- Email: [email protected]
Histopathological and immunohistochemical features associated with clinical response to neoadjuvant gefitinib therapy in early stage non-small cell lung cancer
Lung Cancer 76 (2012) 235–241
Contents lists available at SciVerse ScienceDirect
Lung Cancer journal homepage: www.elsevier.com/locate/lungcan
Histopathological and immunohistochemical features associated with clinical response to neoadjuvant gefitinib therapy in early stage non-small cell lung cancer Humberto Lara-Guerra a,b,1,2 , Catherine T. Chung c,d,1,3 , Joerg Schwock c,d , Melania Pintilie e , David M. Hwang c,d , Natasha B. Leighl f,g , Thomas K. Waddell a,b , Ming-Sound Tsao c,d,∗ a
Division of Thoracic Surgery, University Health Network, Toronto, ON, Canada Department of Surgery and Institute of Medical Sciences, University of Toronto, Toronto, ON, Canada c Department of Pathology, University Health Network and Princess Margaret Hospital, Toronto, ON, Canada d Department of Laboratory Medicine and Pathobiology, University of Toronto, Toronto, ON, Canada e Department of Biostatistics, University Health Network and Princess Margaret Hosital, Toronto, ON, Canada f Division of Medical Oncology and Haematology, University Health Network and Princess Margaret Hospital, Toronto, ON, Canada g Department of Medicine, University of Toronto, Toronto, ON, Canada b
a r t i c l e
i n f o
Article history: Received 20 July 2011 Received in revised form 23 October 2011 Accepted 25 October 2011 Keywords: Neoadjuvant therapy Tyrosine kinase domain Mutation Gefitinib Response biomarker Proliferation marker Fibrosis Inflammatory infiltrate
a b s t r a c t To define the pathological features associated with response to epidermal growth factor receptor (EGFR) tyrosine kinase inhibitors (TKIs) in NSCLC, we have evaluated tumor histopathological features and immunohistochemical markers of proliferation (Ki-67) and epithelial mesenchymal transition (EMT) in 36 resected early stage NSCLC from patients treated preoperatively with gefitinib for 28 days. Tumors studied included 7 squamous cell carcinoma, 27 adenocarcinoma (ADC), one adenosquamous carcinoma, and one large cell carcinoma. Six of the ADC harboured an EGFR tyrosine kinase domain (TKD) mutation; five were the sensitizing type. Five ADC with TKD mutation demonstrated non-mucinous lepidic growth pattern as the dominant histological feature. Post-gefitinib treated EGFR TKD mutant tumors demonstrated lower tumor cellularity and proliferative index compared to wild type ADC and non-ADC cases, features correlating with clinical response. Responding tumors also showed large areas of fibrosis, within which focal residual viable tumor cells were noted. However, there was no significant correlation between the degree of fibrosis and radiological changes in tumor size. Expression of EMT markers was not associated with significant change in tumor size. The results suggest that radiologically assessed response to EGFR TKI in NSCLC is related to loss of tumor cellularity and reduced tumor cell proliferation, but residual viable tumor cells may persist even after prolonged treatment. Neoadjuvant studies in early stage NSCLC offer a unique opportunity to evaluate pathological and biomarker changes induced by targeted drugs. © 2011 Elsevier Ireland Ltd. All rights reserved.
1. Introduction It is now established that epidermal growth factor receptor (EGFR) tyrosine kinase inhibitors (TKIs) can improve the survival of previously treated and untreated advanced non-small cell lung cancer (NSCLC) patients [1,2]. Higher response rates to EGFR TKI
∗ Corresponding author at: Department of Pathology, University Health Network, 200 Elizabeth St., Toronto, ON, Canada M5G 2C5. Tel.: +1 416 340 4737; fax: +1 416 340 5517. E-mail address: [email protected] (M.-S. Tsao). 1 These authors contributed equally to this work. 2 Department of Thoracic and Cardiovascular Surgery, The University of Texas, MD Anderson Cancer Center, Houston, TX, USA. 3 Division of Pathology, The Hospital for Sick Children, Toronto, ON, Canada. 0169-5002/$ – see front matter © 2011 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.lungcan.2011.10.020
therapy have been observed in NSCLC patients who are East Asian, never-smokers, female or present with adenocarcinoma (ADC) histology [1,3,4]. Among ADC patients, better response rate was also reported in tumors with prominent lepidic (formerly bronchioloalveolar carcinoma or BAC) growth pattern [5,6]. However, molecular studies have demonstrated that the presence of mutations in the tyrosine kinase domain (TKD) of the EGFR gene is a better predictor of response than ADC histology [4,7,8]. In contrast, epithelial–mesenchymal transition (EMT) is a potential marker of resistance to EGFR TKI therapy [9,10]. Met is a tyrosine kinase receptor involved in EMT whose activation or gene amplification has been associated with resistance to EGFR TKIs in advanced NSCLC [11,12]. Due to the lack of surgical therapy offered to patients with advanced disease, detailed correlation of the pathological changes
236
H. Lara-Guerra et al. / Lung Cancer 76 (2012) 235–241
of tumors associated with EGFR TKI therapy has been limited. We have previously reported the feasibility of neoadjuvant gefitinib therapy in early stage NSCLC. Thirty-six patients received preoperative treatment. Tumor shrinkage was more frequent among women and nonsmokers. Partial response was seen in four patients (11%), and disease progression was seen in three patients (9%). All EGFR TKD mutant cases experienced some degree of tumor shrinkage with half of them reaching partial response by RECIST [14]. The strongest predictor of response was EGFR mutation [13]. We demonstrated that EGFR TKD mutations are the strongest predictor of clinical response. This trial has provided us with the unique opportunity to examine the histology of NSCLC tumors post-EGFR TKI therapy. We report here the histopathological findings and selected molecular features of NSCLC exposed to EGFR TKI treatment and their correlation with clinical response.
2. Materials and methods The study protocol was approved by the University Health Network Research Ethics Board. Thirty-six patients with biopsyproven clinical stage I NSCLC underwent 4 weeks preoperative treatment with gefitinib as part of a phase II clinical trial. Radiologic tumor reduction was determined by comparing CT scans of the chest before and after gefitinib regimen. Partial response by RECIST criteria was defined as at least 30% reduction in tumor diameter [14]. Study design, procedures, and association with conventional molecular predictors of response were reported previously [13]. For each case, the haematoxylin and eosin (H&E) slides and their corresponding surgical pathology blocks were retrieved. All evaluations of the histopathology and immunohistochemistry (IHC) were blinded to clinical data on response. Tumors were re-staged according to the 7th edition of the TNM classification [15]. Tumor features assessed included histological type according to 2004 WHO classification of lung cancer [16] and adapted to the recently published International Association for the Study of Lung Cancer/American Thoracic Society/European Respiratory Society classification of lung ADC [17], tumor differentiation/grade, tumor cellularity, presence of local invasion (lymphatic, vascular, pleural), intra- and peri-tumoral inflammatory response and lymphocytic aggregates, elastin staining pattern, and degree of fibrosis by Masson’s trichrome staining. For ADC, the proportion of various histological patterns (lepidic, acinar, papillary, and solid) was estimated and the pattern with the highest percentage of the tumor area was designated as the predominant pattern. Since the effect of prolonged (4 weeks) gefitinib therapy on ADC with predominantly lepidic type is unknown, we decided not to use the categories of adenocarcinoma in situ or minimally invasive adenocarcinoma in our reclassification of the tumors. The blocks with the greatest tumor cellularity were selected for IHC studies. Briefly, four micron sections were dried overnight at 60 ◦ C, deparaffinised with xylene, transferred through changes of ethanol and rehydrated using standard histological protocol. After blocking of endogenous hydrogen peroxide activity by 10 min incubation in 3% hydrogen peroxide buffer and microwave antigen retrieval in MicroMed T/T oven (Hacker Instruments, Fairfield, NJ) for 10 min at 120 ◦ C, the sections were incubated with primary antibodies for Ki-67 (clone MIB-1, mouse monoclonal, Dako 1:200; 1 h), Met (clone DL-21, mouse monoclonal, Upstate Biotechnology 1:200; overnight), E-cadherin (clone 36B5, mouse monoclonal, Vector lab 1:100; overnight), vimentin (clone VIM 3B4, mouse monoclonal, Dako 1:300; 1 h, pepsin antigen retrieval), and Snail (goat polyclonal, R&D 1:1000; overnight). After washing and incubation with respective secondary antibodies and color development using NovaRed solution (Vector Laboratories), the slides were counterstained with Mayer’s haematoxylin. Negative control slides
omitting the primary antibodies were included in all staining procedures. The cellular compartment, staining intensity of the signal and percentage of tumor cells stained were recorded independently by three observers (CTC, JS and HLG) after an initial training session (led by the senior pulmonary pathologist MST) and without knowledge of the clinical data. The mean of the three scores was calculated. When a discrepancy (≥20% of tumor cells or 2 points of intensity) among the initial three scorers was observed, the case was reviewed together by four observers (CTC, JS, HLG, MST) using a multi-headed microscope and consensus was obtained. Associations between variables and molecular markers were analyzed using Wilcoxon 2-sample test between a continuous variable and a binary variable, and Fisher’s exact test between categorical variables. Associations with radiologic measurements of tumor change were tested using linear regression. P values < 0.05 were considered significant. All analyses were performed by the study biostatistician (MP) using SAS version 9 (SAS Institute Inc, Cary, NC). 3. Results 3.1. Histopathological assessment and IHC markers in early NSCLC treated with neoadjuvant gefitinib The median number of tumor slides available and reviewed per case was 4 (range 2–10). Although patients were clinical stage I preoperatively, the pathologic stages of the resected tumors were: 19 (53%) stage IA (14 pT1a, 5 pT1b; N0), 9 (25%) stage IB (pT2aN0), 1 (3%) stage IIA (pT2bN0), 6 (16%) stage IIB (3 pT2bN1, 3 pT3N0), 1 (3%) stage IIIA (pT3N1). Pleural (n = 10), vascular (n = 7) or lymphatic (n = 1) invasion was observed in 36% of cases. Twenty-seven of the 36 cases were ADC. Among these included nine (33%) nonmucinous with lepidic predominant pattern and two (7%) mucinous with predominant lepidic pattern. Among the remaining 16 ADC, the predominant histological patterns were acinar in 11 and papillary in 5 (Table 1). Tumors showed varying degrees of inflammatory cell infiltrates with 11 (31%) demonstrating moderate to marked infiltration. Seven tumors showed prominent lymphocytic aggregates (Table 1) and one tumor showed a marked eosinophilic infiltrate. The distribution of tumor cellularity and percentages of tumor fibrosis and necrosis are shown in Fig. 1. Intratumoral fibrosis was observed in 53% of the cases, with the mean area of fibrosis being 16% (range 0–100%). Thirteen (36%) cases demonstrated tumor necrosis, but
Table 1 Histopathological features observed in 36 surgically resected NSCLC specimens after neoadjuvant gefitinib treatment. Tumors and features Squamous cell carcinoma Large cell carcinoma Adenosquamous carcinoma Adenocarcinomas (n = 27) Well differentiated Moderately differentiated Poorly differentiated Non-mucinous predominantly lepidic Mucinous predominantly lepidic Predominant acinar pattern Predominant papillary pattern All cases (n = 36) Intra-/peri-tumoral inflammatory cell infiltrate Minimal – mild Moderate – severe Lymphocytic aggregates Fibrosis Necrosis
Frequency 7 1 1 12 12 3 9 2 11 5
25 11 7 19 13
H. Lara-Guerra et al. / Lung Cancer 76 (2012) 235–241
237
Fig. 1. Distribution of tumor histopathological features among early stage NSCLC patients treated with neoadjuvant gefitinib.
all involved less than 25% of the total tumor area. The mean tumor cellularity was 52% (range 5–95%). Inter-observer correlation coefficients were above 0.85 for all IHC markers except vimentin (Supplementary Table S1). For Met and E-cadherin, percent membranous staining was chosen as it provided the highest scoring reliability and is the main localization involved with their molecular role. The medians of percent tumor cells stained (range) were: Ki-67 proliferative index 26.7% (1–76%), Met 1.1% (0–60%), E-cadherin 70.8% (10–95%), vimentin 0% (0–15%), Snail 0.3% (0–70%) (Fig. 1). 3.2. Association of adenocarcinoma growth patterns with molecular markers of response to EGFR TKIs We previously reported the molecular markers associated with clinical benefit to EGFR TKI therapy in this trial [13]. The six cases with EGFR TKD mutations were ADC. EGFR TKD mutations were
present in a significantly higher proportion of tumors with nonmucinous lepidic predominant histology (56%) (Table 2). Seven out of 9 (78%) non-mucinous ADC with lepidic predominant pattern were also EGFR FISH+ but EGFR amplification was found mainly in ADC with papillary/acinar predominant patterns, and not in those with a lepidic predominant pattern. In contrast, KRAS mutations were found in all patterns of ADC. The two cases of mucinous ADC with lepidic pattern did not harbour aberrations in any of the molecular markers tested (Table 2). 3.3. Histopathological features in adenocarcinomas with EGFR TKD mutations Four of the six ADC with EGFR TKD mutation harboured an exon 19 deletion. Two of these fulfilled the criteria of radiological partial response (PR) according to the RECIST criteria [14]. The other two cases demonstrated 10% and 27% reduction in tumor
238
H. Lara-Guerra et al. / Lung Cancer 76 (2012) 235–241
Table 2 Correlation between predominantly histological patterns of adenocarcinoma and molecular markers. Predominant Pattern
N
Partial response
EGFR FISH+
EGFR Amp
KRAS Mutation
EGFR TKD Mutation
Total Mucinous lepidic Non-mucinous lepidic Acinar Papillary
27 2 9 11 5
4 (15%) 0 (0%) 2 (22%) 1 (9%) 1 (20%)
18 (69%) 0 (0%) 7(78%) 8 (73%) 3 (60%)
4 (15%) 0 (0%) 0 (0%) 2 (18%) 2 (40%)
6 (22%) 0 (0%) 1 (11%) 3 (27%) 2 (40%)
6 (22%) 0 (0%) 5 (56%)* 1 (9%) 0 (0%)
Amp, amplification; TKD, tyrosine kinase domain; EGFR, epidermal growth factor receptor; FISH+, high polysomy and amplification by fluorescence in situ hybridization. * p = 0.031.
diameter, which did not reach the RECIST criteria level for PR. All four tumors demonstrated extensive central fibrosis with marked loss of cellularity (Fig. 2, A–F). Residual viable tumor cells were present focally within the fibrous stroma and particularly in areas with marked lymphocytic infiltrates (Fig. 2, F and H). Residual tumor cells demonstrated low proliferative activity (Ki-67 staining) in the fibrous areas (Fig. 2C) and high activity in lymphocyte rich areas (Fig. 2I). One ADC with radiological PR and harbouring an exon 21 L858R mutation demonstrated lepidic predominant histology (Fig. 2J) with tumor cells showing a low cuboidal appearance (Fig. 2K). This tumor also showed a focal area of collapse with alveolar hemorrhage. A tumor with exon 21 L833V/H835L mutation did not respond to gefitinib treatment; this tumor also showed a non-mucinous lepidic predominant histology but with marked lymphocytic infiltration of the alveolar interstitium (Fig. 2L). However, lymphocytic aggregate or inflammatory response was not correlated with clinical response to changes in tumor diameter assessed by CT scan (Table 3).
3.4. Association of EGFR TKD mutant adenocarcinomas with histopathological features and IHC markers Although EGFR TKD mutant ADC showed extensive fibrotic changes (mean: 32.8% of tumor area), this was not significantly different from EGFR wild type ADC (19.9%) or non-ADC tumors (7.5%) (Fig. 3). In contrast, EGFR TKD mutant ADC cases demonstrated significantly lower cellularity (mean: 24.2% of tumor area) and Ki-67 proliferative index (mean: 4.6%) compared to EGFR wild type ADC (cellularity: 58.6%, p = 0.01; Ki-67: 31.4%, p = 0.002) and non-ADC tumors (cellularity: 55%, p = 0.026; Ki-67: 49.8%, p = 0.001). Met membranous staining levels were low in all groups, with no significant differences detected between ADC with EGFR TKD mutation (mean 2.5% of tumor cells), EGFR wild type (7.9%) or non-ADC (1.8%) tumors (Fig. 3). In contrast, E-cadherin levels were significantly higher in EGFR TKD mutant ADC (70.4% of tumor cells) compared to non-ADC tumors (44.4%, p = 0.026), but were similar when compared to EGFR wild type ADC (64%) (Fig. 3). Vimentin
Fig. 2. Histology of four tumors associated with partial response to gefitinib. (A)–(C) An Exon 19 L747 P752 deletion tumor shows fibrosis with extensive loss of tumor cells (A) and focal residual viable tumor cells (B) with low proliferative activity (C). (D)–(F) An Exon 19 E746 S750insV tumor shows focal intense lymphocytic infiltrates (D), prominent fibrosis (E) and focal residual viable tumor cells (F). (G)–(I) An EGFR wild type but amplified tumor shows large areas of tumor cell loss and fibrosis (G) and areas with marked lymphocytic infiltration (G and H) and foci of residual highly proliferative tumor cells (I) within. (J)-(K) An exon 21 L858R mutant tumor shows a predominant lepidic growth pattern (J) with low cuboidal tumor cells growing along the pre-existing alveolar framework (K). (L) An exon 21 L833V/H835L tumor not responsive to gefitinib shows a predominantly lepidic pattern and prominent stromal lymphocytic infiltrate.
H. Lara-Guerra et al. / Lung Cancer 76 (2012) 235–241
239
Fig. 3. Association between histopathological features and immunohistochemical markers with EGFR tyrosine kinase domain genotypes. Values on Y-axis represent distribution of averaged percentages of features/markers observed in each tumor assessed. In the box plots, top represents the 75th percentile (upper quartile), bottom the 25th percentile (lower quartile), and the middle line the 50th percentile (median). Whiskers represent the highest and lowest values excluding outliers. Outliers (values 1.5 or more times the interquartile range) are represented by circles beyond the whiskers. The p values were assessed by Wilcoxon–Mann–Whitney 2-samples. For Met, vimentin and Snail, the log 10 of percentages were shown, due to their low values.
240
H. Lara-Guerra et al. / Lung Cancer 76 (2012) 235–241
Table 3 Correlation of histopathological features and molecular markers with percent change in maximum tumor diameter on CT scans before and after gefitinib treatment. Characteristic
R2
ˇ (SE)
p
Predominantly lepidic Predominantly acinar Predominantly papillary Tumor grade Inflammatory response Lymphocytic aggregate Pathologic tumor diameter (cm) Fibrosis (%) Necrosis (%) Cellularity (%) Ki-67 index Met membranous staining E-cadherin membranous staining Vimentin cytoplasmic staining Snail nuclear staining
0.200 0.013 0.044 0.166 0.005 0.05 0.041 0.07 0.109 0.217 0.152 0.024 0.057 0 0.047
−4.588 (1.6) 1.076 (1.65) 3.035 (2.46) 9.684 (3.77) 1.657 (3.98) −9.964 (7.53) 2.748 (2.32) −0.197 (0.125) 0.904 (0.45) 0.31 (0.102) 0.289 (0.119) 0.242 (0.27) −0.17 (0.12) 0.005 (1.11) 0.328 (0.258)
0.007 0.52 0.227 0.015 0.68 0.195 0.244 0.125 0.053 0.005 0.02 0.375 0.167 0.997 0.213
R2 , coefficient of determination (proportion of the variation explained by the model); , standardized coefficient (Relative importance of the contribution of the predictor to the model. A negative value means larger the predictor, larger tumor reduction; positive value means larger the predictor, larger tumor growth); SE, standard error.
and Snail levels were low in all groups with the only significant difference observed between Snail in EGFR TKD mutant ADC (0.6%) compared to non-ADC tumors (2.7%) due to a single outlier in the last group (70%). 3.5. Correlation of histopathological features and IHC markers with clinical response Among the histopathological features, gefitinib response assessed by radiologic tumor diameter reduction was significantly associated with non-mucinous lepidic predominant pattern, higher tumor grade and lower tumor cellularity (Table 3). Among IHC markers, only low proliferative index (Ki-67) was significantly associated with radiologic response to pre-operative gefitinib therapy. Radiologic response did not correlate significantly with either acinar or papillary growth patterns in ADC, extent of tumor fibrosis, inflammatory response or tumor necrosis. 4. Discussion It is well established that EGFR TKI therapy can induce dramatic tumor shrinkage in a NSCLC subpopulation mainly defined by the presence of EGFR TKD mutations, but the histopathological correlates of this response have not been previously described. This is because EGFR TKIs have been used primarily to treat advanced NSCLC patients in whom histopathological evaluation of treated tumors is not feasible since surgical resection is not part of their standard of care. We recently conducted a neoadjuvant trial to assess gefitinib-induced clinical tumor response, drug toxicity, and selected clinical and molecular predictive markers of response in early NSCLC [13]. This study allowed us to assess the histopathological correlates of response in early NSCLC after EGFR TKI therapy. Our results demonstrate that significant radiologic tumor response is observed mainly in non-mucinous ADC with a lepidic-predominant growth pattern, which is also most often found in tumors with EGFR TKD mutations. Radiological tumor reduction was associated with a decrease in tumor cellularity and proliferation (Ki-67 index), but not with other factors evaluated, including acinar or papillary histology, inflammatory infiltration, fibrosis, and markers of EMT phenotype (Met/hepatocyte growth factor receptor, E-cadherin, vimentin and Snail).
Previous reports have documented the histological features found in ADC that are more commonly associated with EGFR TKD mutations in untreated tumors. These include low grade (well/moderate) and tumors with prominent lepidic (previously called BAC pattern) and/or papillary or micropapillary growth patterns [18–21]. These histological patterns have also been associated with a higher response rate to EGFR TKI therapy [5,6,22]. However, other reports have not found significant associations of these or other specific histological patterns with EGFR TKD mutations [23–25]. In our small series, 5 of 6 (83%) EGFR TKD mutant ADC were non-mucinous ADC with a predominant lepidic pattern, while the sixth tumor had a predominant acinar pattern. With one-quarter of our cohort being ADC with lepidic predominant pattern, the prevalence of EGFR TKD mutations in this ADC subtype accounts for their association with clinical response. One of the most important aspects of this report is the description of histopathological changes in tumors showing significant clinical response to gefitinib therapy. These tumors were characterized by extensive fibrosis and loss of tumor cells. However, there was no significant association between the extent of fibrotic changes and clinical response. This discrepancy is likely due to the difficulty in distinguishing between treatment and non-treatment related fibrosis, as focal fibrosis occurs commonly during cancer development [26]. Response to gefitinib therapy was, however, significantly associated with low tumor cellularity and a low proliferative index. Not surprisingly, significantly lower tumor cellularity and proliferative index were also found in ADC harbouring EGFR TKD mutations. These findings are in keeping with previous reports indicating that EGFR TKI therapy reduces cellular proliferation in advanced NSCLC [27,28]. Because EMT is a potential marker of resistance to EGFR TKI therapy, we assessed the expression of various markers of EMT in our cohort. Met gene amplification has been identified as potential predictor of resistance to EGFR TKI in different advanced NSCLC cohorts [29], but in our cohort of early NSCLC, we found uniformly low expression of Met, with no association of Met protein expression with radiologic response to therapy. E-cadherin expression was significantly higher in EGFR TKD mutant ADC compared with nonADC cases but not when compared with EGFR TKD wild-type ADC. Taken together with the lack of differences in expression of other EMT markers, this suggests variability of expression between histologic NSCLC subtypes, rather than variation based on genotypic differences between ADC tumors. Despite limited by a lack of corresponding pre-treatment data, this study is to the best of our knowledge the first to report the specific histopathological findings in NSCLC, particularly in EGFR TKD mutant ADC, demonstrating clinical response to treatment with an EGFR TKI. While the original trial was designed for hypothesis generating and the exploratory findings reported here require further validation, this assessment was enabled by the unique design of our neoadjuvant study, suggesting that similar designs may offer important opportunities for future studies with novel drugs aiming to correlate clinical response, histopathological changes, biomarker expression, and functional imaging techniques. This will lead to a better understanding of the tumor response to targeted therapies.
Author contributions HLG, CTC, NBL, TKW, and MST were responsible for the concept and design of the study. HLG, CTC, JS, MP, DMH, and MST were responsible for data acquisition, analysis and interpretation of results. HLG and MST were in charge of article drafting. All authors were involved in revising the article critically and approved its final version.
H. Lara-Guerra et al. / Lung Cancer 76 (2012) 235–241
Conflict of interest statement No conflicts of interest are declared from authors. The clinical trial from which samples analyzed are derived was supported in part by the Princess Margaret Hospital/University Health Network through Junior Investigator Awards (NBL, TKW), and in part and in part by AstraZeneca Canada (Mississauga, ON). AstraZeneca Canada was not involved in the design of the study; collection, analysis or interpretation of data; nor in the writing process or in the decision to submit for publication of this article. Acknowledgments Dr. Lara-Guerra was supported by fellowship from the Canadian Institute for Health Research (CIHR). This work was supported by grant (07NOV-78) from the Ontario Institute of Cancer Research; Dr. Waddell is the R. Fraser Elliott Chair in Transplantation Research; Dr. Tsao is the Choksi Chair in Lung Cancer Translational Research. The NIFTEE study was supported in part by the Princess Margaret Hospital/University Health Network and in part by AstraZeneca Canada (Mississauga, ON). We thank James Ho, Jing Xu and Olga Ludkovski from the Applied Molecular Profiling Laboratory for their technical assistance in the molecular studies and Dr. Frances Shepherd for her support. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.lungcan.2011.10.020. References [1] Shepherd FA, Rodrigues Pereira J, Ciuleanu T, Tan EH, Hirsh V, Thongprasert S, et al. Erlotinib in previously treated non-small-cell lung cancer. N Engl J Med 2005;353:123–32. [2] Mok TS, Wu YL, Thongprasert S, Yang CH, Chu DT, Saijo N, et al. Gefitinib or caboplatin-paclitaxel in pulmonary adenocarcinoma. N Engl J Med 2009;361:947–57. [3] Sakurada A, Shepherd FA, Tsao MS. Epidermal growth factor receptor tyrosine kinase inhibitors in lung cancer: impact of primary or secondary mutations. Clin Lung Cancer 2006;7:138–44. [4] Ellis PM, Morzycki W, Melosky B, Butts C, Hirsh V, Krasnoshtein F, et al. The role of the epidermal growth factor receptor tyrosine kinase inhibitors as therapy for advanced, metastatic, and recurrent non-small-cell lung cancer: a Canadian national consensus statement. Curr Oncol 2009;16:27–48. [5] Miller VA, Kris MG, Shah N, Patel J, Azzoli C, Gomez J, et al. Bronchioloalveolar pathologic subtype and smoking history predict sensitivity to gefitinib in advanced non-small-cell lung cancer. J Clin Oncol 2004;22:1103–9. [6] Miller VA, Riely GJ, Zakowski MF, Li AR, Patel JD, Heelan RT, et al. Molecular characteristics of bronchioloalveolar carcinoma and adenocarcinoma, bronchioloalveolar carcinoma subtype, predict response to erlotinib. J Clin Oncol 2008;26:1472–8. [7] Zhu CQ, da Cunha Santos G, Ding K, Sakurada A, Cutz JC, Liu N, et al. Role of KRAS and EGFR as biomarkers of response to erlotinib in National Cancer Institute of Canada Clinical Trials Group Study BR.21. J Clin Oncol 2008;26:4268–75. [8] Shepherd FA, Tsao MS. Unraveling the mystery of prognostic and predictive factors in epidermal growth factor receptor therapy. J Clin Oncol 2006;24:1219–20. [9] Yauch RL, Januario T, Eberhard DA, Cavet G, Zhu W, Fu L, et al. Epithelial versus mesenchymal phenotype determines in vitro sensitivity and predicts clinical activity of erlotinib in lung cancer patients. Clin Cancer Res 2005;11:8686–98.
241
[10] Miyanaga A, Gemma A, Ando M, Kosaihira S, Noro R, Minegishi Y, et al. Ecadherin expression and epidermal growth factor receptor mutation status predict outcome in non-small cell lung cancer patients treated with gefitinib. Oncol Rep 2008;19:377–83. [11] Zucali PA, Ruiz MG, Giovannetti E, Destro A, Varella-Garcia M, Floor K, et al. Role of cMET expression in non-small-cell lung cancer patients treated with EGFR tyrosine kinase inhibitors. Ann Oncol 2008;19:1605–12. [12] Cappuzzo F, Jänne PA, Skokan M, Finocchiaro G, Rossi E, Ligorio C, et al. MET increased gene copy number and primary resistance to gefitinib therapy in non-small-cell lung cancer patients. Ann Oncol 2009;20:298–304. [13] Lara-Guerra H, Waddell TK, Salvarrey MA, Joshua AM, Chung CT, Paul N, et al. Phase II study of preoperative gefitinib in clinical stage I non-small-cell lung cancer. J Clin Oncol 2009;27:6229–36. [14] Therasse P, Arbuck SG, Eisenhauer EA, Wanders J, Kaplan RS, Rubinstein L, et al. New guidelines to evaluate the response to treatment in solid tumors European Organization for Research and Treatment of Cancer, National Cancer Institute of the United States, National Cancer Institute of Canada. J Natl Cancer Inst 2000;92:205–16. [15] Union Internacional Contra el Cancer. In: Wittekind C, Sobin LH, editors. TNM Classification of Malignant Tumors. 7th edition Wiley-Liss; 2009. [16] World Health Organization/International Agency for Research on Cancer. In: Brambilla E, Muller-Hermelink HK, Harris CC, Travis WD, editors. World Health Organization classification of tumors: pathology and genetics of tumors of the lung, pleura, thymus and heart. IARC Press; 2004. [17] Travis WD, Brambilla E, Noguchi M, Nicholson AG, Geisinger KR, Yatabe Y, et al. International Association for the Study of Lung Cancer/American Thoracic Society/European Respiratory Society International multidisciplinary classification of lung adenocarcinoma. J Thoracic Oncol 2011;6:244–85. [18] Motoi N, Szoke J, Riely GJ, Seshan VE, Kris MG, Rusch VW, 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 2008;32:810–27. [19] Haneda H, Sasaki H, Shimizu S, Endo K, Suzuki E, Yukiue H, et al. Epidermal growth factor receptor gene mutation defines distinct subsets among small adenocarcinomas of the lung. Lung Cancer 2006;52:47–52. [20] Matsumoto S, Iwakawa R, Kohno T, Suzuki K, Matsuno Y, Yamamoto S, et al. Frequent EGFR mutations in noninvasive bronchioloalveolar carcinoma. Int J Cancer 2006;118:2498–504. [21] Ninomiya H, Hiramatsu M, Inamura K, Nomura K, Okui M, Miyoshi T, et al. Correlation between morphology and EGFR mutations in lung adenocarcinomas: significance of the micropapillary pattern and the hobnail cell type. Lung Cancer 2009;63:235–40. [22] Kim YH, Ishii G, Goto K, Nagai K, Tsuta K, Shiono S, et al. Dominant papillary subtype is a significant predictor of the response to gefitinib in adenocarcinoma of the lung. Clin Cancer Res 2004;10:7311–7. [23] Chantranuwat C, Sriuranpong V, Huapai N, Chalermchai T, Leungtaweeboon K, Voravud N, et al. Histopathologic characteristics of pulmonary adenocarcinomas with and without EGFR mutation. J Med Assoc Thai 2005;88:S322–9. [24] Shigematsu H, Lin L, Takahashi T, Nomura M, Suzuki M, Wistuba II, et al. Clinical and biological features associated with epidermal growth factor receptor gene mutations in lung cancers. J Natl Cancer Inst 2005;97:339–46. [25] Dacic S, Kelly L, Shuai Y, Nikiforova MN. miRNA expression profiling of lung adenocarcinomas: correlation with mutational status. Mod Pathol 2010;23:159–68. [26] Goldstein NS, Mani A, Chmielewski G, Welsh R, Pursel S. Prognostic factors in T1 NO MO adenocarcinomas and bronchioloalveolar carcinomas of the lung. Am J Clin Pathol 1999;112:391–402. [27] Baselga J, Rischin D, Ranson M, Calvert H, Raymond E, Kieback DG, et al. Phase I safety, pharmacokinetic, and pharmacodynamic trial of ZD1839, a selective oral epidermal growth factor receptor tyrosine kinase inhibitor, in patients with five selected solid tumor types. J Clin Oncol 2002;20:4292–302. [28] Felip E, Rojo F, Reck M, Heller A, Klughammer B, Sala G, et al. A phase II pharmacodynamic study of erlotinib in patients with advanced non-small cell lung cancer previously treated with platinum-based chemotherapy. Clin Cancer Res 2008;14:3867–74. [29] Bean J, Brennan C, Shih JY, Riely G, Viale A, Wang L, et al. MET amplification occurs with or without T790M mutations in EGFR mutant lung tumors with acquired resistance to gefitinib or erlotinib. Proc Natl Acad Sci USA 2007;104:20932–7.
Contents lists available at SciVerse ScienceDirect
Lung Cancer journal homepage: www.elsevier.com/locate/lungcan
Histopathological and immunohistochemical features associated with clinical response to neoadjuvant gefitinib therapy in early stage non-small cell lung cancer Humberto Lara-Guerra a,b,1,2 , Catherine T. Chung c,d,1,3 , Joerg Schwock c,d , Melania Pintilie e , David M. Hwang c,d , Natasha B. Leighl f,g , Thomas K. Waddell a,b , Ming-Sound Tsao c,d,∗ a
Division of Thoracic Surgery, University Health Network, Toronto, ON, Canada Department of Surgery and Institute of Medical Sciences, University of Toronto, Toronto, ON, Canada c Department of Pathology, University Health Network and Princess Margaret Hospital, Toronto, ON, Canada d Department of Laboratory Medicine and Pathobiology, University of Toronto, Toronto, ON, Canada e Department of Biostatistics, University Health Network and Princess Margaret Hosital, Toronto, ON, Canada f Division of Medical Oncology and Haematology, University Health Network and Princess Margaret Hospital, Toronto, ON, Canada g Department of Medicine, University of Toronto, Toronto, ON, Canada b
a r t i c l e
i n f o
Article history: Received 20 July 2011 Received in revised form 23 October 2011 Accepted 25 October 2011 Keywords: Neoadjuvant therapy Tyrosine kinase domain Mutation Gefitinib Response biomarker Proliferation marker Fibrosis Inflammatory infiltrate
a b s t r a c t To define the pathological features associated with response to epidermal growth factor receptor (EGFR) tyrosine kinase inhibitors (TKIs) in NSCLC, we have evaluated tumor histopathological features and immunohistochemical markers of proliferation (Ki-67) and epithelial mesenchymal transition (EMT) in 36 resected early stage NSCLC from patients treated preoperatively with gefitinib for 28 days. Tumors studied included 7 squamous cell carcinoma, 27 adenocarcinoma (ADC), one adenosquamous carcinoma, and one large cell carcinoma. Six of the ADC harboured an EGFR tyrosine kinase domain (TKD) mutation; five were the sensitizing type. Five ADC with TKD mutation demonstrated non-mucinous lepidic growth pattern as the dominant histological feature. Post-gefitinib treated EGFR TKD mutant tumors demonstrated lower tumor cellularity and proliferative index compared to wild type ADC and non-ADC cases, features correlating with clinical response. Responding tumors also showed large areas of fibrosis, within which focal residual viable tumor cells were noted. However, there was no significant correlation between the degree of fibrosis and radiological changes in tumor size. Expression of EMT markers was not associated with significant change in tumor size. The results suggest that radiologically assessed response to EGFR TKI in NSCLC is related to loss of tumor cellularity and reduced tumor cell proliferation, but residual viable tumor cells may persist even after prolonged treatment. Neoadjuvant studies in early stage NSCLC offer a unique opportunity to evaluate pathological and biomarker changes induced by targeted drugs. © 2011 Elsevier Ireland Ltd. All rights reserved.
1. Introduction It is now established that epidermal growth factor receptor (EGFR) tyrosine kinase inhibitors (TKIs) can improve the survival of previously treated and untreated advanced non-small cell lung cancer (NSCLC) patients [1,2]. Higher response rates to EGFR TKI
∗ Corresponding author at: Department of Pathology, University Health Network, 200 Elizabeth St., Toronto, ON, Canada M5G 2C5. Tel.: +1 416 340 4737; fax: +1 416 340 5517. E-mail address: [email protected] (M.-S. Tsao). 1 These authors contributed equally to this work. 2 Department of Thoracic and Cardiovascular Surgery, The University of Texas, MD Anderson Cancer Center, Houston, TX, USA. 3 Division of Pathology, The Hospital for Sick Children, Toronto, ON, Canada. 0169-5002/$ – see front matter © 2011 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.lungcan.2011.10.020
therapy have been observed in NSCLC patients who are East Asian, never-smokers, female or present with adenocarcinoma (ADC) histology [1,3,4]. Among ADC patients, better response rate was also reported in tumors with prominent lepidic (formerly bronchioloalveolar carcinoma or BAC) growth pattern [5,6]. However, molecular studies have demonstrated that the presence of mutations in the tyrosine kinase domain (TKD) of the EGFR gene is a better predictor of response than ADC histology [4,7,8]. In contrast, epithelial–mesenchymal transition (EMT) is a potential marker of resistance to EGFR TKI therapy [9,10]. Met is a tyrosine kinase receptor involved in EMT whose activation or gene amplification has been associated with resistance to EGFR TKIs in advanced NSCLC [11,12]. Due to the lack of surgical therapy offered to patients with advanced disease, detailed correlation of the pathological changes
236
H. Lara-Guerra et al. / Lung Cancer 76 (2012) 235–241
of tumors associated with EGFR TKI therapy has been limited. We have previously reported the feasibility of neoadjuvant gefitinib therapy in early stage NSCLC. Thirty-six patients received preoperative treatment. Tumor shrinkage was more frequent among women and nonsmokers. Partial response was seen in four patients (11%), and disease progression was seen in three patients (9%). All EGFR TKD mutant cases experienced some degree of tumor shrinkage with half of them reaching partial response by RECIST [14]. The strongest predictor of response was EGFR mutation [13]. We demonstrated that EGFR TKD mutations are the strongest predictor of clinical response. This trial has provided us with the unique opportunity to examine the histology of NSCLC tumors post-EGFR TKI therapy. We report here the histopathological findings and selected molecular features of NSCLC exposed to EGFR TKI treatment and their correlation with clinical response.
2. Materials and methods The study protocol was approved by the University Health Network Research Ethics Board. Thirty-six patients with biopsyproven clinical stage I NSCLC underwent 4 weeks preoperative treatment with gefitinib as part of a phase II clinical trial. Radiologic tumor reduction was determined by comparing CT scans of the chest before and after gefitinib regimen. Partial response by RECIST criteria was defined as at least 30% reduction in tumor diameter [14]. Study design, procedures, and association with conventional molecular predictors of response were reported previously [13]. For each case, the haematoxylin and eosin (H&E) slides and their corresponding surgical pathology blocks were retrieved. All evaluations of the histopathology and immunohistochemistry (IHC) were blinded to clinical data on response. Tumors were re-staged according to the 7th edition of the TNM classification [15]. Tumor features assessed included histological type according to 2004 WHO classification of lung cancer [16] and adapted to the recently published International Association for the Study of Lung Cancer/American Thoracic Society/European Respiratory Society classification of lung ADC [17], tumor differentiation/grade, tumor cellularity, presence of local invasion (lymphatic, vascular, pleural), intra- and peri-tumoral inflammatory response and lymphocytic aggregates, elastin staining pattern, and degree of fibrosis by Masson’s trichrome staining. For ADC, the proportion of various histological patterns (lepidic, acinar, papillary, and solid) was estimated and the pattern with the highest percentage of the tumor area was designated as the predominant pattern. Since the effect of prolonged (4 weeks) gefitinib therapy on ADC with predominantly lepidic type is unknown, we decided not to use the categories of adenocarcinoma in situ or minimally invasive adenocarcinoma in our reclassification of the tumors. The blocks with the greatest tumor cellularity were selected for IHC studies. Briefly, four micron sections were dried overnight at 60 ◦ C, deparaffinised with xylene, transferred through changes of ethanol and rehydrated using standard histological protocol. After blocking of endogenous hydrogen peroxide activity by 10 min incubation in 3% hydrogen peroxide buffer and microwave antigen retrieval in MicroMed T/T oven (Hacker Instruments, Fairfield, NJ) for 10 min at 120 ◦ C, the sections were incubated with primary antibodies for Ki-67 (clone MIB-1, mouse monoclonal, Dako 1:200; 1 h), Met (clone DL-21, mouse monoclonal, Upstate Biotechnology 1:200; overnight), E-cadherin (clone 36B5, mouse monoclonal, Vector lab 1:100; overnight), vimentin (clone VIM 3B4, mouse monoclonal, Dako 1:300; 1 h, pepsin antigen retrieval), and Snail (goat polyclonal, R&D 1:1000; overnight). After washing and incubation with respective secondary antibodies and color development using NovaRed solution (Vector Laboratories), the slides were counterstained with Mayer’s haematoxylin. Negative control slides
omitting the primary antibodies were included in all staining procedures. The cellular compartment, staining intensity of the signal and percentage of tumor cells stained were recorded independently by three observers (CTC, JS and HLG) after an initial training session (led by the senior pulmonary pathologist MST) and without knowledge of the clinical data. The mean of the three scores was calculated. When a discrepancy (≥20% of tumor cells or 2 points of intensity) among the initial three scorers was observed, the case was reviewed together by four observers (CTC, JS, HLG, MST) using a multi-headed microscope and consensus was obtained. Associations between variables and molecular markers were analyzed using Wilcoxon 2-sample test between a continuous variable and a binary variable, and Fisher’s exact test between categorical variables. Associations with radiologic measurements of tumor change were tested using linear regression. P values < 0.05 were considered significant. All analyses were performed by the study biostatistician (MP) using SAS version 9 (SAS Institute Inc, Cary, NC). 3. Results 3.1. Histopathological assessment and IHC markers in early NSCLC treated with neoadjuvant gefitinib The median number of tumor slides available and reviewed per case was 4 (range 2–10). Although patients were clinical stage I preoperatively, the pathologic stages of the resected tumors were: 19 (53%) stage IA (14 pT1a, 5 pT1b; N0), 9 (25%) stage IB (pT2aN0), 1 (3%) stage IIA (pT2bN0), 6 (16%) stage IIB (3 pT2bN1, 3 pT3N0), 1 (3%) stage IIIA (pT3N1). Pleural (n = 10), vascular (n = 7) or lymphatic (n = 1) invasion was observed in 36% of cases. Twenty-seven of the 36 cases were ADC. Among these included nine (33%) nonmucinous with lepidic predominant pattern and two (7%) mucinous with predominant lepidic pattern. Among the remaining 16 ADC, the predominant histological patterns were acinar in 11 and papillary in 5 (Table 1). Tumors showed varying degrees of inflammatory cell infiltrates with 11 (31%) demonstrating moderate to marked infiltration. Seven tumors showed prominent lymphocytic aggregates (Table 1) and one tumor showed a marked eosinophilic infiltrate. The distribution of tumor cellularity and percentages of tumor fibrosis and necrosis are shown in Fig. 1. Intratumoral fibrosis was observed in 53% of the cases, with the mean area of fibrosis being 16% (range 0–100%). Thirteen (36%) cases demonstrated tumor necrosis, but
Table 1 Histopathological features observed in 36 surgically resected NSCLC specimens after neoadjuvant gefitinib treatment. Tumors and features Squamous cell carcinoma Large cell carcinoma Adenosquamous carcinoma Adenocarcinomas (n = 27) Well differentiated Moderately differentiated Poorly differentiated Non-mucinous predominantly lepidic Mucinous predominantly lepidic Predominant acinar pattern Predominant papillary pattern All cases (n = 36) Intra-/peri-tumoral inflammatory cell infiltrate Minimal – mild Moderate – severe Lymphocytic aggregates Fibrosis Necrosis
Frequency 7 1 1 12 12 3 9 2 11 5
25 11 7 19 13
H. Lara-Guerra et al. / Lung Cancer 76 (2012) 235–241
237
Fig. 1. Distribution of tumor histopathological features among early stage NSCLC patients treated with neoadjuvant gefitinib.
all involved less than 25% of the total tumor area. The mean tumor cellularity was 52% (range 5–95%). Inter-observer correlation coefficients were above 0.85 for all IHC markers except vimentin (Supplementary Table S1). For Met and E-cadherin, percent membranous staining was chosen as it provided the highest scoring reliability and is the main localization involved with their molecular role. The medians of percent tumor cells stained (range) were: Ki-67 proliferative index 26.7% (1–76%), Met 1.1% (0–60%), E-cadherin 70.8% (10–95%), vimentin 0% (0–15%), Snail 0.3% (0–70%) (Fig. 1). 3.2. Association of adenocarcinoma growth patterns with molecular markers of response to EGFR TKIs We previously reported the molecular markers associated with clinical benefit to EGFR TKI therapy in this trial [13]. The six cases with EGFR TKD mutations were ADC. EGFR TKD mutations were
present in a significantly higher proportion of tumors with nonmucinous lepidic predominant histology (56%) (Table 2). Seven out of 9 (78%) non-mucinous ADC with lepidic predominant pattern were also EGFR FISH+ but EGFR amplification was found mainly in ADC with papillary/acinar predominant patterns, and not in those with a lepidic predominant pattern. In contrast, KRAS mutations were found in all patterns of ADC. The two cases of mucinous ADC with lepidic pattern did not harbour aberrations in any of the molecular markers tested (Table 2). 3.3. Histopathological features in adenocarcinomas with EGFR TKD mutations Four of the six ADC with EGFR TKD mutation harboured an exon 19 deletion. Two of these fulfilled the criteria of radiological partial response (PR) according to the RECIST criteria [14]. The other two cases demonstrated 10% and 27% reduction in tumor
238
H. Lara-Guerra et al. / Lung Cancer 76 (2012) 235–241
Table 2 Correlation between predominantly histological patterns of adenocarcinoma and molecular markers. Predominant Pattern
N
Partial response
EGFR FISH+
EGFR Amp
KRAS Mutation
EGFR TKD Mutation
Total Mucinous lepidic Non-mucinous lepidic Acinar Papillary
27 2 9 11 5
4 (15%) 0 (0%) 2 (22%) 1 (9%) 1 (20%)
18 (69%) 0 (0%) 7(78%) 8 (73%) 3 (60%)
4 (15%) 0 (0%) 0 (0%) 2 (18%) 2 (40%)
6 (22%) 0 (0%) 1 (11%) 3 (27%) 2 (40%)
6 (22%) 0 (0%) 5 (56%)* 1 (9%) 0 (0%)
Amp, amplification; TKD, tyrosine kinase domain; EGFR, epidermal growth factor receptor; FISH+, high polysomy and amplification by fluorescence in situ hybridization. * p = 0.031.
diameter, which did not reach the RECIST criteria level for PR. All four tumors demonstrated extensive central fibrosis with marked loss of cellularity (Fig. 2, A–F). Residual viable tumor cells were present focally within the fibrous stroma and particularly in areas with marked lymphocytic infiltrates (Fig. 2, F and H). Residual tumor cells demonstrated low proliferative activity (Ki-67 staining) in the fibrous areas (Fig. 2C) and high activity in lymphocyte rich areas (Fig. 2I). One ADC with radiological PR and harbouring an exon 21 L858R mutation demonstrated lepidic predominant histology (Fig. 2J) with tumor cells showing a low cuboidal appearance (Fig. 2K). This tumor also showed a focal area of collapse with alveolar hemorrhage. A tumor with exon 21 L833V/H835L mutation did not respond to gefitinib treatment; this tumor also showed a non-mucinous lepidic predominant histology but with marked lymphocytic infiltration of the alveolar interstitium (Fig. 2L). However, lymphocytic aggregate or inflammatory response was not correlated with clinical response to changes in tumor diameter assessed by CT scan (Table 3).
3.4. Association of EGFR TKD mutant adenocarcinomas with histopathological features and IHC markers Although EGFR TKD mutant ADC showed extensive fibrotic changes (mean: 32.8% of tumor area), this was not significantly different from EGFR wild type ADC (19.9%) or non-ADC tumors (7.5%) (Fig. 3). In contrast, EGFR TKD mutant ADC cases demonstrated significantly lower cellularity (mean: 24.2% of tumor area) and Ki-67 proliferative index (mean: 4.6%) compared to EGFR wild type ADC (cellularity: 58.6%, p = 0.01; Ki-67: 31.4%, p = 0.002) and non-ADC tumors (cellularity: 55%, p = 0.026; Ki-67: 49.8%, p = 0.001). Met membranous staining levels were low in all groups, with no significant differences detected between ADC with EGFR TKD mutation (mean 2.5% of tumor cells), EGFR wild type (7.9%) or non-ADC (1.8%) tumors (Fig. 3). In contrast, E-cadherin levels were significantly higher in EGFR TKD mutant ADC (70.4% of tumor cells) compared to non-ADC tumors (44.4%, p = 0.026), but were similar when compared to EGFR wild type ADC (64%) (Fig. 3). Vimentin
Fig. 2. Histology of four tumors associated with partial response to gefitinib. (A)–(C) An Exon 19 L747 P752 deletion tumor shows fibrosis with extensive loss of tumor cells (A) and focal residual viable tumor cells (B) with low proliferative activity (C). (D)–(F) An Exon 19 E746 S750insV tumor shows focal intense lymphocytic infiltrates (D), prominent fibrosis (E) and focal residual viable tumor cells (F). (G)–(I) An EGFR wild type but amplified tumor shows large areas of tumor cell loss and fibrosis (G) and areas with marked lymphocytic infiltration (G and H) and foci of residual highly proliferative tumor cells (I) within. (J)-(K) An exon 21 L858R mutant tumor shows a predominant lepidic growth pattern (J) with low cuboidal tumor cells growing along the pre-existing alveolar framework (K). (L) An exon 21 L833V/H835L tumor not responsive to gefitinib shows a predominantly lepidic pattern and prominent stromal lymphocytic infiltrate.
H. Lara-Guerra et al. / Lung Cancer 76 (2012) 235–241
239
Fig. 3. Association between histopathological features and immunohistochemical markers with EGFR tyrosine kinase domain genotypes. Values on Y-axis represent distribution of averaged percentages of features/markers observed in each tumor assessed. In the box plots, top represents the 75th percentile (upper quartile), bottom the 25th percentile (lower quartile), and the middle line the 50th percentile (median). Whiskers represent the highest and lowest values excluding outliers. Outliers (values 1.5 or more times the interquartile range) are represented by circles beyond the whiskers. The p values were assessed by Wilcoxon–Mann–Whitney 2-samples. For Met, vimentin and Snail, the log 10 of percentages were shown, due to their low values.
240
H. Lara-Guerra et al. / Lung Cancer 76 (2012) 235–241
Table 3 Correlation of histopathological features and molecular markers with percent change in maximum tumor diameter on CT scans before and after gefitinib treatment. Characteristic
R2
ˇ (SE)
p
Predominantly lepidic Predominantly acinar Predominantly papillary Tumor grade Inflammatory response Lymphocytic aggregate Pathologic tumor diameter (cm) Fibrosis (%) Necrosis (%) Cellularity (%) Ki-67 index Met membranous staining E-cadherin membranous staining Vimentin cytoplasmic staining Snail nuclear staining
0.200 0.013 0.044 0.166 0.005 0.05 0.041 0.07 0.109 0.217 0.152 0.024 0.057 0 0.047
−4.588 (1.6) 1.076 (1.65) 3.035 (2.46) 9.684 (3.77) 1.657 (3.98) −9.964 (7.53) 2.748 (2.32) −0.197 (0.125) 0.904 (0.45) 0.31 (0.102) 0.289 (0.119) 0.242 (0.27) −0.17 (0.12) 0.005 (1.11) 0.328 (0.258)
0.007 0.52 0.227 0.015 0.68 0.195 0.244 0.125 0.053 0.005 0.02 0.375 0.167 0.997 0.213
R2 , coefficient of determination (proportion of the variation explained by the model); , standardized coefficient (Relative importance of the contribution of the predictor to the model. A negative value means larger the predictor, larger tumor reduction; positive value means larger the predictor, larger tumor growth); SE, standard error.
and Snail levels were low in all groups with the only significant difference observed between Snail in EGFR TKD mutant ADC (0.6%) compared to non-ADC tumors (2.7%) due to a single outlier in the last group (70%). 3.5. Correlation of histopathological features and IHC markers with clinical response Among the histopathological features, gefitinib response assessed by radiologic tumor diameter reduction was significantly associated with non-mucinous lepidic predominant pattern, higher tumor grade and lower tumor cellularity (Table 3). Among IHC markers, only low proliferative index (Ki-67) was significantly associated with radiologic response to pre-operative gefitinib therapy. Radiologic response did not correlate significantly with either acinar or papillary growth patterns in ADC, extent of tumor fibrosis, inflammatory response or tumor necrosis. 4. Discussion It is well established that EGFR TKI therapy can induce dramatic tumor shrinkage in a NSCLC subpopulation mainly defined by the presence of EGFR TKD mutations, but the histopathological correlates of this response have not been previously described. This is because EGFR TKIs have been used primarily to treat advanced NSCLC patients in whom histopathological evaluation of treated tumors is not feasible since surgical resection is not part of their standard of care. We recently conducted a neoadjuvant trial to assess gefitinib-induced clinical tumor response, drug toxicity, and selected clinical and molecular predictive markers of response in early NSCLC [13]. This study allowed us to assess the histopathological correlates of response in early NSCLC after EGFR TKI therapy. Our results demonstrate that significant radiologic tumor response is observed mainly in non-mucinous ADC with a lepidic-predominant growth pattern, which is also most often found in tumors with EGFR TKD mutations. Radiological tumor reduction was associated with a decrease in tumor cellularity and proliferation (Ki-67 index), but not with other factors evaluated, including acinar or papillary histology, inflammatory infiltration, fibrosis, and markers of EMT phenotype (Met/hepatocyte growth factor receptor, E-cadherin, vimentin and Snail).
Previous reports have documented the histological features found in ADC that are more commonly associated with EGFR TKD mutations in untreated tumors. These include low grade (well/moderate) and tumors with prominent lepidic (previously called BAC pattern) and/or papillary or micropapillary growth patterns [18–21]. These histological patterns have also been associated with a higher response rate to EGFR TKI therapy [5,6,22]. However, other reports have not found significant associations of these or other specific histological patterns with EGFR TKD mutations [23–25]. In our small series, 5 of 6 (83%) EGFR TKD mutant ADC were non-mucinous ADC with a predominant lepidic pattern, while the sixth tumor had a predominant acinar pattern. With one-quarter of our cohort being ADC with lepidic predominant pattern, the prevalence of EGFR TKD mutations in this ADC subtype accounts for their association with clinical response. One of the most important aspects of this report is the description of histopathological changes in tumors showing significant clinical response to gefitinib therapy. These tumors were characterized by extensive fibrosis and loss of tumor cells. However, there was no significant association between the extent of fibrotic changes and clinical response. This discrepancy is likely due to the difficulty in distinguishing between treatment and non-treatment related fibrosis, as focal fibrosis occurs commonly during cancer development [26]. Response to gefitinib therapy was, however, significantly associated with low tumor cellularity and a low proliferative index. Not surprisingly, significantly lower tumor cellularity and proliferative index were also found in ADC harbouring EGFR TKD mutations. These findings are in keeping with previous reports indicating that EGFR TKI therapy reduces cellular proliferation in advanced NSCLC [27,28]. Because EMT is a potential marker of resistance to EGFR TKI therapy, we assessed the expression of various markers of EMT in our cohort. Met gene amplification has been identified as potential predictor of resistance to EGFR TKI in different advanced NSCLC cohorts [29], but in our cohort of early NSCLC, we found uniformly low expression of Met, with no association of Met protein expression with radiologic response to therapy. E-cadherin expression was significantly higher in EGFR TKD mutant ADC compared with nonADC cases but not when compared with EGFR TKD wild-type ADC. Taken together with the lack of differences in expression of other EMT markers, this suggests variability of expression between histologic NSCLC subtypes, rather than variation based on genotypic differences between ADC tumors. Despite limited by a lack of corresponding pre-treatment data, this study is to the best of our knowledge the first to report the specific histopathological findings in NSCLC, particularly in EGFR TKD mutant ADC, demonstrating clinical response to treatment with an EGFR TKI. While the original trial was designed for hypothesis generating and the exploratory findings reported here require further validation, this assessment was enabled by the unique design of our neoadjuvant study, suggesting that similar designs may offer important opportunities for future studies with novel drugs aiming to correlate clinical response, histopathological changes, biomarker expression, and functional imaging techniques. This will lead to a better understanding of the tumor response to targeted therapies.
Author contributions HLG, CTC, NBL, TKW, and MST were responsible for the concept and design of the study. HLG, CTC, JS, MP, DMH, and MST were responsible for data acquisition, analysis and interpretation of results. HLG and MST were in charge of article drafting. All authors were involved in revising the article critically and approved its final version.
H. Lara-Guerra et al. / Lung Cancer 76 (2012) 235–241
Conflict of interest statement No conflicts of interest are declared from authors. The clinical trial from which samples analyzed are derived was supported in part by the Princess Margaret Hospital/University Health Network through Junior Investigator Awards (NBL, TKW), and in part and in part by AstraZeneca Canada (Mississauga, ON). AstraZeneca Canada was not involved in the design of the study; collection, analysis or interpretation of data; nor in the writing process or in the decision to submit for publication of this article. Acknowledgments Dr. Lara-Guerra was supported by fellowship from the Canadian Institute for Health Research (CIHR). This work was supported by grant (07NOV-78) from the Ontario Institute of Cancer Research; Dr. Waddell is the R. Fraser Elliott Chair in Transplantation Research; Dr. Tsao is the Choksi Chair in Lung Cancer Translational Research. The NIFTEE study was supported in part by the Princess Margaret Hospital/University Health Network and in part by AstraZeneca Canada (Mississauga, ON). We thank James Ho, Jing Xu and Olga Ludkovski from the Applied Molecular Profiling Laboratory for their technical assistance in the molecular studies and Dr. Frances Shepherd for her support. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.lungcan.2011.10.020. References [1] Shepherd FA, Rodrigues Pereira J, Ciuleanu T, Tan EH, Hirsh V, Thongprasert S, et al. Erlotinib in previously treated non-small-cell lung cancer. N Engl J Med 2005;353:123–32. [2] Mok TS, Wu YL, Thongprasert S, Yang CH, Chu DT, Saijo N, et al. Gefitinib or caboplatin-paclitaxel in pulmonary adenocarcinoma. N Engl J Med 2009;361:947–57. [3] Sakurada A, Shepherd FA, Tsao MS. Epidermal growth factor receptor tyrosine kinase inhibitors in lung cancer: impact of primary or secondary mutations. Clin Lung Cancer 2006;7:138–44. [4] Ellis PM, Morzycki W, Melosky B, Butts C, Hirsh V, Krasnoshtein F, et al. The role of the epidermal growth factor receptor tyrosine kinase inhibitors as therapy for advanced, metastatic, and recurrent non-small-cell lung cancer: a Canadian national consensus statement. Curr Oncol 2009;16:27–48. [5] Miller VA, Kris MG, Shah N, Patel J, Azzoli C, Gomez J, et al. Bronchioloalveolar pathologic subtype and smoking history predict sensitivity to gefitinib in advanced non-small-cell lung cancer. J Clin Oncol 2004;22:1103–9. [6] Miller VA, Riely GJ, Zakowski MF, Li AR, Patel JD, Heelan RT, et al. Molecular characteristics of bronchioloalveolar carcinoma and adenocarcinoma, bronchioloalveolar carcinoma subtype, predict response to erlotinib. J Clin Oncol 2008;26:1472–8. [7] Zhu CQ, da Cunha Santos G, Ding K, Sakurada A, Cutz JC, Liu N, et al. Role of KRAS and EGFR as biomarkers of response to erlotinib in National Cancer Institute of Canada Clinical Trials Group Study BR.21. J Clin Oncol 2008;26:4268–75. [8] Shepherd FA, Tsao MS. Unraveling the mystery of prognostic and predictive factors in epidermal growth factor receptor therapy. J Clin Oncol 2006;24:1219–20. [9] Yauch RL, Januario T, Eberhard DA, Cavet G, Zhu W, Fu L, et al. Epithelial versus mesenchymal phenotype determines in vitro sensitivity and predicts clinical activity of erlotinib in lung cancer patients. Clin Cancer Res 2005;11:8686–98.
241
[10] Miyanaga A, Gemma A, Ando M, Kosaihira S, Noro R, Minegishi Y, et al. Ecadherin expression and epidermal growth factor receptor mutation status predict outcome in non-small cell lung cancer patients treated with gefitinib. Oncol Rep 2008;19:377–83. [11] Zucali PA, Ruiz MG, Giovannetti E, Destro A, Varella-Garcia M, Floor K, et al. Role of cMET expression in non-small-cell lung cancer patients treated with EGFR tyrosine kinase inhibitors. Ann Oncol 2008;19:1605–12. [12] Cappuzzo F, Jänne PA, Skokan M, Finocchiaro G, Rossi E, Ligorio C, et al. MET increased gene copy number and primary resistance to gefitinib therapy in non-small-cell lung cancer patients. Ann Oncol 2009;20:298–304. [13] Lara-Guerra H, Waddell TK, Salvarrey MA, Joshua AM, Chung CT, Paul N, et al. Phase II study of preoperative gefitinib in clinical stage I non-small-cell lung cancer. J Clin Oncol 2009;27:6229–36. [14] Therasse P, Arbuck SG, Eisenhauer EA, Wanders J, Kaplan RS, Rubinstein L, et al. New guidelines to evaluate the response to treatment in solid tumors European Organization for Research and Treatment of Cancer, National Cancer Institute of the United States, National Cancer Institute of Canada. J Natl Cancer Inst 2000;92:205–16. [15] Union Internacional Contra el Cancer. In: Wittekind C, Sobin LH, editors. TNM Classification of Malignant Tumors. 7th edition Wiley-Liss; 2009. [16] World Health Organization/International Agency for Research on Cancer. In: Brambilla E, Muller-Hermelink HK, Harris CC, Travis WD, editors. World Health Organization classification of tumors: pathology and genetics of tumors of the lung, pleura, thymus and heart. IARC Press; 2004. [17] Travis WD, Brambilla E, Noguchi M, Nicholson AG, Geisinger KR, Yatabe Y, et al. International Association for the Study of Lung Cancer/American Thoracic Society/European Respiratory Society International multidisciplinary classification of lung adenocarcinoma. J Thoracic Oncol 2011;6:244–85. [18] Motoi N, Szoke J, Riely GJ, Seshan VE, Kris MG, Rusch VW, 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 2008;32:810–27. [19] Haneda H, Sasaki H, Shimizu S, Endo K, Suzuki E, Yukiue H, et al. Epidermal growth factor receptor gene mutation defines distinct subsets among small adenocarcinomas of the lung. Lung Cancer 2006;52:47–52. [20] Matsumoto S, Iwakawa R, Kohno T, Suzuki K, Matsuno Y, Yamamoto S, et al. Frequent EGFR mutations in noninvasive bronchioloalveolar carcinoma. Int J Cancer 2006;118:2498–504. [21] Ninomiya H, Hiramatsu M, Inamura K, Nomura K, Okui M, Miyoshi T, et al. Correlation between morphology and EGFR mutations in lung adenocarcinomas: significance of the micropapillary pattern and the hobnail cell type. Lung Cancer 2009;63:235–40. [22] Kim YH, Ishii G, Goto K, Nagai K, Tsuta K, Shiono S, et al. Dominant papillary subtype is a significant predictor of the response to gefitinib in adenocarcinoma of the lung. Clin Cancer Res 2004;10:7311–7. [23] Chantranuwat C, Sriuranpong V, Huapai N, Chalermchai T, Leungtaweeboon K, Voravud N, et al. Histopathologic characteristics of pulmonary adenocarcinomas with and without EGFR mutation. J Med Assoc Thai 2005;88:S322–9. [24] Shigematsu H, Lin L, Takahashi T, Nomura M, Suzuki M, Wistuba II, et al. Clinical and biological features associated with epidermal growth factor receptor gene mutations in lung cancers. J Natl Cancer Inst 2005;97:339–46. [25] Dacic S, Kelly L, Shuai Y, Nikiforova MN. miRNA expression profiling of lung adenocarcinomas: correlation with mutational status. Mod Pathol 2010;23:159–68. [26] Goldstein NS, Mani A, Chmielewski G, Welsh R, Pursel S. Prognostic factors in T1 NO MO adenocarcinomas and bronchioloalveolar carcinomas of the lung. Am J Clin Pathol 1999;112:391–402. [27] Baselga J, Rischin D, Ranson M, Calvert H, Raymond E, Kieback DG, et al. Phase I safety, pharmacokinetic, and pharmacodynamic trial of ZD1839, a selective oral epidermal growth factor receptor tyrosine kinase inhibitor, in patients with five selected solid tumor types. J Clin Oncol 2002;20:4292–302. [28] Felip E, Rojo F, Reck M, Heller A, Klughammer B, Sala G, et al. A phase II pharmacodynamic study of erlotinib in patients with advanced non-small cell lung cancer previously treated with platinum-based chemotherapy. Clin Cancer Res 2008;14:3867–74. [29] Bean J, Brennan C, Shih JY, Riely G, Viale A, Wang L, et al. MET amplification occurs with or without T790M mutations in EGFR mutant lung tumors with acquired resistance to gefitinib or erlotinib. Proc Natl Acad Sci USA 2007;104:20932–7.