Gemcitabine and vinorelbine in patients with advanced lung cancer: preclinical studies and report of a phase I trial.

PURPOSE: This study was designed to assess the efficacy of gemcitabine plus vinorelbine using the mouse Lewis lung carcinoma model and to translate this regimen to a phase I clinical study of these two agents in patients with advanced lung cancer. MATERIALS AND METHODS: Using the mouse Lewis lung cancer model, employing growth delay and isobologram analysis, we demonstrated that gemcitabine, in combination with vinorelbine, produced additive activity with little increased toxicity over a wide range of doses. At the highest dose level studied, antagonism was observed. Based on these results, we initiated a phase IGemcitabine and vinorelbine in patients with advanced lung cancer: preclinical studies and report of a phase I trial study of this combination at the Dana Farber Cancer Institute (DFCI) in patients with untreated or once pretreated non-small-cell lung cancer (NSCLC) or once pretreated small-cell lung cancer (SCLC). Vinorelbine (given in an intravenous bolus) and gemcitabine (given in a 30-min infusion) were initially administered to patients at a dose of 15 mg/m2 and 500 mg/m2, respectively, on days 1, 8, and 15 of a 28-day cycle. Seven dose levels were subsequently explored over the course of the study. There was no intrapatient dose escalation. RESULTS: From November 1996 to March 1998, 40 patients were enrolled: 32 had NSCLC, 5 had SCLC and 3 had mixed disease (both SCLC and NSCLC). The patients were evenly divided by gender, the median age was 58 years (range 38 to 73 years), and the median ECOG performance status was 1 (range 0 to 2). All patients had normal renal and hepatic function and none had previously received gemcitabine or vinorelbine. Toxic reactions included mild to moderate fatigue, nausea, constipation, and, most significantly, neutropenia and thrombocytopenia. Phlebitis was a major problem when central venous lines were not used with 15% grade 1/2 events. The day-15 dose was held in 43% of patients at the expanded dose. No true maximum tolerated dose was reached after completion of seven dose levels. Dose level 4 (22.5 mg/m2 vinorelbine and 1,000 mg/m2 gemcitabine) was chosen for expansion and future study due to the potential increased ability of patients to receive the full doses on time. CONCLUSIONS: We conclude that this drug combination and dosage are feasible and have potential as either a front- or second-line chemotherapeutic regimen for advanced lung cancer, and phase II/III trials should be performed. However, hematologic toxicities, as found in this study, could probably be reduced with treatment on days 1 and 8 every 21 days, and current literature would suggest this to be the preferred schedule.

Lung cancer in patients under age 40.

A retrospective review of patients <40 years (n=91) seen at the Dana-Farber Cancer Institute and Brigham and Women's Hospital from January 1, 1983-January 1, 1993 was carried out. Of 91 patients, there were 43 men and 48 women with a median age of 36 years (range 28-39). Eighty percent of patients were cigarette smokers for a median of 25 pack years (range 2-68). Ninety-one percent were symptomatic at presentation. The ECOG performance status (PS) was 0 or 1 in 83%. At the time of diagnosis 15% had stage I/II, 17% stage IIIA, 22% stage IIIB and 45% stage IV disease. The most common histopathology was adenocarcinoma (46%), followed by small cell carcinoma (14%), squamous cell carcinoma (12%), large cell undifferentiated (8%) and other types (20%). The median survival for all 91 patients was 1 year with 2 and 5 year survivals of 30% and 18% respectively. Five year survival was related to stage of disease: 60% for patients with stage I, 58% for stage II, 36% for stage IIIA, 10% for stage IIIB, and 3% for stage IV disease. Factors that had no significant effect on overall survival included gender, histologic subtype, degree of differentiation, presence or absence of symptoms, and sites of metastases. Factors that adversely affected survival by univariate analysis included advanced stage of disease, poor PS, duration of symptoms for more than 3 months, and 5% or greater body weight loss. By multivariate analysis only stage (P<0.001) and weight loss (P=0.02) affected survival. This data plus results of other published studies show that young patients under age 40 with lung cancer, compared to the more common older patients, have an increased percentage of women, have a longer duration of symptoms, more often have adenocarcinoma with lower frequency of squamous cell carcinoma and sometimes small cell carcinoma, and more often present with advanced disease. Despite these differences, overall patient survival remains poor and is similar to that of older patients.

Role of serum tumor markers CA 125 and CEA in non-small cell lung cancer.

BACKGROUND: CA 125 and CEA are valuable serum tumor markers that can be used to monitor response to therapy in patients with various solid tumors. Systemic studies of CA125 and CEA have not been evaluated in lung cancer. In this study, we report the serum levels of CA 125 and compared it to CEA in newly diagnosed lung cancer and analyzed the serum levels of these markers pre- and post-therapy. MATERIALS AND METHODS: Two hundred and sixteen patients with newly diagnosed non-small lung cancer were evaluated. CA 125 and CEA levels were correlated with stage and histopathology. RESULTS: CA 125 levels and CEA levels were shown to be lower in patients with early stage disease as compared to patients with unresectable or metastatic disease. CEA levels were significantly higher among patients with adenocarcinoma, while there was no statistically significant relationship between histology and CA 125. There was a statistically significant difference in the CEA and CA 125 levels dependent on tumor size. Thirty-seven patients were analyzed for responses to chemotherapy and responders are more likely to have decreases in CA 125 or CEA. CONCLUSION: When abnormally elevated inpatients witlrlung cancer, CA 125 and CEA are useful indicators of disease extent, a useful clinical therapeutic marker, and may potentially have important prognostic value.

Use of cytokeratins 7 and 20 in determining the origin of metastatic carcinoma of unknown primary, with special emphasis on lung cancer.

Metastatic carcinoma of unknown primary is a common problem, accounting for up to 10-15% of all solid tumours at presentation. Proper identification of the site of origin has prognostic and therapeutic significance. Prior immunohistochemical methods to identify the site of origin have been useful in a limited number of cases. Differential cytokeratin staining may be useful in this setting, particularly in identifying metastases from lung cancer. We have identified 144 cases of metastatic carcinoma of unknown primary to bone, lung or liver at Brigham and Women's Hospital between 1 January 1997 and 1 July 1998. Cytokeratin (CK) 7 and CK20 were used in 75 of these cases to narrow down the possible sites of the primary tumours. All of these cases were ambiguous as to the site of the primary tumour. Forty-five cases were CK7+/CK20-, 15 cases were CK7-/CK20-, 9 cases were CK7-/CK20+ and 6 cases were CK7+/CK20+. Three of the cases were selected for detailed presentation and discussion as well as a discussion of the pertinent literature. Overall, the CK7+/CK20- phenotype favours a lung primary, the CK7+/CK20+ phenotype strongly favours transitional cells (urothelial) carcinoma, the CK7-/CK20+ phenotype favours colorectal carcinoma, while the CK7-/CK20- profile is not helpful.

Malignant pleural mesothelioma.

Malignant pleural mesothelioma remains a difficult tumor to treat, much less cure. Currently, the best chance for long-term survival lies with early diagnosis and aggressive surgical extirpation, but given the typically long delay between the onset of symptoms and diagnosis, this is only possible with a high index of suspicion and an aggressive diagnosis workup. Early referral to a tertiary center experienced in the treatment of MPM may be important for several reasons: (1) decreased risk of tumor spread along multiple thoracenesis/biopsy tracts, (2) the availability of specialized pathologic assays for definitive diagnosis, (3) the availability of critical staging modalities (aggressive mediastinoscopy +/- thoracoscopy, MRI scans performed according to specific mesothelioma protocols, and perhaps PET scans), (4) surgical experience with pleurectomy/decortication and/or extrapleural pneumonectomy, that may decrease morbidity and mortality, and (5) the availability of novel adjuvant protocols. Single-modality therapy is unlikely to result in long-term survival. Aggressive surgery is required for optimal debulking, and extrapleural pneumonectomy may offer better local control compared with pleurectomy/ecortication. Delivery of optimal radiation schedules, which may involve large fractions as well as large total doses, is limited by the presence of nearby dose-limiting structures. Current chemotherapy is severely lacking in producing objective responses and improved survival although gemcitabine and IL-2 may be active agents to be combined with radiation and/or other agents. Hyperthermia, photodynamic therapy, intracavitary therapy, and gene therapy are all relatively new techniques under active investigation that should be supported by enrollment in on-going protocols. Predictably, many of these techniques provide greater benefit when used in the setting of adjuvant protocols or minimal residual disease, emphasizing the importance of multimodality therapy.

Dose-intensive therapy for limited-stage small-cell lung cancer: long-term outcome.

PURPOSE: To determine progression-free survival (PFS) and overall long-term survival for limited-stage small-cell lung cancer (SCLC) patients aged 60 years or younger who respond to first-line chemotherapy followed by high-dose combination alkylating agents (cyclophosphamide 5,625 mg/m(2), cisplatin 165 mg/m(2), and carmustine 480 mg/m(2)) with hematologic stem-cell support and chest and prophylactic cranial radiotherapy. PATIENTS AND METHODS: Patients were selected on the basis of their continued response to first-line therapy, their relative lack of significant comorbidity, and their ability to obtain financial clearance. RESULTS: Of 36 patients with stage III SCLC, nine patients (25%) had achieved a complete response (CR), 20 had achieved a near-CR, and seven had achieved a partial response before undergoing high-dose therapy. Toxicity included three deaths (8%). For all patients, the median PFS was 21 months. The 2- and 5-year survival rates after dose intensification were 53% (95% confidence interval [CI], 39% to 72%), and 41% (95% CI, 28% to 61%). Of the 29 patients who were in or near CR before undergoing high-dose therapy, 14 remain continuously progression-free a median of 61 months (range, 40 to 139 months) after high-dose therapy. Actuarial 2- and 5-year PFS rates were 57% (95% CI, 41% to 79%) and 53% (95% CI, 38% to 76%). By multivariate analysis, short intensive induction chemotherapy was associated with favorable outcome (P <.05). CONCLUSION: Use of high-dose systemic therapy with intensive local-regional radiotherapy was associated with manageable treatment-related morbidity and mortality. Patients who were in or near CR before intensification are enjoying an unmaintained 5-year PFS rate of 53%. Late complications were infrequent, and most patients returned to full-time work and activity. A randomized comparison of this approach and conventional-dose therapy should define the use of dose intensification with hematopoietic support in patients with responding limited-stage SCLC.

Analysis of neuroendocrine markers, HER2 and CEA before and after chemotherapy in patients with stage IIIA non-small cell lung cancer: a Cancer and Leukemia Group B study.

Several studies have suggested that biochemical or molecular markers examined in non-small cell lung cancer carry prognostic or treatment response information. Non-small cell lung cancer patients whose tumors have neuroendocrine (NE) features may be more responsive to chemotherapy. In addition, increased expression of HER2 (c-erbB-2), a membrane-bound receptor with tyrosine kinase activity, has been associated with shortened survival. The Cancer and Leukemia Group B (CALGB) performed a study of patients with stage IIIA (N2 nodes positive) non-small cell lung cancer in which patients received initial chemotherapy followed by surgery, then post-operative therapy consisting of sequential chemotherapy and radiation therapy. Since all patients underwent mediastinoscopy, this provided an opportunity to compare pre- and post-chemotherapy tumor specimens to test the hypothesis that these proteins would predict treatment response. In particular, we hypothesized that the post-chemotherapy specimens would be enriched for NE marker negative cells because of the increased sensitivity of NE positive cells to chemotherapy. We performed immunohistochemical analysis for a panel of NE markers [neuron-specific enolase (NSE), Leu-7, chromogranin A (ChrA), synaptophysin (Syn)], HER2 and CEA to determine if there was an effect of therapy on the percentage of cells expressing these markers. Secondary endpoints were a correlation with chemotherapy response and survival. Slides were scored for intensity (0-4) and percentage of cells positive (0-4). Of 61 eligible patients, there were 38 with both pre- and post-chemotherapy specimens. When both intensity of staining and percentage of positive cells were considered, post-chemotherapy specimens had a higher percentage of positive NE markers compared with pre-chemotherapy. In addition, there was no correlation between NE marker, HER2 or CEA expression (prior to or post treatment) and response to chemotherapy or survival. These data do not support the hypothesis that NE positive tumor cells are preferentially killed by chemotherapy in patients with stage IIIA non-small cell lung cancer.

Molecular abnormalities in lung cancer.

PURPOSE: To review several recently described molecular abnormalities in lung cancer and discuss their potential diagnostic and therapeutic relevance. DESIGN: Articles were identified through a Medline search (1966 to 1997) and studies, including reviews, were cited in the references. RESULTS: Molecular mechanisms altered in lung cancer include induced expression of oncogenes, such as RAS, MYC, c-erbB-2, and BCL-2, and loss of tumor-suppressor genes, such as RB, p53, and p16INK4A. RAS is a 21-kd G protein and up to 30% of adenocarcinomas show mutations in K-RAS oncogene. MYC encodes a transcriptional activator and amplification may adversely affect survival in small-cell lung cancer (SCLC). The growth factor receptor c-erbB-2 is overexpressed in up to 25% of non-small-cell lung cancer (NSCLC) cases. BCL-2, a negative regulator of apoptosis, is expressed differently in some NSCLCs. Abnormalities of RB, a key regulator of cell cycle, are detected in greater than 90% of SCLCs. There is an inverse relationship in lung cancer cells between expression of RB and p16INK4A, an upstream regulator of RB. Mutations of p53, with frequencies up to 50% in NSCLC and 80% in SCLC, can lead to loss of tumor-suppressor function, cellular proliferation, and inhibition of apoptosis. The identified molecular abnormalities in lung cancer are currently used to develop diagnostics for detecting early disease, as well as to identify targets for gene therapy. CONCLUSION: Genetic abnormalities involved in the pathogenesis of lung cancer are rapidly being delineated. Understanding molecular abnormalities in lung cancer could potentially lead to earlier diagnosis and the development of novel investigational approaches to the treatment of lung cancer.

Cancer-associated retinopathy in a patient with non-small-cell lung carcinoma.

Cancer-associated retinopathy (CAR) is a paraneoplastic syndrome most often associated with small-cell lung carcinoma (SCLC), and it has been reported in patients with other malignancies. Antibodies against recoverin, a 23-kDa protein, have been found in patients with CAR suggesting an autoimmune phenomenon. Herein is the first report of a patient with non-small-cell lung cancer (NSCLC) in whom anti-recoverin antibodies were detected in the serum. Steroid therapy, chemotherapy, and radiation therapy did not help the patient's vision. Progressive loss of vision in patients with lung cancer should, potentially, be tested for CAR.

Randomized trial of chemotherapy and radiation therapy with or without warfarin for limited-stage small-cell lung cancer: a Cancer and Leukemia Group B study.

PURPOSE: Studies by the Veterans Administration Cooperative Studies Program and Cancer and Leukemia Group B (CALGB) suggested that the addition of warfarin to chemotherapy might enhance response and/or survival in small-cell lung cancer (SCLC). This randomized study evaluated the effect of warfarin with chemotherapy and radiation therapy in limited-stage SCLC. PATIENTS AND METHODS: Patients were randomized to receive warfarin or no warfarin. All patients received three cycles of doxorubicin, cyclophosphamide, and etoposide (ACE). Cycles 4 and 5 (cisplatin, cyclophosphamide, and etoposide [PCE]) were given concurrently with radiation therapy. Three cycles of ACE were given after chemoradiation therapy, but were discontinued due to a high rate of pulmonary toxicity. RESULTS: There were no significant differences in response rates, survival, failure-free survival, disease-free survival, or patterns of relapse between the warfarin-treated and control groups. In patients treated according to the initial design, an increase in failure-free survival seen with warfarin treatment approached significance (P = .07). Preamendment results, while not significant, did not have superimposable treatment survival curves. A landmark analysis at 8 months showed a median survival time after the landmark for complete responders of 33 months with warfarin treatment compared with < or = 13.75 months for complete or partial responders not treated with warfarin (P = .05). Differences between the complete responders in this preamendment population were not significant (P = .103). CONCLUSION: Warfarin does not appear to improve outcome significantly in limited-stage SCLC. However, the differences in some variables between populations before the protocol amendment correspond to the favorable effects of anticoagulants observed in previous studies.

Non-Hodgkin's lymphomas: current classification and management.

Considerable progress has been made in the classification of non-Hodgkin's lymphomas during the past 15 years, and the use of specific monoclonal antibodies directed against cell surface antigens has contributed to the understanding of the immunology of the disease. Early-stage indolent lymphoma is treated with radiotherapy; treatment of advanced-stage indolent lymphoma varies. Aggressive lymphomas are treated with combination chemotherapy with or without regional radiotherapy, and highly aggressive lymphomas are treated with regimens similar to those for children with leukemia.

Ifosfamide/carboplatin/etoposide/paclitaxel in advanced lung cancer: update and preliminary survival analysis.

The primary objective of this study was to define the maximum tolerated dose and toxicity profile of paclitaxel (Taxol; Bristol-Myers Squibb Company, Princeton, NJ), given as a 24-hour infusion, in conjunction with ifosfamide/carboplatin/etoposide (ICE) chemotherapy in patients with advanced lung cancer. Paclitaxel was escalated from 75 to 225 mg/m2 in 25-mg/m2 increments. All patients received granulocyte colony-stimulating factor 5 microg/kg/d from day 4 until the neutrophil count was > or = 10,000/microL. The study population consisted of 41 patients with a median age of 60 years and a median follow-up of 20.7 months. Stage distribution included 5% stage IIIA, 46% stage IIIB, and 49% stage IV. Histology consisted of 61% adenocarcinoma, 12% squamous cell carcinoma, 10% large cell carcinoma, 15% small cell carcinoma, and 2% mixed. The predominant toxicity was hematologic; 63% of patients experienced grade 4 neutropenia and 49% developed grade 4 thrombocytopenia. Fever and neutropenia occurred in 34% of patients. Hematologic toxicity was, in all cases, short-term and reversible and was not dose related. With few exceptions, nonhematologic toxicity was not clinically important. Among 39 patients evaluable for response, 36% achieved a remission (8% complete, 28% partial, 41% had stable disease, and 23% experienced disease progression). Among 33 patients with non-small cell lung cancer, the response rate was 27% (one complete response, eight partial responses, 15 had stable disease, and nine had progressive disease). Among six patients with small cell carcinoma, the response rate was 83% (two complete responses, three partial responses, and one had stable disease). The median survival of all 41 patients was 13.6 months. Survival was almost identical between stage IIIA and stage IV subsets. We conclude that it is possible to safely administer full-dose single-agent paclitaxel with granulocyte colony-stimulating factor support in conjunction with full-dose ifosfamide/carboplatin/etoposide chemotherapy. While response rates observed were not particularly notable, median survival is considerably longer than that usually achieved with combination chemotherapy in advanced lung cancer.

Chemotherapy for advanced non-small cell lung cancer.

Lung cancer is the most lethal cancer in both men and women. Given that chemotherapy for advanced disease is marginally beneficial and noncurative, its use must be governed judiciously, with each decision being evaluated individually for each patient. Chemotherapy for lung cancer has progressed over the past decade, and with the advent of new agents, its future looks promising.

Neoadjuvant therapy for surgically staged IIIA N2 non-small cell lung cancer (NSCLC).

INTRODUCTION: Neoadjuvant therapy in patients with Stage IIIA NSCLC is associated with a 50-70% resection rate and a 3-5 year survival of 20-32%, but few trials have required meticulous staging of the mediastinum to ensure homogeneity of the study population. Continuous infusion cisplatin 25 mg/m2/day 1-5, 5-fluorouracil 800 mg/m2/day 2-5, and high-dose leukovorin 500 mg/m2/day 1-5 (PFL) given every 4 weeks achieved a 41% response rate in metastatic NSCLC (Lynch TJ, Kalish LA, Kass F, Strauss G, Elias A, Skarin A, Shulman L, Sugarbaker D, Frei E. Continuous infusion cisplatin, 5-fluorouracil, and leukovorin for advanced non-small cell lung cancer. Cancer 1994; 73: 1171-1176). The regimen was therefore evaluated in 34 patients with pathologic Stage IIIA N2 disease between 3/91 and 10/92. METHODS: Staging consisted of chest, liver, brain computerized tomography and bone scan, bronchoscopy and surgical mediastinal node mapping. Patients received PFL for 3 cycles, followed by thoracotomy and thoracic radiotherapy (TRT) to 54-60 Gy. RESULTS: Median age was 57 (42-68) years. Demographic factors included: male 56%; adenocarcinoma 59%, squamous cell carcinoma 24%; Stage T3N2 26%, T2N2 56%, and T1N2 18%. No treatment related deaths occurred. Radiographically defined response to PFL was 65% (6% complete). Thoracotomy was performed in 28 patients (82%) (6 had no attempt due to disease progression). Complete resection was achieved in 21 (75%) and seven were unresectable. Pathologic complete response was observed in five patients (15%) and an additional unresectable patient had fibrosis-only documented at thoracotomy for an overall clinicopathologic response rate of 76% (18% pathologic CR). Another ten patients had residual primary with or without hilar disease with resolution of previously documented mediastinal involvement. Six (18%) patients remain alive and disease-free with a median follow-up of 46 (33-50) months, four of whom had achieved pathologic complete response at time of surgery. CONCLUSIONS: Long-term event-free survival was associated with complete surgical resection which in turn was associated with clinical response to chemotherapy. There was a possible trend associating pathologic downstaging (absent residual disease in mediastinal nodes), particularly pathologic complete response observed in patients with non-bulky mediastinal disease, with improved event-free survival. Pathologic downstaging might therefore be a useful surrogate endpoint in trials evaluating the preoperative activity of new chemotherapy regimens. While radiographic response generally correlated with findings at surgery, response as determined by histologic examination of resected tissue was generally more extensive and may more accurately reflect the systemic impact of the chemotherapy regimen.

Molecular and pathologic markers in stage I non-small-cell carcinoma of the lung.

PURPOSE: Although standard treatment of stage I non-small-cell lung cancer (NSCLC) consists of surgical resection alone, approximately 50% of clinical stage I and 30% to 40% of pathologic stage I patients have disease recurrence and die following curative resection. A large number of traditional pathologic and newer molecular markers have been identified, which appear to have important prognostic significance in this population. This review attempts to summarize these data comprehensively. METHODS: Criteria for study selection were English-language reports, identified using Medline and Cancerline, through the fall of 1994. Abstracts from the American Society of Clinical Oncology (ASCO) and the International Association for the Study of Lung Cancer (IASLG) were also reviewed. RESULTS: Molecular markers are classified as molecular genetic markers, differentiation markers, proliferation markers, and markers of metastatic propensity. A number of these markers have been reported to be highly predictive of outcome in stage I NSCLC, and several reports conclude that a specific biomarker may be, aside from clinical stage, the most powerful determinant of prognosis in NSCLC. However, little has been done to clarify the relationships between these newer biologic markers, classic clinicopathologic variables, and clinical outcome. CONCLUSION: At present, a firm conclusion regarding which biomarkers are most important in predicting outcome is not possible, and a model that reliably integrates all independent prognostic variables cannot be developed. A prospective trial is mandatory to address this issue, and a study design is suggested that would facilitate the development of a prognostic index, while simultaneously asking a therapeutic question. The development of a prognostic index would facilitate future trials in which only high-risk stage I patients could be targeted for investigation of postresection adjuvant treatment strategies.

Use of tumor markers in lung cancer.

Biomarkers relevant to lung cancer can be classified into three broad categories: serum markers, tissue markers, and sputum markers. The predominant contribution of serum biomarkers will be in documenting disseminated disease and response to therapy. The most promising role for tissue biomarkers, at least in non-small cell lung cancer, is for the determination of prognosis in resectable patients. Sputum biomarkers offer considerable promise as early detection tools, which may enhance the sensitivity and specificity of screening when used in conjunction with periodic screening chest radiographs.

Neoadjuvant treatment of stage IIIA non-small cell lung cancer. Long-term results.

The multimodality approach to locally advanced Stage III non-small cell lung cancer is continuing to evolve. In this trial, 54 patients with surgically staged IIIA disease were treated with neoadjuvant chemotherapy, surgical resection, and chest radiotherapy. Response to four cycles of CAP chemotherapy (cyclophosphamide, doxorubicin, cisplatin) was 39% (8% complete responses). One septic death occurred. Thoracotomy was performed on 31 patients, of whom 29 (56%) were resected and 24 (44%) were completely resected. Complete resections were more frequently observed in chemotherapy responders. Extranodal mediastinal extension in nonresponding patients was the most frequent reason not to attempt thoracotomy. The overall median times to progression and survival were 11.6 (.7-66.5) and 17.9 (2.8-71.4) months. Long-term disease-free survival was observed in 11 patients (20%) with a median follow-up of 46.5 (24-71) months. All these patients underwent complete resection and constitute 46% of the patients undergoing complete resection. Median times to progression and survival were 33.4 (5.0-66.5) and 33.5 (10-71.4) months for completely resected patients. Although the ability to perform surgery identified a population that has favorable locoregional control and disease-free survival, distant relapse continues to represent the major obstacle to enhanced survival in resected patients. Unresected patients, however, are likely to relapse in both local and distant sites. Response to chemotherapy may not only enhance systemic control, but may also increase the probability of complete resection. Randomized trials should be conducted to evaluate the role of individual modalities (surgery, chemotherapy, or radiotherapy) while applying the remaining modalities maximally. The temptation to compare different treatment approaches should be resisted.