Introduction
Combination of chemotherapy and external beam radiotherapy is the standard treatment for locally advanced non-small cell lung cancer (NSCLC) patients. Traditionally, external beam radiotherapy for patients with locally advanced NSCLC targets the primary tumor as well as the ipsilateral hilar and mediastinal nodal stations, and sometimes the supraclavicular fossa nodal stations, even if there is no evidence of clinical involvement of all nodal stations. This technique is known as elective nodal irradiation (ENI). However, ENI will limit dose escalation because of pulmonary and esophageal toxicities, and lower radiation dose is unfavorable to local tumor control [
1]. This calls for the omission of ENI and use of another radiation treatment technique, which is called involved field radiotherapy (IFRT). IFRT allows higher radiation dose to the primary tumor with the goal of reducing local failure.
However, the trend of using IFRT has generated concern about the increasing incidence of nodal failure in untreated nodal stations. Some studies reported the crude incidences of elective nodal failure (ENF) were below 10 % when IFRT were applied [
2‐
4], particularly in positron emission tomography-computed tomography (PET-CT) staged patients [
5,
6], while existing evidences for this were weak until now. Now the National Comprehensive Cancer Network (NCCN) guidelines recommend IFRT omitting ENI is category 2A [
7]. We, therefore, carried out a systematic review and meta-analysis to provide more reliable and up-to-date evidence on the incidence of ENF between IFRT and ENI and to identify whether the IFRT was as safe as reported.
Materials and methods
Study design
This was a systematic review carried out in accordance with the Cochrane Collaboration Handbook for Systematic Reviews of Interventions [
8]. The manuscript was arranged according to the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) statement [
9].
Search strategy
In order to achieve the maximum sensitivity, two review authors (RJL and LY) independently searched through the PubMed (1966 to February 2016), Embase (1988 to February 2016), and Cochrane Central Register of Controlled Trials (February 2016, Issue 2) databases to find relevant articles using the following search strategy: (“Non-small cell lung cancer” [all fields]) AND ( (“Involved field radiotherapy” [all fields] OR “Involved field irradiation” [all fields]) OR (“Elective nodal irradiation” [all fields] OR “Selective nodal irradiation” [all fields])). The reference lists of relevant articles were further explored manually.
Inclusion and exclusion criteria
According to the purpose of this meta-analysis, ENF was defined as an uninvolved nodal failure without local failure, that is, any lymph nodes of failure in region only got prophylactic irradiation in the ENI and corresponding region initially uninvolved in the IFRT. Failure in the uninvolved lymph nodes that occurred with distant metastases without local failure was also considered as ENF.
Only studies that investigated the locally advanced NSCLC patients who received 3-dimensional conformal or intensity modulated radiotherapy (concurrent chemotherapy, sequential chemotherapy, or not) with the use of IFRT or ENI were eligible for inclusion in our meta-analysis. Only the photon therapy was allowed. Studies with no ENF data available, with single treatment arm and containing less than 20 patients in each treatment arm were excluded. Expert opinions, reviews, and letters were excluded in case of publication bias. Besides, the studies were limited to English publications in humans. All the articles were filtered by inclusion and exclusion criteria.
Selection of studies and data collection
Three reviewers (SXL, XD and LXY) independently assessed the eligibility of abstracts identified by the search. The full article that appeared to meet the inclusion criteria was retrieved for closer examination. Disagreement over eligibility of a study was resolved by consensus. The meta-analysis was performed to compare IFRT with ENI by estimating the risk ratio (RR) of ENF. Our primary outcome was the incidence of ENF. For each study, we extracted the key information as following: first author’s name, year of publication, number of patients, regimens for intervention and control arms, duration of follow up, as well as the incidence of ENF. According to the inclusion criteria, only the data of locally advanced NSCLC patients were eligible for extraction. If the data extraction of locally advanced was unable to be finished, the study was excluded.
Assessment of the risk of bias in included studies
Two reviewers (LNW and WYL) used the Cochrane risk of bias table [
8] to randomized controlled trial (RCT) and the Newcastle-Ottawa Scale [
10] to cohort study to assess the risk of bias independently. Six domains were employed in the Cochrane Collaboration guidelines including random sequence generation, allocation concealment, blinding of participants or outcome assessment, incomplete outcome data, selective outcome reporting, and other sources of bias. The Newcastle-Ottawa Scale assigned a maximum score of 4 for selection, 2 for comparability, and 3 for outcome. The quality score was ranked as low (≤5 points) or high (≥6 points). As a result, studies ranked as low quality level will be excluded. If necessary, a third reviewer (BSL) would solve disagreements.
Statistical analysis
We calculated the RR for dichotomous data with proper algorithm. RR and 95 % confidence interval (CI) were calculated for each study in an intent-to-treat analysis. Heterogeneity across studies was assessed with a forest plot and the inconsistency statistic (I2). If the heterogeneity was moderate or severe (I2 ≥ 50 %), a random-effects analysis model would be applied; otherwise, the fixed-effects analysis model would be applied. All calculations were performed using Review Manager (version 5.3, the Cochrane Collaboration). A 2-sided p < 0.05 was considered statistically significant. Graphical funnel plot was generated to visually inspect for publication bias.
Discussion
Radiotherapy with ENI has been used since the 2-dimensional radiotherapy era for the improvement of patient’s local control and survival. But emerging evidences from RTOG trials showed that elective irradiation of mediastinal, contralateral hilar and supraclavicular lymph nodes may not be necessary in the treatment of unresectable NSCLC [
17]. In addition, the rate of ENF was really low in clinical practice without ENI [
2‐
6]. Thus, more and more treatment guidelines [
7,
18] do not recommend the use of ENI in NSCLC. We believe that it may be attributed to the following reasons: (1) The development of 3-dimensional conformal and intensity modulated radiotherapy techniques enable dose escalation especially in use of IFRT, and dose escalation is no doubt favorable to local control [
19,
20]. Moreover, IFRT can reduce toxicity by virtue of a decrease in radiation volume. (2) Higher sensitivity of imaging for involved lymph nodes, especially the use of PET-CT [
21], which is supposed to lead to better radiation volume tailoring [
22]. (3) Concurrent or sequential use of chemotherapy with radiotherapy may also help control local diseases. (4) Although the elective nodal regions are not included in IFRT, they are incidentally irradiated with prophylactic doses, and the prophylactic doses may contribute to the low incidence of ENF [
23‐
26]. Before this meta-analysis, despite the widespread abandoning of ENI for NSCLC patients, we still take a conservative approach in our department because existing evidence for this is insufficient. It consists of only three small RCTs and three cohort studies. We believe that meta-analysis can serve as a valuable tool for studying rare and unintended effects of a treatment, and it can extend prior randomized and nonrandomized studies by permitting synthesis of data and providing more stable estimate of effect. Based on the above, we decided to carry out this meta-analysis.
In this meta-analysis, there were three RCTs and three cohort studies. There was no significant difference in incidence of ENF between IFRT and ENI either among RCTs (RR = 1.38, 95 % CI: 0.59–3.25, p = 0.46) or among cohort studies (RR = 0.99, 95 % CI: 0.46–2.10, p = 0.97) assuming a fixed-effects model. The Cochran’s Q test resulted in a p = 0.48 and a p = 0.92 respectively, and the quantity I
2 both were 0 %, indicating that the studies were homogeneous. So we performed a combined analysis of RCTs and cohort studies, and there was also no significant difference in incidence of ENF between IFRT and ENI assuming a fixed-effects model (RR = 1.15, 95 % CI: 0.65–2.01, p = 0.64). The quantity I
2 was also 0 %, indicating that there was no evidence of heterogeneity among all the included studies.
Our meta-analysis demonstrated that the difference in incidence of ENF was not significant between IFRT and ENI. This meta-analysis had two main strengths. Firstly, it summarized the highest quality data available comparing IFRT with ENI for locally advanced NSCLC patients until now. Secondly, there was no heterogeneity in the results of the included studies. Nevertheless, this meta-analysis had several limitations. Firstly, the quality of the data was different. For example, this meta-analysis included three cohort studies, which lacked the experimental random allocation of the intervention in contrast to RCTs. Secondly, we compared outcomes across studies but not within studies, so the balance of baseline characteristics between the treatment groups might be neglected. For example, the IFRT group received higher radiation doses than the ENI group in most studies, and statistical analysis showed that the difference was significant in some studies [
12,
14]. Thirdly, the follow-up time was relatively short in some studies [
13], and some patients might have not suffered from ENF or have died before they suffered from ENF.
Conclusion
To the best of our knowledge, this is the first meta-analysis of published studies to evaluate the difference in incidence of ENF between IFRT and ENI. It provides more reliable and stable evidence that there is no significant difference in incidence of ENF between IFRT and ENI. IFRT will inevitably take the place of ENI.
Acknowledgements
None declared.
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (
http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (
http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.