Abstract
Background/Aim: To select and stratify patients for optimal treatment plans is challenging. Identification of cancer-related biomarkers that serve as predictors for prognosis and treatment response is essential to better predict treatment outcome and find future targets for therapy. Previous data has suggested ARHGAP4 as a relevant biomarker in colorectal cancer (CRC). The purpose of this study was to assess how ARHGAP4 expression affected patients undergoing surgery for colon liver metastasis (CLM) in terms of overall survival (OS). Patients and Methods: A total of 251 patients undergoing resection of CLM from 2006 to 2017 were included. Corresponding resected tumor specimens were examined for ARHGAP4 expression levels by immunohistochemistry (IHC). The correlation between ARHGAP4 expression and postoperative survival was analyzed. Results: High expression levels of ARHGAP4 were seen in 60% of patients. High expression levels of ARHGAP4 were correlated with adverse prognosis after hepatectomy due to CLM. Survival data generated using Cox proportional hazard model showed a statistically significant difference between high and low ARHGAP4 expression groups by univariate (HR=1.5, 95% CI=1.1-2.2) and multivariate (HR=1.5, 95% CI=1.0-2.1) analysis. In multivariate Cox regression, high ARHGAP4 expression, preoperative CEA levels and presence of vascular invasion by pathological examinations were independent predictive factors of overall survival. Conclusion: ARHGAP4 is a novel prognostic biomarker after resection of CLM.
Colorectal cancer (CRC) represents the third most common cause of cancer-related mortality (1). Death from CRC is typically associated with metastatic disease and approximately 50% of patients with CRC will at some point develop distant metastatic lesions (2). Currently, surgical resection of all tumor lesion(s) is the only treatment option offering a potential cure. For patients suffering from metastasis to the liver CLM, a cure is possible if the primary tumor and liver metastasis are surgically removed, but only 20% of patients with CRLM are candidates for this treatment regimen (3). Surgical resection is only advised in patients where all tumor(s) can be resected (R0 resection) and where functional liver volume can be maintained after surgery (3-6). Potential adverse effects of both surgery and chemotherapy can be severe. In patients undergoing resection of both primary CRC and CRLM, the 5-year survival rate is approximately 40% (7-10). Surgical mortality rates are approximately 2% (11, 12).
In addition to surgical resection of the malignant tumors, modern treatment of CRLM in most cases consists of systemic chemotherapy administrated both before and after surgery, ablative techniques of liver tumors, and radiation of primary rectal cancers (3-5, 13). The process of selecting operative techniques, chemotherapy cycles, timing, and additional treatment approaches is complex and requires a high level of competence. Therefore, CRLM is often treated at highly specialized centers where a multidisciplinary conference recommends viable treatment options for the patient based on patient factors and motivation as well as tumor-derived factors (5). The preoperative evaluation is extensive.
Even if viable treatment options in addition to surgery, anesthesiologic patient optimization, and centralization of CRLM to high-volume centers have increased over the last decades, the selection of patients who benefit from specific treatment strategies remains challenging (5). Risk scores, primarily based on tumor-derived factors have been developed to better stratify patients who benefit from surgery (14, 15). However, current knowledge of factors affecting survival after surgery is imprecise and does not help clinicians select patients precisely enough (5). Consequently, interest in tumor biology, including genetic and proteomic analysis, has increased over the last decades.
Rho GTPase-activating protein 4 (ARHGAP4) encoded by the ARHGAP4 gene is localized in the Golgi complex with the potential to redistribute to microtubules (16). The Rho GTPase-activating protein family recognizes and stimulates hydrolysis of guanosine triphosphate to generate guanosine diphosphate (16). ARHGAP4 contains three main functional domains; Fes/Cip4 homology Bin/amphiphysin/Rvs (F-BAR), Ras homology GTPase activating protein (Rho-GAP), and Src Homology 3 (SH3) (16). These domains mediate membrane movement, intracellular vesicle transport, and endocytosis (17). The protein coded by the ARHGAP4 gene has been found to potentially regulate interactions between rat sarcoma (RAS) family and GTPase which has been displayed to affect the prevalence of several types of cancers (18-20). Rho family GTPases have vital effects in various cell functions (21-23). Dysregulation of Rho GTPases have been associated with development and progression of malignant properties in cells. Research studying the mechanisms of ARHGAP4 in cancer is limited, but it is suggested that ARHGAP4 can contribute to tumorigenesis by dysregulating signaling cell growth, actin cytoskeleton remodeling, cell proliferation, and epithelial-mesenchymal transition (16, 20-23). It is, however, important to note that the mechanisms of how ARHGAP4 leads to cancer progression may vary between tumor types. The Human Protein Atlas indicates that mRNA expression levels of ARHGAP4 may influence survival in colorectal cancer (24). A prior study using online Gene Expression Profiling Interactive Analysis (GEPIA) found that high levels of ARHGAP4 expression was correlated with adverse prognosis in CRC (25). So far, there is no report exploring the impact of ARHGAP4 expression on patient outcome in CRLM.
The aim of the study was to investigate if ARHGAP4 protein expression can predict outcome in patients with resected colon liver metastasis (CLM).
Patients and Methods
Data collection. Patients who underwent liver resection due to metastasis of colon cancer at the Skåne University Hospital, Sweden, from 2006 to 2017 were identified in local operative registers. National ID numbers were collected to access medical records. Patient characteristics collected from pre, per, and postoperative settings were extracted. The following data were collected: Date of liver resection, age at the time of liver resection, date of primary cancer diagnosis, date of primary cancer resection, sex, localization of primary cancer within the colon (proximal or distal of the hepatic flexure), TNM-stage according to TNM8, R0 resection (defined as minimum 1 mm margin), lymphovascular and/or perineural invasion in the primary tumor, synchronous/metachronous (cut-off being metastasis diagnosed within 12 months of primary cancer diagnosis), presence of lung metastasis at the time of liver resection, number of liver metastases, largest liver tumor size, administration of neoadjuvant/adjuvant chemotherapy and type of chemotherapy, preoperative carcinoembryonic antigen (CEA) levels, recurrence and date of recurrence or last follow-up for recurrence, death and date of death or last follow-up for death. Follow-up data were collected from medical records and from the Swedish population register. Patients with cancers from the rectum were excluded to make included patient tumor specimen more homogenous and to not include radiated tumors. Resected tumor specimens from included patients were collected from the regional biobank. Patients were included if the preoperative assessment suspected liver metastasis from colon and the postoperative histopathological assessment were consistent. Mucinous tumors (classified as 50% or more of malignant epithelium to contain extracellular mucin according to local routines and WHO criteria) were excluded (26). Patients that did not survive the initial postoperative period (14 days) after surgery were excluded as tumor biology was suspected to have less importance in these cases.
The inclusion/exclusion criteria were pre-established. The study is based on the same individuals as a previous article published by our research group (27).
Histopathological examination. Tissue microarray blocks were generated from paraffin blocks from resected CLM using an automated tissue micro arrayer Minicore 3, Alphelys, Plaisir, France, 4 cores of diameter 2 mm from appropriately marked cancerous areas. The tissue was embedded into a recipient block and finally cut in 4 μm sections. The tumor specimens were incubated in 60° Celsius for 1 h and then deparaffinized in Xylene and rehydrated in descending graded series of ethanol. Prior to immunohistochemical staining, the sections were subject to heat-induced antigen retrieval using EnVision FLEX Target Retrieval Solution low pH (K8005, Dako) in 90° celcius for 20 min in automated PT Link (Dako, Glostrup, Denmark). Afterwards, the slides were three times washed in phosphate-buffered saline (PBS) and subsequently blocked against endogenous peroxidase activity for 10 min in 0.3% hydrogen peroxide and 1% methanol in PBS. Next step was blocking 5% goat normal serum for 1 h. The sections were then incubated in human, polyclonal antibody against ARHGAP4 (dilution 1:200) over night at 4° Celsius. Next day, sections were incubated in goat anti-rabbit biotinylated secondary antibody (Vector Technologies, Burlingame, CA, cat no BA-1000), (dilution 1:200) for 1 h. Afterwards, they were incubated in Avidin/biotin Complex, Vectastain Elite kit (Vector Laboratories, cat no PK-6100) for 30 min. For visualization, the slides were exposed to chromogen Diaminobenzidine (DAB Substrate kit, Peroxidase HRP) (Vector Laboratories, cat no SK-4100) for 5 min and finally washed. The counterstaining for nuclei was performed with hematoxylin. Negative controls were generated by omitting the primary antibody. The slides were scanned for evaluation using Aperio Scanscope scanner (Leica Biosystems, Germany). The tumor specimens were then assessed for staining expression of ARHGAP4 by a senior pathologist who was blinded from outcome data. Sample immunohistochemical images of ARHGAP4 expression are shown in Figure 1. ARHGAP4 levels were divided by staining intensity into two groups (high and low) based on no or mild staining intensity versus moderate or strong intensity. A positive staining reaction was determined if a minimum of 10% of tumor epithelial cells showed a positive staining reaction to ARHGAP4 antibodies. The Anti-ARHGAP4 antibody used was an IgG polyclonal antibody with rabbit host. The antibody and product specifications are available at atlasantibodies.com product number HPA001012.
Representative images of ARHGAP4 high defined as minimum of 10% of tumor epithelial cells showing a positive staining reaction to ARHGAP4 antibodies and low expression defined as less than 10% of tumor epithelial cells showing a positive staining reaction to ARHGAP4 antibodies. (A) High expression of ARHGAP4. (B) Low expression of ARHGAP4.
Statistical analysis. R version 3.5.2 was used to conduct statistical analysis and survival graphs. Hazard ratios (HR) and 95% confidence intervals were calculated using Cox proportional hazard models. Primary outcome was overall survival (OS). HR values were accounted for as the values of the follow-up period. Values with p<0.05 were considered statistically significant. Variables with p<0.05 in univariate analysis were further analyzed in multivariate analysis. In all statistical analyses, a dichotomized variable of low and high ARHGAP4 expression was applied (cut-off 10% of tumor epithelial cells showing positive staining reaction). The data met the assumptions of the statistical tests used in the statistical analysis.
Ethical considerations. This project has ethical approval in accordance with Swedish law (2003:460 §3-4). The research plan was tested and approved by the Regional Ethical Review Board of Lund University (2018/1113). Tumor specimens were collected after approval of collection from the regional biobank representatives, Regionalt biobankscentrum Södra sjukvårdsregionen (SC2879). The study was completed in accordance with the Helsinki Declaration of the World Medical Association. Management of data is consistent with GDPR. The article was written according to the STROBE guidelines (28). All datasets on which the conclusions of the report rely are available on request by contact with corresponding author.
Results
Patient inclusion and exclusion process. A total of 325 patients underwent resection of liver metastasis with the primary cancer located in the colon. Twelve patients were excluded due to ‘disappearing liver metastasis’, i.e., postoperative histopathological examination could not identify any remaining tumor cells after neoadjuvant chemotherapy treatment. Four patients were excluded as they were lost during the follow-up period. A total of 49 patients were excluded during the process of histopathological examination of our research group during the histopathological examination process. This was either because the tumor material could not be collected from the local biobank because the tumor was of mucinous type or because only single cells or small groups of malignant cells remained in the tumor specimen. Nine patients were excluded for procedural reasons during the process of staining for ARHGAP4. A total of 251 patients remained after all incompatible patients were excluded. A flowchart of exclusion process is presented in Figure 2.
Flowchart representing the patient inclusion and exclusion process.
Patient baseline characteristics. Of the included patients, the study cohort consisted of 145 males (58%) and 106 females (42%). The median age at the time of liver resection was 70 years (range=41-87 years). 176 patients (70%) patients underwent neoadjuvant chemotherapy and 175 (70%) adjuvant chemotherapy. Median overall survival was 4.6 years. For patients who did not die during follow up the median follow-up time was 5.1 years. One hundred patients were found to express low levels of ARHGAP4 in the resected tumor specimen (40%) and 151 patients expressed high levels of ARHGAP4 (60%). Baseline characteristics of the 251 final patients are presented in Table I.
Baseline characteristics of included patients collected from medical journals.
Association of ARHGAP4 expression with clinicopathological factors. The associations between ARHGAP4 expression and clinicopathological factors are shown in Table I. ARHGAP4 was significantly associated with site of primary tumor (left/right). ARHGAP4-status was not significantly associated with sex, age, primary node status, number of liver metastases, size of largest tumor, R0-status, synchronous/metachronous disease, presence of extrahepatic disease, vascular/lymphatic invasion, perineural invasion, preoperative CEA levels or administration of chemotherapy in the pre- or postoperative settings.
Prognostic value of ARHGAP4 expression. In Cox univariate analysis comparing low versus high levels of ARHGAP4 in comparison to overall survival by univariate analysis, a statistically significant difference was found (HR=1.5, 95% CI=1.1-2.2). The univariate Cox analysis also found preoperative CEA levels (HR=2.2, 95% CI=1.2-4.0), extrahepatic disease (HR=1.9, 95% CI=1.2-3.0), perineural invasion (HR=1.7, 95% CI=1.2-2.5) and vascular/lymphatic invasion (HR=1.7, 95% CI=1.2-2.4) to be statistically significant for survival after CLM resection. Univariate Cox survival data generated using Cox proportional hazard model are presented in Table II.
Survival data of 251 patients included in final analysis, generated by using Cox proportional hazard model.
Cox multivariate analysis for ARHGAP4 was also found to be statistically significant for survival after surgery (HR=1.5, 95% CI=1.0-2.1). The multivariate Cox analysis also found preoperative CEA levels (HR=2.0, 95% CI=1.1-3.7) and vascular/lymphatic invasion (HR=1.5, 95% CI=1.1-2.2) to be statistically significant for survival after CRLM resection. Multivariate survival data generated using Cox proportional hazard model are presented in Table II.
Kaplan-Meier figures portraying survival probability after surgery in ARHGAP4 low/high groups are presented in Figure 3. The figure displays adverse survival for patients over time if the tumor specimen was found to express high levels of ARHGAP4.
Kaplan-Meier representing patient survival probability in years in association with high respectively low expression levels of ARHGAP4.
Discussion
In this article, we present data from our study analyzing ARHGAP4 protein levels assessed by IHC in a large cohort of CLM and its statistical relationship with survival after surgery. We found ARHGAP4 staining intensity to have a significant correlation with OS by univariate and multivariate analysis. This is the first study to investigate ARHGAP4 as a prognostic marker in CLM. Previously, ARHGAP4 RNA levels have been found to correlate with survival in CRC presented by The Human Protein Atlas (24). A study published in 2022 using GEPIA consistently found ARHGAP4 expression to correlate with adverse prognosis in CRC (25). A previous study has also found indications that high ARHGAP4 levels by immunohistochemical analysis is an independent predictor for poor prognosis in scirrhous-type gastric cancer (29). This study further suggests ARHGAP4 as a potential molecular target for treatment. The ARHGAP4 gene and associated RNA/protein is relatively unstudied so far both as comes intra- and extracellular function and as a potential biomarker in cancer diseases. The described mechanisms of ARHGAP4 in addition to available data suggesting that ARHGAP4 is indicated to have prognostic relevance in CRC as well as other types of cancer reasoned us to believe there is a potential correlation between ARHGAP4 expression and postoperative survival in CLM. The hypothesis that ARHGAP4 expression could serve as a potential marker for prognosis in CLM initiated us to conduct this study. In similarity to previous data, our results indicate that ARHGAP4 could be useful as a novel biomarker in CRC/CLM.
There is currently no reliable method available to sub-categorize patients with CLM based on biological markers. Most of the thoroughly studied biomarkers are still not accessible for use in clinical decision-making. As many existing studies are small and there are significant inter-study differences, thus it can be challenging to interpret and compare study findings between biological markers and clinical outcomes. Possibly, the integration of future projects assessing pre-clinical biological data from CRC/CRLM with clinical outcomes in translational projects could help us better understand the influence of molecular tumor traits on patient outcomes. There is currently a significant unmet need in the production of scientific data of biological markers of CRLM and its prediction and overall understanding of prognosis, therapeutic response and possibly even to identify targets for novel treatments. Due to a lack of knowledge about the biological traits of individual tumors, stratifying patients with CLM for the best possible therapy program remains a consistent challenge. There are data indicating the typical molecular transformation to CRC only occurs in 10% of tumors (typically established as key mutations including oncogenes APS, KRAS, DDC, and tumor suppressor p53) (30). The heterogeneity seen in data from biological research of CRC/CRLM and patient survival may be explained by a number of different molecular mutation pathways. As both the primary and the metastatic tumor separately continue to accrue mutations after the metastatic lesion has developed, there is evidence that CRC and CLM have a complex biological relationship (30). Biological inferences from CRC to CRLM should perhaps be cautiously generalized. Therefore, when future studies of novel biological characteristics are conducted, it could be essential to study both the tumor biology of the primary tumor and the metastatic lesion. A study describing that zinc finger and BTB domain-containing 7A is involved in development and cancer progression of CRC has highlighted the importance of identifying cancer-related genes as prognostic markers to enable individualization of therapy in CRC (31).
Certain limitations in our research must be addressed. There is a potential risk of inconsistencies in collected tumor specimen as they were stored for a long period of time, which in turn could alter the IHC process and in turn evaluation results. Furthermore, neoadjuvant chemotherapy and the liver-first surgery strategy were progressively used more frequently during the study period which, in turn, increased heterogeneity within the group of studied individuals. Most patients received 5-fluorouracil-based chemotherapy in both neoadjuvant and adjuvant settings, but during the later stages of the study combination therapy with Oxaliplatin, Irinotecan, or both became more common. In the later stages of the study period, combined surgical and ablative approaches also increased which further restricts how our study’s findings can be interpreted.
The research studying exact mechanisms of how ARHGAP4 leads to cancer development and progression is limited. Further research is needed to discover the importance of ARHGAP4 expression levels, cellular interactions and signaling pathways influence tumorigenesis in colon liver metastases. The potential of blocking ARHGAP4 signaling as a novel therapeutic strategy is noteworthy. Due to limited research with targeting ARHGAP4 in cancer, no conclusions can be made. However, based on the known functions of ARHGAP4 and its involvement in regulating Rho GTPases, it is possible that modulating ARHGAP4 could have an impact on cancer cell properties. It is, therefore, possible that modulating ARHGAP4 also could have an impact on patient outcome if used as a novel drug. It is important to note that for example blocking of ARHGAP4 signaling could have different consequences in different cancers based on underlying molecular characteristics. Consequently, the potential therapeutic effects of inhibiting ARHGAP4 would depend on presence of ARHGAP4 associated inhibitors and modulators. The efficiency and safety profiles can only be speculated of. Further experimental research is needed to learn more of the effects of ARHGAP4 inhibition and determine its potential as a therapeutic agent.
Conclusion
The present study identified ARHGAP4 as a potential novel prognostic marker for overall survival in resected CLM. Further studies of ARHGAP4 are necessary to further investigate the importance of ARHGAP4 before these findings can be interpreted. There is a substantial need for a more personalized medicine approach in patients with CRLM and further studies of tumor biology in colon cancer and CLM are needed to provide such methods. Biological markers could potentially help clinicians provide tailor-fit patient treatment regimens and manage surgical strategies based on a better prediction of prognosis and outcome after treatment. Further proteogenomic studies may also discover targets for future novel treatments.
Footnotes
Authors’ Contributions
William Torén: Conceptualization (formulation and evaluation of overarching research goals and aims), methodology (development and design of methodology), formal analysis (application of statistical techniques to analyze and synthesize study data), investigation (conducting investigation process and data collection), resources (provision of study materials, reagents, materials, patients), writing – original draft, writing – review and editing, visualization (preparation and creation of the published work, project administration (management and coordination responsibility for the research activity planning and execution. Agata Sasor: Conceptualization (formulation and evaluation of overarching research goals and aims), methodology (development and design of methodology), investigation (conducting investigation process and data collection), writing – review and editing, visualization (preparation and creation of the published work). Daniel Ansari: Conceptualization (formulation and evaluation of overarching research goals and aims), methodology (development and design of methodology), formal analysis (application of statistical techniques to analyze and synthesize study data), resources (provision of study materials, reagents, materials, patients), writing – review and editing, visualization (preparation and creation of the published work), supervision (oversight and leadership responsibility for the research activity), project administration (management and coordination responsibility for the research activity planning and execution). Roland Andersson: Conceptualization (formulation and evaluation of overarching research goals and aims), methodology (development and design of methodology), formal analysis (application of statistical techniques to analyze and synthesize study data), investigation (conducting investigation process and data collection), resources (provision of study materials, reagents, materials, patients), writing – review and editing, visualization (preparation and creation of the published work, supervision (oversight and leadership responsibility for the research activity, including mentorship external to the core team), project administration (management and coordination responsibility for the research activity planning and execution), funding acquisition.
Conflicts of Interest
The Authors declare that no conflict of interest exists.
- Received March 5, 2024.
- Revision received March 29, 2024.
- Accepted April 8, 2024.
- Copyright © 2024 The Author(s). Published by the International Institute of Anticancer Research.
This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY-NC-ND) 4.0 international license (https://creativecommons.org/licenses/by-nc-nd/4.0).
References
In this issue
Jump to section
Related Articles
Cited By...
- No citing articles found.