Handbook of Cancer Chemotherapy (Lippincott Williams & Wilkins Handbook Series), 8th Ed.

6. Carcinoma of the Lung

David E. Gerber and Joan H. Schiller

Carcinoma of the lung is responsible for more than 165,000 deaths each year in the United States. This represents one-third of all deaths due to cancer and more than the number of deaths due to breast, colon, and prostate cancers combined. Because early-stage lung tumors are often asymptomatic and there has been no proven approach to radiographic screening, most patients are diagnosed with advanced stage disease. Approximately 85% of cases are histologically classified as non-small-cell lung cancer (NSCLC), of which adenocarcinoma, squamous cell carcinoma, and large cell carcinoma are the primary subtypes. Small-cell lung cancer (SCLC) accounts for the remaining 15% of cases. The biology, staging, and treatment of SCLC differ substantially from NSCLC. Thus, these two groups are addressed in two separate sections.

I. ETIOLOGY

Lung cancer is predominantly a disease of smokers. Eighty-five percent of lung cancer occurs in active or former smokers, and an additional 5% of cases are estimated to occur as a consequence of passive exposure to tobacco smoke. Tobacco smoke causes an increased incidence of all four histologic types of lung cancer, although adenocarcinoma (particularly the bronchioloalveolar variant) is also found in nonsmokers. Other risk factors for lung cancer include exposure to asbestos or radon. Familial factors such as polymorphisms in carcinogen-metabolizing hepatic enzyme systems may also play a role in determining an individual's propensity to develop lung cancer.

II. MOLECULAR BIOLOGY

Numerous genetic changes have been associated with lung tumors. Most common among these include activation or overexpression of the myc family of oncogenes in SCLC and NSCLC and of the KRAS oncogene in NSCLC, particularly adenocarcinoma. Inactivation or deletion of the p53 and retinoblastoma tumor suppressor genes and a tumor suppressor gene on chromosome 3p (the FHTT gene) have been found in 50% to 90% of patients with SCLC. Abnormalities of p53 and 3p have been associated with 50% to 70% of cases of NSCLC. The KRAS mutation is more frequently found in smokers, those with adenocarcinoma, and those with poorly differentiated tumors. It is also associated with poor prognosis.

Recently, abnormalities in the epidermal growth factor receptor (EGFR) pathway have been identified, making EGFR an attractive molecular target for anticancer therapy. EGFR is expressed or overexpressed in the majority of NSCLC tumors. Binding of ligand to the extracellular domain of EGFR causes receptor dimerization, which in turn activates an intracellular tyrosine kinase domain. Autophosphorylation of the receptor induces a cascade of signal transduction events leading to cell proliferation, inhibition of apoptosis, angiogenesis, and invasion, all resulting in tumor growth and spread. Agents targeting EGFR include tyrosine kinase inhibitors (TKIs), such as gefitinib and erlotinib, and anti-EGFR monoclonal antibodies (MoAbs), such as cetuximab and pa-nitumumab. Tumors harboring activating EGFR gene mutations, which render the cancer highly dependent on EGFR for proliferation and survival, often have dramatic and sustained responses to EGFR TKIs. KRAS and EGFR mutations are rarely found in the same tumor.

Still more recently, abnormalities in the anaplastic lymphoma kinase (ALK) gene have been identified in a subset of lung cancers. These molecular aberrations, which appear mutually exclusive with both EGFR and KRAS gene mutations, appear to render cancers highly responsive to ALK inhibitors in phase I studies. These agents are currently undergoing phase II and III testing.

III. SCREENING

Three U.S. randomized screening studies in the 1980s failed to detect an impact on mortality of screening high-risk patients with chest radiography or sputum cytology, although earlier-stage cancers were detected in the screened groups. Since then, however, low-dose spiral computed tomography (CT) has emerged as a possible new tool for lung cancer screening. Spiral CT is CT imaging in which only the pulmonary parenchyma is scanned, thus negating the use of intravenous contrast medium and the necessity of a physician having to be present. This type of scan can usually be done quickly (within one breath) and involves low doses of radiation. In a nonrandomized, controlled study from the Early Lung Cancer Action Project, low-dose CT was shown to be more sensitive than chest radiography in detecting lung nodules and lung cancer at early stages. However, despite these promising results, it is unclear whether screening with spiral CT will result in a reduction in lung cancer mortality. Potential methodologic issues pertaining to this and other screening studies include lead-time bias, length bias, and overdiagnosis. Overdiagnosis refers to the possibility that small tumors preferentially detected by screening would otherwise remain clinically silent until death from other causes. Furthermore, in some geographic regions, such as the midwestern United States, the incidence of benign nodules is relatively high, resulting in frequent follow-up biopsies that contribute to both the cost and morbidity of screening. To address these issues, the National Cancer Institute has recently completed a large randomized controlled trial (the Lung Screening Study) involving at least 15,000 participants over several years, and initial trial results found 20% fewer cancer deaths among the CT screened participants compared with those screened with chest radiographs.

IV. NSCLC

A. Histology

Until recently, the histologic subtype of NSCLC, while thought to influence the presentation and natural history ofdisease, did not affect patient management. Due to differences in both efficacy and safety, histologic designation now represents a primary consideration in treatment selection. For instance, bronchioloalveolar carcinoma, which has a predilection for younger, never-smoking women, is frequently associated with EGFR gene mutations and thus a high likelihood of response to EGFR inhibitors. Pemetrexed, a multitargeted antifolate, has greater efficacy for the treatment of nonsquamous tumors, presumably due to higher thymidylate synthase levels in squamous cell cancers. Bevacizumab, a MoAb directed against vascular endothelial growth factor (VEGF), is contraindicated in patients with squamous cell tumors due to unacceptably high rates of life-threatening hemoptysis in early phase clinical trials. These histology-dependent safety and efficacy distinctions have highlighted the importance of accurate pathologic classification.

B. Staging

The prognosis and treatment of NSCLCs are dependent primarily on stage of disease at the time of diagnosis. Major changes in the staging of lung cancer were adopted in 2009. These changes are based on an analysis of68,463 patients with NSCLC worldwide; by contrast, the previous (1997 and 2002) editions of the TNM classification of lung cancer were based on data from 5,319 patients in North America. The 2009 TNM staging classification is shown in Tables 6.1 and 6.2; the changes are summarized in Table 6.3. Major changes include subclassification of T1 and T2 by tumor size, reclassification of additional nodule(s) in the same lobe or another ipsilateral lobe, and reclassification of malignant effusions as M1a. This reflects the similar prognosis and treatment (typically chemotherapy alone) of patients with malignant effusions and patients with distant disease. As before, a pleural or pericardial effusion is considered malignant if it has any of the following characteristics: positive cytology, exudative, or hemorrhagic. Survival based on the 2009 staging classification is shown in Table 6.4.

C. Pretreatment evaluation

The diagnosis oflung cancer is usually made by bronchial biopsy or percutaneous needle biopsy. A CT scan of the chest is necessary to evaluate the extent of the primary disease, mediastinal extension or lymphadenopathy, and the presence or absence of other paren-chymal nodules in patients in whom surgical resection is a consideration. The upper abdomen is included to evaluate for hepatic or adrenal metastases. Bone scans should be obtained for the patient with bone pain or an elevated calcium or alkaline phosphatase level. Head CT or magnetic resonance imaging (MRI) is not routinely done in the absence of central nervous system (CNS) signs or symptoms. Because the presence of mediastinal nodal metastases is a key factor in determining tumor resectability, lymph node sampling by medi-astinoscopy, Chamberlain procedure (which samples station 5 and 6 nodes not accessible by mediastinoscopy), and/or endobronchial ultrasound is recommended in most instances when there is not clear evidence of distant disease.

f0097-01

f0098-01

Positron emission tomography (PET), a metabolic imaging scan using fluorodeoxyglucose (FDG), has emerged as a useful staging modality. PET scans are more sensitive and specific than CT scans and could thus potentially save patients with advanced disease, either within or outside of the chest, from unnecessary invasive procedures. However, it is not yet clear as to whether PET scanning can replace mediastinal lymph node sampling, as the scan can be falsely positive in inflammatory processes and falsely negative in lung tumors with low metabolic activity such as bronchioloalveolar carcinoma or carcinoid tumors. Furthermore, due to high background FDG uptake in the brain, PET-CT scans are generally not sufficient to evaluate for brain metastases, and a head CT or MRI should be performed. PET scans are frequently performed in conjunction with CT imaging (PET-CT scans) to provide enhanced anatomic detail. Randomized clinical trials have demonstrated that use of PET-CT scans decreases the total number of thoracotomies and number of futile thoracotomies performed. Population-based studies suggest that increasing use of PET-CT scans may have resulted in stage shifting.

f0099-01

f0099-02

Pulmonary function testing is generally recommended before surgery and, if severe pulmonary disease is clinically apparent, before radiation therapy. Increased postoperative morbidity is associated with a predicted postoperative 1-second forced expiratory volume of less than 800 to 1000 mL, a preoperative maximum voluntary ventilation less than 35% of predicted, a carbon monoxide diffusing capacity less than 60% of predicted, and an arterial oxygen pressure of less than 60 mm Hg or a carbon dioxide pressure of more than 45 mm Hg.

D. Management of early stage NSCLC

1. Stage I disease. Surgical resection is the mainstay of treatment for stage I NSCLC, with cure rates of 60% to 80%. Lobectomy is considered superior to smaller procedures such as wedge resection. An exception may be bronchioloalveolar cancer, which spreads by lepidic (airway) growth rather than hematogenous or lymphatic spread and may be adequately treated with more focal excision. Clinical studies evaluating this approach are ongoing. If not performed preoperatively, it is recommended that mediastinal lymph nodes be sampled at the time of resection to complete staging.

In patients with medical contraindications to surgery but with adequate pulmonary function, conventional fractionated radiotherapy (e.g., 6000 cGy in 30 fractions of 200 cGy each) results in cure in about 20% of patients. Recently, advances in imaging and radiation delivery have led to the use of stereotactic radiation therapy for lung tumors. With this technology, radiation delivery to surrounding normal lung parenchyma is substantially less than that occurring with conventional radiation. Thus, it is possible to give much higher, “ablative” radiation doses over a small number of fractions (e.g., 20 Gy per fraction for three fractions). To date, outcomes with this technique appear promising, with 2-year local control rates in excess of 90%. Clinical trials of stereotactic radiation for early stage lung cancer in both medically operable and inoperable patients are ongoing.

The rationale for adjuvant chemotherapy in patients with early-stage lung cancer is based on the observation that distant metastases are the most common site of failure following potentially curative surgery. Interest in this treatment strategy grew after a 1995 meta-analysis of over 4300 patients, in which those who received cisplatin-based regimens had a survival benefit nearing statistical significance (p= 0.07). Since then, a number of randomized clinical trials have evaluated the role of adjuvant chemotherapy following resection of early-stage NSCLC (see Table 6.5). In a pooled analysis of five of these trials, the hazard ratio for death was 0.89 (95% confidence interval [CI], 0.82–0.96; p = 0.005), corresponding to a 5-year absolute benefit of 5.4% from chemotherapy. Importantly, the benefit of chemotherapy varied considerably by stage. For stage IA NSCLC, adjuvant chemotherapy resulted in a trend toward worse survival (hazard ratio [HR] for death 1.40; 95% CI, 0.95–2.06). For stage IB disease, the HR was 0.93 (95% CI, 0.78–1.10). Similarly, in the CALGB 9633 trial of patients with stage IB disease randomized to surgery alone or surgery followed by carboplatin-paclitaxel, only those patients with tumors ≥4 cm demonstrated a significant survival difference in favor of adjuvant chemotherapy (HR, 0.69; 95% CI, 0.48–0.99; p = 0.04). Given these data, it seems reasonable to discuss the option of adjuvant platinum-based doublet chemotherapy with good performance status (PS) patients with completely resected stage IB disease, particularly those with tumors ≥4 cm. Additional studies are needed for stage IA disease before adjuvant therapy can be routinely recommended for this group of patients.

f0101-01

Patients with resected stage I NSCLC are also at high risk for the development of second lung cancers (about 2% to 3% per year). To date, however, secondary prevention efforts have proven unsuccessful. Neither vitamin A nor its derivatives, β-carotene or os-retinoic acid, have been found to have any benefit in chemoprevention, and contrary to predictions, they may even be deleterious. A recent phase III trial of selenium for secondary prevention was closed early when an interim analysis showed no benefit.

2. Stage II disease. Surgical resection is a standard component of the treatment of stage II NSCLC. Patients with peripheral chest wall invasion (T3N0) should undergo resection of the involved ribs and underlying lung. Chest wall defects are then repaired with chest wall musculature or surgical mesh and methylmethacrylate. Postoperative radiotherapy is often given. Five-year survival rates as high as 50% have been reported.

The role of adjuvant chemotherapy for resected stage II NSCLC is clearer than for stage I disease. In the Adjuvant Navelbine International Trialist Association (ANITA) trial, patients with stage IB to IIIA NSCLC were randomized to surgery alone versus surgery followed by four cycles of cisplatin plus vi-norelbine. Overall survival was significantly improved at 5 years (51% versus 43%), although the survival benefit was restricted to patients with stage II and IIIA disease. In the pooled analysis of cisplatin-based adjuvant chemotherapy trials, patients with stage II NSCLC had a significant survival benefit (HR 0.83; 95% CI, 0.73–0.95). Accordingly, adjuvant chemotherapy is generally recommended following complete resection of stage II NSCLC.

Issues in the use of adjuvant therapy have also included identification of the most appropriate drugs. Given the toxicity and tolerability of cisplatin, there was an interest in substituting carboplatin for adjuvant treatment of NSCLC. Based on available data, it is generally accepted that carboplatin should not routinely be used in lieu of cisplatin, but can be considered in patients that would be considered high risk for cisplatin.

The International Adjuvant Lung Cancer, BR10, and ANITA trials all used vinca alkaloids in combination with cisplatin. Given that there is no major difference in chemotherapy doublets in advanced disease, many clinicians have extrapolated that data in advanced disease to earlier stage disease and are using other “third generation” drugs in combination with cis-platin (such as pemetrexed, docetaxel, and gemcitabine), albeit without level 1 data.

Neoadjuvant (preoperative) chemotherapy has also been studied for resectable NSCLC (see Section IV.D.3). Compared to adjuvant chemotherapy, it offers the potential advantages of reducing tumor volume before surgery (which might simplify resection), demonstrating in vivo chemosensitivity, addressing mi-crometastatic disease earlier, and possibly being better tolerated. Although phase III trials comparing neoadjuvant platinum-based regimens with surgery alone have demonstrated the feasibility of this approach, there is no level 1 data showing a benefit for neoadjuvant compared to adjuvant therapy. In addition, patients who undergo a pneumonectomy following induction chemoradiation have a higher incidence of treatment-related deaths.

a. Pancoast tumors. Pancoast tumors are upper lobe tumors that adjoin the brachial plexus and are frequently associated with Horner syndrome or shoulder and arm pain; the latter is due to rib destruction, involvement of the C8 or T1 nerve roots, or both. These tumors are often treated with preopera-tive chemoradiation, surgery, and then additional postoperative chemotherapy. With this approach, the preoperative chemoradiation facilitates resection in an area where neural structures might otherwise limit surgical options. Five-year survival rates range from 25% to 50%.

3. Locally advanced (stage IIIA and IIIB) disease. Treatment of locally advanced NSCLC is one of the most controversial issues in the management of lung cancer. Interpretation of the results of clinical trials involving patients with locally advanced disease has been clouded by a number of issues including changing diagnostic techniques, different staging systems, and heterogeneous patient populations that may have disease that ranges from “nonbulky” stage IIIA (clinical N1 nodes, with microscopic N2 nodes discovered only at the time of surgery or mediastinos-copy) to “bulky” N2 nodes (enlarged adenopathy clearly visible on chest radiographs or multiple nodal level involvement) to clearly inoperable stage IIIB disease.

a. Nonbulky stage IIIA disease. The optimal treatment for non-bulky stage IIIA generally consists of a local approach (surgery or radiation therapy) plus a systemic treatment (chemotherapy). Current investigational efforts are directed at identifying the optimal combined-modality approach. Possibilities include surgery followed by adjuvant chemotherapy, preoperative (neoadjuvant) chemotherapy followed by surgery, chemotherapy plus radiation therapy (either concurrent or sequential), or a trimodality approach.

The potential benefit of adding surgery to combined chemoradiation for stage IIIA NSCLC has been evaluated in a recent randomized phase III intergroup trial. In this study, 396 patients with stage T1–3N2M0 NSCLC were randomized to concurrent chemoradiation (45 Gy) with cisplatin-etoposide followed by either surgical resection or continuation of radiation therapy to 61 Gy total. Although progression-free survival was significantly longer in the surgery arm (12.8 months versus 10.5 months; p = 0.02), there was no significant difference in overall survival (23.6 months versus 22.2 months; p =0.24). There were greater treatment-related mortalities in the surgery arm as compared to the chemoradiation alone arm (8% versus 2%), particularly for patients undergoing a pneumonectomy.

Several studies have shown, in subset analyses, that those patients receiving neoadjuvant therapy who subsequently have their N2 nodes “cleared” with preoperative therapy do better than those who do not. As of this writing, there is no level 1 evidence to recommend neoadjuvant chemotherapy over adjuvant chemotherapy, although several theoretical reasons for doing so include the fact that patients are more likely to tolerate preoperative chemotherapy over postoperative chemotherapy.

Occasionally, despite preoperative staging, patients thought to have stage I or II disease are found to have N2 nodal involvement at the time of surgery. For these stage III patients, postoperative radiation therapy (PORT; 50–54 Gy) may be considered for fit patients, preferably after completion of adjuvant chemotherapy, based on retrospective and nonrandomized studies demonstrating benefit. Currently, PORT is not recommended for patients with less than N2 nodal involvement.

b. Bulky stage IIIA (N2) and stage IIIB. Bulky stage IIIA and IIIB tumors are generally considered unresectable, with treatment consisting of combined chemoradiation or, in the case of malignant pleural or pericardial effusions (M1a in the new staging classification), chemotherapy alone.

(1) Chemotherapy plus radiation therapy. Chemotherapy plus radiotherapy is the treatment of choice for patients with bulky or inoperable stage IIIA or IIIB disease without pleural effusion. Numerous randomized studies have demonstrated an improvement in median and long-term survival with che-motherapy plus radiation therapy versus radiation therapy alone. Active areas of investigation include choice of che-motherapy, fractionation, and treatment fields.

A randomized Japanese trial reported a 3-month survival advantage with concurrent chemoradiation over a sequential approach. Initial reports from a confirmatory randomized Radiation Therapy Oncology Group trial also showed a trend in favor of concurrent cisplatin and vinblastine with radiation over sequential chemoradiation, albeit with more toxicities, making concurrent chemoradiation therapy the treatment of choice for good PS patients.

Chemotherapy can be given in full “systemic” doses with radiotherapy, in weekly “radiosensitizing” doses, or a combination of both. One of the most commonly used chemotherapy regimens for stage Hi NSCLC is carboplatin in combination with paclitaxel (Table 6.6). Although single agent weekly carboplatin alone has not resulted in a survival benefit when given with radiotherapy, weekly doses of paclitaxel at 50 mg/m2 and carboplatin area under the curve (AUC) 2 with concurrent radiation have proved promising in randomized phase II studies. Generally, concurrent therapy is followed by two cycles of full-dose carboplatin AUC 6 plus paclitaxel at 200 mg/m2 every 21 days to treat micrometastatic disease. By contrast, with concurrent radiation therapy and commonly used cisplatin-etoposide chemotherapy regimens, drug doses during radiation therapy are considered “systemic” and may not require additional chemotherapy after radiation therapy is completed. No survival benefit has been shown for “consolidation” therapy.

f0105-01

4. Stage IV disease. Chemotherapy improves survival in patients with metastatic NSCLC (about 10% 1-year survival rate in untreated patients versus 30% to 35% 1-year survival rate with treatment). The principal factors predicting response to chemotherapy and survival are PS and extent of disease. Patients with a poor PS (Eastern Cooperative Oncology Group [ECOG] PS of 2–4) are less likely to respond to treatment and will tolerate the therapy poorly, although recent subset retrospective analysis has suggested that PS2 patients may also enjoy a modest benefit in survival with treatment. Favorable prognostic factors include female sex, normal serum lactic dehydrogenase level, absence of bone or liver metastases, and absence of weight loss.

a. First-line chemotherapy. Systemic treatment for patients with metastatic NSCLC and adequate PS (ECOG 0–1) generally includes platinum-based doublet chemotherapy. A meta-analysis of large randomized trials indicated that there is a small but significant survival advantage with platinum-based therapy compared with best supportive care. Whereas best supportive care resulted in median survival rates of 4 to 5 months and 1-year survival rates of 5% to 10%, current third-generation regimens of platins combined with paclitaxel and docetaxel, gemcitabine, vinorelbine, and pemetrexed have yielded median survivals of 8 to 9 months and 1-year survivals of 35% to 40%. In addition, randomized studies have shown an improvement in symptoms and quality of life compared with patients treated with best supportive care.

b. Choice of chemotherapy. The common chemotherapyregimens for advanced NSCLC are shown in Table 6.7. Historically, randomized studies have failed to show a major advantage of one new doublet regimen over another. Due to a favorable toxicity profile, carboplatin-paclitaxel was the most frequently used combination in the United States. More recently, however, certain agents have been restricted to specific histologies. For instance, pemetrexed, a multitargeted antifolate, has greater efficacy for the treatment of nonsquamous tumors, presumably due to higher thymidylate synthase levels in squamous cell cancers; pemetrexed is approved only for nonsquamous NSCLC in all settings (first-line, maintenance, and second-line). Bevacizumab, a MoAb directed against VEGF, is contraindicated in patients with squamous cell tumors due to unacceptably high rates of life-threatening hemoptysis in early phase clinical trials.

Although a direct comparison of cisplatin-based therapies and carboplatin-based therapies is limited, meta-analyses have suggested that cisplatin may have a small benefit in terms of survival over carboplatin, albeit with a different toxicity profile. Whereas this small difference may be of limited clinical consequence for patients with metastatic disease, it may be more important in the adjuvant setting, where cure is the goal.

f0107-01

c. Inhibitors of angiogenesis. Inhibition of angiogenesis is based on the observations that neovascularization occurs in tumor tissues and rarely in other physiologic processes except wound healing. Although many antiangiogenesis agents are under investigation, such as VEGF receptor TKIs, the drug that has been shown to have survival benefit in NSCLC is the anti-VEGF MoAb bevacizumab (Avastin).

Bevacizumab has been evaluated in a randomized ECOG trial in combination with standard cytotoxic chemotherapy for advanced NSCLC, in which 878 chemonaïve patients with advanced NSCLC were randomized to receive carboplatin and paclitaxel with or without bevacizumab (15 mg/kg) every 3 weeks for six cycles. Bevacizumab was continued for up to 1 year in patients with nonprogressing disease. Patients with squamous cell histology were excluded based on phase II data showing increased hemorrhagic events as a complication with bevacizumab therapy. The study demonstrated an improvement in median survival (12.3 months versus 10.3 months; p 0.001), overall response rates (35% versus 15%;p 0.001), and progression-free survival (6.2 months versus 4.5 months; p 0.001), favoring the bevacizumab arm. A European study showed a small improvement in progression-free survival with 7.5 mg/kg or 15 mg/kg of bevacizumab plus gemcitabine and cisplatin, compared to gemcitabine plus cisplatin alone, but no improvement in overall survival.

d. Inhibitors of EGFR. EGFR, also known as human epidermal growth factor receptor 1 (HER1) or ErbB1, is a transmembrane receptor tyrosine kinase. On ligand binding, receptor subunits dimerize, resulting in autophosphorylation of intracellular tyrosine residues and initiation of a signal transduction cascade resulting in cellular proliferation, resistance to apoptosis, cellular invasion, metastasis, and angiogenesis. EGFR-inhibiting drugs are classified as TKIs (also called small molecule inhibi-tors because of their low molecular weight) or MoAbs.

TKIs are almost always oral drugs, taken on a daily basis. They are often metabolized via the cytochrome P450 system and therefore prone to drug-drug interactions. Most of these drugs have names ending in “-ib.” MoAbs are always administered intravenously and may be associated with acute infusion reactions. They have a molecular weight of approximately 150,000 daltons and have names ending in “-ab.” From a mechanistic perspective, EGFR TKIs and anti-EGFR MoAbs inhibit the activation of EGFR via direct binding to the kinase activation site or by blocking ligand-receptor binding, respectively. MoAbs may potentially also exert anticancer effects via recruitment of endogenous immune functions (e.g., antibody-dependent cellular cytotoxicity and complement-mediated cytotoxicity). The primary toxicities of EGFR TKIs are acneiform rash and diarrhea. Anti-EGFR MoAbs are also associated with acneiform rash.

Four clinical trials have randomized over 4000 total patients to platinum doublet chemotherapy with or without an EGFR TKI (two studies with gefitinib; two studies with erlotinib). None of these studies demonstrated a survival benefit. Potential explanations for these disappointing results include pharmacodynamic antagonism (EGFR TKI-induced cell-cycle arrest could reduce the efficacy of cytotoxic chemotherapy agents) and enrollment of a nonen-riched population. However, in the first-line Iressa Pan-Asia Study (IPASS), in which over 1200 previously untreated nonsmokers or former light smokers in East Asia who had advanced adenocarcinoma NSCLC were randomized to ge-fitinib at 250 mg orally daily or carboplatin-paclitaxel, progression-free survival was superior in the gefitinib arm (HR 0.74; p0.001), with 12-month progression-free survival rates 25% for gefitinib and 7% for carboplatin-paclitaxel. Among those patients whose tumors harbored EGFR mutations, progression-free survival was significantly longer with gefitinib (HR 0.48; p 0.001); by contrast, in the mutation-negative group, gefitinib resulted in significantly shorter progression-free survival compared with carboplatin-pa-clitaxel (HR 2.85;p 0.001). Although mature survival data is pending, these results suggest that single-agent gefitinib may be an appropriate first-line therapy in patients with EGFR mutations.

A phase III clinical trial of cisplatin-vinorelbine with or without the anti-EGFR antibody cetuximab in patients with EGFR-expressing NSCLC (80% to 90% of cases) demonstrated a modest but statistically significant increase in survival: 11.3 months versus 10.1 months (p = 0.04). Whether the discrepant results with EGFR TKIs or an anti-EGFR antibody combined with chemotherapy reflect differences in drug efficacy, study design, or other factors is not clear.

e. Duration of therapy/maintenance therapy. Four randomized studies failed to show a survival difference with “prolonged” (more than six) cycles of chemotherapy compared with a fewer (four to six) number of cycles. Thus, until recently, continuing chemotherapy until progression has not been routinely recommended.

This approach has been challenged by recent randomized clinical trials of “maintenance” chemotherapy. The term maintenance in this context requires clarification. In some instances, it may refer to continuation of initial therapy (as in the studies described previously). In other instances, maintenance refers to continuation of a component of initial therapy (as with the ongoing administration of bevacizumab or cetuximab following completion of a maximum of six cycles of combination chemotherapy). Finally, maintenance may refer to immediate introduction of another agent on completion of first-line chemotherapy (sometimes called “switch” maintenance). It is this concept that has been evaluated in two recent phase III studies.

In one trial, 309 patients were randomized to immediate versus delayed (instituted at time of disease progression) docetaxel at 75 mg/m2 every 21 days on completion of four cycles of carboplatin-gemcitabine. Median progression-free survival was significantly longer in the immediate docetaxel group (5.7 months versus 2.7 months; p = 0.0001). Median overall survival was also longer, although the difference was not statistically significant (12.3 months versus 9.7 months; p = 0.09). Quality of life did not differ significantly between the two groups (p= 0.76). In another trial, 663 patients who had not progressed on four cycles of platinum-based chemotherapy were randomized 2:1 to maintenance pemetrexed at 500 mg/m2 every 21 days or placebo until progression. Pemetrexed significantly improved median progression-free survival (4.3 months versus 2.6 months;p 0.0001) and median overall survival (13.4 months versus 10.6 months; p = 0.01), benefits that were limited to nonsquamous tumors.

Although a substantial proportion of thoracic oncologists have adopted maintenance chemotherapy, a number of questions remain. First, how maintenance therapy fits into the increasingly complex overall treatment paradigm for advanced NSCLC is not clear, as some first-line regimens (such as those incorporating bevacizumab or cetuximab) continue treatment until progression. Second, it is not clear whether these studies demonstrate a benefit from early use of docetaxel or pemetrexed, or rather a benefit from exposure— regardless of timing—to these drugs.

f. Non-platinum–based regimens. Given the toxicities associated with platinum-based chemotherapy, particularly cisplatin, there is considerable interest in combining two nonplatinum drugs. The majority of recent randomized trials have failed to show a significant difference in survival with platinum regimens compared with non-platinum-based regimens, although the toxicity profile is different.

g. Patients with poor performance status. Patients with ECOG PS2 may be treated with single agent cytotoxic chemotherapy. The most commonly studied agents include vinorelbine and the taxanes. Patients with ECOG PS3–4 are generally not offered chemotherapy. A potential exception is a patient with a tumor harboring an activating EGFR mutation (see subsequent discussion). For such a patient, an EGFR TKI could result in a dramatic response without conveying substantial toxicity.

h. Oligometastatic disease. In certain circumstances, definitive local therapy of both thoracic disease and a metastatic site may be considered. In patients with controlled disease outside of the brain who have an isolated cerebral metastasis in a resectable area, resection followed by whole brain radiation is superior to whole brain radiotherapy alone. This oligometastatic approach appears most beneficial in patients with stage I thoracic disease, where survival approximates that of stage I patients without brain metastases. In patients with locally advanced thoracic disease, the benefit of metastasectomy is less clear. Stereotactic radiosurgery may be another option for these patients. Adrenalectomy for an isolated adrenal metastasis may be associated with up to 25% 5-year survival. Outcomes appear to be superior for patients with a metachronous rather than a synchronous metastasis.

5. Second-line chemotherapy. Docetaxel, pemetrexed, and erlotinib are currently approved by the U.S. Food and Drug Administration (FDA) for second-line monotherapy for patients with meta-static NSCLC.

a. Docetaxel. There have been two randomized trials evaluating second-line docetaxel versus best supportive care in patients who have failed first-line therapy. Docetaxel at a dose of 75 mg/m2 every 3 weeks significantly prolongs survival in comparison with best supportive care and, in comparison with either vinorelbine or ifosfamide, improves time to progression and 1-year survival. Moreover, it also improves quality of life. It was noted that previous paclitaxel exposure did not affect patients' response to docetaxel, suggesting no cross-resistance between the two taxane agents.

b. Pemetrexed. Pemetrexed has similar antitumor activity as doc-etaxel in the second-line setting but with less toxicity. In a randomized trial, patients were treated with pemetrexed at 500 mg/m2 or docetaxel at 75 mg/m2 every 3 weeks. Overall response rates were similar (9.1% versus 8.8% for pemetrexed and docetaxel, respectively) with no differences in median survival (8.3 months versus 7.9 months for pemetrexed and docetaxel, respectively). Docetaxel was associated with higher rates of neutropenia, neu-tropenic fever, and hospitalization due to neutropenic events or other drug-related adverse events as compared to pemetrexed. A post hoc analysis by histology demonstrated a selective benefit for pemetrexed in nonsquamous histology.

c. EGFR TKIs

(1) Gefitinib. Gefitinib was granted FDA approval for previously treated advanced NSCLC based on the outcomes of phase II studies. However, the subsequent phase III Iressa Survival Evaluation in Lung Cancer (ISEL) trial of over 1,600 patients failed to demonstrate a significant survival benefit compared to placebo (median 5.6 months versus 5.1 months; p = 0.09). A subgroup analysis showed significant benefit for patients of Asian origin and in nonsmokers (median survival 8.9 months versus 6.1 months; p = 0.012). Following the negative survival result, gefitinib was relabeled for use restricted to patients already receiving and benefiting from the drug or patients participating in clinical trials. This essentially removed the drug from the American and European markets, although it remained approved and widely used in Asia. Recently, interest in gefitinib has risen again in response to highly encouraging outcomes in clinically and molecularly enriched patient populations.

(2) Erlotinib. Erlotinib is a similar orally available EGFR/TKI. Single-agent treatment with erlotinib for previously treated advanced NSCLC was evaluated in the National Cancer Institute of Canada BR21 trial. This study randomized 731 patients with stage IIIB or IV NSCLC who had failed one or two prior treatment regimens in a 2:1 ratio to receive erlotinib at 150 mg orally daily versus placebo. The overall response rate for erlotinib was 9% versus 1% for placebo (p 0.001). Stable disease was observed in 35% of patients on the erlotinib arm as compared to 27% of patients on placebo. In contrast to the results with gefitinib, there was a significant improvement in progression-free survival (2.2 months versus 1.8 months; p0.001) and overall survival (6.7 months versus 4.7 months; p 0.001), in favor of erlotinib. While all patient subtypes in BR21 derived a survival benefit, the hazard ratio of 0.4 for never-smokers was statistically significantly different than the hazard ratio of 0.9 for smokers (p 0.001).

A number of explanations for the different outcomes of the phase III ISEL (gefitinib) and BR21 (erlotinib) trials have been proposed. Erlotinib may have been administered at a more biologically effective dose, as evidenced by higher rates of rash and diarrhea in the BR21 study. The ISEL study may have selected for patients with particularly aggressive disease, as subjects were required to have disease relapse or progression within 90 days after prior platinum chemotherapy. The BR21 study did not specify timing of disease relapse or progression.

(3) Identification of the target population. Clinical parameters that appear to predict response to EGFR TKIs include never-smoking history, East Asian ethnicity, adenocarci-noma histology (particularly tumors with bronchioloal-veolar features), and female gender. Molecular analysis of tumor specimens from individuals with these characteristics has revealed high rates of activating mutations in the EGFR tyrosine kinase domain. These mutations hyperac-tivate the EGFR tyrosine kinase, rendering cancer cells highly dependent on EGFR oncogenic pathways and thus exquisitely sensitive to EGFR inhibition. Among patients with “classic“ EGFR mutations (exon 19 in-frame deletions of amino acids 747–750 and exon 21 L858R substitutions), response rates to EGFR TKIs exceed 60% and median survival exceeds 2 years. EGFR amplification and gene copy number (determined by fluorescence in situ hybridization) have also been employed to predict a survival benefit to these therapies. EGFR protein expression, determined by immunohistochemistry, is not as strongly correlated with outcomes. Additionally, development of acneiform rash has been associated with improved survival for both EGFR TKIs and anti-EGFR MoAbs in multiple disease settings, including colorectal cancer, pancreatic cancer, and NSCLC.

Clinical and molecular enrichment strategies are exemplified by the recent phase III first-line IPASS trial. In this trial, over 1200 previously untreated nonsmokers or former light smokers in East Asia who had advanced adenocarcinoma NSCLC were randomized to gefitinib at 250 mg orally daily or carboplatin-paclitaxel. Despite a highly clinically enriched population, only 60% of patients had tumors harboring EGFR mutations. Among these patients, progression-free survival was significantly longer with gefitinib (HR 0.48; p 0.001); by contrast, in the mutation-negative group, gefitinib resulted in signifi-cantly shorter progression-free survival compared with carboplatin-paclitaxel (HR 2.85; p 0.001). Early survival results did not show a difference between the gefitinib arm and the carboplatin-paclitaxel arm, presumably because patients were allowed to crossover to the other arm on progression. With outcomes such as these supporting EGFR mutations as strong predictors of benefit from EGFR TKIs but data from second-line trials demonstrating a survival benefit from EGFR TKIs in unselected populations, it remains debated whether and in which patients EGFR mutation testing should be routinely performed.

V. SCLC

SCLC differs from NSCLC in a number of important ways. First, it has a more rapid clinical course and natural history, with the rapid development of metastases, symptoms, and death. Untreated, the median survival time for patients with local disease is typically 12 to 15 weeks and for those with advanced disease 6 to 9 weeks. Second, it exhibits features of neuroendocrine differentiation in many patients (which may be distinguishable histopathologically) and is more commonly associated with paraneoplastic syndromes. Third, unlike NSCLC, SCLC is exquisitely sensitive to both chemotherapy and radiotherapy, although resistant disease often develops. Because of the rapid devel-opment of distant disease and its extreme sensitivity to the cytotoxic effects of chemotherapy, this mode of therapy forms the backbone of treatment for this disease, irrespective of stage.

A. Staging

Although SCLC has a propensity to metastasize quickly and micro-metastatic disease is presumed to be present in all patients at the time of diagnosis, this disease is usually classified into either a local or an extensive stage. Local disease is typically defined as disease that can be encompassed within one radiation port, usually considered limited to the hemithorax and to regional nodes, including mediasti-nal and ipsilateral supraclavicular nodes. Extensive-stage disease is usually defined as disease that has spread outside those areas.

B. Pretreatment evaluation

Common sites of metastases for SCLC include the brain, liver, bone marrow, bone, and CNS. For this reason, a complete staging work-up has traditionally consisted of a complete blood cell count; liver func-tion tests; CT or MRI of the brain; CT of the chest and abdomen; bone scan; and bone marrow aspiration and biopsy. As for NSCLC, PET-CT scans are now routinely employed in the initial staging of SCLC and may be considered in place of the CT chest/abdomen and bone scan. However, this complete staging work-up need not be undertaken unless the patient is a candidate for combined-modality treatment with chest radiation and chemotherapy, the patient is being evaluated for a clinical study, or the information is helpful for prognostic reasons. If the patient is not a candidate for combined-modality treatment or a clinical study, stopping the staging at the first evidence of extensive-stage disease is usually appropriate. Given that isolated bone marrow metastases are rare, bone marrow biopsies and aspirates are not usually done.

C. Prognostic factors

As in NSCLC, the major pretreatment prognostic factors are stage, performance status, and bulky disease. Hepatic metastases also confer a poorer prognosis. Due to the chemosensitivity of SCLC, if a patient's initial poor PS is due to the underlying malignancy, these symptoms often disappear quickly with treatment, resulting in a net improvement in quality of life. However, major organ dys-function from nonmalignant causes often results in an inability of the patient to tolerate chemotherapy.

D. Therapy

1. Combination chemotherapeutic regimens.A number of combination chemotherapeutic regimens are available for SCLC (Table 6.8). No clear survival advantage has been consistently demonstrated for any one regimen over another. With these chemotherapy regimens, overall response rates of 75% to 90% and complete response rates of 50% for localized disease can be anticipated. For extensive-stage disease, overall response rates of about 75% and complete response rates of 25% are common. Despite these high response rates, however, the median survival time remains about 14 months for limited-stage disease and 8 to 9 months for extensive-stage disease. Less than 5% of patients with extensive-stage disease survive more than 2 years.

At present, either cisplatin or carboplatin together with etoposide are the standard of care in North America for the treatment of SCLC. Generally, cisplatin is preferred if given with thoracic radiation for limited-stage disease. For extensive-stage disease, both cisplatin and carboplatin are widely used.

2. Dose intensity. A dose intensity meta-analysis of chemotherapy in SCLC, which evaluated doses not requiring bone marrow trans-plantation support, showed no consistent correlation between dose intensity and outcome. There have been several phase I and II clinical trials evaluating the role of marrow-ablative doses of chemotherapy with subsequent progenitor cell replacement (e.g., autologous bone marrow transplantation) with disap-pointing survival results. In a randomized phase III study, when compared with conventional-dose chemotherapy, a high-dose regimen with stem cell support prolonged relapse-free but not overall survival.

f0115-01

3. Duration of therapy. Most randomized studies do not show a survival benefit for prolonged administration of chemotherapy. Several studies have demonstrated no survival benefit of prolonged first-line treatment over treatment on relapse. The optimal duration of treatment for SCLC is four to six cycles.

4. Second-line therapy. No curative regimens for patients with recurrent disease have been identified. The only drug approved for second-line therapy of SCLC is topotecan, which has a 20% to 40% response rate in patients with sensitive SCLC (those patients who relapsed 2 to 3 or more months after their first-line therapy), with a median survival of 22 to 27 weeks. Other options for patients with sensitive disease include oral etoposide, the combination of cyclophosphamide, doxorubicin, and vincristine, or a return to the first treatment regimen. For patients with refractory disease (progressed through or within 3 months of completion of first-line therapy), the response rate in phase II studies is only between 3% and 11%, and median survival is about 20 weeks.

E. Chemotherapy plus chest irradiation

Numerous studies combining chemotherapy and thoracic radiation therapy have been performed in patients with limited-stage SCLC. Conflicting results have been attributed to differences in chemotherapy regimens and different schedules integrating chemotherapy and thoracic radiation (concurrent, sequential, and “sandwich” approach). Two meta-analyses concluded that thoracic irradiation does result in a small but significant improvement in survival and major control of the disease in the chest, although no conclusions could be made regarding the optimal sequencing of chemotherapy and thoracic radiation. In one randomized study, twice-daily hyperfractionated radiation was compared with a once-daily schedule; both were given concurrently with four cycles of cisplatin and etoposide. Survival was significantly higher with the twice-daily regimen (median survival of 23 months versus 19 months, 5-year survival of 26% versus 16%), albeit at the expense of more grade III esophagitis. In another randomized trial, early administration of thoracic irradiation in the combined-modality therapy of limited-stage SCLC was superior to late or consolidative thoracic irradiation. These data suggest that patients with good PS and with limited disease should receive concurrent chemoradiation, preferably with twice-daily hyperfractionation.

For patients with extensive-stage disease who have complete response to chemotherapy at extrathoracic disease sites, consolidative chest radiation therapy may be considered. In a randomized study, patients who had a complete extrathoracic response and a partial or complete thoracic response after three cycles of cisplatin-etoposide were randomized to either (1) 54 Gy thoracic radiation followed by two additional cycles of cisplatin-etoposide or (2) four additional cycles of cisplatin-etoposide. Patients who received radiation had a significantly longer overall survival (17 months versus 11 months; p = 0.04).

F. Prophylactic cranial irradiation

Given the propensity of SCLC to metastasize to the brain and the resultant morbidity of this event, prophylactic cranial irradiation has been offered to patients with limited-stage disease who have an excellent response to chemoradiation. In a meta-analysis of seven trials, prophylactic cranial irradiation decreased the risk of brain metastasis, prolonged disease-free survival, and significantly increased 3-year overall survival, with a net gain of 5.4%. More recently, a randomized trial of patients with extensive stage SCLC who had had a response to chemotherapy were randomized to prophylactic cranial irradiation or observation. The trial met its primary endpoint with a reduced rate of symptomatic brain me-tastasis (HR 0.27;p 0.001). Furthermore, prophylactic cranial irradiation significantly extended disease-free and overall survival; the 1-year survival rate was 27% in the irradiation group and 13% in the control group.

VI. PALLIATION

A. Radiotherapy

Palliative radiotherapy is often helpful in controlling the pain of bone metastases or neurologic function in patients with brain metastases. Chest radiotherapy may help control hemoptysis, superior vena cava syndrome, airway obstruction, laryngeal nerve compression, and other local complications. For patients with bronchial obstruction who have received maximum external-beam radiotherapy, the use of high-dose endobronchial irradiation may be of temporary benefit.

B. Pleural effusions

For pleurodesis, common sclerosing agents include doxycycline, talc, and bleomycin. The disadvantage of bleomycin is its cost; talc, although effective, has the disadvantage of requiring a thoracoscopy and general anesthesia for insufflation. Alternatively, an indwelling pleural drainage catheter may be placed.

C. Colony-stimulating factors Filgrastim (granulocyte colony-stimulating factor) decreases the incidence of neutropenic fevers, the median duration of neutropenia, days of hospitalization, and days of antibiotic treatment in patients. However, the clinical benefit of maintaining a dose-intense approach in the treatment of patients with lung cancer has not been established. In addition, caution must be exercised when using colony-stimulating factors in patients receiving combined-modality treatment with both chemotherapy and thoracic irradiation. A randomized study by the Southwest Oncology Group found that patients receiving sargra-mostim (granulocyte–macrophage colony-stimulating factor) and chemotherapy with concurrent thoracic irradiation had a significant increase in thrombocytopenia over patients receiving concurrent chemotherapy and radiation therapy without growth factor.

Selected Readings

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 randomized controlled trial. Lancet.2009;374:379–386.

Arriagada R, Bergman B, Dunant A, et al. Cisplatin-based adjuvant chemotherapy in patients with completely resected non-small-cell lung cancer. N Engl J Med. 2004;350:351–360.

Cappuzzo F, Hirsch FR, Rossi E, et al. Epidermal growth factor receptor gene and protein and gefitinib sensitivity in non-small-cell lung cancer. J Natl Cancer Inst. 2005;97:643–655.

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.2009.374:1432–1440.

Dillman RO, Herndon J, Seagren SL, Eaton WL Jr, Green MR. Improved survival in stage III non-small cell lung cancer: seven-year follow-up of CALGB 8433. J Natl Cancer Inst. 1996;88:1210–1215.

Douillard JY, Rosell R, De LM, et al. Adjuvant vinorelbine plus cisplatin versus observation in patients with completely resected stage IB-IIIA non-small-cell lung cancer (Adjuvant Navelbine International Trialist Association [ANITA]): a randomised controlled trial. Lancet Oncol. 2006;7:719–727.

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. 2009;27:591–598.

Furuse K, Fukuoka M, Kawahara M, et al. Phase III study of concurrent versus sequential thoracic radiotherapy in combination with mitomycin, vindesine, and cisplatin in unresectable stage III non-small cell lung cancer. J Clin Oncol. 1999;17;2692–2699.

Gilligan D, Nicolson M, Smith I, et al. Preoperative chemotherapy in patients with resectable non-small cell lung cancer: results of the MRC LU22/NVALT 2/EORTC 08012 multicentre randomised trial and update of systematic review. Lancet. 2007;369:1929–1937.

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.2004;22:1589–1597.

Klasa R, Murray N, Coldman A. Dose-intensity meta-analysis of chemotherapy regimens in small-cell carcinoma of the lung. J Clin Oncol. 1991;9:499–508.

Lynch TJ, Bell DW, Sordella R, et al. Activating mutations in the epidermal growth factor receptor underlying responsiveness of non-small-cell lung cancer to gefitinib. NEngl J Med. 2004;350:2129–2139.

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

Mulshine JL, Sullivan C. Clinical practice. Lung cancer screening. N Engl J ME2005;352:2714-2720.

Murray N, Coy P, Pater J, et al. Importance of timing for thoracic irradiation in the combined modality treatment of limited-stage small-cell lung cancer. J Clin Oncol.1993;11:336–344.

Neal CR, Amdur RJ, Mendenhall WM, Knauf DF, Block AJ, Million RR. Pancoast tumor: radiation therapy alone versus preoperative radiation therapy and surgery. Int J Radiat Oncol Biol Phys.1991;21:651–660.

Non–Small Cell Lung Cancer Collaborative Group. Chemotherapy in non–small cell lung cancer: a meta-analysis using updated data on individual patients from 52 randomized clinical trials. BMJ.1995;311:899–909.

Pignon JP, Arriagada R, Ihde DC, et al. A meta-analysis of thoracic radiotherapy for small-cell lung cancer. N Engl J Med. 1992;327:1618–1624.

Pignon JP, Tribodet H, Scagliotti GV, et al. Lung adjuvant cisplatin evaluation: a pooled analysis by the LACE Collaborative Group. J Clin Oncol. 2008;26:3552–3559.

Pirker R, Pereira JR, Szczesna A, et al. Cetuximab plus chemotherapy in patients with advanced non-small cell lung cancer (FLEX): an open-label randomized phase III trial. Lancet. 2009;373:1525–1531.

Rami-Porta R, Crowley JJ, Goldstraw P. The revised TNM staging system for lung cancer. Ann Thorac Cardiovasc Surg. 2009;15:4–9.

Rosell R, Gómez-Codina J, Camps C, et al. A randomized trial comparing preoperative chemotherapy plus surgery with surgery alone in patients with non–small-cell lung cancer. N Engl J Med.1994;330:153–158.

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

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.2008;26:3543–3551.

Schiller JH, Harrington D, Belani C, et al. Comparison of four chemotherapy regimens for advanced non–small cell lung cancer. NEngl J Med. 2002;346:92–98.

Shepherd FA, Rodrigues Pereira J, Ciuleanu T, et al. Erlotinib in previously treated non-small-cell lung cancer. N Engl J Med. 2005:353:123–132.

Slotman B, Faivre-Finn C, Kramer G, et al. Prophylactic cranial irradiation in extensive small-cell lung cancer. N Engl J Med. 2007;357:664–672.

Strauss GM, Herndon JE 2nd, Maddaus MA, et al. Adjuvant paclitaxel plus carboplatin compared with observation in stage IB non-small-cell lung cancer: CALGB 9633 with the Cancer and Leukemia Group B, Radiation Therapy Oncology Group, and North Central Cancer Treatment Group Study Groups. J Clin Oncol. 2008;26:5043–5051.

Thatcher N, Chang A, Parikh P, et al. Gefitinib plus best supportive care in previously treated patients with refractory advanced non-small cell lung cancer: results from a randomized, placebo-controlled multicentre study (Iressa Survival Evaluation in Lung Cancer). Lancet. 2005;366:1527–1537.

The PORT Meta-analysis Trialists Group. Postoperative radiotherapy in non-small-cell lung cancer: systematic review and meta-analysis of individual patient data from nine randomised controlled trials. Lancet. 1998;352:257–263.

Tsao MS, Sakurada A, Cutz JC, et al. Erlotinib in lung cancer—molecular and clinical predictors of outcome. N Engl J Med 2005;353:133–144.

Turrisi AT, Kim K, Blum R, et al. Twice-daily compared with once-daily thoracic radiotherapy in limited small-cell lung cancer treated concurrently with cisplatin and etoposide. N Engl J Med. 1999;340:265–271.

Von Pawel J, Schiller JH, Shepherd FA, et al. Topotecan versus cyclophosphamide, doxorubicin, and vincristine for the treatment of recurrent small-cell lung cancer. J Clin Oncol. 1999;2:658–667.

Warde P, Payne D. Does thoracic irradiation improve survival and local control in limited-stage small-cell carcinoma of the lung? A meta-analysis. j Clin Oncol 1992;10:890–895.

Winton T, Livingston R, Johnson D, et al. Vinorelbine plus cisplatin vs. observation in resected non-small-cell lung cancer. N Engl J Med. 2005;352:2589–2597.