Issues in Nonoperative Management of Locally Advanced Non-Small-Cell Lung Cancer
Issues in Nonoperative Management of Locally Advanced Non-Small-Cell Lung Cancer
The term locally advanced non-small-cell lung cancer (NSCLC) is used to describe disease that is too extensive for primary surgical resection, is limited to the thorax, and, technically, allows inclusion of the entire tumor within a reasonable radiation field. This definition typically includes patients with stage IIIB and bulky stage IIIA lesions and usually excludes patients with a malignant pleural effusion. In the most recent revision of the American Joint Committee on Cancer staging system for lung cancer, T3N0 tumors were reassigned from stage IIIA to stage IIB, due to their distinctively more favorable prognosis, when compared with lymph-node-positive subgroups of stage IIIA disease. Controversy exists regarding the use of surgery as a component of the initial management of patients with clinical stage IIIA disease. This discussion, however, addresses issues related only to the nonsurgical management of patients.
Patients with NSCLC were expected to comprise 25% to 40% of the 178,100 new lung cancer patients diagnosed in 1997 in the United States. Historically, the standard treatment administered to those patients has been a 6-week course of fractionated external-beam thoracic radiation therapy to 60 Gy. The dose of photon irradiation necessary to provide durable intrathoracic control has been investigated in trials conducted by the Radiation Therapy Oncology Group (RTOG). For example, RTOG 73-01, randomized 551 patients to treatment with four arms of thoracic radiotherapy: 40 Gy delivered in a continuous fashion (2 Gy daily, 5 days a week, for 4 weeks), 40 Gy in a split course, 50 Gy as a continuous dose, or 60 Gy as a continuous dose. Patients assigned to the 60-Gy arm achieved the highest response rate (55%), the lowest rates of local tumor failure at 3 years (36% vs 63% in the other arms), and the best 3-year survival rate (20% vs 10% for the other arms). Unfortunately, these tumor-control and survival advantages were lost by 5-year follow-up, with estimated local failures and survival rates (70% and 7%, respectively) identical in the 60- and 40-Gy arms. In response to the therapeutic advantage seen at 3 years in this study, 60 Gy/6 weeks was adopted as the standard dose for definitive radiotherapy of patients with NSCLC. It became evident, however, that higher radiation doses are necessary to control tumors and further improve the survival results.
In a randomized Southeastern Cancer Study Group trial involving 319 patients whose NSCLC was treated either with thoracic radiotherapy, single-agent vindesine, or a combination of the two, the main conclusion drawn was that standard thoracic radiotherapy did not provide a survival benefit. The overall response rate was superior in both radiotherapy arms (30% vs 10%; P = .001), but median survival time was 8.4 months for patients receiving radiotherapy alone, 9.4 months for those receiving radiotherapy/vindesine, and 10.1 months for those receiving vindesine alone (P = .58). The study was criticized because a large proportion of patients on the vindesine arm (37%) received delayed radiotherapy, thus resulting in a study of immediate vs delayed thoracic radiotherapy.
Reports on the capability of any nonsurgical therapy to control NSCLC vary markedly, depending on the nature of the assessment and the time interval since therapy. When posterior-anterior and lateral chest radiographs were used in RTOG 73-01 and a cross-section of the tumor or the pulmonary shadow was recorded, a complete response was reported for 24% of patients treated with 60 Gy, a partial response was noted in 32%, 35% had stable disease, and only 9% were shown to have progressive disease. One has to note, however, that two-dimensional measurements may not reflect true volumetric responses.
If a complete response is defined rigorously as absence of tumor by clinical, radiographic, and bronchoscopic examination, with a negative endoscopic biopsy, and evaluation of response is repeated every 6 months, only 16% to 20% of patients could be said to have had a complete response. Further, when evaluated in accordance with these assessments, only 15% had a partial response, 16% to 20% had stable disease, and 45% to 53% had progressive disease 3 months after the completion of radiotherapy. At 3 years, local control rates are only 7% to 8%.
More recently, an impressive bronchoscopically verified local control rate of 71% at 2 years was reported by King et al, who used a novel hyperfractionated accelerated radiotherapy regimen to a total dose of 73.6 Gy directed to the primary tumor and adjacent enlarged lymph nodes.
As argued elegantly by Emami, tumor control probability for bronchogenic carcinoma can be estimated at 10% for tumors > 4 cm at a dose of 80 Gy, and the probability of controlling an average-sized lung cancer with even 100 Gy is estimated at 50% to 80%. This is consistent with the original observations of Fletcher. Therefore, if local tumor control is a prerequisite for improved survival, one may expect to start seeing the influence of improved local control rates on survival rates only when eradication of the tumor is possible in over 50% of treated patients.
In addition to external-beam radiotherapy, there may be a role for endobronchial brachytherapy as a means of delivering radiotherapy dose escalation to a bulky parabronchial tumor. In one prospective randomized study, local control was improved with the addition of two sessions of high-dose endobronchial brachytherapy to a standard thoracic radiotherapy regimen (P = .05).
Radiotherapy can effectively control small lung tumors. There are several reports of durable intrathoracic control achieved in patients with clinical stage I (T1 or T2) tumors who could not be treated with surgery because of coexisting medical conditions or refusal.[11-17] Such patients provide an opportunity to better assess the effectiveness of radiotherapy, since their longer survival time is due to the lower stage of disease and local control can therefore be evaluated with less of a competing risk of distant failure.
Precise data on the relationship between tumor size (or volume) and degree of local control are lacking in the radiotherapy literature, but it appears that the rate of local failure with standard thoracic radiotherapy increases sharply when the largest tumor diameter exceeds 3 or 4 cm. For example, the intrathoracic failure rate at 3 years was only 4% (1 of 24) in medically inoperable patients whose stage I tumors measured no more than 4 cm, but the rate increased to 47.8% (11 of 23) in patients with larger tumors treated with a hyperfractionated course of radiotherapy to a dose of 48 or 56 Gy.
Similarly, Kupelian et al quoted a 3-year local failure rate of 11% for patients with T1 lesions and 39% for those with T2 tumors. Significant favorable prognostic factors for local control included tumor size of 4 cm or smaller, no chest wall invasion, a radiation dose of at least 60 Gy, and a complete response at 6 months after radiotherapy. It appears, however, that with longer follow-up, local failure rates increase significantly, even for those with small tumors. Nevertheless, definitive radiotherapy can provide 3-year cause-specific survivals of 30% to 49% [13,15,17] for patients with small tumors and no radiographic evidence of lymph-node involvement, serving as a noninvasive equivalent of wedge resection.
A clear-cut dependence of local control and disease-free survival of T1 tumors on radiotherapy dose is evident in several reports,[12,13,17] with a 90% disease-free survival at 3 years when doses of 65 Gy or higher are used, compared with 29% if delivered doses are between 60 and 65 Gy (P = .0611). Overall, it appears that the dose-response relationship in NSCLC is evident only for tumors 3 cm or smaller, at least within the range of 60- to 65-Gy doses. In those patients with larger tumors, doses much higher than 65 Gy would have to be considered to expect local control. This is difficult to achieve with larger tumors because of the constraints of toxicity to the surrounding normal tissues, most notably lung, spinal cord, and heart.
Results achieved with definitive radiotherapy cannot be directly compared with those of surgical resection, since the pathologic status of regional lymph nodes is not routinely investigated prior to initiation of radiotherapy, and patients frequently do not undergo the rigorous systemic staging before radiotherapy that is standard before surgery.
The realization that local control of lung cancer with conventional radiotherapy (2 Gy daily, 5 days per week) remains unsatisfactory has led to various efforts of optimizing radiotherapy, including altering the radiotherapy fractionation schedule. One such alteration, called hyperfractionation, refers to delivery of a larger number of smaller radiation fractions and may allow delivery of a higher total dose to the tumor, resulting in improved local control with the same probability of late effects to surrounding normal tissues.
Hyperfractionation was investigated by the RTOG in a dose-seeking phase II trial (83-11) designed to identify the maximum tolerable dose of hyperfractionated irradiation and to evaluate tumor control at each dose level. Patients with favorable performance and minimal weight loss were randomized to receive 60, 64.8, 69.6, 74.4, or 79.2 Gy, in two daily fractions of 1.2 Gy. Although the median survival time (13 months) and 2-year survival rates (29%) in the 69.6-Gy arm appeared superior to the benchmark standard fractionation results, there was no apparent improvement in 5-year survival results, which ranged between 6% and 8% at all dose levels. Again, as in the RTOG 73-01 trial, a survival benefit was seen for short-term but not for long-term survival, suggesting the need for more aggressive therapies.
In a phase III study coordinated by the RTOGRTOG 88-08/Eastern Cooperative Oncology Group (ECOG) 4588the 69.6-Gy hyperfractionated dose was tested against both standard once-daily radiotherapy and induction chemotherapy/standard radiotherapy. In that study, the hyperfractionated radiotherapy produced an early survival result that was intermediate between that of the combined-modality arm and the standard radiotherapy, with 1-year survival rates of 59%, 51%, and 46%, respectively. This study is discussed in greater detail later in this section.
It has long been recognized that cells in rapidly proliferating normal tissues and in tumors are not only able to divide during a course of radiotherapy but even divide more rapidly than normal, in a process of accelerated repopulation. This is beneficial in the case of normal tissues, allowing for the healing of acute reactions, but it may be detrimental in the tumor, where such proliferation impairs eradication of disease tissue. A fractionation scheme in the form of accelerated hyperfractionated radiotherapy, ie, delivery of more than one standard-sized (1.6-2.0 Gy) fraction daily, may minimize tumor cell repopulation by shortening the overall treatment time, thereby increasing the probability of tumor control for a given dose level.
CHART (Continuous Hyperfractionated Accelerated Radiation Therapy) is a continuous-treatment regimen that tests the hypothesis that tumor-cell repopulation is an important cause of failure in conventional radiotherapy. To counteract repopulation, CHART was designed to deliver 1.5 Gy three times per day for 12 consecutive days, to a total dose of 54 Gy. An interval of at least 6 hours is maintained between radiotherapy fractions to avoid late toxicity in slowly repairing tissue, such as spinal cord.
Preliminary results have been published of a randomized trial comparing CHART with standard radiotherapy with 66.0 Gy in 563 patients with locally advanced NSCLC and a good performance status. With a minimum potential follow-up of 2 years, this study showed significant improvement in survival rates for the CHART-treated patients over conventionally treated patients (30% vs 20%; P = .006). Although the incidence of significant acute esophagitis was higher in the CHART arm (40% vs 19%), it subsided quickly in both arms and without apparent chronic sequelae. These results are exciting, but longer observation will be necessary before final conclusions can be drawn.
In the United States, thrice-daily radiotherapy (1.1 Gy tid, 5 days a week, to 79.2 Gy) was tested in the RTOG 92-05 trial. Results of this study are pending. An ECOG pilot study also was completed in which 30 patients were treated with 1.5 Gy delivered three times daily, to a total dose of 57.6 Gy. The 1-year survival rate of 63% provided the basis for a larger trial to assess the true efficacy of such a regimen.