[ Original Research Lung Cancer ]

Transcription

1 [ Original Research Lung Cancer ] The New Histologic Classification of Lung Primary Adenocarcinoma Subtypes Is a Reliable Prognostic Marker and Identifies Tumors With Different Mutation Status The Experience of a French Cohort Audrey Mansuet-Lupo, MD ; Antonio Bobbio, MD, PhD ; Hélène Blons, PharmD, PhD ; Etienne Becht ; Hanane Ouakrim ; Audrey Didelot ; Marie-Christine Charpentier, MD ; Serge Bain ; Béatrice Marmey ; Patricia Bonjour ; Jérôme Biton, PhD ; Isabelle Cremer, PhD ; Marie-Caroline Dieu-Nosjean, PhD ; Catherine Sautès-Fridman, PhD ; Jean-François Régnard, MD ; Pierre Laurent-Puig, MD, PhD ; Marco Alifano, MD, PhD, FCCP ; and Diane Damotte, MD, PhD BACKGROUND: Histologic classification of lung adenocarcinoma subtype has a prognostic value in most studies. However, lung adenocarcinoma characteristics differ across countries. Here, we aimed at validating the prognostic value of this classification in a large French series of lung adenocarcinoma. METHODS: We reviewed 407 consecutive lung adenocarcinomas operated on between 2001 and 2005 and reclassified them according to the International Association for the Study of Lung Cancer (IASLC)/American Thoracic Society (ATS)/European Respiratory Society (ERS) classification and subsequently graded them into low, intermediate, and high grade. We analyzed the relevance of this classification according to clinical, pathologic, and molecular analysis. RESULTS: Patients (median age, 61 years; 288 men) underwent lobectomy (n 5 378) or pneumonectomy (n 5 29). Patients overall survival at 5 and 10 years was 53.2% and 32.6%, respectively. Union for International Cancer Control stage distribution was 189 stage I, 104 stage II, 107 stage III, and seven stage IV. Low-grade tumor was found in one patient, intermediate grade in 275 patients, and high grade in 131 patients. KRAS and EGFR mutations were detected in 34% and 9.6%, respectively. Histologic grade was significantly correlated with extent of resection ( P 5.01), thyroid transcriptional factor-1 expression ( P ), vascular emboli ( P 5.03), and EGFR mutations ( P 5.01). Mucinous adenocarcinomas were associated with KRAS mutations ( P 5.003). At univariate analysis, age, extent of resection, histologic grade, pleural invasion, vascular emboli, pathologic T and N, and stage were predictive of survival. At multivariate analysis, age ( P ), histologic grade ( P 5.03), and stage ( P ) were independent prognostic factors. CONCLUSIONS: IASLC/ATS/ERS classification of lung adenocarcinomas predicts survival in French population. Histologic grade correlates with clinical, pathologic and molecular parameters suggesting different oncogenic pathways. CHEST 2014; 146(3): Manuscript received October 27, 2013; revision accepted March 3, 2014; originally published Online First March 27, ABBREVIATIONS: AIS 5 adenocarcinoma in situ; ATS 5 American Thoracic Society; ERS 5 European Respiratory Society; HPF 5 highpowered field; IASLC 5 International Association for the Study of Lung Cancer; MIA 5 minimal invasive adenocarcinoma; TTF 5 thyroid transcriptional factor; UICC 5 Union for International Cancer Control AFFILIATIONS: From INSERM (Drs Mansuet-Lupo, Biton, Cremer, Dieu- Nosjean, Sautès-Fridman, and Damotte, Mr Becht, and Mss Ouakrim, Marmey, and Bonjour) U1138, Team Cancer, Immune Control, and journal.publications.chestnet.org 633

2 Lung cancer is the leading cause of death by cancer in western countries, 1 adenocarcinomas being the most frequent histologic type. 2 According to the 2004 World Health Organization classification, most adenocarcinomas are classified as mixed, because 80% of them display at least two different architectural patterns. 3 In 2011, the International Association for the Study of Lung Cancer (IASLC), the American Thoracic Society (ATS), and the European Respiratory Society (ERS) jointly proposed a new and more satisfactory histologic classification for lung adenocarcinomas, based on the semiquantitative identification of the predominant histologic subtype. This classification describes two new entities: adenocarcinoma in situ (AIS) with pure lepidic growth and minimal invasive adenocarcinoma (MIA) for lepidic adenocarcinomas measuring, 3 cm with predominant lepidic growth and a small stromal invasion (, 0.5 cm).4 Moreover, invasive adenocarcinomas should be divided into lepidic- (formerly nonmucinous bronchiolar alveolar carcinoma), acinar-, papillary-, solid-, micropapillary-, and mucinous- (formerly mucinous bronchiolar alveolar carcinoma) predominant subtype. The main interest of this new classification was its prognostic impact as reported by several studies Yoshizawa et al 10 identified three groups of tumors with different survival in stage I cancer: low grade with AIS and MIA (100% 5-year disease-free survival); intermediate grade with nonmucinous lepidic predominant, papillary predominant, and acinar predominant (90%, 83%, and 84% 5-year disease-free survival, respectively); and high grade with invasive mucinous adenocarcinoma, solid and micropapillary predominant (75%, 70%, and 67% 5-year disease-free survival, respectively). Interestingly, histologic grade frequency is heterogeneous, especially for AIS and MIA tumors, as underlined by two reports, one concerning 500 white patients with neither AIS nor MIA subtypes 8 and another study on 440 Asian patients reporting 12% of AIS and MIA. 11 Several factors could be at the origin of these discrepancies, including ethnic origin, quality of radiologic detection, and extent of resection. The improved quality of CT scan imaging allows the detection of small ground-glass opacities, which correspond to minimally invasive or in situ adenocarcinoma and are mainly treated by sublobar resections (wedge resections or segmentectomy). 13,14 Ethnic origin of patients could be an additional explanation for grade 1 frequency discrepancy, as suggested by extensive analysis of KRAS and EGFR mutations in lung adenocarcinoma for targeted treatment administration, underlining a clear differences in EGFR mutation frequency between Asian and white series. 19 Altogether, these different studies suggest that lung adenocarcinoma histologic and molecular heterogeneity may reflect different tumoral pathways, impacting patients optimal management. The aim of our study was to assess the prognostic value of the IASLC/ATS/ ERS classification in our own population and to correlate the histologic grade to a broad spectrum of clinical, pathologic, and molecular parameters. Materials and Methods Patients We retrospectively reviewed all consecutive patients treated by primary surgery in curative intent for primitive lung carcinoma between June 2001 and June 2005 at the Thoracic Surgery Department of Hôtel-Dieu Hospital in Paris, France. We excluded all nonprimary lung cancers, patients who received neoadjuvant therapies (chemotherapy and/or radiotherapy), and those treated by sublobar resection. Patients characteristics, treatment procedures, and short-term outcomes were prospectively collected using a standardized case report form ( Table 1 ). Long-term outcome was retrospectively obtained by direct patient or family interview and by interrogation of municipality registers to ascertain possible date of death. At last, 407 patients with primary lung adenocarcinoma were enrolled. The research was conducted according to the recommendations outlined in the Helsinki declaration. Institutional review board approval was obtained (CPP Ile de France II, and ). Escape, Centre de Recherche des Cordeliers; the Université Pierre et Marie Curie-Paris 6 (Drs Mansuet-Lupo, Biton, Cremer, Dieu-Nosjean, Sautès-Fridman and Damotte, Mr Becht, and Mss Marmey and Bonjour); the Université Paris Descartes-Paris 5 (Drs Mansuet-Lupo, Blons, Biton, Cremer, Dieu-Nosjean, Sautès-Fridman, Régnard, Laurent- Puig, Alifano, and Damotte, Mr Becht, and Mss Marmey and Bonjour); the Université Denis Diderot-Paris 7 (Drs Mansuet-Lupo and Damotte and Mr Becht); Service de Pathologie (Drs Mansuet-Lupo and Damotte, Ms Charpentier, and Mr Bain), Hôpitaux Universitaire Paris Centre, AP-HP; the Service de Chirurgie Thoracique (Drs Bobbio, Régnard, and Alifano), Hôpitaux Universitaire Paris Centre, AP-HP; the Service de Biochimie (Drs Blons and Laurent-Puig), Hôpital Européen Georges Pompidou, AP-HP; and UMR-S775 (Drs Blons and Laurent-Puig and Ms Didelot), INSERM, Faculté de médecine des Saints Pères, Paris, France. Drs Mansuet-Lupo and Bobbio contributed equally to this work. FUNDING/SUPPORT: This work was supported by the Institut National de la Santé et de la Recherche Médicale (INSERM), Université Paris- Descartes, Université Pierre et Marie Curie, Institut National du Cancer and Cancéropole Ile de France [Grants PLBIO-06-INSERM 6-1, PLBIO-07-INSERM 6-1, PLBIO-03-INSERM 6-1, 11LAXE62_9UMRS872 FRIDMAN] and Fondation ARC pour la Recherche sur le Cancer [Grant SL ]. CORRESPONDENCE TO: Diane Damotte, MD, PhD, Service de Pathologie, Hôpitaux Universitaire Paris Centre, 1, Place du Parvis de Notre Dame, 75181, Paris, France; diane.damotte@htd.aphp.fr 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: /chest Original Research [ 146 # 3 CHEST SEPTEMBER 2014 ]

3 TABLE 1 ] Clinical and Pathologic Characteristics of Patients With Lung Adenocarcinoma, According to the Histologic Grade Characteristic Total (N 5 407) Intermediate Grade (n 5 275) High Grade (n 5 131) P Value Men 288 (71) 190 (69) 97 (74) ns Age, median (range) 61 (19-84) 61 (19-84) 62 (37-83) ns Tobacco pack-y, median (range) 39 (0-100) 38 (0-90) 41 (0-100) ns COPD, yes; no 156 (42); 214 (58) 95 (39); 151 (61) 60 (49); 63 (51) ns Lobectomy; pneumonectomy 378 (93); 29 (7) 262 (95); 13 (5) 115 (88); 16 (12).01 Histologic subtype (predominant pattern) na Lepidic 11 Acinar 191 Papillary 73 Solid 109 Invasive mucinous 16 Micropapillary 4 Signet ring cells 2 TTF1, yes; no 368 (90); 39 (10) 265 (96); 10 (4) 102 (78); 29 (22) Pleural invasion, yes; no 232 (57); 175 (43) 159 (58); 116 (42) 72 (55); 59 (45) ns Vascular emboli, yes; no 229 (59); 160 (41) 146 (55); 117 (45) 83 (66); 42 (34).03 Tumor stage.05 pt pt pt pt Nodal status pn pn pn UICC stage I II III IV ns ns Data are presented as No. or No. (%) unless otherwise noted. Since only one patient had low-grade tumor (minimally invasive adenocarcinoma), we have compared only intermediate- with high-grade group. na 5 not applicable; ns 5 not significant; TTF 5 thyroid transcriptional factor; UICC 5 Union for International Cancer Control. Histologic Evaluation For the 407 patients, all the slides containing tumoral areas (one to seven for each tumor) were independently reviewed by two expert pathologists (D. D. and A. M.-L.), who were blinded to the patient s clinical outcomes. The predominant pattern was defined as the morphologic subtype present in the greatest proportion. In case of discordant results (n 5 7), slides were jointly reviewed and discussed to consensual diag nosis, which was achieved in all these cases. All adenocarcinomas were graded according to their predominant pattern as: low grade (AIS, MIA); intermediate grade (lepidic, papillary, or acinar predominant); and high grade (solid, micropapillary, mucinous, or solid with signet ring cells predominant contingent). 10 To substratify acinar tumors, which represented the most frequent pattern, we evaluated nuclear atypia, mitotic count, and presence of cribriform area, these parameters representing potential aggressiveness of the tumor. 20,21 We applied the Kadota criteria for nuclear atypia and cribriform pattern, 20,21 whereas we adapted mitosis count as follows: low, 0 to 4 mitoses/10 high-powered fields (HPFs); intermediate, 5 to 10 mitoses/10 HPFs; and high,. 10 mitoses/10 HPFs. For all tumors, we assessed the expression of the thyroid transcriptional factor (TTF) 1 (SP24 clone [Novocastra]), the pleural invasion, and the presence of vascular emboli. Tumor stages were attributed in accordance with the seventh edition of the TNM classification of the Union for International Cancer Control (UICC). 22 journal.publications.chestnet.org 635

4 Molecular Analysis For each tumor, the block containing the largest surface at least 30% of tumoral cells was selected for molecular analysis. DNA was extracted from formalin-fixed, paraffin-embedded blocks using a kit (illustra DNA extraction kit BACC2; General Electric Company ), according to the manufacturer s instructions. The most common molecular alter ations (ie, KRAS and EGFR mutations) were analyzed using fragment analysis for insertions and deletions and Taqman probes for point mutations as previously described 23 : in EGFR (exon 19 deletions [DEL19], exon 20 insertions [INS20], and p.l858r), and in KRAS (p.g12a, p.g12c, p.g12r, p.g12v, p.g12s, p.g12d, and p.g13d). Statistical Analysis Categorical variables were compared by x 2 or Fisher exact tests as appropriate, whereas Student or Wilcoxon signed-rank tests were used for continuous variables, as appropriate. Survival rates were calculated using the Kaplan-Meier method, and the differences were analyzed by log-rank test. All the variables associated with survival with a P value,.05 in univariate analysis were entered in a multivariate analysis using Cox proportional hazards regression model, except for pathologic T and N stages, which are included through UICC stage. A P value,.05 was considered significant. All the statistics were computed using the R software ( and survival analyses were carried out with the R survival package. Results Patient and Tumor Characteristics Four hundred seven consecutive patients with primary lung adenocarcinoma underwent surgery (378 lobectomies and 29 pneumonectomies) with a curative intent between 2001 and There were 288 men and 119 women. Median age was 61 years (range, years). Smokers (. 15 pack-years) accounted for 81% of cases. One hundred fifty-six patients had stage I to III COPD according to the GOLD (Global Initiative for Obstructive Lung Disease) classification. The pathologic stages according to the 2009 UICC TNM classification were I, II, III, and IV in 189, 104, 107, and seven patients, respectively ( Table 1 ). According to the IASLC/ATS/ERS 2011 criteria, no AIS and only one MIA were found. Final histologic predominant pattern was acinar in 191 (47%), solid in 109 (27%), papillary in 73 (18%), mucinous in 16, lepidic in 11, micropapillary in four, and solid with signet ring cells in two. A strong positive concordance was found among the two pathologists: In particular, it was 98.4% (95% CI, 95.5%-99.7%; P,.001) for the acinar subtype, 97.3% (95% CI, 90.4%-99.7%; P,.001) for the papillary subtype, 98.2% (95% CI, 93.5%-99.8%; P,.001) for the solid subtype, and 100% for the remaining, more rare subtypes. Only one patient had low-grade adenocarcinoma, whereas 275 had intermediate- and 131 had high-grade adenocarcinoma. TTF1 antigen expression was found in 368 tumors (90%) ( Table 1 ). Histologic high-grade tumors expressed less frequently TTF1 antigen ( P ), had less frequently EGFR mutations ( P 5.01), displayed more frequently vascular emboli ( P 5.03), and were more frequently treated by pneumonectomy ( P 5.01) as compared with intermediate grade ( Tables 1, 2 ). Since acinar-predominant pattern was the most frequent histologic subtype, we investigated if nuclear atypia, mitotic count, and presence of cribriform pattern could substratify them. Nuclear atypia and mitosis count were significantly associated ( P ) and were both associated with pleural invasion ( P 5.04 and P 5.02, respectively) and lower frequency of EGFR mutations ( P and P 5.02, respectively). Mitosis count was also significantly associated with vascular emboli ( P 5.001), and cribriform pattern was significantly associated with nuclear atypia ( P ), mitosis ( P ), and vascular emboli ( P 5.005) ( e-table 1 ). Molecular Analysis of the Tumors and Correlation With the Clinical and Histologic Parameters KRAS and EGFR mutations were analyzed for 397 tumors; 10 out of the 407 tumors could not be amplified. KRAS and EGFR mutations occurred respectively in 34% (136 of 397) and 9.6% (38 of 397) of samples. The most frequent KRAS mutation was p.g12c (59 of 136) (43%). Among the 38 tumors with EGFR tyrosine kinase domain mutations, 17 were exon 19 deletions, 18 were p.l858r, and three were exon 20 insertions. EGFR mutations were correlated to the histologic grade, as these mutations occurred more frequently in intermediate-grade adenocarcinomas ( P 5.01) ( Table 2 ). KRAS mutations were not significantly associated with the histologic grade, but, interestingly, in 12 out of the 16 mucinous tumors we found a KRAS mutation, and, conversely, we never found EGFR mutation in mucinous adenocarcinoma ( Table 2 ). In the KRAS-mutated acinarpredominant adenocarcinoma, we did not find any specific histologic characteristic; for the EGFR-mutated acinar tumors, we found less nuclear atypia ( P 5.008) and a lower mitotic count ( P 5.02) ( e-table 1 ). Survival Analysis Global overall survival was 53.2% (95% CI, ) and 32.6% (95% CI, ) at 5 and 10 years, respectively. At univariate analysis, age ( P ), extent of resection ( P 5.003), pathologic T ( P ), pathologic N ( P ), pathologic 2009 UICC stage ( P ), pleural invasion ( P 5.008), and vascular emboli ( P ) were associated with survival. 636 Original Research [ 146 # 3 CHEST SEPTEMBER 2014 ]

5 TABLE 2 ] Clinical and Pathologic Characteristics According to KRAS and EGFR Status Characteristic EGFR WT KRAS WT (n 5 223) KRAS Mut (n 5 136) P Value EGFR Mut (n 5 38) P Value Men 163 (73) 108 (79) ns 10 (26) ns Age median (range) 60 (19-83) 61 (38-83) ns 68 (46-83) Tobacco, yes; no 184 (84); 34 (16) 124 (93); 10 (7) (22); 28 (78) Pack-y median (range) 42 (0-100) 42 (0-100) ns 6.9 (0-35) COPD, yes; no 86 (42); 119 (58) 58 (48); 63 (52) ns 9 (26); 26 (74) ns Lobectomy - pneumonectomy 204 (91); 19 (9) 127 (93); 9 (7) ns 37 (97); 1 (3) ns Histologic subtype (predominant pattern) Lepidic 6 (2.7) 3 (2) ns 2 (5) ns Acinar 104 (47) 60 (44) ns 21 (56) ns Papillary 38 (17) 24 (18) ns 10 (26) ns Solid 67 (30) 36 (26) ns 3 (8).003 Invasive mucinous 4 (1.8) 12 (9) ns Micropapillary 1 (0.5) 1 (1) ns 2 (5) ns Signet ring cells 2 (1) 0 ns 0 ns Histologic grade ns.01 Intermediate 148 (67) 87 (64) 33 (87) High 74 (33) 49 (36) 5 (13) TTF1, yes; no 201 (90); 22 (10) 121 (89); 15 (11) ns 37 (97); 1 (3) ns Pleural invasion, yes; no 122 (55); 101 (45) 82 (60); 54 (40) ns 23 (60); 15 (40) ns Vascular emboli, yes; no 126 (61); 82 (39) 79 (59); 54 (41) ns 18 (47); 20 (53) ns Tumor stage ns ns pt1 58 (26) 32 (24) 8 (21) pt2 95 (43) 71 (52) 22 (58) pt3 50 (22) 23 (17) 7 (18) pt4 20 (9) 10 (7) 1 (3) (Continued) journal.publications.chestnet.org 637

6 TABLE 2 ] (continued) Characteristic EGFR WT KRAS WT (n 5 223) KRAS Mut (n 5 136) P Value EGFR Mut (n 5 38) P Value Nodal status ns ns pn0 153 (68) 85 (63) 23 (60) pn1 35 (16) 28 (20) 3 (8) pn2 35 (16) 23 (17) 12 (32) UICC stage ns ns I 103 (46) 63 (46) 19 (50) II 59 (27) 35 (26) 7 (18) III 57 (25) 35 (26) 12 (32) IV 4 (2) 3 (2) 0 Data are presented as No. (%) unless otherwise noted. See Table 1 legend for expansion of abbreviations. EGFR or KRAS mutations were not significantly associated with survival ( P 5.6 and P 5.3, respectively) (Table 3 ). According to the IASLC/ATS/ERS classification, we found that patients with intermediate histologic grade tumors had a significantly better survival compared with high-grade tumors ( P 5.001), with 5-year survival rates of 58.6% and 46.1%, respectively, and 10-year survival rates of 42.4% and 21.4%, respectively ( Fig 1 ). Histologic grade remained a significant prognostic factor even in patients with stage I adenocarcinoma ( e-fig 1, e-tables 2, 3 ). Lepidic-, acinar-, and papillarypredominant patterns were associated with 55.6%, 58.2%, and 58.8% survival at 5 years, respectively, as compared with solid-, mucinous-, micropapillary-, and solid with signet ring cells-predominant patterns having 45.1%, 26.7%, 50%, and 0% survival rates at 5 years, respectively. We did not find any significant association between survival and nuclear atypia, mitosis count, or presence of cribriform areas, respectively. It is noticeable that the 5-year survival rate for patients with acinar adenocarcinomas containing. 10% cribriform areas (51.3%) was intermediate between the remaining acinarpredominant adenocarcinomas (58.2%) and the solidpredominant adenocarcinomas (45.1%). As expected, histologic grade combined with the 2009 UICC stage improved the survival prediction, as patients with intermediate grade/stage I disease had a better 5-year survival rate (71.9%) than high grade/stage III to IV (28.2%). At multivariate analysis (model including grade, stage of disease, age, extent of resection, pleural invasion, and vascular emboli), age, pathologic stage, and histologic grade were identified as the independent prognostic factors ( Table 4 ). Discussion The objective of our study was to assess the prognostic value of IASLC/ATS/ETS histologic classification of lung adenocarcinoma in a consecutive series of French patients. We correlated histologic and molecular characteristics of the tumors to clinical data, aiming at determining significant association between different parameters. The patients were operated on in a single surgical department from 2001 to 2005, and the tumors were reclassified according to the latest 2011 IASLC/ATS/ERS adenocarcinoma classification and the UICC 2009 staging system. We were able to demonstrate that (1) lung adenocarcinoma predominant histologic grade was an independent and reliable survival predictor; (2) EGFR and KRAS mutations were significantly 638 Original Research [ 146 # 3 CHEST SEPTEMBER 2014 ]

7 TABLE 3 ] Univariate Survival Analysis for Overall Survival Parameter 5-y Survival, % 10-y Survival, % Hazard Ratio P Value Age a Sex Men Women REF Pack-y COPD Yes No REF Extent of resection Pneumonectomy a Lobectomy REF Histologic GRADE Intermediate REF.001 a High Tumor stage T REF T T3-T Nodal status N REF N N UICC stage I REF II a III-IV Vascular emboli Yes a No REF Pleural invasion Yes a No REF TTF-1 expression Yes No REF KRAS mutation Yes No REF EGFR mutation Yes No REF Age and pack-y consumption were analyzed as continuous variables. REF 5 reference group in the Cox regression analysis. See Table 1 legend for expansion of other abbreviations. a Significant P value ( P,.05). journal.publications.chestnet.org 639

8 Figure 1 Kaplan-Meier survival curves (log-rank test) for histologic grade. The tumors were divided into two histologic grades according to International Association for the Study of Lung Cancer/American Thoracic Society/European Respiratory Society. Classification: intermediate-grade group (lepidic-, acinar-, and papillary-predominant pattern) and high-grade group (solid-, mucinous-, micropapillary-, and solid with signet ring cellspredominant pattern ). associated with histologic subtype, but they had no prognostic value; and (3) several clinical and histologic parameters had a prognostic value in univariate (age, extent of resection, pt, pn, ptnm UICC stage, histologic grade, pleural invasion, and vascular tumoral emboli) and multivariate (age, UICC stage, and histologic grade) analysis. Histologic predominant subtype distribution is highly variable in the literature, probably because of epidemiologic differences in the studied populations. 5-7,10-12,24-26 In our study, the acinar, solid, and papillary patterns (48%, 26%, and 18%, respectively) represented. 90% of all adenocarcinomas, whereas micropapillary- and solid with signet ring cells-predominant patterns were rare (4% and 2%, respectively), even if some foci of such architectural patterns could be found more frequently. Our findings are in accordance with Russell et al, 5 who found acinar, solid, and papillary types as the predominant histologic patterns in, respectively, 40%, 23%, and 12% of 210 adenocarcinomas. As already reported by others, in our population MIA was rare (n 5 1), and AIS was never found. 8 The performance of CT imaging in the period of the study ( ) was lower, 27 and probably some lesions, such as small ground-glass opacities, were not diagnosed. Furthermore, we excluded the sublobar resections and subsequently some MIA and AIS, which could have been preferentially treated by conservative surgery. The ethnic differences between Asian and white population could be also an explanation, as it was already reported for EGFR mutation distribution. Previous studies showed that lepidic-predominant pattern lung adenocarcinomas had a higher frequency of EGFR mutation,28-32 but they were carried out in Asian populations, who are enriched in lepidic-predominant pattern. In our study, we did not reach significance for correlation between the lepidic-predominant pattern and EGFR mutations, although these mutations were more frequent in the whole intermediate-grade group ( P 5.01), probably because we had a low frequency of lepidic-predominant pattern (2.7%). We found in our population, like others reports, a trend toward a significant association ( P 5.057) between micropapillary subtype and EGFR mutations. 30,32-34 Nevertheless, the prognostic value of histologic grade remains valid independently of patient s origin, which strengthens the clinical impact of the IASLC/ATS/ERS 640 Original Research [ 146 # 3 CHEST SEPTEMBER 2014 ]

9 TABLE 4 ] Multivariate Survival Analysis for Overall Survival Parameter Hazard Ratio P Value Age, y Extent of resection Pneumonectomy Lobectomy REF UICC stage I REF II III-IV Histologic grade Intermediate REF.03 a High 1.36 Vascular emboli Yes No REF Pleural invasion Yes No REF Age was analyzed as a continuous variable. See Table 1 and 3 legends for expansion of abbreviations. a Significant P value ( P,.05). classification. The precise histologic subtype within each group of histologic grade seems less important, as we found that patients with lepidic-predominant adenocarcinoma had similar survival rates compared with patients with papillary- and acinar-predominant adenocarcinoma, all belonging to the intermediate group. As already reported, patients with invasive mucinous adenocarcinoma had the worse prognosis (36.7% survival at 5 years), and, interestingly, their tumors were often KRAS mutated.7,11,35 The prognostic value of the histologic grade may be related to the tumor behavior according to histologic subtype. Indeed, high histologic grade adenocarcinomas had significantly higher frequency of vascular emboli ( P 5.03) and pneumonectomy ( P 5.01), reflecting a more aggressive behavior. Interestingly, in the acinarpredominant pattern histologic group, which has an intermediate prognosis, the mitotic count and the cribriform pattern determination allowed the identification of patients with poorer survival, as already reported by Kadota et al. 20,21 In conclusion, we confirmed the prognostic value of the histologic grade based on the IASLC/ATS/ERS classification of lung adenocarcinoma. The more accurate predictor of long-term survival was the combined analysis of the histologic grade and the UICC stage. Histologic grade was significantly associated with different clinical and molecular characteristics, suggesting different oncogenic pathways. The impact of this lung cancer subclassification on personalized therapy needs to be demonstrated in adequately designed prospective studies. journal.publications.chestnet.org 641

10 Acknowledgments Author contributions: D. D. is guarantor of the article. A. M.-L. and A. B. contributed to conception and design; A. M.-L., A. B., H. B., E. B., and M. A. contributed to data analysis and interpretation; D. D. contributed to data interpretation; A. M.-L. and A. B. contributed to the drafting of the manuscript; A. M.-L., A. B., H. O., A. D., S. B., B. M., P. B., J. B., and C. S.-F. contributed to the revision of the article; M.-C. C., I. C., M.-C. D.-N., J.-F. R., P. L.-P., and D. D. contributed to article redaction; H. B., E. B., and M. A. contributed to article preparation; H. O., A. D., M.-C. C., S. B., B. M., P. B., J. B., I. C., M.-C. D.-N., J.-F. R., and P. L.-P. contributed to data collection; and D. D. contributed to leading the work and designing the study. 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 the preparation of the manuscript. Other contributions: We thank The French National Cancer Institute (INCa) for its financial support (Plan cancer ), and the (Cancer Research for Personalized Medicine) CARPEM for its support. Additional information: The e-figure and e-tables can be found in the Supplemental Materials section of the online article. References 1. La Vecchia C, Bosetti C, Lucchini F, et al. Cancer mortality in Europe, , and an overview of trends since Ann Oncol ;21 (6 ): Curado MP, Edwards B, Shin HR, Storm H, Ferlay J, Heanue M, Boyle P, eds. Cancer Incidence in Five Continents, Volume IX. Lyon, France: International Agency for Research on Cancer ; Beasley MB, Brambilla E, Travis WD. The 2004 World Health Organization classification of lung tumors. Semin Roentgenol ;40(2): Travis WD, Brambilla E, Noguchi M, et al. International Association for the Study of Lung Cancer/American Thoracic Society/European Respiratory Society international multidisciplinary classification of lung adenocarcinoma. J Thorac Oncol ;6(2): Russell PA, Wainer Z, Wright GM, Daniels M, Conron M, Williams RA. Does lung adenocarcinoma subtype predict patient survival?: a clinicopathologic study based on the new International Association for the Study of Lung Cancer/ American Thoracic Society/European Respiratory Society international multidisciplinary lung adenocarcinoma classification. J Thorac Oncol ; 6 (9): Song Z, Zhu H, Guo Z, Wu W, Sun W, Zhang Y. Prognostic value of the IASLC/ATS/ERS classification in stage I lung adenocarcinoma patients based on a hospital study in China. Eur J Surg Oncol ;39(11): Tsuta K, Kawago M, Inoue E, et al. The utility of the proposed IASLC/ATS/ERS lung adenocarcinoma subtypes for disease prognosis and correlation of driver gene alterations. Lung Cancer ;81(3): Warth A, Muley T, Meister M, et al. The novel histologic International Association for the Study of Lung Cancer/American Thoracic Society/European Respiratory Society classification system of lung adenocarcinoma is a stage-independent predictor of survival. J Clin Oncol ;30(13): Xu L, Tavora F, Battafarano R, Burke A. Adenocarcinomas with prominent lepidic spread: retrospective review applying new classification of the American Thoracic Society. Am J Surg Pathol ;36(2): Yoshizawa A, Motoi N, Riely GJ, et al. Impact of proposed IASLC/ATS/ERS classification of lung adenocarcinoma: prognostic subgroups and implications for further revision of staging based on analysis of 514 stage I cases. Mod Pathol ;24(5): Yoshizawa A, Sumiyoshi S, Sonobe M, et al. Validation of the IASLC/ATS/ERS lung adenocarcinoma classification for prognosis and association with EGFR and KRAS gene mutations: analysis of 440 Japanese patients. J Thorac Oncol ;8(1): Zhang J, Wu J, Tan Q, Zhu L, Gao W. Why do pathological stage IA lung adenocarcinomas vary from prognosis?: a clinicopathologic study of 176 patients with pathological stage IA lung adenocarcinoma based on the IASLC/ATS/ERS classification. J Thorac Oncol ;8(9): Tsutani Y, Miyata Y, Nakayama H, et al. Appropriate sublobar resection choice for ground glass opacity-dominant clinical stage IA lung adenocarcinoma: wedge resection or segmentectomy. Chest ;145(1): Altorki NK, Yip R, Hanaoka T, et al ; I-ELCAP Investigators. Sublobar resection is equivalent to lobectomy for clinical stage 1A lung cancer in solid nodules. J Thorac Cardiovasc Surg ;147(2): Coate LE, John T, Tsao MS, Shepherd FA. Molecular predictive and prognostic markers in non-small-cell lung cancer. Lancet Oncol ;10(10): Lynch TJ, Bell DW, Sordella R, et al. Activating mutations in the epidermal growth factor receptor underlying responsiveness of non-small-cell lung cancer to gefitinib. N Engl J Med ;350(21): Paez JG, Jänne PA, Lee JC, et al. EGFR mutations in lung cancer: correlation with clinical response to gefitinib therapy. Science ;304(5676): Pao W, Miller V, Zakowski M, et al. EGF receptor gene mutations are common in lung cancers from never smokers and are associated with sensitivity of tumors to gefitinib and erlotinib. Proc Natl Acad Sci U S A ;101(36): Couraud S, Zalcman G, Milleron B, Morin F, Souquet PJ. Lung cancer in never smokers a review. Eur J Cancer ;48(9): Kadota K, Yeh YC, Sima CS, et al. The cribriform pattern identifies a subset of acinar predominant tumors with poor prognosis in patients with stage I lung adenocarcinoma: a conceptual proposal to classify cribriform predominant tumors as a distinct histologic subtype. Mod Pathol ;27(5): Kadota K, Suzuki K, Kachala SS, et al. A grading system combining architectural features and mitotic count predicts recurrence in stage I lung adenocarcinoma. Mod Pathol ; 25 ( 8 ): Sobin LH, Compton C. TNM Seventh Edition: what s new, what s changed: communication from the International Union Against Cancer and the American Joint Committee on Cancer. Cancer ; 116 (22): Didelot A, Le Corre D, Luscan A, et al. Competitive allele specific TaqMan PCR for KRAS, BRAF and EGFR mutation detection in clinical formalin fixed paraffin embedded samples. Exp Mol Pathol ;92(3): Gu J, Lu C, Guo J, et al. Prognostic significance of the IASLC/ATS/ERS classification in Chinese patients: a single institution retrospective study of 292 lung adenocarcinoma. J Surg Oncol ;107(5): Russell PA, Barnett SA, Walkiewicz M, et al. Correlation of mutation status and survival with predominant histologic subtype according to the new IASLC/ATS/ERS lung adenocarcinoma classification in stage III (N2) patients. J Thorac Oncol ;8(4): Woo T, Okudela K, Mitsui H, et al. Prognostic value of the IASLC/ATS/ERS classification of lung adenocarcinoma in stage I disease of Japanese cases. Pathol Int ;62(12): Aberle DR, Adams AM, Berg CD, et al ; National Lung Screening Trial Research Team. Reduced lung-cancer mortality with low-dose computed tomographic screening. N Engl J Med ;365(5): Li H, Pan Y, Li Y, et al. Frequency of well-identified oncogenic driver mutations in lung adenocarcinoma of smokers varies with histological subtypes and graduated smoking dose. Lung Cancer ;79(1): Shim HS, Lee H, Park EJ, Kim SH. Histopathologic characteristics of lung adenocarcinomas with epidermal growth factor receptor mutations in the International Association for the Study of Lung Cancer/American Thoracic Society/European Respiratory Society lung adenocarcinoma classification. Arch Pathol Lab Med ; 135 ( 10 ): Original Research [ 146 # 3 CHEST SEPTEMBER 2014 ]

11 30. Song Z, Zhu H, Guo Z, Wu W, Sun W, Zhang Y. Correlation of EGFR mutation and predominant histologic subtype according to the new lung adenocarcinoma classification in Chinese patients. Med Oncol ;30(3): Sun PL, Seol H, Lee HJ, et al. High incidence of EGFR mutations in Korean men smokers with no intratumoral heterogeneity of lung adenocarcinomas: correlation with histologic subtypes, EGFR/TTF-1 expressions, and clinical features. J Thorac Oncol ;7(2): Warth A, Penzel R, Lindenmaier H, et al. EGFR, KRAS, BRAF and ALK Gene alterations in lung adenocarcinomas: patient outcome, interplay with morphology and immunophenotype. Eur Respir J ; 43 ( 3 ): Motoi N, Szoke J, Riely GJ, et al. Lung adenocarcinoma: modification of the 2004 WHO mixed subtype to include the major histologic subtype suggests correlations between papillary and micropapillary adenocarcinoma subtypes, EGFR mutations and gene expression analysis. Am J Surg Pathol ; 32 ( 6 ): Ninomiya H, Hiramatsu M, Inamura K, et al. Correlation between morphology and EGFR mutations in lung adenocarcinomas: significance of the micropapillary pattern and the hobnail cell type. Lung Cancer ;63(2): Zhang Y, Sun Y, Pan Y, et al. Frequency of driver mutations in lung adenocarcinoma from female never-smokers varies with histologic subtypes and age at diagnosis. Clin Cancer Res ;18(7): journal.publications.chestnet.org 643