If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
The study objective was to determine differences in survival depending on adjuvant therapy type, timing, and sequence in node-negative disease with positive margins after non–small cell lung cancer resection.
Methods
The National Cancer Database was queried for patients with positive margins after surgical resection of treatment-naïve cT1-4N0M0 pN0 non–small cell lung cancer who underwent adjuvant radiotherapy or chemotherapy from 2010 to 2016. Adjuvant treatment groups were defined as surgery alone, chemotherapy alone, radiotherapy alone, concurrent chemoradiotherapy, sequential chemotherapy then radiotherapy, and sequential radiotherapy then chemotherapy. The impact of adjuvant radiotherapy initiation timing on survival was evaluated using multivariable Cox regression. Kaplan–Meier curves were generated to compare 5-year survival.
Results
A total of 1713 patients met inclusion criteria. Five-year survival estimates differed significantly between cohorts: surgery alone, 40.7%; chemotherapy alone, 47.0%; radiotherapy alone, 35.1%; concurrent chemoradiotherapy, 45.7%; sequential chemotherapy then radiotherapy, 36.6%; and sequential radiotherapy then chemotherapy, 32.2% (P = .033). Compared with surgery alone, adjuvant radiotherapy alone had a lower estimated survival at 5 years, although overall survival did not differ significantly (P = .8). Chemotherapy alone improved 5-year survival compared with surgery alone (P = .0016) and provided a statistically significant survival advantage over adjuvant radiotherapy (P = .002). Compared with radiotherapy-inclusive multimodal therapies, chemotherapy alone yielded similar 5-year survival (P = .066). Multivariable Cox regression showed an inverse linear association between time to adjuvant radiotherapy initiation and survival, but with an insignificant trend (10-day hazard ratio, 1.004; P = .90).
Conclusions
In treatment-naïve cT1-4N0M0 pN0 non–small cell lung cancer with positive surgical margins, only adjuvant chemotherapy was associated with a survival improvement compared with surgery alone, with no radiotherapy-inclusive treatment providing additional survival benefit. Delayed timing of radiotherapy initiation was not associated with a survival reduction.
The efficacy of adjuvant therapy focusing on node-negative disease requires greater evidence to support the role of PORT in the context of positive surgical margins. The findings from this study raise a potential paradigm shift and reevaluation of the routine practice of using radiotherapy in the care of patients who have positive margins after resections for NSCLC.
See Commentary on page XXX.
For medically operable patients with early stage non–small cell lung cancer (NSCLC), curative intent complete surgical resection is considered the standard treatment.
A small proportion of surgical resections can be associated with positive surgical margins, which can be classified as microscopic or macroscopic residual disease.
NCCN Clinical Practice Guidelines in Oncology (NCCN Guidelines): Non-Small Cell Lung Cancer. National Comprehensive Cancer Network (NCCN). Version 3.2021. 2021
For patients with positive surgical margins, re-resection of the tumor or postoperative radiotherapy (adjuvant radiotherapy [PORT]) is thought to be indicated, whereas adjuvant chemotherapy (ACT) is often recommended for more advanced stages.
NCCN Clinical Practice Guidelines in Oncology (NCCN Guidelines): Non-Small Cell Lung Cancer. National Comprehensive Cancer Network (NCCN). Version 3.2021. 2021
PORT has been associated as an avenue of treatment for positive margins after surgical resection, but there are limited data that detail the optimal time to initiate PORT. In the LungART trial, study participants were advised to start PORT 4 weeks after surgery at the earliest, but there was no indication of the timing associated with best outcomes.
Postoperative radiotherapy versus no postoperative radiotherapy in patients with completely resected non-small-cell lung cancer and proven mediastinal N2 involvement (Lung ART): an open-label, randomised, phase 3 trial.
By using a cutoff point analysis, a time to radiotherapy of 8 weeks or more with sequential chemotherapy in margin-negative N2 disease has been associated with improved survival.
In early-stage disease, although PORT has been shown to reduce local recurrence, the impact of PORT on overall survival remains uncertain with some studies suggesting that PORT may have a detrimental effect on survival.
demonstrated that the administration of both chemotherapy and radiotherapy is associated with improved survival in patients with microscopically positive margins but that chemotherapy or radiotherapy alone was less consistently associated with improved outcomes. Although these recommendations pertain to the population level, there is limited guidance on the specific treatment of node-negative disease and the timing of PORT. Additional clarity is also needed on the relative efficacy of unimodal versus combination treatment modalities. For example, the appropriate sequence of multimodal therapy—concurrent versus sequential chemoradiotherapy—is not known and has been debated, with published evidence supporting both options.
Much of the existing literature comparing the efficacy of various adjuvant therapies focuses on lymph node–positive N2, margin-negative disease, neoadjuvant-inclusive, or otherwise multimodal therapies. More data are needed to support the role of radiotherapy in the context of positive surgical margins either with or without ACT. The objective of this study was to identify the most effective adjuvant therapy and the optimal timing for PORT specifically for patients with positive margins in treatment-naïve node-negative disease using a national cancer registry.
Materials and Methods
Data Source
The National Cancer Database (NCDB) is a hospital-based clinical oncology tumor registry maintained as a joint effort of the American College of Surgeons and the American Cancer Society. The NCDB contains more than 34 million historical records of patients with cancer obtained from more than 1500 Commission of Cancer–accredited facilities, representing approximately 70% of patients diagnosed annually with cancer. The NCDB maintains that “the data used in the study are derived from a de-identified NCDB file. The American College of Surgeons and the Commission on Cancer have not verified and are not responsible for the analytic or statistical methodology employed, or the conclusions drawn from these data by the investigator.”
The Institutional Review Board or equivalent ethics committee of the University of Southern California approved the study protocol and publication of data (Number HS-16-00,906; approval date: December 19, 2016). Patient written consent for the publication of the study data was waived by the Institutional Review Board because this study is not considered human subjects research.
Study Population
The NCDB Participant User Data File was used to identify treatment-naïve patients (ie, no preoperative therapy) with clinical stage T1-4N0M0 and eventual pathologic N0 NSCLC who underwent surgical resection from 2010 to 2016, were found to have positive margins, and received PORT or ACT. Only patients who underwent lobectomy, extended lobectomy, pneumonectomy, extended pneumonectomy, or extended radical pneumonectomy within 180 days after diagnosis were included. Patients with clinical T descriptors missing or T0 or with metastatic disease (M1) were excluded. Only clinical node-negative and pathological node-negative patients were included. Patients whose PORT lasted greater than 60 days or who received radiotherapy dosages less than 4500 cGy or greater than 7000 cGy were excluded to ensure guideline-concordant care. Only radiotherapy of chest/lung and lung (limited) and the following radiotherapy modalities were included: external beam not otherwise specificed, photons (6-10 MV), photons (mixed energies), intensity-modulated radiotherapy, conformal or 3-dimensional therapy, or proton therapy. Patients receiving single-agent ACT were excluded as were those with unknown sequence of adjuvant therapy (Figure 1).
Figure 1Schematic diagram showing the cohort selection process. The final cohort contained 1713 patients who were categorized into 6 separate cohorts based on the adjuvant therapy administered. NSCLC, Non–small cell lung cancer; NCDB, National Cancer Database.
Demographic variables included sex, age, race, income, education, insurance type, urban-rural, facility type, and distance from nearest facility. Tumor and clinical specific variables included histology, Charlson-Deyo score, American Joint Commission on Cancer (AJCC) clinical TNM stage, AJCC pathological TNM stage, tumor size (≤3 cm, 3-5 cm, 5-7 cm, and >7 cm), surgical margin, and procedure type (Table 1). Histology was subdivided into adenocarcinoma, squamous cell carcinoma, and other. The tumor size cutoffs were chosen in accordance with the Eighth Edition TNM NSCLC staging classification.
Table 1Demographics and clinical characteristics of National Cancer Database patients
Characteristic
SA
PORT
ACT
cCRT
sCRT
sRTC
P value
(N = 1027)
(N = 220)
(N = 230)
(N = 177)
(N = 31)
(N = 28)
No.
%
No.
%
No.
%
No.
%
No.
%
No.
%
Sex
.29
Male
547
53.3
131
59.5
116
50.4
102
57.6
19
61.3
17
60.7
Female
480
46.7
89
40.5
114
49.6
75
42.4
12
38.7
11
39.3
Age, y
<.0001
≤61
195
19.0
42
19.1
79
34.3
63
35.6
7
22.6
10
35.7
61-68
209
20.4
59
26.8
70
30.4
40
22.6
9
29.0
9
32.1
68-75
293
28.5
56
25.5
56
24.3
52
29.4
9
29.0
6
21.4
>75
330
32.1
63
28.6
25
10.9
22
12.4
6
19.4
3
10.7
Race
.73
Missing
3
0.3
1
0.5
0
0
0
0
0
0
0
0
Black
88
8.6
16
7.3
20
8.7
12
6.8
2
6.5
0
0
Other
30
2.9
7
3.2
5
2.2
3
1.7
2
6.5
0
0
White
906
88.2
196
89.1
205
89.1
162
91.5
27
87.1
28
100
Histology
.46
Adenocarcinoma
537
52.3
102
46.4
134
58.3
89
50.3
17
54.8
12
42.9
Other
34
3.3
8
3.6
10
4.3
7
4
1
3.2
2
7.1
Squamous cell
456
44.4
110
50
86
37.4
81
45.8
13
41.9
14
50
Charlson-Deyo Score
.006
0
469
45.7
91
41.4
115
50
96
54.2
16
51.6
12
42.9
1
359
35
77
35
90
39.1
63
35.6
10
32.3
11
39.3
2
138
13.4
30
13.6
19
8.3
16
9
3
9.7
5
17.9
≥3
61
5.9
22
10
6
2.6
2
1.1
2
6.5
0
0
Income
.19
Missing
3
0.3
0
0
1
0.4
1
0.6
0
0
0
0
≥$38,000
813
79.2
176
80
193
83.9
133
75.1
28
90.3
21
75
<$38,000
211
20.5
44
20
36
15.7
43
24.3
3
9.7
7
25
Education
.2
Missing
3
0.3
0
0
1
0.4
1
0.6
0
0
0
0
<20.9% no high school
834
81.2
187
85
202
87.8
143
80.8
25
80.6
24
85.7
>21% no high school
190
18.5
33
15
27
11.7
33
18.6
6
19.4
4
14.3
Insurance
<.0001
Missing
6
0.6
3
1.4
4
1.7
1
0.6
0
0
0
0
Not insured
11
1.1
5
2.3
12
5.2
1
0.6
0
0
0
0
Private insurance/managed care
218
21.2
65
29.5
90
39.1
60
33.9
12
38.7
11
39.3
Medicaid
59
5.7
8
3.6
9
3.9
15
8.5
3
9.7
1
3.6
Medicare
721
70.2
133
60.5
114
49.6
95
53.7
15
48.4
15
53.6
Other government
12
1.2
6
2.7
1
0.4
5
2.8
1
3.2
1
3.6
Urban/rural
.9
Missing
25
2.4
2
0.9
4
1.7
3
1.7
0
0
0
0
Metro
803
78.2
168
76.4
181
78.7
132
74.6
25
80.6
22
78.6
Urban
177
17.2
45
20.5
40
17.4
35
19.8
6
19.4
6
21.4
Rural
22
2.1
5
2.3
5
2.2
7
4
0
0
0
0
Facility type
.06
Missing
3
0.3
0
0
0
0
1
0.6
0
0
0
0
Community cancer program
79
7.7
22
10
24
10.4
16
9
3
9.7
1
3.6
Comprehensive community cancer program
466
45.4
101
45.9
106
46.1
97
54.8
16
51.6
18
64.3
Academic/research program
330
32.1
66
30
68
29.6
34
19.2
4
12.9
7
25
Integrated network cancer program
149
14.5
31
14.1
32
13.9
29
16.4
8
25.8
2
7.1
Distance
.49
Missing
4
0.4
0
0
0
0
0
0
0
0
0
0
>12.5 miles
551
53.7
127
57.7
133
57.8
95
53.7
19
61.3
19
67.9
≤12.5 miles
472
46
93
42.3
97
42.2
82
46.3
12
38.7
9
32.1
AJCC Clinical T
<.0001
c1
465
45.3
81
36.8
41
17.8
29
16.4
4
12.9
6
3.6
c2
386
37.6
96
43.6
108
47.0
68
38.4
13
41.9
16
57.1
c3
140
13.6
37
16.8
68
29.6
64
36.2
10
32.3
2
7.1
c4
36
3.5
6
2.7
13
5.7
16
9
4
12.9
4
14.3
AJCC Pathologic T
<.0001
p1
315
30.7
55
25.0
8
3.5
9
5.1
1
3.2
1
3.6
p2
413
40.2
81
36.8
94
40.9
35
19.8
10
32.3
11
39.3
p3
249
24.2
72
32.7
108
47
102
57.6
16
51.6
12
42.9
p4
45
4.4
11
5
20
8.7
30
16.9
4
12.9
4
14.3
pIS
3
0.3
0
0
0
0
0
0
0
0
0
0
pX
2
0.2
1
0.5
0
0
1
0.6
0
0
0
0
Tumor size
<.0001
≤3 cm
461
51.9
76
41.8
52
25.6
42
28.2
4
14.3
5
21.7
3-5 cm
240
27.0
70
38.5
66
32.5
53
35.6
7
25.0
12
52.2
5-7 cm
108
12.1
26
14.3
40
19.7
29
19.5
12
42.9
5
21.7
>7 cm
80
9.0
10
5.5
45
22.2
25
16.8
5
17.9
1
4.3
Surgical margins
.22
Residual tumor, NOS
411
40
78
35.5
99
43
73
41.2
12
38.7
6
21.4
Microscopic
578
56.3
139
63.2
122
53
96
54.2
19
61.3
21
75
Macroscopic
38
3.7
3
1.4
9
3.9
8
4.5
0
0
1
3.6
Procedure type
.34
Lobectomy or bilobectomy
976
95
212
96.4
212
92.2
169
95.5
31
100
27
96.4
Pneumonectomy
51
5
8
3.6
18
7.8
8
4.5
0
0
1
3.6
SA, Surgery alone; PORT, adjuvant radiotherapy; ACT, adjuvant chemotherapy; cCRT, sequential chemotherapy then radiotherapy; sCRT, sequential chemotherapy then radiotherapy; sRTC, sequential radiotherapy then chemotherapy; AJCC, American Joint Commission on Cancer; NOS, not otherwise specified.
Patients were grouped into 6 cohorts based on their adjuvant therapy sequence: surgery alone (SA), surgery followed by ACT only, surgery followed by PORT only, surgery followed by concurrent chemoradiotherapy (cCRT; defined as chemotherapy and radiotherapy starting within 14 days of each other), surgery followed by sequential radiotherapy then chemotherapy (sRTC; defined as chemotherapy starting after completion of radiotherapy), and surgery followed by sequential chemotherapy then radiotherapy (sCRT; defined as readiotherapy starting at least 100 days after initiation of chemotherapy, per guidelines used in the ANITA and Lung ART trials
Postoperative radiotherapy versus no postoperative radiotherapy in patients with completely resected non-small-cell lung cancer and proven mediastinal N2 involvement (Lung ART): an open-label, randomised, phase 3 trial.
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.
). The primary outcome of interest was 5-year overall survival, defined as the interval from the date of surgery to the date of last contact or last vital status.
Statistical Analysis
Descriptive analysis was used to report frequencies and percentages for all categorical variables for each cohort. Demographic and clinical characteristics (Table 1) were compared using a chi-square test or Fisher exact test when appropriate. The log-rank test with Bonferroni adjustment for P values was used to compare the overall 5-year survival of groups, with Kaplan–Meier estimators calculated for the 5-year time point. Cox regression was used for the association of days between surgery and radiotherapy initiation to 5-year overall survival. Days between surgery and radiotherapy were evaluated as a continuous variable. Tumor size, age, binary gender, quartile of median income, facility location, histology, and surgical margin were included in the time to radiotherapy model as covariates. Proportional hazard assumptions were evaluated using Schoenfeld residuals. TNM clinical staging was omitted because a test based on scaled Schoenfeld residuals showed that the proportional hazards assumption was not supported. All analyses were conducted using R version 4.1.2 (R Foundation for Statistical Computing).
Propensity Matching Survival Analysis
A propensity score matching (PSM) method (nearest-neighbor matching without replacement, using a logistic propensity model and a caliper of 0.2) was used to control for age (in terciles), sex, pathologic stage group (0, 1, 2, 3, or 4), and tumor size (discretized into 4 levels according to Eighth Edition AJCC). Clinical stage group was also viewed as a possible confounder, but was not selected due to moderate rank correlation with pathologic stage group. More complex propensity models, which would have allowed controlling for more variables, were not supported. The log-rank test was used to compare the entire survival experiences of the samples resulting from the PSM procedure. With the same PSM samples, a bootstrap method (the ordinary method, with 10,000 replicates) was used with Kaplan–Meier estimators at 5 years to estimate ratios of 5-year survival probabilities and calculate 95% confidence intervals (CIs).
Results
Of the 114,313 patients with cT1-4N0 pN0 NSCLC, 3280 (3%) had positive margins and 1713 (1%) met the cohort-specific inclusion criteria for treatment-naïve surgical intervention from 2010 to 2016. Of those 1713, 60.0% (1027) underwent SA and 40.0% (686) received some form of adjuvant therapy. For those patients receiving adjuvant therapy, 13.4% (230) underwent ACT, 12.8% (220) underwent PORT, 10.3% (177) underwent cCRT, 1.8% (31) underwent sCRT, and 1.6% (28) underwent RTC.
Evaluating Survival Across Cohorts
ACT yielded the greatest survival, whereas sRTC yielded the lowest estimated 5-year survival probability: ACT, 47.0%; cCRT, 45.7%; sCRT, 36.6%; PORT, 35.1%; sRTC, 32.2%; and SA, 40.7% (P = .033) (Figure 2). Compared with SA, the 5-year survival of PORT was not significantly different (P = .8) (Figure 3, A), although PORT was associated with an estimated 5-point reduction in 5-year survival probability. Between the 2 unimodal therapies, ACT provided a significant improvement in survival over PORT (P = .0021) (Figure 3, B), which remained significant after correcting for multiple testing (P = .032). ACT was the only adjuvant treatment group that improved 5-year survival compared with SA (P = .0016) (Figure 3, C); this statistical difference remained significant after multiple-testing correction (P = .024).
Figure 2Kaplan–Meier survival analysis demonstrating 5-year overall survival of patients with positive margins for all adjuvant therapy treatments with 95% CIs. Overall survival was highest for patients receiving ACT (47.0%) and lowest for those undergoing sRTC (32.2%) (P = .033). ACT, Adjuvant chemotherapy; cCRT, sequential chemotherapy then radiotherapy; PORT, adjuvant radiotherapy; SA, surgery alone; sCRT, sequential chemotherapy then radiotherapy; sRTC, sequential radiotherapy then chemotherapy.
Figure 3Kaplan–Meier survival analysis demonstrating 5-year overall survival of patients with positive margins with 95% CIs according to different adjuvant therapy treatments for (A) PORT and SA; (B) ACT and PORT; (C) ACT and SA; and (D) ACT and the 3 radiotherapy-inclusive combination therapies of cCRT and sCRT and sRTC. PORT, Adjuvant radiotherapy; SA, surgery alone; ACT, adjuvant chemotherapy.
For multimodal combination therapies, cCRT and [sRTC + sCRT] yielded statistically similar 5-year survival (P = .88) when compared with each other. Because of their small sample sizes, sRTC and sCRT were grouped into a single sequential adjuvant therapy cohort for the purposes of statistical analysis. In comparison with unimodal therapy options, neither cCRT nor [sRTC + sCRT] was associated with a significant survival advantage over PORT (P = .33 and P = .41, respectively) or ACT (P = .089 and P = .23, respectively). ACT did not differ from the radiotherapy-inclusive combination therapies (cCRT + sRTC + sCRT) in prolonging survival (P = .066) (Figure 3, D). Additionally, there were no 30-day or 90-day mortalities for any of the 5 adjuvant treatment cohorts.
Subgroup Analyses
To confirm the relative effectiveness of ACT, a propensity-matched analysis was performed. Controlling for age, sex, pathological stage group, and tumor size, ACT provided a statistically significant survival advantage over PORT (unadjusted P < .001; 95% CI, 1.74 [1.15-2.64]) and statistically equivalent survival compared with cCRT (unadjusted P = .29; 95% CI, 1.05 [0.73-1.48]) (Table 2), both of which are consistent with the prior results.
Table 2Output from propensity-matched analysis
Comparison
PSM sample size
PSM log-rank test P value
PSM bootstrap ratio of 5-y survival probabilities (95% CI)
The relative adjuvant treatment efficacy across individual clinical T stages yielded no significant survival differences for the cT1 subgroup, with SA (51.9%) yielding the highest survival. In cT2 patients, ACT (57.5%) had the highest 5-year survival and provided statistically significant survival advantage over cCRT (50.5%), cCRT + sRTC + sCRT (44.4%), SA (36.5%), and PORT (30.8%) (unadjusted P = .03, P = .03, P < .01, P < .01, respectively). For cT3 + cT4 (combined due to small cT4 sample size), sCRT + sRTC (40.4%), cCRT (37.6%), and ACT (30.3%) individually provided significant survival benefit compared with SA (22.6%) (unadjusted P = .04, P < .01, P < .01, respectively). However, no therapy proved more statistically effective than the others (all P > .64).
As a proxy for disease-specific survival, a subgroup analysis was performed for those patients with Charlson-Deyo score equal to 0 whereby the only statistically significant comparison was ACT (45.3%) versus SA (42.4%) (unadjusted P = .047), with no significant differences between the different adjuvant treatment groups (ie, cCRT, 45.2%; PORT, 41.7%; sCRT + sRTC, 37.6%).
Time to Adjuvant Radiotherapy
To evaluate the relationship between PORT timing and survival, a multivariable Cox model was fitted using the PORT and cCRT cohorts combined. The sRTC and sCRT cohorts were excluded because of small sample size and likelihood that the 100-day minimum before initiation of radiotherapy would skew the analysis, respectively. Using time as a continuous variable and the log of the hazard ratios (HRs) derived from the Cox model, a negative, statistically nonsignificant linear relationship was observed between time to PORT initiation and survival (10-day HR, 1.004, 95% CI, 0.949-1.062, P = .90) (Figure 4); thus, the data did not support a significant association between delayed start to PORT and survival. Replicating this analysis with PORT and cCRT treatment cohorts individually yielded the same outcome, with no estimated survival benefit associated with PORT initiation timing (10-day HR, 1.000, 95% CI, 0.937-1.068, P = .99 for PORT; 10-day HR, 1.018, 95% CI, 0.897-1.157, P = .78 for cCRT). Given the linear relationship between time and the log HR, time to radiotherapy initiation was neither dichotomized according to a single cut-point nor evaluated within discrete time intervals.
Figure 4Modeling the relationship between time to initiation of PORT and mortality risk for cCRT and PORT using the log of the HRs derived from a multivariable Cox model, shown on the Y-axis. The 95% CIs of the adjusted HRs are represented by the shaded area (10-day HR, 1.004, 95% CI, 0.949-1.062, P = .90).
The median time to PORT initiation was 49.0 days (interquartile range [IQR], 42-67) for the combined PORT + cCRT group. For PORT, the median time was 55.5 days (IQR, 42-75); for cCRT, the median time was 48.0 days (IQR, 39-61).
Discussion
The current study of 1713 patients from the NCDB with treatment-naïve cT1-4N0M0 pN0 NSCLC and positive margins addresses a current gap in knowledge in which there is a paucity of data to support current therapeutic recommendations. Across the entire cohort, ACT alone was the only treatment statistically associated with an improvement in survival compared with SA. ACT also provided a significant survival benefit compared with PORT (Figure 5). Further, there was no evidence that radiotherapy-inclusive treatment provided any additional survival benefit compared with SA or ACT. For combination adjuvant radiotherpay and chemotherapy modalities, neither concurrent nor sequential therapies were found to be superior. These results were reinforced by the propensity-matched analyses, demonstrating overall equivalence between ACT and cCRT, and clinical T-stage analyses, whereby ACT was equivalent to radiotherapy-inclusive treatment in cT1 and cT3+cT4 and at least equivalent, if not better, in cT2 disease. This suggests that tumor size impacted survival for ACT. Although not the same tumor dimensions (ie, 3-5 cm), these results harken to the CALGB 9633 trial that identified strong overall survival signals favoring ACT for tumors greater than 4 cm.
Adjuvant paclitaxel plus carboplatin compared with observation in stage IB non-small-cell lung cancer: CALGB 9633 with the cancer and Leukemia group B, radiation therapy oncology group, and North Central cancer treatment group study groups.
Figure 5Summary of the methods, results, and implications of this study. NSCLC, Non–small cell lung cancer; NCDB, National Cancer Database; 5-yr, 5-year overall survival.
In this analysis, the use of PORT alone was not associated statistically with any survival benefit compared with SA, and in fact was estimated to lead to a worse 5-year survival. If confirmed by a future study with a larger sample, this lack of efficacy would be inconsistent with many studies that have recommended the use of PORT in the settings of margin-negative and incompletely resected N0-2 disease.
However, the current study is consistent with other literature in not showing a statistically significant survival impact of PORT in both margin-positive and node-negative disease.
Although the findings regarding the relative efficacies of PORT and ACT might challenge the existing National Comprehensive Cancer Network guidelines, which do not recommend ACT alone for any margin-positive patients,
NCCN Clinical Practice Guidelines in Oncology (NCCN Guidelines): Non-Small Cell Lung Cancer. National Comprehensive Cancer Network (NCCN). Version 3.2021. 2021
who concluded that ACT alone can be more effective compared with PORT alone. The current study goes further with individual comparisons of the different adjuvant treatments, especially unimodal versus multimodal, and survival analysis segmentation by clinical T stage. Taken altogether, these results highlight that chemotherapy has more consistently shown better overall survival than PORT, even in N0 patients.
When assessing the multimodal combination therapies, neither cCRT nor sequential therapy (ie, sCRT + sRTC, combined due to small individual sample sizes) provided any significant survival advantage over SA. Although merely a failure to reject the null hypothesis, this finding is different from some studies that identified a benefit in concurrent over sequential chemoradiotherapy
Phase III study of concurrent versus sequential thoracic radiotherapy in combination with mitomycin, vindesine, and cisplatin in unresectable stage III non-small-cell lung cancer.
In the context of ACT's relative advantage or equivalent efficacy, if the focus is shifted to comparing the efficacy between multimodal and unimodal treatment modalities instead of only multimodal, then the statistically equivalent survival benefit of ACT may prompt a reconsideration to use a unimodal option moving forward.
used the NCDB to define the impact of relative timing of postoperative radiotherapy on overall survival. The study focused on pN2, margin-negative disease and used a dichotomized approach in the analysis with therapy initiation defined as less than or greater than 8 weeks. They found that a longer time to radiotherapy, defined as 8 weeks or more, was associated with improved survival. A similar study has been performed for ACT timing.
For the current study's sample, rather than dichotomizing the patient population according to a single cut-point, a continuous multivariable Cox model was created, which demonstrated that delays in PORT were not associated significantly with a survival detriment. Most mentions of radiotherapy timing are in discussions of sequencing when both ACT and radiotherapy are indicated. Although equivalence testing with a larger sample might be needed to show the absence of an effect, the possible lack of detriment to survival could provide patients the opportunity to delay their radiotherapy in accordance with a slower postoperative recovery or in the context of limited access to radiation oncology. If postoperative radiotherapy is needed, then a time constraint would not impact its efficacy.
The landscape of available treatments may soon shift now that landmark studies such as Checkmate 816, ADAURA, and IMpower010 have demonstrated the benefit of using targeted therapy and immunotherapy in both neoadjuvant and adjuvant settings. Although certainly ushering in new NSCLC treatment paradigms, the focus on disease-free survival, neoadjuvant, and margin-negative disease makes their conclusions less applicable to the current study.
Adjuvant atezolizumab after adjuvant chemotherapy in resected stage IB-IIIA non-small-cell lung cancer (IMpower010): a randomised, multicentre, open-label, phase 3 trial.
Unfortunately, only a limited number of patients in the NCDB fit this study's inclusion criteria and received adjuvant immunotherapy. Thus, the relative efficacy and synergy of these new treatments could not be explored.
The observation that chemotherapy provides a consistent benefit over radiotherapy in the context of positive margins, in this sample, is somewhat counterintuitive given that positive margins represent a risk for local recurrence or invasion, which theoretically would be treated by a local therapy such as radiotherapy. In the case of postoperative locoregional recurrence, radiotherapy is widely considered standard of care. However, there is evidence suggesting that the addition of chemotherapy provides an additional survival benefit compared with radiotherapy alone, a potential explanation being that local treatments are limited by the risk of distant failure, which can be mitigated by inclusion of chemotherapy.
Nonetheless, the results outlined add support for the reevaluation of current recommendations regarding adjuvant therapy.
Study Limitations
There are several limitations associated with using the NCDB for this retrospective study. The NCDB does not capture many important patient and tumor attributes regarding clinical reasoning that may have dictated the selection and sequencing of adjuvant therapy. For instance, these data could help elucidate the clinical rationale behind 1027 patients, 60.0% of this study's total sample, undergoing SA in the context of positive margins. Therefore, the assignment of patients to treatment groups may be susceptible to uncontrolled confounding factors.
Perhaps most limiting is that the NCDB does not capture re-resections after initial resection, which is a locoregional treatment option for many patients with positive margins.
NCCN Clinical Practice Guidelines in Oncology (NCCN Guidelines): Non-Small Cell Lung Cancer. National Comprehensive Cancer Network (NCCN). Version 3.2021. 2021
The routine performance of re-resections for positive margins is unclear in the present day given the confluence of a thoracic surgical oncology era that emphasizes minimally invasive surgery and is witnessing improved systemic therapies. Although the NCDB does not capture frozen-section data that would drive re-resection at the time of the index operation, the notion of re-resection hints at the return to the operating room in a delayed fashion based on a final pathology result for a substantial operation that would likely occur with low frequency in the current epoch. In addition, the NCDB does not distinguish whether specific treatments are therapeutic or salvage in nature nor does it identify the cause of death, both of which could potentially help assess the relative benefit and underlying disease progression of each treatment group. Moreover, the NCDB lacks both disease-free and disease-specific survival data, limiting any survival analysis to overall survival only. Therefore, survival analysis may be susceptible to confounding factors such as age and underlying health. However, using a Charlson-Deyo score of 0 as a proxy for disease-specific survival, a subgroup analysis revealed no statistically significant differences in survival among the adjuvant therapy cohorts, thus decreasing the likelihood that patients' underlying health served as a confounder for the survival differences presented in this article.
Other limitations of the study pertain to the small cohort sizes, especially for the sequential therapy treatment groups. These small cohorts were accentuated when comparing adjuvant therapy cohorts sRTC and sCRT, which limited the analysis of the multimodal therapy options. With the P value adjustment method used (Bonferroni), the large number of treatment groups increased the risk of type I error associated with multiple comparisons, with future prospective studies needed to corroborate the findings of this study. In addition, larger sample sizes would help with using equivalence testing to detect the absence of associations or negligible associations. It is also possible for patients to undergo adjuvant therapy at a different, non-NCDB facility and thus would not be captured in this study's cohorts.
Conclusions
Chemotherapy alone was the only adjuvant therapy that consistently delivered a significant survival improvement to patients with treatment-naïve cT1-4N0M0 pN0 disease and positive surgical margins. Radiotherapy alone was not found to be significantly associated with survival benefit compared with SA. Rather, PORT may serve as an adjunct to modern systemic therapies after incomplete resection with the goal of achieving local disease control and without the expectation that PORT will impact survival. At the same time, delays to adjuvant radiotherapy initiation were not associated significantly with a reduction in survival. There was no significant statistical difference between the concurrent and sequential combination therapy cohorts, and neither provided a statistically significant survival benefit over SA. Taken together, the findings of this study suggest a potential paradigm shift in the care of margin-positive, node-negative patients, as the rapidly expanding evolution of targeted therapies and immunotherapies may also add more dimensions to consider.
Conflict of Interest Statement
The authors reported no conflicts of interest.
The Journal policy requires editors and reviewers to disclose conflicts of interest and to decline handling or reviewing manuscripts for which they may have a conflict of interest. The editors and reviewers of this article have no conflicts of interest.
References
Detterbeck F.C.
Lewis S.Z.
Diekemper R.
Addrizzo-Harris D.
Alberts W.M.
Executive summary: diagnosis and management of lung cancer, 3rd ed: American College of Chest Physicians evidence-based clinical practice guidelines.
NCCN Clinical Practice Guidelines in Oncology (NCCN Guidelines): Non-Small Cell Lung Cancer. National Comprehensive Cancer Network (NCCN). Version 3.2021. 2021
Postoperative radiotherapy versus no postoperative radiotherapy in patients with completely resected non-small-cell lung cancer and proven mediastinal N2 involvement (Lung ART): an open-label, randomised, phase 3 trial.
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.
Adjuvant paclitaxel plus carboplatin compared with observation in stage IB non-small-cell lung cancer: CALGB 9633 with the cancer and Leukemia group B, radiation therapy oncology group, and North Central cancer treatment group study groups.
Phase III study of concurrent versus sequential thoracic radiotherapy in combination with mitomycin, vindesine, and cisplatin in unresectable stage III non-small-cell lung cancer.
Adjuvant atezolizumab after adjuvant chemotherapy in resected stage IB-IIIA non-small-cell lung cancer (IMpower010): a randomised, multicentre, open-label, phase 3 trial.
J.Y. is supported by Grants UL1TR001855 and UL1TR000130 from the National Center for Advancing Translational Science of the US National Institutes of Health. Funding was also provided by the Jeffrey P. Smith Endowed Chair in Surgery Fund.
The Institutional Review Board or equivalent ethics committee of the University of Southern California approved the study protocol and publication of data (Number HS-16-00906; approval date: December 19, 2016). Patient written consent for the publication of the study data was waived by the Institutional Review Board because this study is not considered human subjects research.
00033-5&id=fx1.jpg){kind=link}
00033-5&id=fx2.jpg){kind=link}
00033-5&id=gr1.jpg){kind=link}
00033-5&id=gr2.jpg){kind=link}
00033-5&id=gr3.jpg){kind=link}
00033-5&id=gr4.jpg){kind=link}
00033-5&id=gr5.jpg){kind=link}