Abstract
Background/Aim: For patients with local gastrointestinal stromal tumor (GIST), risk stratification is used to assess the prognosis and identify patients to offer adjuvant treatment. For patients with advanced or metastatic GIST, no such risk stratification exists. This study aimed to investigate the prognostic value of 31 different plasma small extracellular vesicles’ (SEVs) surface proteins in GIST patients. Materials and Methods: GIST patients from the two sarcoma centers in Denmark were included. Patients were divided into three groups; group 1: patients undergoing radical surgery; group 2: patients with local, locally advanced, or metastatic GIST; and group 3: patients without evidence of disease after radical surgery. Protein microarray technology was used for the analysis of plasma SEVs. The median plasma SEV marker level was used when comparing groups of patients. The primary endpoint was the progression of GIST. Iterative statistical modeling was used to identify a SEV marker profile/model with a prognostic value. Results: A total of 157 patients were included, with a median follow-up time of 2.05 years. In group 2, a high level of carcinoembryonic antigen (CEA) and a low level of glucose transporter 1 (GLUT-1) were found to be poor prognostic factors [univariate analysis; GLUT-1: hazard ratio (HR)=0.47, 95% confidence interval (CI)=0.22-0.98; CEA: HR=2.12, 95%CI=1.02-4.44]. Composing a model consisting of CEA and GLUT-1 adjusted for age at inclusion was found to have a prognostic value (HR=4.93, 95%CI=2.30-10.57, p<0.0001). Conclusion: Plasma SEVs in GIST showed that CEA and GLUT-1 might be of prognostic value. However, external validation is needed.
Gastrointestinal stromal tumor (GIST) is a mesenchymal tumor with an annual incidence of 10-15 per million inhabitants (1). The primary treatment for GIST patients is surgical resection and adjuvant treatment with imatinib, a tyrosine kinase inhibitor (TKI), for most patients (2, 3). GIST cells often harbor a mutation located in the tyrosine-protein kinase (KIT) or platelet-derived growth factor receptor A (PDGFRA) genes (4). After surgery, well-established risk stratification systems based on tumor size, mitotic rate, location (5), and surgery-related factors are prognostic with regard to the risk of relapse. The risk stratification is used to identify patients with local GIST eligible for adjuvant treatment (3). The first-line treatment for patients with recurrent or metastatic GIST is imatinib, followed by other TKIs at progression (2, 3). However, there is no risk stratification for patients with advanced or metastatic GIST. Furthermore, no soluble biomarkers exist for GIST that can monitor disease activity, help clinical decision-making, and identify patients with poor prognosis.
One interesting biomarker in oncology is small extracellular vesicles (SEVs). SEVs, often termed exosomes, are lipid bilayers containing mRNA, proteins, DNA fragments, and surface proteins reflecting the cells from which they arise (6) and are released into several different types of fluid such as blood (7), bronchoalveolar lavage fluid (8), ascites (9), and cerebrospinal fluid (10). SEVs are identified by surface proteins expressed independently of the cell of origin, such as tetraspanins (CD63, CD9, and CD81) (11) (Figure 1).
An illustration of a tumor and a blood vessel containing several biomarkers: tumor cells, circulating tumor DNA, and small extracellular vesicles. The figure was created with BioRender.com.
It is suggested that SEVs are responsible for removing excess components from the cells and intercellular communication (6). In cancer, SEVs are proposed to influence tumor growth, microenvironment, resistance to oncological treatment (12), immune suppression, and promote tumor cell invasion and metastasis (13). Possible clinical implications of SEVs are diagnostics, prognostics, and assessing the treatment effect (12).
One of the only studies of SEVs in GIST patients was reported by Atay et al. in 2018 (7) and showed that patients with GIST had twice as many SEVs in the blood compared to age-matched healthy controls. Furthermore, compared to primary localized GIST, a significantly higher number of SEVs was found in patients with metastatic disease. However, the surface composition of the individual SEVs in GIST patients has not been described.
This study aimed to investigate the surface composition (phenotype) of the individual plasma SEVs and the prognostic value of SEV surface markers in patients with GIST.
Materials and Methods
This is a prospective, non-randomized, non-interventional, explorative study investigating the prognostic value of plasma SEV phenotypes in patients with GIST.
The study protocol was approved by the Regional Ethics Committee (H-18029854) and the Head of the Knowledge Centre on Data Protection Compliance (P-2019-706). The study was performed with the Good Clinical Practice standard, according to the latest revised Helsinki declaration, and according to national laws. All patients provided signed informed consent before inclusion.
Patients. Patients were included at the Department of Oncology at Herlev & Gentofte Hospital, Department of Oncology at Aarhus University Hospital, and Department of Surgery and Transplantation at Rigshospitalet from January 2019 to December 2021. Patients planned for surgery for local disease with a GIST of ≥2 cm had blood samples collected pre-operative and one day post-operative. Patients diagnosed with GIST were included independent of disease or treatment status in the oncological departments with the following exceptions: patients who stopped adjuvant treatment more than two years ago were excluded, only patients starting on adjuvant treatment or having a maximum of six months left of the adjuvant treatment were enrolled. Blood samples were collected at inclusion and synchronized with every control scan, usually every third month.
The patients were allocated into three groups (Figure 2). Group 1 included patients undergoing radical surgery, group 2 included patients with local, locally advanced, or metastatic disease, and group 3 included patients with no sign of disease (patients in adjuvant treatment after radical surgery or in control after ended adjuvant treatment). Group 1A included the pre-operative blood samples from patients in group 1, and group 1B included the postoperative blood samples in group 1.
Flow chart of included patients. Group 1: patients undergoing radical surgery; group 2: patients with local, locally advanced, or metastatic disease; and group 3: patients with no sign of disease (patients in adjuvant treatment after radical surgery or in control after ended adjuvant treatment). *Thirty-one patients progressed, but only progression samples from 11 patients were available. **Three patients progressed, but only progression samples from two patients were available.
Blood sampling. Blood was collected from the patients in 3.5 ml sodium citrate tubes. All blood samples were handled by the Danish CancerBiobank, Bio- and GenomeBank, Denmark. The maximum time from blood sampling to centrifugation was 4 h; subsequently, the time to storage was 1 to 2 h. Plasma was isolated through centrifugation at 2,000 or 2,500 g for 10 min. After centrifugation, plasma was transferred to another tube and mixed lightly. The plasma was stored at −80°C until use.
EV array. The analysis for SEVs was performed using the extracellular vesicle (EV) array, based on protein microarray technology (14). The method is used to determine the phenotype of unpurified plasma SEVs or other EVs. In this study, the phenotype is defined as the protein composition of the individual SEV.
Production of antibody microarray. The antibody microarrays were produced on epoxy-coated slides (75.6×25.0 mm; SCHOTT Nexterion), and the printing of the antibodies was performed with a sciFLEXARRAYER S12 micro-array printer installed with a piezo dispense capillary (PDC) size 60 with coating type 3 (Scienion AG, Berlin, Germany). Printing buffer consisted of 50 mM trehalose in phosphate-buffered saline (PBS) throughout the experiment. As positive controls (Figure 2), 10 or 20 μg/ml biotinylated goat anti-mouse IgG (H+L) antibody (Novus Biologicals, Centennial, CO, USA) was printed, and a printing buffer was used as a negative control. The 34 anti-human antibodies used are listed in Table I and were printed at 200 μg/ml. See Figure 3 to visualize the print.
The function of the protein markers used in the analysis of the small extracellular vesicles.
Visualization of the extracellular vesicle array print. (A) The antibodies or controls were printed in triplicates in each well. The positive controls K10 and K20 (green) consist of goat anti-Mouse IgG (H+L) secondary antibody (biotin) (NB7537) from Novus with known concentrations. The blank well of the print was the negative control. CD9 and CD81 (blue) are the SEV markers used to normalize the data. (B) An example of the visualization of the spots after the scan. PBS: Phosphate-buffered saline; CD: cluster of differentiation; CEA: carcinoembryonic antigen; Hsp90: heat shock protein 90; PLAP: placental alkaline phosphatase; NY-ESO-1: New York esophageal squamous cell carcinoma 1; TSG101: tumor susceptibility gene 101; CTLA-4: cytotoxic T-lymphocyte-associated protein 4; VEGFR: vascular endothelial growth factor receptor; PD-L1: programmed death-ligand 1; AREG: amphiregulin; GLUT: glucose transporter; TGF: transforming growth factor; TRAIL: TNF-related apoptosis-inducing ligand; L1CAM: L1 cell adhesion molecule.
Catching and visualization of SEVs. The EV Array analysis was performed as described by Jørgensen et al. (15) with modifications. In short, the printed microarray slides were initially blocked (50 mM ethanolamine, 100 mM Tris, 0.1% SDS, pH 9.0) before incubation with a 15 μl plasma sample diluted to 100 μl in wash-buffer (0.05% Tween20 in PBS). The same volume of plasma was used from each patient. The incubation was performed in Multi-Well Hybridization Cassettes (ArrayIt Corporation, Sunnyvale, CA, USA) at room temperature for 2 h, followed by overnight incubation at 4°C. After the 31 cancer-specific antibodies coated on the microarray slides caught the SEVs, biotinylated detection antibodies (antihuman-CD9, -CD63, and -CD81, LifeSpan BioSciences, Seattle, WA, USA) diluted 1:1,500 in a wash-buffer followed by 30 min incubation with Cy5-labelled streptavidin (Life Technologies, ThermoFisher Scientific, Waltham, MA USA) diluted 1:1,500 were used for visualization. Before scanning, the slides were washed in wash buffer, then in ultrapure/deionized water, and finally dried using a Microarray High-Speed Centrifuge (ArrayIt Corporation).
Slides were scanned in an InnoScan 710 AL microarray scanner (Innopsys, Carbonne, France) with the following settings: 532 nm at 10 V, PTM at 100%, and 5 μm resolution. The spots were visualized in Mapix (microarray analysis software, Innopsys) (Figure 3). For analyzing the total intensity at a spot, a GenePix Array List (GAL) file containing the data was used with a constant diameter (Ø135 μm). Through manual examination, contaminated spots were identified and deleted.
Data normalization. Quality control of raw data ensured that the intensity of the triplicate of a protein marker was within a reasonable range. The mean intensity for each protein marker for each sample was calculated. If the relation between the positive control (K20) and the negative control (blank) was >0.97, the sample was considered acceptable.
The total intensity of a protein marker was calculated as follows: the patient sample’s intensity at a specific protein marker minus the blank well’s intensity on the slide divided by the patient’s background intensity at the negative control in the well (blank spot). Intensity values lower than 1, meaning that the signal for a protein marker was lower than the background signal for a patient, were removed from the dataset. Subsequently, the data were log2 transformed. Since CD81, CD9, and CD63 were used to identify the SEVs, data regarding these SEV markers are not reported. The plasma SEV levels of the different SEV markers were normalized to CD9 and CD81 by dividing each SEV marker level with the geometric mean of CD9 and CD81. Performing a t-test showed no difference in the mean of CD9 (p=0.56) and CD81 (0.47) when comparing group 1B and group 2.
Statistical analyses. The primary endpoint was the progression of GIST as per Response Evaluation Criteria in Solid Tumours 1.1. (16) and GIST-related death. The data analysis cut-off date was 30 March 2022.
The Kruskal–Wallis H test was used to compare continuous variables through the one-way analysis of variance between groups. The t-test was used to compare categorical variables between groups. The Wilcoxon matched pairs signed rank test was used to compare categorical variables at repeated measurements for patients since the difference in means between the measurements were not normally distributed.
Univariate and multivariate analyses were conducted using the Cox regression model. The multivariate analyses included the individual SEV markers, age at inclusion (continuous variable), and sex (categorical variable) were incorporated. For each SEV profile, for which the development is described below, different prognostic models were tested against each other.
The signal intensity of each SEV marker was categorized into a low and high value based on the median value (intensity of the spot) in the group investigated. A poor prognosis was assigned a value of 1, and a good prognosis was assigned a value of 0. The sum of the assigned values for each SEV included in the profile was calculated, and the profile was then dichotomized. Only SEV markers significant in the univariate analysis were incorporated in a prognostic profile.
In the prognostic models confounding variables, age at inclusion and sex were included along with the SEV profile. We used Harrell’s C statistics, Akaike information criterion (AIC), Bayesian information criterion (BIC), and the likelihood ratio to find the best prognostic model. The total number of events (n=31) restricted the maximum number of variables (n=3) incorporated in the model testing analysis.
Since we performed multiple testing, a significance level of 0.0015 would be preferable (0.05 divided by 31 SEV markers used in the study). However, since this is an explorative and hypothesis-generating study, we accepted 0.05 as a significance level.
Stata v. 17 (StataCorp LLC, College Station, TX, USA) was used for the data analysis, and Graphpad Prism v. 9 (San Diego, CA, USA) was used for ROC curve analysis.
Results
Patient characteristics. A total of 157 patients were included in this study, with a median follow-up time of 2.05 years.
Patient and disease characteristics at the inclusion time are summarized in Table II. For patients with either local, locally advanced, or metastatic disease (group 2), 31 progressed or died during the follow-up. In group 3, three patients progressed during the follow-up, while no patient progressed in group 1.
Patient characteristics at the time of inclusion for patients undergoing radical surgery (group 1), patients with active gastrointestinal stromal tumors (GIST) (group 2), and patients without evidence of disease (group 3).
SEVs. Univariate and multivariate analyses of risk for progression were performed for each protein investigated (Table III). Glucose transporter 1 (GLUT-1) and carcinoembryonic antigen (CEA) were the only SEV markers found significant on a 0.05 level in the univariate- [GLUT-1; hazard ratio (HR)=0.47, 95% confidence interval (CI)=0.22-0.98, p=0.043, CEA; HR=2.12, 95%CI=1.02-4.44, p=0.045] and multivariate analysis (GLUT-1; HR=0.39, 95%CI=0.18-0.85, p=0.018, CEA; HR=2.14, 95%CI=1.02-4.47, p=0.044). A low level of GLUT-1 and a high level of CEA was related to poor prognosis in the uni- and multivariate analyses. The Kaplan–Meier plots of CEA low/high and GLUT-1 low/high are shown in Figure 4. There was no relation between the value of GLUT-1 or CEA and patient age at study inclusion.
Univariate and multivariate analyses of the small extracellular vesicle (SEV) markers at the time of inclusion for patients with gastrointestinal stromal tumors (group 2).
Kaplan–Meier plots for time to progression from the time of inclusion for patients in group 2. (A) All patients in group 2. (B) The patients were stratified into CEA low (↓) and CEA high (↑). (C) The patients were stratified into GLUT-1 low (↓) and GLUT-1 high (↑). (D) Profile A1 (good prognosis): one of the following scenarios is present; the GLUT-1 was high (GLUT-1↑) and the CEA was low (CEA↓), or GLUT-1 was high (GLUT-1↑) and the CEA was high (CEA↑), or the GLUT-1 was low (GLUT-1↓) and the CEA was low (CEA↓). Profile A2 (poor prognosis): GLUT-1 was low (GLUT-1↓), and the CEA was high (CEA↑). *Low (↓): The SEV marker value was lower or equal to the median value of the SEV marker in group 2. High (↑): The SEV marker value was higher than the median value of the SEV marker in group 2.
The SEV markers from the univariate analysis with a p-value ≤0.05 were incorporated into a profile. Patients with active GIST (group 2) were divided into a good (profile A1) and poor (profile A2) prognosis profile. A comparison of profiles A1 and A2 showed that the profile is of prognostic value (HR=4.17, 95%CI=1.99-8.74, p<0.0001). The accompanying Kaplan–Meier plot illustrating profiles A1 and A2 is shown in Figure 4.
Of the 122 patients in group 2, 98 patients belonged to profile A1, and 19 of these had disease progression after inclusion into the study (19.38%). Of the 24 patients belonging to profile A2, 12 progressed (50%).
A profile B containing the five SEV markers with a p≤0.2 from the univariate analysis: CEA, GLUT-1, Flotillin-1, PD-L1, and CD105 was created. We divided patients into a poor (profile B1) and a good (profile B2) prognosis profile (Table IV). A comparison of profiles B1 and B2 showed that the profile is of prognostic value (HR=0.37, 95%CI=0.18-0.76, p=0.006).
Comparison of prognostic profiles/models A-E for patients with gastrointestinal stromal tumors (group 2).
To find the best possible model based on profiles A and B, different models were tested against each other, and model E was the best obtainable prognostic model. Age at inclusion was not a significant factor in univariate analysis (p=0.16), but in a multivariate analysis in model E, age at inclusion was of significant prognostic value (p=0.036). Sex was not an independent prognostic factor in univariate analysis (p=0.86).
The CD163 intensity on SEVs was found to be significantly lower at progression than at the time of inclusion (p=0.032) for patients with matched samples at the time of inclusion and at tumor progression (13 samples) (Table V). Two SEV markers, CD163 (p=0.027) and CD56 (p=0.029), were found to have significantly lower intensity post-operative than preoperative in patients undergoing radical surgery for GIST (n=15) (Table V).
Small extracellular vesicle (SEV) marker levels.
The median SEV marker levels in group 1B were compared with those in patients with active GIST (group 2) (Table VI). The median intensity of CD42a in SEVs from patients with active GIST was significantly higher than that in patients without evidence of disease (p=0.008) (Figure 5). The receiver operating characteristics (ROC) curve for CD42a is shown in Figure 6, where the area under the curve was 0.70 (95%CI=0.58-0.82, p=0.011), indicating that it is a moderate marker.
Small extracellular vesicle (SEV) marker level at the time of inclusion for patients that had undergone radical surgery and were without evidence of disease (group 1B) vs patients with active gastrointestinal stromal tumors (GIST) (group 2).
The plot visualizes the CD42a values in the group of postoperative samples from patients that had undergone radical resection of gastrointestinal stromal tumor (GIST) (group 1B) and the group of patients with active GIST (group 2). The horizontal lines mark the median values in the two groups.
The receiver operating characteristics curve for CD42a.
Discussion
In this national study, we investigated the prognostic value of the phenotype of plasma SEVs in patients with GIST. Our study showed that patients with active GIST having a high CEA value and/or a low GLUT-1 have a significantly higher risk of progression or death. Age and sex, which are possible confounders, were incorporated in the multivariate analysis with the SEV markers. Disease status, another potential confounder, was not incorporated in the analysis since we already had selected the groups of interest for the analysis based on disease status. We also found a highly significant prognostic model, including CEA, GLUT-1, and age at inclusion.
In GIST, risk stratification is based on tumor size, location, mitotic count (5), and KIT and PDGFRA mutational status (17, 18), which are factors proved to have a prognostic value in patients with resectable GIST. Only KIT and PDGFRA mutational status (17, 18) is a factor with prognostic value in patients with advanced and metastatic GIST. However, no soluble biomarker is known to be of prognostic value for patients with GIST.
Most cells produce SEVs (19), and it has been proposed that cancer patients have more SEVs than healthy individuals (20). The SEVs have been shown to have clinical utility in some cancer types. For example, the National Comprehensive Cancer Network has included SEV-derived biomarkers (RNA) from urine to be considered for early detection of prostate cancer (21). SEV phenotyping has been shown to separate patients diagnosed with advanced-stage non-small cell lung cancer from matched controls with 75.3% accuracy (22). The prognostic value of SEV phenotyping has not been investigated in GIST.
The scope of this study was to investigate the prognostic potential of SEV phenotypes in patients with GIST based on a blood sample and not to gain a quantitative measure of the SEVs. Therefore, it was chosen to use an already established and verified technology and not to focus on the EV characteristics despite the recommendations by the minimal information for studies of extracellular vesicles guidelines (23). The protein microarray (EV Array) technology used in our study did not allow quantitative measurements of the SEVs in contrast to the Nanoparticle Tracking Analysis (NTA) used in the study by Atay et al. (7).
The two SEV surface proteins found in this study to have prognostic importance were CEA and GLUT-1. CEA is an unspecific biomarker, elevated in several cancer types and other conditions such as uremia, lung fibrosis, and is also associated with age (24). CEA belongs to the family with the same name, which in turn belongs to the immunoglobulin superfamily (25). CEA is widely used in the surveillance of colorectal cancer as a prognostic biomarker. However, the specificity and sensitivity are too poor to function as a diagnostic biomarker (24). In colon cancer, overexpression of CEA increases the adhesion of the cancer cells through selectins and, thereby, enhances the metastatic process (26).
GLUT-1 is a glucose transporter belonging to the family GLUT (27). Glucose transport into the cell is essential to maintain a high cell proliferation rate (27). GLUT-1 is often overexpressed in cancer cells (27) and associated with poorer survival in patients with solid tumors (28). Examples of cancer types with overexpression of GLUT-1 are colorectal cancer, breast cancer, and lung cancer (27). The expression levels of GLUT-1 and the SEV levels of GLUT-1 cannot be compared directly.
The SEV intensity of CEA and GLUT-1 was not significantly different when comparing samples obtained at the time of inclusion with samples obtained at image-verified tumor progression.
Another surface protein, CD163, was, however, found to have a significantly lower intensity at the time of progression than at the time of inclusion. Furthermore, the CD163 and the CD56 SEV intensity was significantly lower postoperatively compared to preoperatively. CD163 belongs to the scavenger receptor cysteine-rich receptors and is often expressed on macrophages (29). In a laboratory study, the CD163-induced activation of macrophages has been associated with tumor development (30) and is believed to have an immunosuppressive effect. Furthermore, a high rate of CD163 expressing tumor-associated macrophages has also been associated with poorer overall survival in sarcoma patients (30).
CD56 is a member of the immunoglobulin superfamily (31) and is primarily expressed on natural killer cells where it is believed to aid in the adhesion to target cells (32). CD56 expression has been correlated to a poor prognosis in patients with renal cell carcinoma (31) and non-small cell lung cancer (33). Our findings regarding SEV CD163 and CD56 intensity are non-conclusive due to the small number of patients. The relation of these SEV markers to the innate immune response (macrophages and natural killer cells) in patients with GIST should be further investigated.
Apart from the SEV markers with prognostic potential in patients with GIST, we also found that patients with active GIST had a significantly higher median level of plasma SEV CD42a than the group of postoperative samples from patients radically resected for GIST. These results could imply that a higher plasma SEV CD42a level indicates the presence of GIST cells in the body. However, no such relation was found when comparing SEV marker levels between pre- and postoperative samples, keeping the small number of patients in mind together with the early drawing of the post-operational blood samples. The antigen CD42a is also called glycoprotein IX (GPIX). Together with the GPIbα, GPIbβ, and GPV, GPIX constitutes the GPIb-IX-V complex, which is expressed on the platelet surface (34). The GPIb-IX-V complex is essential in platelet functions in adhesion, activation, and aggregation, which is essential for hemostasis (35). Platelets are reported to play a role in tumor growth and metastasis (36). In mice models, platelets are believed to promote metastases, for example, by hiding the tumor cells from the natural killer cells together with fibrin (37).
This study indicates that SEV phenotyping may have prognostic value in patients with GIST. The relatively short follow-up time (median 2.05 years) and the small number of events (31 patients with progression or death due to GIST after inclusion) affect the power of the results. The results should therefore be interpreted with caution.
Some limitations should be considered. The patient cohort is heterogenous regarding disease status, treatment status, and the time for inclusion in the disease course. The inclusion in the study at different time points in the patients’ disease courses has led to a high number of included patients but complicates data interpretation. The disease status ranges from patients in adjuvant treatment after a radical surgery to patients in lifelong treatment due to metastatic GIST. We used the SEV marker levels in postoperative blood samples from patients that had undergone radical surgery and were without evidence of disease (group 1B) as a comparison to the SEV marker levels in patients with active GIST in this study. This comparison was made to investigate if the SEV marker levels could distinguish patients with GIST cells from patients not having GIST cells in the body and, thereby, if any of the SEV marker levels investigated could be a potential diagnostic biomarker. During the study’s follow-up time, none of the patients undergoing surgery had a relapse of the disease. The trauma caused by surgery could, however, potentially influence the amount and the phenotype of the SEVs. Due to the lack of a healthy control group, we cannot conclude anything regarding the diagnostic potential of SEV markers in patients with GIST.
The present study has several strengths. This study is a nationwide study, including a high number of patients in a clinical setting. This is also the first study investigating the plasma SEV phenotype within patients with GIST. The study of SEVs could also help the understanding of the immunological status of GIST patients. This study suggests that the innate immune system plays a role in tumor progression and initiation. This could explain the lack of immune checkpoint inhibitor effect in GIST. We will investigate this in future studies.
Since no risk stratification is available for patients with advanced or metastatic GIST, a soluble prognostic biomarker would be of great interest to this group of patients. Our study results, however, need external validation in a larger, well-defined independent cohort to validate the prognostic role of CEA and GLUT-1 on SEVs in GIST patients; if confirmed, this could be a new prognostic marker for metastatic GIST.
Conclusion
This is the first study investigating the phenotype of plasma SEVs in GIST patients. The study showed that a high CEA and/or a low GLUT-1 is associated with a poor prognosis. We also report a highly statistically significant prognostic model containing CEA, GLUT-1, and age at inclusion in the study. However, external validation is needed.
Acknowledgements
The Authors would like to thank the Department of Surgery and Transplantation at Rigshospitalet, Copenhagen, Denmark, for the collaboration with this study and acknowledge the help the nurses at the department provided for handling the logistics during the study. Furthermore, we would like to thank the clinical research unit at Aarhus University Hospital for handling the logistics at the Department of Oncology at this site. The Danish CancerBiobank is acknowledged for handling and storing biological material. Figure 1 was created with BioRender.com.
Footnotes
Authors’ Contributions
Conceptualization, C.M.B., A.K.H., N.A.P., and E.H.; methodology, C.M.B., N.A.P., A.K.H., E.H., R.B. and M.M.J.; validation, C.M.B., and N.A.P.; formal analysis, C.M.B., and N.A.P.; investigation, C.M.B., N.A.P., A.K.H., B.E.E., P.B.R., P.D.H., L.P., and H.J.M.; resources, M.M.J., R.B., and H.J.M.; data curation, C.M.B., N.A.P., and A.K.H.; writing—original draft preparation, C.M.B., and N.A.P.; writing—review and editing, C.M.B., N.A.P., A.K.H., E.H., B.E.E., P.B.R., H.J.M., P.D.H., L.P., M.M.J., and R.B.; visualization, C.M.B., N.A.P., A.K.H., and E.H.; supervision, N.A.P., A.K.H., and E.H.; project administration, C.M.B., N.A.P., A.K.H., and E.H.; funding acquisition, C.M.B., and A.K.H. All Authors have read and agreed to the published version of the manuscript.
Conflicts of Interest
The Authors declare no conflicts of interest in relation to this study.
Funding
This research was funded by Candys Foundation, grant number 2019-332, and the Danish Cancer Society, grant number R248-Ai4683.
- Received October 20, 2022.
- Revision received November 1, 2022.
- Accepted November 2, 2022.
- Copyright © 2022 International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved.
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.