Management of Locally Advanced Non–Small-Cell Lung Cancer | 7 |
Locally advanced non–small-cell lung cancer (LA-NSCLC) refers to disease that is unresectable or not curable with surgery alone. This currently includes stages IIIA (T4 N0 M0; T3-4 N1 M0; T1-3 N2 M0) and IIIB (T4 N2 M0; T1-4 N3 M0). In the upcoming revision of the TNM staging criteria (AJCC, eighth edition), patients with more extensive T3-T4 N3 M0 disease will be reclassified as stage IIIC, with a similar prognosis as patients with stage IVA disease (1). Selected patients with limited-extent stage IIIA NSCLC may be treated surgically, whereas patients with more advanced, unresectable disease or those who are medically inoperable are primarily treated with concurrent chemotherapy and radiation (2). Radiation therapy (RT) alone or sequential chemoradiotherapy is appropriate for patients who are not expected to tolerate concurrent chemoradiotherapy due to poor performance status, multiple or severe comorbid conditions, excessive weight loss, or large volume disease (2).
MEDICALLY INOPERABLE OR UNRESECTABLE LA-NSCLC
The majority of patients with LA-NSCLC have unresectable disease. This includes patients with T4 disease due to invasion of local structures, N3 disease, or multilevel or bulky (>3 cm) N2 lymph node involvement. Medically inoperable patients have potentially resectable disease, but have poor cardiopulmonary function (e.g., FEV1 < 2.0 L/sec or predicted postoperative functional lung volume <800 mL) or other comorbid conditions that may preclude surgery. In addition to resectable stage IIIA disease, this population may also include medically inoperable patients with stage II disease (primary tumor >5 cm and/or N1 disease).
For patients who are not fit for surgery or have unresectable disease, local-regional disease control and long-term survival are best achieved with concurrent chemotherapy and radiation. Concurrent chemotherapy augments the activity of radiation to improve local-regional control, while also decreasing the risk of distant metastases. The standard radiation dose is 60 Gy delivered in 2 Gy per fraction prescribed to a volume that encompasses the primary tumor and involved regional lymph nodes. RT is generally delivered concurrently with a platinum-based, two-drug chemotherapy regimen.
Patient Selection
Historically, RT alone was prescribed for patients with unresectable lung cancer. Clinical trials to improve local control and decrease distant failures have focused on increasing the intensity of both local and systemic therapies through the use of sequential chemotherapy, concurrent chemoradiotherapy, escalated dose RT, and altered RT fractionation. Trials evaluating aggressive multimodality therapy for unresectable lung cancer have usually included only patients with a favorable Karnofsky performance status (KPS ≥ 70) and limited weight loss, generally defined as greater than 10% in the month prior to diagnosis. A general guideline for baseline pulmonary function that would allow for definitive RT is FEV1 greater than 1.2 L/s or ≥ 50% predicted. However, treatment eligibility should be individualized based on disease volume and the volume of lung that will be included in the RT portals. Although age is not usually a specific eligibility criterion, the majority of patients enrolled on multimodality trials have been less than 70 years old. Older patients on these trials tend to do as well as younger patients, likely due to careful selection of only the fittest elderly patients with limited comorbidities. Although data are inconsistent, increased esophageal, pulmonary, and hematologic toxicity has been reported in older patients (3, 4). Additionally, older patients tend to have less physical and psychosocial reserve, do not tolerate the side effects of therapy as well, and frequently need more supportive care than younger patients. Careful patient selection, aggressive supportive care, and conservative RT dose parameters for normal tissues should be considered for elderly patients or patients with poor performance status.
Radiation Therapy
The standard RT dose is based on the outcomes of a phase III trial published in 1980 that randomized patients to receive either 40 Gy split–course RT or 40 Gy, 50 Gy, or 60 Gy continuous RT (5). The median survival with RT alone was 12 months with improved survival in the higher dose arm. RT-based strategies to improve local-regional control and survival rates have included dose escalation, altered fractionation, and the addition of concurrent or sequential chemotherapy.
Hyperfractionated RT
Hyperfractionated RT is the delivery of radiation at a lower than standard dose per fraction with more than one fraction per day. The lower dose per fraction decreases the risk of late radiation complications and allows for a higher total dose of RT to the target without increasing toxicity. Radiation Therapy Oncology Group (RTOG) 8311, a phase I/II trial, randomized patients to 60 Gy, 64.8 Gy, 69.6 Gy, 74.4, or 79.2 Gy delivered at 1.2 Gy per fraction given twice daily. Overall, there was no difference in acute or late toxicity among the five arms. In a subset of patients with KPS ≥70 and weight loss of less than 6%, survival was significantly better with 69.6 Gy than with the lower doses (P = .02). However, there was no survival advantage for doses higher than 69.6 Gy (6). Based on these results, 69.6 Gy at 1.2 Gy per fraction twice daily was used in RTOG 8808, an Intergroup phase III study comparing hyperfractionated RT alone, standard fractionation RT alone and standard RT with neoadjuvant chemotherapy (discussed in the following section).
Sequential Chemoradiotherapy
A randomized phase III trial conducted by Cancer and Leukemia Group B (CALGB 8433) compared induction chemotherapy followed by standard RT to RT alone (7). Patients were randomized to receive cisplatin plus vinblastine followed by RT (60 Gy at 2.0 Gy per fraction) or the same RT alone. Median survival for the sequential chemotherapy plus RT group was 13.7 months compared to 9.6 months for the RT alone group (P = .012). RTOG 8808 confirmed these results while also evaluating hyperfractionated RT (8). Patients were randomized to cisplatin plus vinblastine followed by standard once-daily RT (60 Gy at 2.0 Gy per fraction), standard once-daily RT alone, or hyperfractionated, twice-daily RT alone (69.6 Gy at 1.2 Gy per fraction). Median survival with standard RT alone was 11.4 months; with sequential chemotherapy plus standard RT, 13.2 months; and with hyperfractionated RT, 12 months.
Concurrent Chemoradiotherapy
The follow-up study to RTOG 8808 was RTOG 9410, a phase III trial comparing concurrent chemotherapy plus either standard or hyperfractionated RT to sequential chemotherapy plus standard RT (9). Patients were randomized to either sequential cisplatin plus vinblastine with standard RT beginning after chemotherapy, cisplatin plus vinblastine with standard RT beginning on day 1 of chemotherapy, or cisplatin plus oral etoposide with hyperfractionated RT beginning on day 1 of chemotherapy. Median survival favored concurrent chemotherapy with once-daily RT to 60 Gy (17 months vs. 14.6 months with sequential therapy [P = .046] vs. 15.6 months with concurrent hyperfractionated RT [P = .46]).
High-Dose RT With Concurrent Chemotherapy
Three single-arm studies that evaluated dose escalation of standard fractionation RT with concurrent chemotherapy reported that 74 Gy was safe and resulted in a median overall survival of 24 to 37 months (10–12). These findings compared favorably to the historical median overall survival of 16.5 to 17 months reported with concurrent chemotherapy and standard dose (60 Gy) RT (9,13).
RTOG 0617 was designed to compare concurrent chemotherapy to standard RT 60 Gy to concurrent chemotherapy plus RT 74 Gy in addition to evaluating the potential benefit of concurrent cetuximab, an anti-epidermal growth factor receptor (EGFR) monoclonal antibody. All patients were to receive consolidative chemotherapy. This two-by-two factorial, phase III trial randomly assigned patients to receive concurrent chemotherapy plus either RT 60 Gy in 30 daily fractions, RT 74 Gy in 37 daily fractions, RT 60 Gy plus cetuximab, or RT 74 Gy plus cetuximab. The study did not show improved survival with 74 Gy and, in fact, suggested that high-dose RT might be detrimental (14). Median overall survival for the standard RT dose group was 28.7 months and for the high-dose group was 20.3 months (P = .008). On multivariate analysis, factors predicting overall survival were RT dose (60 Gy), maximum esophagitis grade, planning target volume (PTV), heart V5 (volume of heart receiving 5 Gy or more), and heart V30. The addition of cetuximab to concurrent chemoradiotherapy provided no overall survival benefit.
The median overall survival of 28 months reported in RTOG 0617 with concurrent chemoradiotherapy set a new benchmark (14) when compared to 17 months achieved in RTOG 9410 with the use of a similar regimen. This improvement is likely due to several factors, including stage migration with the incorporation of PET into the staging evaluation, advances in radiation technology, more careful patient selection, and an increased focus on supportive care. Advances in RT delivery include motion management, multimodality imaging for tumor definition, consistent use of three-dimensional conformal radiation therapy (3D-CRT) planning, intensity modulated radiotherapy (IMRT), and image guided radiation therapy (IGRT).
Impaired Patients
For patients who may not tolerate concurrent therapy or who have large volume disease that cannot be treated without exceeding normal tissue RT tolerance, sequential induction chemotherapy followed by definitive RT is a reasonable option with an expected median survival of 13 to 15 months (8,9,13). Induction chemotherapy may decrease disease volume, allowing for more focused delivery of RT to gross tumor with avoidance of normal tissue and less acute radiation toxicity.
For patients who cannot tolerate or refuse chemotherapy, RT alone can be used with an expected median survival of 11 to 18 months (8,15). Hypofractionated therapy regimens deliver a higher dose per fraction of RT to a lower total dose, allowing for a shorter course of therapy with an RT dose that is biologically equivalent to standard treatment. The shorter treatment course decreases the burden of therapy from 6 weeks of RT to 3 weeks, and 45 Gy in 15 3 Gy fractions has been shown to be effective in phase II prospective trials and institutional reviews (16–18).
Radiation Therapy Techniques
RT Dose
The generally accepted standard dose of RT for patients with LA-NSCLC is 60 Gy in 30 fractions delivered at 2 Gy per fraction, once a day, 5 days per week with concurrent chemotherapy. With such treatment, expected median survival is 28.7 months and the 2-year overall survival rate is 57.6% for patients with PET-staged stage III disease, performance status 0 to 1, weight loss less than 10% in 1 month, FEV1 ≥ 1.2 L/s, and no contralateral hilar or supraclavicular lymph nodes (10). For patients with poor performance status, extensive weight loss, or other co-morbidities precluding aggressive multimodality therapy, hypofractionated RT to 45 Gy at 3 Gy per fraction should be considered.
RT Simulation
CT simulations should be performed with the patient in a reproducible treatment position with appropriate immobilization. IV contrast is recommended and is particularly useful for target definition of centrally located disease. The use of PET/CT for treatment planning has been shown to improve target accuracy and local control (19). Assessment of three dimensional (3D) tumor and organ motion during simulation is essential for optimal treatment planning. Acceptable methods for motion assessment include fluoroscopy, CT at inhalation and exhalation, slow CT scan, and 4D-CT. It is important to note that respiratory motion changes over time, so reassessment during treatment is required. Motion control methods, including abdominal compression, breath hold, and gating, can be used to decrease the treatment volume needed to cover disease throughout the respiratory cycle. The American Association of Physics in Medicine (AAPM) Task Force 76 report on the management of respiratory motion in radiation oncology is recommended for guidelines on treatment planning and delivery (20).
RT Planning
Gross tumor volume (GTV) consists of all disease that is visible on imaging studies, including the primary tumor and all lymph nodes that are PET positive (SUV >3.0), biopsy-proven, or greater than 1 cm on short axis. The GTV is then expanded to include suspected subclinical disease to create the clinical target volume (CTV). Although the CTV expansion is typically 0.5 cm in lung parenchyma, it should also include involved nodal regions and be confined by normal anatomic boundaries with exclusion of tissues that are clearly not involved with disease, such as vertebral bodies, heart, and great vessels. Uninvolved nodal regions should not be included. The internal target volume (ITV) is technically defined by further expansion of the CTV to account for internal motion due to breathing and heart movement. This expansion ensures that all disease is included in the treatment field throughout the breathing cycle. The PTV is a final expansion, generally 3 to 5 mm that accounts for variations in daily treatment set up. The radiation dose is prescribed to the PTV with a minimum of 95% coverage. Any variations in coverage should be confined to regions of the PTV adjacent to critical structures. The maximum dose should be within the PTV and less than 110% of the prescribed dose.
Maximum acceptable doses to organs at risk (OAR) are listed in Table 7.1. Dose constraints should be tailored to the individual patient. Consideration should be given to factors that may influence the ability to tolerate therapy, such as comorbidities, prior weight loss, and age. Efforts should be made to confine the RT dose to the involved lobes of the ipsilateral lung in patients being treated preoperatively.
Three-Dimensional Conformal Radiation Therapy
3D-CRT uses individualized 3D digital data sets of a patient’s tumor and normal adjacent anatomy to develop a radiation plan that typically uses greater than three beams or fields. Each field is individually shaped and directed to create a conformal dose distribution to the PTV, while sparing surrounding normal structures.
Intensity Modulated Radiation Therapy
Intensity modulated radiation therapy (IMRT) uses a similar data set to create the RT plan. The goal is to conform the dose to the disease and avoid normal tissues, similar to 3D-CRT, although the radiation intensity of each IMRT beam is divided into small segments and modulated throughout the treatment by the multileaf collimator (MLC) attached to the linear accelerator. With multiple modulated radiation segments, IMRT plans are much more conformal and typically result in better sparing of normal tissues.
IMRT usually provides dosimetric and clinical advantages for patients with LA-NSCLC being treated with curative intent RT. IMRT is particularly useful for tumor coverage of disease in close proximity to critical structures and in patients with large treatment volumes due to bulky disease or multilevel or contralateral lymph node involvement. IMRT plans typically will improve the ability to spare normal tissue, thereby minimizing toxicity and improving quality-of-life. In one study comparing IMRT to 3D-CRT, fewer patients who received IMRT had a clinically meaningful decline in quality-of-life 12 months after treatment (21% vs. 46%; P = .003) (21). Liao et al. showed an improvement in overall survival in 2 cohorts of patients treated with IMRT versus 3D-CRT; however, the patients treated with IMRT also had four dimensional CT (4D-CT) planning and quite possibly had better tumor coverage (22). Many insurance companies will not approve IMRT for lung cancer if published normal tissue dose constraints can be met with a 3D-CRT plan. Careful target volume definition utilizing CT and PET data with compensation for motion and set up variation and minimization of normal tissue doses with conformal RT fields with or without IMRT is critical for optimizing local-regional tumor control, overall survival, and quality-of-life.
Chemotherapy
Concurrent chemotherapy and RT is preferred over sequential and induction approaches. A meta-analysis of six trials including 1,024 patients demonstrated a 10% absolute overall survival benefit at 2 years with concurrent chemoradiotherapy when compared to sequential chemoradiotherapy (23). Induction chemotherapy, defined as two cycles of full-dose chemotherapy before concurrent chemoradiotherapy, is also not a recommended strategy. In CALGB 39801, 366 patients were assigned to either two doses of carboplatin plus paclitaxel (CP) followed by concurrent weekly CP plus RT or concurrent weekly CP plus RT without induction therapy. There was no significant improvement in survival with induction therapy (median, 14 vs. 12 months; P = .3), but toxicity was greater in the induction arm (24).
Selection of Chemotherapy Regimen
When choosing a chemotherapy regimen to combine with RT, agents that enhance the effects of radiation and provide effective systemic therapy without excessive toxicity are preferred. Toxicities that are unique to concurrent thoracic chemoradiotherapy include esophagitis and pneumonitis, so drugs that specifically increase pulmonary toxicity, such as gemcitabine, should be avoided. Common chemotherapy regimens that are utilized concurrently with RT are listed in Table 7.2. Although few clinical trials have directly compared chemotherapy regimens in patients with stage III NSCLC, the efficacy of each of these regimens appears to be similar. Therefore, the selection of an appropriate chemotherapy regimen should be guided by an individual patient’s comorbidities, organ function, risk for specific toxicities, and clinician comfort level with the regimen.
The two regimens that are used most frequently with concurrent RT are cisplatin plus etoposide (EP) and weekly CP (2). While there are no high-quality randomized trials comparing EP to CP, retrospective and pooled analyses suggest equivalent survival outcomes, but increased toxicity with EP. A retrospective analysis of a Veterans’ Administration (VA) database reported on 1,842 patients with stage III NSCLC who underwent concurrent therapy, 27% of whom received EP and 73% CP. Survival was comparable between the two treatments, but EP was associated with increased incidence of hospitalization, infection, acute kidney injury, mucositis, and esophagitis (25). A more comprehensive pooled analysis compared data from 32 studies (3,194 patients) utilizing EP and 51 studies (3,789 patients) utilizing CP along with RT for patients with stage III NSCLC. Response rates (65% vs. 56%, P = .6), progression-free survival (11.5 vs. 9.3 months, P = .2), and overall survival (19.8 vs. 18.4 months, P = .48) were similar in EP and CP treated patients, respectively. Selected toxicities, including cytopenias, nausea, and vomiting, were more common with EP, but both regimens resulted in equivalent rates of esophagitis and pneumonitis (26). Median overall survival in this pooled analysis was less than 24 months for both regimens.
Other Chemotherapy Regimens
Cispaltin plus pemetrexed (PC) is a well-tolerated and effective first-line regimen for treatment of metastatic non-squamous NSCLC (27), and pemetrexed has been shown to be an effective radiosensitizer (28). In the phase III PROCLAIM study, RT plus concurrent PC followed by consolidation therapy with single-agent pemetrexed was compared to RT plus concurrent EP followed by consolidation with a nonpemetrexed containing regimen in patients with stage III non-squamous NSCLC (29). An interim analysis showed that the overall survival in the PC arm was not superior to that in the EP arm (median, 26.8 vs. 25 months, P = .83), so the study was discontinued for futility. PC had a more favorable toxicity profile with a lower incidence of grade 3 to 4 adverse events (64% vs. 76.8%, P = .001). The only exception was a higher rate of pneumonitis in the CP arm (17% vs. 11%, P = .37), but the incidence of severe pneumonitis remained low in both arms (<3%). Phase II studies evaluating carboplatin plus pemetrexed (PCb) given concurrently with RT have also shown an acceptable toxicity profile with a 2-year overall survival rate of 48% (30), offering another option for patients with non-squamous NSCLC who cannot tolerate cisplatin.
Consolidation Chemotherapy
Consolidation chemotherapy with docetaxel after EP is not recommended. A Hoosier Oncology Group study (LUN 01-24) reported that concurrent EP plus RT followed by consolidation docetaxel did not improve survival over concurrent EP plus RT alone. However, consolidation docetaxel was associated with high rates of pneumonitis (9.7%), hospitalization (29%), and treatment-related mortality (5.5%) (31).
In another randomized trial, consolidation cisplatin plus docetaxel after concurrent RT and weekly cisplatin plus docetaxel was compared to the same concurrent chemoradiotherapy regimen without consolidation. Both progression-free survival (median, 9.1 vs. 8.1 months, P = .36) and overall survival (21.8 vs. 20.6 months, P = .44) were similar in both arms of the study (32). However, cisplatin plus docetaxel is not a commonly used regimen and it is unclear if these results can be extrapolated to the more popular weekly CP regimen.
A pooled analysis of 41 studies including 3,400 patients suggested no benefit from consolidation chemotherapy, but was limited by the heterogeneity of study regimens and patient populations (33). Although supporting data are scarce, for patients receiving low-dose, “radiosensitizing” chemotherapy, such as weekly CP, consolidation chemotherapy may offer some benefit by providing higher-dose systemic therapy (34).
Targeted Therapy
Cetuximab, a monoclonal antibody targeting EGFR, was assessed in combination with concurrent CP and RT in RTOG 0617, and did not improve overall survival when compared to chemoradiotherapy alone (25 vs. 24 months, P = .29). However, cetuximab did increase grade 3 or higher toxicity (86% vs. 70%, P < .0001) (14). Gefitinib, an EGFR tyrosine kinase inhibitor, was tested against placebo as a maintenance strategy in patients with stage III NSCLC who completed concurrent EP plus RT followed by consolidation docetaxel. Importantly, patients were not selected for EGFR mutational status. Gefitinib had a negative impact on survival when compared to placebo (23 vs. 35 months, P = .013), mainly due to recurrent lung cancer (35).
In studies in both NSCLC and SCLC, the addition of bevacizumab, an antiangiogenic monoclonal antibody targeting vascular endothelial growth factor (VEGF), to concurrent chemoradiotherapy greatly increased the risk of tracheaesophageal fistula formation, and such combinations should be avoided (36).
RESECTABLE LA-NSCLC
Primary surgical resection is considered the standard treatment option for patients with stage IIIA—T3N1 NSCLC. The role of surgical resection for patients with more advanced LA-NSCLC remains controversial, but continues to be discussed due to the local failure rate of 30% to 40% for patients treated with standard concurrent chemoradiotherapy (15). When surgery is employed, it is generally preceded by neoadjuvant chemoradiotherapy or chemotherapy alone with or without postoperative RT. Neoadjuvant therapy followed by resection is most often considered in patients with minimal N2 disease involving nonbulky (<3 cm) lymph nodes in 1 to 2 nodal stations.
Patient Selection
Determination of the role of surgery in a patient with N2 lymph node involvement should be made prior to the initiation of any therapy by a multidisciplinary team including a board-certified thoracic surgeon with expertise in thoracic oncology (37). Appropriate patients should have technically resectable disease with low-volume lymph node involvement before the initiation of any therapy. Attempts to convert an unresectable cancer into a resectable cancer with neoadjuvant therapy are rarely successful and should be discouraged.
N2 Disease Burden
Predictors of a favorable outcome with surgery for patients with N2 lymph node involvement include low-volume N2 disease (<3 cm), single-station N2 disease, and nodal down-staging with preoperative therapy. A nonsurgical, combined-modality approach is recommended for patients with a high N2 disease burden.
Occult vs. clinical: Occult or “minimal” N2 disease refers to pathological involvement of the mediastinal lymph nodes found at resection in patients with negative radiographic and pathologic preoperative staging. Clinical N2 disease refers to patients with mediastinal lymph nodes that are enlarged by CT criteria (>1 cm in short axis), metabolically active by PET, and/or biopsy-proven during preoperative staging. The 5-year survival rate after primary surgical resection is 25% to 40% for patients with minimal N2 disease, but only 3% to 10% for those with clinical N2 disease (38–40). Therefore, primary surgical resection is not optimal therapy for patients with clinical N2 lymph node involvement. Patients with a high probability of N2 disease due to a central primary tumor or hilar lymph node involvement (N1) should undergo adequate mediastinal staging prior to resection. If N2 disease is detected, then appropriate treatment would be either definitive chemoradiotherapy or neoadjuvant therapy depending on the N2 disease burden.
Multistation N2 disease: Multistation nodal disease prior to therapy is a poor prognostic factor with a 5-year survival rate of 11% to 17% compared to 34% to 39% for single-station involvement (39–41). The number of lymph node stations involved is a key prognostic factor for both occult and clinical N2 disease. In a study of 702 patients with resected N2-positive NSCLC, the 5-year overall survival rate ranged from 3% in multistation disease to 8% with single-station involvement in patients with clinical N2 involvement compared to 11% and 34%, respectively, in the occult N2 group (40). Isolated N2 disease without N1 involvement, called skip N2, refers to a unique subset of patients with a favorable prognosis similar to those with only N1 involvement (42). In addition to the number of nodal stations involved, the absolute number of lymph nodes with metastatic foci also correlates with prognosis (43).
Location of nodes: Depending on the primary tumor location, a distinct pattern of regional spread to mediastinal nodes has been described (44). Nonregional spread is associated with a poorer prognosis (45). For example, left upper lobe tumors commonly spread to the aortopulmonary lymph nodes and patients with left upper lobe tumors and isolated spread to these nodes have a better prognosis with a 5-year survival rate of about 30% after resection compared to patients with subcarinal lymph node involvement who did not survive beyond 3 years from resection (46).
Mediastinal Staging
Mediastinal staging at diagnosis is important, not only for confirming suspicious nodal disease found on CT and PET, but also for evaluating for occult N3 and multistation N2 disease. Many surgeons have adopted the strategy of invasive mediastinal restaging after neoadjuvant therapy followed by resection only for patients who have been down-staged. Postinduction therapy mediastinal staging by CT, PET, endoscopic ultrasound (EUS), endobronchial ultrasound (EBUS), or repeat mediastinoscopy has a false-negative rate of 20% to 30% (47–50). Repeat mediastinoscopy is a technically difficult procedure, so initial, pretreatment mediastinal staging should be performed with EBUS and/or EUS, reserving mediastinoscopy for restaging after neoadjuvant therapy (51).
Pneumonectomy
Many studies have shown a high early mortality rate after pneumonectomy done following neoadjuvant therapy due to acute respiratory distress syndrome (ARDS) and other factors (52). This risk is higher in patients requiring a right pneumonectomy (53). The risk of surgery after concurrent chemoradiotherapy can be minimized with careful attention to radiotherapeutic and surgical technique, and optimal supportive care. Although pneumonectomy poses a higher risk, the potential need for a pneumonectomy is not an absolute contraindication to preoperative chemoradiotherapy for appropriately selected patients in experienced hands (54,55).
Preoperative Chemotherapy
Two small, landmark phase III trials from the 1990s that compared surgery alone to chemotherapy followed by surgery in patients with stage III NSCLC demonstrated a significant overall survival advantage with induction chemotherapy (median, 26–64 vs. 8–11 months) (56,57). Both studies closed early at interim analysis due to improved survival with chemotherapy. However, as noted earlier, surgery alone is inadequate therapy for most patients with stage III disease, and randomized trials comparing neoadjuvant chemotherapy to chemotherapy plus RT have not been performed. A recent meta-analysis showed a 13% reduction in the relative risk of death with preoperative chemotherapy in patients with stage I to III NSCLC (58).
While early neoadjuvant trials used obsolete, three-drug chemotherapy regimens, more recent studies have typically used standard, platinum-based, two-drug regimens for three cycles. With such regimens, response rates have ranged from 44% to 70% and pathologic compete response rates from 12% to 16%, with only 10% to 14% of patients having disease progression (59–61). A meta-analysis has shown that neoadjuvant chemotherapy provides an overall survival benefit that is comparable to that achieved by adjuvant chemotherapy (62). If induction chemotherapy is to be used, current clinical guidelines recommend a cisplatin-based, two-drug regimen similar to those used in the adjuvant setting with carboplatin substitution for patients who cannot tolerate cisplatin (2).
Preoperative Chemoradiotherapy
Pathologic complete response (pCR) and mediastinal down-staging after preoperative therapy are associated with improved survival (41,59). The addition of RT to chemotherapy in phase II trials increases the rate of mediastinal down-staging when compared to historical results with chemotherapy alone (63). Although phase III trials have shown an advantage for preoperative chemotherapy over surgery alone, concurrent chemoradiotherapy has not been adequately compared to chemotherapy alone in the preoperative setting. Recently, Pless et al. randomized 232 patients with resectable stage IIIA NSCLC to induction chemotherapy with cisplatin plus docetaxel followed by RT versus chemotherapy alone followed by resection and demonstrated no significant difference in event-free survival (12.8 vs. 11.6 months, P = .67) or overall survival (61). However, in this study, chemotherapy and RT were given sequentially rather than concurrently and the RT dose was 44 Gy delivered over 3 weeks.
The North American Intergroup trial 0139 is the largest phase III study that has prospectively addressed whether or not survival is improved with surgery after induction chemoradiotherapy for patients with stage IIIA (T1-3 pN2 M0) resectable NSCLC (64). All patients received concurrent chemoradiotherapy with EP and RT to 45 Gy. In the absence of progression, patients were randomized to undergo surgery or continuation of RT to 61 Gy. Both groups received consolidation chemotherapy with two cycles of EP. Mediastinal nodal clearance was noted in 46% of patients and pCR in 18%. Treatment-related mortality was higher in the surgery arm (7% vs. 1.6%). Most postoperative deaths were due to ARDS, predominantly after pneumonectomy. Disease-free survival was significantly prolonged in the surgery arm (14.0 vs. 11.7 months, P < .02), but overall survival was similar in both arms (22.2 vs. 23.6 months, P = .24). In an exploratory analysis, overall survival was improved with surgery in patients undergoing a lobectomy, but not in those undergoing pneumonectomy.
It is unlikely that another prospective trial will be completed that directly compares definitive chemoradiotherapy to trimodality therapy. Appropriate patients should be counselled about the potential risks and benefits of definitive chemoradiotherapy with and without a surgical resection (preferably by lobectomy). Additionally, INT 0139 emphasized the need for meticulous RT planning, surgical technique, and postoperative care in patients undergoing aggressive trimodality therapy.
The appropriate dose of preoperative RT is currently being reevaluated. Excellent outcomes have been reported in a single-institution trial that utilized neoadjuvant concurrent chemotherapy plus standard-dose RT (>59 Gy) followed by resection incorporating vascularized muscle flaps to cover the bronchial stump, limitation of intraoperative fluids, and avoidance of postoperative barotrauma (65). RTOG 0229 was a multicenter phase II trial that evaluated full-dose RT plus concurrent chemotherapy followed by surgical resection (63). Clearance of mediastinal lymph nodes was noted in an impressive 63% of patients after full-dose chemoradiotherapy, meeting the study endpoint, and the incidence of grade 3 postoperative pulmonary complications was only 14%.
Although mediastinal clearance and pCR rates appear to be improved with standard-dose RT and concurrent chemotherapy, the appropriate dose of RT remains clear. If preoperative chemotherapy plus RT to 45 Gy is planned, it is imperative that surgical evaluation occur rapidly so as not to delay the continuation of RT to a definitive dose if surgery is not possible. An alternative strategy is to use preoperative chemotherapy alone with postoperative RT provided for residual N2 disease or positive surgical margins. If RT is to be delivered, full dose RT (60 Gy) is preferred if surgeons are experienced with resection after high-dose RT.
Adjuvant Therapy for Primarily Resected Occult LA-NSCLC
In surgical series, about 20% of patients who are clinically staged as node-negative are found to have ipsilateral mediastinal lymph node involvement (N2) on pathologic evaluation of the surgical sample (66). These patients have varying prognoses depending on their burden of N2 disease and T stage (40). Randomized trials have shown that adjuvant chemotherapy can improve long-term overall survival in this setting, while retrospective analyses suggest that postoperative RT (PORT) may also favorably impact long-term outcome.
Adjuvant Chemotherapy
Adjuvant chemotherapy is indicated for patients with completely resected stage II and III NSCLC. Three landmark randomized trials, JBR-10, International Adjuvant Lung Cancer Trial (IALT), and Adjuvant Navelbine International Trialist Association (ANITA), showed improvements in overall survival, ranging from 4% to 15%, with adjuvant chemotherapy, as summarized in Table 7.3 (67–69). The Lung Adjuvant Cisplatin Evaluation (LACE) meta-analysis included five large, randomized trials published after 1995 that randomized 4,584 patients who underwent complete resection of stage I to III NSCLC to cisplatin-based, two-drug chemotherapy versus no chemotherapy. LACE demonstrated a 5.3% absolute benefit in overall survival with adjuvant chemotherapy (HR 0.89, 95% CI 0.82–0.96). The significant benefit was confined to patients with stage II (HR 0.83, 95% CI 0.73–0.95) and stage III (HR 0.83, 95% CI 0.72–0.94) disease, with potential detriment in stage IA (HR 1.4, 95% CI 0.95–2.06) and no significant benefit in stage IB (HR 0.93, 95% CI 0.78–1.10) (70). This benefit was validated by a larger meta-analysis including 26 trials with 8,447 patients which showed an absolute overall survival benefit of 4% at 5 years, (HR 0.86, 95% CI 0.81–0.92) P <.0001 (71). Appropriate patients for adjuvant chemotherapy should have good performance status and an uncomplicated recovery from surgery.
Adjuvant chemotherapy regimens: In a planned subset analysis of the LACE database, patients receiving cisplatin plus vinorelbine (PVb) had an absolute 5-year survival benefit of 8.9% compared to observation, which appeared to be indirectly superior to the benefit derived from other cisplatin-based combinations (72). However, PVb is associated with significant toxicity and compliance to therapy is poor compared to newer regimens such as cisplatin plus pemetrexed (CP). This was demonstrated in the TREAT study, where 132 patients with resected non-squamous NSCLC were randomized to receive adjuvant CP or PVb with a primary endpoint of clinical feasibility, which was defined as no grade 4 cytopenia, no grade 3 to 4 febrile neutropenia or nonhematological toxicity, and absence of death or premature study withdrawal. The clinical feasibility rate with CP was superior to that with PVb (96% vs. 75%, P = .001) (73). Current guidelines recommend four cycles of a cisplatin-based, two-drug regimen for eligible patients and carboplatin-based, two-drug regimens for patients who cannot tolerate cisplatin (2). In the recently concluded trial E1505, which evaluated the role of adjuvant bevacizumab in 1,501 patients, oncologists were allowed to choose between four adjuvant chemotherapy regimens, 25% of patients received CVb, 19% cisplatin plus gemcitabine, 23% cisplatin plus docetaxel, and 33% CP (74). Bevacizumab was not found to be beneficial in the adjuvant setting. However, pooled analysis showed that within the squamous and non-squamous subgroups, disease-free and overall survival were similar with each of the chemotherapy regimens. In the non-squamous cohort, there was significantly less grade 3 to 5 toxicity in the CP arm, with increased neutropenia and thrombocytopenia in the CVb and cisplatin plus gemcitabine arms, respectively (74).
Timing of chemotherapy: Most adjuvant chemotherapy studies mandated that patients start treatment within 60 days of surgery, and no randomized trials have addressed the question of delayed adjuvant therapy. A retrospective analysis of the Ontario Cancer Registry showed that delayed initiation of adjuvant chemotherapy, defined as chemotherapy given between 10 and 16 weeks after resection, occurred in one-third of patients and was still associated with a positive effect on survival (75). If both adjuvant chemotherapy and PORT are planned, then they should be delivered in a sequential manner with chemotherapy given first in order to avoid unnecessary treatment delays and interruptions that might be caused by poor tolerance of concurrent therapy.
Elderly patients: Patients over 70 years old accounted for only 9% of those in the LACE database, while patients over 75 years old accounted for only 1.3%. An age-specific pooled analysis showed that elderly patients seemed to derive benefit from adjuvant chemotherapy despite receiving lower doses and fewer cycles of therapy (76). Data in octogenarians is quite limited, but there is some evidence that adjuvant chemotherapy may be detrimental (77). The use of adjuvant chemotherapy in older patients must be individualized with consideration of each patient’s potential risks and benefits.
Postoperative RT
The role of postoperative RT (PORT) in patients with completely resected NSCLC remains controversial, but the American Society for Radiation Oncology’s (ASTRO) evidence-based clinical practice guidelines state that PORT can be used to optimize local control in patients with positive surgical margins, gross residual disease, or mediastinal lymph node involvement (78).
Incomplete resection: In an evaluation of over 100,000 patients undergoing lung cancer resection, surgical margins were positive in 4.7% and this was associated with a worse 5-year overall survival (59% vs. 34%) (79). An incomplete resection adversely affected prognosis irrespective of the stage and the magnitude of this effect was comparable to shifting the patient to the next higher stage of disease. Factors associated with positive margins include higher stage, advanced tumor grade, squamous cell histology (probably due a central location), tumor involving multiple lobes, and lower socioeconomic status (79). Repeat resection is recommended when feasible, but in the absence of resection with negative margins, PORT is indicated to improve local-regional control.
pN2 disease: The PORT meta-analysis published in 1998 included nine studies, with 2,128 patients, in which surgery followed by PORT was compared to surgery alone. This meta-analysis concluded that PORT produced a significant detrimental effect on survival (80). However, this detrimental effect was confined to patients with pN0 and pN1 disease. Despite antiquated RT techniques, there was a nonsignificant trend toward improved survival with PORT in patients with stage IIIA-pN2 disease. A more recent retrospective analysis of the Surveillance Epidemiology and End Results (SEER) database included 5,600 patients who underwent complete resection of NSCLC and reported a lower overall survival for patients with pN0 or pN1 disease who received PORT. However, the pN2 subgroup of patients had a significant 3-year overall survival benefit (81).
In the ANITA adjuvant chemotherapy trial, the use of postoperative RT for patients with stage III-pN2 NSCLC was determined by individual institutions before initiating the study. If given, PORT (45–60 Gy in 2 Gy fractions) was administered within 2 weeks of completion of chemotherapy or surgery depending on the treatment arm. Overall survival was improved in patients with pN2 disease who received PORT, both in the chemotherapy arm (median, 47 vs. 24 months) and the observation arm (median, 23 vs. 13 months) (82). A retrospective review of the National Cancer Database evaluated the impact of modern PORT (≥45 GY) in 4,483 patients with stage III-N2 NSCLC treated between 2006 and 2010 with surgery and adjuvant chemotherapy. The use of PORT was associated with an increase in median and 5-year overall survival compared with no PORT (median, 45 vs. 41 months; 5-year, 39% vs. 35%; P = .014) (83).
CONCLUSION
LA-NSCLC is a heterogeneous disease consisting of subsets of patients ranging from those with resectable or minimal N2 disease with a relatively good prognosis to those with unresectable, bulky, extensive intrathoracic disease with a poor prognosis similar to that of stage IVA patients. Patients with LA-NSCLC are ideally managed by a multidisciplinary team of thoracic surgeons, radiation oncologists, and medical oncologists who can individualize therapy based on each patient’s clinical and pathological characteristics.
