Diseases of of the the Esophagus Esophagus (2014) (2015) ••, 28,••–•• 750–756 Diseases DOI: 10.1111/dote.12248 10.1111/dote.12248 DOI:

Original article

Clinicopathological and prognostic significance of epidermal growth factor receptor overexpression in patients with esophageal adenocarcinoma: a meta-analysis Y.-M. Guo,1 W.-W. Yu,2 M. Zhu,3 C.-Y. Guo1 Department of Gastroenterology, Shanghai Tenth People’s Hospital of Tongji University, 2Department of Radiation Oncology, Shanghai Jiao Tong University Affiliated Sixth people’s Hospital, and 3Department of Biostatistics, School of Public Health, Fudan University, Shanghai, China

1

SUMMARY. The prognostic significance of epidermal growth factor receptor (EGFR) overexpression in patients with esophageal adenocarcinoma (EAC) remains controversial. Eligible studies that investigated the association between survival in EAC and the expression status of EGFR were identified by an electronic search of PubMed, EMBASE, and ISI Web of Science. A meta-analysis was performed to clarify the impact of EGFR overexpression on clinicopathological parameters or overall survival (OS) in EAC. A total of seven studies including 1028 patients were subjected to the final analysis. The overall results suggested that overexpression of EGFR was significantly correlated with not only the depth of invasion, lymph node status, and tumors stage of EAC, with a pooled odds ratio of 2.99 (95% confidence interval [CI]: 1.07–8.35; Z = 2.09; P = 0.037), 3.05 (95% CI: 1.77–5.27; Z = 4.00; P < 0.001), and 5.37 (95% CI: 2.49–11.57; Z = 4.29; P < 0.001), respectively, but also the poorer OS with a pooled hazard ratio of 2.20 (95% CI: 1.47–3.31; Z = 3.79; P < 0.001). Overexpression of EGFR correlates with not only the clinicopathological features, but also the worse OS, and it might be useful as a predictive biomarker in clinical practice, yet the clinicopathological and prognostic role of EGFR in EAC still needs further confirmation by well-designed prospective studies. KEY WORDS: epidermal growth factor receptor, esophageal adenocarcinoma, meta-analysis, prognosis.

INTRODUCTION The incidence of esophageal adenocarcinoma (EAC) has been increasing substantially during the last few decades and will continue to rise over the coming decades. Moreover, EAC has become the dominant histological type of esophageal cancer in western society.1 EAC is characterized by some striking features: lacking of preventive measures, highly aggressive, and worse prognosis than that for most other types of tumor.2 Despite growing knowledge of its

Address correspondence to: Dr Chuan-Yong Guo, PhD, Department of Gastroenterology, Shanghai Tenth People’s Hospital of Tongji University, 301 Yanchang Road, Shanghai 200072, China. Email: [email protected] Yan-Mei Guo and Wei-Wei Yu – The first two authors contributed equally to this work, and both should be considered first author. Conflicts of interest and informed consent: The authors have declared that no conflicts of interests exist.

7502014 International Society for Diseases of the Esophagus ©

biology, no molecular biomarkers are currently used in routine clinical practice to determine prognosis. Epidermal growth factor receptor (EGFR), also known as HER-1 or erbB-1, is a tyrosine kinase receptor. Upon activation by specific ligands, it can trigger a cascade of biophysiological signaling reaction related to the behavior of malignant cells, including cell proliferation, apoptosis arrest, invasion, and angiogenesis, such as endothelial proliferation or migration.3,4 Up to now, overexpression of EGFR has been identified in a wide variety of solid human tumors, such as non-small cell lung cancer, colon cancer, and hepatocellular carcinoma, conferring a poor prognosis.5–7 EGFR is also a key mediator in the progression from normal esophageal tissue to Barrett’s esophagus and finally to EAC.8 Increased expression of EGFR has been detected in EAC, particularly in poorly differentiated tumors.9,10 A majority of studies have found a significant association of EGFR C 2014 International Society for Diseases of the Esophagus V 1

2

Diseases of the Esophagus

overexpression with lymph node involvement, metastasis occurrence, and poorer overall survival (OS) in EAC,10–17 but conflicting results have been reported from various laboratories, and the clinical importance of EGFR overexpression remains unsettled. We therefore carried out a meta-analysis of published studies to clarify this.

MATERIALS AND METHODS Search strategy and selection criteria Systematic computerized searches of the PubMed and Medline databases (up to July 5, 2013) were performed. The following MESH headings, key words, and text words were used: (i) esophagus OR esophageal and adenocarcinoma, and (ii) epidermal growth factor receptor OR EGFR OR ErbB1. The reference lists of retrieved articles and previous narrative reviews were scanned for other potentially relevant articles; investigators of eligible studies were contacted for supplement of additional data relevant to meta-analysis. When multiple articles pertained to overlapping populations of patients, only the newest, largest, or most informative single article was selected. Studies were included in the meta-analysis if they met the following criteria: (i) expression of EGFR was evaluated in the primary EAC tissue as opposed to sera or metastatic tissue or in tissue adjacent to the tumor; (ii) the expression of EGFR was measured by immunohistochemistry (IHC) of protein only; (iii) analysis of the association between EGFR expression and clinicopathological parameters or OS; (iv) studies were published as a full paper in English; and (v) median follow-up time exceeded 2 years in studies for survival analysis. Reviews and conference abstracts were excluded because of limited data for evaluation. We created two major groups according to study objective. The first one included studies investigating the association between EGFR overexpression and clinicopathological parameters, which included depth of invasion (T), lymph node status (N), tumor stage, and tumor differentiation grade. The second included studies analyzing the association between the overexpression of EGFR and OS in patients with EAC. Data extraction Data were carefully extracted independently by two investigators (Guo and Yu), and controversies were resolved by discussion according to the criteria listed above. If they could not reach an agreement, an expert was invited to the discussion. The following information was extracted from each study: first author, year of publication, number of eligible patients, stage and tumor differentiation grade of C 2014 International Society for Diseases of the Esophagus V

The role of EGFR overexpression in EAC 751

enrolled patients, antibody for IHC, definition of EGFR overexpression, and OS results. Statistical analysis The impact of EGFR overexpression on clinicopathological parameters was estimated for each study by odds ratio (OR), with its 95% confidence interval (CI), respectively. According to clinical characteristics, well and moderate differentiation were combined; T1 and T2 were combined; T3 and T4 were combined; stage I and stage II were combined; and stage III and stage IV were combined. For the quantitative aggregation of OS results, the impact of EGFR overexpression on OS was estimated for each study by the hazard ratio (HR), with its 95% CI, respectively, according to the methods described by Parmar et al.18 When HRs and their 95% CIs were described in text or tables, we obtained these values directly. When these statistical variables were not given explicitly in an article, they were calculated from available numerical data using methods reported by Parmar et al.18 In summary, HRs and CIs were calculated from differences between numbers of observed and expected events concerning the primary end-point of length of survival. When the only available data were in the form of graphical representations, they were calculated from Kaplan–Meier survival curves; in these cases, after dividing the time axis into non-overlapping intervals, log HR and its variance for each interval were calculated. These estimated values were combined in a stratified manner to obtain the overall HR and 95% CI. The Kaplan– Meier survival curves were read by two persons using Engauge Digitizer version 4.1 (free software downloaded from http://sourceforge.net) independently to reduce inaccuracy in extracted survival rates. All the calculations for HRs were performed by calculations spreadsheet in Microsoft Excel (Microsoft Corporation, Redmond, WA, USA) reported by Tierney et al.19 By convention, an observed HR of >1 implied a worse prognosis for the group with EGFR overexpression. The impact of EGFR on survival was considered to be statistically significant if the 95% CI did not overlap with 1. For these analyses, a P-value 0.1 indicates a lack of heterogeneity among studies, so the fixed-effects model was used to combine OR or HR. Otherwise, the random-effects model was used.22 Evidence of publication bias was evaluated by using the Begg and Egger tests.23,24. For these analyses, a P-value below 0.05 was considered representative of statistically significant publication bias. All statistical analyses were performed by STATA 10.0 software (Stata Corporation, College Station, TX, USA). © 2014 International Society for Diseases of the Esophagus

752 Diseases of the Esophagus

The role of EGFR overexpression in EAC

3

Table 1 Main characteristics and results of the eligible studies Gender First author

Year of publication

Number of patients

Stage

M

F

Clinicopathological features

EGFR effect on survival

Paterson et al.13 Navarini et al.14 Goh et al.15 Marx et al.16 Wang et al.17 Wilkinson et al.10 Yacoub et al.25

2013 2012 2011 2010 2007 2004 1997

367 37 359 112 103 38 20

I–IV I–IV I–IV I–IV I–IV I–IV I–III

293 31 — 92 93 34 18

74 6 — 20 10 4 2

Differentiation, T, N Differentiation, T, N, stage — Differentiation, T, N Differentiation, T, N, stage Differentiation, N, stage Stage

NS S (worse) S (worse) S (worse) NS — S (worse)

EGFR, epidermal growth factor receptor; N, lymph node status; NS, non-significant; S, significant; T, depth of invasion.

RESULTS

author, 160 patients (15.6%) in this meta-analysis had EGFR overexpression, with a range of 10–45.5%.

Study characteristics The characteristics of the eligible studies are summarized in Table 1. A total of seven studies10,13–17,25 published from 1997 to 2013 met the criteria for this review. A total of 1028 patients were subjected to the final analysis (mean: 146 per study; range: 20–363). All of the studies were based on the data of retrospective analysis. Six studies investigated the correlation of EGFR overexpression with clinicopathological parameters. Among the six studies dealing with OS, a significant association between EGFR overexpression and OS was found in four, with the worse survival. The immunohistochemical technique for EGFR expression detection was summarized in Table 2. The expression of EGFR was measured by IHC in all publications, but the techniques used varied widely between studies, with a wide range of dilutions (from 1 : 10 to 1 : 200). According to the cut-off values for EGFR overexpression, as defined by each study’s

EGFR overexpression and clinicopathological characteristics Four studies investigated the association of EGFR overexpression with tumor depth of invasion (T). We observed a statistically significant effect of EGFR overexpression on depth of invasion with a pooled OR of 2.99 (95% CI: 1.07–8.35; Z = 2.09; P = 0.037), with a heterogeneity (P = 0.085), so the random model was slected (Fig. 1, Table 3). Five studies assessed the relationship of EGFR overexpression to lymph node status (N). We observed a statistically significant effect of EGFR on lymph node status with OR 3.05 (95% CI: 1.77–5.27; Z = 4.00; P < 0.001), without heterogeneity (P = 0.517) (Fig. 2, Table 3), Four eligible studies analyzed the association of EGFR overexpression and tumor stages. We observed a statistically significant impact of EGFR overexpression on tumor stage of EAC with OR 5.37

Table 2 Immunohistochemical technique used in studies Definition of EGFR overexpression

EGFR overexpression (%)

Intensity of EGFR membrane staining

≥score 2

37 (10.2)

Membrane tumor cell stained in brown color Intensity of EGFR membrane staining

>1

15 (45.5)

≥score 2

36 (10.0)

Combination of staining intensity and percentage of positive cells Percentage of positive cells

≥score 2

16 (14.3)

>5%

33 (32.0)

1 μg/mL

Intensity of EGFR membrane staining

≥score 1

15 (39.5)

1/75

Percentage of positive cells

≥25%

First author

Antibody and source

Dilution

Counting method

Paterson et al.13

NCL-L-EGFR antibody (Novocastra, Newcastle upon Tyne, UK) Anti-EGFR MAb, clone H11 (Dako, Ely, UK) Anti-EGFR antibody (Novocastra; Leica, Milton Keynes, UK) Anti-EGFR MAb, clone 2-18C9 (Dako, Glostrup, Denmark) Anti-EGFR MAb (clone 31G7; Zymed Laboratories, Inc., San Francisco, CA, USA) Anti-EGFR MAb (clone 31G7; Zymed Laboratories, Inc.) Anti-EGFR MAb (Biogenex, San Ramon, CA, USA)

1/10 1/200

Navarini et al.14 Goh et al.15 Marx et al.16 Wang et al.17

Wilkinson et al.10 Yacoub et al.25

1/10 — —

8 (40.0)

EGFR, epidermal growth factor receptor. © 2014 International Society for Diseases of the Esophagus

C 2014 International Society for Diseases of the Esophagus V

4

The role of EGFR overexpression in EAC 753

Diseases of the Esophagus

Fig. 1 Forest plot showing the association between epidermal growth factor receptor (EGFR) overexpression and depth of invasion. CI, confidence interval; OR, odds ratio.

(95% CI: 2.49–11.57; Z = 4.29; P < 0.001), without heterogeneity (P = 0.272) (Fig. 3, Table 3). Four eligible studies investigated the relationship between the EGFR overexpression and the tumor differentiation grade in EAC, the test of heterogeneity was significant (P = 0.022) and hence, the OR was calculated using a random-effect model. However, we did not find statistical significance between EGFR overexpression with the tumor differentiation grade, with OR 1.99 (95% CI: 0.87–4.61; Z = 1.62; P = 0.105) (Table 3).

The correlation of EGFR overexpression and OS The meta-analysis was conducted on six studies dealing with OS and EGFR overexpression, including a total of 990 patients. As there is heterogeneity (P = 0.016), random effect was selected. The pooled HR was 2.20 (95% CI: 1.47–3.31; Z = 3.79; P < 0.001), which illustrated that overexpression of EGFR was significantly correlated with poor prognosis (Fig. 4). Begg’s funnel plot and Egger’s test were performed to assess the publication bias of literatures. The shapes

Table 3 OR value and 95% CI of clinicopathological parameters in gastric cancer Heterogeneity Clinicopathological features

Number of studies

Number of patients

Pooled OR

Depth of invasion (T) Lymph node status (N) Stage Differentiation

4 5 4 5

600 608 187 626

2.99 3.05 5.37 1.99

95% CI

Overall effect Z

P-value for OR

I2 (%)

P-value

1.07–8.35 1.77–5.27 2.49–11.57 0.86–4.61

2.09 4.00 4.29 1.62

0.037