Prognostic Significance of the Extent of Visceral Pleural Invasion in Completely Resected Node-Negative Non-small Cell Lung Cancer

CHEST Original Research Prognostic Significance of the Extent of Visceral Pleural Invasion in Completely Resected Node-Negative Non-small Cell Lung Cancer Jung-Jyh Hung, MD, PhD ; Wen-Juei Jeng, MD ; Wen-Hu Hsu, MD ; Teh-Ying Chou, MD, PhD ; Shiou-Fu Lin, MD; and Yu-Chung Wu, MD LUNG CANCER Objective: Visceral pleural invasion (VPI) has been defined as invasion of the tumor beyond the elastic layer (PL1), including invasion to the visceral pleural surface (PL2). The aim of this study was to evaluate the prognostic factors and patterns of recurrence in resected node-negative nonsmall cell lung cancer (NSCLC) with VPI. Methods: We retrospectively reviewed the clinicopathologic characteristics of 355 patients with resected node-negative NSCLC with VPI at Taipei Veterans General Hospital between 1990 and 2006. The prognostic value and patterns of recurrence were analyzed and compared between PL1 and PL2 groups. Results: The median follow-up time was 54.2 months. The 5-year overall survival rate and probability of freedom from recurrence were 61.9% and 66.2%, respectively. The extent of VPI was PL1 in 300 patients (84.5%) and PL2 in 55 (15.5%). During follow-up, 107 patients (30.1%) developed recurrence. The patterns of recurrence included local recurrence only in 20 patients (18.7%), distant metastasis only in 59 (55.1%), and both local recurrence and distant metastasis in 28 (26.2%). Thirteen of the 107 patients (12.1%) with recurrence developed malignant pleural effusion. The percentage of malignant pleural effusion in the PL2 group was significantly higher than that in the PL1 group ( P 5.006). Patients with PL2 had significantly worse overall survival ( P 5.046) and lower probability of freedom from recurrence ( P 5.028) in multivariate analysis. Conclusions: PL2 was a significant prognostic factor for recurrence and worse overall survival in node-negative NSCLC with VPI. This information is important for further design of clinical trials for aggressive adjuvant therapy. CHEST 2012; 142(1):141 150 Abbreviations: HR 5 hazard ratio; IASLC 5 International Association for the Study of Lung Cancer; NSCLC 5 nonsmall cell lung cancer; PL1 5 visceral pleural invasion beyond the elastic layer; PL2 5 visceral pleural invasion to the visceral pleural surface; VPI 5 visceral pleural invasion Lung cancer is the leading cause of cancer death worldwide. Surgical resection is the treatment of choice for early stage non-small cell lung cancer Manuscript received October 6, 2011; revision accepted December 14, 2011. Affiliations: From the Division of Thoracic Surgery (Drs Hung, Hsu, and Wu), Department of Surgery, Taipei Veterans General Hospital and School of Medicine, National Yang-Ming University; Department of Internal Medicine (Dr Jeng), Chang Gung Memorial Hospital, Chang Gung University; Institute of Clinical Medicine (Dr Chou), National Yang-Ming University; and Department of Pathology and Laboratory Medicine (Drs Chou and Lin), Taipei Veterans General Hospital, Taipei, Taiwan. Drs Hung and Wu contributed equally to this article. Funding/Support: This work was supported in part by National Science Council [NSC-100-2314-B-075-001] (Dr Hung) and NSC-100-2314-B-075-004-MY2 (Dr Wu)], Center of Excellence (NSCLC). 1,2 Visceral pleural invasion (VPI) was first included as a non-size-based T2 descriptor in the fifth edition of the TNM classification for lung cancer published in 1997. 3 Although VPI has generally been for Cancer Research at Taipei Veterans General Hospital [DOH100- TD-C-111-007] (Dr Wu), Yen Tjing Ling Medical Foundation [CI-100-15] (Dr Hung) and Taipei Veterans General Hospital [V101B-038] (Dr Hung). Correspondence to: Jung-Jyh Hung, MD, PhD, Division of Thoracic Surgery, Department of Surgery, Taipei Veterans General Hospital, No. 201, Section 2, Shih-Pai Rd, Taipei 112, Taiwan; e-mail: bradley.hung@gmail.com 2012 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.11-2552 journal.publications.chestnet.org CHEST / 142 / 1 / JULY 2012 141

reported as a poor prognostic factor, 4-9 some studies have demonstrated that VPI did not have an impact on survival. 10-15 Hammar 16,17 suggested a classification of pleural invasion as follows: Px and P0, lack of pleural invasion beyond the elastic layer; P1, invasion beyond the elastic layer; P2, invasion to the surface of the visceral pleura; and P3, invasion of the parietal pleura, chest wall, or both. The seventh edition of the TNM classification for lung cancer was published in 2010. 18,19 The International Association for the Study of Lung Cancer (IASLC) proposed the classification of pleural invasion as follows 20 : PL0, tumor within the subpleural lung parenchyma or invading superficially into the pleural connective tissue beneath the elastic layer; PL1, invasion beyond the elastic layer; PL2, invasion to the surface of the visceral pleura; and PL3, invasion into any compartment of the parietal pleura. According to the IASLC proposal, VPI was defined as invasion beyond the elastic layer (PL1), Table 1 Characteristics and Status of Recurrence in 355 Patients With Resected Node-Negative Non-small Cell Lung Cancer With Visceral Pleural Invasion and the Relationship Between Extent of Visceral Pleural Invasion and Clinicopathologic Variables Extent of Visceral Pleural Invasion Variable Total (N 5 355) PL1 (n 5 300) PL2 (n 5 55) P Value Age at operation, y 67.0 10.2 67.1 10.3 66.4 10.0.640 Sex.674 Male 241 (67.9) 205 (68.3) 36 (65.5) Female 114 (32.1) 95 (31.7) 19 (34.5) Smoking index, pack-y 22.7 27.2 23.2 27.9 19.9 23.6.479 Tumor size.004 2 cm 72 (20.3) 68 (22.7) 4 (7.3). 2 to 3 cm 106 (29.8) 79 (26.3) 27 (49.1). 3 to 5 cm 137 (38.6) 116 (38.7) 21 (38.2). 5 to 7 cm 27 (7.6) 25 (8.3) 2 (3.6). 7 cm 13 (3.7) 12 (4.0) 1 (1.8) Histologic type.485 Squamous cell carcinoma 75 (21.1) 66 (22.0) 9 (16.4) Adenocarcinoma 263 (74.1) 219 (73.0) 44 (80.0) Large cell carcinoma 10 (2.8) 8 (2.7) 2 (3.6) Others 7 (2.0) 7 (2.3) 0 (0.0) Extent of pulmonary resection.829 Sublobar resection 37 (10.5) 31 (10.3) 6 (10.9) Lobectomy 279 (78.6) 237 (79.0) 42 (76.4) Bilobectomy 29 (8.1) 23(7.7) 6 (10.9) Pneumonectomy 10 (2.8) 9 (3.0) 1 (1.8) Extent of visceral pleural invasion PL1 300 (84.5) PL2 55 (15.5) T descriptor.333 T2a 315 (88.7) 263 (87.7) 52 (88.7) T2b 27 (7.6) 25 (8.3) 2 (7.6) T3 13 (3.7) 12 (4.0) 1 (3.7) Number of mediastinal LNs dissected 19.3 11.0 19.4 11.1 18.9 10.4.780 Angiolymphatic invasion a.476 Absent 184 (51.8) 154 (76.6) 30 (71.4) Present 59 (16.7) 47 (23.4) 12 (28.6) Unknown 112 (31.5) Histologic grade a.716 Well differentiated 22 (6.2) 19 (8.8) 3 (7.2) Moderately differentiated 185 (52.1) 156 (72.6) 29 (69.0) Poorly differentiated 50 (14.1) 40 (18.6) 10 (23.8) Unknown 98 (27.6) Status of recurrence a.007 Absent 228 (64.2) 201 (71.0) 27 (51.9) Present 107 (30.1) 82 (29.0) 25 (48.1) Unknown 20 (5.7) Data are presented as mean SD or No. (%). LN 5 lymph node; PL1 5 visceral pleural invasion beyond the elastic layer; PL2 5 visceral pleural invasion to the visceral pleural surface. a Data are lacking in some patients for these variables. 142 Original Research

T3. The IASLC also recommended that elastic stains be used in cases when the distinction between PL0 and PL1 is not clear based on evaluation of tumor sections by hematoxylin and eosin sections. 20 Although the prognostic value of VPI has been widely reported in the literature, 4-15 the patterns of recurrence in patients with VPI and comparison of prognostic value between PL1 and PL2 have not been well demonstrated. 7,8,16 In this report, we focus on prognostic factors and patterns of recurrence in patients with completely resected node-negative NSCLC with VPI. We also compare the prognostic value and patterns of recurrence between PL1 and PL2 groups. Materials and Methods Figure 1. A, Cumulative probability of overall survival in patients with node-negative non-small cell lung cancer with visceral pleural invasion. B, Cumulative probability of freedom from recurrence in patients with node-negative non-small cell lung cancer with visceral pleural invasion (log-rank test). including invasion of tumor to the visceral pleural surface (PL2). 20 If a tumor achieves T2 status based on VPI and is 5 cm in size, it is classified as T2a. If such a tumor is. 5 cm but 7 cm in size, it is classified as T2b. If the tumor is. 7 cm, it is classified as Between January 1990 and December 2006, 355 patients who underwent resections for pathological node-negative NSCLC with VPI at Taipei Veterans General Hospital were included in this retrospective study. The study was approved (201110132GB) by the Institutional Review Board of Taipei Veterans General Hospital, and patient consent was waived. The preoperative staging work-up was done as previously described. 21,22 All patients underwent complete resection of lung cancer with mediastinal lymph node dissection/sampling as previously described. 21,22 Histologic typing was determined according to the World Health Organization classification. 23 Determination of disease stages were based on the TNM classification (seventh edition) of the American Joint Committee on Cancer and the International Union Against Cancer. 18,19 Determination of VPI was done as previously described. 22 VPI was classified according to the proposal of IASLC. 20 PL1 and PL2 indicate VPI and are a T2 descriptor. VPI was first examined in tumor sections with hematoxylin and eosin stain. Elastic stains were performed in tumor sections when the status of VPI was indeterminate by hematoxylin and eosin stains. All patients were followed up in our outpatient department as previously described. 21,22 To investigate their impact on overall survival and recurrence-free survival, clinicopathologic factors were analyzed in univariate and multivariate analyses. The length of overall survival was defined as the interval between the date of surgical resection and the date of either death or the last follow-up. The period of freedom from recurrence (local recurrence or distant metastasis) was defined as the interval between the date of surgical resection and the date of the first recurrence or the last Table 2 The Relationship Between Characteristics of Disease Failure and Extent of Visceral Pleural Invasion in 107 Patients Who Developed Recurrence Extent of Visceral Pleural Invasion Variable Total (N 5 107) PL1 (n 5 82) PL2 (n 5 25) P Value Disease-free interval, mo 21.6 17.3 21.7 17.6 21.0 16.8.869 Patterns of recurrence.013 Local recurrence only 20 (18.7) 18 (22.0) 2 (8.0) Distant metastasis only 59 (55.1) 48 (58.5) 11 (44.0) Local recurrence and distant metastasis 28 (26.2) 16 (19.5) 12 (48.0) Malignant pleural effusion.006 Absent 94 (87.9) 76 (92.7) 18 (72.0) Present 13 (12.1) 6 (7.3) 7 (28.0) Data are presented as mean SD or No. (%). See Table 1 legend for expansion of abbreviations. journal.publications.chestnet.org CHEST / 142 / 1 / JULY 2012 143

Table 3 Univariate and Multivariate Analyses for Overall Survival in 355 Patients With Resected Node-Negative Non-small Cell Lung Cancer With Visceral Pleural Invasion Univariate Analysis Multivariate Analysis Variable HR 95% CI P Value HR 95% CI P Value Age at operation, y a 1.025 1.006-1.043.008 1.012 0.971-1.054.573 Sex Female 1 1 Male 2.024 1.357-3.021.001 1.560 0.667-3.650.305 Smoking index, pack-y b 1.006 1.000-1.012.041 0.996 0.983-1.009.534 Tumor size 2 cm 1 1. 2 to 3 cm 1.399 0.797-2.458.243 4.316 1.230-15.140.022. 3 to 5 cm 1.945 1.154-3.278.012 4.398 1.270-15.232.019. 5 to 7 cm 3.627 1.896-6.936,.001 15.832 3.156-79.428.001. 7 cm 3.326 1.492-7.415.003 9.095 0.817-101.190.072 Histologic type Squamous cell carcinoma 1 1 Adenocarcinoma 0.633 0.442-0.907.013 1.019 0.435-2.385.965 Others 0.892 0.400-1.986.779 Extent of pulmonary resection Sublobar resection 1 1 Lobectomy 0.588 0.356-0.971 0.038 0.532 0.210-1.348.183 Bilobectomy or pneumonectomy 0.991 0.531-1.849 0.997 0.790 0.246-2.539.692 Extent of visceral pleural invasion PL1 1 1 PL2 1.529 1.012-2.308.044 2.094 1.013-4.329.046 Number of mediastinal LNs dissected c 0.989 0.972-1.006.208 Angiolymphatic invasion Absent 1 1 Present 1.685 1.053-2.697.030 1.513 0.727-3.148.268 Histologic grade Well differentiated 1 1 Moderately differentiated 1.492 0.644-3.456.350 3.509 0.784-15.175.101 Poorly differentiated 3.256 1.348-7.866.009 5.212 1.105-24.579.037 HR 5 hazard ratio. See Table 1 legend for expansion of other abbreviations. a The HR associated with age is that the increase in hazard is associated with a 1-y increase in age. b The HR associated with smoking index is an increased hazard per 1 pack-y of additional smoking. c The HR associated with number of mediastinal LNs dissected is an increased hazard per one LN of additional dissection. follow-up. An observation was censored at the last follow-up session when the patient was alive with recurrence-free status or died without recurrence. Secondary primary lung cancer was differentiated from recurrent NSCLC according to the criteria proposed by Martini et al 14 and Martini and Melamed. 24 Local recurrence was defined as tumor recurrence in a contiguous anatomic site, including the ipsilateral hemithorax and mediastinum, after surgical resection. Distant metastasis was defined as tumor recurrence in the contralateral lung or outside the hemithorax and mediastinum after surgical resection. The overall survival and probability of freedom from recurrence were calculated by the Kaplan-Meier method. 25 The x 2 test and independent sample t test were used to compare groups with respect to categorical and continuous variables, as appropriate. Univariate and multivariate analyses were performed by means of the Cox proportional hazards model using SPSS, version 16.0 (SPSS Inc) statistical software. Only variables with P,.1 after the univariate analysis were entered into the multivariate analysis. Statistical significance was defined as P,.05. Results The median follow-up time for all 355 patients with surgically resected node-negative NSCLC was 54.2 months (mean, 54.0 29.3 months). The characteristics of these 355 patients are listed in Table 1. Eight (2.3%) postoperative deaths occurred. The 5- and 10-year overall survival rates were 61.9% and 33.5%, respectively ( Fig 1A ). The 5- and 10-year probabilities of freedom from recurrence were 66.2% and 62.9%, respectively ( Fig 1B ). The extent of VPI was PL1 in 300 patients (84.5%) and PL2 in 55 patients (15.5%). The relationship between extent of VPI and clinicopathologic variables are also shown in Table 1. The PL1 group had more tumors measur ing 2 cm and. 5 cm, whereas the PL2 group had more tumors measuring. 2 to 3 cm ( P 5.004). No relationship between other clinicopathologic characteristics and extent of VPI was detected. Patterns of Recurrence Among the 355 patients, 228 (64.2%) were free of tumor recurrence, 107 (30.1%) developed recurrence, and 20 (5.7%) were with unknown recurrence 144 Original Research

metastasis in 28 (26.2%). The percentages of local recurrence only in the PL1 and PL2 groups were 22.0% and 8.0%, respectively. The percentages of both local recurrence and distant metastasis in the PL1 and PL2 groups were 19.5% and 48.0%, respectively. The PL1 group comprised more patients with local recurrence only, whereas the PL2 group comprised more patients with both local recurrence and distant metastasis ( P 5.013). Among the 107 patients with recurrence, 13 (12.1%) developed malignant pleural effusion. The percentage of malignant pleural effusion in the PL2 group (28.0%) was significantly higher than that in PL1 group (7.3%; P 5.006). Overall Survival Univariate analysis for overall survival indicated that older age ( P 5.008), male sex ( P 5.001), increased smoking index ( P 5.041), larger tumor size, squamous cell carcinoma ( P 5.013), sublobar resection ( P 5.038), presence of angiolymphatic invasion ( P 5.030), and a histologic grade of poor differentiation ( P 5.009) were predictors of worse overall survival ( Table 3 ). Patients with PL2 had significantly worse overall survival than those with PL1 ( P 5.044) ( Fig 2A ). Larger tumor size, PL2 (hazard ratio [HR], 2.094; 95% CI, 1.013-4.329; P 5.046), and a histologic grade of poor differentiation (HR, 5.212; 95% CI, 1.105-24.579; P 5.037) were still significant predictors of worse overall survival in multivariate analysis ( Table 3 ). Figure 2. A, Cumulative probability of overall survival in patients with node-negative non-small cell lung cancer with visceral pleural invasion stratified by PL1 and PL2. B, Cumulative probability of freedom from recurrence in patients with node-negative nonsmall cell lung cancer with visceral pleural invasion stratified by PL1 and PL2 (log-rank test). PL1 5 visceral pleural invasion beyond the elastic layer; PL2 5 visceral pleural invasion to the visceral pleural surface. status after surgical resection. The overall recurrence rates for PL1 and PL2 were 29.0% (82 of 283 patients) and 48.1% (25 of 52 patients), respectively ( Table 1 ). The overall recurrence rate in patients with PL2 was significantly higher than in those with PL1 ( P 5.007). The relationship between characteristics of disease failure and extent of VPI are shown in Table 2. The median time to initial recurrence for these 107 patients was 17.5 months (mean, 21.6 17.3 months). There was no significant difference in the disease-free interval between the PL1 and PL2 groups ( P 5.869). The patterns of recurrence included local recurrence only in 20 patients (18.7%), distant metastasis only in 59 (55.1%), and both local recurrence and distant Probability of Freedom From Recurrence Univariate analysis for probability of freedom from recurrence indicated that larger tumor size and histologic grades of moderate ( P 5.049) and poor differentiation ( P 5.005) were predictors of significantly lower probability of freedom from recurrence ( Table 4 ). Patients with PL2 had significantly lower probability of freedom from recurrence than those with PL1 ( P 5.003) ( Fig 2B ). Larger tumor size, PL2 (HR, 1.810; 95% CI, 1.065-3.078; P 5.028) and a histologic grade of poor differentiation (HR, 7.044; 95% CI, 1.628-30.472; P 5.009) were still significant predictors for lower probability of freedom from recurrence in multivariate analysis ( Table 4 ). Probability of Freedom From Local Recurrence and Distant Metastasis We further performed univariate and multivariate analyses for probability of freedom from local recurrence and distant metastasis, respectively ( Table 5 ). Tumor size, sublobar resection (vs lobectomy, P,.001; vs bilobectomy or pneumonectomy, P 5.016), and PL2 ( P 5.048) were significant predictors for lower probability of freedom from local recurrence in multivariate analysis. Tumor size, PL2 ( P 5.004), and a histologic journal.publications.chestnet.org CHEST / 142 / 1 / JULY 2012 145

Table 4 Univariate and Multivariate Analyses for Probability of Freedom From Recurrence in 355 Patients With Resected Node-Negative Non-small Cell Lung Cancer With Visceral Pleural Invasion Univariate Analysis Multivariate Analysis Variable HR 95% CI P Value HR 95% CI P Value Age at operation, y a 1.013 0.993-1.033.209 Sex Female 1 Male 1.212 0.801-1.835.362 Smoking index, pack-y b 1.003 0.996-1.011.386 Tumor size 2 cm 1 1. 2 to 3 cm 2.232 1.128-4.417.021 2.217 0.951-4.761.066. 3 to 5 cm 2.728 1.418-5.248.003 2.937 1.349-6.394.007. 5 to 7 cm 4.078 1.767-9.413.001 4.632 1.319-16.270.017. 7 cm 2.077 0.579-7.449.262 1.620 0.193-13.628.657 Histologic type Squamous cell carcinoma 1 Adenocarcinoma 1.062 0.663-1.701.802 Others 0.881 0.304-2.558.816 Extent of pulmonary resection Sublobar resection 1 1 Lobectomy 0.605 0.337-1.087.093 0.563 0.291-1.091.089 Bilobectomy or pneumonectomy 0.607 0.266-1.385.235 0.794 0.303-2.083.640 Extent of visceral pleural invasion PL1 1 1 PL2 1.952 1.247-3.056.003 1.810 1.065-3.078.028 Number of mediastinal LNs dissected c 1.001 0.981-1.022.891 Angiolymphatic invasion Absent 1 Present 1.239 0.741-2.073.414 Histologic grade Well differentiated 1 1 Moderately differentiated 4.121 1.004-16.912.049 4.109 0.992-17.026.051 Poorly differentiated 8.010 1.873-34.253.005 7.044 1.628-30.472.009 See Table 1 and 3 legends for expansion of abbreviations. a The HR associated with age is that the increase in hazard is associated with a 1-y increase in age. b The HR associated with smoking index is an increased hazard per 1 pack-y of additional smoking. c The HR associated with number of mediastinal LNs dissected is an increased hazard per one LN of additional dissection. grade of poor differentiation ( P 5.023) were significant predictors for lower probability of freedom from distant metastasis in multivariate analysis. Comparison of Survival and Probability of Freedom From Recurrence in T2 Status Stratified by T2a ( 5 cm and PL1), T2a ( 5 cm and PL2), and T2b (. 5 to 7 cm) To compare survival difference among T2a with PL1, T2a with PL2, and T2b, we divided patients with T2 status (n 5 342) into three groups: T2a ( 5 cm and PL1), n 5 263; T2a ( 5 cm and PL2), n 5 52; and T2b (. 5 to 7 cm), n 5 27. Patients with T2a ( 5 cm and PL2) had significantly worse overall survival than those with T2a ( 5 cm and PL1) ( P 5.044) ( Fig 3A ). However, there was no significant difference in overall survival between patients with T2a ( 5 cm and PL2) and those with T2b (. 5 to 7 cm) ( P 5.157) ( Fig 3A ). The results in probability of freedom from recurrence were similar ( Fig 3B ). Discussion This study investigated the prognostic role of conventional clinicopathologic factors and patterns of recurrence in patients with resected node-negative NSCLC with VPI. Larger tumor size, PL2, and a histologic grade of poor differentiation were significant predictors for both worse overall survival and lower probability of freedom from recurrence in multivariate analysis. The PL2 group comprised significantly more patients who developed malignant pleural effusion. Tumor size is a significant prognostic factor for stage I NSCLC. 4,10-14,26-29 In the seventh edition of the TNM classification for lung cancer, T1 descrip tor was classified into T1a ( 2 cm) and T1b (. 2 to 3 cm). 18,19,30 Size-based T2 descriptor was clas sified into T2a (. 3 to 5 cm), T2b (. 5 to 7 cm), and T3 (. 7 cm). 18,19,30 Stage I NSCLC was subdivided into IA (T1aN0M0, T1bN0M0) and IB (T2aN0M0). 18,19,31 T2bN0M0 was reclassified as stage IIA. 18,19,31 In the present study, 146 Original Research

Table 5 Univariate and Multivariate Analyses for Probability of Freedom From Local Recurrence or Distant Metastasis in 355 Patients With Resected Node-Negative Non-small Cell Lung Cancer With Visceral Pleural Invasion Local Recurrence Distant Metastasis Univariate Analysis Multivariate Analysis Univariate Analysis Multivariate Analysis Variable HR 95% CI P Value HR 95% CI P Value HR 95% CIP Value HR 95% CIP Value Age at operation, y a 1.009 0.981-1.038.530 1.011 0.989-1.032.332 Sex Female 1 1 Male 1.590 0.827-3.058.165 1.117 0.712-1.751.630 Smoking index, pack-y b 1.011 1.002-1.021.021 1.008 0.997-1.019.172 0.998 0.988-1.007.613 Tumor size 2 cm 1 1 1 1. 2 to 3 cm 2.898 1.082-7.763.034 3.374 1.036-10.988.044 2.187 1.025-4.668.043 1.843 0.773-4.397.168. 3 to 5 cm 2.243 0.832-6.046.110 3.235 1.002-10.449.050 2.713 1.316-5.594.007 2.408 1.049-5.528.038. 5 to 7 cm 3.479 0.933-12.966.063 3.585 0.752-17.097.109 4.622 1.877-11.384.001 6.148 1.780-21.235.004. 7 cm 3.203 0.621-16.517.164 3.147 0.323-30.656.323 1.739 0.376-8.052.479 2.102 0.256-17.267.489 Histologic type Squamous cell carcinoma 1 1 Adenocarcinoma 0.690 0.362-1.315.259 1.254 0.723-2.212.410 Others 1.123 0.320-3.943.856 1.291 0.428-3.891.650 Extent of pulmonary resection Sublobar resection 1 1 1 Lobectomy 0.361 0.172-0.754.007 0.168 0.063-0.451,.001 0.869 0.418-1.810.708 Bilobectomy or pneumonectomy 0.455 0.152-1.361.159 0.189 0.049-0.733.016 0.960 0.370-2.493.934 Extent of visceral pleural invasion PL1 1 1 1 1 PL2 2.716 1.457-5.063.002 2.310 1.008-5.292.048 2.360 1.465-3.802,.001 2.258 1.292-3.945.004 Number of mediastinal LNs dissected c 0.985 0.953-1.017.349 1.001 0.979-1.025.909 Angiolymphatic invasion Absent 1 1 Present 1.382 0.661-2.892.390 1.240 0.709-2.167.451 Histologic grade Well differentiated 1 1 Moderately differentiated N/A d 3.339 0.809-13.777.095 3.513 0.845-14.612.084 Poorly differentiated 5.824 1.334-25.423.019 5.573 1.262-24.602.023 N/A 5 not applicable. See Table 1 and 3 legends for expansion of other abbreviations. a The HR associated with age is that the increase in hazard is associated with a 1-y increase in age. b The HR associated with smoking index is an increased hazard per 1 pack-y of additional smoking. c The HR associated with number of mediastinal LNs dissected is an increased hazard per one LN of additional dissection. d The analysis was not performed because no patient who developed local recurrence had a histologic grade of well differentiation. journal.publications.chestnet.org CHEST / 142 / 1 / JULY 2012 147

Figure 3. A, Cumulative probability of overall survival in patients with node-negative non-small cell lung cancer with visceral pleural invasion stratified by T2a ( 5 cm and PL1), T2a ( 5 cm and PL2), and T2b (. 5 to 7 cm). Patients with T2b (. 5 to 7 cm) and T2a ( 5 cm and PL2) had significantly worse overall survival than those with T2a ( 5 cm and PL1) ( P,.001 and P 5.044, respectively). There was no significant difference in overall survival between patients with T2b (. 5 to 7 cm) and those with T2a ( 5 cm and PL2) ( P 5.157). B, Cumulative probability of freedom from recurrence in patients with node-negative non-small cell lung cancer with visceral pleural invasion stratified by T2a ( 5 cm and PL1), T2a ( 5 cm and PL2), and T2b (. 5 to 7 cm). Patients with T2b (. 5 to 7 cm) and T2a ( 5 cm and PL2) had significantly lower probability of freedom from recurrence than those with T2a ( 5 cm and PL1) ( P 5.017 and P 5.013, respectively). There was no significant difference in probability of freedom from recurrence between patients with T2b (. 5 to 7 cm) and those with T2a ( 5 cm and PL2) ( P 5.650) (log-rank test). See Figure 2 legend for expansion of abbreviations. tumor size categorized according to the seventh edition of the TNM classification for lung cancer was a significant prognostic factor for both overall survival and probability of freedom from recurrence. VPI was shown to correlate with a higher frequency of mediastinal lymph node involvement and, thus, a poor survival. 5,6,9 However, the impact of VPI in nodenegative NSCLC is less clear. Several reports demonstrated that VPI is a poor prognostic factor in stage I NSCLC.4-6,9 Manac h et al 5 reported that the 5- and 10-year overall survival rates were 45.6% and 37%, respectively, in resected N0 NSCLC with VPI. In the present study, the 5- and 10-year overall survival rates were 61.9% and 33.5%, respectively. The 5- and 10-year probabilities of freedom from recurrence were 66.2% and 62.9%, respectively. Only a few studies demonstrated the patterns of recurrence in patients with resected NSCLC with VPI. 5,10 In the study by Jones et al, 10 who analyzed prognostic factors in stage IB NSCLC, VPI was identified in two of six patients (33%) with locoregional recurrence and eight of 22 patients (36%) with distant recurrence. In the study by Manac h et al, 5 patterns of recurrence in resected N0 NSCLC with VPI included local recurrence in 5.1% of patients, single organ metastasis in 48.7%, multiple organ metastases in 43.6%, and both local and distant metastasis in 2.6%. Among 96 patients with VPI (including those with N0, N1, N2) who developed a recurrence, only 6 (6.3%) developed malignant pleural effusion. 5 I n the present study, patterns of recurrence in resected node-negative NSCLC with VPI included local recurrence only in 18.7% of patients, distant metastasis only in 55.1%, and both local recurrence and distant metastasis in 26.2%. Thirteen of 107 patients (12.1%) with recurrence developed malignant pleural effusion. The percentage of malignant pleural effusion was significantly higher in the PL2 group. The prognostic effect of the extent of VPI (PL1 vs PL2) has not been well demonstrated. Only two stud ies analyzed data comparing PL1 vs PL2. Osaki et al 7 and Shimizu et al 8 both showed that there was no difference in survival between patients with PL1 and patients with PL2. However, there were only 78 patients in the study by Osaki et al 7 and 207 in Shimizu et al. 8 In the current study, 355 patients with resected node-negative NSCLC with VPI were included. Patients with PL2 had significantly worse overall survival and lower probability of freedom from recurrence than those with PL1. We further showed that patients with PL2 had significantly lower probabilities of freedom from both local recurrence and distant metastasis than those with PL1. According to the proposal by the IASLC, VPI is defined as PL1 and PL2. 20 If a tumor achieves T2 status based on VPI, and it is 5 cm in size, it is classified as T2a. If such a tumor is. 5 to 7 cm in size, it is T2b. The present study further shows that overall survival and probability of freedom from recurrence in patients with T2b (. 5 to 7 cm) and those with T2a ( 5 cm and PL2) were similar. There are some limitations of the present study that should be mentioned. As a retrospective study, patient selection bias and data lacking in some patients for some variables are inevitable. Prospective multi in stitutional studies are mandatory to validate further the prognostic value of PL2 and the patterns of recurrence in patients with PL2 (vs PL1). In conclusion, patients with PL2 had significantly worse overall survival and lower probability of freedom 148 Original Research

from recurrence than those with PL1 after resection of node-negative NSCLC. There was no significant difference in overall survival and prob ability of freedom from recurrence between patients with T2b (. 5 to 7 cm) and those with T2a ( 5 cm and PL2). Upgrading T status in tumors of PL2 and 5 cm in size from T2a to T2b should be studied thoroughly in the revision of the next edition of TNM classification for lung cancer. This information is important for further design of clinical trials for aggressive adjuvant chemotherapy and radiation therapy. Acknowledgments Author contributions: Drs Hung and Wu had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. Dr Hung: contributed to the study design; data management; data analysis; and writing, revision, and approval of the manu script. Dr Jeng: contributed to the study design, data management, data analysis, and approval of the manuscript. Dr Hsu: contributed to the study design, data management, data analysis, and approval of the manuscript. Dr Chou: contributed to the study design, data management, data analysis, and approval of the manuscript. Dr Lin: contributed to study design, data analysis and approval of the manuscript, and participated in pathologic slides review for determination of the extent of visceral pleural invasion. Dr Wu: contributed to the study design, data management, data analysis, and revision and approval of the manuscript. Financial/nonfinancial disclosures: The 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. Role of sponsors: The sponsors had no role in the design of the study, the collection and analysis of the data, or in the preparation of the manuscript. Other contributions: We are grateful to Liang-Shun Wang, MD, of Shuang Ho Hospital for contributing to this article. We also thank Jung-Hsing Lin, MS, for his assistance in data collection. References 1. Spira A, Ettinger DS. Multidisciplinary management of lung cancer. N Engl J Med. 2004 ;350(4):379-392. 2. Scott WJ, Howington J, Feigenberg S, Movsas B, Pisters K ; American College of Chest Physicians. Treatment of nonsmall cell lung cancer stage I and stage II: ACCP evidencebased clinical practice guidelines (2nd edition). Chest. 2007 ; 132 (suppl 3 ):234S-242S. 3. Mountain CF. Revisions in the international system for staging lung cancer. Chest. 1997 ;111(6):1710-1717. 4. Harpole DH Jr, Herndon JE II, Young WG Jr, et al. Stage I non-small cell lung cancer. Cancer. 1995 ;76(5):787-796. 5. 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. 6. Kang JH, Kim KD, Chung KY. Prognostic value of visceral pleura invasion in non-small cell lung cancer. Eur J Cardiothorac Surg. 2003 ;23(6):865-869. 7. Osaki T, Nagashima A, Yoshimatsu T, Yamada S, Yasumoto K. Visceral pleural involvement in nonsmall cell lung cancer: prognostic significance. Ann Thorac Surg. 2004 ;77(5):1769-1773. 8. 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. 9. 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. 10. Jones DR, Daniel TM, Denlinger CE, Rundall BK, Smolkin ME, Wick MR. Stage IB nonsmall cell lung cancers: are they all the same? Ann Thorac Surg. 2006 ; 81 ( 6 ): 1958-1962. 11. Carbone E, Asamura H, Takei H, et al. T2 tumors larger than five centimeters in diameter can be upgraded to T3 in non-small cell lung cancer. J Thorac Cardiovasc Surg. 2001 ; 122 (5 ):907-912. 12. Takeda S, Fukai S, Komatsu H, Nemoto E, Nakamura K, Murakami M ; Japanese National Chest Hospital Study Group. Impact of large tumor size on survival after resection of pathologically node negative (pn0) non-small cell lung cancer. Ann Thorac Surg. 2005 ;79 (4 ):1142-1146. 13. López-Encuentra A, Gómez de la Cámara A, Rami-Porta R, Duque-Medina JL, de Nicolás JL, Sayas J ; Bronchogenic Carcinoma Cooperative Group of the Spanish Society of Pneumology and Thoracic Surgery (GCCB-S). Previous tumour as a prognostic factor in stage I non-small cell lung cancer. Thorax. 2007 ;62 (5 ):386-390. 14. Martini N, Bains MS, Burt ME, et al. Incidence of local recurrence and second primary tumors in resected stage I lung cancer. J Thorac Cardiovasc Surg. 1995 ;109 (1 ):120-129. 15. Padilla J, Calvo V, Peñalver JC, Sales G, Morcillo A. Surgical results and prognostic factors in early non-small cell lung cancer. Ann Thorac Surg. 1997 ;63 (2 ):324-326. 16. Hammar SP. Common tumors. In: Dail DH, Hammar SP, eds. Pulmonary Pathology. 2nd ed. New York, NY : Springer- Verlag ; 1994 :1123-1278. 17. Hammar SP. Common tumors. In: Dail DH, Hammar SP, eds. Pulmonary Pathology. 1st ed. New York, NY : Springer- Verlag ; 1988 :727-845. 18. American Joint Committee on Cancer. AJCC Cancer Staging Manual. 7th ed. New York, NY : Springer ; 2010. 19. International Union Against Cancer. TNM Classification of Malignant Tumours. 7th ed. Oxford, England : Wiley-Blackwell ; 2009. 20. Travis WD, Brambilla E, Rami-Porta R, et al ; International Staging Committee. 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. 21. Hung JJ, Wang CY, Huang MH, Huang BS, Hsu WH, Wu YC. Prognostic factors in resected stage I non-small cell lung cancer with a diameter of 3 cm or less: visceral pleural invasion did not influence overall and disease-free survival. J Thorac Cardiovasc Surg. 2007 ;134 (3 ):638-643. 22. Hung JJ, Jeng WJ, Hsu WH, et al. Prognostic factors in pathological stage IB nonsmall cell lung cancer greater than 3 cm. Eur Respir J. 2010 ;36 (6 ):1355-1361. 23. Travis WD, Colby TB, Corrin B, et al. Histological typing of tumors of lung and pleura. In: Sobin LH, ed. World Health Organization International Classification of Tumours. 3rd ed. New York, NY : Springer-Verlag ; 1999 :21-47. 24. Martini N, Melamed MR. Multiple primary lung cancers. J Thorac Cardiovasc Surg. 1975 ;70 (4 ):606-612. 25. Kaplan EL, Meier P. Nonparametric estimation for incomplete observations. J Am Stat Assoc. 1958 ;53 (282 ):457-481. 26. Padilla J, Calvo V, Peñalver JC, et al. Survival and risk model for stage IB non-small cell lung cancer. Lung Cancer. 2002 ; 36 (1 ):43-48. 27. Sculier JP, Chansky K, Crowley JJ, Van Meerbeeck J, Goldstraw P ; International Staging Committee and Participating Institutions. The impact of additional prognostic factors on survival and their relationship with the anatomical journal.publications.chestnet.org CHEST / 142 / 1 / JULY 2012 149

extent of disease expressed by the 6th edition of the TNM Classification of Malignant Tumors and the proposals for the 7th edition. J Thorac Oncol. 2008 ;3 (5 ):457-466. 28. Birim O, Kappetein AP, Takkenberg JJ, van Klaveren RJ, Bogers AJ. Survival after pathological stage IA nonsmall cell lung cancer: tumor size matters. Ann Thorac Surg. 2005 ; 79 (4 ):1137-1141. 29. Port JL, Kent MS, Korst RJ, Libby D, Pasmantier M, Altorki NK. Tumor size predicts survival within stage IA non-small cell lung cancer. Chest. 2003 ;124 (5 ):1828-1833. 30. Rami-Porta R, Ball D, Crowley J, et al ; International Staging Committee ; Cancer Research and Biostatistics ; Observers to the Committee ; Participating Institutions. The IASLC Lung Cancer Staging Project: proposals for the revision of the T descriptors in the forthcoming (seventh) edition of the TNM classification for lung cancer. J Thorac Oncol. 2007 ; 2 (7 ):593-602. 31. Goldstraw P, Crowley J, Chansky K, et al ; International Association for the Study of Lung Cancer International Staging Committee ; Participating Institutions. The IASLC Lung Cancer Staging Project: proposals for the revision of the TNM stage groupings in the forthcoming (seventh) edition of the TNM Classification of malignant tumours. J Thorac Oncol. 2007 ;2 (8 ):706-714. 150 Original Research