Prognostic significance of Human epidermal growth factor receptor-2 expression in patients with resectable gastric adenocarcinoma

Background The purpose of this study was to investigate the correlation between human epidermal growth factor receptor 2 (HER2) overexpression and clinicopathologic factors and overall survival rate in patients who underwent curative gastrectomy for gastric adenocarcinoma. Methods Among patients who underwent curative gastrectomy for gastric adenocarcinoma at Inje University Paik Hospital from January 2012 to December 2015, 782 patients underwent an immunohistochemical analysis to evaluate HER2 expression levels. Clinicopathologic records that were collected from a gastric cancer database were retrospectively reviewed to identify clinicopathologic factors and survival rates of the patients. Results HER2 overexpression was detected in 166 patients (21.2%). There was a statistically significant correlation between HER2 expression level and sex (p = 0.013), histologic differentiation (p < 0.001), Lauren classification (p < 0.001), and T pathologic stage (p = 0.022). There were no statistically significant relationships between HER2 expression level and overall 5-year survival rate (p = 0.775) and overall 5-year survival rate of gastric adenocarcinoma classified according to the TNM stage (stage I: p = 0.756, stage II: p = 0.571, stage III: p = 0.704). The HER2 expression level was not affected by the overall 5-year survival rate in the uni- and multivariate analyses. Conclusions In this study, the HER2 overexpression rate in gastric adenocarcinoma was 21.2% and was observed in well- and moderately differentiated types according to histologic differentiation, intestinal type according to the Lauren classification, male, and low T stage. There was no correlation between HER2 expression level and overall 5-year survival rate, and HER2 expression level was not associated with independent prognostic factors.


Introduction
In Korea, the detection rate of early gastric cancer has increased from 24.8 to 57.6% over the past 20 years due to the implementation of national checkups and the development of diagnostic techniques [1]. Furthermore, the overall 5-year survival rate of gastric cancer has increased from 42.8 to 74.4% due to the standardization of gastrectomy, increased operator expertise, and development of surgical tools [2]. However, the mortality rate due to gastric cancer remains high (16.7 per 100,000 patients) [3] because not all gastric cancers are diagnosed at early stages and various genetic mutations are present in gastric cancer, leading to various anti-cancer treatment responses. The overexpression or abnormal expression of tyrosine kinase receptors plays crucial roles in the survival and proliferation of gastric cancer cells. Based on this observation, receptor ligands, receptors, tyrosine kinase domains in the lower portions of the receptors, and signaling molecules have been targeted for treating cancer [4].
Human epidermal growth factor receptor 2 (HER2), a target used in breast cancer treatment, is the second member of the epidermal growth factor receptor (EGFR) family and is a type of receptor tyrosine kinase. It is a 185-kD glycoprotein that was discovered when the NEU oncogene was studied in rats. HER2 is encoded by the ERBB gene that is located on human chromosome 17, q21 [5]. HER2 is a prognostic factor for invasive breast cancer and shows gene amplifications in approximately 20% of breast cancers, resulting in its overexpression and providing a treatment target [6]. In addition to its significance in breast cancer, HER2 gene amplification and protein overexpression have been observed in esophageal, gastric, and colon cancers, and HER2 plays an important role in cancer cell proliferation, apoptosis, adhesion, angiogenesis, and aggressiveness [5].
The purpose of this study was to investigate HER2 overexpression by immunohistochemical (IHC) analysis in gastric adenocarcinoma tissues that were obtained by curative gastrectomy. In addition, we compared clinicopathologic factors and survival rates with HER2 expression level to identify factors associated with HER2 overexpression. Furthermore, HER2 overexpression was examined to determine its value as an independent prognostic factor.

Study participants
Among 810 patients who underwent curative gastrectomy to treat gastric adenocarcinoma at Inje University Paik Hospital, from January 2012 to December 2015, this study included 782 patients after excluding 28 patients. Twenty-eight patients who were excluded from our study underwent curative gastrectomy after endoscopic submucosal dissection and did not undergo IHC analysis because of the absence of residual lesion in the pathologic result. Fresh gastric adenocarcinoma tissue obtained during curative gastrectomy was created simultaneously with slides for the pathological stage as well as slides for IHC analysis. The clinicopathologic records were collected from a gastric cancer database and were analyzed retrospectively to identify pathologic features of the patients. The pathologic stage was defined by the 7th edition of the TNM classification system from the American Joint Committee on Cancer [24]. Patients with TNM stage Ib were treated with oral 5-fluorouracil, and patients with TNM stages II and III were treated with S-1 [25] and xelox (intravenous oxaliplatin and oral capecitabine) [26] as adjuvant chemotherapy.
This study was approved by the Institutional Review Board of the Inje University Paik Hospital (IRB No. 17-0104).

IHC methods
After the curative gastrectomy, gastric adenocarcinoma tissues were sent to the department of pathology to identify the gastric adenocarcinoma pathologic stage. In brief, 4-μm-thick samples were cut and slices were prepared on glass slides, to which aminoprophyltriethoxysilane (APEAS, Sigma, USA) was applied. The slices were deparaffinized twice in 100% xylene for 10 min; hydrated with 90%, 85%, 80%, 70%, 60%, and 50% alcohol for 10 min; and washed with distilled water. The 10-mM citrate buffer (pH 6.0) was heated at 750 W for 5 min in a microwave. The slides were then heated twice for 5 min to expose antigens and left at room temperature for 20 min to gradually lower the temperature. Subsequently, the slides were washed with Tris-buffered saline (TBS, 50 mM, pH 7.4), treated with 0.3% hydrogen peroxide for 15 min, and washed thrice with Tris-buffered solution. To block non-specific antigens in the tissues, the slides were treated with normal horse serum for 30 min. The primary antibody that was used was rabbit antihuman c-erbB-2 oncoprotein (DAKO, Denmark, 1:100), and the slides were incubated overnight in a chamber in a 4°C bath and washed three times with TBS. Subsequently, the slides were treated with a secondary antibody (Vector Elite kit, Vector Laboratories, USA) for 30 min and then reacted with avidin-biotin conjugate (ABC) reagent at room temperature for 45 min. After washing the specimens with TBS and Tris-HCl (pH 7.6) buffered solution, the slides were treated with diaminobenzidine tetrachloride (DAB, Sigma Chemicals, USA) for 2-3 min, washed with distilled water, and stained with 10% Mayer's hematoxylin.

Evaluation and classification of IHC
IHC was assessed according to the Hoffmann criteria [27]. Specimens with no staining or staining in less than 10% of all tumor cells were classified as 0, those with faint or hardly noticeable staining in 10% or more of the tumor cells were classified as 1+, those with mild to moderate staining in 10% or more of the tumor cells were classified as 2+, and those with moderate to strong staining in 10% or more of the tumor cells were classified as 3+ and defined as overexpression.

Statistical analysis
Statistical analyses were performed using SPSS (version 23, SPSS, Chicago, IL, USA). Independent two-sample t tests and Mann-Whitney tests were used for continuous data, and the chi-square test was used for categorical data. Survival curves were analyzed using the Kaplan-Meier survival analysis, and the log-rank test was used to compare survival rates in each group. In all cases, p values < 0.05 were considered as statistically significant.
Of the patients who underwent curative gastrectomy, 690 (88.2%) survived and the 5-year overall survival rate was 83.0%. HER2 expression level and its relationship with clinicopathologic factors (Table 2) HER2 overexpression was significantly high in males (p = 0.013), and HER2 expression was significantly more frequent in patients with differentiated gastric adenocarcinoma (well-and moderately differentiated adenocarcinomas) than in those with undifferentiated gastric adenocarcinoma (poorly differentiated adenocarcinoma and signet ring cell carcinoma) (p < 0.001). According to the Lauren classification, HER2 expression was significantly more frequent in intestinal-type adenocarcinoma than in diffuse-type adenocarcinoma (p < 0.001). With respect to the T stage, the lower the pathologic stage, the greater was the HER2 expression level and this was statistically significant (p = 0.022

Discussion
In our study, HER2 overexpression was found in 166 of 782 patients (21.2%) who underwent curative gastrectomy for gastric adenocarcinoma. HER2 overexpression was predominantly observed in males, well-and moderately differentiated type according to the histologic differentiation, the intestinal-type according to Lauren classification, and low T stage. There was no correlation between HER2 expression level and overall 5-year survival rate, regardless of the classification method. The clinicopathologic factors that affected overall 5-year survival rate in the univariate analysis were age, cancer size, histologic differentiation, Lauren classification, T stage, N stage, TNM stage, lymphatic invasion, vascular invasion, and perineural invasion. In the multivariate analysis, the important clinicopathologic factors were age, cancer size, N stage, and perineural invasion. In other studies, HER2 overexpression was defined as HER2 3+ in IHC and as HER2 2+ in IHC and specimens with amplified HER chromosome levels based on fluorescence in situ hybridization (FISH) or silver-enhanced in situ hybridization (SISH). HER2 overexpression rates ranged from 6 to 35% [8-12, 15-18, 20, 22, 28, 29]. IHC and FISH/SISH showed a concordance rate of HER2 expression to be 88.5-93.5% [11,18,25,30,31]. The rates of specimens with HER2 2+ in IHC and amplified HER2 gene levels based on FISH (or SISH) were 4% [11], 7.7% [8], and 20% [9], respectively. In this study, FISH was not performed in the HER2 2+ IHC specimens due to the high cost of implementing FISH. Based on the high concordance of HER2 expression level in IHC and FISH/SISH and the low HER2 gene amplification rate in FISH/SISH and HER2 2+ in IHC, we defined only HER2 3+ in IHC as HER2 overexpression. Thus, HER2 3+ in IHC was defined as HER2 overexpression among the 782 patients who underwent curative gastrectomy for gastric adenocarcinoma and the HER2 overexpression rate was 21.2%. Instead, the patients were divided into overexpression, non-expression, and other groups to analyze whether the survival rate and prognostic factors were related. If FISH was performed on the specimens from the 171 patients with HER2 2+, the HER2 overexpression rate would have been a little higher.
A multicenter study by Seo et al. [23] showed that HER2 overexpression was related to the intestinal type according to the Lauren classification. Kataoka et al. [15] reported that HER2 overexpression was related to males, old age, and intestinal-type adenocarcinoma. Shen et al. [18] claimed that HER2 expression was associated with sex, tumor location, histologic differentiation, and TNM   stage. Begnami et al. [19] reported that HER2 overexpression was related to the degree of tumor invasion, lymph node metastasis, and tumor pathologic stage. Kim et al. [20] reported that age, histologic differentiation, lymphovascular invasion, and lymph node metastasis correlated with HER2 overexpression. Kurokawa et al. [22] reported that histologic differentiation, intestinaltype adenocarcinoma, and pathologic stage correlated with HER2 expression, and Kim et al. [28] reported that sex, age, and intestinal-type adenocarcinoma were associated with HER2 expression. In this study, HER2 overexpression in gastric adenocarcinoma was observed in patients with differentiated-type gastric adenocarcinoma (well-and moderately differentiated adenocarcinoma) according to the histologic differentiation and intestinaltype adenocarcinoma according to the Lauren classification. HER2 overexpression was more frequently found in males than in females and in low T stages. However, unlike other studies, this study did not investigate HER2 expression according to age, cancer location, cancer size, N stage, TNM stage, lymphatic invasion, vascular invasion, and perineural invasion. Among T stages, N stages, and TNM stages related to the progression of gastric adenocarcinoma, the HER2 overexpression was only related to the low T stages. Therefore, HER2 expression seems to show a high level of expression in gastric adenocarcinoma which is in the low stages and has good differentiation.
Although some studies reported that HER2 overexpression led to a poor survival rate [13,14,19,22], others demonstrated that HER2 expression level was not related to overall survival rates [12,[15][16][17][18]. Terashima et al. [16] reported no relationship between relapse-free survival and HER2 expression level, and Shen et al. [18] reported that HER2 overexpression was not associated with disease-free survival. In this study, there was no correlation between HER2 expression level and overall 5-year survival rate in patients with resectable gastric adenocarcinoma. The HER2 expression level was also independent of the overall 5-year survival rate when classified according to the pathologic stage. Thus, HER2 expression level did not affect survival rates. This was because HER2 overexpression is not closely related to TNM stages, which was strongly related to the survival rate of gastric adenocarcinoma.
There have been various opinions on the relationship between HER2 expression level and prognostic factors. Patients with gastric cancer with HER2 overexpression have been reported to show poor prognosis [10], and HER2 overexpression has been reported to be an independent prognostic factor for patients with gastric cancer [14,[20][21][22]. On the other hand, some studies have indicated that the HER2 expression level is not related to the prognosis of gastric adenocarcinoma [12,16,18]. In this study, HER2 overexpression did not correlate with the clinicopathologic factors that affect overall 5-year survival Unlike other studies that carried out a retrospective IHC analysis on prepared tissue samples to evaluate HER2 expression level, in this study, we performed a prospective IHC analysis on pathologic tissues obtained from patients who underwent curative gastrectomy for gastric adenocarcinoma, to assess HER2 expression level. In other words, fresh adenocarcinoma tissue obtained during curative gastrectomy was made simultaneously with slides for the pathological stage as well as slides for IHC analysis. However, the HER2 expression level was assessed only by IHC analysis, so patients with HER2 2+ in IHC and HER2 gene amplification in FISH were excluded from the HER2 overexpression group. As a result, this study has shown that HER2 was neither an independent prognostic factor nor an informative prognostic biomarker in resectable gastric adenocarcinomas. However, there may be some bias in the relationship between HER2 expression levels and clinicopathologic factors and in the investigation of the impact of these factors on 5-year survival rates and prognostic factors. In contrast, the ToGA study has shown that the use of trastuzumab in unresectable HER2 positive gastric cancer is effective and patients with HER2-positive breast cancer were treated with trastuzumab. Based on the ToGA study, we will have to compare the effects and side effects of S-1 or xelox treatment and trastuzumab treatment as adjuvant therapy in resectable HER2-positive gastric cancer (stage II and III). An alternative adjuvant therapy should be to treat patients having HER2-positive gastric adenocarcinoma with trastuzumab.

Conclusion
In this study, IHC analysis of gastric adenocarcinoma tissues from patients who underwent curative gastrectomy was performed to investigate HER2 expression level. The HER2 overexpression rate was 21.2% for patients with gastric adenocarcinoma, and HER2 overexpression was predominantly seen in males, well-and moderately differentiated adenocarcinomas, intestinal type according to the Lauren classification, and low T stage. HER2 expression level did not affect the overall 5year survival rate of patients with gastric adenocarcinoma and was not correlated with independent gastric adenocarcinoma prognostic factors.
Authors' contributions KHK contributed to the study conception and design. KHK and SHS contributed to the development of methodology. HK and SHS contributed to the acquisition of data. SHS contributed to the analysis and interpretation of data. KHK contributed to the writing, review, and/or revision of the manuscript. KHK and SHO contributed to the administrative, technical, or material support. KHK contributed to the study supervision. MSA, HJB, and CSC had other relevant contributions to the study. All authors read and approved the final manuscript.

Funding
This work was supported by the 2017 Inje University Busan Paik Hospital research grant.

Availability of data and materials
All data generated or analyzed during this study are included in this published article (and its supplementary information files).