IJC International Journal of Cancer
Tumor-associated macrophages correlate with response to epidermal growth factor receptor-tyrosine kinase inhibitors in advanced non-small cell lung cancer Fu-Tsai Chung1,2,3*, Kang-Yun Lee1,3*, Chih-Wei Wang4, Chih-Chen Heh1, Yao-Fei Chan1, Huan-Wu Chen5, Chih-Hsi Kuo1,3, Po-Hao Feng1,2,3, Ting-Yu Lin1,3, Chun-Hua Wang1,3, Chun-Liang Chou1,3, Hao-Cheng Chen1,3, Shu-Min Lin1,3 and Han-Pin Kuo1,3 1
Pulmonary Disease Research Center, Chang Gung Memorial Hospital, Linkou, Taiwan, Republic of China Graduate Institute of Clinical Medical Sciences, College of Medicine, Chang Gung University, Taoyuan, Taiwan, Republic of China 3 Department of Internal Medicine, Chang Gung Memorial Hospital at Linkou, Chang Gung University, College of Medicine, Taipei, Taiwan, Republic of China 4 Department of Pathology, Chang Gung Memorial Hospital at Linkou, Chang Gung University, College of Medicine, Taipei, Taiwan, Republic of China 5 Department of Radiology, Chang Gung Memorial Hospital at Linkou, Chang Gung University, College of Medicine, Taipei, Taiwan, Republic of China
Our study investigated whether tumor-associated macrophages (TAMs) in advanced non-small cell lung cancer (NSCLC) are related to treatment response to epidermal growth factor receptor-tyrosine kinase inhibitors (EGFR-TKIs) and may be a predictor of survival. Of 206 advanced NSCLC patients treated (first-line) with an EGFR-TKI at the study hospital from 2006 to 2009, 107 with adequate specimens for assessing CD68 immunohistochemistry as a marker of TAMs were assessed. After EGFR-TKI treatment, response was observed in 55 (51%) patients, and the median follow-up period was 13.5 months. Most TAMs were located in the tumor stroma (>95%) and positively costained with the M2 marker CD163. TAM counts were significantly higher in patients with progressive disease than in those without (p < 0.0001), a trend that remained in patients with known EGFR mutation status (n 5 59) and those with wild-type EGFR (n 5 20). High TAM counts, among other factors (e.g., wild-type EGFR), were significantly related to poor progression-free survival (PFS) and overall survival (OS) (all p < 0.0001 for TAMs). Multivariate Cox analyses showed that high TAM counts and EGFR mutations were both independent factors associated with PFS [odds ratio (OR), 8.0; 95% confidence interval (CI), 2.87–22.4; p 5 0.0001 and OR, 0.03; 95% CI, 0.003– 0.31; p 5 0.003, respectively] and OS (OR, 2.641; 95% CI, 1.08–6.5; p 5 0.03 and OR, 0.14; 95% CI, 0.03–0.56; p 5 0.006, respectively). TAMs are related to treatment response irrespective of EGFR mutation and can independently predict survival in advanced NSCLC treated with an EGFR-TKI.
Key words: TAM, EGFR-TKIs, response, outcome, advanced NSCLC Abbreviations: CI: conﬁdence interval; CR: complete response; EGFR-TKI: epidermal growth factor receptor-tyrosine kinase inhibitors; HGF: hepatocyte growth factor; HR: hazards ratio; IPASS: Iressa Pan-Asian Study; IQR: interquartile range; NSCLC: non-small cell lung cancer; OR: odds ratio; OS: overall survival; PCR: polymerase chain reaction; PD: progressive disease; PFS: progression-free survival; PR: partial response; PS: performance status; SD: stable disease; TAMs: tumor-associated macrophages; TKI: tyrosine kinase inhibitor Additional Supporting Information may be found in the online version of this article. Grant sponsor: Chang Gung Memorial Hospital; Grant numbers: CMRPG391211, CMRPG391221, CMRPG300181 *F.T.C. and K.Y.L. contributed equally to this work DOI: 10.1002/ijc.27403 History: Received 5 Aug 2011; Accepted 2 Dec 2011; Online 15 Dec 2011 Correspondence to: Han-Pin Kuo, Department of Thoracic Medicine, Chang Gung Memorial Hospital, 199 Tun Hwa N. Road, Taipei, Taiwan, Tel.: 3281200 ext. 8467, Fax: (886)-3-3272474, E-mail: [email protected]
C 2011 UICC Int. J. Cancer: 131, E227–E235 (2012) V
Introduction Epidermal growth factor receptor-tyrosine kinase inhibitors (EGFR-TKIs), e.g. geﬁtinib and erlotinib, have shown clinical activity against non-small cell lung cancer (NSCLC). However, phase III clinical trials have demonstrated that these drugs have comparable efﬁcacy to chemotherapy in previously treated NSCLC patients, with objective response rates of only 8.99.1%.1,2 There is compelling evidence that sensitivity to EGFR-TKIs correlates very strongly with the presence of somatic activating mutations in the EGFR kinase domain.3,4 Such mutations are more frequent in patients with characteristics previously shown to correlate with clinical response to EGFR-TKIs treatment: nonsmokers, women, individuals of Asian ethnic background and those with adenocarcinoma.4,5 The recent Iressa Pan–Asian study (IPASS) trial demonstrated the superiority of geﬁtinib to carboplatin–paclitaxel [in terms of progression-free survival (PFS)] for initial treatment of pulmonary adenocarcinoma among nonsmokers and former light smokers in East Asia.6 Previous subgroup analysis indicated that the presence of EGFR mutation in a tumor is a strong predictor of better outcome upon geﬁtinib
treatment, a conclusion supported by other phase III trials comparing geﬁtinib to platinum-doublet chemotherapy for adenocarcinoma patients selected on the basis of EGFR mutations.7,8 Nevertheless, in the IPASS trial, EGFR mutation data were evaluated in only 437 out of 1217 patients (35.9%). Controversy has also surrounded the predictive value of the data in unselected patients, such that 1020% of patients who had a partial response (PR) to geﬁtinib did not have detectable EGFR mutations.9 Similarly, in the BR.21 trial, EGFR mutational status did not show signiﬁcant association with responsiveness to erlotinib.10 Notably, the EGFR mutation incidence was only around 1723% in Caucasians10,11 and 4060% in East Asians.6,7,12 Therefore, identiﬁcation of a novel biomarker other than EGFR mutation for predicting the efﬁcacy of EGFR-TKIs will contribute to further individualizing NSCLC treatment. A tumor’s behavior is affected, not only by its intrinsic genetic/metabolic makeup but also by its microenvironment.13 Macrophages are a major component of immune cell inﬁltration in the tumor microenvironment. Such tumorassociated macrophages (TAMs) were originally proposed to play a role in antitumor immunity, but substantial clinical and experimental evidence support their tumor-supporting role in most cases.14 In established tumors, TAMs seem to have a skewed M2 phenotype, which is alternatively activated in contrast to the inﬂammatory M1 phenotype.15 However, recent gene proﬁling experiments revealed that TAMs often share features of both types, albeit with a shift to this immunoregulatory type.16,17 The prognostic value of TAMs in NSCLC has been evaluated in patients who have undergone surgery. Chen et al. reported that the density of TAMs negatively correlates with patient survival, which is associated with tumor angiogenesis18 and promotion of tumor invasiveness and matrix-degrading activity.19 TAMs have since been shown to differentially correlate with survival of NSCLC patients according to their microanatomic localization. As such, although stromal TAMs are frequently negatively associated with patient survival,20,21 tumor islet TAMs are consistently positively correlated with patient survival.21,22 Nevertheless, the fact that the majority of TAMs localize in the tumor stroma2 and express protumor M2 markers20 still supports a protumor role of TAMs in most NSCLC patients, as observed in earlier studies.18,19 Furthermore, this is supported by a recent report from Ohtaki et al.,23 who demonstrated an association between CD204-positive stromal macrophages and poor outcomes, indicating the protumor effect of stromal macrophages. Although there have been extensive studies on the prognostic value of TAMs, the potential use of TAMs as a predictor for treatment response, particularly to EGFR-TKIs, has not been evaluated. Thus, this retrospective study, taking advantage of samples from treated naı¨ve patients, aimed to investigate the correlation of TAMs with treatment response to EGFR-TKIs as ﬁrst-line therapy for advanced NSCLC, and
TAM and EGFR-TKIs in advanced NSCLC
their predictive value for PFS and overall survival (OS). These values, in the context of known or unknown EGFR mutation status, are also addressed.
Material and Methods Patients
All patients with advanced NSCLC (stage IIIb or IV) who had been treated with an EGFR-TKI (geﬁtinib or erlotinib) as ﬁrstline treatment at the Linkou Branch of Chang Gung Memorial Hospital between 2006 and 2009 were included in our study. The EGFR mutation status of most patients was unknown when they received the EGFR-TKI treatment. This retrospective study was reviewed and approved by the ethical review board of our hospital (IRB no. 100-0522B). The study involved patients treated from 2006 to 2009, when using EGFR mutation examination as a guide for the ﬁrst-line EGFR-TKI was not conclusive. All patients in our study were those who refused chemotherapy as ﬁrst-line therapy and asked instead for EGFR-TKI as ﬁrst-line treatment for advanced NSCLC. The decision of using an EGFR-TKI as the ﬁrst-line treatment was also a consensus by the clinician and was approved by team conference (including oncologists, pulmonologists, pathologists, radiologists, surgeons, radiation oncologists and nuclear medicine specialists). Sixty-seven patients were lost to follow-up or were excluded, and 32 patients did not have adequate pathologic specimens for study. Thus, our ﬁnal analysis included 107 patients (Supporting Information Fig. 1). The tumor response was evaluated using computed tomography according to the Response Evaluation Criteria in Solid Tumors, and was classiﬁed as complete response (CR), PR, stable disease (SD) or progressive disease (PD). Clinicopathologic information, including patient characteristics, treatment, clinical staging results, histologic subtype and survival data, was recorded. Sample preparation
All tissue specimens were ﬁxed in 10% buffered formalin and embedded in parafﬁn according to standard procedures. All of the tissues were ﬁxed immediately after biopsy, with the time from tissue acquisition to ﬁxation as short as possible. Serial sections (4-lm thickness) placed on positively charged slides (Menzel-Glaser, Germany) were used for immunohistochemistry and detection of EGFR mutation. Immunohistochemical staining
Tissue sections were cut, placed onto glass slides, de-waxed in xylene and alcohols and then washed with phosphate-buffered saline twice. Mouse anti-human macrophage CD68 monoclonal antibody (Clone PG-M1 from Dako, Glostrup, Denmark.) was used as a speciﬁc marker for macrophages. M1 and M2 polarization of TAMs was determined by CD68/ HLA-DR1 (clone TAL 1B5 from Abcam) and CD68/CD163 (clone 10D6 from Abcam) double staining, respectively. Immunostaining was performed according to the manufacturer’s instructions. C 2011 UICC Int. J. Cancer: 131, E227–E235 (2012) V
Figure 1. Immunohistochemical staining of TAMs. Representative images of TAMs stained with anti-CD68 (brown) [(a) low-power field and (b) high-power field], M1 macrophages double-stained with anti-CD68 (brown) and anti-HLA-DR (red) [(c) low-power field and (d) high-power field], and M2 macrophages double-stained with anti-CD68 (brown) and anti-CD163 (red) [(e) low-power field and (f) high-power field]. In contrast to the generally CD163-stained macrophages (e and f), only rare macrophages were HLA-DR-stained (c and d). T: tumor islet; S: tumor stroma.
Analysis and validation of immunostaining
Determination of EGFR mutations
Analysis of immunostaining was performed blindly with respect to the clinical outcome. The ﬁve most representative high-power ﬁelds (200 magniﬁed) per slide were manually selected using an Olympus BX50 microscope (Olympus, Southall, United Kingdom). The nuclear cells with positive CD68 staining were counted as TAMs. Analysis was carried out independently by two researchers, including a pathologist. The two sets of data were then compared to assess the reproducibility and, hence, the validity of the results. The median number of TAMs was 118 per high power ﬁeld (200), and thus this value was chosen as the cut-off level for determination between high and low TAM counts. Representative pictures of low and high TAM counts are shown in Figures 1a and 1b, respectively.
EGFR mutations were assessed using the EGFR PCR Kit (Qiagen, Valencia, CA) according to the manufacturer’s instructions. By combining two technologies, ARMSV and Scorpions, this kit enabled the detection of the most prevalent somatic mutations in the EGFR gene that are common in human cancers. Surgically resected tissue samples containing a tumor cell proportion of at least 10%, or biopsy samples containing a proportion of at least 30%, were determined as adequate. Tumor cell enrichment was achieved using Laser capture microdissection for those samples with