Background
Completely resected non-small cell lung cancer (NSCLC) patients with pathologically confirmed N2 disease are considered to be a heterogeneous population [
1], showing 5-year survival rates ranging from 10% to 30% [
2]. Systemic recurrence following surgery is one of the major problems in stage IIIA(N2) patients, and the use of postoperative chemotherapy (POCT) in stage IIIA disease prolongs survival [
3]. The value of postoperative radiotherapy (PORT) for completely resected NSCLC remains controversial, as the effect on survival has been inconclusive [
4-
6]. A meta-analysis of PORT published in 1998 [
4] described a relative increase of the risk of death with the addition of PORT for completely resected NSCLC. This detrimental effect was evident among patients who exhibited no mediastinal involvement, whereas in patients with stage III and pN2 disease, a slight increase in survival was detected, although the difference was not statistically significant. Similar results were found when this meta-analysis was updated in 2005 [
5]. Recently, several large retrospective studies and a recently published randomized trial have provided evidence of the possible benefit of PORT in completely resected stage IIIA(N2) patients [
7-
13].
Several limitations of the previous prospective studies included in the PORT meta-analysis have been recognized, including the use of suboptimal radiation techniques and wide irradiation portals. The quality of radiation therapy (RT) was inferior to what is now available, with patients being currently treated using linear accelerators and the three-dimensional conformal radiotherapy (3D-CRT) technique. The irradiation fields employed in most trials have often been large and varying (typically including the entire mediastinum and occasionally the supraclavicular region or contralateral hilum). It has been hypothesized that the toxicity reported in the meta-analysis was related to large field sizes and the use of obsolete radiotherapy techniques [
14-
16].
Currently, growing evidence suggests that PORT administered using the modern 3D-CRT technique has a favourable effect on the survival of patients with pN2 disease [
13,
17]. However, there exists significant heterogeneity within the reported studies with respect to the irradiation fields employed for PORT because there is no clear consensus on the definition of the extent of the clinical target volume (CTV) [
9-
13]. To the best of our knowledge, there is no solid evidence available for the PORT CTV designs used in the currently published prospective trial [
13] and ongoing multi-centre phase III studies. Therefore, we designed a patterns-of-failure study after complete surgery in resected pN2 disease to evaluate the rationale of the proposed PORT CTVs based on the most likely sites of nodal failure, and the institutional standard CTV delineation for PORT was developed in our hospital [
18].
The aim of the present study was to explore the clinical efficacy of PORT administered using 3D-CRT techniques and the institutional standard CTV delineation guideline in our hospital for patients with completely resected pathologic stage IIIA(N2) NSCLC, in attempt to provide evidence for future phase III clinical trials.
Discussion
A growing number of more recent publications have bolstered the use of modern PORT for completely resected stage IIIA(N2) NSCLC [
26,
27]. However, these reports did not contain detailed information regarding RT, especially concerning the PORT treatment volume. To our knowledge, with the introduction of our institutional standard PORT CTV delineation guideline [
18], this is the first report on the clinical efficacy of modern PORT, administered using the 3D-CRT technique and the institutional standard CTV delineation design based on the patterns of local-regional failure data, for completely resected stage IIIA(N2) NSCLC. In the present study, all of the patients included in the PORT group in the analysis were treated using a linear accelerator with the 3D-CRT technique. Moreover, the underlying strength of this study lies in the institutional standard PORT CTV delineation as applied herein [
18]. Potential advantages include the following: (
1) the PORT CTV delineation based on the patterns of failure data might be more reasonable and appropriate; (
2) this CTV delineation guideline strictly defines LNSs included in the CTV, thus making it more consistent and reproducible in clinical practice; and (
3) the design of treatment fields tailored to the area most-at-risk for recurrence will reduce the irradiation volume (not including LNSs 1, 3A, 3P, 8, and 9 in our institutional CTV delineation [
18]). Thus, it would be of value to assess the efficacy of PORT using the 3D-CRT technique and this institutional standard CTV delineation guideline for completely resected stage IIIA(N2) NSCLC patients.
Prior studies reporting the outcomes of completely resected pN2 patients are outlined in Table
4. We found that the patients treated in both groups analysed in our study (5-yr OS, 57.5% for PORT and 35.1% for non-PORT) yielded superior OS compared with those in previously reported studies [
7-
13]. The following may be possible explanations for why our survival results in both treatment groups appear to be better than their corresponding historical controls. First, improved survival might be secondary to better patient selection, as a homogeneous group of patients who underwent complete resection of NSCLC and systematic nodal assessment was selected in our study. Second, this difference might be due to the inclusion of a majority of cases receiving adjuvant chemotherapy in our study (all patients in the PORT group and 85% in the non-PORT group received POCT). The ANITA study [
8] also demonstrated an advantage of adjuvant chemoradiotherapy in completely resected patients with pN2 disease. It was reported that 5-yr OS was 47.4% under the use of adjuvant chemoradiotherapy, which was relatively comparable to the results of our analysis. Third, the improvement of survival observed in the PORT group likely depends on the application of our institutional CTV delineation guideline, leading to relatively small-sized PORT fields tailored to the area most-at-risk for recurrence after surgery, with good consistency in clinical practice. Miles et al. [
14] attempted to estimate the field size dependence of RT-induced mortality and tumour control in the postoperative setting. It has been shown that RT-induced mortality is strongly dependent on the field size, which may partly offset the OS benefit afforded by PORT. The incongruity between an improvement in local control and a decrease in survival may have been secondary to RT-induced complications.
Table 4
Rates of overall survival and locoregional recurrence rates after complete resection in pN2 NSCLC
| 2006 | pN2 | 1987 | 20 | 27 | NS |
| 2008 | pN2 | 106(observation) | 16.6 | 21.3 | 42.1* | 22.1* |
118(chemotherapy) | 34 | 47.4 | 25.7* | 14.6* |
| 2010 | pN2 | 183 | 22.2 | 30.5 | 66 | 27 |
| 2010 | pN2 | 175 | NS | 44 | 20 |
| 2011 | IIIA-N2 | 221 | 30.6 | 36.6 | 53 | 36 |
| 2013 | pN2 | 66 | NS | 37 | NS | 28 |
| 2014 | IIIA-N2 | 135 | 27.5 | 37.9 | 49.3* | 27.3* |
Our results showed that the locoregional recurrence (LRR) rate was reduced from 33.6% to 8.1% with the administration of PORT (P < 0.001), and 5-yr OS for patients who received PORT was 57.5%, which was obviously higher than in patients not receiving PORT (5-yr OS, 35.1%). A similar reduction in LRR rates with a survival benefit was reported by previous retrospective series [
7-
9,
11]. Our findings are also congruent with a recent meta-analysis study [
17] echoing a similar increase in local control and OS for completely resected stage IIIA(N2) NSCLC. In this meta-analysis, it was reported that the application of PORT using modern techniques was estimated to reduce the LRF rate to 10% and increase absolute 5-year OS by 13% [
17].
The type of disease failure pattern predominated by DM is also quite similar to the results reported in other trials [
9-
12]. The patterns of failure outcomes after surgery with or without PORT reported herein are in keeping with the clinical efficacy of PORT as well, demonstrating that PORT is able to reduce locoregional recurrences, but not in supraclavicular nodes or distant metastases. Of note, distant metastases remain more frequent in completely resected pIIIA(N2) disease, despite the addition of PORT, thereby encouraging further exploration. It is possible that patients with NSCLC have occult systemic disease, especially in pN2 stages, and that PORT alone is not adequate to confer a survival benefit without effective systemic control by POCT. In the light of our data, it can be concluded that the major problem for this patient population remains the high risk of distant metastases, indicating the necessity for the development of optimal adjuvant or systemic treatment strategies.
The current study is observational in nature and as a result cannot prove a direct causal relationship between PORT and prolonged survival. However, this link is highly plausible for the following three major reasons. First, it was demonstrated that PORT was independently associated with improved OS according to the multivariate analysis. Second, although the baseline data were not balanced in the two treatment groups, these baseline imbalances might bias our results towards either the PORT or non-PORT group. In our study group, patients whose tumour characteristics (>4 positive lymph nodes and LNR >20%) were perceived to be worse might have been referred for PORT more often. Third, the fact that EGFR-TKI therapy subsequently administered for relapse or progressive disease might obscure improved survival should be considered and accounted for in the evaluation of OS endpoints in current clinical practice. Thus, we attempted to control for the disparity in subsequent EGFR-TKI therapy between the two treatment groups by applying a censoring approach at the OS estimation [
24,
25]. This is one of the main differences between our study and most other studies, including a recently published small randomized trial conducted in China [
13]. In consideration of these factors, the application of PORT using the 3D-CRT technique and our institutional standard CTV delineation guideline might confer a significant survival advantage for completely resected stage IIIA(N2) patients based on our present descriptive data.
In the present study, several clinical parameters (current/heavy ex-smoker, cN2 status and LNR >20%) were identified as indicators of a high risk of LRF after complete surgery in resected pN2 patients. These findings of this study are in line with the results of previous studies on the prognosis of completely resected IIIA(N2) patients [
1,
28-
30]. The identification of high-risk prognostic factors for LRF after complete surgery could be applied to individualized clinical decision making (as completely resected patients with pN2 disease can have different prognoses) and in stratifying the randomization applied in clinical trials.
We acknowledge that there are some limitations inherent to this retrospective study, such as selection bias, missing data and inconsistent follow-up intervals. Comparisons between the PORT and non-PORT groups have been hampered by the retrospective nature of the study and difficulty in controlling for confounding variables. The two populations were not well-balanced with respect to several clinicopathologic factors. In fact, there existed both favourable and adverse prognostic confounders that may have biased the results towards either the PORT or non-PORT group. Furthermore, we could not differentiate which factors among these potential confounders presented larger values and significantly contributed to the outcomes presented herein. Another important limitation was that the subsequent EGFR-TKI therapy administered for relapse or progressive disease after complete surgery was not strictly controlled for and was not well-balanced in the two groups. It was noted that more patients in the PORT group received EGFR-TKI than in the non-PORT group, which might result in a bias towards improved survival results. However, we have taken appropriate steps, including statistical considerations (censoring the analysis at the time of TKI initiation in the OS estimation), in an attempt to control for the potential impact of this disparity on OS. Finally, our study is limited by the relatively small number of patients analysed in the PORT group. Therefore, there is still a need for a prospective study to validate the efficacy of 3D-conformal PORT in accordance with our institutional standard CTV delineation guideline.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made.
The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
XLF designed the study and guaranteed the integrity of the entire study. JQX, YWZ, and HQC contributed to data acquisition. WF collected the data and performed the statistical analyses. XWC, ZFZ, HJY, and XLF undertook the data analyses and interpretation. WF and QZ wrote the report. All authors have read and approved the final manuscript.
Wen Feng and Qin Zhang are co-first authors.