Role of postoperative radiotherapy after curative resection and adjuvant chemotherapy for patients with pathological stage N2 non-small-cell lung cancer: a propensity score matching

Accepted Manuscript

Role of postoperative radiotherapy after curative resection and adjuvantchemotherapy for patients with pathological stage N2 non-small cell lung cancer: Apropensity score matching analysis

Byoung Hyuck Kim , Hak Jae Kim , Hong-Gyun Wu , Chang Hyun Kang , Young TaeKim , Se-Hoon Lee , Dong-Wan Kim

PII: S1525-7304(14)00112-0

DOI: 10.1016/j.cllc.2014.05.005

Reference: CLLC 281

To appear in: Clinical Lung Cancer

Received Date: 16 March 2014

Revised Date: 28 April 2014

Accepted Date: 19 May 2014

Please cite this article as: Kim BH, Kim HJ, Wu H-G, Kang CH, Kim YT, Lee S-H, Kim D-W, Roleof postoperative radiotherapy after curative resection and adjuvant chemotherapy for patients withpathological stage N2 non-small cell lung cancer: A propensity score matching analysis, Clinical LungCancer (2014), doi: 10.1016/j.cllc.2014.05.005.

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service toour customers we are providing this early version of the manuscript. The manuscript will undergocopyediting, typesetting, and review of the resulting proof before it is published in its final form. Pleasenote that during the production process errors may be discovered which could affect the content, and alllegal disclaimers that apply to the journal pertain.

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

1

Role of postoperative radiotherapy after curative resection and adjuvant chemotherapy

for patients with pathological stage N2 non-small cell lung cancer: A propensity score

matching analysis

Byoung Hyuck Kim 1, Hak Jae Kim 1, Hong-Gyun Wu 1,

Chang Hyun Kang 2, Young Tae Kim 2, Se-Hoon Lee 3, Dong-Wan Kim 3

Department of 1 Radiation Oncology, 2 Thoracic Surgery, and 3 Internal Medicine,

Seoul National University College of Medicine, Seoul, Republic of Korea

Corresponding Author: Hak Jae Kim, MD, PhD.

Department of Radiation Oncology, Seoul National University College of Medicine

101 Daehak-ro, Jongno-gu, Seoul 110-744, Korea

Tel: +82-2-2072-2520

Fax: +82-2-765-3317

E-mail address: [email protected]

Conflicts of Interest: none.

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

2

Abstract

Objective: To evaluate the role of postoperative radiotherapy (PORT) in the setting of

adjuvant chemotherapy for pathological stage N2 non-small cell lung cancer (pN2 NSCLC).

Materials and Methods: A retrospective review of 219 consecutive pN2 NSCLC patients

who underwent curative surgery followed by adjuvant chemotherapy was performed. Forty-

one patients additionally received PORT. Propensity scores for PORT receipt were

individually calculated and used for matching to compare the outcome between patients with

and without receiving PORT. 111 patients in PORT (-) group and 38 patients in PORT (+)

group were matched. Clinical and pathologic characteristics were well-balanced.

Results: The median follow-up duration was 48 months. In the matched patients, PORT

resulted in a significantly lower crude loco-regional relapse (43.2% vs. 23.7%, p = 0.032).

Also, PORT was associated with improved loco-regional control rate (LRC) (5-year LRC

63.7% vs. 48.6%, p = 0.036), but not distant metastasis-free survival, disease-free survival

(DFS) and overall survival. An exploratory subgroup analysis suggested a potential DFS

benefit of PORT in patients with multiple station mediastinal lymph node metastases (5-year

DFS 43.2% vs. 16.6%, p = 0.037) and squamous cell carcinoma histology (5-year DFS

70.1% vs. 23.3%, p = 0.011).

Conclusions: Even in the setting of adjuvant chemotherapy, PORT significantly increased

LRC for patients with curatively resected pN2 NSCLC. Some subgroups appear to benefit

from PORT in terms of DFS as well as LRC. Individualized strategies based on risk factors

may be considered.

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

3

Key words: postoperative radiotherapy, adjuvant chemotherapy, non-small cell lung cancer,

propensity score.

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

4

Introduction

Non-small cell lung cancer (NSCLC) shows a considerable rate of recurrence even

after complete resection, and its prognosis remains unsatisfactory. Especially, pathological

stage N2 (pN2) NSCLC is a heterogeneous group considering primary tumor status, clinical

nodal stage and extent of mediastinal lymph node (LN) involvement, with 5-year survival

rates in the range of 5-57% according to prognostic factors.1-4 Postoperative adjuvant

chemotherapy showed its positive impact on overall survival (OS) in several randomized

trials and eventually was accepted as standard treatment for resected pN2 NSCLC. However,

up to 40% of locoregional relapse (LRR) have been reported even after complete resection

followed by chemotherapy.5-7 Based on these considerations, postoperative radiotherapy

(PORT) has been tried to reduce LRR and further to increase OS.

Unfortunately, the PORT Meta-analysis Trialist Group reported that PORT had a

detrimental effect on OS even though patients with pN2 showed non-significantly better

survival with PORT.8 After that, the use of PORT dramatically decreased and consequently

many studies evaluating the role of PORT had great difficulty in accruing patients.9 But, that

meta-analysis has been criticized for several reasons including patient selection, out-of-date

technique, old equipment and inappropriate fraction size which were quite different from

current practice. With rapid improvement in radiotherapy techniques, the interest for PORT

has risen again and recent retrospective studies showed an OS benefit in the pN2 group,

without severe morbidity after the introduction of linear accelerator.10-13

Thus far, no definitive conclusion of the impact of PORT in pN2 NSCLC can be

drawn because no complete prospective randomized controlled study using modern

radiotherapy in the setting of adjuvant chemotherapy has been published. Most reports

concerning PORT have been retrospective analyses of data that included heterogeneous

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

5

treatments with or without adjuvant chemotherapy.10-13 Therefore, we conducted this

retrospective analysis to evaluate the role of PORT in pN2 NSCLC patients who received

adjuvant chemotherapy after curative resection.

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

6

Materials and Methods

Study Population

Between January 2000 and December 2011, 365 consecutive NSCLC patients with

pathologically confirmed N2 nodal involvement after curative resection were identified in our

institution. Of these patients, 29 were excluded because they were lost to follow-up

immediately after surgery (n = 14), died due to postoperative complications (n = 10), or

underwent preoperative concurrent chemoradiation (n = 5). The remained 336 patients were

classified into four groups according to adjuvant treatment: 49 patients did not receive

adjuvant therapy, 68 received PORT alone, 178 received adjuvant chemotherapy alone, and

41 received adjuvant chemotherapy combined with PORT. The latter two groups comprising

219 patients who received adjuvant chemotherapy were included in this study. Reasons why

some patients did not receive adjuvant chemotherapy included poor performance status,

patient refusal, and decision of the medical oncologists because it was not routinely

recommended for pN2 patients in the early period of this study. The medical records were

reviewed retrospectively, and institutional review board approval was obtained. The patient

characteristics are shown in Table 1.

Surgery

All patients underwent lobectomy or pneumonectomy according to the surgeons’

discretion. Methodical mediastinal LN dissection was performed during surgery. Tumor and

nodal status were determined using standard TNM staging according to the American Joint

Committee on Cancer, 7th edition.

Chemotherapy

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

7

Patients with clinically N2 station involvement in the preoperative computed

tomography (CT) and/or positron emission tomography (PET) were received induction

chemotherapy based on physician’s judgment. Induction chemotherapy regimens were

variable, mostly paclitaxel and cisplatin (n = 40) with a median of two cycles (range, 2-4).

After surgery, median 4 cycles (range, 1-6) of adjuvant chemotherapy was

administered sequentially with PORT. Adjuvant chemotherapy regimens varied and included

paclitaxel and carboplatin (n = 111, 50.7%), navelbine and cisplatin (n = 51, 23.3%),

paclitaxel and cisplatin (n = 28, 12.8%), gemcitabine and cisplatin (n = 10, 4.6%), or others

(n = 19, 8.7%).

Radiotherapy

The administration of PORT was based on the decision of radiation oncologist and

referring physician. PORT was delivered using conventional technique (n = 14) before 2005

and three-dimensional conformal technique (n = 27) after that. For conventional technique,

the initial radiation fields covered the entire mediastinum, ipsilateral hilum and bronchial

stump, and then reduced to tumor bed and involved nodal area. For three-dimensional

conformal radiotherapy (3D-CRT), the clinical tumor volume (CTV) included involved

mediastinal LN stations, ipsilateral hilum and bronchial stump. Supraclavicular LN area was

not electively irradiated. Both radiotherapy techniques were delivered with a linear

accelerator using 6 or 10 MV x-rays at 1.8 or 2.0 Gy per fraction, 5 days per week. Median

radiation dose was 54 Gy (range, 50-56).

Follow-Up, Evaluation of Toxicity and Survival

After the completion of treatment, patients were followed up regularly, every 3

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

8

months for the first year and, every 3-6 months thereafter. Radiation pneumonitis and

esophagitis were graded according to Radiation Therapy Oncology Group (RTOG) criteria

and Common Terminology Criteria for Adverse Event v 4.0, respectively. Grades of toxicities

were retrospectively assigned based on medical records at the time of the chart review.

Failure sites were classified as locoregional, distant, or both. Locoregional failure

was defined as a relapse in the surgical bed or in the regional LN regions (including N3 node).

Locoregional control (LRC) and distant metastasis-free survival (DMFS) were calculated

from the date of surgery to the date of each clinical event or last follow-up. Disease-free

survival (DFS) was defined as the time interval between the date of surgery and any failure

after adjuvant treatment or death. Overall survival was defined from the date of surgery to the

date of death from any cause or the last follow-up.

Statistical Analyses

Statistical analyses were performed using the SPSS software, release 18.0.1 (SPSS

Inc. Chicago, IL) and R 2.8.0 statistical package. Because PORT was not randomly assigned

in our cohort, a routinely used multivariate model using Cox regression analysis may not

properly adjust for many confounding factors and possible selection bias. Also, some studies

demonstrated that propensity score (PS) produced results that were more unbiased and robust

than logistic regression results when there were seven or less events per confounder.14 We

identified the number of events of PORT (n = 41) was low relative to the number of potential

confounders, as listed in Table 1. Therefore, we used PS matching analysis to control for

differences in the baseline characteristics between patients with and without receiving PORT,

consequently to observe a less confounded effect of PORT on each clinical endpoint.

Propensity scores for PORT receipt were calculated using multivariate logistic

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

9

regression. The covariates included in the calculation were: age, performance status, clinical

N2 stage, histology, type of surgery, pathologic T stage, tumor size, number of dissected

mediastinal LN, number of positive LN, positive LN ratio, involved number of N2 stations,

and lymphovascular space invasion. Given the PS for all patients, each one patient in PORT

(+) group was matched to maximum four patients who did not receive PORT using nearest

neighbor algorithm.

The differences of clinical and tumor characteristics between two groups were

compared by using Fisher’s exact test or Chi-square test for categorical variables and

independent samples t-test for continuous variables. LRC, DMFS, DFS and OS were

determined through Kaplan-Meier method and compared using log-rank test.

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

10

Results

Patient Characteristics

After PS matching, 111 patients in PORT (-) group and 38 patients in PORT (+)

group were matched. Table 2 lists the characteristics of patients before and after matching.

Prior to matching, patients who received PORT were more likely to have clinical N2 disease

(p = 0.012) and be male gender (p = 0.027). Also, standardized mean differences between two

groups were more than 10% in 11 of 14 variables. However, they were substantially reduced

after matching, and none of the variables was significantly different between two groups

(Table 2).

Patterns of Failure

Median follow-up duration for all patients was 48 months. In the matched patients,

treatment failures were observed in 95 patients, including 19 with LRR alone, 38 with distant

metastases, and 38 with both. Among the 57 patients with LRR, there were 48 (43.2%) in the

PORT (-) group and 9 (23.7%) in the PORT (+) group. PORT resulted in a significantly lower

crude LRR (p = 0.032) after PS matching. But, the difference in distant metastases rate

between two groups was not significant (48.6% vs. 57.9%, p = 0.325). PORT was also

associated with improved LRC, the 5-year LRC 63.7% in the PORT (+) group and 48.6% in

the PORT (-) group (p = 0.036) (Fig. 1). The 5-year DMFS rates were 43.2% and 41.2%,

respectively (p = 0.864).

Survival

The OS and DFS curves according to the PORT receipt in PS-matched patients were

demonstrated in Figure 2. The 5-year OS rates were 58.2% in the PORT (-) group and 49.9%

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

11

in the PORT (+) group (p = 0.466). The 5-year DFS rates were 31.0% in the PORT (-) group

and 38.4% in the PORT (+) group (p = 0.292). No significant impact of PORT on DFS or OS

was observed. Additionally, in the entire patients, the 5-year OS and DFS rates were 53.3%

and 29.4% in the PORT (-) group and 51.0% and 37.7% in the PORT (+) group.

Although this study was not designed to evaluate the impact of PORT on DFS or OS

according to the risk factors, exploratory subgroup analyses were performed. As a result, two

possible factors were found: multiple station mediastinal LN metastases and squamous cell

carcinoma (SqCC) histology. Patients responded differently to PORT depending upon these

two factors.

In patients with multiple station mediastinal LN metastases, 5-year LRC (66.0% vs.

29.4%, p = 0.011) and DFS (43.2% vs. 16.6%, p = 0.037) were significantly higher in

patients who underwent PORT than in those who did not (Fig. 3A, 3B). Similarly, in patients

with SqCC, 5-year LRC (80.8% vs. 45.8%, p = 0.043) and DFS (70.1% vs. 23.3%, p = 0.011)

were significantly higher in patients who underwent PORT than in those who did not (Fig. 3C,

3D). However, whether receiving PORT or not, there were no differences in LRC, DMFS,

DFS and OS in patients with single station mediastinal LN metastasis or adenocarcinoma

histology. Especially, the possible reason of different impact of PORT according to histology

may be assumed to be a discrepancy of failure pattern. In patients with SqCC, LRR (37.5%)

was most common pattern and distant metastasis (32.5%) was also occurred comparably,

whereas distant metastasis was more frequent (59.2%) than LRR (37.7%) in patients with

adenocarcinoma.

Toxicities of PORT

Among 41 patients who received PORT, grade 3 radiation pneumonitis and

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

12

esophagitis was observed in only one patient (2.4%), respectively. These patients recovered

after supportive management. No grade 4 or more acute radiation-induced toxicity was

occurred. Unfortunately, it was not possible to assess the long-term effects of PORT because

causes of death could not be clearly evaluated due to the retrospective nature of this study.

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

13

Discussion

The main purpose of this retrospective study was to evaluate the role of PORT for

pN2 NSCLC. Chemotherapy is accepted as the standard postoperative treatment for high-risk

completely resected NSCLC, so the role of PORT should be discussed separately according

to the administration of adjuvant chemotherapy. To our knowledge, there has been no

complete randomized controlled trial to assess the role of PORT in the setting of adjuvant

chemotherapy. Given the paucity of evidences for the application of PORT, the preferred

recent strategy in our institution has been adjuvant chemotherapy alone. Thus, it is hard to

conduct a new randomized trial, and we used retrospective matching analysis to obtain a less

biased comparison.

Many studies have reporting an OS benefit of PORT for pN2 NSCLC (Table 3). In

the subgroup analysis of the Surveillance, Epidemiology, and End Results (SEER) database,

PORT significantly improved survival for patients with pN2 disease (5-year OS 27% vs. 20%,

HR 0.855, p = 0.008).10 The positive effect of PORT in pN2 disease was also suggested in the

retrospective evaluation of the Adjuvant Navelbine International Trialist Association

(ANITA) trial, which demonstrated that PORT increased median survival both in the

chemotherapy (47.4 vs. 23.8 months) and observation arm (22.7 vs. 12.7 months).11 Recently,

the outcome of an early closed randomized controlled trial comparing postoperative

chemotherapy alone and concurrent chemoradiotherapy in patients with pN2 NSCLC was

reported. It demonstrated that addition of PORT increased both locoregional and distant DFS

rate and produced marginally significant increase in OS (5-year 37.9% vs. 27.5%, p = 0.073),

although it had a relatively small sample size because of slow accrual.15

On the other hand, the present absence of an OS benefit does not coincide with

previous studies. There are some possible reasons. First, OS in our study was relatively very

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

14

high compared to previous reports, as 5-year OS being about 50% in both the PORT (+) and

PORT (-) group (Table 3). Although adjuvant chemotherapy was generally recommended for

pN2 NSCLC patients, patients who did not receive it due to postoperative complications or

poor general condition were not included in our study. This fact could have influenced the

clinical outcomes in comparison to those of previously reported for pN2 NSCLC.

Consequently, the benefit of PORT on OS may have been masked by the high OS of the

PORT (-) group. Second, the composition of patient characteristics may also have masked the

benefit of PORT. Our subgroup analysis suggested that patients with single mediastinal LN

station metastasis or adenocarcinoma histology did not derive benefit from PORT. However,

there were more patients with these characteristics than without them (Table 1). Patients with

adenocarcinoma accounted for 64% of all patients, which was slightly higher than the 41-

53% reported in previous randomized studies for resected NSCLC.5-7 Adenocarcinoma has a

relatively higher possibility of developing distant metastases without local progression in

NSCLC patients.16,17 Our analysis of pattern of failure also identified these observations.

Because radiation therapy is effective in lowering the risk of LRR as a local modality, it

seemed reasonable that those patients who were at high risk for LRR could be the subjects

who were probably to obtain a benefit from PORT.

Some authors suggested different effectiveness of PORT according to the risk factors.

Sawyer et al. found that patients at intermediate or high-risk for local recurrence had

improved LRC and OS with the use of PORT.18 No statistically differences appeared in the

low-risk group between the patients who received PORT and those who did not. Matsuguma

et al. also reported that PORT for completely resected pN2 NSCLC was more effective for

multiple station mediastinal LN metastasis than single station.19 Five-year DFS of the PORT

group (41.7%) was significantly better than that (5.9%) of the non-PORT group in patients

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

15

with multiple station N2 disease (p = 0.022), while there was no difference in single station

N2 disease (p = 0.601). These results are in close agreement with our study. According to the

our subgroup analysis, multiple station mediastinal LN metastasis and SqCC could be the

factors favoring PORT in pN2 NSCLC patients who were received adjuvant chemotherapy

after curative resection.

Although there is still controversy regarding the role of PORT in terms of OS, it is

very likely that PORT could improve LRC in pN2 disease even after adjuvant chemotherapy.

Among the pN2 patients randomized to chemotherapy in the ANITA trial, patients with

PORT showed 14.6% LRR, which was better than 25.7% without PORT.11 Another phase II

study by the RTOG reported comparable intra-thoracic recurrence rate of 15% after

administration of adjuvant paclitaxel and carboplatin with concurrent PORT in resected stage

II and IIIA NSCLC.20 Those results are similar with ours: the LRR rates were 23.7% and

43.2% in the PORT (+) and PORT (-) groups after PS matching, respectively (p = 0.032).

Similar results concerning LRC had also been confirmed in several previous randomized

studies without applying adjuvant chemotherapy.21-23 All the above findings provide

convincing evidence of LRC benefit with the use of PORT for patients with resected pN2

NSCLC.

In terms of surgical techniques, 30 or more mediastinal LN were dissected in about

57% of all the patients, which was more extensive than what were used in other published

PORT studies.12,15 Still, there was no obvious evidence of improvement in survival by

extensive mediastinal LN dissection especially for pN2 NSCLC. But, recent studies

demonstrated that adequate LN dissection could allow for accurate nodal staging and also

reduce the chance of occult nodal metastases.24,25 In our study, the use of additional PORT

improved LRC even after sufficient LN dissection was performed. The impact of PORT

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

16

according to the extent of nodal dissection needs to be further studied.

Our results also showed that PORT-related acute toxicity was mild. Only one patient

suffered from grade 3 radiation pneumonitis and esophagitis, respectively. Other recent

prospective studies including NSCLC patients who were treated with PORT after curative

resection, also reported that there were acceptable rates of grade 3 or more severe

complications associated with PORT.15,20 Concerning late toxicity of PORT, we could not

analyze it because PORT-related death could not be clearly assessed. But, several studies with

contemporary radiotherapy techniques have suggested that some worrying toxicities

associated with pulmonary or cardiac complications decreased greatly after the wide use of

modern treatment systems such as linear accelerators and 3D-CRT.26,27 Also, recent published

meta-analysis demonstrated that modern PORT with linear accelerators was estimated to

increase the absolute 5-year OS by 13% in stage IIIA-N2 NSCLC patients even after

induction chemotherapy and surgery.28 All these findings may support the idea that there was

no significant increase of death resulted from PORT nowadays, but rather a decrease in

cancer-related death resulted from preventing LRR.

There are several limitations to this study. First, its retrospective nature might be a

significant weakness although we tried to obtain unbiased comparison by using PS matching.

PS matching itself could not solve the problem of hidden bias caused by unexamined factors.

In addition, PORT-related toxicity might be underestimated because it was not collected

prospectively. Second, the small number of patients in the PORT (+) group also restricted

statistical power and the number of patients who did not receive PORT might not have been

large enough to allow high-quality matching. Third, because not all patients receive

preoperative PET/CT evaluation and mediastinal staging using endobronchial ultrasound

guided biopsy or mediastinoscopy, patients who could not be eligible for curative resection

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

17

may have been included. Those patients would receive definitive concurrent chemoradiation

instead of surgery followed by adjuvant therapy even if accurate preoperative staging had

been performed, whether or not induction chemotherapy was applied. Lastly, heterogeneous

adjuvant treatments during a relatively long period may also influence clinical outcomes.

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

18

Conclusion

In conclusion, PORT is associated with a higher LRC for patients with resected pN2

NSCLC even in the setting of adjuvant chemotherapy, without severe toxicity. Our data

suggest that patients with multiple mediastinal LN station metastases or SqCC histology

appear to benefit from PORT in terms of DFS as well as LRC. Individualized adjuvant

strategies based on risk factors may be considered. Further prospective studies with a

sufficient sample size are needed to confirm these results.

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

19

Acknowledgments

This study was supported by a grant of the Ministry of Science, Information,

Communication, Technology, and Future Planning Korea Grant No. NRF (National Research

Foundation)-2013R1A1A1007199.

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

20

Conflict of Interest statement

None declared.

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

21

References

1. Lorent N, De Leyn P, Lievens Y, et al. Long-term survival of surgically staged IIIA-N2

non-small-cell lung cancer treated with surgical combined modality approach: analysis of

a 7-year prospective experience. Ann Oncol 2004; 15:1645–53.

2. Casali C, Stefani A, Natali P, et al. Prognostic factors in surgically resected N2 non-small

cell lung cancer: the importance of patterns of mediastinal lymph nodes metastases. Eur

J Cardiothorac Surg 2005; 28:33–8.

3. Andre F, Grunenwald D, Pignon JP, et al. Survival of patients with resected N2 non-

small-cell lung cancer: evidence for a subclassification and implications. J Clin Oncol

2000; 18:2981–9.

4. Suzuki K, Nagai K, Yoshida J, et al. The prognosis of surgically resected N2 non-small

cell lung cancer: the importance of clinical N status. J Thorac Cardiovasc Surg 1999;

118:145–53.

5. Arriagada R, Bergman B, Dunant A, et al. Cisplatin-based adjuvant chemotherapy in

patients with completely resected non-small-cell lung cancer. N Engl J Med 2004;

350:351–60.

6. Winton T, Livingston R, Johnson D, et al. Vinorelbine plus cisplatin vs. observation in

resected non-small-cell lung cancer. N Engl J Med 2005; 352:2589–97.

7. Douillard J-Y, Rosell R, De Lena M, et al. Adjuvant vinorelbine plus cisplatin versus

observation in patients with completely resected stage IB-IIIA non-small-cell lung cancer

(Adjuvant Navelbine International Trialist Association [ANITA]): a randomised

controlled trial. Lancet Oncol 2006; 7:719–27.

8. Postoperative radiotherapy in non-small-cell lung cancer: systematic review and meta-

analysis of individual patient data from nine randomised controlled trials. PORT Meta-

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

22

analysis Trialists Group. Lancet 1998; 352:257–63.

9. Bekelman JE, Rosenzweig KE, Bach PB, et al. Trends in the use of postoperative

radiotherapy for resected non-small-cell lung cancer. Int J Radiat Oncol Biol Phys 2006;

66:492–9.

10. Lally BE, Zelterman D, Colasanto JM, et al. Postoperative radiotherapy for stage II or III

non-small-cell lung cancer using the surveillance, epidemiology, and end results database.

J Clin Oncol 2006; 24:2998–3006.

11. Douillard J-Y, Rosell R, De Lena M, et al. Impact of postoperative radiation therapy on

survival in patients with complete resection and stage I, II, or IIIA non-small-cell lung

cancer treated with adjuvant chemotherapy: the adjuvant Navelbine International Trialist

Association (ANITA) Randomized Trial. Int J Radiat Oncol Biol Phys 2008; 72:695–701.

12. Dai H, Hui Z, Ji W, et al. Postoperative radiotherapy for resected pathological stage IIIA-

N2 non-small cell lung cancer: a retrospective study of 221 cases from a single

institution. The oncologist 2011; 16:641–50.

13. Zou B, Xu Y, Li T, et al. A multicenter retrospective analysis of survival outcome

following postoperative chemoradiotherapy in non-small-cell lung cancer patients with

N2 nodal disease. Int J Radiat Oncol Biol Phys 2010; 77:321–8.

14. Cepeda MS, Boston R, Farrar JT, et al. Comparison of logistic regression versus

propensity score when the number of events is low and there are multiple confounders.

Am J Epidemiol 2003; 158:280–7.

15. Shen W-Y, Ji J, Zuo Y-S, et al. Comparison of efficacy for postoperative chemotherapy

and concurrent radiochemotherapy in patients with IIIA-pN2 non-small cell lung cancer:

An early closed randomized controlled trial. Radiother Oncol 2014; 110:120-5.

16. Cox JD, Scott CB, Byhardt RW, et al. Addition of chemotherapy to radiation therapy

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

23

alters failure patterns by cell type within non-small cell carcinoma of lung (NSCCL):

analysis of radiation therapy oncology group (RTOG) trials. Int J Radiat Oncol Biol Phys

1999; 43:505–9.

17. Perez CA, Pajak TF, Rubin P, et al. Long-term observations of the patterns of failure in

patients with unresectable non-oat cell carcinoma of the lung treated with definitive

radiotherapy. Report by the Radiation Therapy Oncology Group. Cancer 1987; 59:1874–

81.

18. Sawyer TE, Bonner JA, Gould PM, et al. Effectiveness of postoperative irradiation in

stage IIIA non-small cell lung cancer according to regression tree analyses of recurrence

risks. Ann Thorac Surg 1997; 64:1402–1407.

19. Matsuguma H, Nakahara R, Ishikawa Y, et al. Postoperative radiotherapy for patients

with completely resected pathological stage IIIA-N2 non-small cell lung cancer: focusing

on an effect of the number of mediastinal lymph node stations involved. Interact

Cardiovasc Thorac Surg 2008; 7:573–7.

20. Bradley JD, Paulus R, Graham MV, et al. Phase II trial of postoperative adjuvant

paclitaxel/carboplatin and thoracic radiotherapy in resected stage II and IIIA non-small-

cell lung cancer: promising long-term results of the Radiation Therapy Oncology Group-

-RTOG 9705. J Clin Oncol 2005; 23:3480–7.

21. Effects of postoperative mediastinal radiation on completely resected stage II and stage

III epidermoid cancer of the lung. The Lung Cancer Study Group. N Engl J Med 1986;

315:1377–81.

22. Stephens RJ, Girling DJ, Bleehen NM, et al. The role of post-operative radiotherapy in

non-small-cell lung cancer: a multicentre randomised trial in patients with pathologically

staged T1-2, N1-2, M0 disease. Medical Research Council Lung Cancer Working Party.

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

24

Br J Cancer 1996; 74:632–9.

23. Feng QF, Wang M, Wang LJ, et al. A study of postoperative radiotherapy in patients with

non-small-cell lung cancer: a randomized trial. Int J Radiat Oncol Biol Phys 2000;

47:925–9.

24. Boffa DJ, Kosinski AS, Paul S, et al. Lymph node evaluation by open or video-assisted

approaches in 11,500 anatomic lung cancer resections. Ann Thorac Surg 2012; 94:347-53.

25. Xu F, Qi L, Yue D, et al. The effect of the extent of lymph node dissection for stage IA

non-small-cell lung cancer on patient disease-free survival. Clin Lung Cancer 2013;

14:181-7.

26. Lally BE, Detterbeck FC, Geiger AM, et al. The risk of death from heart disease in

patients with nonsmall cell lung cancer who receive postoperative radiotherapy: analysis

of the Surveillance, Epidemiology, and End Results database. Cancer 2007; 110:911–7.

27. Wakelee HA, Stephenson P, Keller SM, et al. Post-operative radiotherapy (PORT) or

chemoradiotherapy (CPORT) following resection of stages II and IIIA non-small cell

lung cancer (NSCLC) does not increase the expected risk of death from intercurrent

disease (DID) in Eastern Cooperative Oncology Group (ECOG) trial E3590. Lung

Cancer Amst Neth 2005; 48:389–97.

28. Billiet C, Decaluwé H, Peeters S, et al. Modern post-operative radiotherapy for stage III

non-small cell lung cancer may improve local control and survival: A meta-analysis.

Radiother Oncol 2014; 110:3-8.

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

25

Figure Legends

Figure 1. Locoregional control curves according to the administration of postoperative

radiotherapy (PORT).

Figure 2. Overall survival (A) and disease-free survival (B) curves according to the

administration of postoperative radiotherapy (PORT).

Figure 3. Locoregional control (A) and disease-free survival (B) curves according to the

administration of postoperative radiotherapy (PORT) in patients with multiple station

mediastinal lymph node metastases. Locoregional control (C) and disease-free survival (D)

curves according to the administration of postoperative radiotherapy (PORT) in patients with

squamous cell carcinoma.

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

Table 1. Patient characteristics

Characteristics No. of patients %

Age (years) Median 60 (range, 34-84)

Gender Male 138 (63.0)

Female 81 (37.0)

Performance (ECOG) 0 18 (8.2)

1 182 (83.1)

2 19 (8.7)

Preoperative clinical stage N2 Yes 90 (41.1)

No 129 (58.9)

Type of surgery Lobectomy 175 (79.9)

Bilobectomy 23 (10.5)

Pneumonectomy 21 (9.6)

Histology Squamous cell carcinoma 61 (27.9)

Adenocarcinoma 141 (64.4)

Others 17 ( 7.8)

Tumor size <3cm 93 (42.5)

≥3cm 126 (57.5)

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

Pathologic T stage T0-1 62 (28.3)

T2 109 (49.8)

T3-4 48 (21.9)

Number of dissected nodes <30 94 (42.9)

≥30 125 (57.1)

Number of positive nodes <6 148 (67.6)

≥6 71 (32.4)

Positive lymph node ratio <0.2 145 (66.2)

≥0.2 74 (33.8)

Number of N2 stations Multiple stations 84 (38.4)

Single station 135 (61.6)

Lymphovascular space invasion Yes 107 (48.9)

No 98 (44.7)

Unknown 14 (6.4)

Induction chemotherapy Yes 51 (23.3)

No 168 (76.7)

Postoperative radiotherapy Yes 41 (18.7)

No 178 (81.3)

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

ECOG, eastern cooperative oncology group.

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

Table 2. Comparison of characteristics before and after propensity score matching

Before matching After matching

Characteristics

PORT (-)

(n = 178)

PORT (+)

(n = 41)

Pa

SMD

PORT (-)

(n = 111)

PORT (+)

(n = 38)

Pa

SMD

Age (yr) 60.2±8.8 57.6±9.4 0.092 0.277 59.0±9.2 57.4±9.7 0.362 0.047

Gender 0.027 0.405 0.135 0.290

Male 106 (59.6%) 32 (78.0%) 70 (63.1%) 29 (76.3%)

Female 72 (40.4%) 9 (22.0%) 41 (36.9%) 9 (23.7%)

ECOG 0.270 0.210 0.265 0.042

0 15 (8.4%) 3 (7.3%) 11 (9.9%) 3 (7.9%)

1 145 (81.5%) 37 (90.2%) 88 (79.3%) 34 (89.5%)

2 18 (10.1%) 1 (2.4%) 12 (10.8%) 1 (2.6%)

Preoperative clinical

stage N2

0.012 0.430 0.482 0.134

Yes 66 (37.1%) 24 (58.5%) 54 (48.6%) 21 (55.3%)

Induction

chemotherapy

0.068 0.301 0.516 0.122

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

Yes 37 (20.8%) 14 (34.1%) 29 (26.1%) 12 (31.6%)

Histology 0.500 0.109 0.925 0.021

SqCC 50 (28.1%) 11 (26.8%) 29 (26.1%) 11 (28.9%)

Adenocarcinoma 116 (65.2%) 25 (61.0%) 74 (66.7%) 24 (63.2%)

Others 12 (6.7%) 5 (12.2%) 8 (7.2%) 3 (7.9%)

Type of surgery 0.092 0.200 0.082 0.185

Lobectomy 142 (79.8%) 33 (80.5%) 90 (81.1%) 30 (78.9%)

Bilobectomy 16 (9.0%) 7 (17.1%) 9 (8.1%) 7 (18.4%)

Pneumonectomy 20 (11.2%) 1 (2.4%) 12 (10.8%) 1 (2.6%)

Pathologic T stage 0.850 0.056 0.858 0.015

T0-1 49 (27.5%) 13 (31.7%) 30 (27.0%) 11 (28.9%)

T2 90 (50.6%) 19 (46.3%) 61 (55.0%) 19 (50.0%)

T3-4 39 (21.9%) 9 (22.0%) 20 (18.0%) 8 (21.1%)

Tumor size (mm) 34.7±15.8 32.7±15.5 0.483 0.124 33.2±15.7 34.2±14.8 0.752 0.095

Number of dissected

nodes

31.2±10.9 34.3±12.5 0.111 0.248 33.8±11.5 34.1±12.8 0.882 0.016

Number of positive 5.4±5.4 5.9±6.5 0.579 0.084 5.6±5.5 5.9±6.7 0.767 0.007

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

nodes

Positive LN ratio 0.2±0.2 0.2±0.2 0.901 0.020 0.2±0.2 0.2±0.2 0.727 0.009

Number of N2

stations

0.128 0.253 0.730 0.017

Multiple 64 (36.0%) 20 (48.8%) 49 (44.1%) 18 (47.4%)

Lymphovascular

space invasion

0.576 0.108 0.864 0.041

Yes 77 (43.3%) 21 (51.2%) 54 (48.6%) 17 (44.7%)

No 90 (50.6%) 17 (41.5%) 50 (45.0%) 19 (50.0%)

Unknown 11 (6.2%) 3 (7.3%) 7 (6.3%) 2 (5.3%)

a Chi-square test for nominal variables, independent samples t-test for continuous variables.

PORT, postoperative radiotherapy; SMD, standardized mean differences (absolute value); ECOG, eastern cooperative

oncology group; SqCC, squamous cell carcinoma; LN, lymph node.

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

Table 3. Comparison of outcomes of previous studies with or without postoperative radiotherapy for pathological stage N2 non-

small cell lung cancer

Study

PORT

No. of

patients

Disease-free survival Overall survival Percentage of patients

receiving chemotherapy 5-year rate P 5-year rate P

Lally et al. 200610a No 755 (38%) NR - 20.0% 0.008 NR

Yes 1232 (62%) NR 27.0%

Douillard et al. 200811a No 70 (59%) NR - 34.0% NR 100%

Yes 48 (41%) NR 47.4%

Zou et al. 201013 No 79 (43%) 9.3% 0.003 14.4% 0.007 100%

Yes 104 (57%) 22.2% 30.5%

Dai et al. 201112 No 125 (57%) 16.5% 0.009 30.6% 0.046 73%

Yes 96 (43%) 32.1% 36.6%

Shen et al. 201315 No 69 (51%) 18.8% 0.041 27.5% 0.073 100%

Yes 66 (49%) 30.3% 37.9%

Current study No 178 (81%) 29.4% 0.226 53.3% 0.824 100%

Yes 41 (19%) 37.7% 51.0%

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

a Results of subgroup analysis.

PORT, postoperative radiotherapy; NR, not reported.

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT