each station, degree of nodal involvement, and lymph node capsular invasion. There is no evidence that mediastinoscopic biopsy spreads tumor within the mediastinum, worsens the prognosis, or renders eventual surgical mediastinal dissection difficult. The sensitivity of mediastinoscopy is estimated to be between 85% and 92%, with a specificity of 100% and a false-negative rate of 3% to 10%.

With the advent of videoscopic technology, the video-assisted mediastinal lymphadenectomy (VAMLA) and the transcervical extended mediastinal lymphadenectomy (TEMLA) have been described, whereby extensive mediastinal evaluation and lymph node resection are performed. The additional 1 to 2 hours required to perform these procedures and slightly increased patient discomfort are offset by the significant improvement in the access to the nodal stations, acquisition of lymph nodes, and improved staging and therapeutic potential. The sensitivity of TEMLA (94%) is superior to that of mediastinoscopy, with a negligible false-negative rate.

TABLE 2: Selective indications for mediastinoscopy

Selective indications for mediastinoscopy are listed in Table 2. To assess response to therapy, repeated mediastinoscopy has been performed, with few complications. Patients with N2 disease may potentially benefit from neoadjuvant treatment. Patients with N3 disease are considered to be stage IIIB and are less likely to benefit from surgical resection. A few retrospective reports have demonstrated survival from induction therapy in patients with microscopic N3 involvement.

Endobronchial ultrasonography and endoesophageal ultrasonography staging

Endobronchial ultrasonography (EBUS) is a real-time means of directly evaluating mediastinal masses and lymph nodes to direct transtracheal and transbronchial fine-needle and core aspiration/biopsy. Often performed under general anesthesia and combined with EUS, it provides access to all of the paratracheal, subcarinal, and proximal hilar lymph node stations. It may reduce the need for mediastinoscopy.

Real-time transesophageal biopsies by endoesophageal ultrasonography (EUS) can be performed on mediastinal tumors and the lymph nodes with sensitivity and accuracy comparable to that of EBUS. EUS provides access to lymph node stations in the posterior mediastinum in the mid to distal periesophageal lymph nodes, involved in 10% to 15% of patients. Both are minimal-access procedures that can be performed with little risk. In the prospective, multi-institutional ASTER study (Assessment of Surgical Staging vs Endosonographic Ultrasound in Lung Cancer: A Randomized Clinical Trial), initiating mediastinal staging with EBUS improved the detection of mediastinal involvement and reduced the need for futile thoracotomies while decreasing the complication rate. Mediastinal staging is considered to be unnecessary in patients with obvious T4 or bulky mediastinal disease or peripheral lesions smaller than 3 cm with no PET or CT evidence of mediastinal disease.

In a group of 425 NSCLC patients at high risk for mediastinal nodal involvement, EBUS and EUS were falsely negative in 21% and 20%, respectively (Cerfolio et al: Ann Thorac Surg 2010).

Thoracentesis and thoracoscopy

Individuals who have pleural effusions should undergo thoracentesis. Video-assisted thoracoscopic surgery (VATS) should be used to assess patients who have cytology-negative effusions. Sixty percent of patients with known pleural disease and effusions will have cytology-negative effusions. However, lung cancer patients with exudative cytology-negative effusions and their cytology-positive counterparts appear to have equally poor survival. VATS permits direct visualization of the pleural surface, enables direct biopsy of the pleural nodules, and also may facilitate biopsy of ipsilateral mediastinal lymph nodes.

Measurement of serum tumor-associated antigens

This has no current role in the diagnosis or staging of NSCLC.

Several serum and blood biomarkers have been evaluated and have variable and potential roles. Serum neuron-specific enolase (NSE) levels greater than 100 μg/L are highly correlated with SCLC. NSE has been used to assess treatment response and recurrence in patients who have elevated NSE levels. Serum samples must be separated from cellular elements and run right away to avoid erroneously high levels. Levels of serum carcinoembryonic antigen (CEA) are elevated in patients with adenocarcinoma and large-cell lung cancer. CEA has a half-life of 1 to 4 days; this is delayed in patients with renal and liver dysfunction. Forty percent of adenocarcinoma patients have an elevated CEA level. Meta-analysis has found CEA to be prognostic and predictive of recurrence. Carbohydrate antigen (CA) 125 is another serum marker that is highly correlated with NSCLC, but it is not specific. Adenocarcinoma and large-cell lung cancer are correlated with serum CA 125 levels greater than 100 U/mL and serum CEA levels greater than 10 μg/L. Cytokeratin 19 fragment CYFRA 21-1 is the most sensitive serum marker for NSCLC and is particularly sensitive for squamous cell lung cancer. Elevated levels are independently prognostic and may be used to monitor response to therapy. The half-life is 1.5 to 3 hours. Levels may be erroneously low if the collection tubes are agitated, and higher in end-stage renal disease. Markers appear to be correlated with tumor burden; higher levels are more specific for disease and may be used to monitor therapy. Serum proteomics, autoantibodies, DNA fragments, miRNA, and circulating tumor cells, among other circulating biomarkers, are all under investigation for the early diagnosis, prognostication, and therapeutic monitoring.

Diagnosis and Evaluation of Suspected Carcinoid Tumor

A carcinoid tumor of the lungs may be suspected in a patient with a slowly enlarging pulmonary mass and a prolonged history of respiratory symptoms. Patients usually have no symptoms. Most tumors are located centrally and exist endobronchially. When they occur, symptoms may include wheezing, recurrent pneumonia, dyspnea, and potentially paraneoplastic syndromes. Bronchoscopy frequently assists in diagnosing the lesion. The finding of a polypoid, pale, firm mass should not lull the bronchoscopist into taking a large-forceps biopsy. First, these masses are frequently vascular and massive bleeding has been reported; second, the entire mass can be accidentally removed, making it difficult to identify the site of the original location of the polyp.

Given the bleeding potential, rigid bronchoscopy may be a better way to assess these lesions. Especially with cytology, carcinoid tumors may be difficult to differentiate from small-cell and atypical carcinoid tumors. Carcinoid tumor should be suspected when a small-cell tumor diagnosis by fine-needle aspiration does not respond to therapy. True carcinoids will have metastatic nodal disease in 5% to 10% of patients and have an excellent prognosis with surgical resection. Atypical carcinoid tumors are differentiated from typical tumors in that they have more than 5 to 10 mitotic figures per high-power field and areas of necrosis.

Unlike its infradiaphragmatic counterpart, pulmonary carcinoid tumors rarely present with paraneoplastic syndromes, including carcinoid syndrome, acromegaly, and Cushing syndrome. Therefore, it is only necessary to measure urinary 5-hydroxyindoleacetic acid (5-HIAA) excretion before surgery in symptomatic patients. Less than 3% of all patients with pulmonary carcinoid tumors have detectable urinary 5-HIAA.

Intraoperative Staging

Intraoperative staging is an integral part of any operation for lung cancer. In addition to the thorough visual and tactile inspection of the lungs, diaphragm, and pleura, the ipsilateral mediastinal lymph nodes must be either completely removed or at least sampled.

The American Thoracic Society has assigned numbered levels to locations in which lymph nodes are regularly found, defined by their relation to constant anatomic structures. A complete mediastinal lymph node dissection is associated with little morbidity and lengthens the operation only slightly.

Pulmonary Evaluation

To determine the volume of lung that can be removed without rendering the patient a pulmonary cripple and to identify individuals at risk for postoperative complications, each patient may undergo pulmonary function testing, spirometry, and potentially, an assessment of diffusing capacity. The results of pulmonary testing should be referenced to the normal values for ethnicity, height, age, and gender rather than the absolute values.

Forced expiratory volume in 1 second

Postoperative respiratory failure rarely occurs if the postresection forced expiratory volume in 1 second (FEV1) is greater than 30% of predicted. Regardless of the extent of the scheduled resection (lobectomy or pneumonectomy), if the preoperative FEV1 is less than 60% of predicted, a split-function perfusion scan may be obtained to determine the contribution of each lung region to overall pulmonary function. This information may be critical when an unplanned pneumonectomy is required to achieve complete tumor resection.

An analysis of 18,800 patients from the Society of Thoracic Surgeons National Database showed the odds ratios for morbidity and mortality for surgical resection of NSCLC patients to be unavailable and 3.9, respectively, for pneumonectomy, 2.5 and 3.1 for poor performance status (Zubrod ≥ 3), 2 and 3.6 for poor physical status (American Society of Anesthesiologists (ASA) ≥ 4), 2 and 2.1 for induction therapy; 1.9 and unavailable for lobectomy vs wedge resection, 1.7 and 2.5 for renal insufficiency, 1.6 and 1.9 for treatment with steroids, 1.6 and unavailable for thoracotomy vs VATS, 1.5 and unavailable for heart failure; 1.5 and unavailable for smoking within 30 days, 1.2 and 1.8 for every 10 years of age, unavailable and 1.7 for urgent surgery, 1.1 and 1.1 for every 10% reduced in FEV1, unavailable and 1.4 for male gender, and 0.8 and 0.7 for every increase of 10 kg/m2 in body mass index (Kozower et al: Ann Thorac Surg 2010).

Other pulmonary function tests

A diffusing capacity of the lung for carbon monoxide (DLCOa; a = adjusted for the patient's hemoglobin level) less than 60% of the predicted value or a maximum voluntary ventilation of less than 35% is associated with increased postoperative morbidity. All presurgical patients should have DLCO measured, because a surprising number have a normal FEV1 and an abnormally low DLCO that might have an impact on the surgical management. Patients with a baseline oximetry saturation of less than 90% and those who desaturate more than 4% with exercise have a greater likelihood of postoperative complications. An arterial Pco2 greater than 45 mm Hg is an independent risk factor for increased operative morbidity and mortality.

In patients with borderline lung function, further physiologic testing may be required to better estimate pulmonary reserve before and after surgery. Quantitative pulmonary ventilation-perfusion scanning may assist in this endeavor. The perfusion portion is used to calculate the percentage of lung to be removed and the estimated postoperative percentage of normal. An additional test is exercise pulmonary function testing. Patients are monitored for heart rate, rhythm, blood pressure, and oxygen consumption. Patients who reach their target heart rate and exercise capacity and who have a maximal oxygen consumption of more than 15 mL/kg/min are less likely to have a postoperative complication and mortality. In a recent report, Taylor et al concluded that marginal pulmonary function should not preclude patients from undergoing a lobectomy.

TABLE 3: WHO and IASLC guidelines for the histologic classification of lung cancer

Pathology

The World Health Organization (WHO) and the IASLC have devised guidelines for the histologic classification of lung cancer, which are revised as necessary (Table 3). The different cell-type classifications are performed using light microscopy and do not require electron microscopy or immunohistochemistry. There are variations in the natural history of the different cell types and potential differences in response to treatment and survival. Overall, 90% to 95% of NSCLCs are adenocarcinomas, squamous cell carcinomas, or large-cell carcinomas, with 3% to 4% being mixed tumors, such as adenosquamous carcinomas. Three major types of tumors are included under the NSCLC category: adenocarcinomas, squamous cell carcinomas, and large-cell carcinomas.

Adenocarcinomas

These are the most common type of NSCLC, accounting for approximately 30% to 40% of cases. Of all the types of lung cancer, adenocarcinomas are most likely to occur in nonsmokers or former smokers. They are also the most common tumors in women.

Typically, adenocarcinomas present as small peripheral lesions that have a high propensity to metastasize to both regional lymph nodes and distant sites. Because of the tendency of the primary tumor to occur in peripheral locations, it frequently produces no symptoms. In contrast with their metastatic lesions, the primary adenocarcinoma tumor is histologically heterogeneous in 80% of patients, consisting of numerous histologic subtypes.

The terms “bronchioloalveolar adenocarcinoma” and “mixed subtype adenocarcinoma” should no longer be used. Instead, new terms include “adenocarcinoma in situ” (AIS) for adenocarcinomas with pure lepidic growth and “minimally invasive adenocarcinoma” (MIA) for adenocarcinomas with predominantly lepidic growth and 5 mm or more of invasion. Both AIS and MIA are rarely mucinous. Invasive cancers are classified according to the predominant histologic pattern: lepidic, acinar, papillary, micropapillary, and solid. Variant classifications include mucinous adenocarcinomas and colloid, fetal, and enteric adenocarcinomas.

Immunohistochemistry that is specific for adenocarcinoma includes positive staining for thyroid transcription factor 1 (TTF-1), high molecular weight cytokeratin, napsin A, and mucicarmine. For pulmonary adenocarcinoma, the stain is positive for cytokeratin 7 and negative for cytokeratin 20.

Squamous cell carcinomas

These account for approximately 30% of all cases of lung cancer. These tumors tend to occur in a central location and tend to spread to regional lymph nodes; they are the most likely of all lung cancers to remain localized and to cavitate. In fact, autopsy studies have shown that about 15% to 30% of patients with squamous cell carcinoma may die of local disease without evidence of distant metastases. Immunohistochemistry for squamous cell is p63-positive, cytokeratin 5/6–positive, and TTF-1–negative.

Large-cell carcinomas

These account for approximately 10% to 15% of all lung cancers. They tend to present as relatively large peripheral lesions and, like adenocarcinomas, have a high propensity to metastasize to regional lymph nodes and distant sites.

Treatment

In operable candidates, clinically staged IA, IB, IIA, and IIB NSCLC should undergo anatomic complete surgical resection. Primarily, patients with stages IIIB and IV disease are treated nonoperatively, although some patients with metastases to the brain and adrenals alone may be candidates for curative resection if the metastatic disease can be controlled. Although multimodality therapy is routinely recommended for stage IIIA disease, it is unclear which treatment regimen is best; surgical resection does provide a clear survival advantage, and the decision to use it in these patients should be thoroughly considered.

Surgical Approach

Over the past two decades, the new subspecialty of general thoracic surgery has evolved in the United States. Surgeons in this specialty have completed their cardiothoracic surgery training and have obtained further subspecialty training in malignant and benign diseases of the chest.

The appropriate surgical treatment of NSCLC is resection of the lobe containing the tumor to achieve a negative surgical margin (R0), 16 to 20 lymph nodes pathologically assessed from at least 3 lymph node stations (2 from the ipsilateral and contralateral mediastinum), and a pleural wash of the affected side. Tumor spillage is avoided. Occasionally, a bilobectomy or pneumonectomy is required. Mortality approximates 3% following lobectomy and 7% following pneumonectomy. For peripheral lesions smaller than 3 cm with AIS or MIA, wedge resection with negative margins provides excellent long-term survival. For central lesions or lesions larger than 2 cm, a wedge or segmental resection is associated with a three- to five-times-higher incidence of local recurrence and a lower 5-year survival rate than a lobectomy. Therefore, if the patient can tolerate the procedure, the standard operation should be a lobectomy, rather than a wedge resection or segmentectomy. As a curative resection, segmentectomy has not been sufficiently evaluated, and more recent investigation suggests that in select tumors, when the bronchus and vascular supply are individually ligated with a regional node resection, survival appears to be comparable and it salvages lung parenchyma.

A retrospective evaluation based on the Surveillance, Epidemiology, and End Results (SEER) database found that in stage I lung cancer, the difference in long-term survival between a wedge resection and lobectomy was negligible after 71 years of age.

A radiologic and PET evaluation has demonstrated that patients with lesions smaller than 2 cm, peripheral ground-glass opacity with a PET maximum SUV of less than 2.5, and no evidence of metastatic disease may be considered for a more limited resection if the surgical margin is approximately the size of the lesion. A prospective randomized clinical trial (CALGB 140503; clinicaltrials.gov NCT00499330) is accruing patients to assess whether anatomic lobectomy provides superior survival compared with sublobar resection in peripheral stage I NSCLC lesions smaller than 2 cm.

At the 2014 Chicago Multidisciplinary Symposium on Thoracic Oncology, Husain et al reported a study of more than 112,000 patients with a diagnosis of stage I NSCLC using the National Cancer Data Base (NCDB). While the overall postsurgical 30-day mortality rates were low, the authors had an interesting finding when age and comorbidity were examined together. In patients younger than 75 years with medical comorbidities, the 30-day postoperative mortality rate was only 1.8%. However, in elderly patients (> 75 years) with comorbidities, this rate nearly tripled, to 4.6% (P < .01). Much of this difference was driven by elderly patients with comorbidities who underwent lobectomy, in whom the 30-day postoperative mortality rate was 5.1%, as opposed to 3.1% among patients treated with sublobar resection (P < .01). Of note, patients who underwent lobectomy were excluded from this analysis.

VATS and Robotic Resection

Traditionally, lung cancers have been resected through a posterolateral thoracotomy incision. Muscle-sparing incisions may reduce pain. The current trend is toward an even less invasive approach: lobectomy and lymph node dissection with VATS. This approach appears to offer the same cancer operation and survival with perhaps lower morbidity.

Two VATS methods have been described: the mass hilar ligation technique and individual ligation of the vasculature and airway. Patients with peripheral tumors up to 4 to 6 cm without clinical hilar or mediastinal adenopathy appear to be good candidates for a VATS procedure. Conversion rates to open thoracotomy are 10%, and hospital stays are usually 3 to 5 days.

The results of several VATS series show lower complication rates than for reported series for thoracotomy; granted, a selection bias may have occurred. One small randomized trial showed a significant benefit favoring VATS. Patients have better shoulder function, better performance on the 6-minute walk, and less impairment of vital capacity after VATS than after thoracotomy. A VATS approach may be better tolerated than other approaches for older patients.

Patients with pathologic stage IA disease have a 70% to 80% 5-year survival rate after resection, whereas 5-year survival rates are 60% in those with stage IB disease and 40% to 50% in those with stage IIA/IIB disease. Patients found to have N2 (stage IIIA) disease located at a single nodal level have a 25% to 30% 5-year survival rate. Patterns of recurrence were studied in the American College of Surgeons Oncology Group Z0030 study by Su et al that included 1,018 patients; local recurrence-free survival at 5 years was 95% for stage T1 and 91% for stage T2, and there were no differences between patients undergoing VATS or open lobectomy.

Robotic technology has been used for over 10 years to perform lung cancer resections. The expense and limited availability of the equipment continue to be obstacles, yet the number and percentage of anatomic resections continue to increase. Short- and long-term outcomes are similar. As experience increases, we may see opportunities to reduce the need for single lung ventilation, greater precision in limited resections and lymphadenectomies, and possibly reduced discomfort by minimizing the perturbation of the chest wall.

Mediastinal lymph node involvement

The standard lung cancer operation should include sampling or dissection of mediastinal lymph nodes. The presence of metastases in any of the mediastinal lymph nodes (N2 and/or N3 disease) is indicative of advanced disease and is thought by some to represent a contraindication to surgery. Resection of mediastinal disease may have prognostic significance, implications for postoperative care, and potential therapeutic value. Some series of patients with N2 disease have shown a 5-year survival rate of 20% to 30%, but patients in these series are highly selected.

Patients with N2 disease may potentially benefit from neoadjuvant treatment. Patients with N3 disease are considered to be stage IIIB and less likely to benefit from surgical resection. A few retrospective reports have demonstrated survival from induction therapy in patients with microscopic N3 involvement. The American College of Surgeons has completed accrual to a randomized, prospective study comparing survival following mediastinal lymph node sampling vs dissection. Complications and operative mortality appear equivalent between the sampling and dissection groups. Long-term survival is under investigation. Also, clinical trials are currently testing preoperative chemotherapy and chemoradiation therapy in patients with mediastinal node involvement.

Preoperative histologic assessment of the mediastinal lymph nodes is essential if multilevel metastases are suspected, because there have been few long-term survivors among patients with metastatic disease at more than one level. Nonsurgical treatment appears preferable, or patients should be offered participation in a trial designed to assess the benefits of neoadjuvant therapy. Although patients with stage IIIB tumors are usually treated with irradiation and chemotherapy (see later discussions), the occasional patient with isolated involvement of the vena cava or atrium can undergo resection.

Sidebar: At the 2013 ASCO meeting, Sorensen et al presented the results of a prospective randomized phase III trial by the Nordic Thoracic Oncology Group, studying the role of surgery for pathologically verified N2 NSCLC. The chemotherapy involved paclitaxel and carboplatin. Of note, the study was stratified by histology. Overall, the trial showed no improvement in survival with the addition of surgery to chemotherapy and radiation alone. However, there was a significant benefit in survival in the surgical arm for patients with adenocarcinoma (as well those with stage T1N2 disease). Of note, the chemotherapy and radiation in this trial were administered sequentially, not concurrently. While N2 status was confirmed preoperatively, the N2 nodal localizations and numbers were not recorded.

Carcinoid tumors

Although the majority of carcinoid tumors remain localized, regional lymph node metastases are identified in 5% to 10% of patients. The surgical approach, therefore, should be similar to that used in NSCLC. If a small tumor in a proximal airway is identified and there is no histologic evidence of lymph node disease, then a bronchoplastic procedure with preservation of lung tissue can sometimes be performed. Rates of survival at 10 years are greater than 90% for patients with stage I disease and 60% for patients with stage II disease.

Adjuvant Therapy

Chemotherapy

Classic postoperative adjuvant chemotherapy has been tested in three randomized trials conducted by the Lung Cancer Study Group (LCSG). For almost 20 years, the relative value of adjuvant chemotherapy for resectable NSCLC has been disputed and debated. A randomized prospective study involving 488 patients (Keller et al: N Engl J Med 2000) showed no benefit to adjuvant chemotherapy. The ALPI (Adjuvant Lung Project Italy) study of 1,209 patients also showed no survival benefit. In contrast, the IALT (International Adjuvant Cancer Trial) randomized 1,867 patients to receive cisplatin-based adjuvant chemotherapy or no treatment. At 5 years, the treatment arm showed a survival advantage of 4.1% (P = .003), compared with the observation arm. However, that benefit appeared to diminish with longer follow-up. With 3 additional years of follow-up, there was a significant difference between the results of overall survival before and after 5 years (HR = 0.86; P = .01 vs HR = 1.45; P = .04). Similar results were observed with disease-free survival and confirmed the efficacy of chemotherapy for the first 5 years after surgery; however, they suggested possible late adjuvant chemotherapy-related mortality and underscored the need for long-term follow-up of adjuvant lung cancer trials.

Two trials, the Canadian JBR-10 and CALGB 9633, both demonstrated initial clinically significant improvement in survival, with minimal chemotherapy side effects; cisplatin and vinorelbine were used in the BR-10 trial and carboplatin and paclitaxel were used in CALGB 9633. These results, in combination with the recent positive findings of the adjuvant trial of UFT (a drug composed of tegafur and uracil mixed at a 1:4 ratio) in patients with stage IB NSCLC, increase the likelihood of adjuvant platinum-based therapy becoming the standard of treatment for patients with stages IB-IIIB NSCLC. Recent results of a prospective phase III trial from the Spanish Lung Cancer Group (SLCG) NATCH trial (Neoadjuvant Taxol Carboplatin Hope) compared the use of adjuvant and neoadjuvant and surgery alone in 616 patients with stages I (> 2 cm), II, and T3 N1. For those who received chemotherapy, three 3-week cycles of paclitaxel (200 mg/m2) and carboplatin (at an AUC of 6) were given. Despite an excellent safety profile, a major radiologic response at 59%, and a complete response of 9% of induction patients, there was no difference found in survival. Again, it appears that the survival advantage in the treatment arm was lost, except in patients with larger primary tumors (> 4 cm) or more advanced disease (stage IIIA).

A meta-analysis by the Lung Adjuvant Cisplatin Evaluation (LACE) group found that in completely resected NSCLC patients, the 5-year overall survival benefit of cisplatin-containing adjuvant regimens was 5.4% (HR = 0.89; P = .004). There appeared to be a survival advantage of 8.9% for cisplatin-vinorelbine, which was superior to other regimens containing cisplatin. The greatest benefit of adjuvant therapy was seen in patients having disease of a higher stage (14.7% for stage III, 11.6% for stage II, 1.8% for stage I; Douillard et al: J Clin Oncol 2010).

Based on these results from adjuvant chemotherapy trials, the initial enthusiasm has been somewhat tempered. The potential benefits are higher efficacy of chemotherapy early in the natural history of disease, facilitation of subsequent local therapy, and early eradication of distant micrometastases. Further investigation in the use of adjuvant therapy is under way to determine which patients will benefit and what the optimal treatment is.

Stage I disease. Currently approved chemotherapeutic regimens do not appear to provide any survival benefit in the stage IA patient population and are not advised outside of the clinical trial setting. From the subset analysis of CALGB 9335, patients with stage IB tumors larger than 4 cm may have some benefit, given the likelihood for recurrence. Currently under investigation is the use of tumor histology and marker evaluation to select the appropriate patient and the ideal chemotherapeutic plan with the greatest efficacy and least toxicity.

JBR.10, however, which included 532 patients with stage IB or II completely resected cancers who were randomized to cisplatin/vinorelbine or observation, demonstrated that the survival benefit appears to be confined to N1 patients over 9 years of prolonged follow-up.

The current trend is to provide involvement in a chemopreventive clinical trial for patients with stage IA disease. For patients with stage IB disease, particularly those with tumors larger than 4 cm in diameter, platinum-based adjuvant chemotherapy (cisplatin or carboplatin combined with a taxane or vinorelbine) or involvement in an adjuvant therapy clinical trial should be strongly considered.

Stage II/III disease. In two trials, postoperative adjuvant chemotherapy with six courses of CAP (cyclophosphamide, Adriamycin [doxorubicin], and Platinol [cisplatin]), given alone in one study and following postoperative radiation therapy in the other, resulted in a modest improvement in median survival but had no impact on long-term survival. In contrast, the IALT, ANITA, CALGB 9623, and Canadian BR-10 trials demonstrated an initial clinically significant survival advantage, justifying consideration for adjuvant chemotherapy. In contrast to IALT, which included all resectable disease and showed a decrease in survival after 5 years, JBR.10, which included stage IB/II patients, continues to show a survival benefit with over 9 years of follow-up. While there does not appear to be a clear survival advantage to adding adjuvant mediastinal radiotherapy to chemotherapy, this is reasonable to consider when there are particularly high-risk factors for local recurrence. Potential indications for postoperative radiotherapy include:

• Mediastinal involvement

• Multiple positive lymph nodes or lymph node stations

• Extracapsular nodal extension

• Bulky nodal disease

• Grossly positive surgical margins

• Close or microscopically positive margins

Radiation therapy

A trial conducted by the LCSG showed that in patients with squamous cell carcinoma of the lungs and resected N1/N2 disease, administration of postoperative radiation therapy reduced the risk of recurrence in the chest from 20% to 1%. Although there was no improvement in overall survival, postoperative irradiation was associated with a significant improvement in disease-free survival for patients with N2 disease. A trial by the British Medical Research Council (BMRC) reached similar conclusions.

A meta-analysis of nine randomized trials assessing postoperative radiation therapy in lung cancer reported a 21% increase in mortality in patients receiving this therapy. However, many of the patients in these trials had N0 disease, for which few would advocate radiation therapy. Also, many of the patients were treated with cobalt-60 beams and technically limited treatment planning, not with modern radiation therapy techniques.

The impact of radiation therapy on survival was reanalyzed in the ANITA trial (Adjuvant Navelbine International Trialist Association), in which postoperative radiation therapy was optional, based on each center's practice. Interestingly, in patients with N2 (mediastinal) disease, adjuvant radiation therapy improved overall survival in both arms: 21% vs 17% in the observation arms and 47% vs 34% in the adjuvant chemotherapy arms (with or without radiation therapy, respectively). Based on this descriptive analysis, radiation therapy appeared to provide a benefit in addition to adjuvant chemotherapy in patients with N2 disease (whereas the converse was true in patients with N1 disease). A randomized trial with and without postoperative radiation therapy in patients with resected N2 disease is under way in Europe.

These results created a lack of consensus about treatment recommendations, with some experts advocating the use of postoperative radiation therapy to reduce local recurrence, and others avoiding it because of the absence of an effect on survival.

At present, therefore, the appropriate role of postoperative radiation therapy remains controversial. Such therapy should be seriously considered, however, in patients at high risk for locoregional relapse, in patients with resected N2 disease, or in those with other possible indications as outlined above. In patients with no known residual disease after surgery who are receiving adjuvant chemotherapy, it is reasonable to administer the chemotherapy first (because it has been associated with a survival benefit) followed by radiation therapy (for enhanced local tumor control).

Moreover, in a randomized trial (Keller et al: N Engl J Med 2000), no benefit was shown for concurrent chemoradiation therapy over radiation therapy alone in the adjuvant setting for completely resected patients with stage II/IIIA NSCLC. An exception may be in patients with microscopic residual disease (ie, positive margins or extracapsular extension), for whom a delay in radiation therapy may be detrimental.

Neoadjuvant Chemotherapy or Chemoradiation Therapy

During the past decade, numerous phase II trials showed that, in general, it is feasible to perform pulmonary resection following chemotherapy or chemoradiation therapy. Although surgery can be more difficult after preoperative treatment, morbidity and mortality are acceptable in experienced hands.

Stage IIIA/IIIB disease

The greater effectiveness of current chemotherapeutic regimens to reduce disease bulk suggested that their use before surgery, either alone or in combination with radiation therapy, might increase both resectability and survival in patients with stage IIIA NSCLC. For these patients, such an approach has been shown to be feasible; however, it is not clear whether such a strategy improves median or long-term survival over nonsurgical chemoradiotherapy .

In 2009, Albain et al reported on Intergroup 0139, a phase III study of 202 patients with stage IIIA disease that showed improved progression-free survival in the trimodality group with neoadjuvant chemoradiation followed by surgery (12.8 months vs 10.5 months; P = .017) but no difference in overall survival. The lack of survival benefit was thought to be possibly related to the high mortality rate following pneumonectomies, particularly on the right side.

Current recommendations

In select patients, preoperative treatment may have a favorable effect on outcome in surgically resectable stage IIIA NSCLC. Although aggressive neoadjuvant approaches may have increased treatment-associated mortality, in experienced institutions, potential benefits seem to outweigh the risks. The results of the Intergroup randomized trial comparing preoperative chemoradiation therapy with definitive chemoradiation therapy (in pathologic N2 disease) showed a significant improvement in progression-free survival (but not overall survival) in the surgical arm (P = .02). This approach, however, may not be optimal if a pneumonectomy is required, because this procedure was associated with a high rate of treatment-related deaths (> 20%). In a subset analysis, survival was improved for patients who underwent lobectomy, but not pneumonectomy, compared with chemotherapy plus radiation alone. In this trial, it is not clear whether patients with persistently positive N2 disease after neoadjuvant therapy will benefit from surgical resection.

Stage I-IIIA disease

Neoadjuvant chemotherapy may have a role in early-stage disease. A multicenter trial from France randomized 373 patients with stage I to IIIA NSCLC to undergo either surgery alone or chemotherapy (mitomycin [6 mg/m2 on day 1], ifosfamide [1.5 g/m2 on days 1 to 3], and cisplatin [30 mg/m2 on days 1 to 3]) at 3-week intervals for three cycles followed by surgery. Disease-free survival was significantly longer in the patients randomized to receive neoadjuvant chemotherapy than in those treated with surgery alone (P = .02). The most striking benefit of chemotherapy was seen in patients who had minimal lymphadenopathy (either N0 or N1; P = .008). No excessive complications were seen in the chemotherapy-treated patients.

Cumulative data support a difference in treatment, but no one study is statistically significant. Pooled analyses suggest a positive benefit from neoadjuvant therapy, with all individual studies showing a trend toward improved overall survival, and significant differences in disease-free survival.

Following on previous trials was a study presented by Scagliotti in which 270 patients with stage IB-IIIA NSCLC were randomized to undergo surgery and follow-up or induction chemotherapy with cisplatin and gemcitabine (Gemzar) followed by surgery. This trial closed early because of accrual problems, after only 270 of a projected 700 patients enrolled. Overall survival trends at 3 years favored the chemotherapy plus surgery arm (P = .053). Overall survival at 3 years for the neoadjuvant arm was 67% vs 60% for surgery alone (P = .053).

A subset analysis suggested that for stage IIb-IIIa, 3-year overall survival was 70% vs 40% in favor of chemotherapy (P = .001), whereas there was no difference for patients with stage Ib-IIa who were treated. This continues to support data suggesting that for stage IIb-IIIa disease, chemotherapy, either in the adjuvant or neoadjuvant setting, provides a significant survival advantage. In 2015, Pless et al reported the findings of the Swiss SAKK Lung Cancer Project Group's phase III study in patients with stage IIIA (N2) disease, including 7% with bulky mediastinal disease. Operable patients were randomized to induction cisplatin and docetaxel followed by surgical resection with or without concurrent induction radiation therapy, at 44 Gy. Median overall survival time was 26.2 months for the no induction radiation group and 37.1 months for the induction radiation group; although radiation therapy appeared to confer superior survival, the difference in survival outcomes between the two groups was not statistically significant. Median event-free survival time was 22 months for patients in the induction radiation therapy arm vs 15.4 months for patients in the arm with no induction radiotherapy. The R0 rate appeared to be greater for the concurrent-radiation group, at 90% vs 81% for the chemotherapy group. Nodal downstaging was 64% in the induction radiation therapy group vs 53% in the non–induction radiation group. Pathological complete response rates were 16% in the induction group vs 12% in the non–induction radiation therapy group, respectively. The authors found that the addition of concurrent induction radiation therapy did not appear to offer any survival advantage for the stage IIIA (N2) patients treated with induction chemotherapy followed by surgery. They concluded that either local therapy should be used, radiotherapy or surgery; but not both. The accompanying editorial questioned the validity of the results.

Also in 2015, in Thorax, McElnay et al published results of a meta-analysis and systematic review of randomized trials assessing the impact of surgical resection in patients with IIIA (N2) disease. After elimination of some of the trials based on exclusion criteria, the authors evaluated a total of six trials with 868 patients. There appeared to have a trend toward superior survival among patients randomized to surgery after chemoradiotherapy, compared with patients who were randomized to chemoradiation alone (HR = 0.87; P = .068).

Treatment of Patients With Medically Inoperable Stage I/II Disease

Some patients with resectable stage I or II NSCLC are high-risk operative candidates because of poor cardiopulmonary function or other medical problems. Other patients refuse to undergo surgery despite the recommendation of their treating physicians. In such patients, an attempt should be made to optimize pulmonary function by encouraging smoking cessation and initiating vigorous treatment, such as with bronchodilators. Evaluation by an experienced general thoracic surgeon is warranted before discounting a surgical approach. For many of these patients, stereotactic body radiation therapy (SBRT) has emerged as a safe and effective noninvasive treatment option (see discussion below).

Radiation therapy

Several institutions have reported their experience with conventional radiation therapy for such patients. Although the results are not as good as those reported in patients selected for surgery (possibly due to differences in patient selection and between clinical vs pathologic staging), patients with medically inoperable early-stage NSCLC clearly should be offered radiation therapy, with a reasonable expectation of cure. For patients with early-stage (I/II) NSCLC, surgical resection is the standard of care. However, many patients with lung cancer are not eligible for surgery because of various comorbidities. In the past, the main treatment alternative for these medically inoperable patients has been conventional radiation therapy, typically requiring once-daily treatments every weekday for approximately 7 weeks. Results with conventional radiation therapy have not been encouraging, with local tumor control rates of only 50%.

SBRT has evolved as another treatment option for patients with medically inoperable, peripheral stage I NSCLC. SBRT involves the delivery of high doses of highly precise radiation therapy in a hypofractionated scheme generally over three to five fractions, administering biologically equivalent doses (BEDs) greater than 100 Gy over 1 to 2 weeks, with promising local tumor control rates of greater than 90%.

Onishi et al reported the clinical outcomes of approximately 250 patients treated in Japan with SBRT for stage I NSCLC. They reported a cumulative 5-year local tumor control rate of 84% for those treated with a BED of greater than 100 Gy vs 37% for those who received a BED of less than 100 Gy (P < .001). Interestingly, among operable patients, the 5-year overall survival rate was 72% for those treated with a BED of greater than 100 Gy vs 50% for those who received a BED of less than 100 Gy (P < .05)

Videtic et al recently published the American College of Radiology (ACR) Appropriateness Criteria for early-stage NSCLC, with a focus on patients with localized, mediastinal lymph node–negative NSCLC. For medically operable patients, surgical resection is the standard of care. For those designated high-risk surgical patients due to medical comorbidities, there is emerging evidence demonstrating the availability of low-toxicity curative therapies, such as stereotactic body radiation therapy (SBRT). The American College of Radiology Appropriateness Criteria are evidence-based guidelines that are reviewed by a multidisciplinary expert panel. The guideline development and review processes include an extensive analysis of current medical literature from peer-reviewed journals and the application of a well-established consensus methodology, to rate the appropriateness of imaging and treatment procedures by the panel. Other ACR Appropriateness Criteria recently published for NSCLC include guidelines by Chang et al in 2014, on locally advanced NSCLC with good performance status/definitive intent, and Appropriateness Criteria by Willers et al in 2015, regarding induction and adjuvant therapy for N2 NSCLC.

Timmerman et al initially reported the results of a phase I study of SBRT in patients with medically inoperable stage I NSCLC. SBRT was delivered in three fractions over 2 weeks, with a starting dose of 800 cGy per fraction. The dose was escalated to 2,000 cGy per fraction, for three fractions. Recently, Timmerman et al reported in JAMA on a phase II North American multicenter study (RTOG 0236) of 59 patients with biopsy-proven, peripheral T1-2, N0, M0 NSCLC tumors (measuring < 5 cm in diameter) and medical conditions precluding surgical treatment. Patients were treated with a total of 54 Gy in three fractions over a 1- to 2- week period, with a median follow-up of 34.4 months. The rates for disease-free survival and overall survival at 3 years were 48.3% (95% CI, 34.4%–60.8%) and 55.8% (95% CI, 41.6%–67.9%), respectively. The estimated 3-year primary tumor control was 97.6%, and the median overall survival was 48.1 months. The 2-year local control rate was excellent at 94%.

At ASTRO 2014, Timmerman et al presented the long-term results of RTOG 0236 (a phase II trial of SBRT for medically inoperable stage I NSCLC). Among the 55 evaluable patients with a median follow-up of 4 years (7 years for surviving patients), the median overall survival was 4 years in this group of medically inoperable patients. The estimated 5-year primary tumor failure rate was 7%. A total of nine additional patients had a recurrence within the involved lobe with a 5-year primary, and an involved lobe failure rate of 20%. The 5-year disseminated failure rate was 31%. The rates of disease-free and overall survival at 5 years were 26% and 40%, respectively. An excess of late apparent toxicity was not observed, and there were no grade 5 adverse events.

Although this treatment (ie, SBRT using three fractions of 18 to 20 Gy each) was safe and effective for patients for peripheral lung lesions, there was a significantly higher rate of grade 3 toxicity in patients with central lung lesions. Other institutional studies, such as those conducted at Henry Ford Hospital, suggest that early central NSCLC lesions can be treated safely and effectively with SBRT by using smaller fraction sizes of 10 to 12 Gy per session for three or four sessions. A prospective trial, RTOG 0813, is studying this issue further in a dose-escalation trial (ClinicalTrials.gov Identifier: NCT00750269).

The role of lung SBRT for operable patients is not known. A recent retrospective analysis demonstrated a promising 5-year overall survival of 51% in patients with clinical stage I operable NSCLC (with a median age of 76 years) with a 30-day post-SBRT mortality rate of 0%. Unfortunately a lung cancer trial (ACOSOG Z4099/RTOG 1021), randomizing patients with high-risk stage I NSCLC to sublobar resection (with or without brachytherapy) or SBRT, closed due to poor accrual, so this important question remains unanswered.

At the 2014 Chicago Multidisciplinary Symposium on Thoracic Oncology, Mancini et al reported that elderly patients undergoing SBRT for medically inoperable early-stage NSCLC achieve outcomes similar to those observed in younger patients. They identified 253 patients with a median age of 75 years. At median follow-up of 25 months, they found no significant difference in 2-year survival or local recurrence–free survival between patients older vs younger than 75 years of age. There was a significant difference in 2-year distant recurrence–free survival (90% vs 75%, P = .02), favoring patients older than 75 years. There was no difference in acute or late grade 2 and 3 toxicity by age. Overall, elderly patients (> 75 years) undergoing SBRT achieved similar outcomes and had similar toxicity rates, compared with younger patients.

Radiofrequency Ablation

Radiofrequency ablation is being actively studied in patients who are not operative candidates. While radiofrequency ablation may be an option for select patients who have peripheral lesions less than 3 cm without mediastinal disease and who are poor surgical candidates, further studies are needed with longer follow-up. Although there is considerable experience with use of radiofrequency ablation in patients with cancer in other organs, further evaluation of radiofreqency ablation and cryotherapeutic techniques for lung cancer is under way to assess complications related to the therapy (including the known risk of pneumothorax), local recurrence, and survival.

In an analysis comparing radiofrequency ablation with stereotactic body radiation therapy, Bilat et al reported that while survival at 1 and 3 years was similar between patient groups, the 5-year survival was higher with stereotactic body radiation therapy (47%) than with radiofrequency ablation (20% to 27%). Local progression rates were lower in patients treated with stereotactic body radiation therapy (4% to 15% vs 24% to 43%). Pneumothorax (19% to 63%) was the most common complication following radiofrequency ablation. Pneumonitis (2% to 12.5%) and chest wall pain (3% to 12%) occurred following stereotactic body radiation therapy. In a separate study from Harvard by Lanuti et al, 55 radiofrequency ablation procedures were performed in 45 patients with stage I NSCLC. At a median follow-up of 32 months, locoregional recurrence occurred in 21 patients (38%). Eighty percent of tumors larger than 3 cm were associated with locoregional recurrence.

Treatment of Patients With Stage IIIA/IIIB Disease

Radiation therapy

In the past, radiation therapy was considered the standard therapy for patients with stage IIIA or IIIB disease. Long-term survival was poor, in the range of 5%, with poor local tumor control and early development of distant metastatic disease.

A randomized trial compared standard daily radiation therapy (66 Gy) with a continuous hyperfractionated accelerated radiation therapy (CHART) regimen that delivered 54 Gy over 2.5 weeks. The altered fractionation schedule resulted in improved 2-year survival.

Various efforts have assessed combining altered fractionation schema with chemotherapy. Although the results of RTOG 94-10 did not favor altered fractionation (see section on “Concurrent vs sequential chemoradiation therapy”), the long-term results of another study support this strategy.

A randomized trial of approximately 200 patients (Jeremic et al: Proc Am Soc Clin Oncol 2000) compared hyperfractionated radiation therapy (twice daily, to 69.6 Gy) and concurrent low-dose daily carboplatin/etoposide with or without weekend carboplatin/etoposide. Although investigators found no benefit with the addition of weekend carboplatin/etoposide, both arms demonstrated promising median survival times of 20 and 22 months and excellent 5-year survival rates of 20% and 23%.

Conformal radiation therapy

In the Michigan phase I dose-escalation trial of three-dimensional (3D) conformal radiation therapy for NSCLC (Hayman et al: J Clin Oncol 2001), the radiation dose was escalated based on the effective volume of irradiated lung (up to 102.9 Gy). Such doses produced acceptable toxicity and no cases of isolated failures in purposely nonirradiated, clinically uninvolved nodal regions.

Socinski et al reported a dose-escalation radiotherapy (from 60 Gy up to 74 Gy) trial, using 3D computer-assisted planning techniques, in patients receiving induction carboplatin and paclitaxel and concurrent weekly carboplatin/paclitaxel. Ninety-seven percent of the patients (31 of 32) completed therapy to 74 Gy, as planned. The grade 3/4 esophagitis rate overall was relatively low at 11%. Moreover, the results showed a promising median survival of 26 months and a 3-year survival of 47%. Recently, studies are exploring newer radiotherapy techniques, such as proton therapy. In one prospective study, 44 patients were treated with 74 Gy (RBE) proton therapy with weekly carboplatin/paclitaxel and had a promising median survival of 29 months. Future studies are needed to further test such new technologies.

Indeed, several analyses supported the importance of radiation therapy dose escalation in stage III NSCLC. For example, researchers at the University of Michigan found a significantly improved survival in patients treated with a higher BED. To test this important hypothesis prospectively, an RTOG/Intergroup phase III trial has compared concurrent chemoradiation therapy with either 60 Gy or 74 Gy with or without cetuximab (Erbitux). However, the data monitoring committee for this study closed the high-dose arms because of a futility analysis that showed no benefit to a higher radiation dose (74 Gy vs 60 Gy) in this setting. RTOG is exploring new strategies (involving functional imaging) to more effectively escalate the radiation dose in this setting.

In 2015, Bradley et al published the results of RTOG 0617, a randomized phase III study of standard-dose vs high-dose conformal radiation with concurrent and consolidated carboplatin plus paclitaxel with or without cetuximab for patients with stage III NSCLC. They randomly assigned patients to receive either 60 Gy (standard dose), 74 Gy (high dose), 60 Gy plus cetuximab, or 74 Gy plus cetuximab (in addition to concurrent chemotherapy). Overall, they found that high-dose (74 Gy) radiation with concurrent chemotherapy was not better than 60 Gy plus concurrent chemotherapy for patients with stage III NSCLC, and might be potentially harmful (given that it was associated with a lower survival rate). The addition of cetuximab to concurrent chemoradiation and consolidation treatment provided no benefit in overall survival for these patients.

Movsas et al presented the quality-of-life (QOL) analysis of the RTOG 0617 study during the plenary session at ASTRO 2013. QOL data were collected prospectively by a validated lung cancer instrument: the Functional Assessment of Cancer Therapy-Trial Outcome Index (FACT-TOI), which includes the Physical Well-Being (PWB), Functional Well-Being (FWB), and the Lung Cancer Subscale (LCS). The primary QOL hypothesis predicted for a clinically meaningful decline (CMD) in the LCS on the high-dose radiation therapy arm at 3 months. Indeed, the CMD in LCS for the 74 Gy arm was significantly higher at 3 months than for the 60 Gy arm (P = .024), confirming the primary QOL hypothesis. This difference in QOL decline between arms resolved by 12 months. The baseline QOL significantly predicted for survival on multivariate analysis. While this study was not randomized by technology, intensity modulated radiation therapy (IMRT) was used to treat patients at higher cancer and with larger-volume tumors. Nevertheless, significantly fewer patients who received IMRT (vs 3D conformal radiation therapy) had a clinically meaningful decline in the LCS at 12 months (P = .005), suggesting that improved RT treatment techniques may help enhance the therapeutic window of patients with lung cancer.

at tumor radia

At the 2014 Chicago Multidisciplinary Symposium on Thoracic Oncology, Rimner et al. analyzed the optimal radiation dose in stage III non-small cell lung cancer patients treated without concurrent chemotherapy. In an analysis of almost 600 patients, they found that higher radiation doses appeared to be associated with improved overall survival in patients who were not eligible for concurrent chemoradiation. Patients who received ≥ 70 Gy had the highest 3-year overall survival of 42% vs 37% for those who received > 60 Gy and vs 26% for those who received < 60 Gy. On multi-variable analysis for those treated with radiation alone, only higher radiation dose remained as a strong predictor for better overall survival (P = .02). For patients treated with concurrent chemoradiation, however, there appeared an improvement in overall survival with increasing radiation doses from < 60 Gy to 60 Gy (3-year overall survival 32% vs 41%), but not as much for dose escalation > 60 Gy (3-year overall survival of 45%). Radiation dose was not a significant predictor for outcome on multivariable analysis in patients receiving chemoradiation.

At ASTRO 2014, Pham et al studied whether radiation appears to attribute to cardiac specific death (CSD) in patients with stage III lung cancer. RTOG 0617 reported that a factor predictive of less favorable survival on multi-variable analysis included the heart volume treated to greater than 5 Gy. Using the SEER database, they analyzed over 60,000 patients & found that the median cardiac cause specific (CCS) survival was not significantly different between patients who received definitive radiation compared to definitive surgery. Multi-variable analysis revealed increased age, African American race, & male gender were associated with an increased risk of CCS mortality. However, the risk of CCS death due to radiation was not increased compared to those who received surgery in patients with stage III lung cancer. Adjunct radiation did not increase the risk to CCS death in patients who had surgery for lung cancer.

Chemoradiation therapy

Chemoradiation therapy vs radiation therapy alone. At least 11 randomized trials have compared thoracic irradiation alone with chemoradiation therapy in patients with stage III NSCLC. Several meta-analyses have demonstrated a small but statistically significant improvement in survival with the combined-modality regimens.

Concurrent vs sequential chemoradiation therapy. A phase III study (Furuse et al: J Clin Oncol 1999) evaluated mitomycin, vindesine, and cisplatin (MVP), administered either concurrently or before thoracic irradiation (56 Gy), in patients with unresectable stage III NSCLC. With more than 300 patients randomized, survival favored concurrent over sequential therapy (median survival, 16.5 months vs 13.3 months, and 5-year survival rates, 15.8% vs 8.9%; P = .04). The study also reported the patterns of failure, which demonstrated a benefit of concurrent chemoradiotherapy in improving local relapse-free survival (P = .04) but not the distant relapse-free survival (P = .6).

Curran et al presented long-term results of a larger randomized trial (> 600 patients) comparing sequential and concurrent chemoradiotherapy (RTOG 9410). The 4-year survival with concurrent cisplatin/vinblastine and once-daily irradiation was 16% vs 10% with sequential treatment (P = .04). The result in the third treatment arm (concurrent cisplatin/oral etoposide and hyperfractionated irradiation) was intermediate, with a 4-year survival of 17%.

The Hoosier Oncology Group (HOG) reported the results of the phase III trial of cisplatin plus etoposide plus concurrent chest radiation therapy with or without consolidation docetaxel (Taxotere) in patients with inoperable stage III NSCLC. In a study of 203 patients (of whom 147 were randomized to receive or not receive docetaxel) with a median follow-up of 25.6 months, they reported a median survival time in the docetaxel arm of 21.6 months vs 24.2 months in the observation arm (P = .94). They concluded that consolidation docetaxel does not further improve survival, is associated with significant toxicity (including an increased rate of hospitalization and premature deaths), and should not be used in this setting.

At the 2014 Chicago Multidisciplinary Symposium on Thoracic Oncology, Kim et al analyzed the outcomes of stage III NSCLC treated with carboplatin or cisplatin-based concurrent chemoradiation therapy using SEER-Medicare data. In an analysis of more than 2,800 patients, carboplatin vs cisplatin-based concurrent chemoradiation regimen had comparable overall survival and cause-specific survival. The mostly commonly used regimens (carboplatin/paclitaxel vs cisplatin/etoposide) also had comparable outcomes.

Current treatment recommendations

At present, it is reasonable to consider concurrent chemoradiation therapy (with once-daily radiation therapy) as a treatment paradigm in stage III (inoperable) lung cancer patients with an ECOG performance status of 0/1 who have not lost more than 5% of their usual body weight. A radiation dose of 74 Gy was associated with a lower survival than 60 Gy. The results of a meta-analysis using individual patient data from randomized clinical trials showed that concomitant chemoradiation therapy, as compared with sequential chemotherapy and radiation therapy, improved survival of patients with locally advanced NSCLC, primarily due to a decrease in locoregional tumor progression. Overall, there was an absolute benefit in survival of 6% at 3 years.

Novel biologic agents. SWOG 0023 was a randomized phase III trial of gefitinib (Iressa) vs placebo maintenance after chemoradiation therapy followed by docetaxel in patients with stage III NSCLC. With a median follow-up of 27 months in 243 randomized patients, the median survival for the gefitinib arm (n = 118) was 23 months vs 35 months for the placebo arm (n = 125; P = .013). The authors concluded that in this unselected population, gefitinib did not improve survival. The decrease in survival was likely due to cancer, not to gefitinib toxicity. A secondary analysis of SWOG 0023 found that V20 (which represents the percentage of total normal lung volume that received > 20 Gy of radiation) was the greatest predictor of radiation pneumonitis. This finding is consistent with other studies showing V20 to be a reliable predictor of pneumonitis in patients who are treated with radiation therapy for lung cancer. They found that patients with a V20 of greater than 35% had a higher rate of grade 3 or higher pneumonitis than patients with a V20 of less than 35% (14% vs 6%, respectively; P = .02) and a shorter median survival (11 vs 23 months, respectively; P = .0001). Of note, there was no imbalance of the V20 parameter between the two arms of SWOG 0023. It is now considered standard for radiation oncologists to formally evaluate dose vs volume parameters in the treatment of lung cancer.

Blumenschein et al combined cetuximab with paclitaxel and carboplatin with radiation therapy. This phase II RTOG trial accrued 87 patients and reported a median survival of 23 months. However, this novel strategy of adding cetuximab to chemoradiation therapy in locally advanced NSCLC was tested in a randomized phase III study (RTOG 0617) and showed no survival advantage by adding cetuximab to standard chemoradiation.

Prophylactic Cranial Irradiation

RTOG 0214 studied the role of prophylactic cranial irradiation (PCI) in patients with stage III NSCLC. The total accrual was 356 patients (of the targeted 1,058), of whom 340 were eligible. One-year overall survival (75.6% vs 76.9%; P = .86) and 1-year disease-free survival (56.4% vs 51.2%; P = .11) for PCI vs observation, respectively, were not significantly different. However, the incidence of CNS metastases at 1 year was 7.7% for PCI vs 18% for observation (P = .004). While PCI in stage III NSCLC significantly decreased the risk of CNS metastases, there was no significant difference in overall survival or disease-free survival. Moreover, PCI was associated with some decline in immediate and delayed recall at 1 year, and thus is not considered to be standard of care in patients with NSCLC.

Treatment of Patients With Stage IV Disease

In 2014, Ashworth et al reported an individual patient data meta-analysis of outcomes and prognostic factors after treatment of limited or oligometastatic NSCLC. They obtained individual data on more than 750 NSCLC patients with one to five synchronous or metachronous metastases treated with surgical metastectomy, stereotactic radiotherapy/radiosurgery, or radical external-beam radiation, as well as curative treatment for the primary lung cancer. The authors reported median overall survival of 26 months, 1-year survival of 70%, and a 5-year survival of 29.4%. In the recursive partitioning analysis (RPA), they found three risk groups: low-risk, metachronous metastases (5-year survival: 48%); intermediate risk, synchronous metastases and N0 disease (5-year survival: 36%); and high risk, synchronous metastases and N1/N2 disease (5-year survival: 14%). This study suggests that long-term survival is common in selected patients with metachronous oligometastases.

Until relatively recently, there was considerable controversy over the value of treating stage IV NSCLC patients with chemotherapy. Treatment with older cisplatin-containing regimens, such as cisplatin/etoposide, showed only a modest effect on survival, improving median survival by approximately 6 weeks, according to a meta-analysis, and yielding a 1-year survival rate of approximately 20% (as compared with a rate of approximately 10% for supportive care).

TABLE 4: Active newer agents for NSCLC chemotherapy

TABLE 5: Results of selected randomized trials evaluating chemotherapy regimens of newer agents in advanced NSCLC

However, several chemotherapeutic agents have produced response rates in excess of 20% in NSCLC (Table 4). The potentially useful newer cytotoxic agents include the taxanes (paclitaxel and docetaxel), vinorelbine, gemcitabine, irinotecan, and pemetrexed. Several of these drugs have unique mechanisms of action. Paclitaxel and docetaxel increase the polymerization of tubulin, gemcitabine is an antimetabolite, and irinotecan is a topoisomerase I inhibitor.

Furthermore, randomized trials demonstrated that a combination of a newer agent plus cisplatin significantly improves the response rate over cisplatin monotherapy (historically considered the most active agent against NSCLC). This increase in response rate translates into a significant, although modest, improvement in survival.

Optimal chemotherapy for advanced NSCLC

Until the early 1990s, regimens of cisplatin plus a vinca alkaloid or etoposide were most common. More recently, regimens that employ newer agents are more widely used. However, choosing one regimen from many options is a difficult task because there is no survival advantage documented for one regimen over another or for the standard regimen vs regimens containing newer agents.

Perhaps most importantly, it is now recognized that NSCLC consists of molecular subtypes driven by genetic aberrations. This has led to approval of several agents specifically targeting these “driver mutations” in the setting of first-line treatment for advanced NSCLC. These agents include erlotinib, approved for approximately 10% to 15% of NSCLC patients with tumors harboring an EGFR mutation, and crizotinib, which targets translocations in ALK present in 2% to 5% of cases of NSCLC. Treatment outcome differences have been noted between two different EGFR tyrosine kinase mutations; exon 19 deletions do better than the L858R mutations. Many studies have shown the outcomes to be better in patients treated with EGFR tyrosine kinase inhibitors in the first-line setting than in patients treated with cytotoxic therapy. These agents are discussed further in the “Personalized therapies” section of this chapter.

Table 5 summarizes the results of select randomized trials in which combination cytotoxic regimens containing a newer agent are compared with old “standard” regimens or regimens containing another newer agent. Subtle differences in the eligibility criteria (eg, inclusion of patients with stage III tumors or those with poor performance status) make it difficult to directly compare the trial results. Nevertheless, there is a trend indicating that regimens containing newer agents yield higher response rates and better survival outcomes in some series than do older regimens.

Vinorelbine plus cisplatin combination. Vinorelbine was the first agent that demonstrated improved activity against NSCLC in combination with cisplatin. A European multicenter trial (Le Chevalier et al: J Clin Oncol 1994) showed the results favoring a cisplatin plus vinorelbine combination (vinorelbine, 30 mg/m2 weekly; cisplatin, 120 mg/m2 on days 1 and 29, then every 6 weeks) over a vindesine plus cisplatin combination (vindesine, 3 mg/m2 weekly; cisplatin, 120 mg/m2 on days 1 and 29, then every 6 weeks) and vinorelbine alone (30 mg/m2 weekly). The median survival duration of 40 weeks in the vinorelbine/cisplatin treatment arm was significantly longer than the 32 weeks seen in the vindesine/cisplatin arm (P = .04) and 31 weeks in the vinorelbine monotherapy arm (P < .001). This trial, however, did not confirm the role of vinorelbine in NSCLC therapy, even though it confirmed the role of cisplatin.

To address this issue, SWOG conducted a study comparing cisplatin alone (100 mg/m2 every 4 weeks) with the vinorelbine/cisplatin combination (cisplatin, 100 mg/m2 every 4 weeks; vinorelbine, 25 mg/m2 weekly × 3 every 4 weeks). Survival outcome was analyzed for 415 patients, 92% with stage IV tumors. The vinorelbine/cisplatin treatment significantly improved the disease progression–free survival (median, 2 months vs 4 months; P = .0001) and overall survival (median, 6 months vs 8 months; 1-year survival 20% vs 36%; P = .0018).

Paclitaxel plus platinum compound. A number of studies demonstrate promising results with paclitaxel in combination with cisplatin or carboplatin and other agents. Two large randomized trials compared paclitaxel plus cisplatin with standard regimens. In a three-arm, randomized trial (ECOG 5592) (Bonomi et al: J Clin Oncol 2000), 600 eligible patients with chemotherapy-naive stages IIIB-IV NSCLC were randomly assigned to receive a combination of cisplatin (75 mg/m2) plus etoposide (100 mg/m2 daily on days 1 to 3) or either low-dose (135 mg/m2 over 24 hours) or high-dose (250 mg/m2 over 24 hours with growth factor) paclitaxel plus cisplatin (75 mg/m2). The response rates for the low-dose and high-dose paclitaxel arms were 26.5% and 32.1%, respectively, significantly better than for the cisplatin/etoposide arm (12%). Superior survival was observed with the combined paclitaxel regimens (median survival time, 9.99 months; 1-year survival rate, 39%) compared with etoposide plus cisplatin (median survival, 8 months; 1-year survival rate, 32%; P = .048). Comparing survival rates for the two dose levels of paclitaxel revealed no significant differences.

A European trial of similar design (Giaccone et al: J Clin Oncol 1998) reported that cisplatin/paclitaxel improved the response rate and quality-of-life parameters. There was no improvement in overall survival, however, compared with a standard regimen of cisplatin/teniposide (Vumon).

Paclitaxel/carboplatin has been the most widely favored regimen for first-line chemotherapy in all NSCLC stages among US medical oncologists, mainly because of promising phase II trial results and the ease of administration for outpatients, with manageable toxicity profiles compared with those of cisplatin-containing regimens. One of the early phase II trials, for example, reported a response rate of 62%, a median survival duration of 53 weeks, and a 1-year survival rate of 54%. However, a randomized trial sponsored by the manufacturer of paclitaxel failed to demonstrate a survival advantage over the standard cisplatin plus etoposide regimen. Nevertheless, paclitaxel plus carboplatin may remain a community standard because a SWOG trial reported results equivalent to the time-tested vinorelbine/cisplatin regimen (see Table 5).

Combination vs single-agent chemotherapy. A randomized phase III study conducted by CALGB further supported the superiority of combination chemotherapy over single-agent therapy. Previous trials had indicated that a platinum plus a novel agent was superior to a platinum alone. Lilenbaum et al demonstrated that for patients with stage IIIB or IV NSCLC, carboplatin and paclitaxel are superior to paclitaxel alone, even for patients with an ECOG performance status of 2. This randomized trial showed a median survival advantage for the combination therapy.

Gemcitabine plus cisplatin. Gemcitabine has also been approved by the FDA for use against NSCLC, based on a series of successful phase II trials of cisplatin/gemcitabine and three major phase III trials. The HOG study, reported by Sandler et al, compared gemcitabine/cisplatin with cisplatin alone and showed a modest improvement in median and 1-year survival comparable to rates seen in the vinorelbine trials. The Spanish and Italian trials, reported by Cardenal et al and Crinò et al, compared gemcitabine plus cisplatin with standard-regimen cisplatin plus etoposide and mitomycin plus ifosfamide plus cisplatin, respectively. Although there was a significant improvement in overall response, these two studies failed to demonstrate a survival benefit.

Because gemcitabine is relatively well tolerated without dose-limiting myelosuppression, it is being evaluated for use as a single agent or in combination with other agents in older or medically compromised patients. Italian investigators report that gemcitabine combined with the vinorelbine regimen is associated with significantly better survival than single-agent vinorelbine in elderly patients with NSCLC.

Other combination regimens that contain cisplatin plus newer agents, such as docetaxel or irinotecan, also showed similar results when compared with other two-drug regimens of either two newer or two older agents.

Major randomized trials comparing cytotoxic regimens

To identify a better chemotherapy regimen for advanced-stage NSCLC, the US cooperative study groups conducted large phase III trials. The SWOG investigators compared paclitaxel/carboplatin with vinorelbine/cisplatin (the time-tested regimen in previous European and SWOG trials). A total of 404 evaluable patients were randomized to receive either paclitaxel (225 mg/m2 over 3 hours) plus carboplatin (at an AUC of 6 mg/mL/min on day 1) every 21 days or vinorelbine (25 mg/m2 weekly) plus cisplatin (100 mg/m2 on day 1) every 28 days. Overall response rates were 27% for both groups. The median survival times were also identical (8 months), with virtually identical 1-year survival rates (35% and 33%, respectively). Although both regimens provided effective palliation for advanced NSCLC, the investigators identified paclitaxel/carboplatin for future studies because of a favorable toxicity profile and better tolerability and compliance.

The ECOG 1594 trial compared three platinum-based regimens containing new agents in the treatment of NSCLC with a control arm of cisplatin and paclitaxel. The regimens were gemcitabine (1,000 mg/m2 on days 1, 8, and 15) plus cisplatin (100 mg/m2 on day 1) every 4 weeks, docetaxel (75 mg/m2) plus cisplatin (75 mg/m2 on day 1) every 3 weeks, and paclitaxel (225 mg/m2 over 3 hours) plus carboplatin (at an AUC of 6 mg/mL/min on day 1) every 21 days; the reference regimen was paclitaxel (175 mg/m2 over 24 hours) plus cisplatin (75 mg/m2 on day 1) every 21 days.

Analysis of 1,163 eligible patients showed no statistically significant differences in overall response, median survival, and 1-year survival rates when compared with the control arm, paclitaxel and cisplatin. Gemcitabine plus cisplatin was associated with a statistically significant prolongation of time to disease progression when compared with the control arm (4.5 months vs 3.5 months; P = .002) but was also associated with a higher percentage of grade 4 thrombocytopenia, anemia, and renal toxicity.

Randomized trials suggest that differentiating treatment regimens by histology may be important. A phase III trial by Scagliotti et al randomized 1,725 patients to receive a combination of cisplatin and gemcitabine vs cisplatin and pemetrexed. In patients who did not have squamous cell histology, the combination of cisplatin and pemetrexed had a superior overall survival. On the other hand, in patients with squamous cell histology, the combination of cisplatin and gemcitabine had a superior survival. These and other data suggest that pemetrexed and platinum appear to be superior in non–squamous cell populations. Overall survival was statistically superior for cisplatin/pemetrexed than for cisplatin/gemcitabine in patients with adenocarcinoma vs squamous histology (12.6 months vs 10.9 months; P = .03). In addition, cisplatin/pemetrexed provided efficacy similar to that of cisplatin/gemcitabine, with better tolerability, a reduced need for supportive therapies, and more convenient administration than cisplatin. This was the first phase III study in NSCLC to prospectively demonstrate a survival difference between platinum doublets based on histology. On the basis of the results of this study, the FDA approved pemetrexed, in combination with cisplatin, as the first-line treatment of locally advanced and metastatic NSCLC, for patients with a non-squamous histology.

Because all the regimens showed similar efficacy, quality of life becomes a critical issue in choosing a particular regimen. The decision to use one regimen over another will depend not only on ease of administration and side effects but also on the personal preference and experience of the treating oncologist.

Second-line chemotherapy for NSCLC

Before the new generation of more effective agents became available, few, if any, significant benefits were expected from second-line chemotherapy. As a result, reports in the literature seldom address this issue specifically or systematically. The most experience with second-line chemotherapy in NSCLC is with docetaxel, which has received FDA approval for this indication based on two randomized phase III trials confirming the promising phase II results of docetaxel monotherapy in patients with advanced NSCLC previously treated with platinum-based chemotherapy.

In a multicenter US trial (Fossella et al: J Clin Oncol 2000), 373 patients were randomized to receive either docetaxel (100 mg/m2 [D100] or 75 mg/m2 [D75]) or a control regimen of vinorelbine (30 mg/m2/wk) or ifosfamide (2 g/m2 × 3 days) every 3 weeks. Overall response rates were 10.8% with D100 and 6.7% with D75, each significantly higher than the 0.8% response of the control arm (P = .001 and P = .036, respectively). Although overall survival was not significantly different among the three groups, the 1-year survival was significantly higher with D75 than with the control treatment (32% vs 19%; P = .025).

In a phase III trial led by Hanna et al, 572 previously treated patients were randomized to receive pemetrexed with vitamin B12, folic acid, and dexamethasone or docetaxel with dexamethasone. The overall response rate in the pemetrexed arm was higher, 9.1% vs 8.8%. The median disease progression–free survival was 2.9 months for each arm, and the median survival favored pemetrexed (8.3 months vs 7.9 months). The 1-year survival for each arm was 29.7%. Adverse reactions (grade 3/4) were more severe with docetaxel.

The second trial (Shepherd et al: J Clin Oncol 2000) compared single-agent docetaxel with best supportive care. The initial docetaxel dose was 100 mg/m2, which was changed to 75 mg/m2 midway through the trial because of toxicity. A total of 204 patients were enrolled; 49 received D100, 55 received D75, and 100 received best supportive care. Treatment with docetaxel was associated with significant prolongation of survival (7 months vs 4.6 months; log-rank test, P = .047) and time to disease progression (10.6 weeks vs 6.7 weeks; P < .001).

Hanna et al conducted a noninferiority trial to test whether pemetrexed was equivalent to docetaxel in the second-line setting. They randomized 571 patients to receive 500 mg/m2 of pemetrexed or 75 mg/m2 of docetaxel preceded by a vitamin B12 injection every two cycles. The median survival on the two arms was equivalent, but the pemetrexed arm had significantly less grade 3/4 neutropenia and fewer hospitalizations overall. On the basis of this study, the FDA approved pemetrexed for second-line therapy for NSCLC.

Duration of chemotherapy

Maintenance single-agent chemotherapy or biologic therapy may be considered for patients with responsive or stable disease after an initial four to six cycles of platinum doublet-based chemotherapy. Data on maintenance treatment are discussed below.

Evidence for maintenance therapy

Several trials have suggested that there may be some benefit to maintenance chemotherapy or targeted biologic agents in patients who have response or stable disease after front-line therapy.

Switch maintenance. Fidias et al suggested that adding docetaxel as maintenance therapy after platinum-based front-line chemotherapy can delay tumor progression and lead to improvements in disease-free survival.

Ciuleanu et al demonstrated an improvement in progression-free survival (4.3 months vs 2.6 months; P < .001) and overall survival (13.4 months vs 10.6 months; P = .012) for switch maintenance chemotherapy with pemetrexed after four cycles of a non–pemetrexed-containing platinum doublet.

SATURN (Sequential Tarceva in Unresectable NSCLC) was an international phase III study of 889 patients that compared maintenance erlotinib with placebo in patients who had received four cycles of front-line therapy. Cappuzzo et al demonstrated that maintenance erlotinib given immediately after first-line chemotherapy significantly increased progression-free survival in both wild-type and EGFR mutation–positive patients, with HRs of 0.78 and 0.10, respectively. Maintenance erlotinib is a reasonable option for patients without an EGFR mutation, but the benefit is not quite as dramatic as results seen for patients with the mutation.

The BETA (Bevacizumab/Tarceva) trial had previously compared bevacizumab in combination with erlotinib vs erlotinib alone for the treatment of advanced NSCLC after failure of first-line chemotherapy. This phase III trial demonstrated improved progression-free survival in the combination arm. Thus, ATLAS (A Study Comparing Bevacizumab Therapy With or Without Erlotinib for First-Line Treatment of Non-Small Cell Lung Cancer) was designed to compare maintenance therapy using bevacizumab and erlotinib with bevacizumab alone following first-line therapy using bevacizumab and a platinum-containing doublet in patients with stage IIIB/IV NSCLC. The data safety monitoring committee stopped the trial early, because it met the primary endpoint, showing that the median progression-free survival was 4.8 months for the combination arm and 3.7 months for the bevacizumab-only arm. Bevacizumab has known side effects, including impaired wound healing, proteinuria, and hypertension. Survival data will be available in the next few years.

Continuation maintenance. The PARAMOUNT study demonstrated an improvement in progression-free survival and overall survival for patients with advanced non-squamous NSCLC who responded to four cycles of induction cisplatin/pemetrexed chemotherapy and who continued on maintenance pemetrexed until disease progression. The median progression-free survival, measured from randomization, was 4.1 months (95% CI, 3.2–4.6) for pemetrexed and 2.8 months (95% CI, 2.6–3.1) for placebo. Median overall survival was increased from 14 months for placebo to 16.9 months for patients who received maintenance pemetrexed.

Maintenance therapy with pemetrexed and bevacizumab compared with bevacizumab alone prolongs progression-free survival (7.4 months vs 3.7 months, P < .001), however an overall survival benefit for this strategy has yet to be demonstrated.

Promising agents and personalized therapies

Several novel agents are being developed for the treatment of solid tumors, including lung cancer. These include signal transduction inhibitors, such as tyrosine kinase inhibitors (eg, erlotinib, gefitinib); small-molecule tyrosine kinase inhibitors of VEGF; and antiangiogenic agents, including monoclonal antibodies, such as bevacizumab, cetuximab, and trastuzumab (Herceptin). Many of these novel agents are being tested in combination with chemotherapeutic agents, because their mechanisms of action suggest that they may be far more effective as long-term inhibitors of cancer progression than as classic cytotoxics.

Combinations of cetuximab with chemotherapy also appear promising. Pirker et al showed that adding cetuximab to cytotoxic chemotherapy with cisplatin and vinorelbine improved median survival from 10.1 months to 11.3 months in a randomized trial of 1,125 patients with advanced and metastatic NSCLC. This statistically significant (P = .044) 1.2-month improvement in median survival was not duplicated when cetuximab was added to taxane/platinum-based chemotherapy. However, in both of these studies, adding cetuximab to chemotherapy improved response rate, and these data are being evaluated in their totality as physicians integrate this EGFR antibody into the therapeutic paradigm for NSCLC.

Gefitinib and erlotinib. To date, most phase I studies of previously mentioned biologic compounds have suffered from a difficulty in developing pharmacologically or molecularly driven endpoints that will serve as reasonable intermediate biomarkers of efficacy or even surrogates for toxicity. Further research has focused on the novel small-molecule tyrosine kinase inhibitors erlotinib and gefitinib. Two phase II trials of gefitinib in the second- and third-line settings were conducted in Europe and Japan. Patients were randomized to receive gefitinib at either 250 or 500 mg daily. The drug was found to be active, with an 11% to 18% response rate, and there was no superiority for the higher dose. Similar data were seen for erlotinib.

Unfortunately, randomized combination trials of gefitinib at 250 and 500 mg daily with cytotoxic chemotherapy, either paclitaxel and carboplatin in one trial or gemcitabine and cisplatin in the other study vs placebo in front-line therapy, failed to demonstrate any survival advantage. These results have cast a pall over the development of tyrosine kinase inhibitors in combination with chemotherapy. Regardless, the question of whether or not to approve these new agents for third-line therapy of lung cancer in cisplatin/docetaxel-refractory patients remains open for debate.

The activity demonstrated by gefitinib and erlotinib as single agents in advanced NSCLC generated much optimism, including a recent trial indicating that erlotinib improved survival (by a median of 2 months) in the second- and third-line treatments of NSCLC, with a substantial increase in 1-year survival (P < .001). After the initial successes with chemotherapy and promising results from targeted tyrosine kinase inhibitors in stages IIIB and IV NSCLC, several large phase III trials of either gefitinib or erlotinib in combination with conventional platinum-based chemotherapy were conducted as the INTACT (Iressa NSCLC Trial Assessing Combination Treatment) 1 and INTACT 2 trials or the TALENT (Tarceva Lung Cancer Investigation) and TRIBUTE (Tarceva Responses in Conjunction with Paclitaxel and Carboplatin) trials.

The phase III trials of single-agent erlotinib and gefitinib, when compared with best supportive care in the second- and third-line settings, had dramatically different results. Shepherd et al conducted a randomized trial of erlotinib (150 mg daily) vs placebo in 723 patients previously treated with front- or second-line chemotherapy for NSCLC. A total of 488 patients received erlotinib and 243 received placebo.

The median overall survival significantly favored the erlotinib arm (6.7 months vs 4.7 months), with a reduction in the HR of 0.73 (P > .001), as did the 1-year survival rate (31% vs 21%; P > .01). Quality of life and median time to deterioration also significantly favored erlotinib. On the other hand, the phase III trial of gefitinib vs best supportive care in more than 1,600 patients was reported and showed no significant advantage to gefitinib at 250 mg daily in a patient population with a significant number of active smokers.

For patients who have the classic exon 19 or 21 mutation, erlotinib and gefitinib provide a significant advantage over standard chemotherapy. However, some patients are resistant or develop a T790M mutation; this mutation adds a bulky methionine group in the adenosine triphosphate binding pocket, which sterically hinders the attachment of tyrosine kinase inhibitors.

IPASS (Iressa Pan-Asia Study) was a phase III, multicenter, randomized, open-label, parallel-group study of 1,217 select never- or light-smoking patients in East Asia who had advanced NSCLC and received either gefitinib or carboplatin plus paclitaxel as first-line treatment. The primary endpoint was progression-free survival; the secondary endpoints included overall survival, for which the analysis is ongoing. The 12-month rates of progression-free survival were 24.9% with gefitinib and 6.7% with carboplatin-paclitaxel (P < .001). On the basis of these data and those of SATURN, the FDA approved erlotinib for first-line maintenance treatment of locally advanced or metastatic NSCLC in patients whose disease has not progressed after four cycles of platinum-based chemotherapy. Subsequent data analysis has led to the approval of erlotinib as first-line therapy for stage IV NSCLC with sensitizing mutations of EGFR.

Two trials reported in 2011 provided evidence of the superiority of erlotinib in EGFR mutation–positive patients, one in a Chinese population (OPTIMAL) and a second in a European population (EURTAC). OPTIMAL demonstrated that erlotinib improved progression-free survival and quality of life compared with gemcitabine/carboplatin (13.7 months vs 4.6 months). Similarly, EURTAC showed progression-free survival of 9.4 months in the erlotinib arm vs 5.2 months in the platinum-based chemotherapy arm. EURTAC also investigated overall survival, which was not significantly different; this finding is possibly explained by the fact that crossover was allowed on progression.

Sidebar: Presented at ASCO 2014, and published in 2015 in the Journal of Clinical Oncology by Kelly et al, were the final results of the RADIANT phase III trial in patients with stage IB–IIIA NSCLC treated postsurgically with or without adjuvant chemotherapy and the addition of erlotinib. The median disease-free survival was 50.5 months in the erlotinib arm vs 48.2 months in the placebo arm. The study was insufficiently powered to assess the role of erlotinib in the EGFR-mutated patients. As a result, the ALCHEMIST trial was designed to answer the question of using molecularly targeted agents in the postsurgical adjuvant setting.

Bevacizumab. When combined with chemotherapy, bevacizumab appears to result in a significant survival advantage for patients with advanced NSCLC. On the basis of promising phase II data of bevacizumab in combination with carboplatin and paclitaxel, Sandler et al randomized 878 patients to receive either bevacizumab with carboplatin and paclitaxel or carboplatin and paclitaxel alone.

There was a significant increase in the response rate for the combination (27.2%) with bevacizumab vs paclitaxel and carboplatin alone (10%; P < .001), and both median and progression-free survival significantly favored the bevacizumab combination arm (overall survival: 12.5 months vs 10.2 months; P = .007; HR = 0.77; 95% CI, 0.65–0.93; progression-free survival: 6.4 months vs 4.5 months; P < .001; HR = 0.62; 95% CI, 0.53–0.72). As in the phase II trial, this study reported a total of eight treatment-related deaths in the bevacizumab arm vs two in the paclitaxel and carboplatin arm. A second phase III randomized trial testing the value of adding bevacizumab to gemcitabine and cisplatin therapy showed an improvement in disease-free survival, but not overall survival.

Other promising novel agents now under investigation in lung cancer include inhibitors of the insulin-like growth factor (IGF) receptor. Karp et al showed that these agents appear to be particularly promising in patients with squamous cell lung tumors, with response rates in a small study as high as 70%. As more pathways involved in NSCLC are found, more clinical trials involving inhibitors of phosphoinositide 3-kinase, IGF, and c-MET (where sensitivity is predicted by a mutation in BRAF [B-Raf proto-oncogene serine/threonine-protein kinase]) are being initiated. MET expression is associated with poor outcome in NSCLC. A recent study compared erlotinib plus a monoclonal antibody the binds to the Met receptor (MetMAb) with erlotinib plus placebo in patients with increased expression of c-MET measured by immunohistochemistry. Patients in the MetMAb arm had progression-free survival of 3 months vs 1.5 months for patients in the placebo arm. Overall survival was 12.6 months vs 4.6 months for the placebo arm.

Personalized therapies

Evaluation of NSCLC involves distinction between a squamous lesion or adenocarcinoma and of actual genetic differences. Pemetrexed and bevacizumab are used in non-squamous histology, and erlotinib is given for EGFR mutations, but other oncogene addictions are becoming apparent. ALK gene rearrangement occurs in 4% of NSCLC patients with EML4 (echinoderm microtubule-associated protein-like 4) as its most common fusion partner. EML4-ALK is seen in up to 20% of nonsmokers and in about 30% of patients who are never-smokers and EGFR wild-type. It is strongly associated with signet ring cell histology. Crizotinib, an ALK inhibitor, has been found to improve overall survival in EML4-ALK mutation–positive patients in both first-line and second-line treatment. In first-line therapy, 2-year overall survival was 64% vs 33% for historical controls. In second- and third-line therapy, 2-year overall survival was 61% vs 9% for historical controls. Early results from a second, ongoing study in patients who had progressed after at least one chemotherapy regimen reported 83% of patients had tumor shrinkage and most patients had clinically significant improvements in pain, cough, dyspnea, and fatigue. Recent data suggest that crizotinib also has activity in NSCLC harboring a ROS1 gene rearrangement.

Sidebar: At ASCO 2014, the results of the ASCEND-1 trial were updated. Ceritinib, an advanced anaplastic lymphoma kinase (ALK) inhibitor, demonstrated a response rate of 59% overall and 55% response in patients with metastatic NSCLC who were previously treated with other ALK inhibitors. There was a 50% overall response in the patients with brain metastases. At ASCO 2015, results were reported of a phase II study using alectinib, a new ALK inhibitor, in ALK-mutation positive patients who were crizotinib-resistant. The rate of response was 50%, with a duration of 11.2 months, and the rate of adverse grade III and IV events was less than 5%.

Sidebar: At ASCO 2015, the third-generation irreversible EGFR tyrosine kinase inhibitor AZD9291 (osimertinib) was studied as a first-line treatment at two dosages, 80 mg and 160 mg, in EGFR-mutant metastatic NSCLC, with an overall response rate of 73% and a disease control rate of 97%. Outcomes appeared to be superior at the higher dose, with approximately a 30% improvement in response. The progression-free survival rate at 1 year was 72%. A new trial, the FLAURA study, is a phase III trial comparing AZD9291 vs gefitinib or erlotinib in patients with metastatic EFGR-mutant NSCLC. The endpoint is progression-free survival.

Very encouraging data with the anti-programmed death-1 (anti-PD-1) antibody, nivolumab, and other agents targeting the PD1 axis have also been reported. PD-1 is an immune checkpoint molecule involved in suppression of T-cell–mediated anti-tumor immunity. Strategies being investigated prospectively to maximize the effectiveness of PD-1 inhibition may include sequencing with chemotherapy, co-inhibition with other immune checkpoint inhibitors, and combination with other targeted agents.

Sidebar: At ASCO 2015, investigators reported results of the CheckMate 017 trial, a phase III study comparing nivolumab with docetaxel in 272 patients with metastatic squamous cell cancers who demonstrated progressive disease after platinum doublet therapy. Survival at 1 year was 42% for patients treated with nivolumab vs 24% for those treated with docetaxel. Median overall survival was 9.2 months with nivolumab and 6 months with docetaxel. The response duration was 17.2 months for nivolumab-treated patients and 5.6 months for those who received docetaxel. To date, there is no good biomarker to predict response to nivolumab. A similar study design was performed in the non–squamous-cell patients in CheckMate 057. Median survival was 12.2 months with nivolumab and 9.4 months with docetaxel. Rates of 1-year survival were 51% with nivolumab and 39% with docetaxel. There were more adverse events due to docetaxel, 88% vs 69% for any grade and 54% vs 10% for grades 3 and 4.

Sidebar: At ASCO 2014, investigators presented updated results of the combined prospective phase III trials LUX-Lung 3 and LUX-Lung 6. These studies compared afatinib, an irreversible EGFR and human epidermal growth factor receptor (HER2, also known as erbB-2) tyrosine kinase, vs cytotoxic therapies, specifically cisplatin and pemetrexed in LUX-Lung 3 and cisplatin and gemcitabine in LUX-Lung 6. Patients with exon 19 deletion showed superior outcomes from chemotherapy, compared with L858R-mutant patients (HR = 0.54; P = .0015). Also at ASCO 2014, AZD9291 appeared effective in preclinical studies against exon 19 deletion, L858R and T790m, a resistant mutation that occurs in 50% of patients treated with inhibitors; AZD9291 yielded an overall response rate of 51% and 64% in the EGFR T790m patients. Neither rash nor diarrhea were noted, unlike the experience with other EGFR tyrosine kinase inhibitors. Similarly, in a study of rociletinib, which inhibits T790m and other activating mutations while sparing wild-type EGFR, median progression-free survival was not reached and will likely exceed 1 year.

TABLE 6: Single-agent chemotherapy regimens for NSCLC

TABLE 7: Combination chemotherapy regimens recommended for NSCLC

Current treatment recommendations

It is important to note that patients who have lost significant amounts of weight or who have a poor performance status are at greater risk for toxicity, including a higher likelihood of lethal toxicity, when they are treated with modest doses of chemotherapy. On the basis of currently available data, a reasonable approach for patients with stage IV NSCLC who have a good performance status (ECOG performance status 0/1) and have not lost a significant amount of weight (< 5% of usual weight) would be to encourage them to participate in a clinical trial. However, it would also be appropriate to treat this group of patients with etoposide plus cisplatin or with one of the newer combination regimens, such as gemcitabine/cisplatin, vinorelbine/cisplatin, paclitaxel/cisplatin, paclitaxel/carboplatin, or docetaxel/cisplatin (Tables 6 and 7).

Role of Photodynamic Therapy

Photodynamic therapy (PDT), which combines porfimer sodium (Photofrin), a hematoporphyrin derivative in which the less active porphyrin monomers have been removed with an argon-pumped dye laser, has been explored in a variety of different tumors, with varying results. Several investigators have reported excellent results with PDT in early-stage head and neck cancers as well as intrathoracic tumors. However, initial studies have involved a limited number of patients.

Although this novel technique seems to be extremely promising, it appears to be applicable to only a minority of patients with NSCLC. Nevertheless, PDT appears to be particularly useful for the treatment of early-stage lung cancer for a variety of reasons. First, it appears to preserve lung function and can be repeated as additional tumors appear-an important consideration, because such patients appear to be at high risk for developing other new tumors. Second, this technique does not preclude ultimate surgical intervention when deemed necessary.

Results in early-stage NSCLC

In a prospective phase II study (Furuse et al: J Clin Oncol 1993), 54 patients with 64 early-stage lung cancers received Photofrin (2 mg/kg) and 630-nm illumination of 100 to 200 J/cm2. Of 59 assessable tumors, 50 responded completely and 6 showed partial responses. Five patients with complete response developed recurrences 6 to 18 months after treatment.

The major predictor of response in this study was tumor length. The likelihood of achieving a complete response was 97.8% if the tumor was smaller than 1 cm, as opposed to only 42.9% if the lesion was larger than 1 cm. The overall survival rate in these patients was 50% at 3 years.

A similar study by Kato et al also indicated a 96.8% complete response rate for tumors smaller than 0.5 cm but only a 37.5% rate for tumors larger than 2 cm. The overall 5-year survival rate for the 75 patients treated in this study was 68.4%, which is acceptable by current standards.

Results in advanced-stage NSCLC

Two prospective randomized trials (European and US/Canadian) compared PDT with the neodymium:yttrium-aluminum-garnet (Nd:YAG) laser for treatment of partially obstructive, advanced NSCLC. Investigators analyzed results from the two trials both individually and collectively. Collective analysis included data from 15 centers in Europe and 20 centers in the United States and Canada, and it involved a total of 211 patients. In the European trial, 40% of the patients had received previous therapy, whereas in the US/Canadian trial, all of the patients had received previous treatment.

Tumor response was similar for both therapies at 1 week. At 1 month, however, 61% and 42% of the patients treated with PDT in the European and US/Canadian trials, respectively, were still responding, compared with 36% and 19% of patients who underwent laser therapy in the two trials.

PDT also produced more dramatic improvements in dyspnea and cough than did Nd:YAG therapy in the European trial, but the two treatments had similar effects on these symptoms in the US/Canadian trial. Both sets of investigators concluded that PDT appears to be superior to laser therapy for the relief of dyspnea, cough, and hemoptysis. Also, the overall incidence of adverse reactions was similar with the two therapies (73% for PDT vs 64% for Nd:YAG therapy).

Palliation of Local and Distant Symptoms

Radiation therapy

Many patients with lung cancer experience distressing local symptoms at some time. They may arise from airway obstruction by the primary tumor, compression of mediastinal structures by nodal metastases, or metastatic involvement of distant organs. Radiation therapy is effective in palliating most local symptoms as well as symptoms at common metastatic sites, such as bone and brain. For select patients with a solitary brain metastasis and controlled disease in other sites, resection followed by irradiation appears to be superior to radiation therapy alone in improving both survival and quality of life. (For more information regarding management of brain metastases, see the “Primary and Metastatic Brain Tumors” chapter.)

An American Society for Radiation Oncology (ASTRO) committee cochaired by Rodrigues and Movsas developed an evidence-based guideline for palliative radiotherapy in cancer. Overall, higher-dose/fractionation palliative external beam radiation therapy (EBRT) regimens (eg, 30 Gy/10 fractions equivalent or greater) are associated with modest improvements in survival and total symptom score, particularly in patients with good performance status. Because these improvements are associated with an increase in esophageal toxicity, however, various shorter EBRT dose/fractionation schedules (eg, 20 Gy in five fractions, 17 Gy in two weekly fractions, 10 Gy in one fraction), which provide good symptomatic relief with fewer side effects, can be used for patients who request a shorter treatment course and/or in those with a poor performance status. Integration of concurrent chemotherapy with palliative intent/fractionated radiotherapy is not currently supported by the medical literature.

Endobronchial irradiation. Cobalt-60 or iridium-192 has been used to palliate symptoms arising from partial airway obstruction, including cough, dyspnea, and hemoptysis. The dosimetric advantage of being able to deliver a high radiation dose to the obstructing endobronchial tumor while sparing adjacent normal structures, such as the lungs, spinal cord, and esophagus, has clear appeal, particularly in the patient whose disease has recurred following prior EBRT. Although good rates of palliation have been reported with endobronchial irradiation, significant complications, including fatal hemoptysis, are seen in 5% to 10% of patients. It remains unclear, however, how often this complication is actually due to the irradiation or the underlying disease itself.

Endobronchial brachytherapy (EBB) should be considered as one of several approaches (including laser excision, cryotherapy, and stent placement) used in the management of patients with symptomatic airway obstruction, and management should be individualized. All of these approaches are more suitable for partial than for complete airway obstruction. Overall, EBB can be a reasonable option for the palliation of endobronchial lesions that cause obstructive symptomatology, including lung collapse, or for hemoptysis, after EBRT failure.

Chemotherapy

Several trials have explored the use of chemotherapy to palliate specific symptoms in patients with lung cancer. In general, these trials have found that rates of symptomatic improvement were considerably higher than objective response rates and were not dissimilar to symptomatic response rates with local radiation therapy.

Thus, although radiation therapy remains the most appropriate modality for treatment of problems such as superior vena cava obstruction, spinal cord compression, brain metastases, or localized bone pain, patients who have more extensive disease without these local emergencies may be considered for palliative chemotherapy, which may relieve local symptoms and prolong survival.

Bisphosphonates

Approximately 30% to 65% of patients with advanced lung cancer develop bone metastases. The median survival following development of bone metastases is 6 months. Bone disease is associated with significant morbidity, including severe pain, hypercalcemia of malignancy, pathologic fracture, and spinal cord or nerve root compression. Treatment of bone metastases may include surgical intervention, radiation therapy, and chemotherapy. Bisphosphonate treatment can decrease skeleton-related complications, delay progressive disease, and relieve bone pain.

Bisphosphonates such as clodronate, pamidronate, and zoledronic acid (Zometa) exhibit strong affinity for the hydroxyapatite crystal of bone and preferentially accumulate at sites of active bone remodeling, where they prevent bone resorption. They provide effective treatment for hypercalcemia of malignancy and have been shown to delay the onset of progressive bone disease and relieve bone pain in studies largely performed in patients with metastases secondary to breast cancer, multiple myeloma, and prostate cancer. Nitrogen-containing bisphosphonates, such as pamidronate and zoledronic acid, appear to exert antitumor effects. Zoledronic acid has also been found to have pronounced anti-nociceptive effects, which have been absent with other bisphosphonates in preclinical studies. Zoledronic acid is the only bisphosphonate shown to be effective in reducing skeletal complications in patients with bone metastases from lung cancer and solid tumors other than breast and prostate cancers.

Follow-Up of Long-Term Survivors

ASTRO and ASCO have been developing cancer survivorship plans. At present, no standard follow-up protocol exists for patients with cured NSCLC or SCLC. However, long-term follow-up should at least include serial physical examinations once the patient has reached the 5-year mark. There is some controversy about the value of using CT scanning or even chest radiography for the long-term follow-up of these patients.

In this vein, retrospective reviews of the literature have revealed that patients with SCLC appear to have the highest rate of second primary tumor development-as high as 30%-over the course of their lifetime, with some studies reporting annual second primary tumor rates of 5% to 10%. Therefore, the concept of chemoprevention appears to have particular merit in these patients.

A randomized chemoprevention study of patients with stage I NSCLC showed a surprisingly high annual recurrence rate of 6.5% in patients with T1 tumors, as opposed to 11.2% in patients with T2 tumors. There is clearly a need for effective chemoprevention for both of these tumor subsets, as well as the establishment of consistent guidelines for routine long-term follow-up.

Sidebar: Long-term lung cancer survivors who engaged in regular physical activity for 30 minutes or more 5 days per week had higher QOL scores and better symptom control than their less active counterparts (Solberg et al: Lung Cancer 2012).

Suggested Reading

Albain KS, Swann RS, Rusch VW, et al: Radiotherapy plus chemotherapy with or without surgical resection for stage III non-small-cell lung cancer: A phase III randomised controlled trial. Lancet 374:379–386, 2009.

Annema JT, van Meerbeeck JP, Rintoul RC, et al: Mediastinoscopy vs endosonography for mediastinal nodal staging of lung cancer: A randomized trial. JAMA 304:2245–2252, 2010.

Arriagada R, Dunant A, Pignon JP, et al: Long-term results of the international adjuvant lung cancer trial evaluating adjuvant cisplatin-based chemotherapy in resected lung cancer. J Clin Oncol 28:35–42, 2010.

Ashworth AB, Senan S, Palma DA, et al: An individual patient data metaanalysis of outcomes and prognostic factors after treatment of oligometastatic non–small-cell lung cancer. Clin Lung Cancer 15:346–355, 2014.

Azzoli CG, Temin S, Aliff T, et al: 2011 Focused update of 2009 American Society of Clinical Oncology clinical practice guideline update on chemotherapy for stage IV non-small-cell lung cancer. J Clin Oncol 29:3825–3831, 2011.

Bach PB, Jett JR, Pastorino U, et al: Computed tomography screening and lung cancer outcomes. JAMA 297:953–961, 2007.

Bergethon K, Shaw AT, Ou SH, et al: ROS1 rearrangements define a unique molecular class of lung cancers. J Clin Oncol 30:863–870, 2012.

Bilal H, Mahmood S, Rajashanker B, et al: Is radiofrequency ablation more effective than stereotactic ablative radiotherapy in medically inoperable non-small cell lung cancer? Interact Cardiovasc Thorac Surg 15:258–265, 2012.

Bradley JD, Paulus R, Komaki R, et al: Standard-dose versus high-dose conformal radiotherapy with concurrent and consolidation carboplatin plus paclitaxel with or without cetuximab for patients with stage IIIA or IIIB non-small-cell lung cancer (RTOG 0617): A randomised, two-by-two factorial phase 3 study. Lancet Oncol 16:187–199, 2015.

Brenner DJ, Hall EJ: Computed tomography-An increasing source of radiation exposure. N Engl J Med 357:2277–2284, 2007.

Cappuzzo F, Ciuleanu T, Stelmakh L, et al: SATURN: A double-blind, randomized, phase III study of maintenance erlotinib versus placebo following non-progression with first-line platinum-based chemotherapy in patients with advanced NSCLC. J Clin Oncol 27(15S): abstract 8001, 2009.

Chang JY, Kestin LL, Barriger RB, et al: ACR Appropriateness Criteria® nonsurgical treatment for locally advanced non-small-cell lung cancer: Good performance status/definitive intent. Oncology (Williston Park) 28:706–710, 2014.

Ciuleanu T, Brodowicz T, Zielinski C, et al: Maintenance pemetrexed plus best supportive care versus placebo plus best supportive care for non-small-cell lung cancer: a randomised, double-blind, phase 3 study. Lancet 374:1432–1440, 2009.

Crinò L, Kim D, Riely JG, et al: Initial phase II results with crizotinib in advanced ALK-positive non-small cell lung cancer (NSCLC): PROFILE 1005. J Clin Oncol 29[suppl]: abstract 7514, 2011.

Curran WJ Jr, Paulus R, Langer CJ, et al: Sequential vs. concurrent chemoradiation for stage III non-small-cell lung cancer: Randomized phase III trial RTOG 9410. J Natl Cancer Inst 103:1452–1460, 2011.

Douillard JY, Tribodet H, Aubert D, et al: Adjuvant cisplatin and vinorelbine for completely resected non-small cell lung cancer: Subgroup analysis of the Lung Adjuvant Cisplatin Evaluation. J Thorac Oncol 5:220–228, 2010.

Fidias PM, Dakhil SR, Lyss AP, et al: Phase III study of immediate compared with delayed docetaxel after front-line therapy with gemcitabine plus carboplatin in advanced non-small-cell lung cancer. J Clin Oncol 27:591–598, 2009.

Gore EM, Bae K, Wong SJ, et al: Phase III comparison of prophylactic cranial irradiation versus observation in patients with locally advanced non-small-cell lung cancer: Primary analysis of radiation therapy oncology group study RTOG 0214. J Clin Oncol 29:272–278, 2011.

Grønberg BH, Bremnes RM, Fløtten O, et al: Phase III study by the Norwegian lung cancer study group: Pemetrexed plus carboplatin compared with gemcitabine plus carboplatin as first-line chemotherapy in advanced non-small-cell lung cancer. J Clin Oncol 27:3217–3224, 2009.

Grunnet M, Sorensen JB: Carcinoembryonic antigen (CEA) as tumor marker in lung cancer. Lung Cancer 76:138–143, 2012.

Hanna N. Neubauer M, Yiannoutso SC, et al: Phase III study of cisplatin, etoposide, and concurrent chest radiation with or without consolidation docetaxel in patients with inoperable stage III non-small-cell lung cancer: The Hoosier Oncology Group and US Oncology. J Clin Oncol 26:5755–5760, 2008.

Hanna N, Shepherd FA, Fossella FV, et al: Randomized phase III trial of pemetrexed versus docetaxel in patients with non–small-cell lung cancer previously treated with chemotherapy. J Clin Oncol 22:1589–1597, 2004.

Hughes JR, Stead LF, Lancaster T: Antidepressants for smoking cessation. Cochrane Database Syst Rev Jan 24;(1):CD000031, 2007.

International Early Lung Cancer Action Program Investigators; Henschke CI, Yankelevitz DF, Libby DM, et al: Survival of patients with stage I lung cancer detected on CT screening. N Engl J Med 355:1763–1771, 2006.

Karp DD, Paz-Ares LG, Novello S: High activity of the anti-IGF-IR antibody CP-751,871 in combination with paclitaxel and carboplatin in squamous NSCLC. J Clin Oncol 26(15 Suppl):8015, 2008.

Karp DD, Lee SJ, Shaw Wright GL, et al: A phase III, intergroup, randomized, double-blind, chemoprevention trial of selenium (SE) supplementation in resected stage I non-small cell lung cancer (NSCLC). J Clin Oncol 28(18S): abstract 7004, 2010.

Kelly K, Altorki NK, Eberhardt WEE, et al: Adjuvant erlotinib versus placebo in patients with stage !B-IIIA non–small-cell lung cancer (RADIANT): A randomized, double-blind, phase III trial. J Clin Oncol August 31, 2015. [Epub before print]

Kim E, Biswas T, Dowlati A, et al: Comparison of overall and cause-specific survivals of stage III non-small cell lung cancer (NSCLC) treated with carboplatin- or cisplatin-based concurrent chemoradiation therapy using SEER-medicare data (1992-2007). Int J Radiat Oncol Biol Phys 90(suppl): S19.

Kratz JR, He J, Van Den Eeden SK, et al: A practical molecular assay to predict survival in resected non-squamous, non-small-cell lung cancer: Development and validation studies. Lancet 379:823–832, 2012.

Kumar S, Movsas B, Brown SL, et al: Prognostic biomarkers in non-small cell lung cancer patients treated with radiation therapy. Int J Radiat Oncol 90:S25–S26, 2014.

Lanuti M, Sharma A, Willers H, et al: Radiofrequency ablation for stage I non-small cell lung cancer: management of locoregional recurrence. Ann Thorac Surg 93:921–927, 2012.

Lilenbaum RC, Herndon JE 2nd, List MA, et al: Single-agent versus combination chemotherapy in advanced non-small-cell lung cancer: The Cancer and Leukemia Group B (study 9730). J Clin Oncol 23:190–196, 2005.

Mancini BR, Park HS, Harder E, et al: Elderly patients undergoing SBRT for inoperable early-stage NSCLC achieve similar operable outcomes to younger patients. Int J Radiat Oncol Biol Phys 90:S11, 2014.

McElnay PJ, Choong A, Jordan E, et al: Outcome of surgery versus radiotherapy after induction treatment in patients with N2 disease: Systematic review and meta-analysis of randomised trials. Thorax May 12, 2015. [Epub before print]

Miller VA, O'Connor P, Soh C, et al: A randomized, double-blinded, placebo-controlled, phase IIIb trial (ATLAS) comparing bevacizumab therapy with or without erlotinib after completion of chemotherapy with B for first-line treatment of locally advanced, recurrent, or metastatic non-small cell lung cancer. J Clin Oncol 27(18S): abstract LBA8002, 2009.

Mok TS, Wu YL, Thongprasert S, et al: Gefitinib or carboplatin-paclitaxel in pulmonary adenocarcinoma. N Engl J Med 361:947–957, 2009.

Movsas B, Moughan J, Sarna L, et al: Quality of life supersedes the classic prognosticators for long-term survival in locally advanced non-small-cell lung cancer: an analysis of RTOG 9801. J Clin Oncol 27:5816–5822, 2009.

Okamoto H, Atagi S, Kawahara M, et al: Updated results of a phase III trial comparing standard thoracic radiotherapy (RT) with or without concurrent daily low-dose carboplatin in elderly patients (pts) with locally advanced non-small cell lung cancer (NSCLC): JCOG0301. J Clin Oncol 30(suppl): abstract 7017, 2012.

Paz-Ares LG, de Marinis F, Dediu M, et al: Maintenance therapy with pemetrexed plus best supportive care versus placebo plus best supportive care after induction therapy with pemetrexed plus cisplatin for advanced non-squamous non-small-cell lung cancer (PARAMOUNT): a double-blind, phase 3, randomised controlled trial. Lancet Oncol 13:247–555, 2012.

Paz-Ares LG, de Marinis F, Dediu M, et al: PARAMOUNT: Final overall survival (OS) results of the phase III study of maintenance pemetrexed (pem) plus best supportive care (BSC) versus placebo (plb) plus BSC immediately following induction treatment with pem plus cisplatin (cis) for advanced nonsquamous (NS) non-small cell lung cancer (NSCLC). J Clin Oncol 30(18S): abstract LBA7507, 2012.

Peng G, Tinner RG, Wang Y, et al: Comparison of patient outcomes stratified by histology among pemetrexed (P)-treated patients with stage IIIB/IV non-small cell lung cancer (NSCLC) in two phase II trials. J Clin Oncol 26(15S): abstract 8096, 2008.

Pham A, Arora S, Wernicke A, et al: Radiation does not add to cardiac-specific death in patients with lung cancer. Int J Radiat Oncol Biol Phys 90:S59, 2014.

Pirker R, Szczesna A, von Pawel J, et al: FLEX: A randomized, multicenter, phase III study of cetuximab in combination with cisplatin/vinorelbine (CV) versus CV alone in the first-line treatment of patients with advanced non-small cell lung cancer (NSCLC). J Clin Oncol 26(15S): abstract 3, 2008.

Pisters K, Vallieres E, Bunn PA, et al: S9900: Surgery alone or surgery plus induction (ind) paclitaxel/carboplatin (PC) chemotherapy in early stage non-small cell lung cancer (NSCLC): Follow-up on a phase III trial. J Clin Oncol 25(18S): abstract 7520, 2007.

Pless M, Stupp R, Ris HB, et al; SAKK Lung Cancer Project Group: Induction chemoradiation in stage IIIA/N2 non-small-cell lung cancer: A phase 3 randomised trial. Lancet 1049–1056, 2015.

Raaschou-Nielsen O, Andersen ZJ, Beelen R, et al: Air pollution and lung cancer incidence in 17 European cohorts: prospective analyses from the European Study of Cohorts for Air Pollution Effects (ESCAPE). Lancet Oncol 14:813–822, 2013.

Ramlau R, Zatloukal P, Jassem J, et al: Randomized phase III trial comparing bexarotene (L1069-49)/cisplatin/vinorelbine with cisplatin/vinorelbine in chemotherapy-naive patients with advanced or metastatic non-small-cell lung cancer: SPIRIT I. J Clin Oncol 26:1886–1892, 2008.

Riely GJ, Brahmer JR, Planchard D, et al: A randomized discontinuation phase II trial of ridaforolimus in non-small-cell lung cancer (NSCLC) patients with KRAS mutations. J Clin Oncol 30(15S): abstract 7531, 2012.

Rimner A, Din S, Shaikh F, et al: Stage III non-small cell lung cancer treated without concurrent chemotherapy: What is the optimal radiation dose? Int J Radiat Oncol Biol Phys 90:S23, 2014.

Rodrigues G, Videtic G, Sur R, et al: Palliative thoracic radiotherapy in lung cancer: An American Society for Radiation Oncology evidence-based clinical practice guideline. Pract Radiat Oncol 1:60–71, 2011.

Sandler A, Gray R, Perry MC, et al: Paclitaxel-carboplatin alone or with bevacizumab for non-small-cell lung cancer. N Engl J Med 355:2542–2550, 2006.

Scagliotti GV, Parikh P, von Pawel J, et al: Phase III study comparing cisplatin plus gemcitabine with cisplatin plus pemetrexed in chemotherapy-naive patients with advanced-stage non-small-cell lung cancer. J Clin Oncol 26:3543–3551, 2008.

Senan S, Verstegen NE, Haasbeek CJ, et al: Outcomes of stereotactic body radiotherapy in patients with clinical stage I non-small cell lung cancer who are fit to undergo surgery. J Clin Oncol 29(suppl): abstract 7051, 2011.

Shaw AT, Yeap BY, Mino-Kenudson M, et al: Clinical features and outcome of patients with non-small-cell lung cancer who harbor EML4-ALK. J Clin Oncol 27:4247–4253, 2009.

Shaw AT, Yeap BY, Solomon BJ, et al: Impact of crizotinib on survival in patients with advanced ALK-positive NSCLC compared with historical controls. J Clin Oncol 29(suppl): abstract CRA7507, 2011.

Siegel R, Naishadham D, Jemal A, et al: Cancer statistics, 2012. CA Cancer J Clin 62:10–29, 2012.

Simon G: Personalized chemotherapy may favorably alter intrinsic disease biology to produce a higher proportion of long term survivors in patients with advanced NSCLC (abstract D7.6). Presented at the Annual Meeting of the International Association for the Study of Lung Cancer; San Francisco; July 31–August 4, 2009.

Spigel DR, Ervin TJ, Ramlau R, et al: Final efficacy results from OAM4558g, a randomized phase II study evaluating MetMAb or placebo in combination with erlotinib in advanced NSCLC. J Clin Oncol 29(suppl): abstract CRA7505, 2011.

Su S, Scott WJ, Allen MS, et al: Patterns of survival and recurrence after surgical treatment of early stage non–small cell lung carcinoma in the ACOSOG Z0030 (ALLIANCE) trial. J Thorac Cardiovasc Surg 147:747–753, 2014.

Sun A, Bae K, Gore EM, et al: Phase III trial of prophylactic cranial irradiation compared with observation in patients with locally advanced non-small-cell lung cancer: Neurocognitive and quality-of-life analysis. J Clin Oncol 29:279–286, 2010.

Taylor MD, La Par DJ, Isbell JM, et al: Marginal pulmonary function should not preclude lobectomy in selected patients with non-small cell lung cancer. J Thorac Cardiovasc Surg 147:738–744, 2014.

Timmerman RD, Hu C, Michalski J, et al: Long-term results of RTOG 0236: A phase II trial of stereotactic body radiation therapy (SBRT) in the treatment of patients with medically inoperable stage I non-small cell lung cancer. Int J Radiat Oncol Biol Phys 90:S30, 2014.

Topalian SL, Hodi FS, Brahmer JR, et al: Safety, activity and immune correlates of anti-PD-1 antibody in cancer. N Engl J Med 28:2443–2454, 2012.

Travis WD, Brambilla E, Noguchi M, et al: International Association for the Study of Lung Cancer/American Thoracic Society/European Respiratory Society international multidisciplinary classification of lung adenocarcinoma. J Thorac Oncol 6:244–285, 2011.

Videtic GM, Chang JY, Chetty IJ, et al; Expert Panel on Radiation Oncology-Lung: ACR Appropriateness Criteria® early-stage non-small-cell lung cancer. Am J Clin Oncol 37:201–207, 2014.

Vincent MD, Butts C, Seymour L, et al: Updated survival analysis of JBR.10: A randomized phase III trial of vinorelbine/cisplatin versus observation in completely resected stage IB and II non-small cell lung cancer (NSCLC). J Clin Oncol 27(15S): abstract 7501, 2009.

Willers H, Stinchcombe TE, Barriger RB, et al; Expert Panel on Radiation Oncology-Lung: ACR Appropriateness Criteria® and adjuvant therapy for N2 non-small-cell lung cancer. Am J Clin Oncol 38:197–205, 2015.