Human Epidermal Growth Factor Receptor-2(HER-2) expression in Sri Lankan gastric adenocarcinoma patients: An analysis of the concordance of HER2 expression and survival based on silver in-situ hybridization (SISH)

Background: Accurate HER2 status is crucial for gastric adenocarcinoma (GC) patient selection for antiHER2 therapy. It is assessed immunohistochemically (IHC) for protein expression, and by silver in-situ hybridization (SISH) for gene copy number. This study aimed to evaluate the concordance of HER2 status by IHC/SISH analyses and HER2-SISH based survival. Methods: This prospective study includes 145 GC’s (excluding gastro-oesophageal-junction tumours) from the National Hospital of Sri Lanka with dened demographic, clinical-radiological-pathological characteristics. HER2-IHC was assessed by DAKO A0485, RealTM Envision system and interpreted using Ruschoff criteria. HER2-SISH was assessed with INFORM HER2 dual ISH DNA Probe Cocktail. Concordance between HER2 IHC/SISH results was determined by Cohens kappa statistics. The association between the survival and HER2-SISH positivity was evaluated using the cox-regression method. Adjustments were done for age, gender, Lauren classication, tumour location and the tumour staging. Results: Of the 69 gastrectomies and 76 biopsies, 8.3% (n = 12) were HER2-IHC positive (n = 7, + 2 and n = 5, + 3). HER2-SISH positivity was 4.8 % (n = 7). All IHC + 3 were SISH positive, while two + 2 cases were SISH positive. Concordance for IHC 0, + 1, +3 were 100%. There was a signicant overall correlation (kappa = 0.72, p < 0.001) between HER2-IHC and HER2-SISH indicating substantial concordance. The mean overall survival of HER2-SISH negative and positive patients were 41.7 (0-210) and 14.6 (3–51) weeks respectively, after a mean duration of patient follow up for 40.4 weeks (range 0-210). Survival was relatively lower (p = 0.001) in the group with HER2-SISH positivity. Conclusion: for 0, + 1, +3 scores and could be used for treatment and prognostication in low resource settings, where SISH facility is unavailable. HER2-IHC + 2, without gene amplication may be due to transcriptional activation by other genes or post-transcriptional events, mandating further evaluation by SISH. Survival of GC patients is signicantly affected by HER2-SISH positive status. HER2 SISH as the probe. quantitatively detect amplication by light microscopy of the HER2 gene and the centromere portion of chromosome 17(CEP17), via two colour chromogenic in situ hybridization(ISH) in FFPE GC tissue, following staining on Ventana automated slide strainers. During SISH signal counting, a discrete signal was counted as a single copy of HER2 or CEP17. HER2 SISH signals (black) are typically smaller in size and more discrete in appearance than CEP17 SISH signals (red), due to differences in target sizes and detection chemistries. During the signal interpretation process 20 cells were counted for red (CEP 17-Chromosome 17 centromere) and black (HER2 signals). HER2 gene status was classied as non-amplied (HER2/Chr17 ratio < 2.0) or amplied (HER2/Chr17 ratio ≥ 2.0).


Introduction
HER2 positive gastric carcinomas (GC) are usually associated with more aggressive biological behavior [1][2][3][4][5][6] . The Trastuzumab for Gastric Cancer (ToGA) trial revealed trastuzumab in combination with chemotherapy, signi cantly improves the overall survival of patients with HER2 positive advanced GC or gastroesophageal junction carcinoma (GOJC) 6 . Based on the results of the ToGA study in 2010, the Food and Drug Administration (FDA) 7 in USA and Therapeutic Goods Administration (TGA) in Australia granted approval 8 for trastuzumab in combination with chemotherapy for the treatment of patients with GC and GOJC. Subsequently this therapy has been approved globally for GC and GOJC 8 .
The reported frequency of HER2 protein overexpression by immunohistochemistry (IHC) in GC ranges from 8.2% to 40.3% [1][2][3][4][5][6][9][10][11][12] . Only a single documentation is available from Sri Lanka of HER2 expression by IHC in Sri Lankan GC patients. 12 HER2 status is determined by immunohistochemistry (IHC) and in-situ hybridization (ISH) techniques. IHC is semi-quantitative, evaluating HER2 protein expression. IHC has the advantage of being relatively inexpensive and is widely used in clinical laboratories on formalin-xed para n-embedded tissue sections (FFPE). In contrast, in-situ hybridization (ISH) is quantitative, evaluating the HER2 gene copy number. Three types of ISH methods are used for HER2 status determination; uorescence in situ hybridization (FISH), chromogenic in situ hybridization (CISH) and silver in situ hybridization (SISH) 13,14 . SISH is the preferred method among these different techniques 15 . HER2-SISH status data of Sri Lankan GC patients is unavailable, with paucity of similar data from South Asia. Sri Lanka has a low incidence of GC in comparison to global and regional countries, with an incidence of 1.2 per 100,000 population and an age adjusted mortality rate of 6.7 16  In all patients the diagnosis was con rmed by upper gastrointestinal endoscopy (UGIE) and biopsy. Only the gastric resection specimen was included in patients who underwent curative surgery following biopsy. The endoscopic biopsy was included in patients with advanced tumours, who did not undergo gastric resection. Age at diagnosis, gender, type of specimen, tumour location (proximal/distal stomach) and radiological stage assessed by contrast enhanced computerized tomography (CECT) of the abdomen and thorax were documented using a structured data sheet. Radiological data was used to determine the T (tumour), N (nodal with enlargement > 1 cm) and M (metastasis) stages of patients who only had biopsies without resections. Pathological data were used to determine the T and N stages in patients who underwent resections. The TNM stage was determined in accordance with the seventh edition of the UICC guidelines 17 .
Pathology: FFPE tumour samples were stained with haematoxylin and eosine. Lauren's classi cation for GC was used for histological typing (diffuse, intestinal or mixed) 18 .
Immunohistochemistry: FFPE tumour tissue sections were stained for HER2 protein expression using polyclonal rabbit anti-human c-erB-2 oncoprotein (Dako A0485) and Dako Real TM Envision systems. Breast cancer tissue with a HER2-IHC +3 score was used as the positive control. According to Ruschoff scoring [19] a score of IHC 0 or +1 was considered negative for HER2 overexpression, whereas a score of IHC +3 was considered strongly positive. A score of IHC +2 was also considered positive for HER2 overexpression based on criteria by Ruschoff et al.
Silver in situ hybridization: Tissue microarrays (TMAs) were prepared at the Department of Anatomical Pathology, Pathwest QE II Medical Centre, Perth, Australia, using two tissue cores with a diameter of 0.6 mm extracted from each tumor using the TMA arrayer (TMA Master 1.16 SP1). Each TMA also contained various non-gastric tissue samples as controls. Sections of 4μm obtained from the TMA blocks were used for SISH/IHC. The slides were stained using automatic staining devices: the Benchmark Ultraview (Ventana Medical Systems).
The INFORM HER2 dual ISH DNA Probe Cocktail was used for HER2 SISH as the probe. This was designed to quantitatively detect ampli cation by light microscopy of the HER2 gene and the centromere portion of chromosome 17(CEP17), via two colour chromogenic in situ hybridization(ISH) in FFPE GC tissue, following staining on Ventana automated slide strainers. During SISH signal counting, a discrete signal was counted as a single copy of HER2 or CEP17. HER2 SISH signals (black) are typically smaller in size and more discrete in appearance than CEP17 SISH signals (red), due to differences in target sizes and detection chemistries. During the signal interpretation process 20 cells were counted for red (CEP 17-Chromosome 17 centromere) and black (HER2 signals). HER2 gene status was classi ed as non-ampli ed (HER2/Chr17 ratio < 2.0) or ampli ed (HER2/Chr17 ratio ≥ 2.0).
Statistics: The statistical software program SPSS 21(SPSS Inc., Chicago, IL, USA) was used for data analysis.

Survival analysis:
Initially survival analysis was performed by the Kaplan-Meier method with log-rank test comparing the survival of HER2-SISH positive and negative groups. Subsequently the association between the survival time and HER2-SISH positivity was evaluated using the cox-regression method. Adjustments were done for age, gender, Lauren classi cation, tumour location and tumour staging.
HER2-SISH expression and survival: The mean duration of patient follow up (89% patients for ve years or until death as the end point) was 40.4 weeks (range 0-210). All have been treated with standard adjuvant chemotherapy with none receiving anti HER2 (Tratuzumab) therapy. The mean overall survival of HER2-SISH negative and positive patients were 41.7 (0-210) weeks and 14.6 (3-51) weeks respectively. The signi cantly poor overall survival (p=0.018) of HER2-SISH positive patients is illustrated in Figure 6. SISH-positivity and advanced age were observed to be independently having a higher hazard ratio for mortality as shown in Table 3.

Discussion
GC accounts for about 10% of cancer related deaths worldwide with a case fatality rate of 70% [20]. Most (70%) GCs occur in developing countries with half the world's total cases occurring in East Asia 21,22 . The majority are diagnosed at an advanced stage. Despite multiple therapeutic strategies including radical surgery combined with chemotherapy, improving the outcome of advanced stage GC remains a challenge. In recent years, HER2 receptor status has been identi ed as a novel molecular targeted therapy for GC 6 . Molecular target therapy for GC depends on the accurate evaluation of the HER2 receptor status or the target gene. Therefore, accurate determination of HER2 status is important for molecular targeted therapy.
HER2 positivity by IHC was 8.2% (n=12), in the study. HER2-IHC has been documented as 9% in the rst 100 patients of this patient cohort (12). IHC +2, +3 were considered as positive in both these studies. While HER2-IHC positivity rates in GC have generally ranged from 8.2% to 40.3% 1-10 1-6, 9-12 , Asian studies have reported rates ranging from 11.7% to 15.74% [23][24][25] . The explanation for this variation is likely to be multifactorial including differences in the populations studied, use of non-standardized assays using different antibodies and the application of different scoring criteria for interpretation. GOJC's are also reported to have higher HER2 expression rates 26 . The GOJC's were excluded from the current study as these are now considered to be a distinct entity 27 Exclusion of GOJC and the use of a polyclonal antibody may have contributed to the lower HER2 positivity rate the study encountered. Alternatively, the low HER2-IHC positivity rate could be a re ection of a genuine difference in tumour biology of the local population or due to heterogeneity of HER2 receptor expression which is a well-recognized phenomenon in GC 28 .
Most comprehensive data on the concordance between HER2 status by IHC and SISH methods comes from the systematic review and meta-analysis by Pyo et al 30 , which included a total of 12,679 cases from 45 individual studies. Very high concordance rates were found between 0/1+ and 3+ as was expected. Concordance between HER2 IHC2+ and HER2 gene ampli cation was more variable. The pooled sensitivity and speci city of IHC positivity (when both 2+ and 3+ were considered positive), in predicting ISH con rmed HER2 positivity, were 0.86 and 0.91, respectively. In the current study all cases underwent HER2 ampli cation testing by SISH. 5 of the 12 IHC+ cases were SISH negative; All 5 SISH negative cases were IHC +2. HER2 gene was ampli ed in all cases with overexpression of the HER2 protein at +3 level, and HER2 gene was not ampli ed at 0/+1 level, which is consistent with the reported data 31 . Therefore in this cohort of Sri Lankan GC patients HER2-IHC was well concordant with HER2-SISH for 0, +1, +3 scores. These ndings are valuable as HER2-IHC could provide useful information in limited resource settings for treatment and prognostication for patients with HER2 scores of 0, +1, +3. This is in keeping with the current recommendation that all patients with GC should have their tumours tested for HER2 status at the time of the initial diagnosis with IHC as the rst screening method. Those cases with results considered equivocal for HER2 overexpression (2+) should be referred for HER2 analysis by in situ hybridization 14 . This is endorsed by the study ndings where, of the 7 IHC +2 cases, SISH positivity was observed in only 2 cases.
The subset of cases showing IHC/SISH discordant results could also be due to heterogeneous HER2 protein expression. Chromosomal instability is probably one of the major causes for this heterogeneity. Kameda et al detected HER2 overexpression without ampli cation and considered that this may indicate that gene ampli cation may not be the primary mechanism by which the HER2 protein is over expressed in GC 32 . HER2 overexpression may occur by a number of different mechanisms, including transcriptional activation by other genes or post-transcriptional events 33 Therefore, HER2-IHC 2+ cases represent a subset of patients with uncertain diagnostic characteristics in terms of HER2 status. IHC 2+ cases should be interpreted carefully and necessary con rmatory tests for gene ampli cation should be carried out.
Many studies have examined the hypothesis that HER2 status could be a predictor of the survival rate in patients with GC. Andreas et al 34 have shown that there is a signi cant association between the level of HER2 expression and faster achievement of cancer-free and overall survival, without adjusting for the tumour stage. Brien et al 35 have shown that the pathological stage and HER2 gene ampli cation are independent prognostic factors of survival in multivariate analysis. According to the ToGA study 6 median overall survival was 13.8 months in HER2 positive GC treated with trastuzumab compared to 11.1 months in the chemotherapy only arm. In keeping with these ndings, HER2-SISH positivity and advancing age were associated with a signi cantly low overall survival in the present study independently of tumour stage and other confounding factors.
In conclusion, HER2 prevalence by IHC and SISH were 8.2% and 4.8% respectively in this Sri Lankan GC patient cohort. The reliability between HER2-IHC and HER2-SISH is satisfactory with substantial agreement. In view of this, HER2 by IHC could be a satisfactory alternative for GC with 0, +1 and +3 scores, in settings where HER2 by SISH is unavailable. Patients with HER2-IHC +2 scores however require con rmatory testing for gene ampli cation. Additionally, SISH HER2 positivity is associated with a low overall survival among Sri Lankan GC Funding-This study was funded by National Research Council, Sri Lanka by providing the research grant .
Overseas phase of this study was funded by National Science Foundation for providing me a travel grant (OSTP/2016/03).
Availability of Data -The datasets used and analyzed during the current study are available from the corresponding author on reasonable request.

Ethics approval and consent to participate
This study has been approved by the Ethics review committee of the Faculty of Medicine, University of Colombo (EC 11-139) and the Ethics Review committee of the NHSL. All patients gave their written informed consent to collection and use of their data for research purposes. No individual patients can be identi ed by the anonymous data used in this study.
Consent for publication -Not applicable. Tables   Table 01-