Sublobar resection with intraoperative brachytherapy versus sublobar resection alone for early-stage non-small-cell lung cancer: a meta-analysis

Abstract

 

OBJECTIVES

The purpose of this study was to compare the clinical outcomes for sublobar resection (SR) or SR plus intraoperative brachytherapy (SRB) for clinical stage I non-small-cell lung cancer.

METHODS

A systematic search was performed in the EMBASE, PubMed and Cochrane Library databases to identify related studies comparing SR to SRB. Data were collected on local recurrence (LR) as a primary outcome and regional or distant recurrence, overall survival and disease-free survival (DFS) as secondary outcomes. Meta-analysis was carried out using Stata 12.0.

RESULTS

A total of 476 patients received SRB, and 617 received SR across 5 studies. Meta-analysis of LR, regional or distant recurrence, overall survival and disease-free survival rates showed no significant difference between SRB and SR groups. However, when biologically effective dose (BED) was >100 Gy, LR rate was lower in the SRB group than in the SR group (Relative risk [RR] = 0.143, 95% confidence interval [CI]: 0.051-0.397) (p < 0.001). When BED was <100 Gy, no significant difference was found in LR rate between SRB and SR groups (SRB versus SR: RR = 1.132, 95%CI: 0.704-1.821) (p = 0.608).

CONCLUSIONS

Intraoperative brachytherapy was not associated with reduced risk of regional or distant metastasis or improved outcomes for patients with clinical stage I non-small-cell lung cancer; however, it might reduce the LR rate when BED was >100 Gy.

INTRODUCTION

Lung cancer is among the most common tumours worldwide and is the main cause of cancer-related mortality in humans [1–3]. Non-small-cell lung cancer (NSCLC) already accounts for over 80% of all lung cancers, of which ∼15–20% of cases presented with early-stage lung cancer [4]. Formal anatomic pulmonary lobectomy with mediastinal lymph node evaluation has been the standard procedure for early-stage NSCLC. Sublobar resection (SR) therapy including wedge and segmental resection may be used to treat patients who are not appropriate candidates for standard lobectomy. Yet, a major issue with SR for NSCLC is the higher local recurrence (LR) rates compared to standard lobectomy [5]. Thus, this method is typically preferable for patients with increased risks for lobectomy including advanced age or significant impairment in cardiopulmonary reserve.

To compensate for less extensive surgical resection, numerous scholars used radioactive Iodine-125 (¹²⁵I) or Cesium-131 (¹³¹Cs) seed implanted within a polyglycolic mesh to help reduce LR of the tumour in such patients. Such adjuvant intraoperative brachytherapy techniques allow the delivery of high doses of radiation to the high-risk staple line and tumour bed sustainably with the unique advantages of more conformal potent dose of radiation to the target with minimized radiation toxicity, which are resulted from a decrease in the volume of normal lung exposed to radiation [6]. Voynov et al. evaluated the delivery dose of 100–120 Gy with the use of ¹²⁵I seed to the staple line in 110 early NSCLC cases. The results showed the 5-year local control (LC) rate of ∼90% for all patients, and loco-regional control of over 60% [7]. Chen et al. [8] also reported satisfactory LC rates in early NSCLC using SR plus intraoperative brachytherapy (SRB) with a dose of 100 to 120 Gy. Nevertheless, several scholars had come to opposite conclusions. They had described satisfactory results with SR alone and suggested that intraoperative ¹²⁵I seed implantation brachytherapy was not necessary as a local adjuvant therapy. The efficacy of the intraoperative brachytherapy to reduce the incidence rate of LR in early NSCLC still remains controversial.

Therefore, in the present study, we aimed to explore if the addition of brachytherapy to SR in early-stage NSCLC could improve clinical outcomes by conducting meta-analysis. Also, to our knowledge, the prescribed doses (PD) among previous studies were different. We assumed that such differences might have impacts on LC when SRB was performed. Here, we reported the results of this meta-analysis, which had been registered in the PROSPERO database with the registration number CRD42020203427.

MATERIALS AND METHODS

This meta-analysis was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses reporting guidance. Literature retrieval was performed in EMBASE, PubMed and Cochrane Library databases up to August 2020. Several keywords and combinations were applied to search for relevant studies: (sublobar resection OR wedge resection OR segmentectomy OR limited resection) AND (brachytherapy OR radiotherapy, interstitial) AND (NSCLC OR Carcinomas, Non-Small-Cell Lung OR Lung Carcinoma, Non-Small-Cell). No language or publishing period restrictions was applied.

Selection criteria

The outcome measures were LR, regional or distant recurrence (RDR), overall survival (OS) and disease-free survival (DFS) in comparison of SR versus SRB for the treatment of early NSCLC. The inclusion criteria were: (i) clinical trials or retrospective studies treating stage I NSCLC with SRB and (ii) provision of at least one of the outcome measures. Reviews, case reports, letters, editorials and cell culture or animal studies were excluded.

Data extraction and quality assessment

Two reviewers (Enli Chen and Chenfei Jia) independently performed a literature search, screened results and carried out the data extraction, and 2 reviewers solved any disagreements by consultation with a third analyst (Xueya Min). The following important data were extracted from each included research: publication details; study design; clinical data; sample size; and oncological outcomes including LR, RDR, OS and DFS rates.

The Cochrane Risk of Bias assessment tool [9] was employed to assess methodological quality of the included studies on RCTs, considering random sequence generation, allocation concealment, blinding of participants and investigators, incomplete outcome data, selective reporting of outcomes, and other bias.

We applied the Newcastle–Ottawa Scale for methodological quality assessment as most of the included studies were retrospective [10]. Two researchers (Enli Chen and Juan Wang) performed the scoring independently and debated until a full agreement was reached. Studies with a score of >7 were considered as high quality, 4–6 as moderate quality and <4 as low quality.

Statistical analysis

The biologically effective dose (BED) formula was used; BED = D [1 + R0/(μ + λ)(α/β)] [11]. D is the total dose for brachytherapy; R0 is the initial dose rate (R0 = D × λ); μ is a tissue repair constant (μ  =  0.5 h⁻¹); λ is a decay constant, that is λ = ln(2)/half-life; α/β = 3 for lung tissue or late responding tissue, 10 for tumour tissue. Using this formula, the 100 Gy for ¹²⁵I seed brachytherapy is equivalent to a BED of ∼100 Gy; the 80 Gy for ¹³¹Cs seed brachytherapy is equivalent to a BED of ∼81 Gy.

Meta-analyses and forest plots were prepared in Stata 12.0. The relative risk (RR) with 95% confidence interval (CI) was used for dichotomous outcome in the inverse variance method. For time-to-event data such as OS and DFS, hazard ratio (HR) with 95% CI was used in the inverse variance method. If HR was unavailable in the included studies, we estimated HR and corresponding standard errors from survival curves as previously described [12]. Heterogeneity analysis was carried out with I² test [13]. P-values <0.1 in Q-test or I² > 50% in I² statistic was defined as significant heterogeneity. We took a fixed effect approach once I² < 50% and P > 0.1. Otherwise, a random-effect method was adopted [14]. LR rate was analysed as a primary outcome, and a sensitivity analysis was performed to test the stability and reliability of the results. Subgroup analysis was carried out to explore the source of heterogeneity. PD of 100 Gy was used as cut-off value for subgroup analysis due to the availability of data in included studies and the references from external radiation therapy. P < 0.05 was considered to indicate statistical significance. Publication bias tests were not carried out because <10 researches were included in the present study.

RESULTS

In total, 124 relevant researches were originally retrieved. After removing case reports, reviews, editorials, letters and consensus, 33 researches were then subjected to abstract assessment. By carefully reading further the abstracts, 21 relevant articles were retained for further assessment. Ultimately, 5 researches with 1093 patients were incorporated into the final analysis, with 476 and 617 patients who underwent SRB and SR, respectively [15–19]. The process of literature retrieval and exclusion is summarized in Fig. 1.

Figure 1:

Flow chart of publication search and selection.

Among 5 studies, 4 were retrospective researches, and 1 included prospective designs [18]. Of them, 4 studies focused on SRB with radioactive ¹²⁵I seed [15–18], and 1 with ¹³¹Cs [19]. The total radiation dose ranged from 80 to 120 Gy. The median follow-up period was 17.5–52.6 months. Comparisons of LR were carried out in these researches; 2 of these (40%) showed lower LR for SRB (P < 0.05) [15, 16], while 3 did not show a statistical difference (P > 0.05) [17–19]. Among the 5 studies that provided RDR, OS and DFS comparisons, no statistical difference was reported. Table 1 summarizes the general information of the included studies.

Table 1:

Characteristics of the included studies

Author, year	Source	Study design	Quality	Seed (dose, BED group)	Number of patients (SRB/SR)	Tumour stage and nodal status	Surgical margins	Measured margins	Intraoperative brachytherapy used cases	Sublobar resection	Follow-up, months
Santos, 2003	Surgery	Retrospective	High	125I (100–120 Gy, >100 Gy)	101/102	SR: pathological stage IA (T1N0M0); SRB: stage I (T1–2N0M0)	Clear grossly and on frozen section	NA	The more recent SR cases	Wedge resection	Median: SR: VATS: 24, open: 29; SRB: 18
Fernando, 2005	J Thorac Cardiovasc Surg	Retrospective	High	125I (100–120 Gy, >100 Gy)	60/64	Pathological T1N0	Pathologically confirmed clear surgical margins	NA	The more recent SR cases	Segmental or wedge resections	Mean: 34.5
Landreneau, 2014	Surgery	Retrospective	High	125I (80–100 Gy, ≤100 Gy)	155/214	All were clinical stage I; pathological stage 1A: 207; 1B: 124; 2A or above: 37	Clear resection margins on frozen section	SR: 17.5 mm (average); SRB: 17.2 mm	Based on the individual surgeon’s preference	Segmentectomy	Mean: 32.9
Fernando, 2014	J Clin Oncol	RCT	Moderate	125I (100 Gy, ≤100 Gy)	108/114	SR: T1N0; SRB: T1N0:104, T2N0:3, T3N0: 1	Only 14 patients had positive staple line cytology	NA	Random	Segmental or wedge resections	Median: 4.38 years
Parashar, 2015	Brachytherapy	Retrospective	High	131Cs (80 Gy, ≤100 Gy)	52/123	Stage I	Brachytherapy was only added to those with suspected close or positive margins	A margin equal to the diameter of the tumour or at least a 1-cm margin	Brachytherapy was determined by surgeon’s assessment of adequacy of the surgical margin	Wedge resection	Median: 17.5

BED: biologically effective dose; ¹³¹Cs: Cesium-131; ¹²⁵I: Iodine-125; SR: sublobar resection; SRB: sublobar resection plus brachytherapy; VATS: video-assisted thoracoscopy; NA = not available.

The overall risks of all types of biases in the Randomized Controlled Trials (RCT) [18] were as follows: random sequence generation: unclear risk; allocation concealment: unclear risk; blinding of participants and personnel: high risk; blinding of outcome assessment: unclear risk; incomplete outcome data and selective reporting: low risk; and other bias: nothing. Stratification analysis was performed on 3 of the 4 retrospective studies, so the Newcastle–Ottawa Scale score of the studies was 1 in terms of comparability. Finally, 4 retrospective studies [15–17, 19] were of high quality with a score of 7 (as shown inSupplementary Material, Table S1).

The result of meta-analysis on LR of the 5 included studies did not show any statistical difference between the groups (RR = 0.512, 95% CI: 0.202–1.298) (P = 0.159) (as shown in Fig. 2). Significant heterogeneities were reported. Sensitive analysis showed that the result of meta-analysis regarding LR was stable (as shown in Fig. 3). We performed a subgroup analysis according to the design of studies, and the result indicated for both RCT and retrospective studies, LR did not show any statistical difference between the groups (RR = 1.188, 0.370; 95% CI: 0.639–2.207, 0.106–1.296) (P = 0.587, 0.120). We conducted a further subgroup analysis according to BED. The result of the meta-analysis on LR of the studies with BED > 100 Gy [15, 16] showed an RR of 0.143 (95% CI, 0.051–0.397, P < 0.001), favouring the SRB arms; however, no significant differences were found between the arms when BED was <100 Gy [16–18] (RR = 1.132, 95% CI: 0.704–1.821, P = 0.608). No significant heterogeneity was recorded (as shown in Fig. 4). We performed another subgroup analysis according to different surgical procedures (wedge or segmental resection). The results of the meta-analysis on LR of the studies with wedge resection and BED > 100 Gy [15] revealed an RR of 0.11 (95% CI, 0.03–0.44, P = 0.002), favouring the SRB groups; however, no significant differences were found between the groups when wedge resection with BED ≤100 Gy or segmentectomy was applied [17–19] (RR = 0.85, 1.52; 95% CI: 0.44–1.68, 0.78-2.97) (P = 0.648, 0.219). The results of the meta-analysis are detailed in Tables 2 and 3.

Figure 2:

Forest plots of meta-analysis regarding local recurrence.

Figure 3:

Leave-one-out sensitive analysis regarding local recurrence.

Figure 4:

Subgroup analysis for biologically effective dose.

Table 2:

Outcomes summary of meta-analysis

	Models	I 2 (%)	Number of included studies	Number of included patients	RR/HR	95% CI	P-value
LR	Random	73.1	5	1093	0.512	0.202–1.298	0.159
	Fixed	73.1	5	1093	0.785	0.510–1.207	0.270
RDR	Fixed	0	4	969	0.903	0.690–1.180	0.454
OS	Fixed	0	2	330	1.07a	0.695–1.647	0.757
1-Year survival	Fixed	0	3	602	0.750	0.413–1.364	0.347
2-Year survival	Fixed	0	3	602	0.747	0.522–1.067	0.109
3-Year survival	Fixed	0	3	602	0.907	0.684–1.203	0.498
4-Year survival	Fixed	0	3	602	0.916	0.723–1.161	0.468
5-Year survival	Fixed	0	2	399	1.136	0.832–1.551	0.423
5-Year DFS	Fixed	0	2	544	1.012	0.811–1.262	0.916

CI: confidence interval; DFS: disease-free survival; HR: hazard ratio; LR: local recurrence; OS: overall survival; RDR: regional or distant recurrence; RR: relative risk.

^aHR value.

Table 3:

Outcomes summary of subgroup analysis for local recurrence

	Subgroup	I 2 (%)	Number of studies	Number of patients	RR	95% CI	P-value	P-value between subgroups
Study design	Prospective studies	–	1	222	1.188	0.639–2.207	0.587	0.068
	Retrospective studies	74.0%	4	871	0.370	0.106–1.296	0.120
BED	>100 Gy	0	2	327	0.143	0.051–0.397	<0.001	<0.001
	≤100 Gy	0	3	766	1.132	0.704–1.821	0.608
Sublobar resection modalities	Wedge resection	73.1%	3	533	0.41	0.11–1.56	0.189	0.039
	Segmental resection	0	2	436	1.52	0.78–2.97	0.219
BED in wedge resection group	>100 Gy	–	1	203	0.11	0.03–0.44	0.002	0.010
	≤100 Gy	0	2	330	0.85	0.44–1.68	0.648

BED: biologically effective dose; CI: confidence interval; RR: relative risk.

The result of meta-analysis on RDR showed no statistical difference between the groups (RR = 0.903, 95% CI: 0.69–1.18) (P = 0.454). Significant heterogeneities were not reported. The results of the meta-analysis of 1-, 2-, 3-, 4- and 5-year OS among the studies were not significantly different between the arms (RR = 0.750, 0.747, 0.907, 0.916 and 1.136, respectively; 95% CI: 0.413–1.364, 0.522–1.067, 0.684–1.203, 0.723–1.161 and 0.832–1.551, respectively) (P = 0.347, 0.109, 0.498, 0.468 and 0.423, respectively). The result of meta-analysis on OS showed no statistical difference between the groups (HR = 1.07, 95% CI: 0.695–1.647) (P = 0.757) (as shown in Fig. 5). The result of the meta-analysis of 5-year DFS did not show a statistical difference between the groups (RR = 1.012, 95% CI: 0.811–1.262) (P = 0.916). There were no obvious heterogeneities reported within each of the meta-analysis. Table 2 shows the results of the meta-analysis.

Figure 5:

Forest plots of meta-analysis regarding overall survival.

DISCUSSION

SR has been typically used an alternative for patients with early-stage NSCLC at high risk. However, a major issue concerning SR is the higher incidence rate of LR than formal lobectomy. Currently, 2 main approaches are available to decrease LR. One is to perform segmental resection instead of wedge resection; the other is to expand the resection margin to 1 or 2 cm or to achieve a margin-to-tumour ratio of ≥1 in order to obtain pathologically confirmed clear surgical margins [20]. However, some patients fail to tolerate the prolonged one-lung ventilation required to perform anatomic segmental resection or to achieve the safer surgical margin. Recently, SRB has become a widely used treatment option for these patients. The radioactive ¹²⁵I and ¹³¹Cs seeds are often used in intraoperative brachytherapy. ¹²⁵I seed is a low-energy radioactive nuclide which can continuously and slowly release low-dose X-rays and gamma rays acting on the DNA double strands of tumour cells to cause apoptosis of tumour cell and destroy their proliferative capacity. Meanwhile, because radiation penetration is extremely weak at a length of 17 mm, it can reduce the dose around the target area rapidly. The damage to surrounding normal tissues and organs is minimal while delivering a high dose to the target area. The effective duration of tumour exposure can be up to 200 days, which increases the effectiveness of radiotherapy. ¹³¹Cs seed has a higher dose rate than ¹²⁵I seed, which offers a shorter half-life (9.7 days) and faster delivery of the total radiation dose (delivers over 90% of total dose in less than 33 days) than ¹²⁵I seed. Two methods of brachytherapy were reported in previous studies [15–19]. Some scholars designed polyglactin sutures which contained ¹²⁵I seed, and placed the sutures parallel to and 5 mm away from the surgical staple line. Then, the ¹²⁵I seed sutures were fixed to the lung surface with 3.0 silk sutures placed 1–2 cm apart. Also, some scholars designed a piece of vicryl mesh where ¹²⁵I sutures were placed at 1-cm intervals. They placed such vicryl mesh implants over the surgical staple line to help reduce LR of the tumour. To date, however, the role of SRB has been still controversial.

In this meta-analysis, we found an equivalent LR, RDR, OS and DFS rates between SRB and SR for patients with early NSCLC who are ineligible for standard lobectomy. There was no significant heterogeneity in the analysis for RDR, OS and DFS, which indicated SRB was not associated with reduced risk of regional or distant metastasis and improved survival for patients with clinical stage I NSCLC. SRB was not superior to SR in these respects. However, the analysis for LR revealed significant heterogeneity. We performed the subgroup analysis to explore the source of heterogeneity, and the results indicated lower heterogeneity in each group, as expected. Interestingly, our subgroup analysis reported that intraoperative brachytherapy could reduce LR rate when the BED was >100 Gy.

In 2003, Santos et al. [15] analysed clinical data for the high-risk patients with early-stage NSCLC with limited cardiopulmonary reserve treated by SR or SRB with a dose of 100–120 Gy. The results showed that there was no improvement in either RDR or OS with the addition of intraoperative ¹²⁵I seed brachytherapy; yet, the LR was decreased from 18.6% to 2% (P = 0.001) with SRB. In 2005, Fernando et al. [16] assessed 291 high-risk patients with early-stage NSCLC, 60 of whom underwent SRB. In this subset of patients, the addition of ¹²⁵I seed brachytherapy was related to a reduced LR rate from 17.2% to 3.3%, which was similar to the prior study. However, in a randomized controlled trial [18], high-risk patients with early-stage NSCLC were assigned randomly to receive SR or SRB. Five-year LR rates were 16.7% and 14.0% with SRB and SR, respectively (P = 0.59). Nevertheless, there was an obvious tendency favouring SRB, of which 14 patients had positive staple line cytology, but the statistically significant differences were not found. The authors stated that the reason for this was closer attention to parenchymal margins by researchers in the trial. It was mystifying that although the surgical margins were all pathologically confirmed clear in the study of Santos and Fernando, the LR could be still reduced with SRB compared to SR alone. Thus, the reasons given by the author of above randomized controlled trial seem a bit of a stretch. Taken together with our results, we speculated that the different PD among such studies might have impacts on LR.

D'Amato et al. [21] reported 14 patients with NSCLC with limited cardiopulmonary reserve underwent video-assisted thoracoscopic (VATS) wedge resection. Pathologically confirmed clear surgical margins were obtained. A mesh containing ¹²⁵I seeds with a radiation dose of 100–120 Gy was used to margins following wedge resection. No LR was reported at mean follow-up of 7 months. A study with a larger sample size by Voynov et al. [7] showed that 110 patients with early-stage NSCLC underwent SRB prescribed to a dose of 100–120 Gy, with a median follow-up of 11 months, and recurrence was 3.6% (4/110) within the radiation volume. However, among our included studies in which the PD was <100 Gy, Landreneau [17] assessed 369 stage I NSCLC cases who underwent segmentectomy with (n = 155) or without (n = 214) ¹²⁵I seed brachytherapy prescribed to a dose of 80–100 Gy; the overall LR rate was 5.4% (SRB: 6.4% versus SR: 4.6%, P = 0.49). The LR rate in the SRB group (6.4%) was higher than that (3.6%) in Voynov et al.’s study. Finally, the randomized controlled trial by Fernando [18] also showed a higher LR rate in SRB (16.7%, 18/108) with the PD of <100 Gy. The aforementioned studies further confirmed that there was a dose–response relationship between BED and LR.

Hiraoka et al. [22] retrospectively analysed 241 patients with early-stage NSCLC from 13 Japanese institutions who were treated with stereotactic body radiation therapy (SBRT). The LR rate reached 20% when BED was <100 Gy, and merely 6.5% when it was over 100 Gy. The 3-year survival rate was ∼42% when it was <100 Gy, and reached 46% when it was over l00 Gy. Guckenberger et al. [23] analysed 124 patients with 159 pulmonary lesions undergoing SBRT. The results showed the radiation dose to the tumour target in view of four-dimensional dose computation was closely related to LC of tumour: LC rates at 36 months were 62% and 89% for BED <100 Gy and BED >100 Gy, respectively. Onishi et al. showed in a large-scale retrospective study that BED of over 100 Gy could achieve better LC and OS than BED <100 Gy. The results from these studies supported this conclusion that doses of >100 Gy BED to the tumour target led to excellent LC rates, which was consistent with our findings. However, our result indicated SRB might be ineffective in reducing the risk of regional or distant metastasis and improving survival of patients. The reasons might be as follows. Intraoperative brachytherapy was a treatment option for local cancer, whereas tumours were systemic diseases with high heterogeneity, possibly leading to a similar rate of systemic failure in each group.

In recent years, the most important prognostic factor for SR appears to be the performance of anatomical segmental resection rather than wedge resection. Research of Chamogeorgakis et al. [24] showed wedge resection was not commensurable with formal lobectomy for early-stage NSCLC. Yet, segmental resection was commensurable with formal lobectomy for peripheral early-stage NSCLC. Koike et al. [25] compared wedge and segmentectomy resection and found the HR for LR and for poorer DFS with the former to be 5.8 and 3.2, respectively. All these studies had indicated that performing anatomical segmentectomy rather than wedge resection was a certain factor potentially associated with better LC. Our results suggested that both segmental resection and wedge resection plus brachytherapy with BED ≤100 Gy were not superior to SR alone in respect of LC. However, wedge resection combined with brachytherapy with BED >100 Gy was related to a better LC. Due to the limited number of studies, it was still unknown whether segmental resection combined with brachytherapy with BED >100 Gy would improve LC.

Limitations

This study has several limitations, including the small number of included studies, different sublobectomy methods (wedge resection and segmentectomy), tumour location, treatment planning system, seed activity and retrospective design of included studies, which possibly led to the intrinsic degree of differences in clinical outcomes for a given dose used. In addition, this study failed to provide specific outcomes due to the different follow-up periods across the studies. Tumour size also had an important impact in determining clinical outcomes, as illustrated by the study, which showed tendency for reduced LC rate for patients with stage IB versus IA [26].

CONCLUSION

In conclusion, results from our meta-analysis demonstrate that for patients with early-stage NSCLC who were not candidates for lobectomy, SR plus intraoperative brachytherapy was not associated with reduced risk of regional or distant metastasis or improved outcomes. However, it may reduce the LR rate when BED was >100 Gy. Future researches with a larger number of cases are required to confirm the conclusions of the present study.

SUPPLEMENTARY MATERIAL

Supplementary material is available at ICVTS online.

Conflict of interest: none declared.

Author contributions

Enli Chen: Conceptualization; Data curation; Software; Writing—original draft; Writing—review & editing. Juan Wang: Conceptualization; Project administration; Resources; Supervision. Chenfei Jia: Data curation; Methodology; Supervision. Xueya Min: Data curation; Investigation; Supervision. Hongtao Zhang: Conceptualization; Investigation; Supervision.

Reviewer information

Interactive CardioVascular and Thoracic Surgery thanks Mohsen Ibrahim, Stefano Margaritora, Mohamed Rahouma and the other, anonymous reviewer(s) for their contribution to the peer review process of this article.

ABBREVIATIONS

BED

Biologically effective dose

CI

Confidence interval

DFS

Disease-free survival

HR

Hazard ratio

LR

Local recurrence

NSCLC

Non-small-cell lung cancer

OS

Overall survival

PD

Prescribed dose

RDR

Regional or distant recurrence

RR

Relative risk

SR

Sublobar resection

SRB

Sublobar resection plus intraoperative brachytherapy