| Background and AimsColorectal cancer (CRC) is one of the most common human neoplasms. Synchronous liver metastases (SLM) account for approximately14.5%of newly diagnosed CRC patients, which are often resistant to conventional therapies and lead to a poor prognosis. Therefore, they need more effective treatments. A better understanding of the mechanism for SLM development will provide a biological foundation for the design of more effective specific drugs.The system of hepatocyte growth factor (HGF) and its receptor of mesenchymal-epithelial transition factor (Met) had been found to play a vital role in metastasis of CRCs. Experimental studies have shown that the pathway was important for proliferation, invasion and migration of colorectal cancer cells, the prerequisites for cancer cell metastasis. When the antagonist Nk4was used to inhibit the ability of HGF in colon cancer cells, metastasis decreased in animal models. Inhibition of HGF and Met was reported to prevent distant metastasis for rectal cancer after preoperative chemoradiotherapy. In clinical reports, Met amplification and overexpression at the levels of DNA, RNA and protein were observed in colonic adenomas, primary tumors and liver metastases while less expression existed in normal colonic tissues. HGF was also detected in primary tumors and metastases of CRCs. But HGF expression showed no significant differences between tumor and normal tissues, indicating that metastatic potential enhanced by HGF and Met was due to the overexpression of Met increasing sensitivity to HGF. Their expression in primary tumors was associated with colorectal cancer metastasis and has been used to evaluate prognosis and metastasis of colorectal cancer. At the treatment levels, the inhibitors of HGF/Met are being developed to be novel targeted drugs for controlling colorectal cancer metastasis.However, it seems that there are some limitations for these reports when considering the relationship between the pathway and SLM. Recent studies suggest that SLM is an independent entity in CRC and is different from other metastases and even metachronous liver metastasis at the levels ot biology and treatment. Previous studies did not distinguish SLM from other metastases. For CRC patients with SLM, some are concurrent with regional lymph node metastases (RLNM). others directly develop into SLM without RLNM involving, which remains an unsolved issue. RLNM and other clinicopathological factors such as age in female patients and primary tumor site have influence on SLM. Some of previous clinical studies didn’t eliminate the influence of these factors when they investigated the relationship between the pathway and colorectal cancer metastasis. Moreover, the inhibitors of Met and HGF may be useful to treat metastatie tumors. Nevertheless, there are heterogeneity between primary CRCs and liver metastases. which will lead to different response to drugs. Evaluation of the response to molecular targets is mainly based on their status in primary tumors. EGFR inhibitors in clinical practice have raised a question whether metastatie sites need biopsies when they arc used to treat advanced stage CRCs. The same question also exists for the inhibitors of1IGF/Met in treatment of CRC metastasis. Several studies have compared I IGF and Met expression between primary CRCs and matched metastases but showed eontroversial results. These questions will make puzzling on understanding of the biology of SLM and treatment of SLM using the inhibitors of HGF and Met.To solve the puzzling, we designed a case-control study and detected expression of HGF and Met in primary CRCs with SLM versus primary CRC without SLM. For patients with metastases, the two molecules were also detected in their primary tumors and corresponding metastases. Based on comparison their expression in primary tumors with SLM versus without SLM. and comparison their expression between primary tumors and corresponding metastases. we aimed to investigate the role of the pathway in SLM and to explore whether their expression was concordant or different between primary CRCs and corresponding metastases.Patients and MethodsBetween June2001and June2010,253patients with SLM were found in1557newly diagnosed CRC patients from Shandong Cancer Hospital. Only30patients with SLM in the pathological database received radical resection and had complete clinical data, which were divided into two subgroups according to the presence of RLNM. One Subgroup included21patients with TxN1-2Mliver and the other subgroup included9patients with TxNOMiliver-In the subgroup with TxN1-2Mliver, each patient was matched to two patients of one with TxN1-2M0and the other with TxNOMO. In the other subgroup, each patient with TxNOMliver was matched to one patient with TxNOMO. Each pair had the same status at age, gender, depth of invasion, differentiation and tumor site so as to eliminate their influence on SLM. Finally,81patients in the total three matched groups were included in the study. The matched groups and their clinicopathological factors were shown in Tablel.51matched patients without liver metastases at initial diagnose were followed up for at least six months after surgery so as to exclude SLMs. They could provide available tissues including primary tumors and matched metastases. None of the cases received adjuvant therapy before surgery. Their formalin-fixed, paraffin-embedded (FFPE) tissues of primary tumors (81specimens), matched lymph node metastases (42specimens) and liver metastases (30specimens) were collected to detect HGF and Met expression at the protein level and RNA level using the methods of immunohistochemistry(IHC) and real-time reverse transcription-polymerase chain reaction(RT-PCR). All patients in the study were consented according to the ethical standards of the Helsinki Declaration of1975.ResultsHGF and Met Protein expression in primary tumors with liver metastasis versus without liver metastasisHGF expression was observed in membrane and cytoplasm of tumor cells. Statistical analyses showed that its expression had no relation with clinicopathological factors. In the subgroup of TxNOM1versus TxNOMO (respectively9cases),67%(6/9) versus56%(5/9) showed positive in primary tumors with SLM compared with those without any metastases. It didn’t reach significant statistics (P=0.734). In the other subgroup of21matches, HGF expression in primary tumors was different (p=0.003) among three matched groups. HGF expression in primary tumors with SLM showed stronger than that in those without SLM of the other two groups. In the total three matched groups, it exhibited positive in70%(21/30) of primary tumors of SLM group,42.9%(9/21)of primary tumors with RLNM, and only30%(9/30) of primary tumors without any metastases, which reached a significant statistical ditterence (P=0.007). Figure1A, ID and IF were primary tumors lrom a matched pair of three patients, which respectively showed positive, negative and negative in expression of HGF.Met immunoreactivity was observed in the cytoplasm and plasma membrane of tumor cells. It had correlation with lymph node stage (r=0.381, P=0.000). The intensity of Met expression in primary tumors with N2stage showed stronger than those with N1and NO stage. Its expression in primary tumors showed in Table2. In the subgroup of TxNOM1versus TxNOMO, Met expression showed positive in89%(8/9) of primary tumors with SLM and67%(6/9) of primary tumors without metastases. It didn’t reached significant (p=0.436). In the other subgroup of21matches, Met expression (positive and negative) in primary tumors showed different (P=0.001). The intensity of Met expression in primary tumors of TxN1-2M1and TxN1-2M0showed stronger than that in primary tumors without any metastases. There were no significant difference between primary tumors of TxN1-2M1and TxN1-2M0. In the total three groups, it (identified to be positive and negative) showed positive in90%(27/30)of primary tumors in SLM group,86%(18/21) of primary tumors in LN group and50%(15/30) of primary tumors in PT group. The results reached significance (p=0.004).RT-PCR of HGF and Met mRNA expression in primary tumors and liver metastasesThere was significant correlation between the levels of protein and RNA for HGF (r=0.600, P=0.003) and Met r(r=0.724. P=0.001) expression in primary tumors (81tissues) and liver metastases(IIGF:r=0.523P=0.032; Met:r=0.439, p=0.037). Their expression in primary tumors with SLM and those without SLM were listed in table5and table6.In the subgroup of TxNOMl and TxN0M0(9matches expression of HGF and Met mRNA in primary tumors had no significance between TxNOMl and TxNOMO (independent t test, HGF:p=0.855. Met:p=0.646). In the other subgroup of TxNl-2ML TxNl-2M0and TxNOMO (21matches reached significance (One way ANOVA, HGF:p=0.038, Met:p=0.0082). In the total three matched groups, they also reached significance (One way ANOVA, HGF:p=0.035, Met:p=0.0079).When comparing HGF and Met mRNA expression between primary tumors and matched liver metastasis, they had no significant significance. Both the two subgroups also had no significance. However, there were significant differences for some pairs when viewing of HGF and Met expression in each pair. In one subgroup of TxNOM1and TxN0M0(9pairs), there were respective four cases for HGF and six cases for Met to have similar relative copies, and five and three cases to show significant differences for HGF and Met mRNA. For HGF, four cases had higher copies of HGF mRNA in liver metastases than that in matched primary tumors and one had lower copies in liver metastasis than that in matched primary tumor(≥or<20relative copies). For Met mRNA, two cases showed higher relative copies in liver metastases than that in matched primary tumors and one case showed lower relative copies in liver metastasis than that in matched primary tumor(≥or<300relative copies). In the other subgroup, their expression in lymph node metastases was not detected because the paucity of extraction RNA. In the21patients of paired primary tumors and liver metastases,18cases for HGF and17cases for Met had similar relative copies while5cases and2cases showed different. Totally,22cases had similar relative copies of HGF mRNA and23cases had similar relative copies of Met mRNA. The concordance rates between primary tumors and matched liver metastases were73.3%for HGF and76.7%for Met.ConclusionHGF and Met might play the role in development of SLM when concurrent with RLNM from CRC but have little influence on SLM without involvement of RLNM. Major concordance and minor difference exist between primary tumors and matched mestases, which provide evidence for further using inhibitors of HGF and Met in CRCs with SLM. |
PDF Full Text Request