INTRODUCTION

Worldwide, nearly 1 million people die annually from non–small cell lung cancer (NSCLC). Traditionally, complete surgical resection has been considered the only treatment with a curative potential in the 35% or so of these patients with early-stage disease (i.e., stages IA–IIIA). However, more than one half ultimately die from recurrent disease resulting from micrometastases not apparent at the time of surgery. Hence, there is a clear need for systemic treatment in the form of chemotherapy and perhaps newer biologically targeted agents to compliment the critical local role of surgery.

SURVIVAL FOLLOWING COMPLETE SURGICAL RESECTION IS RELATED TO STAGE AND RISK OF MICROMETASTASES

• ▪
  

  Five-year survival rates in completely resected NSCLC ranges from 70% in patients with stage I disease to about 10% in patients with stage IIIA disease.

  
  
• ▪
  

  In most cases, the failure of surgery to cure localized NSCLC results from the growth and progression of micrometastatic cancer not detected at the time of surgery.

POST-OPERATIVE RADIATION THERAPY MAY HAVE A DELETERIOUS EFFECT ON SURVIVAL

• ▪
  

  The Post-Operative Radiation Therapy (PORT) Meta-Analysis Trialists Group first published a meta-analysis based on individual data on 2128 patients from nine randomized clinical trials comparing postoperative observation with postoperative radiotherapy in NSCLC in 1998. In 2005, the PORT group updated their meta-analysis with the addition of individual patient data from a 10th clinical trial, bringing the patient total to 2232. The results indicated a deleterious effect of radiotherapy on survival in patients with stage I or II lung cancer (18% increased risk of death over 5 years and an absolute decrease in 5-year survival of 6%). No definite adverse effect of PORT could be identified in patients with stage III (N2) NSCLC.

CHEMOTHERAPY AFTER SURGERY (ADJUVANT) OR BEFORE SURGERY (NEOADJUVANT)

• ▪
  

  Risks versus benefits: If the primary goal of adding chemotherapy to surgery is to eradicate micrometastatic disease, biologically there should be no major difference in whether the chemotherapy is given after or before the surgery. However, because surgery provides the only curative potential for patients with clinically localized lung cancer, the risks of delaying surgery, increasing surgical morbidity or mortality, or, worse, making surgery impossible because of a severe chemotherapy-related adverse event, most chemotherapy studies over the past 15 years have been in the postoperative (adjuvant) setting. As postoperative chemotherapy becomes better established and more effective, it is rational for chemotherapy to be more commonly administered in the preoperative setting when patients are better able to tolerate it and when it might enhance the feasibility and limit the extent of surgery by tumor downstaging.

CHEMOTHERAPY AFTER SURGERY (ADJUVANT) OR BEFORE SURGERY (NEOADJUVANT)

• ▪
  

  Risks versus benefits: If the primary goal of adding chemotherapy to surgery is to eradicate micrometastatic disease, biologically there should be no major difference in whether the chemotherapy is given after or before the surgery. However, because surgery provides the only curative potential for patients with clinically localized lung cancer, the risks of delaying surgery, increasing surgical morbidity or mortality, or, worse, making surgery impossible because of a severe chemotherapy-related adverse event, most chemotherapy studies over the past 15 years have been in the postoperative (adjuvant) setting. As postoperative chemotherapy becomes better established and more effective, it is rational for chemotherapy to be more commonly administered in the preoperative setting when patients are better able to tolerate it and when it might enhance the feasibility and limit the extent of surgery by tumor downstaging.

POSTOPERATIVE (ADJUVANT) CHEMOTHERAPY

• ▪
  

  Goal: Improve survival following complete surgical resection of early stage NSCLC by administration of systemic therapy that eradicates microscopic residual or metastatic disease.

  
  
• ▪
  

  Early efforts with first-generation platinum-based chemotherapy regimens: Clinical trials before 1985 were generally small, poorly designed, and used chemotherapy regimens with minimal efficacy. After 1985, cisplatin-based chemotherapy regimens that appeared to improve survival in patients with stage IV NSCLC began to be used in the adjuvant setting.

  
  
  
  
  • ○
    

    Lung Cancer Study Group (LCSG) trial 772: 141 patients with completely resected stage II or III adenocarcinoma or large cell undifferentiated carcinoma were randomized to receive six cycles of cyclophosphamide, doxorubicin, and cisplatin (CAP) chemotherapy or putative immunotherapy (intrapleural BCG and levamisole). Patients in the adjuvant chemotherapy arm achieved significant delay in time to recurrence of lung cancer and improved cancer-free survival at 1 year (77% versus 62%) but no statistically significant long-term or overall survival advantage.

    
    
  • ○
    

    LCSG trial 791: Patients with positive surgical margins or metastases to high paratracheal lymph nodes (level 2) randomized to post-operative radiation therapy alone or radiation therapy followed by adjuvant CAP chemotherapy. One-year survival rates favored chemotherapy arm, but survival at 2 years and overall survival were not different between the two arms of the trial.

    
    
  • ○
    

    LCSG trial 801: Patients with completely resected T2N0 or T2N1 NSCLC were randomized postoperatively to observation or four cycles of CAP chemotherapy. There was no difference in survival, but nearly half of the patients in the adjuvant chemotherapy arm failed to receive the planned four cycles of treatment. Subset analysis suggested that patients receiving more than 80% of planned treatment may have a survival advantage. There is uncertainty whether this represents an effect of chemotherapy or an effect of a better patient subset. This clinical trial demonstrated the difficulty patients have in tolerating the toxic side effects of some chemotherapy regimens.

    
    
  • ○
    

    A Finnish trial demonstrated an 11% improvement in 5-year survival in patients with completely resected T1-3N0 NSCLC receiving adjuvant CAP chemotherapy compared with observation. More than 60% of patients received less than the planned four cycles of CAP because of severe hematologic or gastrointestinal side effects.

    
    
  • ○
    

    The Non–Small Cell Lung Cancer Collaborative Group conducted a meta-analysis of 52 randomized clinical trials performed between 1965 and 1991 with a total of 9387 patients with completely resected NSCLC. Overall, adjuvant chemotherapy improved the 5-year survival rate by 5%, representing a 13% reduction in the relative risk of death. Chemotherapy regimens that used primarily alkylating agents worsened 5-year survival (15% relative increase risk of death), whereas platinum-based chemotherapy regimens produced a 13% relative decrease in the risk of death.

  
  
  
  
• ▪
  

  More successful results with second- and third-generation platinum-based regimens: In the 1980s, new agents such as etoposide, vinblastine, and vindesine demonstrated improved outcomes combined with a platinum in patients with advanced or metastatic NSCLC (second-generation platinum-based regimens). In the 1990s, newer agents such as gemcitabine, paclitaxel, docetaxel, and vinorelbine (third-generation platinum-based regimens) appeared to improve survival in the advanced disease setting further still. These new regimens set the stage for more promising randomized clinical trials in the adjuvant disease setting.

  
  
  
  
  • ○
    

    The International Adjuvant Lung Trial (IALT) randomized 1867 patients with completely resected stage I, II, or IIIA NSCLC to observation or adjuvant chemotherapy with cisplatin combined with etoposide or vinblastine or vindesine or vinorelbine. The chemotherapy arms resulted in an absolute 4.1% improvement in the proportion of patients surviving 5 years (44.5% versus 40.4%), translating to a 14% decrease in the relative risk of dying.

    
    
  • ○
    

    The National Cancer Institue of Canada (NCI-C) JBR-10 trial randomized 482 completely resected patients with stage IB-II NSCLC patients to observation versus four cycles of adjuvant cisplatin and vinorelbine. Although only 58% of patients were able to tolerate three or more of the planned cycles of treatment, there was a 15% absolute improvement in 5-year survival favoring the chemotherapy arm (69% versus 54%, P =0.03) representing a 40% reduction in the relative risk of dying.

    
    
  • ○
    

    The Cancer and Leukemia Group B (CALGB) trial 9633 evaluated the role of adjuvant carboplatin and paclitaxel in 344 patients with surgically staged IB NSCLC randomized to observation versus chemotherapy. Unfortunately, an 8% improvement in survival at 3 and 4 years (79% versus 71% and 69% versus 61%, respectively) narrowed to only a 2% advantage at 5 years (59% versus 57%). The overall hazard ratio for survival of 0.8, suggesting a 20% decreased risk of death, was not statistically significant ( P = 0.10).

    
    
  • ○
    

    The Adjuvant Navelbine International Trialist Association (ANITA) trial randomized 840 patients with completely resected stage IB-IIIA NSCLC to postoperative observation versus four cycles of cisplatin and vinorelbine (identical to NCI-C JBR-10 trial). Again only 61% of patients completed three or more cycles of the planned chemotherapy regimen because of the severity of treatment-related side effects. However, the adjuvant chemotherapy arm was associated with an absolute 8.6% improvement in 5-year survival and an improvement in median survival from 43.7 months to 65.7 months ( P = 0.017).

  
  
  
  
• ▪
  

  Uracil/Tegafur-Based Adjuvant Trials in Japan

  
  
  
  
  • ○
    

    Uracil/Tegafur (UFT) is an oral prodrug of 5-fluorouracil that has been studied in six randomized clinical trials in Japan.

    
    
  • ○
    

    A meta-analysis of 2003 Japanese patients with stage I NSCLC randomized to postoperative observation versus UFT demonstrated a 26% decreased risk of death with UFT taken orally for 2 years. UFT was associated with an absolute 5% improvement in survival at 5 years (81.5% versus 76.5%, P =.011) and 7.7% at 7 years (77.2% versus 69.5%, P =.01).

    
    
  • ○
    

    These results have not been confirmed in randomized clinical trials outside Japan.

  
  
  
  
• ▪
  

  Current standard of care for adjuvant chemotherapy in NSCLC

  
  
  
  
  • ○
    

    Patients with completely resected stage IIA, IIB, or IIIA NSCLC should be offered platinum-based adjuvant chemotherapy

    
    
    
    
    • ▪
      

      The data support the preferential use of cisplatin in most patients.

      
      
    • ▪
      

      Carboplatin is acceptable in patients who are intolerant of cisplatin and who are unable to tolerate or accept risks of nephrotoxicity (e.g., creatinine clearance below 60 mL/min, congestive heart failure, or cardiovascular condition intolerant of copious normal saline hydration and mannitol-forced diuresis, or diabetes), ototoxicity (e.g., pre-existing hearing loss or tinnitius) or peripheral neuropathy (e.g., pre-existing peripheral neuropathy, diabetic neuropathy, or trade or activity dependent on fine manual motor skills), and the elderly.

      
      
    • ▪
      

      The data most strongly support the use of vinorelbine added to cisplatin. Based on the results of randomized trials in patients with advanced stage IIIB or IV NSCLC, other third-generation agents, such as gemcitabine, paclitaxel, or docetaxel are probably at least equally efficacious, but direct data in the adjuvant setting are limited.

    
    
    
    
  • ○
    

    Subset analyses suggest that improvement in survival is greatest in patients with the most advanced stage of localized disease (stage IIIA) and decreases proportionately with earlier stage. For example, no survival benefit can be convincingly demonstrated in patients with stage IB disease except in the T2N0 subgroup with tumor diameter greater than 4 cm.

  
  
  
  
• ▪
  

  Future directions

  
  
  
  
  • ○
    

    Patient selection: Use of genomic analyses of resected tumor specimens to identify patients at highest risk of disease recurrence thus sparing chemotherapy toxicities in patients more likely to be cured with surgery alone.

    
    
  • ○
    

    Chemotherapy selection: Use of biochemical, biologic or genomic analyses of resected tumor specimens to identify drug sensitivity or resistance features (e.g., high ERCC-1 tumor expression predicts unfavorable response to platinum).

    
    
  • ○
    

    Biologically targeted therapy: The addition of biologically targeted agents to chemotherapy. Current studies exploring the use of the epidermal growth factor receptor–targeted agent, erlotinib, or the vascular endothelial growth factor–targeted agent, bevacizumab.

  
  

PREOPERATIVE (NEOADJUVANT) CHEMOTHERAPY

• ▪
  

  Goal: Reduce locoregional disease to improve surgical resectability and eradicate micrometastatic disease to improve survival.

  
  
  
  
  • ○
    

    Patients with pathologic N2 NSCLC have a survival rate of 5% to 15% following complete surgical resection, depending on the extent of mediastinal nodal involvement and less than 5% following incomplete surgical resection.

    
    
  • ○
    

    The fact that two thirds of first recurrences following complete surgical resection occur at distant sites reflects the micrometastatic nature of NSCLC and the potential benefit from effective systemic therapy.

    
    
  • ○
    

    Stage IIIA (N2) NSCLC is a heterogeneous disease and can be divided into three general groups :

    
    
    
    
    • ▪
      

      Patients with minimal N2 involvement found incidentally during or after surgery

      
      
    • ▪
      

      Patients with limited N2 disease (one or two nodal stations) diagnosed preoperatively with imaging or surgical procedures (mediastinoscopy or transbronchial/transesophageal biopsy)

      
      
    • ▪
      

      Patients with multistation or bulky mediastinal nodal involvement

    
    
    
    
  • ○
    

    The potential benefit of preoperative chemotherapy intended to improve the rate of complete surgical resection and reduce micrometastatic disease must be weighed against the potential growth of tumor that may occur by delaying surgery and the potential adverse effects of cytotoxic therapy on perioperative morbidity and mortality

  
  
  
  
• ▪
  

  Early feasibility studies: Several small phase II studies of preoperative chemotherapy performed in the early 1990s.

  
  
  
  
  • ○
    

    Established the safety of this approach (no apparent increase in surgical morbidity or mortality)

    
    
  • ○
    

    Suggested that chemotherapy effectiveness, as measured by objective response rates, was significantly better in this earlier disease setting than in stage IIIB/IV disease (response rates of 50% to 70% versus 20% to 30%, respectively)

    
    
  • ○
    

    Answered valid concerns regarding the frequency of chemotherapy-resistant tumors progressing to an unresectable stage by delaying surgery (less than 10%)

    
    
  • ○
    

    Demonstrated that 5% to 15% of patients with pathologically documented N2 involvement could be downstaged to pN0 with possible improved survival outcomes

  
  
  
  
• ▪
  

  Initial phase III randomized trials: Several pioneering phase III randomized trials were conducted in the early to mid 1990s. Although some of the results were promising, these first efforts suffered from several problems including trial size, variable pretreatment assessment of mediastinal nodal involvement, or variable stages of disease, the noncontrolled use of postoperative chemotherapy or radiation therapy and the use of older second-generation, platinum-based regimens.

  
  
  
  
  • ○
    

    A randomized trial conducted in Spain selected patients with stage IIIA (not all N2 documented) disease to undergo surgery alone or preoperative chemotherapy with mitomycin, ifosfamide, and cisplatin. All patients received postoperative radiotherapy. The study was closed early, with only 60 total patients (30 per arm) because of a striking survival benefit from chemotherapy. The median survival time was 22 months in the neoadjuvant chemotherapy group versus 10 months in the surgery only group. Survival at 3 years was 20% versus 10%, respectively.

    
    
  • ○
    

    A randomized trial conducted at the MD Anderson Cancer Center selected patients with stage IIIA (not all N2 documented) disease for randomization to surgery alone versus preoperative chemotherapy with etoposide, cyclophosphamide, and cisplatin. Patients with incompletely resected disease received postoperative radiation therapy and patients evaluated to have responded to the induction chemotherapy received three additional cycles of chemotherapy postoperatively. This study was also closed early with only 60 patients because of an impressive survival benefit from neoadjuvant chemotherapy. The median survival time was 21 months in the neoadjuvant chemotherapy group (28 patients) versus 14 months in the surgery only group (32 patients). Survival at 3 years was 43% versus 19%, respectively.

    
    
  • ○
    

    A larger phase III trial in France randomized 355 patients with resectable stage IB to IIIA NSCLC to two cycles of mitomycin, ifosfamide, and cisplatin before surgery (and two additional cycles following surgery) or to surgery alone. Despite a striking difference in median survival between the two arms of the study (37 months versus 26 months, respectively), the difference was not statistically significant ( P = 0.15). The percent of patients surviving 3 years was 52% and 41%, respectively.

    
    
  • ○
    

    A fourth trial conducted in Japan was closed early because of slow accrual and no apparent differences in survival. In this study, 62 patients with documented mediastinal nodal involvement (N2) were randomized to preoperative treatment with vindesine and cisplatin or to surgery alone. The median survival times were 17 months and 16 months, respectively, and the percent of patients surviving 3 years was 23% and 26%, respectively.

  
  
  
  
• ▪
  

  Neoadjuvant trials with newer platinum-based regimens: As was the case in adjuvant chemotherapy trials, the newer and more effective platinum-based regimens that incorporated gemcitabine or one of the taxanes, paclitaxel or docetaxel, began to be tested in the preoperative setting in the late 1990s and early 2000s. Furthermore, lessons learned from the pioneering neoadjuvant trials mentioned earlier led to more standardized patient selection and mediastinal nodal staging.

  
  
  
  
  • ○
    

    At least four phase II trials tested cisplatin and gemcitabine in the preoperative setting in patients with resectable N2 disease or more advanced and unresectable stage IIIB disease. The treatment was well tolerated, and the radiographic objective responses ranged from 57% to 70%, which is considerably higher than what generally occurs in patients with stage IV disease. Furthermore, complete resection rates appeared to be increased and complete pathologic downstaging of mediastinal nodal involvement to pN0 occurred in about 10% of patients overall. Median survival times generally ranged from 19 to 20 months.

    
    
  • ○
    

    The Lung Cancer Project Group of the Swiss Group for Clinical Cancer Research (SAKK) treated 90 patients with stage IIIA (pN2) NSCLC with three consecutive cycles of cisplatin and docetaxel before surgical resection. Overall, 66% of patients achieved a radiographic clinical response (greater than 50% tumor regression). Surgery was performed in 78 of the 90 patients (12 nonsurgical patients included one ineligible, nine progressive disease during or after chemotherapy, one deteriorated lung function, and one refusal); 75 were able to undergo resection (three were found to be unresectable). Radiographic clinical response was associated with a higher complete resection rate (93% versus 57%) and higher rate of nodal downstaging (73% versus 40%). Fourteen of the 75 patients (19%) completing surgical resection achieved pathologic complete response (pCR) with complete tumor necrosis or necrosis plus fibrosis. Multivariate analysis identified mediastinal nodal clearance and complete resection as strongly prognostic for increased survival.

    
    
  • ○
    

    The European Organization for Research and Treatment of Cancer (EORTC) studied carboplatin and paclitaxel in 90 patients with biopsy-proven N2 disease. The objective radiographic response rate was 66%. Patients responding to the neoadjuvant chemotherapy were subsequently randomized to receive either surgery or radiotherapy; therefore, there are limitations in assessing the effects on resectability and survival. The efficacy of the carboplatin and paclitaxel regimen was further investigated in a multicenter phase II study in the United States in 94 patients with stage IB-II NSCLC and negative mediastinoscopy. Patients received two cycles of chemotherapy, followed by surgery, followed by three cycles of chemotherapy. This study demonstrated that patients better tolerate preoperative chemotherapy than postoperative chemotherapy and that relatively high radiographic response rates and complete resection rates could be achieved (60% and 81%, respectively).

  
  
  
  
• ▪
  

  Randomized phase III neoadjuvant trials in early stage lung cancer

  
  
  
  
  • ○
    

    Because the trials summarized above appeared to demonstrate that preoperative chemotherapy could be given safely and effectively to patients with N2 disease or early-stage disease, several randomized trials in patients with stage IB-IIIA NSCLC were initiated.

    
    
    
    
    • ▪
      

      The Southwest Oncology Group (SWOG) 9900 selected patients with clinical stage IB-IIIA (N1) disease for randomization to neoadjuvant carboplatin and paclitaxel, followed by surgery or to surgery alone. Mediastinoscopy was required in all patients with mediastinal adenopathy. However, this study was closed early with 354 patients because of the emerging positive results of adjuvant chemotherapy trials and the ethical problems of not giving adjuvant chemotherapy to the surgery-only group. Despite the early closure, it was the largest randomized trial comparing preoperative chemotherapy to surgery alone. Nearly half of the patients in the neoadjuvant chemotherapy arm achieved greater than 50% radiographic reduction in cancer preoperatively and an improvement in median progression free survival (35 months versus 20 months, P = 0.07).

      
      
    • ▪
      

      Several other neoadjuvant trials are either ongoing or recently closed, including trials in which neoadjuvant chemotherapy is compared directly with adjuvant chemotherapy.

    
    

  
  
  
  
• ▪
  

  Role of radiation therapy in the neoadjuvant (preoperative) setting

  
  
  
  
  • ○
    

    Although radiation therapy and concurrent chemotherapy are the standard of care for patients with unresectable stage IIIA or IIIB NSCLC, the role of radiation therapy combined with neoadjuvant chemotherapy in resectable stage IIIA-N2 NSCLC is more controversial.

    
    
    
    
    • ▪
      

      In a trial headed by the SWOG and participated in by other cooperative groups in the United States (Intergroup Study 0139) , 429 patients with documented N2 disease were treated with concurrent chemotherapy (cisplatin and etoposide) and radiation therapy (45 Gy) and then randomized to surgery or to an additional 18 Gy of radiation therapy (63 Gy total). Both groups were eligible to receive an additional two cycles of cisplatin and etoposide following surgery or radiation therapy. There were no statistically significant differences in survival between the two treatment arms except in subset analyses based on the type of surgery performed (pneumonectomy versus lobectomy). Patients undergoing pneumonectomy demonstrated a significantly inferior survival compared with matched patients undergoing chemotherapy and radiation therapy only in part related to a postoperative mortality rate of 26% in patients undergoing pneumonectomy. Many of the postoperative deaths were due to severe acute respiratory distress syndrome in the remaining lung. On the other hand, patients undergoing lobectomies appeared to have a survival advantage compared with matched patients undergoing chemotherapy and radiation therapy only (median survival 34 months versus 22 months, respectively, and 5-year overall survival of 36% versus 18%, respectively, P =0.002). Most interestingly, the 5-year survival of the 42% of patients achieving pathologic downstaging of mediastinal nodes to N0 was greater than 40%. These provocative results are currently being pursued in further studies.

    
    

  
  
  
  
• ▪
  

  Current standard of care regarding treatment of patients with stage IIIA (N2) NSCLC

  
  
  
  
  • ○
    

    Patients with N2 NSCLC represent a relatively heterogeneous group of patients. Therefore, no one treatment plan fits all.

    
    
    
    
    • ▪
      

      Patients with multistation nodal involvement or bulky nodal metastases are probably best treated with concurrent chemotherapy and radiation therapy. Surgery in this subset is best limited to patients who achieve an excellent radiographic response and in whom all disease can be resected without the need for a pneumonectomy. But this requires that radiation therapy be limited to not more than 45 Gy because of the great difficulty in performing complete surgical resections safely following higher doses of radiation that usually result in considerable fibrosis of soft tissue and lung. Most patients in this category are probably best treated with chemotherapy and concurrent definitive (>60 Gy) radiation therapy.

      
      
    • ▪
      

      Patients with minimal single nodal station N2 disease can be treated either with preoperative (neoadjuvant) or postoperative (adjuvant) chemotherapy, although emerging data make the former a very attractive consideration.

      
      
    • ▪
      

      Patients with intermediate volume N2 metastases involving one or two nodal stations can be treated with either neoadjuvant chemotherapy alone or with concurrent radiation therapy (45 Gy). As noted earlier, radiation therapy should not be performed in patients requiring a pneumonectomy for complete surgical resections.