Stage I Lung Cancer

Stage I disease in NSCLC is completely confined to the lung parenchyma or airway, without nodal spread, and does not arise within 2 cm of the carina. The standard of care for patients with adequate lung function, no other significant comorbid illness (e.g., unstable coronary artery disease, debilitating arthritis), and a nodule with high probability of being lung cancer is surgical resection. Patients with stage I NSCLC who undergo surgery with complete resection of their disease have a high rate of cure with surgery alone (60%-80%).

Based on a randomized trial of NSCLC surgery conducted by the Lung Cancer Study Group, lobectomy is the standard surgery for lung cancer. Beginning in 1982, this study compared outcomes of patients who underwent lobectomy versus segmentectomy or wedge resection. The rate of local recurrence was 75% lower in the lobectomy group, and since then, lobectomy has remained the “gold standard” procedure for healthy patients with stage I disease.

Recent data and advances in imaging have led to a reevaluation of this standard. Specifically, several Japanese studies showed that small (≤2 cm) peripheral tumors with a ground-glass appearance on thin-section computed tomography (CT; ≤3 mm) had a low rate of nodal involvement, with 5-year survival of almost 100%, even after limited anatomic resection. Newer studies are re-addressing the appropriate degree of lung resection for patients with small, stage I tumors. One should also bear in mind that the time frame of the Lung Cancer Study Group trial comparing limited resection with lobectomy predated the widespread use of CT and positron emission tomography (PET) scans. Also, the increased use of CT scans to image the chest results in more tumors being detected at smaller sizes. This smaller average size of tumors should not be interpreted to justify abandoning lobectomy as a standard of care, but randomized controlled trials (RCTs) designed to determine the optimum degree of resection in this more modern setting are ongoing.

For now, the standard procedure for stage I NSCLC remains a lobectomy with lymph node dissection. Exceptions to this rule are usually applied for patients with peripheral tumors and limited pulmonary reserve from emphysema. Many surgeons will offer limited pulmonary resection, with or without concomitant lung volume reduction surgery (LVRS), to patients with severe emphysema. Published series demonstrating the feasibility of LVRS for patients with severe emphysema universally came from groups participating in RCTs of LVRS. This cannot be overlooked as a contribution to the better-than-expected outcomes in these highly select patients. Patients undergoing evaluation for LVRS who also had lung cancer surgery underwent rigorous pulmonary rehabilitation, with surgery in centers with regimented postoperative pain management, early mobilization, chest physiotherapy, and bronchodilator treatment. Lung cancer surgery for patients with severe emphysema should be done in centers with experience in all these strategies. Preoperative staging of these patients must be thorough, to avoid incomplete resection of tumors with nodal metastases.

The presence or absence of nodal involvement is the most important prognostic factor to determine after surgical resection. In addition to lobectomy, mediastinal lymph node dissection is essential to accurate staging of surgically resected lung cancer and must be done as part of any lung cancer resection. Some studies have found a direct correlation between survival and the number of lymph nodes removed at surgery, although this has not been widely validated.

The surgical approach to NSCLC has traditionally employed posterolateral thoracotomy as the standard procedure. The development of video-assisted thoracoscopic surgery (VATS) in the 1990s, followed by its spread through thoracic surgical training programs, has gradually resulted in a greater proportion of NSCLC surgeries being performed by this approach. Various studies show that VATS lobectomy is oncologically equivalent to open lobectomy, such that the tumor-containing lobe is removed intact, and mediastinal lymph node dissection or sampling is completed with equal effectiveness by either approach. VATS lobectomy is associated with less incision pain, reduced length of stay, and faster recovery of preoperative activity levels than open lobectomy. These advantages make VATS the approach of choice for most lobectomies, allowing well-trained thoracic surgeons to offer lobectomy to patients with a broader range of comorbid conditions.

Some patients are not candidates for VATS lobectomy, and occasionally an operation started with the VATS approach is converted to an open procedure at the surgeon’s discretion. The open approach is more suited to large (>3 cm) tumors or central tumors located close to the major blood vessels or airway. Patients with these tumors may benefit from the surgeon’s greater tactile access in an open chest to ensure resection margins are negative and arteries and airways are removed intact. In addition, patients who have had preoperative chemotherapy or radiation, or who have pleural adhesions or prior chest surgery, may not be candidates for VATS. Tumors with chest wall invasion requiring an en bloc resection usually require an open approach as well. When pneumonectomy is required for complete resection, it is generally performed by thoracotomy. In such situations the size of the specimen requires a large incision to facilitate removal of the specimen, reducing the advantages of VATS.

In general, centers that treat the largest number of patients have the best survival outcomes, for all treatment modalities, but especially for the surgical resection of lung cancer. In the United States, approximately 80% of lung cancer resections between 1996 and 2005 were performed by surgeons for whom lung cancer surgery was not the primary focus of their practice. In this same study, thoracic surgeons (those for whom noncardiac thoracic surgery was the primary focus of their practice) had the lowest operative mortality for patients requiring lobectomy or pneumonectomy. Current evidence-based guidelines from the American College of Chest Physicians (ACCP) advocates that patients being considered for lung cancer surgery should be seen by a “thoracic surgical oncologist with a prominent part of his/her practice focused on NSCLC.”

For patients with pathologically confirmed stage I disease, postoperative chemotherapy (adjuvant therapy) is not currently the standard of care. Based on RCT results of postoperative chemotherapy, patients with completely resected, pathologic stage I disease did not have a clear survival benefit with adjuvant chemotherapy. Subgroup analysis of RCTs suggests that patients with the highest risk (based on current understanding of such risk) may benefit from adjuvant therapy. These patients were generally those with large (>4 cm) primary tumors. Many oncologists offer discussions of risks and benefits of adjuvant chemotherapy to patients with large, stage IB tumors (>4 cm) or those with other possible risk factors, such as prominent vascular invasion. On an individualized basis, this practice is reasonable, but adjuvant chemotherapy cannot yet be considered the standard of care for completely resected stage I NSCLC.

For patients with clinical stage I NSCLC who are not candidates for resection because of comorbidity, there are still curative options, and the current standard of care is external beam radiation. An extensive discussion of radiotherapy methods is beyond the scope of this text, but the trend in thoracic oncology for inoperable stage I disease has been toward stereotactic body radiotherapy (SBRT) and away from conventionally fractionated external beam radiotherapy (EBRT). Conventional radiotherapy is given in doses of 60 to 70 grays (Gy), usually divided into 30 fractions of 2 to 2.5 Gy. SBRT is a form of radiotherapy employing many customized lower-dose beams converging into a volume that encompasses the tumor. These treatments are usually given up to a dose of 50 to 60 Gy over 3 to 5 days of treatment. Tumor motion caused by respiratory excursion can be significant, leading to undertreatment of tumor and overexposure of normal tissue to radiation, if the motion of the tumor is not tracked accurately. SBRT can be delivered more accurately with the use of fiducial markers placed in or near the tumor to track it during the respiratory cycle.

Cure rates with SBRT in stage I NSCLC for patients carefully staged can be excellent; 3-year disease-free survival in a large Phase II trial was close to 50%, with most recurrences being distant disease. Patients with large or centrally located tumors are more likely to be treated with conventionally fractionated EBRT; toxicity was greater in those with central tumors in a Phase II study of SBRT. Some groups treating large numbers of patients with SBRT report late development of chest wall pain or rib fractures. The risk factors for rib fracture include tumors within 2 cm of the chest wall and larger treatment volume. Complications in patients treated with SBRT should be tracked closely because it is a relatively new modality used in this patient population.

Thermal ablation is a minimally invasive therapy that can be used for local control of primary lung tumors. Radiofrequency ablation (RFA) uses a radiofrequency probe inserted percutaneously into the tumor, usually under CT guidance, to generate frictional heating that leads to cell death. RFA is the most frequently performed thermal ablation procedure used to treat lung tumors. Although the role of RFA is not clearly defined in lung cancer, small uncontrolled case series support its use for lung metastases. Although RFA is performed more often to treat hepatic and renal tumors, it may be well suited to the lung because of its ability to concentrate thermal energy focally within tumor tissue, with little or no energy spreading to the adjacent aerated normal lung parenchyma. There are no RCTs comparing RFA with radiotherapy in medically inoperable patients.

Based on published data, smaller tumors (<3 cm) may be more effectively treated with RFA than larger tumors. Peripheral tumors surrounded by lung parenchyma and away from hilar structures can be safely treated with RFA. The risk of pneumothorax is significant, as high as 50% in one series. Because this is a newer technology, the role for RFA in treating lung cancer has not been defined by large, well-controlled studies, but centers with experience in this approach can offer this option for inoperable patients, particularly if there is a contraindication to radiotherapy, such as prior irradiation.

A point of emphasis is necessary for defining the term “unresectable” stage I disease. In general, the most qualified person to determine a given patient’s suitability for surgery is an experienced thoracic surgeon with lung cancer as a major focus of practice. The benefits of surgery are being extended to more patients with more severe comorbid conditions because of improvements in preoperative care, surgical techniques, and postoperative care. Patients with severe COPD or other comorbid illnesses should not be denied surgery without at least seeing an experienced thoracic surgeon. Furthermore, decisions about any patient’s operability should be made in the context of optimum therapy for underlying lung disease.

In some patients a trial of bronchodilators and inhaled or oral corticosteroids might result in improvement in lung function sufficient to reduce surgical risk. The nature of these patients and the increased use of multimodality treatment across all stages of NSCLC make a compelling rationale for a multidisciplinary approach to treating patients with lung cancer.

Stage II Lung Cancer

Stage II NSCLC is defined as including mainly (1) tumors with hilar or intralobar nodal metastasis, with tumor confined to one lobe of the lung, or (2) tumors invading the chest wall, without nodal metastasis. Treatment of stage II NSCLC is still primarily surgical resection, and the same principles apply here as in stage I disease. One major difference is that adjuvant chemotherapy has proven survival benefit for patients with stage II disease. If one makes the safe assumption that the risk of systemic chemotherapy is the same whether the patient had stage I or stage II disease, then the favorable risk/benefit data for adjuvant chemotherapy in patients with stage II reflects that occult metastases are more likely to be present in patients with more advanced disease. Therefore, patients with pathologic stage II lung cancer have more to gain with adjuvant chemotherapy than stage I patients. After recovery from surgery, patients who undergo surgical resection of tumors with peribronchial or hilar nodal metastasis (stage II NSCLC) should be referred to a medical oncologist to discuss adjuvant chemotherapy.

Current recommendations suggest all patients with stage II or III disease should receive postoperative chemotherapy. However, metaanalyses of adjuvant trials suggest the number of patients who need to be treated with adjuvant chemotherapy to achieve one additional long-term cure is 24 to 39. To reduce this number, studies looked for additional patient or tumor-specific characteristics that predict benefit (or lack of) from adjuvant treatment. One oft-cited study found tumor biomarkers that may identify patients most likely to benefit from adjuvant chemotherapy. Retrospective examination of tumor specimens from patients enrolled in an adjuvant trial identified excision repair cross complementation 1 (ERCC1), an enzyme involved in repairing cisplatin-induced DNA damage, as a possible prognostic and predictive marker. High expression of ERCC1 was associated with a good prognosis, but also identified a patient subgroup who did not benefit from adjuvant (cisplatin-based) chemotherapy. Similarly, high expression of ribonucleotide reductase M1 (RRM1, which metabolizes gemcitabine) was identified by immunohistochemistry as predicting better overall survival, but poor response to gemcitabine-containing chemotherapy. Insufficient data are available to recommend the use of biomarker-based selection of any chemotherapy, including adjuvant chemotherapy, but studies are underway to determine whether this improves outcomes relative to empirically chosen chemotherapy and allows fewer patients to be treated while achieving the same survival benefit.

Postoperative radiotherapy for completely resected stage II NSCLC does not have proven survival benefit, may be associated with worse outcomes, and in general should not be used. However, for patients with stage II NSCLC who cannot tolerate surgical resection, definitive radiotherapy is the standard of care. Although radiotherapy can provide excellent local control rates, it does not reduce the likelihood of distant metastasis. An unanswered question is whether adjuvant chemotherapy would benefit patients with clinical stage II NSCLC treated with curative radiotherapy. If adjuvant chemotherapy reduces the likelihood of death after surgery for stage II disease, the same approach might benefit patients treated with radiotherapy. No data support this approach, however, and extrapolating surgical data to patients treated with radiotherapy for local control is not appropriate outside of a clinical trial.

Stage III Lung Cancer

Stage III NSCLC makes up the largest proportion of any stage and generally includes patients with mediastinal nodal metastasis. Historically, the treatment for stage III lung cancer was thoracic radiotherapy alone, but long-term survival with this approach was poor (<10%), with most patients succumbing to distant metastases. Studies showed that combining systemic chemotherapy with thoracic radiation could improve long-term survival to close to 20%.

The current standard of care for adequately staged patients with stage III NSCLC depends greatly on the patient’s comorbid conditions and overall health status. The most commonly used global assessment of these factors in clinical practice is performance status (Box 67-1). Standard treatments for those with good to excellent Eastern Cooperative Oncology Group performance status (ECOG 0-1) are chemoradiation therapy administered concurrently. For those with less favorable performance status (ECOG 2), sequential chemotherapy followed by radiation therapy is still possible. The rationale for this is that concurrent therapy provides a survival benefit compared with sequential therapy, but at the cost of a greater likelihood of toxicity. Pulmonary toxicity is similar whether treatment is concurrent or sequential, but the occurrence of esophagitis (which can be severe) is more common with concurrent chemoradiation therapy than with sequential therapy. Those with marginal performance status (ECOG 2) may tolerate sequential therapy but are poor candidates for concurrent chemotherapy with thoracic radiation. In general, patients with ECOG performance status 3 or worse are best treated with a palliative approach focused on managing symptoms (i.e., best supportive care). This can include local radiotherapy to minimize complications such as airway obstruction, chest wall pain, or hemoptysis. Patients with poor performance status do not benefit from systemic chemotherapy and are more likely to experience severe toxicity.

The choice of chemotherapy for stage III NSCLC includes a platinum-based drug (cisplatin or carboplatinum) in combination with one other agent with demonstrated activity in NSCLC (Table 67-1). Gemcitabine, a drug with good single-agent activity often used in adjuvant settings or metastatic disease, is not typically used in combination with radiation because of its tendency to sensitize even normal tissue to the toxic effects of ionizing radiation. The dose of radiation for patients treated concurrently is generally left up to the radiotherapist, but doses used in this setting are lower than those used for primary treatment of early-stage, medically unresectable patients. Most trials delivered radiation doses of about 60 Gy. With better techniques that can limit damage to normal tissues, radiation oncologists are seeking ways to escalate the dose of radiation to increase the rate of cure. PET scans have been used in a course of radiation to adjust the port to a shrinking tumor, permitting a greater dose to metabolically active tumor (and consequently less to healthy tissue), as defined by the PET scan. These ongoing studies are likely to impact on the future treatment of NSCLC.

Table 67-1 Common Chemotherapy Regimens for Non–Small Cell (NSCLC) and Small Cell Lung Cancer

* Patients with activating mutations in epidermal growth factor receptor (EGFR) have better survival when treated with first-line EGFR–tyrosine kinase inhibitors.

† Gefitinib is not approved in the United States.

Given that chemotherapy (typically two cycles) combined with radiation in stage III disease increases local control and reduces distant metastases, one might conclude that more chemotherapy must be better. However, available studies do not support the role of additional “consolidation” chemotherapy after concurrent chemoradiation. One trial randomized patients to either observation or three cycles of docetaxel after patients had completed standard cisplatin and etoposide (two cycles) with radiation. Subjects receiving docetaxel had more treatment-related toxicity, without improved survival, than the control group. Although trials addressing this question continue, for now there is no role for maintenance or consolidation chemotherapy, even for patients with an apparent complete response.

Another area that remains unsettled is whether there are patients with stage IIIA NSCLC who benefit from some combination of surgery with chemotherapy and radiation therapy. This question has engendered significant debate among experts, but RCTs comparing “definitive” chemoradiation therapy with induction chemoradiation followed by surgery show no differences in survival between the two groups. Does this mean that no patients can benefit from “trimodality” therapy, or that the proper subgroup best suited for this most aggressive approach has yet to be identified? All that can be said with certainty is that the use of neoadjuvant therapy followed by surgical resection cannot yet be considered the standard of care for unselected patients with stage IIIA NSCLC.

One readily identifiable but rare subgroup of patients with stage IIIA NSCLC that most experts would agree should receive surgery is those with T3 tumors (either by chest wall invasion or, in the new staging system, with tumors greater than 7 cm) and hilar nodal metastasis. This group of T3N1, stage IIIA patients are surgical candidates with or without neoadjuvant chemotherapy, as long as they have the physiologic reserve to tolerate complete resection. A consistent finding in trials of neoadjuvant chemotherapy followed by surgery is that patients demonstrating a “complete response” in the mediastinal nodes when restaged before surgery have the best prognosis. Whether this simply identifies a population of patients with a good prognosis (and therefore no need for surgery), or whether surgery itself contributes to favorable long-term survival in this group of patients, cannot currently be answered.

Stage IV Lung Cancer

Until the early 1990s, the utility of chemotherapy for metastatic lung cancer was a subject of debate, with a widely held belief being that systemic chemotherapy had no role for patients with stage IV NSCLC. That debate has ended. Studies have demonstrated convincing (but modest) survival benefit with chemotherapy. Metaanalyses show that two-drug chemotherapy regimens are better than single-agent treatment, but the addition of a third agent is associated with greater toxicity, without additional gain in survival. Response to modern combinations of chemotherapy, with more than 50% tumor shrinkage, occurs in 20% to 30% of patients with stage IV NSCLC. Complete clinical remission is rare (<5% of patients), and those who do respond to chemotherapy eventually relapse and die of their disease. Patients with excellent performance status clearly benefit from systemic chemotherapy, but many patients present with ECOG 2 performance status, representing a controversial population regarding the efficacy and risk of systemic chemotherapy for stage IV NSCLC. Subgroup analyses of clinical studies suggest that patients with performance status of 2 can benefit from active treatment, including systemic chemotherapy. Many oncologists treat such patients with a less intense dose schedule.

Despite advances in treatment of advanced NSCLC, therapeutic nihilism has remained prevalent among clinicians, generally more than in other types of solid-organ tumors. A metaanalysis of trials with patients randomly assigned to “best supportive care” versus chemotherapy, including a cisplatin-based regimen, showed a survival benefit in the chemotherapy group. The active chemotherapy group had a reduction in the risk of death of 27% and an absolute improvement in survival of 10% at 1 year. In the 1990s, many promising new chemotherapeutic agents (paclitaxel, docetaxel, irinotecan, vinorelbine, gemcitabine, pemetrexed), each with single-agent activity in advanced disease, were developed for and then used in patients who had stage IV NSCLC. Numerous trials evaluated combinations of one or more of these newer agents with a platinum compound (cisplatin or carboplatinum; see Table 67-1). No one combination of drugs is superior. However, with use of platinum-based combinations, median survival gradually increased to 8 to 9 months, with 1-year survival of 30% to 35%. This compares favorably with untreated patients with stage IV NSCLC, who have median survival of 3 to 4 months, with 1-year survival of approximately 15%. Although no data exist to define the optimum combination of chemotherapeutic agents, newer data are leading to increasingly individualized approaches to care.

Pemetrexed was compared with gemcitabine (both in combination with a platinum compound) in patients with metastatic cancer. The overall population showed no difference between the two drug combinations, but a prespecified subgroup of patients with non–squamous cell histology showed improved outcomes when treated with the platinum-pemetrexed combination. This has resulted in the first histology-based recommendation for systemic chemotherapy in NSCLC.

Erlotinib and gefitinib, oral inhibitors of the epidermal growth factor receptor (EGFR) tyrosine kinase, were initially studied and used as second-line or third-line agents for patients with relapsed or metastatic NSCLC. Responses were dramatic in some cases, and larger trials noted common epidemiologic traits in those with excellent responses to these drugs, including nonsmoking (or light smoking) history, female gender, adenocarcinoma histology, and Asian ethnicity. Further studies that sequenced the EGFR gene in tumors of responders and nonresponders found that the drugs were unusually effective in patients’ tumors and cell lines bearing a few specific activating mutations affecting the tyrosine kinase domain of the EGFR gene. Further epidemiologic studies confirmed that activating mutations conferring sensitivity to EGFR-TKI drugs were also common in those who shared one or more of the demographic or histologic traits previously noted to confer a high likelihood of response: female, Asian, nonsmoking, or adenocarcinoma. A study in a mostly Asian population, for patients with tumors bearing activating mutations in EGFR, showed that first-line treatment with EGFR-TKI resulted in improved response and overall survival compared with standard chemotherapy among patients with mutant EGFR. In contrast, patients with tumors harboring wild-type EGFR fared better when treated with traditional cytotoxic chemotherapy.

Based on preclinical data demonstrating the importance of angiogenesis in supporting solid-tumor growth, targeted drug development for cancer has included anti-angiogenic agents, particularly drugs targeting vascular endothelial growth factor (VEGF) or its receptors. Bevacizumab, a fully human monoclonal antibody against VEGF, is the most advanced antiangiogenic agent in clinical use. In metastatic NSCLC, RCTs have demonstrated a modest but statistically significant survival benefit when bevacizumab was added to standard chemotherapy. In earlier Phase II trials, patients with large central tumors or with squamous cell histology had unacceptably high incidence of serious toxicity, specifically hemoptysis, and were specifically excluded from trials that resulted in the approval of bevacizumab for metastatic NSCLC. Bevacizumab use is also associated with a risk for hypertension and thrombotic complications and is contraindicated in patients with central nervous system (CNS) metastases.

An exciting addition in the field of targeted NSCLC therapy was the report in 2007 of a fusion protein, joining echinoderm microtubule–like 4 (EML4) with the anaplastic lymphoma kinase (ALK). EML4-ALK is an activating mutation capable of driving the proliferation of NSCLC cells. Drugs targeting abnormally activated ALK (e.g., crizotinib) were quickly developed and brought to clinical trials, showing great promise in patients bearing ALK mutations. Detecting ALK fusion can be difficult because it is typically formed by a short-segment chromosomal inversion. Fluorescent in situ hybridization (FISH) reagents developed for this purpose, called “break-apart FISH probes,” can highlight this chromosomal rearrangement. Similar to mutant EFGR tumors, the prevalence of ALK-mutant NSCLC tumors in most series is small (~5%) and most often found in nonsmokers. Although this finding was relevant to only a small number of cancer patients, the exciting nature of this discovery is that within 2 years after the discovery of ALK’s role in transformation of cancer cells, drugs targeting ALK were already in clinical trials, showing dramatic benefit for those with ALK-activating mutations. This pace of progression from discovery to treatment is highly unusual, not just in cancer, but in any field of biomedical research. Additional driving mutations similar to the one that activates ALK will likely be discovered in the future.

In summary, patients with metastatic, stage IV NSCLC and good performance status (ECOG 0-1) are best treated with systemic chemotherapy, with the addition of bevacizumab in those with no contraindication. Patients with ECOG performance status of 2 can also expect to benefit from active treatment, although with a higher risk of toxicity. Diagnoses in all patients should be made with biopsies that yield enough tissue to permit analysis for specific activating mutations in EGFR, or to detect the presence of an activating ALK fusion. Patients with these abnormalities, particularly EGFR mutations, are known (or suspected in the case of ALK) to gain prolonged survival when treated with agents targeted at their specific mutation.