[
Original Research Lung Cancer
]
Prognostic Significance of Visceral Pleural Involvement in Early-Stage Lung Cancer Sameer Lakha, BA; Jorge E. Gomez, MD; Raja M. Flores, MD, FCCP; and Juan P. Wisnivesky, MD, DrPH
Visceral pleural invasion (VPI) may impact non-small cell lung cancer (NSCLC) survival. However, previous studies are mixed as to whether VPI is an independent prognostic factor in early-stage cancers and whether its effect is size dependent. In the current American Joint Committee on Cancer (AJCC) staging system, VPI leads to upstaging of cancers , 3 cm but not of those 3 to 7 cm in size. BACKGROUND:
Using the Surveillance, Epidemiology, and End Results (SEER) registry, we identified 16,315 patients with stage I-II NSCLC treated with lobectomy. We used the Kaplan-Meier method and Cox regression to assess the association of VPI with lung cancer-specific (primary outcome) and overall survival. Based on these results, we created a revised VPI staging classification.
METHODS:
Overall, 3,389 patients (21%) had VPI. Kaplan-Meier analysis stratified by tumor size showed worse cancer-specific survival in patients with VPI (P , .0001). VPI was independently associated with decreased lung cancer-specific survival (hazard ratio, 1.38; 95% CI, 1.29-1.47) after controlling for tumor size and other confounders; this effect was not size dependent. In our revised classification, tumors , 7 cm with VPI were upstaged to the next T category.
RESULTS:
VPI is a prevalent finding associated with worse prognosis in early-stage lung cancer, even among patients with tumors . 3 cm, a factor not captured in the current staging system. Patients with VPI may be considered candidates for more aggressive treatment.
CONCLUSIONS:
CHEST 2014; 146(6):1619-1626
Manuscript received January 23, 2014; revision accepted June 4, 2014; originally published Online First July 17, 2014. ABBREVIATIONS: AJCC 5 American Joint Committee on Cancer; CSE 5 Collaborative Staging Extension; HR 5 hazard ratio; LN 5 lymph node; NSCLC 5 non-small cell lung cancer; SEER 5 Surveillance, Epidemiology, and End Results; VPI 5 visceral pleural involvement AFFILIATIONS: From the Division of General Internal Medicine (Mr Lakha and Dr Wisnivesky), Tisch Cancer Institute (Dr Gomez), Department of Thoracic Surgery (Dr Flores), and Division of Pulmonary, Critical Care, and Sleep Medicine (Dr Wisnivesky), Icahn School of Medicine at Mount Sinai, New York, NY.
journal.publications.chestnet.org
Part of this article was presented in abstract form at the American Thoracic Society International Conference, May 19, 2014, San Diego, CA. FUNDING/SUPPORT: The authors report to CHEST that no funding or other support was received for this study. CORRESPONDENCE TO: Sameer Lakha, BA, Department of Medicine, Icahn School of Medicine at Mount Sinai, One Gustave L. Levy Pl, Box 1087, New York, NY 10029; e-mail: [email protected] © 2014 AMERICAN COLLEGE OF CHEST PHYSICIANS. Reproduction of this article is prohibited without written permission from the American College of Chest Physicians. See online for more details. DOI: 10.1378/chest.14-0204
1619
Accurate staging of lung cancer is critical for optimizing treatment decisions, communicating prognosis, and determining eligibility for clinical trials. Tumor spread to the visceral pleura has been recognized as a poor prognostic factor in non-small cell lung cancer (NSCLC).1-4 Consequently, visceral pleural involvement (VPI) has been incorporated into the American Joint Committee on Cancer (AJCC) TNM staging system alongside tumor size.5 According to these criteria, tumors , 3 cm that would otherwise be classified T1a (, 2 cm) or T1b (2-3 cm) disease based on size alone are considered to be equivalent to more advanced tumors (T2a; 3-5 cm) if they extend into the visceral pleura. Conversely, for larger tumors (ie, . 3 cm), VPI does not result in upstaging.5 Several studies have evaluated whether VPI is an independent prognostic factor in patients with early-stage
Materials and Methods We used data from the latest release of the Surveillance, Epidemiology, and End Results (SEER) registry.21 SEER is a network of cancer registries that covers approximately 28% of the population of the United States and collects data on all new cases of cancer in its coverage regions.22 From SEER, we identified all incident cases of histologically confirmed NSCLC diagnosed between 2004 and 2010. The study included patients for whom NSCLC was their first and only primary malignancy, who had tumors , 7 cm, without lymph node (LN) involvement (N0) or distant metastasis (M0), and who underwent lobectomy. We excluded patients diagnosed on autopsy or death certificate data, as well as cases with missing tumor size and stage information. We further excluded patients with tumors , 3 cm but classified as having T2 disease due to hilar atelectasis or obstructive pneumonitis, patients with T3 or T4 disease regardless of tumor size, and patients treated with preoperative radiotherapy (due to concerns about possible understaging). SEER reports data on sociodemographic characteristics, including age, sex, race/ethnicity, and marital status. Cancer information includes year of diagnosis, tumor location (upper lobe, middle lobe, lower lobe, or central), and histology (adenocarcinoma, large-cell carcinoma, squamous cell carcinoma, or other). The registry includes detailed pathologic staging data collected within 4 months of diagnosis or within completion of the first course of treatment.23,24 Data are available regarding local tumor extension and size, LN involvement, and presence or absence of systemic metastasis. TNM data are also provided according to the AJCC staging system. Based on this classification, cases were grouped according to tumor size into the following categories: , 2 cm, 2 to 3 cm, 3 to 5 cm, and 5 to 7 cm. The presence of VPI was ascertained using Collaborative Staging Extension (CSE) codes (410, 420, 430, and 450). Of note, VPI cases prior to 2010 were classified using the CSE code 450, which includes both VPI and extension to the pulmonary ligament (ie, the fold of parietal pleura that forms and extends inferiorly from the hilum). In 2010, SEER introduced a distinct CSE code (440) for pulmonary ligament involvement. Of the 25,891 NSCLC cases diagnosed in 2010, 2,074 (8.01%) were coded as having VPI (CSE codes 410, 420, 430) compared with only 9 (0.03%) identified as having pulmonary ligament involvement (CSE code 440). Moreover, all nine of these cases were excluded based on other cohort selection criteria. Therefore, we inferred that the number of pulmonary ligament cases among patients with CSE code 450 prior
1620 Original Research
NSCLC, with varying results.2,6-20 Single-center and multicenter series, as well as analyses of patients enrolled in a clinical trial, found that VPI confers worse prognosis among resected tumors , 7 cm.2,9,12,15,17,18,20 However, the prognostic significance of VPI was not confirmed in adjusted analysis in many of these studies. In contrast, several other studies showed either no impact of VPI on prognosis, or a differential effect based on the size of the tumor.6,7,11,12,16,19 Thus, the independent prognostic significance of VPI, as well as its proper role in the staging of these cancers, remains unclear. In this population-based study, we used data from a large cohort of patients with surgically resected stage I and II NSCLC to better characterize the prognostic impact of VPI in tumors , 7 cm and to refine the staging of these cancers.
to 2010 included in the study was negligible, suggesting minimal misclassification. SEER reports data on surgical treatment and radiation therapy use during the first course of cancer-directed treatment. Patients undergoing lobectomy were identified using SEER surgical codes 30 through 48. Data are also provided regarding the number of LNs sampled during surgery. Radiotherapy and radiation-surgery sequence codes were used to ascertain whether patients underwent postoperative external beam radiotherapy. SEER does not report information about chemotherapy use, due to concerns about misclassification. The primary study outcome was lung cancer-specific mortality, as the study goal was to assess the impact of VPI on lung cancer prognosis; secondary analyses used overall survival. Survival time was calculated as the number of months from cancer diagnosis until death or last follow-up (December 31, 2010) for censored observations. Cause of death was coded by SEER using information extracted from death certificate data. Patients who died of non-lung cancer causes were classified as censored at the date of death in analyses of lung cancer-specific survival. Statistical Analysis Differences in the baseline characteristics of patients with and without VPI were compared using t tests or x2 test, as appropriate. The unadjusted association of VPI with survival was assessed using Kaplan-Meier analysis stratifying the sample by tumor size categories (, 2 cm, 2-3 cm, 3-5 cm, and 5-7 cm); tumors . 7 cm or of T3 status were included for comparison. Adjusted analysis was conducted using a Cox proportional hazards model to assess the association between VPI and survival after controlling for age, sex, race/ethnicity, marital status, histology, tumor size, number of LNs examined, and use of adjuvant radiotherapy. The analysis was repeated including interactions between tumor size and VPI to assess if the effect of pleural involvement varied across different tumor size categories. To combine groups with similar prognoses, we fitted a Cox model including one dichotomous indicator for each tumor size category and VPI status (ie, , 2 cm without VPI, , 2 cm with VPI, etc). Based on the findings of this analysis, we grouped cancers with overlapping hazard ratios (HRs), thereby developing a revised prognostic system. Results were deemed significant at a two-sided P value , .05. All statistical analysis was conducted using SAS software, version 9.2 (SAS Institute Inc). The study was determined to be exempt research by the Institutional Review Board of the Icahn School of Medicine at Mount Sinai.
[
146#6 CHEST DECEMBER 2014
]
Results The cohort selection process is shown in Figure 1. Baseline demographic and clinical characteristics of patients with and without VPI are provided in Table 1. Relative to patients without VPI, patients with VPI were older (P 5 .0002), less likely to be white (P 5 .03), had larger tumors (P , .0001), were more likely to present with adenocarcinoma (P , .0001), had more LNs evaluated at surgery (P 5 .01), and were more likely to have undergone adjuvant radiotherapy (P , .0001). There were no significant differences in the distribution of sex (P 5 .24) or marital status (P 5 .85). Lung cancer-specific (P , .0001) and overall (P , .0001) survival were significantly different across tumor size and VPI strata (Fig 2). Five-year survival rates according to tumor size and VPI status are shown in Table 2. Overall, VPI was associated with worse survival for all tumor size groups, with prognosis of patients with VPI being similar to that of patients with the adjacent larger tumor size category. Cox analysis demonstrated that VPI confers a significant lung cancer-specific survival disadvantage
Figure 1 – Cohort selection criteria. The process by which we identified our study cohort is shown. NSCLC 5 non-small cell lung cancer; RT 5 radiotherapy; SEER 5 Surveillance, Epidemiology, and End Results.
journal.publications.chestnet.org
(HR, 1.38; 95% CI, 1.29-1.47), even after controlling for size and other potential confounders. Interactions between tumor size and VPI were not statistically significant (P value range 5 .4-.6), suggesting an equivalent effect of VPI in all tumor size categories. Similar results were observed in analyses with overall survival as the end point. HRs for individual size-VPI strata are shown in Table 3. As with the 5-year survival rates in Table 2, the HRs and their CIs cluster into groups that do not track the current AJCC stage categories. Based on the groups that emerged from our results, we developed a revised classification in which tumors with VPI were upstaged to the next T-status category, resulting in the following groups: , 2 cm without VPI; , 2 cm with VPI and 2 to 3 cm without VPI; 2 to 3 cm with VPI and 3 to 5 cm without VPI; 3 to 5 cm with VPI and 5 to 7 cm without VPI; 5 to 7 cm with VPI and . 7 cm and T3 tumors (Table 4).
Discussion The prognostic impact of VPI in early-stage NSCLC has been unclear. This study confirms a size-independent, negative prognostic effect of VPI in a large populationbased sample. Moreover, a revised tumor classification incorporating this finding had equivalent discrimination and may help to better select VPI patients with poorer prognosis. These results have implications for the treatment of patients with VPI, a frequent finding affecting 25% of patients with early-stage NSCLC. Studies have shown an association between VPI and survival. Koike et al8 found an independent, negative prognostic effect of VPI on survival in a single-center series. In larger registry-based studies, Yoshida et al15 and Kawase et al18 found worse prognosis among patients with VPI and suggested revised staging classifications similar to ours; however, these studies did not adjust for confounders to assess the independence and size-dependence of this association. Other studies show limited or no association between VPI and prognosis. For instance, using a single-center study of adenocarcinoma cases treated with lobectomy or limited resection, Nitadori et al19 found an impact on survival among patients with tumors 2 to 3 cm, but not those , 2 cm. Similarly, David et al16 found no association between VPI and survival in tumors , 5 cm. Yanagawa et al20 found no independent association between VPI and survival. In contrast, we found a size-independent, negative association in all tumors , 7 cm on unadjusted and adjusted analyses. These differences may be due to 1621
TABLE 1
] Baseline Characteristics of Study Subjects According to VPI
Characteristic
No VPI (n 5 12,926)
VPI (n 5 3,389)
Age, y , 50
.02 765 (6)
177 (5)
50-60
2,406 (19)
579 (17)
. 60
9,755 (75)
2,633 (78)
,2
5,138 (40)
910 (27)
2-3
3,657 (28)
1,017 (30)
. 3-5
3,235 (25)
1,110 (33)
. 5-7
896 (7)
, .0001
Tumor size, cm
352 (10)
Sex
.24
Male
6,076 (47)
1,631 (48)
Female
6,850 (53)
1,758 (52)
Race/ethnicity White Black
.03 10,420 (81) 991 (8)
2,657 (78) 275 (8)
Hispanic
601 (5)
179 (5)
Other
914 (8)
278 (8)
7,504 (58)
1,975 (58)
Marital status Married
.85
Single
1,308 (10)
329 (10)
Divorced/separated
1,669 (13)
430 (13)
Widowed
2,081 (16)
565 (17)
364 (3)
90 (3)
Other Tumor site Upper lobe
.007 7,949 (62)
2,058 (61)
Middle lobe
644 (5)
214 (6)
Lower lobe
4,139 (32)
1,054 (31)
Other
194 (2)
63 (2) , .0001
Histology Adenocarcinoma
7,675 (59)
2,109 (62)
Squamous cell
3,499 (27)
774 (23)
Large cell Other
469 (4) 1,283 (10)
129 (4) 377 (11)
LNs examined Mean No.
.01 8.34
8.69 , .0001
Radiation therapy No Yes
P Value
12,612 (98) 314 (2)
3,189 (94) 200 (6)
Data given as No. (%) unless otherwise indicated. LN 5 lymph node; VPI 5 visceral pleural involvement.
sample size, patient selection, or cut points for stratification. To our knowledge, our study is the only one in the literature on VPI to combine a large population-based cohort with granular stratification cut points that match
1622 Original Research
the latest AJCC classifications. In addition to using adjusted analysis to evaluate the impact of VPI on survival, we devised a staging classification to better account for this prognostic effect.
[
146#6 CHEST DECEMBER 2014
]
Figure 2 – A, Lung cancer-specific survival by tumor size and VPI status. B, Overall survival by tumor size and VPI status. VPI 5 visceral pleural involvement.
Our results support the need to revise the AJCC classification to better address the negative impact of VPI on prognosis among tumors , 7 cm. Given the relatively high frequency of VPI in early-stage NSCLC, reclassification is important to convey accurate prognostic information to patients and to select homogeneous groups of patients for inclusion in clinical trials. Presently, recommendations for adjuvant therapy in journal.publications.chestnet.org
patients with T2N0 tumors call for observation or chemotherapy for high-risk patients (where VPI is considered a high-risk factor).25 In the absence of evidence from randomized trials regarding the optimal treatment of this group, these data may help in decisions concerning adjuvant therapy in patients with VPI. Additionally, our proposed modifications to the T-status classifications would result in a number of tumors , 2 cm with VPI 1623
TABLE 2
] Five-Year Survival Rates According to Tumor Size and VPI
5-y Survival (95% CI) Category
Lung Cancer-Specific (%)
Overall (%)
, 2 cm without VPI
88 (86-89)
75 (74-77)
, 2 cm with VPI
84 (81-87)
70 (66-75)
2-3 cm without VPI
79 (77-81)
68 (65-70)
2-3 cm with VPI
71 (67-75)
61 (56-65)
3-5 cm without VPI
72 (70-74)
60 (57-62)
3-5 cm with VPI
66 (62-69)
53 (49-57)
5-7 cm without VPI
67 (63-71)
57 (52-61)
5-7 cm with VPI
62 (55-68)
48 (40-55)
. 7 cm or T3
56 (52-60)
44 (41-47)
See Table 1 legend for expansion of abbreviation.
(presently considered stage IB) to be reassigned to stage IA disease. This change in staging could impact treatment decisions, as adjuvant chemotherapy may presently be considered for patients with stage IB disease with VPI but not for patients with stage IA disease. Moreover, our proposed modifications would result in tumors 3 to 5 cm with VPI being assigned to stage IIA (instead of stage IB), and tumors 5 to 7 cm with VPI being assigned to stage IIB (instead of stage IIA). In both of these categories, adjuvant chemotherapy would be more likely to be recommended. However, information from prospective studies is warranted to determine the best postoperative treatment of these cancers. These results also suggest that VPI should be included as a stratification variable in future randomized trials of adjuvant chemotherapy.
TABLE 3
] Adjusteda Association Between Tumor Size and VPI Categories and Survival
Category , 2 cm without VPI
Lung Cancer-Specific HR (95% CI) Overall HR (95% CI) Reference
Reference
, 2 cm with VPI
1.46 (1.19-1.79) 1.24 (1.05-1.46)
2-3 cm without VPI
1.47 (1.29-1.67) 1.28 (1.16-1.42)
2-3 cm with VPI
2.05 (1.73-2.43) 1.64 (1.43-1.88)
3-5 cm without VPI
1.87 (1.65-2.12) 1.56 (1.41-1.72)
3-5 cm with VPI
2.61 (2.24-3.04) 1.90 (1.67-2.15)
5-7 cm without VPI
2.45 (2.07-2.90) 1.86 (1.63-2.14)
5-7 cm with VPI
3.15 (2.53-3.94) 2.35 (1.95-2.82)
T3
3.89 (3.42-4.43) 2.75 (2.48-3.05)
HR 5 hazard ratio. See Table 1 legend for expansion of other abbreviations. Adjusted for age, sex, race/ethnicity, marital status, tumor size, tumor site, histology, LNs examined, and receipt of postoperative radiation therapy.
a
1624 Original Research
This study has some strengths and limitations worth noting. We used high-quality, representative, populationbased data with a large number of cases; as such, its results are widely generalizable compared with single- or even multicenter studies. In particular, the nature of the SEER registry makes it less susceptible to selection bias or influence of regional or center-specific practice patterns. The large number of cases allowed us to stratify our study by size and VPI while still maintaining adequate cohort sizes. The SEER registry benefits from standardized data collection and excellent ascertainment, as well as long-term follow-up on patients. The limitations of this registry include lack of information on the presence of lymphovascular invasion, a recognized negative prognostic factor.20 Work has shown that depth of VPI is associated with lung cancer outcome, suggesting a role for even more precise characterization of pleural involvement.10,18,26-28 These findings have been incorporated by the recent American College of Chest Physicians (CHEST) guidelines for lung cancer pathologic evaluation.4 Unfortunately, we were unable to assess the depth of VPI, as this information is not provided by SEER. Thus, further work is needed to determine whether the AJCC staging system should be revised to incorporate the depth of VPI. Furthermore, because SEER does not conduct an independent review documenting the nature and/or depth of VPI, we were unable to determine which histologic methods (ie, hematoxylin and eosin staining, elastic staining, immunohistochemistry) were performed to ascertain its presence. Thus, the prevalence of VPI may have been underestimated in our study in cases evaluated with less sensitive techniques. However, underascertainment of VPI would drive our results toward the null and should not explain the observed association of VPI with worse survival. These findings, therefore, reflect the standard practice of cancer care in the United States and the prognostic impact of VPI within this context. We did not have information about the use of adjuvant chemotherapy, though higher use of chemotherapy among patients with VPI would bias the results toward the null. As discussed, prior to 2010, cases were coded identically for VPI or pulmonary ligament involvement, though the potential number of misclassified cases in our study is likely very small. Finally, because the SEER registry uses death certificates to ascertain cause of death, there is a possibility of misclassification. However, our results were consistent when using all-cause mortality as the outcome of interest. It has also been shown that death certificate data reflect lung cancer mortality with high rates of accuracy.29
[
146#6 CHEST DECEMBER 2014
]
TABLE 4
] Current AJCC vs Proposed Revised Classi¿cation Proposed Revision
AJCC 7th Edition5 Category
T Status
Stage
T Status
Stage
, 2 cm without VPI
T1a
IA
T1a
IA
, 2 cm with VPI
T2a
IB
T1b
IA
2-3 cm without VPI
T1b
IA
T1b
IA
2-3 cm with VPI
T2a
IB
T2a
IB
3-5 cm without VPI
T2a
IB
T2a
IB
3-5 cm with VPI
T2a
IB
T2b
IIA
5-7 cm without VPI
T2b
IIA
T2b
IIA
5-7 cm with VPI
T2b
IIA
T3
IIB
T3
IIB
T3
IIB
. 7 cm
AJCC 5 American Joint Committee on Cancer. See Table 1 legend for expansion of other abbreviation.
Conclusions We found that VPI is associated with a consistent, size-independent, negative impact in NSCLC tumors , 7 cm, a prognostic factor not adequately captured in the current AJCC system. We propose an alternate
Acknowledgments Author contributions: J. P. W. had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. S. L. served as principal author. S. L., J. E. G., R. M. F., and J. P. W. contributed to the study concept and design; S. L. and J. P. W. contributed to data analysis and interpretation and drafting of the manuscript; J. E. G., R. M. F., and J. P. W. contributed to critical revision of the manuscript; and S. L., J. E. G., R. M. F., and J. P. W. approved the final version of the manuscript. Financial/nonfinancial disclosures: The authors have reported to CHEST the following conflicts of interest: Dr Wisnivesky is a member of the research board of EHE International, has received honoraria from Merck & Co Inc and IMS Health Inc, and research grants from GlaxoSmithKline plc and Sanofi-Aventis US LLC. The other authors have reported to CHEST that no potential conflicts of interest exist with any companies/organizations whose products or services may be discussed in this article.
References 1. Kodama K, Doi O, Higashiyama M, Yokouchi H, Tatsuta M. Long-term results of postoperative intrathoracic chemo-thermotherapy for lung cancer with pleural dissemination. Cancer. 1993;72(2):426-431. 2. Shimizu K, Yoshida J, Nagai K, et al. Visceral pleural invasion classification in non-small cell lung cancer: a proposal on the basis of outcome assessment. J Thorac Cardiovasc Surg. 2004;127(6):1574-1578.
journal.publications.chestnet.org
T-classification system for future iterations of the AJCC staging guidelines that better accounts for the prognostic significance of VPI. Patients with VPI should be considered candidates for more aggressive management.
3. Brewer LA, Bai AF, Little JN, Rabago y Pardo G. Carcinoma of the lung; practical classification for early diagnosis and surgical treatment. J Am Med Assoc. 1958;166(10):1149-1155. 4. Schwartz AM, Rezaei MK. Diagnostic surgical pathology in lung cancer: diagnosis and management of lung cancer, 3rd ed: American College of Chest Physicians evidence-based clinical practice guidelines. Chest. 2013;143(5_suppl): e251S-e262S.
11.
12.
5. Goldstraw P, Crowley J, Chansky K, et al. AJCC Cancer Staging Manual. New York, NY: Springer; 2010. 6. Jones DR, Daniel TM, Denlinger CE, et al. Stage IB nonsmall cell lung cancers: are they all the same? Ann Thorac Surg. 2006;81(6):1958-1962. 7. Fibla JJ, Cassivi SD, Brunelli A, et al. Re-evaluation of the prognostic value of visceral pleura invasion in stage IB non-small cell lung cancer using the prospective multicenter ACOSOG Z0030 trial data set. Lung Cancer. 2012;78(3):259-262. 8. Koike T, Koike T, Yoshiya K, Tsuchida M, Toyabe S. Risk factor analysis of locoregional recurrence after sublobar resection in patients with clinical stage IA non-small cell lung cancer. J Thorac Cardiovasc Surg. 2013;146(2):372-378. 9. Manac’h D, Riquet M, Medioni J, Le Pimpec-Barthes F, Dujon A, Danel C. Visceral pleura invasion by non-small cell lung cancer: an underrated bad prognostic factor. Ann Thorac Surg. 2001;71(4):1088-1093. 10. Osaki T, Nagashima A, Yoshimatsu T, et al. Visceral pleural involvement in nonsmall
13.
14.
15.
16.
17.
cell lung cancer: prognostic significance. Ann Thorac Surg. 2004;77(5):1769-1773. Ou SH, Zell JA, Ziogas A, Anton-Culver H. Prognostic significance of the nonsize-based AJCC T2 descriptors: visceral pleura invasion, hilar atelectasis, or obstructive pneumonitis in stage IB nonsmall cell lung cancer is dependent on tumor size. Chest. 2008;133(3):662-669. Shim HS, Park IK, Lee CY, Chung KY. Prognostic significance of visceral pleural invasion in the forthcoming (seventh) edition of TNM classification for lung cancer. Lung Cancer. 2009;65(2): 161-165. Shimizu K, Yoshida J, Nagai K, et al. Visceral pleural invasion is an invasive and aggressive indicator of non-small cell lung cancer. J Thorac Cardiovasc Surg. 2005;130(1):160-165. Wang WL, Shen-Tu Y, Wang ZQ. Prognostic factors for survival of stage IB upper lobe non-small cell lung cancer patients: a retrospective study in Shanghai, China. Chin J Cancer Res. 2011;23(4): 265-270. Yoshida J, Nagai K, Asamura H, et al; Japanese Joint Committee for Lung Cancer Registration. Visceral pleura invasion impact on non-small cell lung cancer patient survival: its implications for the forthcoming TNM staging based on a large-scale nation-wide database. J Thorac Oncol. 2009;4(8):959-963. David E, Thall PF, Kalhor N, et al. Visceral pleural invasion is not predictive of survival in patients with lung cancer and smaller tumor size. Ann Thorac Surg. 2013;95(6):1872-1877. Hung JJ, Jeng WJ, Chou TY, et al. Prognostic value of the new International
1625
Association for the Study of Lung Cancer/American Thoracic Society/ European Respiratory Society lung adenocarcinoma classification on death and recurrence in completely resected stage I lung adenocarcinoma. Ann Surg. 2013;258(6):1079-1086. 18. Kawase A, Yoshida J, Miyaoka E, et al. Visceral pleural invasion classification in non-small-cell lung cancer in the 7th edition of the tumor, node, metastasis classification for lung cancer: validation analysis based on a large-scale nationwide database. J Thorac Oncol. 2013;8(5):606-611. 19. Nitadori J, Colovos C, Kadota K, et al. Visceral pleural invasion does not affect recurrence or overall survival among patients with lung adenocarcinoma ⱕ 2 cm: a proposal to reclassify T1 lung adenocarcinoma. Chest. 2013; 144(5):1622-1631. 20. Yanagawa N, Shiono S, Abiko M, Ogata SY, Sato T, Tamura G. Prognostic impact and initial recurrence site of lymphovascular and visceral pleural invasion in surgically resected stage I
1626 Original Research
21.
22.
23.
24.
25. National Comprehensive Cancer Network. Non-small Cell Lung Cancer, Version 2. Fort Washington, PA: National Comprehensive Cancer Network; 2013. 26. Sakakura N, Mori S, Okuda K, et al. Subcategorization of lung cancer based on tumor size and degree of visceral pleural invasion. Ann Thorac Surg. 2008;86(4):1084-1090. 27. Travis WD, Brambilla E, Rami-Porta R, et al. Visceral pleural invasion: pathologic criteria and use of elastic stains: proposal for the 7th edition of the TNM classification for lung cancer. J Thorac Oncol. 2008;3(12): 1384-1390. 28. Warth A, Muley T, Herpel E, et al. A histochemical approach to the diagnosis of visceral pleural infiltration by nonsmall cell lung cancer. Pathol Oncol Res. 2010;16(1):119-123. 29. Percy C, Stanek E III, Gloeckler L. Accuracy of cancer death certificates and its effect on cancer mortality statistics. Am J Public Health. 1981;71(3): 242-250.
non-small-cell lung carcinoma. Eur J Cardiothorac Surg. 2013;44(3):e200-e206. Surveillance, Epidemiology, and End Results Research Data (1973-2010), National Cancer Institute, DCCPS, Surveillance Research Program, Surveillance Systems Branch, released April 2013, based on the November 2012 submission. National Cancer Institute website. http://seer.cancer.gov. Accessed August 21, 2013. Surveillance Epidemiology, and End Results Program. About the SEER Program. National Cancer Institute website. http://seer.cancer.gov/about/. Accessed January 15, 2014. Adamo MB, Johnson CH, Ruhl JL, et al. 2013 SEER Program Coding and Staging Manual. Bethesda, MD: National Cancer Institute; 2013. Collaborative Stage Work Group of the American Joint Committee on Cancer. Collaborative Stage Data Collection System User Documentation and Coding Instructions, Version 02.04.40. Chicago, IL: American Joint Committee on Cancer; 2012.
[
146#6 CHEST DECEMBER 2014
]
Original Research Lung Cancer
]
Prognostic Significance of Visceral Pleural Involvement in Early-Stage Lung Cancer Sameer Lakha, BA; Jorge E. Gomez, MD; Raja M. Flores, MD, FCCP; and Juan P. Wisnivesky, MD, DrPH
Visceral pleural invasion (VPI) may impact non-small cell lung cancer (NSCLC) survival. However, previous studies are mixed as to whether VPI is an independent prognostic factor in early-stage cancers and whether its effect is size dependent. In the current American Joint Committee on Cancer (AJCC) staging system, VPI leads to upstaging of cancers , 3 cm but not of those 3 to 7 cm in size. BACKGROUND:
Using the Surveillance, Epidemiology, and End Results (SEER) registry, we identified 16,315 patients with stage I-II NSCLC treated with lobectomy. We used the Kaplan-Meier method and Cox regression to assess the association of VPI with lung cancer-specific (primary outcome) and overall survival. Based on these results, we created a revised VPI staging classification.
METHODS:
Overall, 3,389 patients (21%) had VPI. Kaplan-Meier analysis stratified by tumor size showed worse cancer-specific survival in patients with VPI (P , .0001). VPI was independently associated with decreased lung cancer-specific survival (hazard ratio, 1.38; 95% CI, 1.29-1.47) after controlling for tumor size and other confounders; this effect was not size dependent. In our revised classification, tumors , 7 cm with VPI were upstaged to the next T category.
RESULTS:
VPI is a prevalent finding associated with worse prognosis in early-stage lung cancer, even among patients with tumors . 3 cm, a factor not captured in the current staging system. Patients with VPI may be considered candidates for more aggressive treatment.
CONCLUSIONS:
CHEST 2014; 146(6):1619-1626
Manuscript received January 23, 2014; revision accepted June 4, 2014; originally published Online First July 17, 2014. ABBREVIATIONS: AJCC 5 American Joint Committee on Cancer; CSE 5 Collaborative Staging Extension; HR 5 hazard ratio; LN 5 lymph node; NSCLC 5 non-small cell lung cancer; SEER 5 Surveillance, Epidemiology, and End Results; VPI 5 visceral pleural involvement AFFILIATIONS: From the Division of General Internal Medicine (Mr Lakha and Dr Wisnivesky), Tisch Cancer Institute (Dr Gomez), Department of Thoracic Surgery (Dr Flores), and Division of Pulmonary, Critical Care, and Sleep Medicine (Dr Wisnivesky), Icahn School of Medicine at Mount Sinai, New York, NY.
journal.publications.chestnet.org
Part of this article was presented in abstract form at the American Thoracic Society International Conference, May 19, 2014, San Diego, CA. FUNDING/SUPPORT: The authors report to CHEST that no funding or other support was received for this study. CORRESPONDENCE TO: Sameer Lakha, BA, Department of Medicine, Icahn School of Medicine at Mount Sinai, One Gustave L. Levy Pl, Box 1087, New York, NY 10029; e-mail: [email protected] © 2014 AMERICAN COLLEGE OF CHEST PHYSICIANS. Reproduction of this article is prohibited without written permission from the American College of Chest Physicians. See online for more details. DOI: 10.1378/chest.14-0204
1619
Accurate staging of lung cancer is critical for optimizing treatment decisions, communicating prognosis, and determining eligibility for clinical trials. Tumor spread to the visceral pleura has been recognized as a poor prognostic factor in non-small cell lung cancer (NSCLC).1-4 Consequently, visceral pleural involvement (VPI) has been incorporated into the American Joint Committee on Cancer (AJCC) TNM staging system alongside tumor size.5 According to these criteria, tumors , 3 cm that would otherwise be classified T1a (, 2 cm) or T1b (2-3 cm) disease based on size alone are considered to be equivalent to more advanced tumors (T2a; 3-5 cm) if they extend into the visceral pleura. Conversely, for larger tumors (ie, . 3 cm), VPI does not result in upstaging.5 Several studies have evaluated whether VPI is an independent prognostic factor in patients with early-stage
Materials and Methods We used data from the latest release of the Surveillance, Epidemiology, and End Results (SEER) registry.21 SEER is a network of cancer registries that covers approximately 28% of the population of the United States and collects data on all new cases of cancer in its coverage regions.22 From SEER, we identified all incident cases of histologically confirmed NSCLC diagnosed between 2004 and 2010. The study included patients for whom NSCLC was their first and only primary malignancy, who had tumors , 7 cm, without lymph node (LN) involvement (N0) or distant metastasis (M0), and who underwent lobectomy. We excluded patients diagnosed on autopsy or death certificate data, as well as cases with missing tumor size and stage information. We further excluded patients with tumors , 3 cm but classified as having T2 disease due to hilar atelectasis or obstructive pneumonitis, patients with T3 or T4 disease regardless of tumor size, and patients treated with preoperative radiotherapy (due to concerns about possible understaging). SEER reports data on sociodemographic characteristics, including age, sex, race/ethnicity, and marital status. Cancer information includes year of diagnosis, tumor location (upper lobe, middle lobe, lower lobe, or central), and histology (adenocarcinoma, large-cell carcinoma, squamous cell carcinoma, or other). The registry includes detailed pathologic staging data collected within 4 months of diagnosis or within completion of the first course of treatment.23,24 Data are available regarding local tumor extension and size, LN involvement, and presence or absence of systemic metastasis. TNM data are also provided according to the AJCC staging system. Based on this classification, cases were grouped according to tumor size into the following categories: , 2 cm, 2 to 3 cm, 3 to 5 cm, and 5 to 7 cm. The presence of VPI was ascertained using Collaborative Staging Extension (CSE) codes (410, 420, 430, and 450). Of note, VPI cases prior to 2010 were classified using the CSE code 450, which includes both VPI and extension to the pulmonary ligament (ie, the fold of parietal pleura that forms and extends inferiorly from the hilum). In 2010, SEER introduced a distinct CSE code (440) for pulmonary ligament involvement. Of the 25,891 NSCLC cases diagnosed in 2010, 2,074 (8.01%) were coded as having VPI (CSE codes 410, 420, 430) compared with only 9 (0.03%) identified as having pulmonary ligament involvement (CSE code 440). Moreover, all nine of these cases were excluded based on other cohort selection criteria. Therefore, we inferred that the number of pulmonary ligament cases among patients with CSE code 450 prior
1620 Original Research
NSCLC, with varying results.2,6-20 Single-center and multicenter series, as well as analyses of patients enrolled in a clinical trial, found that VPI confers worse prognosis among resected tumors , 7 cm.2,9,12,15,17,18,20 However, the prognostic significance of VPI was not confirmed in adjusted analysis in many of these studies. In contrast, several other studies showed either no impact of VPI on prognosis, or a differential effect based on the size of the tumor.6,7,11,12,16,19 Thus, the independent prognostic significance of VPI, as well as its proper role in the staging of these cancers, remains unclear. In this population-based study, we used data from a large cohort of patients with surgically resected stage I and II NSCLC to better characterize the prognostic impact of VPI in tumors , 7 cm and to refine the staging of these cancers.
to 2010 included in the study was negligible, suggesting minimal misclassification. SEER reports data on surgical treatment and radiation therapy use during the first course of cancer-directed treatment. Patients undergoing lobectomy were identified using SEER surgical codes 30 through 48. Data are also provided regarding the number of LNs sampled during surgery. Radiotherapy and radiation-surgery sequence codes were used to ascertain whether patients underwent postoperative external beam radiotherapy. SEER does not report information about chemotherapy use, due to concerns about misclassification. The primary study outcome was lung cancer-specific mortality, as the study goal was to assess the impact of VPI on lung cancer prognosis; secondary analyses used overall survival. Survival time was calculated as the number of months from cancer diagnosis until death or last follow-up (December 31, 2010) for censored observations. Cause of death was coded by SEER using information extracted from death certificate data. Patients who died of non-lung cancer causes were classified as censored at the date of death in analyses of lung cancer-specific survival. Statistical Analysis Differences in the baseline characteristics of patients with and without VPI were compared using t tests or x2 test, as appropriate. The unadjusted association of VPI with survival was assessed using Kaplan-Meier analysis stratifying the sample by tumor size categories (, 2 cm, 2-3 cm, 3-5 cm, and 5-7 cm); tumors . 7 cm or of T3 status were included for comparison. Adjusted analysis was conducted using a Cox proportional hazards model to assess the association between VPI and survival after controlling for age, sex, race/ethnicity, marital status, histology, tumor size, number of LNs examined, and use of adjuvant radiotherapy. The analysis was repeated including interactions between tumor size and VPI to assess if the effect of pleural involvement varied across different tumor size categories. To combine groups with similar prognoses, we fitted a Cox model including one dichotomous indicator for each tumor size category and VPI status (ie, , 2 cm without VPI, , 2 cm with VPI, etc). Based on the findings of this analysis, we grouped cancers with overlapping hazard ratios (HRs), thereby developing a revised prognostic system. Results were deemed significant at a two-sided P value , .05. All statistical analysis was conducted using SAS software, version 9.2 (SAS Institute Inc). The study was determined to be exempt research by the Institutional Review Board of the Icahn School of Medicine at Mount Sinai.
[
146#6 CHEST DECEMBER 2014
]
Results The cohort selection process is shown in Figure 1. Baseline demographic and clinical characteristics of patients with and without VPI are provided in Table 1. Relative to patients without VPI, patients with VPI were older (P 5 .0002), less likely to be white (P 5 .03), had larger tumors (P , .0001), were more likely to present with adenocarcinoma (P , .0001), had more LNs evaluated at surgery (P 5 .01), and were more likely to have undergone adjuvant radiotherapy (P , .0001). There were no significant differences in the distribution of sex (P 5 .24) or marital status (P 5 .85). Lung cancer-specific (P , .0001) and overall (P , .0001) survival were significantly different across tumor size and VPI strata (Fig 2). Five-year survival rates according to tumor size and VPI status are shown in Table 2. Overall, VPI was associated with worse survival for all tumor size groups, with prognosis of patients with VPI being similar to that of patients with the adjacent larger tumor size category. Cox analysis demonstrated that VPI confers a significant lung cancer-specific survival disadvantage
Figure 1 – Cohort selection criteria. The process by which we identified our study cohort is shown. NSCLC 5 non-small cell lung cancer; RT 5 radiotherapy; SEER 5 Surveillance, Epidemiology, and End Results.
journal.publications.chestnet.org
(HR, 1.38; 95% CI, 1.29-1.47), even after controlling for size and other potential confounders. Interactions between tumor size and VPI were not statistically significant (P value range 5 .4-.6), suggesting an equivalent effect of VPI in all tumor size categories. Similar results were observed in analyses with overall survival as the end point. HRs for individual size-VPI strata are shown in Table 3. As with the 5-year survival rates in Table 2, the HRs and their CIs cluster into groups that do not track the current AJCC stage categories. Based on the groups that emerged from our results, we developed a revised classification in which tumors with VPI were upstaged to the next T-status category, resulting in the following groups: , 2 cm without VPI; , 2 cm with VPI and 2 to 3 cm without VPI; 2 to 3 cm with VPI and 3 to 5 cm without VPI; 3 to 5 cm with VPI and 5 to 7 cm without VPI; 5 to 7 cm with VPI and . 7 cm and T3 tumors (Table 4).
Discussion The prognostic impact of VPI in early-stage NSCLC has been unclear. This study confirms a size-independent, negative prognostic effect of VPI in a large populationbased sample. Moreover, a revised tumor classification incorporating this finding had equivalent discrimination and may help to better select VPI patients with poorer prognosis. These results have implications for the treatment of patients with VPI, a frequent finding affecting 25% of patients with early-stage NSCLC. Studies have shown an association between VPI and survival. Koike et al8 found an independent, negative prognostic effect of VPI on survival in a single-center series. In larger registry-based studies, Yoshida et al15 and Kawase et al18 found worse prognosis among patients with VPI and suggested revised staging classifications similar to ours; however, these studies did not adjust for confounders to assess the independence and size-dependence of this association. Other studies show limited or no association between VPI and prognosis. For instance, using a single-center study of adenocarcinoma cases treated with lobectomy or limited resection, Nitadori et al19 found an impact on survival among patients with tumors 2 to 3 cm, but not those , 2 cm. Similarly, David et al16 found no association between VPI and survival in tumors , 5 cm. Yanagawa et al20 found no independent association between VPI and survival. In contrast, we found a size-independent, negative association in all tumors , 7 cm on unadjusted and adjusted analyses. These differences may be due to 1621
TABLE 1
] Baseline Characteristics of Study Subjects According to VPI
Characteristic
No VPI (n 5 12,926)
VPI (n 5 3,389)
Age, y , 50
.02 765 (6)
177 (5)
50-60
2,406 (19)
579 (17)
. 60
9,755 (75)
2,633 (78)
,2
5,138 (40)
910 (27)
2-3
3,657 (28)
1,017 (30)
. 3-5
3,235 (25)
1,110 (33)
. 5-7
896 (7)
, .0001
Tumor size, cm
352 (10)
Sex
.24
Male
6,076 (47)
1,631 (48)
Female
6,850 (53)
1,758 (52)
Race/ethnicity White Black
.03 10,420 (81) 991 (8)
2,657 (78) 275 (8)
Hispanic
601 (5)
179 (5)
Other
914 (8)
278 (8)
7,504 (58)
1,975 (58)
Marital status Married
.85
Single
1,308 (10)
329 (10)
Divorced/separated
1,669 (13)
430 (13)
Widowed
2,081 (16)
565 (17)
364 (3)
90 (3)
Other Tumor site Upper lobe
.007 7,949 (62)
2,058 (61)
Middle lobe
644 (5)
214 (6)
Lower lobe
4,139 (32)
1,054 (31)
Other
194 (2)
63 (2) , .0001
Histology Adenocarcinoma
7,675 (59)
2,109 (62)
Squamous cell
3,499 (27)
774 (23)
Large cell Other
469 (4) 1,283 (10)
129 (4) 377 (11)
LNs examined Mean No.
.01 8.34
8.69 , .0001
Radiation therapy No Yes
P Value
12,612 (98) 314 (2)
3,189 (94) 200 (6)
Data given as No. (%) unless otherwise indicated. LN 5 lymph node; VPI 5 visceral pleural involvement.
sample size, patient selection, or cut points for stratification. To our knowledge, our study is the only one in the literature on VPI to combine a large population-based cohort with granular stratification cut points that match
1622 Original Research
the latest AJCC classifications. In addition to using adjusted analysis to evaluate the impact of VPI on survival, we devised a staging classification to better account for this prognostic effect.
[
146#6 CHEST DECEMBER 2014
]
Figure 2 – A, Lung cancer-specific survival by tumor size and VPI status. B, Overall survival by tumor size and VPI status. VPI 5 visceral pleural involvement.
Our results support the need to revise the AJCC classification to better address the negative impact of VPI on prognosis among tumors , 7 cm. Given the relatively high frequency of VPI in early-stage NSCLC, reclassification is important to convey accurate prognostic information to patients and to select homogeneous groups of patients for inclusion in clinical trials. Presently, recommendations for adjuvant therapy in journal.publications.chestnet.org
patients with T2N0 tumors call for observation or chemotherapy for high-risk patients (where VPI is considered a high-risk factor).25 In the absence of evidence from randomized trials regarding the optimal treatment of this group, these data may help in decisions concerning adjuvant therapy in patients with VPI. Additionally, our proposed modifications to the T-status classifications would result in a number of tumors , 2 cm with VPI 1623
TABLE 2
] Five-Year Survival Rates According to Tumor Size and VPI
5-y Survival (95% CI) Category
Lung Cancer-Specific (%)
Overall (%)
, 2 cm without VPI
88 (86-89)
75 (74-77)
, 2 cm with VPI
84 (81-87)
70 (66-75)
2-3 cm without VPI
79 (77-81)
68 (65-70)
2-3 cm with VPI
71 (67-75)
61 (56-65)
3-5 cm without VPI
72 (70-74)
60 (57-62)
3-5 cm with VPI
66 (62-69)
53 (49-57)
5-7 cm without VPI
67 (63-71)
57 (52-61)
5-7 cm with VPI
62 (55-68)
48 (40-55)
. 7 cm or T3
56 (52-60)
44 (41-47)
See Table 1 legend for expansion of abbreviation.
(presently considered stage IB) to be reassigned to stage IA disease. This change in staging could impact treatment decisions, as adjuvant chemotherapy may presently be considered for patients with stage IB disease with VPI but not for patients with stage IA disease. Moreover, our proposed modifications would result in tumors 3 to 5 cm with VPI being assigned to stage IIA (instead of stage IB), and tumors 5 to 7 cm with VPI being assigned to stage IIB (instead of stage IIA). In both of these categories, adjuvant chemotherapy would be more likely to be recommended. However, information from prospective studies is warranted to determine the best postoperative treatment of these cancers. These results also suggest that VPI should be included as a stratification variable in future randomized trials of adjuvant chemotherapy.
TABLE 3
] Adjusteda Association Between Tumor Size and VPI Categories and Survival
Category , 2 cm without VPI
Lung Cancer-Specific HR (95% CI) Overall HR (95% CI) Reference
Reference
, 2 cm with VPI
1.46 (1.19-1.79) 1.24 (1.05-1.46)
2-3 cm without VPI
1.47 (1.29-1.67) 1.28 (1.16-1.42)
2-3 cm with VPI
2.05 (1.73-2.43) 1.64 (1.43-1.88)
3-5 cm without VPI
1.87 (1.65-2.12) 1.56 (1.41-1.72)
3-5 cm with VPI
2.61 (2.24-3.04) 1.90 (1.67-2.15)
5-7 cm without VPI
2.45 (2.07-2.90) 1.86 (1.63-2.14)
5-7 cm with VPI
3.15 (2.53-3.94) 2.35 (1.95-2.82)
T3
3.89 (3.42-4.43) 2.75 (2.48-3.05)
HR 5 hazard ratio. See Table 1 legend for expansion of other abbreviations. Adjusted for age, sex, race/ethnicity, marital status, tumor size, tumor site, histology, LNs examined, and receipt of postoperative radiation therapy.
a
1624 Original Research
This study has some strengths and limitations worth noting. We used high-quality, representative, populationbased data with a large number of cases; as such, its results are widely generalizable compared with single- or even multicenter studies. In particular, the nature of the SEER registry makes it less susceptible to selection bias or influence of regional or center-specific practice patterns. The large number of cases allowed us to stratify our study by size and VPI while still maintaining adequate cohort sizes. The SEER registry benefits from standardized data collection and excellent ascertainment, as well as long-term follow-up on patients. The limitations of this registry include lack of information on the presence of lymphovascular invasion, a recognized negative prognostic factor.20 Work has shown that depth of VPI is associated with lung cancer outcome, suggesting a role for even more precise characterization of pleural involvement.10,18,26-28 These findings have been incorporated by the recent American College of Chest Physicians (CHEST) guidelines for lung cancer pathologic evaluation.4 Unfortunately, we were unable to assess the depth of VPI, as this information is not provided by SEER. Thus, further work is needed to determine whether the AJCC staging system should be revised to incorporate the depth of VPI. Furthermore, because SEER does not conduct an independent review documenting the nature and/or depth of VPI, we were unable to determine which histologic methods (ie, hematoxylin and eosin staining, elastic staining, immunohistochemistry) were performed to ascertain its presence. Thus, the prevalence of VPI may have been underestimated in our study in cases evaluated with less sensitive techniques. However, underascertainment of VPI would drive our results toward the null and should not explain the observed association of VPI with worse survival. These findings, therefore, reflect the standard practice of cancer care in the United States and the prognostic impact of VPI within this context. We did not have information about the use of adjuvant chemotherapy, though higher use of chemotherapy among patients with VPI would bias the results toward the null. As discussed, prior to 2010, cases were coded identically for VPI or pulmonary ligament involvement, though the potential number of misclassified cases in our study is likely very small. Finally, because the SEER registry uses death certificates to ascertain cause of death, there is a possibility of misclassification. However, our results were consistent when using all-cause mortality as the outcome of interest. It has also been shown that death certificate data reflect lung cancer mortality with high rates of accuracy.29
[
146#6 CHEST DECEMBER 2014
]
TABLE 4
] Current AJCC vs Proposed Revised Classi¿cation Proposed Revision
AJCC 7th Edition5 Category
T Status
Stage
T Status
Stage
, 2 cm without VPI
T1a
IA
T1a
IA
, 2 cm with VPI
T2a
IB
T1b
IA
2-3 cm without VPI
T1b
IA
T1b
IA
2-3 cm with VPI
T2a
IB
T2a
IB
3-5 cm without VPI
T2a
IB
T2a
IB
3-5 cm with VPI
T2a
IB
T2b
IIA
5-7 cm without VPI
T2b
IIA
T2b
IIA
5-7 cm with VPI
T2b
IIA
T3
IIB
T3
IIB
T3
IIB
. 7 cm
AJCC 5 American Joint Committee on Cancer. See Table 1 legend for expansion of other abbreviation.
Conclusions We found that VPI is associated with a consistent, size-independent, negative impact in NSCLC tumors , 7 cm, a prognostic factor not adequately captured in the current AJCC system. We propose an alternate
Acknowledgments Author contributions: J. P. W. had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. S. L. served as principal author. S. L., J. E. G., R. M. F., and J. P. W. contributed to the study concept and design; S. L. and J. P. W. contributed to data analysis and interpretation and drafting of the manuscript; J. E. G., R. M. F., and J. P. W. contributed to critical revision of the manuscript; and S. L., J. E. G., R. M. F., and J. P. W. approved the final version of the manuscript. Financial/nonfinancial disclosures: The authors have reported to CHEST the following conflicts of interest: Dr Wisnivesky is a member of the research board of EHE International, has received honoraria from Merck & Co Inc and IMS Health Inc, and research grants from GlaxoSmithKline plc and Sanofi-Aventis US LLC. The other authors have reported to CHEST that no potential conflicts of interest exist with any companies/organizations whose products or services may be discussed in this article.
References 1. Kodama K, Doi O, Higashiyama M, Yokouchi H, Tatsuta M. Long-term results of postoperative intrathoracic chemo-thermotherapy for lung cancer with pleural dissemination. Cancer. 1993;72(2):426-431. 2. Shimizu K, Yoshida J, Nagai K, et al. Visceral pleural invasion classification in non-small cell lung cancer: a proposal on the basis of outcome assessment. J Thorac Cardiovasc Surg. 2004;127(6):1574-1578.
journal.publications.chestnet.org
T-classification system for future iterations of the AJCC staging guidelines that better accounts for the prognostic significance of VPI. Patients with VPI should be considered candidates for more aggressive management.
3. Brewer LA, Bai AF, Little JN, Rabago y Pardo G. Carcinoma of the lung; practical classification for early diagnosis and surgical treatment. J Am Med Assoc. 1958;166(10):1149-1155. 4. Schwartz AM, Rezaei MK. Diagnostic surgical pathology in lung cancer: diagnosis and management of lung cancer, 3rd ed: American College of Chest Physicians evidence-based clinical practice guidelines. Chest. 2013;143(5_suppl): e251S-e262S.
11.
12.
5. Goldstraw P, Crowley J, Chansky K, et al. AJCC Cancer Staging Manual. New York, NY: Springer; 2010. 6. Jones DR, Daniel TM, Denlinger CE, et al. Stage IB nonsmall cell lung cancers: are they all the same? Ann Thorac Surg. 2006;81(6):1958-1962. 7. Fibla JJ, Cassivi SD, Brunelli A, et al. Re-evaluation of the prognostic value of visceral pleura invasion in stage IB non-small cell lung cancer using the prospective multicenter ACOSOG Z0030 trial data set. Lung Cancer. 2012;78(3):259-262. 8. Koike T, Koike T, Yoshiya K, Tsuchida M, Toyabe S. Risk factor analysis of locoregional recurrence after sublobar resection in patients with clinical stage IA non-small cell lung cancer. J Thorac Cardiovasc Surg. 2013;146(2):372-378. 9. Manac’h D, Riquet M, Medioni J, Le Pimpec-Barthes F, Dujon A, Danel C. Visceral pleura invasion by non-small cell lung cancer: an underrated bad prognostic factor. Ann Thorac Surg. 2001;71(4):1088-1093. 10. Osaki T, Nagashima A, Yoshimatsu T, et al. Visceral pleural involvement in nonsmall
13.
14.
15.
16.
17.
cell lung cancer: prognostic significance. Ann Thorac Surg. 2004;77(5):1769-1773. Ou SH, Zell JA, Ziogas A, Anton-Culver H. Prognostic significance of the nonsize-based AJCC T2 descriptors: visceral pleura invasion, hilar atelectasis, or obstructive pneumonitis in stage IB nonsmall cell lung cancer is dependent on tumor size. Chest. 2008;133(3):662-669. Shim HS, Park IK, Lee CY, Chung KY. Prognostic significance of visceral pleural invasion in the forthcoming (seventh) edition of TNM classification for lung cancer. Lung Cancer. 2009;65(2): 161-165. Shimizu K, Yoshida J, Nagai K, et al. Visceral pleural invasion is an invasive and aggressive indicator of non-small cell lung cancer. J Thorac Cardiovasc Surg. 2005;130(1):160-165. Wang WL, Shen-Tu Y, Wang ZQ. Prognostic factors for survival of stage IB upper lobe non-small cell lung cancer patients: a retrospective study in Shanghai, China. Chin J Cancer Res. 2011;23(4): 265-270. Yoshida J, Nagai K, Asamura H, et al; Japanese Joint Committee for Lung Cancer Registration. Visceral pleura invasion impact on non-small cell lung cancer patient survival: its implications for the forthcoming TNM staging based on a large-scale nation-wide database. J Thorac Oncol. 2009;4(8):959-963. David E, Thall PF, Kalhor N, et al. Visceral pleural invasion is not predictive of survival in patients with lung cancer and smaller tumor size. Ann Thorac Surg. 2013;95(6):1872-1877. Hung JJ, Jeng WJ, Chou TY, et al. Prognostic value of the new International
1625
Association for the Study of Lung Cancer/American Thoracic Society/ European Respiratory Society lung adenocarcinoma classification on death and recurrence in completely resected stage I lung adenocarcinoma. Ann Surg. 2013;258(6):1079-1086. 18. Kawase A, Yoshida J, Miyaoka E, et al. Visceral pleural invasion classification in non-small-cell lung cancer in the 7th edition of the tumor, node, metastasis classification for lung cancer: validation analysis based on a large-scale nationwide database. J Thorac Oncol. 2013;8(5):606-611. 19. Nitadori J, Colovos C, Kadota K, et al. Visceral pleural invasion does not affect recurrence or overall survival among patients with lung adenocarcinoma ⱕ 2 cm: a proposal to reclassify T1 lung adenocarcinoma. Chest. 2013; 144(5):1622-1631. 20. Yanagawa N, Shiono S, Abiko M, Ogata SY, Sato T, Tamura G. Prognostic impact and initial recurrence site of lymphovascular and visceral pleural invasion in surgically resected stage I
1626 Original Research
21.
22.
23.
24.
25. National Comprehensive Cancer Network. Non-small Cell Lung Cancer, Version 2. Fort Washington, PA: National Comprehensive Cancer Network; 2013. 26. Sakakura N, Mori S, Okuda K, et al. Subcategorization of lung cancer based on tumor size and degree of visceral pleural invasion. Ann Thorac Surg. 2008;86(4):1084-1090. 27. Travis WD, Brambilla E, Rami-Porta R, et al. Visceral pleural invasion: pathologic criteria and use of elastic stains: proposal for the 7th edition of the TNM classification for lung cancer. J Thorac Oncol. 2008;3(12): 1384-1390. 28. Warth A, Muley T, Herpel E, et al. A histochemical approach to the diagnosis of visceral pleural infiltration by nonsmall cell lung cancer. Pathol Oncol Res. 2010;16(1):119-123. 29. Percy C, Stanek E III, Gloeckler L. Accuracy of cancer death certificates and its effect on cancer mortality statistics. Am J Public Health. 1981;71(3): 242-250.
non-small-cell lung carcinoma. Eur J Cardiothorac Surg. 2013;44(3):e200-e206. Surveillance, Epidemiology, and End Results Research Data (1973-2010), National Cancer Institute, DCCPS, Surveillance Research Program, Surveillance Systems Branch, released April 2013, based on the November 2012 submission. National Cancer Institute website. http://seer.cancer.gov. Accessed August 21, 2013. Surveillance Epidemiology, and End Results Program. About the SEER Program. National Cancer Institute website. http://seer.cancer.gov/about/. Accessed January 15, 2014. Adamo MB, Johnson CH, Ruhl JL, et al. 2013 SEER Program Coding and Staging Manual. Bethesda, MD: National Cancer Institute; 2013. Collaborative Stage Work Group of the American Joint Committee on Cancer. Collaborative Stage Data Collection System User Documentation and Coding Instructions, Version 02.04.40. Chicago, IL: American Joint Committee on Cancer; 2012.
[
146#6 CHEST DECEMBER 2014
]
