Multidisciplinary Approach to Potentially Curable Non-Small Cell Carcinoma of the Lung

Introduction

Patients with non-small-cell lung cancer (NSCLC) represent an extremely heterogeneous group with respect to several factors, including age, gender, performance status, histologic type, stage, presence of other major comorbid conditions, and degree of enthusiasm about therapy. It is necessary to evaluate all these factors and to appreciate the most recent treatment data before a management plan is recommended for an individual patient.

With regard to curability, we define potentially curable NSCLC patients as those surgically/pathologically staged patients with stage I and II disease and a small minority (about 15% of the entire group) of stage IIIA patients, a group that will be more specifically defined later in this article.

Historically, treatment of stages I, II, and certain subsets of stage IIIA NSCLC has consisted of surgical resection.[1-4] The majority of patients with stages IIIA and IIIB disease are not resectable and have been treated with radiation therapy [5-7]; patients with stage IV disease generally have received either symptomatic care alone or various types of chemotherapy.

The idea of combined-modality therapy for various subsets of patients is not new. The rationale has been present for many years and is based on the fact that micrometastases are usually present at the time of diagnosis, except in patients with stage I disease. We will review the current status of combined-modality therapy for patients with potentially curable NSCLC.

Stage, Surgery, and Prognosis

Risk of Micrometastasis

Assessment of the risk of micrometastasis in patients with NSCLC is similar in many ways to risk assessment in patients with breast carcinoma. The size of the primary lesion and the presence, number, size, and location of local nodal metastases are fairly accurate predictors of the presence of other distant micrometastatic tumors. Adequate surgical-pathologic staging for carcinoma of the breast includes removal of the primary tumor and axillary nodes. This is equally true for patients with NSCLC, but hilar and mediastinal node sampling or dissection has been practiced less rigorously and with little enthusiasm by many thoracic surgeons, in no small part due to the questionable therapeutic value of these procedures and the poor or ineffective systemic therapy available.

Accurate final staging and therapy for patients with clinical stages I and II disease (no enlarged nodes seen in the mediastinum on chest radiograph and CT scan) usually require thoracotomy with complete excision of all known tumor either by lobectomy or pneumonectomy with concomitant extensive node sampling or, ideally, a complete node dissection to identify N1, N2, or N3 disease. These surgical-pathologic data provide an accurate assessment of the degree of risk for distant micrometastatic tumors and, therefore, accurate staging and classification of patients. The success of curative resection can be predicted from these data, and potential candidates for adjuvant therapy can be identified.

The majority of patients do not present with clinical stages I and II disease, but rather, have either clinical stage III (enlarged nodes in the mediastinum on chest radiograph or CT scan) or stage IV disease. Patients with stage III disease are a heterogeneous group. Histologic documentation of N2 or N3 disease (stage IIIB) is desirable and can be performed by transbronchial biopsy, fine-needle aspiration, mediastinoscopy, or mediastinotomy before definitive therapy is planned. Occasionally, it is necessary to perform staging and therapy (resection) at the same time during thoracotomy.

Disease Classification by Prognosis

There are several distinct groups or subsets of patients with stage IIIA disease. It has been recognized for years that patients with chest wall involvement, parietal pleural involvement, and superior sulcus tumors without local node involvement (T3 N0) are potentially curable with resection (about 30% to 40%). Involvement of N1 nodes decreases the cure rate to approximately 15%, and involvement of N2 nodes makes cure unlikely following surgical resection.

Furthermore, the amount of ipsilateral mediastinal node involvement is an important determinant of prognosis following resection. Patients with either clinical stages I or II disease (normal mediastinum on chest radiograph, CT scan, and/or preoperative biopsy) who are found to have minimal N2 disease (stage IIIA) at either thoracotomy (microscopic involvement of one node station/level) or by preoperative biopsy have curative potential (about 15% to 30%) after surgical resection. In contrast, patients with bulky N2 disease ( 2 cm or more nodes in several node stations or levels) cannot be cured, even after radical resection. Patients with smaller amounts or nonbulky but multiple N2 disease (ranging from mi- croscopic involvement of multiple node levels to macroscopic involvement 2 cm or less) have a small chance of cure following resection, with or without postoperative radiation therapy (about 5% to 10%). The T-status is also important; considering the same N status, patients with T2 and T3 lesions are cured less often than are patients with T1 lesions.

Subsets of Stage IIIA Disease--Given the variable possibility of cure following surgical resection, with or without radiation therapy, for patients with stage IIIA disease, and the evolving data favoring multimodality therapy, it is now reasonable to divide these patients into several subsets. This division is particularly appropriate for planning and evaluating multidisciplinary therapy. Precise staging is necessary to design and interpret clinical trials. In addition, knowledge of these data allows for a more logical decision regarding management for these patients.

We propose a division of patients with stage IIIA disease as outlined in Table 1 (subsets IIIA1-IIIA7). This definition was proposed, in part, by Ruckdeschel and associates.[8] These subsets are listed in order from the best to worst prognosis (following surgical resection and/or definitive radiation therapy). It appears that the outcome of multimodality therapy differs among patients in these subsets of stage IIIA disease, making their distinction relevant both prognostically and therapeutically.

Although stage IIIB disease is not included here, groups of patients with stage IIIB disease appear to have a similar prognosis to that of patients in the worst subsets of stage IIIA disease. These patients with stage IIIB disease include some with T4 N0 disease, as well as those with mediastinal and/or hilar N3 disease only (without pleural/pericardial effusions or supraclavicular node involvement). These patients are also candidates for multimodality approaches.

Even patients in the most favorable subsets with stage IIIA (IIIA1-IIIA4) disease have distant micrometastases nearly 60% to 90% of the time. Nearly all patients with stage IIIA disease with either bulky ( 2 cm or less) multiple levels of N2 (IIIA5, IIIA7) or T3 N2 (IIIA6) involvement have distant micrometastasis, and the prospect of curative resection or radiation therapy is very unlikely. These patients are common, representing the largest subgroups of patients with stage IIIA disease. Therefore, from a historical perspective, patients with potentially curable NSCLC represent those patients with surgical-pathologic staged I and II disease (approximately 75% and 50% cure rates) and the small minority (nearly 15% of the entire group) of patients with substages IIIA1-IIIA4 disease (cure rates of 5% to 40%).

When considering multimodality therapy, it is essential to define the patient population and prognosis. Furthermore, the performance status and physiologic status of the heart, lungs, liver, and kidneys are important considerations. Medically inoperable patients generally are not candidates for multimodality therapy. With a few exceptions, nearly all the clinical multimodality trials have included patients with bulky, unresectable stage IIIA (N2) or IIIB disease (mediastinal and/or hilar N3 involvement) easily seen on plain radiographs or CT scans of the chest. In our opinion, the results observed in several recent trials with these patients are also applicable to patients who have surgically resectable, potentially curable tumors.

Chemotherapy

Chemotherapy is now the preferred therapy for ambulatory patients with stage IV disease. Once systemic therapy is "effective," it will also serve as the primary therapy for the majority of patients with stages IB, II, and III disease. Local therapy will then be "adjuvant therapy."

Chemotherapy has developed slowly for NSCLC; many agents active against several other neoplasms have been ineffective in treating NSCLC. We refer to the era of single-agent and combination chemotherapy before the development of cisplatin(Drug information on cisplatin) (Platinol) (1965-1980) as first-generation chemotherapy (Table 2). First-generation chemotherapy was essentially inactive against NSCLC, with no effect on the survival of patients with stage IV disease and no apparent additive effect when combined with surgery or radiation therapy for patients with stage III disease.

Cisplatin became the cornerstone of second-generation chemotherapy (1981-1992). When used usually in combination with other agents, it was capable of prolonging survival, as well as often providing useful palliation for patients with stage IV disease who had a good performance status. The less toxic platinum analog, carboplatin(Drug information on carboplatin) (Paraplatin), has proven to be as useful as, if not more useful than, cisplatin. Cisplatin-based therapy has also prolonged the survival of patients with stage III disease when used in concert with sequential and probably concurrent radiotherapy. Preoperative chemotherapy for patients with resectable stage III disease also looks more promising than resection alone.

The present era (1993-?) of third-generation chemotherapy has essentially just begun with the evaluation of several new, relatively active single agents, including paclitaxel(Drug information on paclitaxel) (Taxol), vinorelbine (Navelbine), docetaxel(Drug information on docetaxel) (Taxotere), irinotecan(Drug information on irinotecan) (Camptosar), gemcitabine(Drug information on gemcitabine) (Gemzar), and topotecan(Drug information on topotecan) (Hycamtin). These agents are rapidly being tested both as single agents[9-18] and in various combinations with other drugs, including cisplatin, carboplatin, and etoposide(Drug information on etoposide) (Etopophos, VePesid) (Table 3).[19-40] Other phase II studies have also evaluated many of these new agents in concert with radiation therapy.[34-36] Although paclitaxel has demonstrated only about a 20% to 25% objective response rate, it has consistently been associated with a 1-year survival rate of 35% to 40%, which is superior to that of other single agents and most combinations for patients with stage IV disease.

Recently, a large, randomized comparison of paclitaxel and cisplatin has shown them to be superior to standard cisplatin and etoposide.[39] These results are particularly impressive in light of the prolonged survival. Several series also have studied paclitaxel and carboplatin, and findings have shown quite impressive 1-year survival rates, in addition to response rates ranging from nearly 30% to 60%.[16-21] The Cancer and Leukemia Group B (CALGB) is currently comparing paclitaxel plus carboplatin vs paclitaxel alone. A randomized comparison performed by the Southwest Oncology Group (SWOG) has shown the superiority of cisplatin and vinorelbine over cisplatin alone.[40] The current SWOG trial is comparing paclitaxel and carboplatin vs cisplatin and vinorelbine. The Eastern Cooperative Oncology Group (ECOG) has prepared a four-arm study of patients with stage IV disease; the regimens being compared are paclitaxel/carboplatin, paclitaxel/cisplatin, docetaxel/cisplatin, and gemcitabine/cisplatin.

With these recent encouraging findings, additional refinements are likely to continue. The combined-modality therapies reported with second-generation chemotherapy for patients with stage III disease (see Table 4, Table 5, and Table 6) and the current ongoing trials must be interpreted in light of the evolving and improved third-generation chemotherapy.[58] Based on the results seen with these newer, active third-generation combinations, a case can be made for their immediate incorporation into combined-modality programs for patients with stages II and III disease.