It has been proven that the expression of HER2 on cancer cells in patients with gastrointestial cancers is not such a good predictor of response to anti-HER2 therapy as in patients with breast cancer. However, the need for evaluation of HER2 protein expression or HER2 gene amplification has become the standard in gastric cancer patients diagnosis. HER2 overexpression on cells of gastric and esophagogastric junction cancers may affect quarter of the patients. Rajagopal et al. indicated that patients with esophagogastric junction cancer showed HER2 expression in 45.5% of cases. Whereas, only 22.4% of gastric cancer patients showed HER2 expression on cancer cells [12]. The authors presented the results obtained using IHC method.
Unfortunately, only a small fraction of our patients were tested for HER2 protein expression by IHC method before surgery (no access to preoperative treatment with trastuzumab in Poland). However, in most of our patients, we examined the HER2 gene amplification using the FISH method. This method is used in routine diagnostics to verify questionable results of the IHC test (low expression of HER2 proteins). HER2 gene amplification is almost always associated with high expression of HER2 protein on tumor cells (we also proved this observation in a small group of patients). We found HER2 gene amplification in 38% of GC patients and 40% of EGJC patients. Thus, we have not confirmed that HER2 gene amplification is more common in EGJC patients than in GC patients.
Pyo et al. and Staněk et al. indicated that the compatibility between results of IHC, FISH and qPCR methods performed in tumor tissue is quite high and ranges from 88% to 100% of cases. Nevertheless, it is pointed out that the qPCR results have less compliance with the results of other methods especially in patients with diffuse gastric cancer [13,14,15]. Also, Zhu et al. found high compliance between IHC and qPCR results in the detection of HER2 abnormalities in cancerous tissue of gastric cancer. Authors stated that qPCR may be an alternative method of HER2 testing in tumor tissue [16]. Studies also demonstrated relatively high concordance between elevated serum HER2 level and positive HER2 status in tumor tissues [17,18,19].
We detected strong relationship between HER2 gene copy number in tissue assessed by qPCR and number of HER2 gene copies in cancer nuclei examined by FISH method. However, we are convinced that quality of FFPE tissue is crucial for obtained reliable results of IHC, FISH and qPCR methods. Therefore, for the first time in the world in GC and EGJC patients qualified for surgery, we have attempted to evaluate the HER2 gene copy number in cf-DNA in liquid biopsy in patients with GC and EGJC. This method is non-invasive, and enables evaluation of genetic changes in free circulating tumor DNA (ct-DNA), which is usually of good quality. In our study, we did not observe statistically significant correlation between serum and tissue HER2 gene copy number measured with qPCR but compatibility of results of these tests was 63% (κ=0.26, data not shown). Therefore, we found that patients with a high copy number of the HER2 gene in the tumor tissue assessed by qPCR (but not by FISH) had significantly more often a high number of HER2 gene copies in liquid biopsy. Moreover, GC and EGJC patients had a higher number of HER2 gene copies in the serum than healthy people. Moreover, we did not notice the difference in HER2 gene copy number between patients with GC and EGJC. We found that HER2 CNV examination could distinguish healthy individuals and patients with gastric or esophagogastric junction cancers with sensitivity and specificity of 58% and 98%. The assessment of HER2 gene copy number in liquid biopsy could supplement IHC and FISH examination. However, we should conclude that it is not possible to replace the diagnosis of HER2 abnormalities in tumor tissue with such diagnosis in liquid biopsy for qualification of patients to anti-HER2 treatment.
Similarly, Sasaki et al. argued that HER2 protein measurement in serum cannot substitute diagnosis of HER2 abnormalities in cancer tissue in advanced gastric cancer patients, despite that they indicated correlation between protein expression in tissue and in serum [20]. They used the chemiluminescent immunoassay test for HER2 detection in serum. However, they have no data regarding the molecular status of HER2 gene in serum.
Liu et al. studied the plasma HER2 gene amplification in gastric cancer patients by droplet digital PCR (ddPCR) during 12 months of chemotherapy with fluorouracil and oxaliplatin combined with trastuzumab [21]. They observed that the concordance rate of results of HER2 gene examination in plasma and FFPE tissue samples by ddPCR was 81.4%, with a sensitivity of 76% and a specificity of 84%. Significant decrease in the plasma HER2 gene copy number was found after two months of treatment [20]. Researchers indicated that serum HER2 gene assessment may serve as a good marker for monitoring the effectiveness of trastuzumab therapy in patients with gastric cancer. However, authors stated that their study involved only small group of 12 patients.
Kinugasa et al. also used ddPCR to assess HER2 gene status in ct-DNA in group of 25 gastric patients [22]. They demonstrated that compliance of results of HER2 gene examination in FFPE tissue samples and in ct-DNA was not high (62.5%). This level of concordance between the results of the HER2 gene tests in cancer tissue and in liquid biopsy was also found in our study and in study by Kim et al. [9]. In addition Kinuasa et al. found that patients who had high number of HER2 gene copies in ct-DNA showed significantly shorter survival compared with HER2-negative patients [21]. Wang et al. indicated that evaluation of ct-DNA in liquid biopsy by next generation sequencing could be useful in diagnosis of trastuzumab resistance in HER2+ metastatic GC [22]. The median HER2 gene copy number in ct-DNA in Kinugasa et al. study was 1.15. They assumed the cut-off point above 1.2 gene copy number was considered as HER2 positivity [23]. Based on this threshold, they found that 7 (29%) out of 24 examined patients were HER2-positive. In our study median HER2 gene copy number was 1.85. Based on this threshold, we found high HER2 gene copy number in 40 (51%) of 78 GC and EGJC patients. Using tissue material, we detected 26 (38%) of 67 patients with HER2 gene amplification (FISH method) and 29 (46%) of 63 patients with high number HER2 gene copies (qPCR method).
There are many more reports on CNV HER2 evaluation in liquid biopsy in patients with breast cancer compared to GC or EGJC patients [24, 25]. They showed high concordance between HER2 gene copy number in tissue and in liquid biopsy in breast cancer which is not always observed in the case of gastric cancer and, primarily, in gastroesophageal junction cancer. EGJC is the least studied cancer for HER2 gene status. Moreover, De Mattos-Arruda indicated that ctDNA from CSF (cerebro-spinal fluid) had the potential to identify brain metastasis-specific actionable genomic alterations that may facilitate the design of personalised treatments in breast cancer patients [26]. There are indication that HER2 gene copy number can be assessed in serum of peripheral blood as a predictive marker for qualification or monitoring of targeted treatment with trastuzumab in breast cancer patients. Moreover, there are studies on HER2-positive free-circulating tumor cells which indicate the possibility of using this diagnostic and monitoring tool in breast cancer patients [27, 28].
In general, a liquid biopsy prior to any treatment implementation can be used as a benchmark for further molecular testing. Due to the lack of reports on the measurement of the HER2 gene copy number in serum or plasma before surgery in patients with GC or EGJC, further research in this area should be carried out. Before surgery, the correlation between CNV of HER2 gene in serum or plasma and in tumor tissue is not known and its clinical significance is also unknown. GC or EGJC patients qualified for surgery with a priori analysis of HER2 gene status could be considered as potential candidates for neoadjuvant treatment with trastuzumab. Clinical trials are conducted with the use of trastuzumab in combination with chemotherapy as adjuvant or neoadjuvant therapy for patients qualified for surgery [29,30]. These studies use the evaluation of HER2 gene status in tumor tissue. HER2 expression may change during preoperative treatment. In qualification to postoperative therapy, the ideal solution would be the evaluation of HER2 gene status in a liquid biopsy. We realize that the introduction of such a method is not easy. However, the availability of material for testing (plasma or serum), without the need to use invasive methods to collect tissue sample, seems to be a great advantage, both in the diagnosis of patients before and after surgery. Further research in this area is definitely necessary.