Db
SSa
b
a
ARRA
KVAPLPC
1
wfiswpimmpma
0h
Lung Cancer 82 (2013) 407– 412
Contents lists available at ScienceDirect
Lung Cancer
j ourna l ho me page: www.elsev ier .com/ locate / lungcan
ifferences in the prognostic implications of vascular invasionetween lung adenocarcinoma and squamous cell carcinoma
hingo Usuia,b, Yuko Minamia,∗, Toshihiro Shiozawaa, Shinji Iyamaa, Kaishi Satomia,hingo Sakashitaa, Yukio Satob, Masayuki Noguchia
Department of Pathology, Faculty of Medicine, University of Tsukuba, 1-1-1 Tennoudai, Tsukuba-shi, Ibaraki, 305-8575, JapanDepartment of Thoracic Surgery, Faculty of Medicine, University of Tsukuba, 1-1-1 Tennoudai, Tsukuba-shi, Ibaraki, 305-8575, Japan
r t i c l e i n f o
rticle history:eceived 1 July 2013eceived in revised form 19 August 2013ccepted 4 September 2013
eywords:ascular invasiondenocarcinomaeripheral-type squamous cell carcinomaarge-vessel invasionrognosisavity formation
a b s t r a c t
Objectives: Vascular invasion (VI) has been accepted as a universally important prognostic factor forpatients with lung carcinoma. However, the clinical significance of VI in each of the histological subtypeshas been unclear. The aim of the present study was to investigate differences in the clinicopathologicalimplications of VI between adenocarcinoma and squamous cell carcinoma.Method: A total of 336 patients were evaluated, of whom 81 were diagnosed as having peripheral-typesquamous cell carcinoma, and 255 as having adenocarcinoma.Result: Among the 336 patients, the five-year survival rates for those who were VI-positive and VI-negativewere 38.4% and 76.3%, respectively, the difference being significant (p < 0.0001). Multivariate analysisidentified VI as an independent prognostic factor (hazard ratio: 1.86). Although the difference in cancer-free survival between VI-positive and -negative patients was statistically significant for adenocarcinoma(p < 0.0001), it was not significant for squamous cell carcinoma (p = 0.086). For adenocarcinoma, thedifference between the survival curves for VI-positive and -negative patients was significant for the sub-types with a predominant lepidic (p < 0.0001), papillary (p = 0.0026), and acinar (p = 0.0060) component,
whereas that for the predominantly solid subtype was not significant (p = 0.58). Squamous cell carcino-mas were then divided into two groups on the basis of the diameter of vessels that had been invaded bythe cancer cells: large-vessel invasion (LVI; 1000 �m or more) and small-vessel invasion (SVI; less than1000 �m). Although there was no difference in the survival curves between the LVI and SVI groups, theLVI group showed a significantly higher incidence of cavity formation and distant metastasis.Conclusion: We conclude that VI is a useful prognostic indicator in lung carcinoma, although the clinicaletwe
implications of VI differ b. Introduction
Lung cancer is the leading cause of cancer-related death world-ide. Even for patients with pathological stage I lung cancer, theve-year survival rate is approximately 66–83%, which is far fromatisfactory [1]. Previously reported prognostic factors for patientsith lung cancer have included age, sex, performance status (PS),eripheral neutrophil and lymphocyte counts, tumor size, pleural
nvasion, lymphatic permeation, vascular invasion (VI), lymph nodeetastasis, and distant metastasis [2–6]. Among them, old age,
ale gender, poor PS, high neutrophil count, large tumor size, andresence of pleural invasion, lymphatic permeation, lymph nodeetastasis and distant metastasis are all indicative of an unfavor-
ble prognosis. VI in particular has been accepted as an important
∗ Corresponding author. Tel.: +81 29 853 3150; fax: +81 29 853 3150.E-mail address: [email protected] (Y. Minami).
169-5002/$ – see front matter © 2013 Elsevier Ireland Ltd. All rights reserved.ttp://dx.doi.org/10.1016/j.lungcan.2013.09.001
en adenocarcinoma and squamous cell carcinoma.© 2013 Elsevier Ireland Ltd. All rights reserved.
prognostic factor for patients with not only all stages [7–9] but alsoearly-stage [10–15] lung cancers.
Among the histological subtypes of lung cancer, a few reportshave indicated that VI is prognostically significant in adeno-carcinoma. For example, Bodendorf et al. have reported thatadenocarcinoma frequently shows distant metastasis rather thanlocal recurrence, and that VI is more significantly associatedwith metastasis than with local recurrence, thus being an indi-cator of poor prognosis [9]. Using a logistic regression model,Ruffini indicated that the differences between adenocarcinomawith and without VI, and between adenocarcinoma and non-adenocarcinoma, were significant [15]. These previous reportssuggested that VI is a unique prognostic indicator in individual
patients. On the other hand, the contribution of VI to the outcomeof squamous cell carcinoma has not been reported.In 2011, a new multidisciplinary classification of pulmonaryadenocarcinoma was proposed [16]. In this classification, invasiveadenocarcinoma was divided into five different subtypes, i.e. those
408 S. Usui et al. / Lung Cancer 82 (2013) 407– 412
Table 1Clinicopathological characteristics of all patients (n = 336).
All cases (n = 336) Ad (n = 255) Sq (n = 81)
Age 69.0 ± 9.9 67.6 ± 10.3 73.4 ± 7.0Sex (male/female) 225/111 153/102 72/9Tumor Size (Av. ± S.D.) 28.0 ± 14.1 26.0 ± 12.6 34.1 ± 16.6Primary site (left/right) 146/190 102/153 44/37Surgical procedure (par/seg/lob/pn) 30/30/274/2 23/23/207/2 7/7/67/0Pathological stage (IA/IB/IIA/IIB/IIIA/IIIB/IV) 131/86/32/30/47/2/8 106/60/25/16/39/1/8 25/26/7/14/8/1/0Pathological T (1a/1b/2a/2b/3/4) 87/55/132/14/46/2 75/40/98/9/31/2 12/15/34/5/15/0Pathological N (0/1/2/3) 254/37/44/1 189/29/37/0 65/8/7/1Presence of distant metastasis (%) 16 (4.8%) 10 (3.9%) 6 (7.4%)Pleural invasion (0/1/2/3) 211/75/24/26 164/57/20/14 47/18/4/12Presence of vascular invasion (%) 160 (47.6%) 115 (45.1%) 45 (55.6%)Presence of lymphatic permeation (%) 145 (43.2%) 114 (44.7%) 31 (38.3%)Recurrence of cancer (%) 99 (29.5%) 67 (26.3%) 30 (37.0%)Cause of death 82 (24.4%) 59 (23.1%) 23 (28.4%)Lung cancer 53 43 10
p d, ade
wscaflV
cIlm
2
2
pmUhtawiaemt
F(
Other 27
Unknown 2
ar, partial resection; seg, segmentectomy; lob, lobectomy; pn, pneumonectomy; A
ith a predominantly lepidic, acinar, papillary, micropapillary orolid component. These five subtypes differ not only histologi-ally, but also prognostically. Nakazato reported that tumors with
predominant lepidic, acinar or papillary component had a moreavorable prognosis than tumors with a predominant micropapil-ary or solid component [17]. However, the clinical implications ofI in these various histological subtypes have not been analyzed.
The aim of the present study was to confirm the clinical signifi-ance of VI in lung adenocarcinoma and squamous cell carcinoma.n addition, we examined the value of VI among the five histo-ogical subtypes of adenocarcinoma and the relationship between
acroscopic cavity formation and VI in squamous cell carcinoma.
. Materials and methods
.1. Case selection
Between January 2001 and December 2010, 420 consecutiveatients with a histological diagnosis of adenocarcinoma or squa-ous cell carcinoma underwent surgical treatment at Tsukubaniversity Hospital (Ibaraki, Japan). Eighty-four patients whoad central-type squamous cell carcinoma (n = 14), preoperativereatment (n = 18), adenocarcinoma in situ, minimally invasivedenocarcinoma and pure mucinous adenocarcinoma (n = 52)ere excluded, leaving a final total of 336 patients for inclusion
n this study. In addition, as cases of adenosquamous carcinomand sarcomatoid carcinoma were very limited, these were alsoxcluded. From among these patients, we obtained 81 speci-ens of peripheral-type squamous cell carcinoma arising beyond
he subsegmental bronchus, and 255 specimens of invasive
ig. 1. Pathologic findings of vascular invasion (VI) in (a) adenocarcinoma and (b) squamoc) large-vessel invasion (LVI) by squamous cell carcinoma (EVG, ×2.5). Each black bar is
16 110 2
nocarcinoma; Sq, squamous cell carcinoma.
adenocarcinoma; the patients’ clinical characteristics wereobtained from the hospital’s database. All of the patients gaveinformed consent for study of their surgical materials beforecollection. The study was approved by the Regional Committeefor Medical Research Ethics (University of Tsukuba). The patients’clinicopathological profiles are summarized in Table 1.
Preoperative assessment of the patients included chest andabdominal computed tomography (CT), bone scintigraphy, brainmagnetic resonance imaging (MRI), and peripheral venous bloodsampling for evaluation of tumor markers. CEA (ng/ml), SCC (ng/ml)and proGRP (pg/ml) were measured by ELISA. NSE (ng/ml) wasmeasured by RIA, and CYFRA (ng/ml) and SLX (ng/ml) were mea-sured by IRMA. All markers were evaluated within a month beforesurgery. Patients in whom distant metastasis was detected withinone month after surgery were recognized as having distant metas-tasis and excluded from survival analysis.
2.2. Histological examination
The surgically resected specimens were fixed routinely with 15%formalin, and cut serially into slices 5–7 mm thick. We preparedparaffin blocks of the largest cut surface of the tumor and exam-ined all of the 4-�m-thick sections obtained. Finally, the sectionswere stained with hematoxylin–eosin, and judged for VI and lym-phatic permeation after elastica van Gieson staining (Fig. 1) and
D2-40 immunostaining, respectively. The adenocarcinomas wereclassified in accordance with the IASLC/ATS/ERS international mul-tidisciplinary classification of lung adenocarcinoma (2011) [16].Invasive adenocarcinoma was divided into five subtypes, i.e. thosewith a predominantly lepidic, acinar, papillary, micropapillary orus cell carcinoma of the lung as revealed by elastica van Gieson staining (×10), and500 �m.
S. Usui et al. / Lung Cancer 82 (2013) 407– 412 409
Table 2Univariate analysis between VI negative and VI positive in all 336 cases.
VI negative (n = 176) VI positive (n = 160) p
Age (Av. ± S.D.) 69.0 ± 9.7 69.7 ± 9.9 0.18Tumor size (Av. ± S.D.) 24.9 ± 12.6 31.4 ± 14.9 0.0001*
Pathological stage (IA/IB/IIA/IIB/IIIA/IIIB/IV) 99/39/8/12/14/2/2 31/47/24/18/33/0/6 0.0001*
Pathological T (1a/1b/2a/2b/3/4) 64/39/48/5/18/1 22/16/84/9/28/1 0.0001*
Pathological N (0/1/2/3) 152/9/14/1 102/28/30/0 0.0001*
Presence of distant metastasis (%) 4 (2.3%) 12 (7.5%) 0.022*
Pleural invasion (0/1/2/3) 142/19/9/6 69/56/15/20 0.0001*
Presence of lymphatic permeation (%) 42 (23.9%) 103 (64.4%) 0.0001*
Predominant of invasive adenocarcinoma (lepi/aci/sol/pap/micropap) 89/16/14/20/1 25/27/38/24/1 0.0001*
CEA (Av. ± S.D.) (ng/ml) 6.3 ± 17.2 18.0 ± 88.0 0.0001*
SLX (Av. ± S.D.) (ng/ml) 24.5 ± 20.1 26.2 ± 14.5 0.062SCC (Av. ± S.D.) (ng/ml) 1.2 ± 1.0 2.7 ± 11.2 0.0001*
CYFRA (Av. ± S.D.) (ng/ml) 1.7 ± 1.9 3.3 ± 10.2 0.001*
NSE (Av. ± S.D.) (ng/ml) 8.7 ± 3.7 8.9 ± 5.6 0.922
V
snmmOd(
2
wwKsso(l
Fcesc
ProGRP (Av. ± S.D.) (ng/ml)
I, vascular invasion.* Indicates significant differences.
olid component. Among the 255 adenocarcinomas, 144 were diag-osed as predominantly lepidic, 52 as acinar, 44 as papillary, 2 asicropapillary and 43 as solid. Histological classification of squa-ous cell carcinoma was carried out according to the World Healthrganization (WHO) classification (2004) [19]. Tumor staging wasone in accordance with the Staging Manual for Thoracic Oncology2009) [18].
.3. Statistical analysis
Statistical analysis was performed using the SPSS 11.5J soft-are package (SPSS Inc., Chicago, IL, USA). Cancer-free survivalas assessed by disease-free survival (DFS) analysis using theaplan–Meier method, and the significance of differences between
urvival curves was assessed using the log-rank test. Cancer-freeurvival was determined from the date of surgery to the datef recurrence or last follow-up, excluding other causes of deathdisease-free survival). To examine associations with clinicopatho-ogical factors, the chi-squared test (or Fisher’s exact test whereig. 2. Disease-free survival curves for patients with and without vascular invasion. (aarcinoma (81 cases). The significance of differences between survival curves was assesseach of the four specific adenocarcinoma subtypes. Differences between the survival curubtype, (e) the acinar-predominant subtype, (f) the papillary-predominant subtype, and
urves was assessed by the log-rank test.
3.3 ± 4.8 26.2 ± 17.1 0.25
appropriate) was used for categorical variables, and the t-test forcontinuous variables. To assess the independent predictive valueof different variables for survival, multivariate analysis was per-formed using the logistic regression analysis model. The result wasconsidered to be significant when the p value was less than 0.05.
3. Results
3.1. Prognostic value of vascular invasion in the cases overall
The associations between VI and various clinicopathologicalprognostic factors are shown in Table 2. VI was significantly associ-ated with tumor size, pathological T, N factor and stage, distantmetastasis, pleural invasion, lymphatic permeation, and several
tumor markers (CEA, SCC and CYFRA). The five-year cancer-free sur-vival rates for VI-positive and -negative cases were 38.4% and 76.3%,respectively (p < 0.0001, Fig. 2a). In addition, all cases positive forlymphatic permeation showed a significantly unfavorable outcome(p < 0.0001), as did cases of invasive adenocarcinoma (p < 0.0001)) Total cases (336 cases), (b) adenocarcinoma (255 cases), and (c) squamous celld by the log-rank test. Disease-free survival curves were obtained for patients withves for VI-positive and -negative cases are shown for (d) the lepidic-predominant(g) the solid-predominant subtype. The significance of differences between survival
410 S. Usui et al. / Lung Cancer 82 (2013) 407– 412
Table 3Multivariate prognostic analysis for cancer free survival.
Variable Hazard ratio (95% CI) p value
Histological subtype (sq vs ad) 0.65 (0.34–1.24) 0.19Tumor size (30 mm< vs 30 mm≥) 1.02 (0.60–1.76) 0.93Pathological T (1 vs 2≥) 1.24 (0.88–1.76) 0.23Pathological N (negative vs positive) 1.89* (1.08–3.30) 0.026*
Pathological stage (I vs II≥) 1.09 (0.71–1.68) 0.69Vascular invasion (negative vs positive) 1.86* (1.08–3.22) 0.026*
Lymphatic permeation (negative vs positive) 1.42 (0.80–2.54) 0.23Pleural invasion (0 vs 1≥) 1.00 (0.76–1.32) 0.99CEA (9 ng/ml< vs 9 ng/ml≥) 1.21 (0.69–2.13) 0.50SCC (1.7 ng/ml< vs 1.7 ng/ml≥) 1.18 (0.65–2.16) 0.50CYFRA (3.0 ng/ml< vs 3.0 ng/ml≥) 0.97 (0.50–1.90) 0.93Solid predominant invasive adenocarcinoma or not 2.32* (1.22–4.41) 0.010*
Sq, squamous cell carcinoma; Ad, adenocarcinoma; CI, confidence interval.* Indicates significant differences.
Table 4Multivariate prognostic analysis of invasive adenocarcinoma.
Variable Hazard ratio (95% CI) p value
Pathological T (1 vs 2≥) 1.14 (0.79–1.14) 0.48Pathological N (negative vs positive) 1.12 (0.62–1.12) 0.71Pathological stage (I vs II≥) 1.53 (0.98–1.53) 0.062solid predominant invasive adenocarcinoma or not 2.13* (1.12–4.06) 0.021*
Vascular invasion (negative vs positive) 2.40* (1.17–4.93) 0.017*
Lymphatic permeation(negative vs positive) 1.21 (0.58–2.54) 0.61Pleural invasion (0 vs 1≥) 1.00 (0.72–1.40) 0.99CEA (9 ng/ml< vs 9 ng/ml≥) 1.28 (0.66–2.51) 0.47SCC (1.7 ng/ml< vs 1.7 ng/ml≥) 1.67 (0.79–3.55) 0.18
C
(ttl(Nirspsstv7nfi4a
i0
Slas
3v
s
significant, whereas that for the solid-predominant subtype wasnot significant (p = 0.58, Fig. 2g). Multivariate prognostic analysisof invasive adenocarcinoma also showed VI to be the factor thatwas best able to separate patients with favorable and unfavorableprognoses. VI and solid-predominant invasive adenocarcinoma
Table 5Univariate analysis between squamous cell carcinoma cases with SVI and those withLVI (n = 45).
SVI group(n = 24)
LVI group(n = 21)
p
Age (Av. ± S.D.) 74.0 ± 8.0 75.0 ± 6.0 0.53Tumor Size (Av. ± S.D.) 34.3 ± 16.0 41.7 ± 20.0 0.24Pathological stage (IA/IB/IIA/IIB/IIIA) 5/9/5/4/1 4/8/1/6/2 0.50Pathological T (1a/1b/2a/2b/3) 2/4/12/1/5 3/1/10/2/5 0.68Pathological N (0/1/2) 19/4/1 16/4/1 0.97Presence of distant metastasis (%) 0% 4 (19.0%) 0.025*
CYFRA (3.0 ng/ml< vs 3.0 ng/ml≥) 1.28
I, confidence interval.* Indicates significant differences.
Table 2, Supplementary Fig. 1). The prognostic factors subjectedo multivariate analysis in the present study were histological sub-ype, tumor size, pathological T and N factor, pathological stage, VI,ymphatic permeation, pleural invasion and tumor marker levelsCEA 9.0 ng/ml, CYFRA 3.0 ng/ml, SCC 1.7 ng/ml). VI, pathological
factor and solid predominant invasive adenocarcinoma weredentified as independent prognostic factors (Table 3). The hazardatios for patients who were VI-positive, N-positive and with theolid subtype were 1.86, 1.89 and 2.32 times higher than those foratients who were VI-negative, N-negative and with a non-solidubtype, respectively. We also compared the cancer-free five-yearurvival rates of VI-positive and VI-negative patients with each ofhe histological subtypes. For adenocarcinoma, the five-year sur-ival rates for VI-positive and -negative patients were 33.4% and7.1%, respectively, and the difference between them was sig-ificant (p < 0.0001) (Fig. 2b). For squamous cell carcinoma, theve-year survival rates for VI-positive and -negative patients were4.0% and 60.5%, respectively. VI-positive cases showed an unfavor-ble prognosis, but not to a significant degree (p = 0.086) (Fig. 2c).
Supplementary data associated with this article can be found,n the online version, at http://dx.doi.org/10.1016/j.lungcan.2013.9.001.
upplementary Fig. I. Disease-free survival curves for patients with and withoutymphatic permeation. (a) Total cases (336 cases), (b) adenocarcinoma (255 cases),nd (c) squamous cell carcinoma (81 cases). The significance of differences betweenurvival curves was assessed by the log-rank test.
.2. Subtype survival analysis according to the presence ofascular invasion after resection of invasive adenocarcinoma
The Kaplan–Meier survival curves of patients with inva-ive adenocarcinoma subtypes based on the IASLC/ATS/ERS
(0.55–3.02) 0.57
multidisciplinary classification were analyzed. In this study, weexcluded micropapillary-predominant adenocarcinomas, sincetheir number was very limited (2 cases). The five-year survival ratesfor patients with the predominantly lepidic, acinar, papillary andsolid subtypes were 83.3%, 27.2%, 33.4% and 55.9%, respectively.The five-year survival rates for VI-positive and -negative patientswere 46.7% and 94.9% for the lepidic subtype, 10.0% and 46.7% forthe acinar subtype, 16.0% and 70.2% for the papillary subtype, and56.5% and 56.6% for the solid subtype, respectively. The differencebetween the survival curves for VI-positive and -negative patientswas most significant for the predominantly lepidic subtype(p < 0.0001, Fig. 2d) and those for the acinar (p = 0.0061, Fig. 2e)and papillary (p = 0.0026, Fig. 2f)-predominant subtypes were also
Pleural invasion (0/1/2/3) 11/8/1/4 8/6/2/5 0.80Presence of lymphatic permeation (%) 11 (45.8%) 9 (42.9%) 0.84Presence of cavity formation (%) 4 (16.7%) 14 (66.7%) 0.001*
SVI, small-vessel invasion; LVI, large-vessel invasion.* Indicates significant differences.
ancer 82 (2013) 407– 412 411
whsw
3p
(ppssios(goMdscffh
4
c
Table 6Univariate analysis between cases with cavity formation and without cavity forma-tion (n = 65).
Cavity formation p
Negative (n = 33) Positive (n = 32)
Tumor size (Av.±S.D.) 27.0 ± 13.4 38.4 ± 17.4 0.004*
Pathological T (1a/1b/2a/2b/3) 8/9/10/0/6 4/3/18/3/4 0.04*
Presence of small VI 15 (45.5%) 18 (56.3%) 0.38*
Ffi
S. Usui et al. / Lung C
ere identified as independent prognostic factors (Table 4). Theazard ratios for patients who were VI-positive and had the solidubtype were 2.40 and 2.13 times higher than those for patientsho were VI-negative and had a non-solid subtype, respectively.
.3. Clinicopathological significance of vascular invasion ineripheral squamous cell carcinoma
Among the peripheral squamous cell carcinomas, tumor sizep = 0.04), pathological N factor (p = 0.02), stage (p = 0.048), andleural invasion (p = 0.015) were significantly correlated with VI-ositivity or -negativity, but none of the other factors showed aignificant association. We then focused on the size of the ves-els that had been invaded by the tumor cells. Cases with vascularnvasion were divided into two groups: large-vessel invasion (LVI;uter short-axis diameter 1000 �m or more) and small-vessel inva-ion (SVI; less than 1000 �m). There was no significant differencep = 0.97) in five-year cancer-free survival between the LVI and SVIroups, but the LVI group showed a significantly higher incidencef distant metastasis and macroscopic cavity formation (Table 5).acroscopic cavity formation is defined as cavity formation evi-
ent upon macroscopic examination of the cut surface of thepecimen. The association between cavity formation and variouslinicopathological prognostic factors is shown in Table 6. Cavityormation was correlated with LVI, tumor size and pathological Tactor. Fig. 3 shows the CT findings, macroscopic cut surface, andistology of a representative case showing LVI.
. Discussion
The clinical implications of VI in adenocarcinoma and squamousell carcinoma were compared using various clinicopathological
ig. 3. A representative case of squamous cell carcinoma showing large-vessel invasion (ndings in the cavity wall (H and E, ×4), and (d) a large-vessel invasion (≥1000 �m in dia
Presence of Large VI 6 (18.2%) 14 (43.8%) 0.026
VI, vascular invasion.* Indicates significant differences.
factors, including disease-free survival. We were able to demon-strate that VI-positive adenocarcinoma showed a significantlyworse outcome than VI-negative adenocarcinoma, whereas no suchdifference was evident for squamous cell carcinoma.
As VI is the initial step of distant metastasis, it would naturally beexpected to be associated with an unfavorite prognosis. However,the molecular mechanism of VI is still unclear. Recently, fibroblastgrowth factor receptor 1 (FGFR1), �-catenin and vascular endothe-lial growth factor (VEGF) have been reported to be associated withVI [20–22]. In order to prevent VI, it is very important to clarify themolecular correlations between vessels and cytokines produced bytumors.
In this study, we counted both arterial and venous invasionaccording to the Staging Manual for Thoracic Oncology (2009) [18]and evaluated both forms of vascular invasion. Hemodynamically,
however, venous invasion would be expected to have a greater pro-gnostic impact than arterial invasion. In the present study, VI wasshown to be a strong prognostic factor without any distinction fromarterial invasion. If venous invasion and arterial invasion could beLVI). (a) Chest computed tomography, (b) cut surface of the tumor, (c) microscopicmeter) found in this tumor (EVG, ×2.5). Each black bar is 1000 �m.
4 ancer
sm
cel[bfbstrpigdVeocttitp
iWvsiccatccIbim
caappcsp
C
[
[
[
[
[
[
[
[
[
[
[
[
[
12 S. Usui et al. / Lung C
eparated easily from each other, the results of this study would beore significant and specific.The histology of invasive adenocarcinoma is heterogeneous and
an be classified into various subtypes. In the present study, wexamined the clinical significance of VI among each of these histo-ogical subtypes according to the IASLC/ATS/ERS new classification16]. Interestingly, a significant difference in cancer-free survivaletween VI-positive and -negative patients was demonstratedor lepidic, acinar and papillary-predominant adenocarcinomas,ut not for solid-predominant adenocarcinoma. In patients witholid-predominant adenocarcinoma, factors other than VI may con-ribute significantly to outcome. Interestingly, Nakazato [23] haseported nuclear grade to be a significant prognostic factor inatients with solid-predominant adenocarcinoma. Nuclear atypia
s a morphological characteristic biologically indicative of variousenetic abnormalities. Atypia associated with cell proliferation orrug resistance may be a more powerful prognostic indicator thanI in solid-predominant adenocarcinoma. As the number of casesxamined was very limited, we were unable to assess the outcomef micropapillary-predominant adenocarcinomas, and thus thelinical implications of VI in these cancers remained unclear. Likehe solid type, this subtype of adenocarcinoma has been reportedo have a high grade, and its unique morphologic characteristics frequent lymphatic permeation. It may be of further interesto compare the clinical significance of VI with that of lymphaticermeation in micropapillary-predominant adenocarcinoma.
In squamous cell carcinoma, there was no significant differencen cancer-free survival between VI-positive and -negative cases.
e then focused on the relationships between the diameter ofessels invaded by tumor cells and clinicopathological findings,ince invasion of large vessels has been detected more frequentlyn squamous cell carcinoma than in adenocarcinoma. Interestingly,ases with LVI showed distant metastasis more frequently thanases without LVI. Macchiarini reported that VI could be used asn indicator of occult metastasis, and that LVI can be an impor-ant indicator of distant metastasis in patients with squamous cellarcinoma [24]. Similarly to distant metastasis, LVI was signifi-antly related to cavity formation in the present study (Table 5).n general, CT is useful for detecting intratumoral cavity formationefore surgery. Our present study revealed that intratumoral cav-
ty formation detected by CT was an indicator of potential distantetastasis.We conclude that VI is a useful prognostic indicator in lung
arcinoma, although the clinical implications of VI differ betweendenocarcinoma and squamous cell carcinoma. Among invasivedenocarcinomas, it was interesting that VI was an independentrognostic factor in tumors with a predominantly lepidic, acinar, orapillary component, but not in those with a predominantly solidomponent. Among squamous cell carcinomas, those with LVI wereignificantly associated with macroscopic cavity formation, and the
resence of distant metastasis.onflict of interest statement
None of the authors have any conflicts of interest to declare.
[
[
82 (2013) 407– 412
References
[1] Asamura H, Goya T, Koshiishi Y. A Japanese lung registry study: prognosis of13,010 resected lung cancers. J Thorac Oncol 2008;3:46–52.
[2] Suzuki K, Nagai K, Yoshida J. Conventional clinicopathologic prognostic factorsin surgically resected non-small cell lung carcinoma. A comparison of pro-gnostic factors for each pathologic TNM stage based on multivariate analyses.Cancer 1999;86:1976–84.
[3] Mountain CF. Lung cancer staging classification. Clin Chest Med 1993;14:43–53.[4] Naruke T, Goya T, Tsuchiya R. Prognosis and survival in resected lung carci-
noma based on the new international staging system. J Thorac Cardiovasc Surg1988;96:440–7.
[5] Paesmans M, Sculier JP, Libert P. Prognostic factors for survival in advanced non-small-cell lung cancer: univariate and multivariate analyses including recursivepartitioning and amalgamation algorithms in 1,052 patients. The EuropeanLung Cancer Working Party. J Clin Oncol 1995;13:1221–30.
[6] Kobayashi N, Usui S, Kikuchi S. Preoperative lymphocyte count is an indepen-dent prognostic factor in node-negative non-small cell lung cancer. Lung Cancer2012;75:223–7.
[7] Kessler R, Gasser B, Massard G. Blood vessel invasion is a major progno-stic factor in resected non-small cell lung cancer. Ann Thorac Surg 1996;62:1489–93.
[8] Bréchot JM, Chevret S, Charpentier MC. Blood vessel and lymphatic vessel inva-sion in resected non-small cell lung carcinoma. Correlation with TNM stage anddisease-free and overall survival. Cancer 1996;78:2111–8.
[9] Bodendorf MO, Haas V, Laberke HG. Prognostic value and therapeutic conse-quences of vascular invasion in non-small cell lung carcinoma. Lung Cancer2009;64:71–8.
10] Mineo TC, Ambrogi V, Baldi A. Prognostic impact of VEGF, CD31, CD34, andCD105 expression and tumour vessel invasion after radical surgery for IB–IIAnon-small cell lung cancer. J Clin Pathol 2004;57:591–7.
11] Khan OA, Fitzgerald JJ, Field ML. Histological determinants of survival in com-pletely resected T1-2N1M0 non-small cell cancer of the lung. Ann Thorac Surg2004;77:1173–8.
12] Gabor S, Renner H, Popper H. Invasion of blood vessels as significant progno-stic factor in radically resected T1-3N0M0 non-small cell lung cancer. Eur JCardiothorac Surg 2004;25:439–42.
13] Tsuchiya T, Akamine S, Muraoka M. Stage IA non-small cell lung cancer: vesselinvasion is a poor prognostic factor and a new target of adjuvant chemotherapy.Lung Cancer 2007;56:341–8.
14] Shoji F, Haro A, Yoshida T. Prognostic significance of intratumoral blood vesselinvasion in pathologic stage IA non-small cell lung cancer. Ann Thorac Surg2010;89:864–9.
15] Ruffini E, Asioli S, Filosso PL. Significance of the presence of microscopic vas-cular invasion after complete resection of stage I–II pT1-T2N0 non-small celllung cancer and its relation with T-size categories: did the 2009 7th Edition ofthe TNM Staging System miss something. J Thorac Oncol 2011;6:319–26.
16] Travis WD, Brambilla E, Noguchi M. International Association for the Study ofLung Cancer/American Thoracic Society/European Respiratory Society Interna-tional Multidisciplinary Classification of Lung Adenocarcinoma. J Thorac Oncol2011;6:244–85.
17] Nakazato Y, Maeshima AM, Ishikawa Y. Interobserver agreement in the nucleargrading of primary pulmonary adenocarcinoma. J Thorac Oncol 2013;8:736–43.
18] Goldstraw P. Staging manual in thoracic oncology. An International Associationfor the Study of Lung Cancer Publication. Florida: Editorial Press; 2009.
19] Travis WD, Bramblilla E. Pathology and genetics of tumours of the lung, pleura,thymus and heart. Lyon: IARC Press; 2004.
20] Tran TN, Selinger CI, Kohonen-Corish MR. Fibroblast growth factor receptor 1(FGFR1) copy number is an independent prognostic factor in non-small celllung cancer. Lung Cancer 2013;81:462–7.
21] Kase S, Sugio K, Yamazaki K. Expression of E-cadherin and beta-catenin inhuman non-small cell lung cancer and the clinical significance. Clin CancerRes 2000;12:4789–96.
22] Han H, Silverman JF, Santucci TS. Vascular endothelial growth factor expressionin stage I non-small cell lung cancer correlates with neoangiogenesis and a poorprognosis. Ann Surg Oncol 2001;8:72–9.
23] Nakazato Y, Minami Y, Kobayashi H. Nuclear grading of primary pulmonaryadenocarcinomas: correlation between nuclear size and prognosis. Cancer2010;116:2011–9.
24] Macchiarini P, Fontanini G, Dulmet E. Angiogenesis: an indicator of metasta-sis in non-small cell lung cancer invading the thoracic inlet. Ann Thorac Surg1994;57:1534–9.
Recommended
