|
Radiotherapy is a critical partner to surgical resection in the treatment of locally advanced lung cancer. Its primary purpose, when used in combination with surgery, is to downstage tumors to facilitate surgical resection and to sterilize areas of microscopic disease in the mediastinum either before or after surgery. This chapter focuses on the role of adjuvant and neoadjuvant radiotherapy in non-small cell lung cancer (NSCLC). |
|
POSTOPERATIVE (ADJUVANT) TREATMENT The first attempts to improve the outcome of resected NSCLC through radiation involved the use of adjuvant (i.e., postoperative) radiotherapy. There are two primary indications for adjuvant radiotherapy for NSCLC. The first is to sterilize the surgical field to minimize microscopic risk in the mediastinum when an N2 lymph node is found at the time of surgery. The second addresses the surgical margin when there is a positive or close margin on bronchial resection. In 1986, the Lung Cancer Study Group conducted a randomized trial to investigate the role of postoperative (adjuvant) radiotherapy (PORT). In that study, patients were randomized to receive either a 50-Gy dose to the entire mediastinum or no further therapy. The treatment volume included the mediastinum beginning at the sternal notch and extending 5 cm below the carina. The study revealed a significant difference in local control. Local disease control was demonstrated in 97% of patients receiving radiotherapy compared with 59% in the no-therapy control group (p < 0.001). However, because of the high rate of distant failures demonstrated in both groups, neither revealed a significant difference in recurrence-free or overall survival.1 It became clear from this trial that while radiotherapy did have the potential to influence local control in resected NSCLC, overall disease outcome was not affected because of the lack of systemic therapy. During this same period, multiple randomized trials investigating the role of postoperative radiation therapy were being conducted with varied results. An influential meta-analysis of these trials, called the PORT meta-analysis, attempted to determine the impact of these earlier trials. This study, published in 1998, included nine randomized trials, mostly from Europe. These trials included patients with stage I–III NSCLC after surgical resection. The median survival of all patients was 25 months. The results of this trial, unfortunately, showed a decrement in the overall survival of the group receiving radiotherapy, which raised questions about the safety of this approach.2 The hazard ratio for the PORT group was 1.21, which corresponded to a 7% decrease in overall survival at 2 years. It is noteworthy that the local failure rate in the PORT group actually was lower than in the non-PORT group (195/1056 = 18.5% for PORT versus 276/1072 = 25.7% for no PORT). However, death from other causes was higher in the PORT group, and this factor influenced the overall hazard ratio. In addition, when the data were evaluated in terms of N1 versus N2 positive nodes, the effect on survival appeared to be less severe for patients with positive N2 nodes as opposed to positive N1 nodes. Although the study concluded that radiotherapy should not be used in resected NSCLC, use of older radiotherapy equipment (cobalt-60) and large-field treatment likely contributed to the increased hazard ratio noted in this report. In addition, the high rate of distant metastases obliterated the benefits derived from local control.2 Recognition by the radiation therapy community that systemic failure was limiting the effectiveness of combined treatment with surgery plus adjuvant radiotherapy led to new trials designed to incorporate systemic therapy in the adjuvant setting. Keller and colleagues published a randomized trial in the New England Journal of Medicine in 2002 investigating the role of postoperative radiotherapy plus or minus the addition of chemotherapy. The radiotherapy dose to mediastinal nodes prescribed in this study was 50.4 Gy, which is similar to that in previous trials, after accounting for differences in conventional techniques. Chemotherapy consisted of four cycles of cisplatin and etoposide given concurrently with radiotherapy. This trial evaluated 488 patients and demonstrated no difference in survival in either arm of the trial. There appeared to be no added benefit to administering chemotherapy. However, the median survival in this group was an impressive 38 months in both arms. Additionally, local control was 77% in both arms, suggesting a benefit to postoperative radiation therapy.3 As a result of the concern raised by the PORT meta-analysis, a subgroup analysis was done to evaluate the number of potential deaths owing to radiotherapy. This study established the safety of radiation therapy, in that no additional radiation toxicity deaths were observed compared with a similar group of patients who did not receive radiotherapy.4 Results of a more recent trial (RTOG 9705) also have been reported. This trial was nonrandomized and treated patients with adjuvant chemoradiotherapy for stage II and stage III NSCLC. The main difference between RTOG 9705 and Keller's trial is the use of carboplatin-paclitaxel chemotherapy as opposed to cisplatin and etoposide. It showed an improved overall outcome with a median survival of 58 months. The findings suggested a benefit in both stage II and stage III resected NSCLC.5 Also, there were no additional deaths from intercurrent illnesses. Machtay and colleagues, who combined the results from these two trials with their own analysis, concluded that the fears of increased morbidity from postoperative radiotherapy had vanished with the modernization of radiation technology.6 It emerges from all these studies that adjuvant radiotherapy can be given safely after surgery and appears to confer a local control benefit but no improvement in overall survival. The questions that remain unresolved include the role of postoperative radiotherapy in stage II NSCLC with N1 nodal involvement. Currently, there are no randomized data demonstrating that these patients benefit from adjuvant radiotherapy. However, the results of the two most recent trials indicate that the outcomes of postoperative chemoradiotherapy in these stage II patients compare favorably with historical data. Whether these patients should be treated routinely depends on the case in question. Currently, at the Dana Farber Cancer Institute, patients with N1 positive nodes are not routinely offered postoperative radiotherapy. However, if there is a question regarding margin status, mediastinal invasion, or extensive multiple positive nodes, it is offered on a case-by-case basis. A second question relates to radiotherapy dose. All the current study designs deliver 50 Gy postoperatively. However, there remains a persistent local failure rate of 15%. Other aerodigestive malignancies often receive higher doses (57–60 Gy) in the postoperative setting. This has led some to question whether an increased dose would be permissive or helpful in this setting. Further studies are needed to answer this question, although our current standard at the Dana Farber is to deliver 54–60 Gy in the postoperative setting. |
|
NEOADJUVANT TREATMENT Neoadjuvant radiotherapy or chemoradiotherapy is used in patients who are primarily unresectable or who have documented N2 nodal involvement. This represents a larger group of patients than those treated in the adjuvant setting. Formerly, these patients were treated with primary radiotherapy or chemoradiotherapy. In the last 15 years, however, data have emerged suggesting that combined neoadjuvant therapy and surgery can improve outcome. The first significant trial of multimodality therapy (chemoradiation followed by surgery) was conducted by the Southwest Oncology Group and published in 1995. This study treated patients with stage IIIA and stage IIIB NSCLC using preoperative cisplatin and etoposide in combination with a 45-Gy dose of thoracic radiotherapy followed by surgical resection. Over 80% of the patients treated with this approach were found to be resectable. Patients in this trial tolerated the treatment remarkably well, with a treatment-related mortality rate of only 10%. The local control was 90%, and the overall survival was 18 months.7 The most significant finding, however, was that patients who had an absence of mediastinal nodes at surgery experienced a median survival of 30 months. These good responders had a 44% 3-year survival. This trial was encouraging in its demonstration of the safety of trimodality therapy. In addition, it raised the possibility that trimodality therapy may produce better outcomes than chemoradiotherapy alone. Additional support for this approach came in the setting of superior sulcus tumors or Pancoast tumors. These lesions, by definition, lie at the apex of the lung and often present as locally unresectable cancers with no mediastinal nodal involvement. The Southwest Oncology Group ran a phase II trial in patients with superior sulcus tumors with stage T3 or T4 disease but without any clinically evident mediastinal nodal involvement. They treated patients with a dose of 45 Gy of radiotherapy combined with cisplatin-etoposide chemotherapy and then resection. In this trial of 102 patients, 75% were found to be resectable after the induction therapy. Sixty-five percent of patients had a pathologic complete response, and the 2-year survival was 77% in patients who underwent complete resection.8 This trial was followed by a phase III randomized clinical trial, Intergroup 0139, that compared the trimodality approach with the standard of chemoradiotherapy. This multi-institutional study took 10 years to accrue. Four-hundred and thirty-eight patients completed treatment on this trial. The trial was limited to patients who were considered at initial evaluation to be primarily resectable. Patients were screened and then randomized to 45 Gy of thoracic radiotherapy combined with cisplatin-etoposide and followed by surgical resection versus 61 Gy of radiotherapy with concurrent cisplatin and etoposide. The results of this trial have been reported preliminarily and indicate an improvement in local control for the trimodality group over the chemoradiotherapy group. There was no difference in overall survival.9 The latter trial was associated with a large treatment-related mortality in the trimodality group, with 16 of 202 (8%) patients dying of postoperative morbidity. Mortality was particularly high in the subset of patients undergoing complex pneumonectomies and right pneumonectomies (14 of 54, or 26%). An exploratory subset analysis of this trial was presented recently, excluding the patients who underwent pneumonectomy. When that group is excluded, there was a significant advantage for the trimodality group. The final results for this trial have not yet been published. A critical piece of data emerging from both these large trials is the prognostic effect of nodal downstaging on long-term survival of patients undergoing trimodality therapy. In the Intergroup trial, patients who were downstaged to N0 status had an impressive 41% survival at 3 years. Similar results have been reported by other investigators, including a large retrospective study published by our group.10,11 These data are being used to guide further improvements in trimodality therapy under the working hypothesis that if the nodal downstaging rate can be increased, overall outcome will improve. The first attempts to improve nodal downstaging rates centered on escalating the neoadjuvant radiotherapy dose. The group at the University of Maryland has been the most active in this effort. They recently published their results with preoperative therapy doses greater than 59 Gy combined with chemotherapy. Using this dose, the rate of complete nodal clearance was 82%, and the pathologic complete response was 45%. This study demonstrates two critical points. First, it suggests that increased-intensity radiotherapy can increase the nodal downstaging rate. Second, it indicates the importance of the surgical arm because there was a 37% difference between the nodal clearance rate and the pathologic complete response rate, indicating that the primary tumors did not resolve completely with these doses of therapy.12 Even more important, there were no postoperative deaths in this cohort. This finding exemplifies the importance of working with experienced surgeons who perform a large volume of thoracic surgery in terms of limiting postoperative complications. At the Brigham and Women's Hospital we have had similar results. In 74 patients treated with preoperative chemoradiotherapy before pneumonectomy, the 30-day postoperative mortality rate was 4%, again underscoring the importance of surgical experience when treating patients undergoing trimodality therapy for NSCLC.13 |
|
TECHNIQUES OF RADIOTHERAPY Adjuvant Therapy Defining the target volume and determining the treatment dose are the essential elements of planning a safe and effective radiotherapy protocol, whether in the adjuvant or neoadjuvant setting. The volume of treatment for adjuvant radiotherapy has changed little over time. In the initial Lung Cancer Study Group trial, the volume of radiotherapy was defined as the entire mediastinum, ipsilateral bronchial stump, and supraclavicular fossa. The same definition of treatment volume was used for the Keller trial, which delivered large-volume radiotherapy in the postoperative setting, with the exception of the supraclavicular fossa. These same volumes were continued in the recently published RTOG 9705 protocol.5 None of these studies used CT-based planning preoperatively, although a postoperative CT scan was used. Large-volume radiotherapy involving not only the primary tumor but also coverage of the regional lymph nodes in the hilum and mediastinum has been the traditional radiotherapy volume. The problem with large-volume radiotherapy is twofold. First, large-volume radiotherapy limits the ability to deliver higher doses of radiotherapy secondary to toxicity. Conversely, techniques designed to make radiotherapy more effective by limiting toxicity are limited when one treats large volumes. In addition, clinicians have hypothesized that the addition of regional nodal radiotherapy may not improve the outcome of the treatment. The concept of eliminating elective nodal radiotherapy has gained momentum in recent years. Multiple investigators14–16 have shown recently that eliminating elective nodal radiotherapy does not change results for local control after radiotherapy and that isolated elective nodal failure is uncommon. These studies have been done in the setting of definitive radiotherapy or chemoradiotherapy for NSCLC. The question remains whether these same principles apply in the neoadjuvant or adjuvant setting. One could make the argument in either setting that there is more reason to omit elective nodal radiotherapy when surgery has been added because the addition of surgery decreases the likelihood of local recurrence, and therefore, there is even less rationale to include elective nodes. However, the converse also could be true, that patients with aggressive therapy including surgery may benefit most from elective nodal radiotherapy. The standard at Dana Farber Cancer Institute (DFCI)/Brigham and Women's Hospital is to treat only gross disease in the neoadjuvant setting and then only the area at highest risk in the adjuvant setting. |
|
CASE HISTORIES Case 1 A 64-year-old woman was a long-time smoker until she developed chest pain that prompted a chest x-ray showing a left upper lobe mass. Complete staging including CT, FDG/PET, and brain MRI showed no additional evidence of disease (Fig. 74-1). Mediastinoscopy was negative for lymph node spread, and pulmonary function was excellent. The patient underwent left upper lobectomy without complications. Pathology revealed a squamous cell carcinoma of 3.2 cm with negative margins. One station 5 lymph node was found to be involved with disease.
The patient was recommended for adjuvant chemotherapy as well as adjuvant radiotherapy (Fig. 74-2). Initially, the patient received four cycles of carboplatin-paclitaxel chemotherapy. After completing chemotherapy, she received thoracic radiotherapy to 54 Gy in 2 Gy fractions.
Case 2 A 54-year-old woman, nonsmoker, developed a cough. She was evaluated by her medical team and found to have a 5.1-cm right upper lobe mass with enlarged mediastinal lymph nodes (Fig. 74-3). She had no significant symptoms, and her pulmonary function tests were excellent. A mediastinoscopy was positive for adenocarcinoma in a station 4R node. The patient was referred for neoadjuvant therapy.
After consultation (Fig. 74-4), the patient received a 54-Gy dose of thoracic radiotherapy with concurrent cisplatin-etoposide chemotherapy. When neoadjuvant therapy was concluded, restaging showed an excellent response. The patient was taken to the OR on week 12 and underwent a left lower lobectomy. Pathology showed a residual focus of active cancer but no pathologic lymph node involvement.
|
|
EDITOR'S COMMENT The role of adjuvant and neoadjuvant radiation therapy with or without chemotherapy has been examined in the recent intergroup studies. Although XRT may help patients with residual disease or positive margins, there is no clear-cut evidence of benefit in other populations. On the other hand, recent studies have shown a benefit for neoadjuvant chemoradiation in stage III NSCLC, especially when patients who undergo pneumonectomy are excluded. Most importantly, these patients are the best survival candidates if there is a complete pathologic response of mediastinal lymph nodes. –MJK |
|
REFERENCES
|