Interactive CardioVascular and Thoracic Surgery (2019) 1–8 ORIGINAL ARTICLE

doi:10.1093/icvts/ivz140

Cite this article as: Lutfi W, Schuchert MJ, Dhupar R, Sarkaria I, Christie NA, Yang C-FJ et al. Sublobar resection is associated with decreased survival for patients with
early stage large-cell neuroendocrine carcinoma of the lung. Interact CardioVasc Thorac Surg 2019; doi:10.1093/icvts/ivz140.

THORACIC
Sublobar resection is associated with decreased survival
for patients with early stage large-cell neuroendocrine
carcinoma of the lung

Downloaded from https://academic.oup.com/icvts/advance-article-abstract/doi/10.1093/icvts/ivz140/5513096 by guest on 23 July 2019

Waseem Lutfi a, Matthew J. Schucherta, Rajeev Dhupar a, Inderpal Sarkariaa, Neil A. Christiea,
Chi-Fu J. Yangb, John Z. Deng c, James D. Luketicha and Olugbenga T. Okusanyaa,*
a
Department of Cardiothoracic Surgery, University of Pittsburgh Medical Center, Pittsburgh, PA, USA
b
Department of Cardiothoracic Surgery, Stanford University Medical Center, Stanford, CA, USA
c
Division of Cardiothoracic Surgery, Duke University Medical Center, Durham, NC, USA

* Corresponding author. Department of Cardiothoracic Surgery, University of Pittsburgh Medical Center, 200 Lothrop St, Pittsburgh, PA 15213, USA.
Tel: 4126232025; e-mail: okusanyaot@upmc.edu (O.T. Okusanya).

Received 27 December 2018; received in revised form 16 April 2019; accepted 1 May 2019

Abstract
OBJECTIVES: Sublobar resection (SLR) for early non-small-cell lung carcinoma (NSCLC) has been shown to have a survival rate similar to
that of lobectomy. Large-cell neuroendocrine carcinoma (LCNEC) of the lung, although treated like an NSCLC, has a poor prognosis com-
pared to NSCLC. We sought to determine if outcomes are poor in patients with early stage LCNEC treated with SLR versus lobectomy.
METHODS: We searched for patients with pathological stage I LCNEC <_3 cm within the National Cancer Database between 2004 and
2014. Propensity score matching was used to compare the 5-year overall survival rate of patients having SLR (wedge or segmentectomy)
to that of patients having a lobectomy. Patients were matched for age, node sampling, comorbidity score, tumour size, insurance status
and other factors. Patients who received neoadjuvant therapy were excluded. Kaplan–Meier methods were used for analysis.

C The Author(s) 2019. Published by Oxford University Press on behalf of the European Association for Cardio-Thoracic Surgery. All rights reserved.
V
2 W. Lutfi et al. / Interactive CardioVascular and Thoracic Surgery

RESULTS: A total of 1011 patients met the inclusion criteria: 263 were treated with SLR (223 wedges and 40 segmentectomies) and 748
patients, with lobectomy. Patients who received SLR were older, had more comorbidities and smaller tumours. On unadjusted Kaplan–
Meier analysis, patients who had SLR had decreased 5-year overall survival compared to those who had a lobectomy (37.9% vs 56.6%,
P < 0.001). Propensity score matching (1:1) across 12 demographic and tumour variables yielded 185 patients per group with 34 segmen-
tectomies and 151 wedge resections in the SLR cohort. On Kaplan–Meier analysis of the matched cohort, patients who had SLR had a
worse 5-year overall survival rate compared to those who had a lobectomy (41.5% vs 60.3%; P = 0.001).
CONCLUSIONS: SLR for early stage LCNEC is associated with a lower 5-year overall survival rate compared to lobectomy on unadjusted
and propensity matched analyses.
Keywords: Large-cell neuroendocrine carcinoma • Sublobar resection • Lobectomy • Overall survival • Lung cancer

Downloaded from https://academic.oup.com/icvts/advance-article-abstract/doi/10.1093/icvts/ivz140/5513096 by guest on 23 July 2019

INTRODUCTION
Large-cell neuroendocrine carcinoma (LCNEC) is a rare tumour
that accounts for approximately 3% of all lung cancer diagnoses
[1]. Originally these tumours were classified as large-cell carcino-
mas within the non-small-cell lung carcinoma (NSCLC) category.
However, in 2015, these tumours were no longer classified with
large-cell carcinomas and were grouped with neuroendocrine
carcinomas such as carcinoid and small-cell [2]. Despite this
reclassification, treatment guidelines do not differ for LCNEC and
NSCLC, even though, compared to NSCLC, LCNEC has a poor
prognosis [3, 4]. Especially for early stage disease there is contro-
versy as to optimal treatment regarding the use of surgery or ad-
juvant chemotherapy, given the aggressive nature of the cancer
[5–8]. The survival rate for patients with early stage LCNEC has
been reported to be as low as 33% [9].
Sublobar resection (SLR) (wedge resection or segmentectomy)
has gained recent support as a viable therapy for early stage
NSCLC. Specifically, several retrospective studies have shown that
for select patients with NSCLC who have a segmentectomy, local
and distant recurrence rates as well disease-free survival and
overall survival rates are not statistically different from those of
patients who have a lobectomy [10–15]. Also, for some early
adenocarcinomas, wedge resection has been shown to have a
high cure rate [16, 17]. These techniques have the advantage of
preserving lung parenchyma but theoretically may increase local
recurrence [18]. However, a debate remains about the appropri-
ateness of SLR as a broadly applicable standard. Specifically, it is
unknown if LCNEC is appropriately treated with SLR. In addition,
many patients have a resection using SLR techniques for diagno-
sis and therapy in the same setting and may not be considered
for lobectomy. Given the generally poor overall survival rate
from LCNEC and the rising utilization of SLR in early stage cancer,
we used the most recent national data to examine the effect on
survival when SLR is performed for early LCNEC.
METHODS
Data source, patient selection and variables studied
The National Cancer Database (NCDB) is a nationwide facility-
based dataset that currently captures 70% of all newly diagnosed
cancers in the United States annually reported from approxi-
mately 1500 hospitals with Commission on Cancer accredited
cancer programmes [19]. The American College of Surgeons and
the Commission on Cancer have not verified and are not
responsible for the analytical or statistical methods used or for
the conclusions drawn from these data by the investigators.
Patients diagnosed with American Joint Commission on
Cancer pathological stage I LCNEC <_3 cm treated surgically with
wedge resection, segmentectomy or lobectomy between 2004
and 2014 were selected for this study. The pathological stage of
the patient’s tumour was defined in the NCDB as the patient’s
American Joint Commission on Cancer stage determined from
surgical specimens after resection. Patients for whom there were
no data on hospital category, chemotherapy, tumour location or
lymph node sampling and those with other malignancies, neoad-
juvant therapy and positive surgical margins were excluded from
analysis. Inclusion criteria are further detailed in Fig. 1.
Surgical resection was stratified into 2 categories: SLR (wedge
resection and segmentectomy) and lobectomy. Patient factors
studied were age, sex, race, zip code-based income, insurance
status and Charlson–Deyo Comorbidity Index. Facility character-
istics included facility type and patient distance to the treating fa-
cility. Tumour factors included tumour size and tumour location.
Treatment variables included lymph node sampling and chemo-
therapy; chemotherapy included single-agent and multiagent
protocols.
Statistical analyses
Patients diagnosed with pathological T1N0M0 LCNEC were
grouped according to whether they received surgical treatment
with SLR or lobectomy. First an unadjusted analysis was per-
formed. Baseline patient, hospital and clinical characteristics for
the SLR and lobectomy groups were evaluated using the
Wilcoxon rank sum test and Pearson’s v2 test as appropriate.
Kaplan–Meier methods with log-rank tests were used to analyse
overall survival. Hazard ratios (HRs) were also generated using
Cox proportional hazard regression and reported with a 95%
confidence interval (CI). A 1-to-1 propensity score matching for
12 patient, hospital and clinical variables between SLR and lobec-
tomy was then conducted using the caliper-matching method
(Table 1). We imposed a 0.01 propensity score tolerance on the
maximum propensity score distance (caliper) in our algorithm to
optimize matching results. After matching, Kaplan–Meier meth-
ods were used to analyse overall survival in our matched cohort.
An additional analysis was performed to ascertain the impact
of histological analysis of LCNEC on overall survival after SLR.
A cohort of patients with adenocarcinoma of the lung treated
with SLR were identified within the NCDB and matched to our
cohort with LCNEC that received SLR using the same algorithm
and covariates in the aforementioned propensity match.
 W. Lutfi et al. / Interactive CardioVascular and Thoracic Surgery
Figure 1: Inclusion criteria. LCNEC: large-cell neuroendocrine carcinoma.
Kaplan–Meier methods were used to analyse overall survival be-
tween the 2 groups.
Statistical analyses were performed using Stata/SE 15.1 statisti-
cal software (College Station, TX, USA). All tests were 2-sided, and
a P-value <0.05 was considered statistically significant.
RESULTS
Sublobar resection versus lobectomy
A total of 1011 patients were identified who met the inclusion
criteria: 263 (26.0%) received SLR (223 wedge resection and 40
segmentectomy) and 748 (74.0%) received lobectomy. Table 1
details the baseline characteristics. Compared to patients receiv-
ing a lobectomy, patients receiving SLR were more likely to have
increased age, have Medicare insurance and have smaller
3
THORACIC
Downloaded from https://academic.oup.com/icvts/advance-article-abstract/doi/10.1093/icvts/ivz140/5513096 by guest on 23 July 2019
tumours and were less likely to have a comorbidity index score
of 0 and lymph node sampling than those receiving lobectomy.
On Kaplan–Meier analysis, patients who had SLR had a de-
creased 5-year overall survival rate compared to patients who re-
ceived a lobectomy (37.8% vs 56.7%; P < 0.001) (Fig. 2), with an
HR of 1.66 (95% CI 1.38–2.00) for SLR compared to lobectomy.
Individually, segmentectomy and wedge resection were associ-
ated with a worse 5-year overall survival rate compared to lobec-
tomy (segmentectomy: 33.2% vs 56.7%; P = 0.003; HR 1.80, 95%
CI 1.22–2.66) (Fig. 3) (wedge resection: 38.6% vs 56.7%; P < 0.001;
HR 1.64, 95% CI 1.34–2.00) (Fig. 3).
Propensity score analysis
After propensity score matching, 370 patients were identified,
with 185 patients in both the SLR (34 segmentectomies and 151
4 W. Lutfi et al. / Interactive CardioVascular and Thoracic Surgery

Table 1: Patient demographics: prematch and matched cohorts

Characteristic Prematch Matched

SLR Lobectomy P-value SLR Lobectomy P-value

Numbers of patients, n (%) 263 (26.0) 748 (74.0) 185 (50.0) 185 (50.0)
Age (years), median (IQR) 70 (63–76) 66 (59–72) <0.001 69 (63–76) 69 (62–76) 0.959
Year of diagnosis, n (%) 0.295 0.255
2004 23 (8.7) 57 (7.6) 18 (9.7) 11 (6.0)
2005 18 (6.8) 63 (8.4) 13 (7.0) 10 (5.4)

Downloaded from https://academic.oup.com/icvts/advance-article-abstract/doi/10.1093/icvts/ivz140/5513096 by guest on 23 July 2019

2006 26 (9.9) 69 (9.2) 15 (8.0) 19 (10.4)
2007 17 (6.5) 55 (7.4) 12 (6.5) 18 (9.7)
2008 23 (8.8) 74 (9.9) 19 (10.4) 12 (6.4)
2009 19 (7.2) 85 (11.4) 12 (6.5) 18 (9.7)
2010 21 (8.0) 76 (10.2) 11 (6.0) 22 (11.9)
2011 24 (9.0) 65 (8.7) 17 (9.2) 19 (10.4)
2012 27 (10.3) 55 (7.4) 21 (11.3) 16 (8.6)
2013 28 (10.7) 79 (10.5) 21 (11.3) 22 (11.8)
2014 37 (14.1) 70 (9.3) 26 (14.1) 18 (9.7)
Sex, n (%) 0.856 1.000
Male 121 (46.0) 349 (46.7) 86 (46.5) 86 (46.5)
Female 142 (54.0) 399 (53.3) 99 (53.5) 99 (53.5)
Race, n (%) 0.918 0.592
White 234 (89.0) 664 (88.8) 163 (88.1) 159 (86.0)
Black 21 (8.0) 64 (8.5) 16 (8.7) 16 (8.7)
Other 8 (3.0) 20 (2.7) 6 (3.2) 10 (5.4)
Median income, n (%) 0.979 0.676
<$38 000 52 (20.0) 148 (20.2) 39 (21.1) 40 (21.6)
$38 000–$47 999 61 (23.5) 163 (22.2) 39 (21.1) 46 (24.9)
$48 000–$62 999 73 (28.1) 212 (28.8) 59 (31.9) 60 (32.4)
>_$63 000 74 (28.4) 212 (28.8) 48 (26.0) 39 (21.1)
Insurance, n (%) 0.009 0.745
Private 62 (23.6) 238 (31.8) 50 (27.0) 56 (30.3)
Medicare 183 (69.6) 440 (58.8) 125 (67.6) 118 (63.8)
None/other 18 (6.8) 70 (9.4) 10 (5.4) 11 (6.0)
Comorbidity index, n (%) 0.001 0.994
0 75 (28.5) 313 (41.8) 55 (29.7) 55 (29.7)
1 125 (47.5) 313 (41.8) 91 (49.2) 90 (48.7)
2 50 (19.0) 95 (12.7) 33 (17.8) 33 (17.8)
>_3 13 (5.0) 27 (3.7) 6 (3.2) 7 (3.8)
Facility type, n (%) 0.163 0.554
Community 13 (4.9) 65 (8.7) 8 (4.3) 8 (4.3)
Comprehensive 126 (47.9) 329 (44.0) 86 (46.5) 90 (48.7)
Academic 91 (34.6) 274 (36.7) 72 (38.9) 61 (33.0)
Integrated cancer network 33 (12.6) 79 (10.6) 19 (10.3) 26 (14.1)
Miles to facility, median (IQR) 10.9 (4.7–25.9) 11.2 (5.3–28.7) 0.307 10.5 (4.7–25.3) 11.8 (5.6–31.6) 0.156
Tumour size (cm), median (IQR) 1.8 (1.4–2.2) 2.0 (1.5–2.5) <0.001 1.7 (1.3–2.3) 1.8 (1.3–2.3) 0.493
Tumour location, n (%) 0.177 0.760
Upper right lobe 82 (31.2) 284 (38.0) 60 (32.4) 58 (31.4)
Middle right lobe 9 (3.4) 35 (4.7) 7 (3.8) 4 (2.2)
Lower right lobe 41 (15.6) 119 (15.9) 28 (15.1) 23 (12.4)
Upper left lobe 89 (33.8) 211 (28.2) 62 (33.5) 69 (37.3)
Lower left lobe 42 (16.0) 99 (13.2) 28 (15.1) 31 (16.8)
Lymph node resection, n (%) <0.001 0.586
No 73 (27.8) 9 (1.2) 6 (3.2) 8 (4.3)
Yes 190 (72.2) 739 (98.8) 179 (96.8) 177 (95.7)
IQR: interquartile range; SLR: sublobar resection.

wedge resections) and lobectomy groups. A comparison of the
12 preoperative variables between the matched groups is shown
in Table 1. Matching yielded no statistically significant differences
between SLR and lobectomy in any of the 12 covariates, confirm-
ing a good balance between the 2 groups. After matching, rates
of adjuvant chemotherapy utilization were found to be similar at
15.7% in patients who had SLR and 16.8% in patients who had a
lobectomy (P = 0.96). Patients receiving SLR had decreased over-
all survival compared to those receiving a lobectomy (P = 0.001)
(Fig. 4) with an HR of 1.59 (95% CI 1.20–2.12) for SLR compared
to lobectomy. The 5-year overall survival rates were 41.5% (95%
CI 33.0–49.7%) for SLR and 60.3% (95% CI 52.1–67.6%) for
lobectomy.
Sublobar resection in large-cell neuroendocrine
carcinoma versus adenocarcinoma
A cohort of patients with adenocarcinomas who received SLR
were identified within the NCDB with the same inclusion criteria
 W. Lutfi et al. / Interactive CardioVascular and Thoracic Surgery 5

THORACIC
Downloaded from https://academic.oup.com/icvts/advance-article-abstract/doi/10.1093/icvts/ivz140/5513096 by guest on 23 July 2019
Figure 3: Unadjusted Kaplan–Meier curves for wedge resection, segmentec-
Figure 2: Kaplan–Meier survival curves for the unmatched cohort: Patients with
tomy and lobectomy. Both segmentectomy and wedge resection individually
early stage large-cell neuroendocrine carcinoma of the lung who had a lobec-
are associated with decreased 5-year survival rates compared to lobectomy.
tomy have improved 5-year survival rates compared to those who had an SLR.
SLR: sublobar resection.

as described in Fig. 1. A similar 1:1 propensity match (including

matching for wedge resection and segmentectomy) between
patients with adenocarcinoma and LCNEC who received SLR was
performed and yielded a balanced match (Table 2). After propen-
sity score matching, 518 patients were identified with 259
patients in both the LCNEC and adenocarcinoma groups.
Kaplan–Meier analysis of the matched cohort demonstrated
worse survival for the patients with LCNEC (P < 0.001; Fig. 5) with
an HR of 1.93 (95% CI 1.52–2.47) for LCNEC compared to adeno-
carcinoma. The 5-year overall survival rates were 38.0% (95% CI
31.1–44.9%) for LCNEC and 63.4% (95% CI 56.5–69.4%) for
adenocarcinoma.

DISCUSSION
LCNEC is believed stage for stage to be a more lethal cancer than
other NSCLCs [20, 21]. Investigations have been made in treating
even early stage LCNEC with adjuvant chemotherapy [22].
However, no current guidelines indicate that surgical treatment
for LCNEC differs from that for NSCLC. In both unadjusted and
propensity matched analyses, this study found that SLR was asso-
ciated with a decreased 5-year overall survival rate compared to
lobectomy for patients with LCNEC. We found a 5-year overall
survival rate of 49% for node negative LCNEC less than 3 cm
(stage IA3) treated with SLR. This is the largest retrospective study
to date, using the most recent national data, to show that SLR is
associated with increased mortality in patients with LCNEC.
This study demonstrated a 5-year overall survival rate of less
than 60% after lobectomy for LCNEC, which was consistent with
results from other published series [6, 23–26]. However, few stud-
ies have yet to evaluate long-term outcomes after SLR. In one of
the largest single-institution studies to evaluate surgical resection
of LCNEC, Veronesi et al. compared results from 15 SLRs to those
from 95 lobectomies and found no difference in survival with a
5-year overall survival rate of 52% for stage I disease [23].
Another study by Filosso et al. found no difference in survival be-
tween 24 SLRs and 81 lobectomies with a 5-year overall survival
rate of 42% for overall stage I disease [26]. Although these studies
demonstrated no difference in the 5-year overall survival rate
when LCNECs were treated with SLR or lobectomy, they are lim-
ited by their small sample size and are likely underpowered to
Figure 4: Kaplan–Meier survival curves for the matched cohort: after propen-
sity matching, the improved survival of patients who had a lobectomy persists.
SLR: sublobar resection.
identify statistical differences in survival. Both of these studies in-
cluded patients who were stage I through III. From their respec-
tive published data the exact stages of the patients who had SLR
and lobectomy are unclear. Additionally, all stages were included
in their respective multivariable Cox regressions that included re-
section type as a covariate. Given that our cohort included only
stage I patients, this may be another reason to further explain
why the conclusions from these other studies were different from
ours.
Other large NCDB studies have shown the 5-year overall sur-
vival rate for SLR with node sampling for NSCLC less than 3 cm
to be 58%, which would be much higher than the survival we
found for LCNEC treated with SLR [27]. The strength of our analy-
sis is the large sample size allowing enough power to propensity
match our data and identify a lower overall survival associated
with SLR compared to lobectomy.
One weakness of this analysis is the inability to control for pul-
monary function testing and other factors that would lead a sur-
geon to select SLR because these variables are not available
within the NCDB. In an attempt to mitigate the effect of selection
bias to perform an SLR, we performed a second propensity
match to patients who received SLR for lung adenocarcinoma.
When compared to SLR for adenocarcinoma, SLR for LCNEC was
6 W. Lutfi et al. / Interactive CardioVascular and Thoracic Surgery

Table 2: Patient demographics for sublobar resection in large-cell neuroendocrine carcinoma and adenocarcinoma: prematch and
matched cohorts

Characteristic Prematch Matched

LCNEC Adenocarcinoma P-value LCNEC Adenocarcinoma P-value

Numbers of patients, n (%) 263 (3.2) 804 (96.8) 259 259

Age (years), median (IQR) 70 (63–76) 71 (65–77) 0.030 70 (63–76) 69 (62–76) 0.370
Year of diagnosis, n (%) 0.467 0.194

Downloaded from https://academic.oup.com/icvts/advance-article-abstract/doi/10.1093/icvts/ivz140/5513096 by guest on 23 July 2019

2004 23 (8.8) 466 (5.8) 22 (8.5) 26 (10.0)
2005 18 (6.8) 556 (6.9) 18 (7.0) 18 (7.0)
2006 26 (9.9) 579 (7.2) 25 (9.7) 22 (8.5)
2007 17 (6.5) 657 (8.2) 17 (6.6) 27 (10.4)
2008 23 (8.8) 657 (8.2) 23 (8.9) 23 (8.9)
2009 19 (7.2) 712 (8.9) 18 (7.0) 30 (11.6)
2010 21 (8.0) 733 (9.1) 21 (8.1) 22 (8.5)
2011 24 (9.1) 879 (10.9) 24 (9.2) 15 (5.8)
2012 27 (10.3) 808 (10.1) 27 (10.4) 20 (7.7)
2013 28 (10.6) 947 (11.7) 27 (10.4) 34 (13.1)
2014 37 (14.0) 1047 (13.0) 37 (14.2) 22 (8.5)
Sex, n (%) 0.304 0.333
Male 121 (46.0) 3443 (42.8) 119 (46.0) 130 (50.2)
Female 142 (54.0) 4598 (57.2) 140 (54.0) 129 (49.8)
Race, n (%) 0.510 0.045
White 234 (89.0) 7156 (89.0) 230 (88.8) 231 (89.2)
Black 21 (8.0) 547 (6.8) 21 (8.1) 27 (10.4)
Other 8 (3.0) 338 (4.2) 8 (3.1) 1 (0.4)
Median income, n (%) 0.170 0.652
<$38 000 52 (20.0) 1425 (17.9) 52 (20.1) 63 (24.3)
$38 000–$47 999 61 (23.5) 1754 (22.1) 61 (23.6) 62 (23.9)
$48 000–$62 999 73 (28.1) 1981 (24.9) 72 (27.8) 68 (26.3)
>_$63 000 74 (28.4) 2792 (35.1) 74 (28.5) 66 (25.5)
Insurance, n (%) 0.859 0.429
Private 62 (23.6) 1841 (22.9) 61 (23.6) 74 (28.6)
Medicare 183 (69.6) 5704 (70.9) 180 (69.4) 168 (64.8)
None/other 18 (6.8) 496 (6.2) 18 (7.0) 17 (6.6)
Comorbidity index, n (%) 0.002 0.676
0 75 (28.5) 3213 (40.0) 75 (29.0) 63 (24.3)
1 125 (47.5) 3214 (40.0) 122 (47.1) 133 (51.4)
2 50 (19.0) 1210 (15.0) 49 (18.9) 50 (19.3)
>_3 13 (5.0) 404 (5.0) 13 (5.0) 13 (5.0)
Facility type, n (%) 0.009 0.055
Community 13 (4.9) 511 (6.4) 13 (5.0) 9 (3.5)
Comprehensive 126 (47.9) 3272 (40.8) 124 (47.9) 129 (49.8)
Academic 91 (24.6) 3500 (43.6) 89 (34.4) 69 (26.6)
Integrated cancer network 33 (12.6) 746 (9.3) 33 (12.7) 52 (20.1)
Miles to facility, median (IQR) 10.9 (4.7–25.9) 10.9 (4.5–27.7) 0.843 10.9 (4.7–25.9) 11.5 (4.5–28.2) 0.802
Tumour size (cm), median (IQR) 1.8 (1.4–2.2) 1.7 (1.2–2.2) 0.060 1.8 (1.4–2.2) 1.8 (1.5–2.4) 0.078
Tumour location, n (%) 0.355 0.442
Upper right lobe 82 (31.2) 2637 (32.8) 81 (31.3) 75 (29.0)
Middle right lobe 9 (3.4) 266 (3.3) 9 (3.5) 17 (6.6)
Lower right lobe 41 (15.6) 1543 (19.2) 39 (15.1) 33 (12.7)
Upper left lobe 89 (33.8) 2298 (28.6) 89 (34.4) 97 (37.4)
Lower left lobe 42 (16.0) 1297 (16.1) 41 (15.7) 37 (14.3)
Lymph node resection, n (%) 0.965 0.768
No 73 (27.8) 2222 (27.6) 73 (29.2) 70 (27.0)
Yes 190 (72.2) 5819 (72.4) 186 (71.8) 189 (73.0)
IQR: interquartile range; LCNEC: large-cell neuroendocrine carcinoma.

associated with decreased 5-year overall survival rate, which is in
keeping with the established knowledge that LCNEC is a poor
histological subtype of NSCLC [3]. The decreased survival for
patients with LCNEC treated with SLR compared to lobectomy
may be partly due to the curative value of lobectomy or the po-
tential or unmeasured confounders such as death related to dis-
eases associated with the decision to perform SLR and the ability
for a patient to tolerate lobectomy. However, this analysis
remains imperfect and is hindered by several factors such as the
lack of disease-specific survival and more discrete comorbidity
information. It should be noted that in some of these patients it
is possible that wedge was chosen purely as a diagnostic proce-
dure. We selected patients who had margin-negative resections
and who were matched for lymph node sampling to select for
patients who had what we can discern to be curative intent sur-
gery, especially because they are all early stage. If we are to as-
sume that these procedures were still diagnostic, then a
reasonable conclusion from this study is that a diagnostic wedge
for early stage LCNEC is insufficient and that either frozen analy-
sis should be used in the operating room to guide dissection or
 W. Lutfi et al. / Interactive CardioVascular and Thoracic Surgery
Figure 5: Kaplan–Meier survival curves for sublobar resection in LCNEC and
adenocarcinoma: Compared to patients with an adenocarcinoma, patients
with a LCNEC continued to have worse survival rates with sublobar resection
after propensity matching. LCNEC: large-cell neuroendocrine carcinoma.
patients found to have LCNEC on a diagnostic wedge should be
considered for further resection. Ultimately, data on this histolog-
ical subtype are limited because its incidence is rare, but the
NCDB offers a relatively large but still limited sample for
comparison.
Our understanding of LCNEC continues to grow over time. In
terms of its behaviour, this tumour is much more closely aligned
with small-cell carcinoma than with large-cell carcinoma because
aggressive behaviour and recurrence even with early stage dis-
ease are common. Most LCNEC tumours do not demonstrate ac-
tionable gene mutations such as epidermal growth factor
receptor (EGFR) and anaplastic lymphoma kinase (ALK). There
are reports of up to 22% PD-1 positivity in these tumours, but
the benefit of anti-PD-1/PD-L1 antibodies is unclear [28]. Also, al-
though the tumour responds to platinum-based chemotherapies,
there is still debate about the use of adjuvant chemotherapy.
Thus, ultimately, optimal surgical therapy is critical to the survival
of patients with this disease.
Limitations
This study has several additional limitations. As with any retro-
spective study, there may be unobserved confounders and selec-
tion bias. Specifically, there may be covariates associated with
mortality that play into a surgeon’s decision to select SLR for a
patient, especially wedge resection. Although our propensity
score match included all of the clinically relevant variables avail-
able in the NCDB, the NCDB does not include granular patient
factors such as the results of pulmonary function tests.
Additionally, physician selection bias may be a factor because
surgeons may only reserve lobectomy for the fittest patients and
use SLR for patients who are not fit for a lobectomy. The NCDB
does not provide data on recurrence, disease-free survival,
disease-specific survival and cause of death. This is a major limi-
tation to the final conclusions because overall survival may be
imperfect for studies analysing cancer-related outcomes. Data
with regards to lymph node sampling are limited to the total
number of lymph nodes sampled but do not include information
regarding lymph node stations. We did investigate whether the
number of lymph nodes sampled was predictive of survival;
7
however, this result did not correlate with survival in patients
who had either SLR or lobectomy (data not shown). An intent-
to-treat analysis was precluded in this study because we used
THORACIC
pathological staging for our analysis. Lastly, the histological cod-
ing may allow for the inclusion of patients with tumours that are
only partial LCNEC mixed with other histological subtypes.
Though the NCDB does allow for the selection of patients with
negative margins, it does not give the distance of this margin,
which is believed to be a factor in the recurrence rate for SLR.
Downloaded from https://academic.oup.com/icvts/advance-article-abstract/doi/10.1093/icvts/ivz140/5513096 by guest on 23 July 2019
CONCLUSION
In conclusion, SLR for early stage LCNEC of the lung is associated
with worse overall survival compared to lobectomy. When com-
pared to patients with early stage adenocarcinomas, patients
with LCNEC treated with SLR also do poorly. In a fit patient with
LCNEC who can tolerate a lobectomy, a lobectomy should
strongly be considered as the primary surgical treatment
modality.
Acknowledgments
The authors would like to recognize the Hillman Cancer Center and
Dr. Stanley Marks for their support of this work.
Conflict of interest: none declared.
REFERENCES
[1] Fasano M, Della Corte CM, Papaccio F, Ciardiello F, Morgillo F.
Pulmonary large-cell neuroendocrine carcinoma: from epidemiology to
therapy. J Thorac Oncol 2015;10:1133–41.
[2] Travis WD, Brambilla E, Burke AP, Marx A, Nicholson AG. Introduction
to the 2015 World Health Organization classification of tumors of the
lung, pleura, thymus, and heart. J Thorac Oncol 2015;10:1240–2.
[3] Lo Russo G, Pusceddu S, Proto C, Macerelli M, Signorelli D, Vitali M et al.
Treatment of lung large cell neuroendocrine carcinoma. Tumor Biol
2016;37:7047–57.
[4] Ettinger DS, Wood DE, Aisner DL, Akerley W, Bauman J, Chirieac LR et al.
Non-small cell lung cancer, Version 5.2017, NCCN clinical practice
guidelines in oncology. J Natl Compr Canc Netw 2017;15:504–35.
[5] Kim KW, Kim HK, Kim J, Shim YM, Ahn MJ, Choi YL. Outcomes of
curative-intent surgery and adjuvant treatment for pulmonary large cell
neuroendocrine carcinoma. World J Surg 2017;41:1820–7.
[6] Sarkaria IS, Iyoda A, Roh MS, Sica G, Kuk D, Sima CS et al. Neoadjuvant
and adjuvant chemotherapy in resected pulmonary large cell neuroen-
docrine carcinomas: a single institution experience. Ann Thorac Surg
2011;92:1180–6; discussion 86–7.
[7] Filosso PL, Guerrera F, Evangelista A, Galassi C, Welter S, Rendina EA
et al. Adjuvant chemotherapy for large-cell neuroendocrine lung carci-
noma: results from the European Society for Thoracic Surgeons Lung
Neuroendocrine Tumours Retrospective Database. Eur J Cardiothorac
Surg 2017;52:339–45.
[8] Iyoda A, Hiroshima K, Moriya Y, Takiguchi Y, Sekine Y, Shibuya K et al.
Prospective study of adjuvant chemotherapy for pulmonary large cell
neuroendocrine carcinoma. Ann Thorac Surg 2006;82:1802–7.
[9] Rossi G, Cavazza A, Marchioni A, Longo L, Migaldi M, Sartori G et al.
Role of chemotherapy and the receptor tyrosine kinases KIT,
PDGFRalpha, PDGFRbeta, and Met in large-cell neuroendocrine carci-
noma of the lung. J Clin Oncol 2005;23:8774–85.
[10] Bedetti B, Bertolaccini L, Rocco R, Schmidt J, Solli P, Scarci M.
Segmentectomy versus lobectomy for stage I non-small cell lung cancer:
a systematic review and meta-analysis. J Thorac Dis 2017;9:1615–23.
[11] Kodama K, Higashiyama M, Okami J, Tokunaga T, Imamura F, Nakayama
T et al. Oncologic outcomes of segmentectomy versus lobectomy for
8 W. Lutfi et al. / Interactive CardioVascular and Thoracic Surgery
clinical T1a N0 M0 non-small cell lung cancer. Ann Thorac Surg 2016;
101:504–11.
[12] Moon MH, Moon YK, Moon SW. Segmentectomy versus lobectomy in
early non-small cell lung cancer of 2 cm or less in size: a population-
based study. Respirology 2018;23:695–703.
[13] Song CY, Sakai T, Kimura D, Tsushima T, Fukuda I. Comparison of peri-
operative and oncological outcomes between video-assisted segmentec-
tomy and lobectomy for patients with clinical stage IA non-small cell
lung cancer: a propensity score matching study. J Thorac Dis 2018;10:
4891–901.
[14] Tsutani Y, Miyata Y, Nakayama H, Okumura S, Adachi S, Yoshimura M
et al. Oncologic outcomes of segmentectomy compared with lobectomy
for clinical stage IA lung adenocarcinoma: propensity score-matched
analysis in a multicenter study. J Thorac Cardiovasc Surg 2013;146:
358–64.
[15] Landreneau RJ, Normolle DP, Christie NA, Awais O, Wizorek JJ, Abbas G
et al. Recurrence and survival outcomes after anatomic segmentectomy
versus lobectomy for clinical stage I non-small-cell lung cancer: a
propensity-matched analysis. J Clin Oncol 2014;32:2449–55.
[16] Tsutani Y, Miyata Y, Nakayama H, Okumura S, Adachi S, Yoshimura M
et al. Appropriate sublobar resection choice for ground glass opacity-
dominant clinical stage IA lung adenocarcinoma: wedge resection or
segmentectomy. Chest 2014;145:66–71.
[17] Zhong C, Sakurai H, Wei S, Fang W, Asamura H. Sublobar resections for
small-sized stage Ia lung adenocarcinoma: a Sino-Japanese multicenter
study. J Thorac Dis 2018;10:991–8.
[18] Ginsberg RJ, Rubinstein LV. Randomized trial of lobectomy versus lim-
ited resection for T1 N0 non-small cell lung cancer. Lung Cancer Study
Group. Ann Thorac Surg 1995;60:615–22; discussion 22–3.
[19] Bilimoria KY, Stewart AK, Winchester DP, Ko CY. The National Cancer
Data Base: a powerful initiative to improve cancer care in the United
States. Ann Surg Oncol 2008;15:683–90.
[20] Takei H, Asamura H, Maeshima A, Suzuki K, Kondo H, Niki T et al. Large
cell neuroendocrine carcinoma of the lung: a clinicopathologic study of
eighty-seven cases. J Thorac Cardiovasc Surg 2002;124:285–92.
[21] Battafarano RJ, Fernandez FG, Ritter J, Meyers BF, Guthrie TJ, Cooper JD
et al. Large cell neuroendocrine carcinoma: an aggressive form of non-
small cell lung cancer. J Thorac Cardiovasc Surg 2005;130:166–72.
[22] Kujtan L, Muthukumar V, Kennedy KF, Davis JR, Masood A, Subramanian
J. The role of systemic therapy in the management of stage I large cell
neuroendocrine carcinoma of the lung. J Thorac Oncol 2018;13:707–14.
[23] Veronesi G, Morandi U, Alloisio M, Terzi A, Cardillo G, Filosso P et al.
Large cell neuroendocrine carcinoma of the lung: a retrospective analy-
Downloaded from https://academic.oup.com/icvts/advance-article-abstract/doi/10.1093/icvts/ivz140/5513096 by guest on 23 July 2019
sis of 144 surgical cases. Lung Cancer 2006;53:111–5.
[24] Iyoda A, Hiroshima K, Moriya Y, Sekine Y, Shibuya K, Iizasa T et al.
Prognostic impact of large cell neuroendocrine histology in patients
with pathologic stage Ia pulmonary non-small cell carcinoma. J Thorac
Cardiovasc Surg 2006;132:312–5.
[25] Iyoda A, Jiang SX, Travis WD, Kurouzu N, Ogawa F, Amano H et al.
Clinicopathological features and the impact of the new TNM classifica-
tion of malignant tumors in patients with pulmonary large cell neuroen-
docrine carcinoma. Mol Clin Oncol 2013;1:437–43.
[26] Filosso PL, Rena O, Guerrera F, Moreno Casado P, Sagan D, Raveglia F
et al. Clinical management of atypical carcinoid and large-cell neuroen-
docrine carcinoma: a multicentre study on behalf of the European
Association of Thoracic Surgeons (ESTS) Neuroendocrine Tumours of
the Lung Working Group. Eur J Cardiothorac Surg 2015;48:55–64.
[27] Speicher PJ, Gu L, Gulack BC, Wang X, D’Amico TA, Hartwig MG et al.
Sublobar resection for clinical stage IA non-small-cell lung cancer in the
United States. Clin Lung Cancer 2016;17:47–55.
[28] Eichhorn F, Harms A, Warth A, Muley T, Winter H, Eichhorn ME. PD-L1
expression in large cell neuroendocrine carcinoma of the lung. Lung
Cancer 2018;118:76–82.