Postoperative radiotherapy in non-small cell lung cancer: systematic review and meta-analysis of individual patient data from nine randomized controlled trials

Author: PORT Meta-analysis Trialists Group
Source:Lancet. 1998 Jul 25;352(9124):257-63

Background

Lung cancer remains the leading cause of cancer death in the United States. Surgery is the treatment of choice for non-small cell lung cancer (NSCLC); however, even in the case of completely resected disease, 5-year survival is only 40%. In an effort to improve local control, radiation therapy is often given post-operatively. A number of randomized trials have been conducted examining the effect of post-operative radiation therapy with conflicting results. Many of these studies were limited by inadequately small patient sample sizes. A meta-analysis was performed to improve the statistical power of these trials and to see if post-operative radiation offered an improvement over surgery alone.

Materials and Methods

A review of both published and unpublished studies comparing surgery alone for NSCLC to surgery with post-operative radiation therapy. All trials included were required to be randomized, include only patients who had a complete resection, and started after 1965. Trials that utilized orthovoltage radiation therapy were excluded. Update information was obtained on survival, recurrence, and date of last follow-up of all patients included in the study. Previously excluded patients were included in the meta-analysis. All analysis was performed on an intent-to-treat basis.

Results

The initial literature search revealed 15 potentially eligible trials. Six of these trials were excluded. The analysis was therefore conducted on 9 trials included a total of 2128 patients. Overall, the majority of patients were male (1795/2128) and early stage (stage I 562/2128, stage II 718/2128). Radiation doses ranged from 30-60 Gy in 10-30 fractions and included a wide variety of radiation techniques. Median follow-up for the surviving patients was 3.9 years. Post-operative radiation was found to decrease overall survival with a hazard ratio of 1.21 (p=0.001). This was equivalent to an absolute decrease in 2-year OS from 55% to 48%. Local recurrence-free survival was also decreased with the addition of radiation with a hazard ratio of 1.16 (p=0.005). The hazard ratio for distant recurrence-free survival was 1.16 (p=0.007) and for overall recurrence-free survival was 1.13 (p=0.018), both in favor of surgery. On subgroup analysis, the detrimental effect of radiation was not seen in patients with stage III disease or with patients with N1 or N2 disease.

Conclusions

Although the results of this study would seem to indicate that post-operative radiation therapy is detrimental for completely resected NSCLC, there are several significant flaws to this analysis. The majority of patients included in the study were stage I or II patients for whom it is already known that radiation therapy is unnecessary in completely resected patients. Inclusion of such a large number of these patients undoubtedly resulted in the worsened outcomes for all patients in the analysis. Also, the studies included in the analysis were all initiated before 1990, and as far back as 1966. The radiation techniques and machines used were all older and resulted in the inclusion of a much higher amount of normal tissue than in techniques that are used today. The outdated treatments certainly contributed to the increased detrimental effect seen in patients treated with radiation. In addition, other published studies including at least one which showed a benefit to treatment with radiation were inexplicably excluded from the study. Overall, this meta-analysis does not reveal anything new about post-operative radiation beyond the already established fact that early-stage, completely resected NSCLC does not need post-operative radiation.