The clinical and prognostic significance of HER2 and Ki67 expression in 2062 Chinese patients with resectable gastric cancer: a retrospective study

Background The purpose of this study was to assess the prognostic value of the expression of human epidermal growth factor receptor 2 (HER2), Ki67, and their combination levels in the prognosis of Chinese patients with resectable gastric cancer (GC). Methods A total of 2062 Chinese GC patients were recruited with HER2 and Ki67 expression being evaluated using immunohistochemistry. Patients were divided into four groups according to HER2 and Ki67 expression. The distributions between HER2 and Ki67 expression levels and clinicopathological characteristics were compared using the Chi-square test. The relationship between HER2 and Ki67 expression level and overall survival were evaluated with the univariate and multivariate Cox regression analysis. Results There was no statistical differences between the overall survival (OS) rate and the expression level of HER2 ( P = 0.748) or Ki67 ( P = 0.063), but there were significant relationships between the OS rates and the combining expression levels of HER2/Ki67 ( P < 0.05). Further, Ki67, sex, T stage, N stage, TNM stage, and adjuvant chemotherapy were significant and independent risk factors for GC survival ( P < 0.05). Conclusions Our study illustrated that Ki67, but not HER2 acted as an independent prognostic factor in Chinese resectable GC patients. The evaluation of the combining expression levels of HER2/Ki67 may be more useful to assess patient prognosis with resectable GC.


Introduction
Gastric cancer (GC) was the fifth most common malignancy in 2018, with 1,033,701 new cases (5.7% of all cancers), and the third leading cause of cancer death worldwide, with 782,685 deaths (8.2% of all cancer deaths). 1 China accounts for nearly one-half of the world's GC cases with over 400,000 new cases and an estimated 290,000 deaths in 2014. 2 The 5-year survival rate of Chinese GC was 31.3%, much lower than North Korea (58%) and Japan (54%), and even lower than Mauritius (41%). 3 The fundamental treatment of GC involves the complete resection of localized tumors, but recurrence and metastatic spread occur frequently, nevertheless. In recent years, although the development of new agents and combination chemotherapies has gradually improved, the overall survival (OS) of advanced GC patients and the effective treatment therapy of advanced GC, especially potential therapeutic targets, is still a great challenge.
Among targeted therapeutic drugs, trastuzumab, a humanized monoclonal antibody, was firstly approved by the US Food and Drug Administration (USFDA) for human epidermal growth factor receptor 2 (HER2)-targeted therapy to treat HER2-positive breast cancer in 1998. 4 Over the last decade, trastuzumab application has significantly improved the outcomes of HER2-positive breast tumors, which reinforces the use of targeted therapies as a standard treatment for HER2-positive breast cancer. In breast cancer, it is well known that HER2 overexpression is an independent prognostic factor and a predictive factor for treatment with trastuzumab. 5,6 It was also showed that trastuzumab combined with chemotherapy increased the life-span of 4.2 months in HER2positive metastatic GC patients. 7 Accordingly, the relationship between the expression of HER2 and the prognosis of GC, which should be fundamental to support the value of trastuzumab as a targeted therapeutic drug for gastric cancer, has concerned many researchers. However, the results from related literature is inconsistent and requires further investigation.
Ki67, another important indicator in the molecular type of breast cancer, is also a prognostic biomarker in GC. Molecular classification provides an excellent prognostic and predictive significance in breast cancer, which were classified into four subtypes: luminal A, luminal B, HER2/neu, and basal cell-like. 8 This approach uses immunohistochemical definitions of estrogen and progesterone receptors, the detection of overexpression and/or amplification of the HER2 oncogene, and Ki67 labeling index as the means of identifying tumor subtypes. 8 Numerous studies reported that luminal A displayed the highest OS, followed by luminal B, HER2, and basal-like. 9-11 Aberrant expression of Ki67 often depicts various malignant tumors, including breast cancer and GC. 12,13 Currently, there is little research available on the molecular classification of GC. Therefore, our research objective focuses on survival analysis of the varying HER2 and Ki67 status, as well as their combinations. We divided 1663 patients with GC into four groups, based on the status of HER2 and Ki67, to explore their associations with the prognosis of GC, which may provide important information on the clinical application of trastuzumab and prognosis prediction for different GC patients.

Patient and Data Collection
The study flowchart was showed in Fig. 1. From January 2013 to December 2018, a total of 3769 consecutive cases of primary stomach or gastro-esophageal junction cancer patients were recruited from the General Hospital of Ningxia Medical University of China. This study included 2062 patients after excluding 1707 patients, who exhibited residual GC and recurrence of GC combined with other tumors. Patients who received preoperative adjuvant chemotherapy or did not undergo immunohistochemistry (IHC) analysis were also excluded from the study. We retrospectively retrieved the demographic and clinicopathological features of each patient from the medical records. Demographic information, including age, gender, ethnicity, cigarette smoking and alcohol history, and blood type were collected. Clinical pathological data, including surgical method, tumor location and size, Bormann's type, tumor differentiation, Lauren classification, extent of vessel invasion, TMN stage, and adjuvant chemoradiotherapy were collected. All resection specimens were pathologically staged according to the seventh edition of the AJCC TNM classification system. 14 Patient follow-up was conducted via telephone call or text message to obtain patient survival information until January 24th, 2019. The study endpoint was OS, which was regarded as the number of months from the date of admission to the date of death or last follow-up. The study was approved by the Ethical Committee of Ningxia Medical University of China (No. 2017-013).

Tissue Processing And IHC Staining
Surgically resected primary gastric tumor tissues were sent to the Department of Pathology, the General Hospital of Ningxia Medical University in China. IHC staining was carried out using HER2 (mouse anti-human monoclonal antibody, ZSGB-BIO, Beijing, China) and Ki-67(mouse anti-human monoclonal antibody, ZSGB-BIO, Beijing, China) as the primary antibody. Tissue specimens were fixed in 10% neutral buffered formalin, embedded in paraffin, and serially sectioned (4 µm thick). Sections were immersed in fresh xylene for 10 min to remove residual paraffin and were deparaffinized three times in 100% xylene for 10 min, and then hydrated three times in 100%, 95%, and 75% alcohol for 3 min. Sections were boiled for 2 min in EDTA antigen retrieval solution (pH = 9.0) and then naturally cooled at room temperature. Sections were immersed with 3% hydrogen peroxide for 10 min to eliminate endogenous peroxidase activity, and then washed three times with phosphatebuffered saline (PBS). Next, sections were incubated with HER2 and Ki-67 for 60 min at 37 °C. After rinsing three times with PBS, sections were incubated with sheep anti-mouse IgG polymer (ZSGB-BIO, Beijing, China) for 30 min at room temperature. After rinsing three times with PBS, sections were incubated with freshly prepared DAB coloring solution for 10 min at room temperature. Thereafter, sections were lightly stained with hematoxylin for 20 seconds. Finally, after dehydrating, sections were sealed with neutral gum.

Immunohistochemical Assessment
We applied the HER2 IHC scoring system for gastric cancer amended by Hoffmann et al.. 15 HER2 expression was scored as follows: No membranous staining or staining in < 10% of all neoplastic cells was defined as 0; Faint or barely noticeable staining in ≥ 10% of neoplastic cells was defined as 1+; Mild to moderate basolateral or complete membranous staining in ≥ 10% of neoplastic cells was defined as 2+; Strong or basolateral or complete membranous staining ≥ 10% of neoplastic cells was defined as 3+. Specimens with an IHC scoring of 0 or 1 + were classified as low HER2 expression (abbreviated as HER2 low ), those with IHC 2 + were classified as "HER2 equivocally", and those with IHC 3 + were classified as high HER2 expression (abbreviated as HER2 high ).

Statistical analysis
Data analysis was conducted using SPSS software (Version 17, Chicago, IL, USA). The chi-squared test was performed to analyze the relationship between the expression of HER2 and Ki67 and each variable. The Kaplan-Meier method was carried out to estimate survival rates, and survival curves were compared with the log-rank test. The hazard ratio (HR) and 95% confidence intervals (CIs) were estimated with a Cox proportional hazard model. Multivariate Cox regression analysis was conducted to adjust for potential confounding factors. P values < 0.05 is considered statistically significant.

Clinical information of the study participants
In total, 2062 patients were involved in this study, including 1626 males and 436 females, with the male-tofemale ratio being 3.72:1. The mean age of all GC patients was 59 years old, ranging from 15 to 91 years old. According to histologic differentiation of GC, 1017 cases (52.1%) displayed a well or moderately differentiated type and 936 cases (47.9%) displayed a poorly differentiated type. According to the Lauren classification, we found 229 intestinal-type cases (46.0%), 123 diffuse-type (24.7%), and 146 mixed-type (29.3%). In terms of T stages, there were 339 T1 cases (16.8%), 288 T2 cases (14.3%), 264 T3 cases (13.0%), and 1128 T4 cases (55.9%). A total of 1209 (59.9%) patients presented with lymph node metastases, and the others did not. Regarding the tumor stage, 1000 (49.5%) patients presented with stage I and II, and 1021 (50.5%) patients with stage III and IV (missing value 41). Additional details are summarized in Supplementary Table 1.

Discussion
The treatment of advanced GC is still a challenge despite various clinical therapies. Molecular target therapy, which has been proved to be very effective in breast cancer, might present a potential therapeutic option for GC as well. In this study, we analyzed 2062 GC patients with resectable GC to identify the relationship between the expression of HER2 and Ki67 and the resultant prognosis.
Firstly, the rate of high HER2 expression in our study was found to be 9.2% (190 of 2062) using IHC, which was consistent with most previous reports (4.4-53.4%). 16  We then divided all patients into four groups based on the expression status of HER2 and Ki67 to further compare their OS rates among four groups. Although there were no statistical differences among the OS rates of four groups, it was found that GC patients in HER2 high /Ki67 low group showed the highest OS of 69.8%, followed by those patients in other three groups with OS rates being 63.2%, 59.7%, and 57.8%. We then conducted subgroup analyses, which stratified the related variables. It is interesting to find that patients over 60 years old had reduced survival time compared with younger patients among three groups. It was also observed that patients with a smoking history exhibited poorer survival in HER2 low /Ki67 low and HER2 high /Ki67 low groups, compared with patients without a smoking history. Moreover, it was clearly showed that the OS time in HER2 low / Ki67 high group was significantly shorter than that in HER2 low /Ki67 low for patients with no-smoking history, blood type AB, blood type O, adenocarcinoma, metastasis, and no-adjuvant chemotherapy. However, the OS time in HER2 high /Ki67 low group was longer than that in HER2 low /Ki67 low for patients with blood type A. All these findings mentioned above support that routine evaluation of combining HER2 and Ki67 expression levels could be useful to assess patient prognosis with resectable GC.
However, our study had several limitations. Firstly, the expression level of HER2 was tested only using IHC. The cases which scored 2 + using IHC were not confirmed using a FISH test. Accordingly, the number of HER2 positive cases were affected, which may be prone to bias. Secondly, most patients received several uncontrolled adjuvants and palliative chemotherapies, which may lead to a protective effect for patients with GC. Lastly, our data were retrieved only from a single agency, indicating that caution should be exercised when interpreting the results.
To the best of our knowledge, this is the first study to evaluate HER2 status combined with Ki67 status in 2062 patients with GC. In conclusion, this large-scale study demonstrated that low Ki67 expression is an independent prognostic factor in GC patients and various combinations of HER2 and Ki67 expression can affect patient prognosis.

Declarations
Ethics approval and consent to participate The study was approved by the Ethical Committee of Ningxia Medical University prior to study conduct (No. 2017-013). Informed consent was obtained from all patients included in the study.

Consent for publication
Written informed consent for publication was obtained from all participants.

Availability of data and materials
The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.
data. WW and WY edited the manuscript. WY and JR supervised the study. All authors read and approved the final manuscript. The expression of HER2 and Ki67 protein detected by IHC. Representative specimen with low HER2 expression (a), high HER2 expression (c) (x200), low Ki67 expression (b), and high Ki67 expression (d) (x200).

Supplementary Files
This is a list of supplementary files associated with this preprint. Click to download. Supplementarydata.pdf