Biomolecules & Therapeutics 2024; 32(3): 379-389  https://doi.org/10.4062/biomolther.2023.159
Metabolomic Profiles in Patients with Cervical Cancer Undergoing Cisplatin and Radiation Therapy
Seo-Yeon Choi1, Suin Kim1, Ji-Young Jeon2, Min-Gul Kim2,3,4, Sun-Young Lee4,5,* and Kwang-Hee Shin1,*
1College of Pharmacy, Research Institute of Pharmaceutical Sciences, Kyungpook National University, Daegu 41566,
2Center for Clinical Pharmacology, Jeonbuk National University Hospital, Jeonju 54907,
3Department of Pharmacology, School of Medicine, Jeonbuk National University, Jeonju 54907,
4Research Institute of Clinical Medicine of Jeonbuk National University-Biomedical Research Institute of Jeonbuk National University Hospital, Jeonju 54907,
5Departments of Radiation Oncology, School of Medicine, Jeonbuk National University, Jeonju 561712, Republic of Korea
*E-mail: sylee78@jbnu.ac.kr (Lee SY), kshin@knu.ac.kr (Shin KH)
Tel: +82-63-250-2183 (Lee SY), +82-53-950-8582 (Shin KH)
Fax: +82-63-250-1192 (Lee SY), +82-53-950-8557 (Shin KH)
Received: September 12, 2023; Revised: November 13, 2023; Accepted: November 30, 2023; Published online: April 9, 2024.
© The Korean Society of Applied Pharmacology. All rights reserved.

This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/4.0/) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.
Abstract
This study was aimed to evaluate endogenous metabolic changes before and after cisplatin and radiation therapy in patients with cervical cancer via untargeted metabolomic analysis using plasma samples. A total of 13 cervical cancer patients were enrolled in this study. Plasma samples were collected from each patient on two occasions: approximately one week before therapy (P1) and after completion of cisplatin and radiation therapy (P2). Of the 13 patients, 12 patients received both cisplatin and radiation therapy, whereas one patient received radiation therapy alone. The samples were analyzed using the Ultimate 3000 coupled with Q ExactiveTM Focus Hybrid Quadrupole-OrbitrapTM mass spectrometry (Thermo Fisher Scientific, Waltham, MA, USA). Chromatographic separation utilized a Kinetex C18 column 2.1×100 mm (2.6 μm) (Phenomenex, Torrance, CA, USA), and the temperature was maintained at 40°C. Following P2, there were statistically significant increases in the concentrations of indoxyl sulfate, phenylacetylglutamine, Lysophosphatidyethanolamine (LysoPE) (18:1), and indole-3-acetic acid compared with the concentrations observed at P1. Specifically, in the human papillomavirus (HPV) noninfection group, indoxyl sulfate, LysoPE (18:1), and phenylacetylglutamine showed statistically significant increases at P2 compared with P1. No significant changes in metabolite concentrations were observed in the HPV infection group. Indoxyl sulfate, LysoPE (18:1), phenylacetylglutamine, and indole-3-acetic acid were significantly increased following cisplatin and radiation therapy.
Keywords: Cervical cancer, Metabolomics, Cisplatin, Biomarker
INTRODUCTION

Cervical cancer is the second most prevalent cancer among women and the third main cause of mortality worldwide (Brown et al., 2019). Patients with progressive or recurrent cervical cancer face a poor prognosis, with a 1-year survival rate ranging from only 10%-20%. Chemotherapy is currently the standard treatment for these patients with poor prognosis (Zhu et al., 2016). Cisplatin, a platinum-based anticancer drug, is commonly used as the first-line treatment for patients with cervical cancer (Brown et al., 2019). However, its efficacy varies, as only 29.1%-67% of patients respond to cisplatin treatment, with a median overall survival of 12.87 months. Especially, the development of acquired resistance to cisplatin was known to significantly compromises its therapeutic effectiveness (Zhu et al., 2016).

Anticancer drug therapeutic effects can be evaluated using established tumor prognostic markers. In a previous study, the concentration of squamous cell carcinoma (SCC) antigen was identified as a relevant tumor marker in cervical cancer (Gadducci et al., 2008). Patients with cervical cancer responsive to anticancer treatment exhibited a decrease in serum SCC antigen concentration, suggesting its potential as a treatment response indicator (Gadducci et al., 2008). Additionally, in gynecological cancer, a decrease in the plasma concentrations of cancer antigen 125 (CA-125) and carcinoma embryonic antigen (CEA) has been used as prognostic biomarkers for anticancer therapy (Dasari et al., 2015).

Cisplatin is known to cause various side effects, as demonstrated in several studies; however, clinical research data on the metabolomics approach concerning cisplatin and radiation therapy remain scarce. The side effects associated with cisplatin include nephrotoxicity, neurotoxicity, bone marrow toxicity, and ototoxicity (Ezaki et al., 2017). Recently, novel biomarkers for cisplatin-induced nephrotoxicity were reported, namely miR-130a and ARL13B, both of which exhibited approximately 10-fold and 15-fold increases after cisplatin treatment relative to the pretreatment levels, respectively (Pabla and Dong, 2008; Ezaki et al., 2017; Holditch et al., 2019). In addition to miR-130a and ARL13B, other markers for cisplatin-induced nephrotoxicity include creatinine, blood urea nitrogen, and cystatin-C in serum (Holditch et al., 2019). The results of these previous studies were all conducted in vivo models.

Metabolomic research plays a crucial role in identifying relevant metabolites and discovering biomarkers, which can facilitate the establishment of a correlation between metabolites and phenotypes across different biological samples. Currently, metabolomic approaches have been extensively studied for various aspects of cancer, such as prevention, diagnosis, treatment, and prediction of disease progression (Zhu et al., 2020). Specifically, lactate, aspartate, proline, and glutamate have been investigated as potential biomarker candidates to differentiate between healthy individuals, individuals with cervical cancer, and those with early-stage cervical intraepithelial neoplasia (CIN) (Khan et al., 2019). These metabolites hold promise in advancing our understanding and management of cervical cancer and its precursor stages.

Recently, mass spectrometry-based high-throughput analytical technology has been widely applied to identify new biomarkers of disease from biological samples (Zhu et al., 2020). Metabolome profiling techniques using biofluids, such as serum or plasma enable systematic screening and fingerprinting of metabolites with molecular weight <1,000 Da associated with the alterations of metabolic signature and pathway in various stages of disease, leading to the discovery of potential biomarkers and their underlying mechanisms (Chen et al., 2022).

Metabolic biomarkers for cervical cancer provide information pertaining to the progression and severity of the disease. For instance, lactate, produced as a byproduct of anaerobic metabolism, shows increased levels in cancer cells due to their high metabolic demands. Elevated levels of lactate have been associated with a higher risk of metastasis in cervical cancer and poorer prognosis (Schwickert et al., 1995). Furthermore, the concentration of tryptophan in patients with cervical cancer was found to be lower compared with the healthy control group. This decrease in tryptophan levels might be attributed to its consumption by cancer cells or altered metabolism within the tumor microenvironment (Hascitha et al., 2016). On the contrary, some metabolites were suggested as a biomarker of the effectiveness of anticancer agents. A study reported urinary diacetylspermine as a metabolic biomarker of doxorubicin effectiveness in triple-negative breast cancer (Velenosi et al., 2022). In another in vitro study, enhanced glycolysis contributed to the reduced sensitivity to arabinofuranosyl cytidine (Ara-C), whereas the inhibition of glycolysis hindered cell proliferation of acute myeloid leukemia (AML) and potentiated the cytotoxicity of Ara-C (Chen et al., 2014).

This study aimed to profile changes in endogenous metabolites before and after cisplatin and radiation therapy using the ultrahigh-performance liquid chromatography-tandem high-resolution mass spectrometry (UHPLC-HRMS) technique, utilizing plasma samples from patients with cervical cancer.

MATERIALS AND METHODS

Study design and sample collection

A total of 27 patients with cervical cancer were enrolled in this study at Jeonbuk National University Hospital. All participants provided written informed consent before their involvement in the study. The protocol of this study was reviewed and approved by the Institutional Review Board of Jeonbuk National University Hospital, Jeonju, Korea (IRB No. CUH 2017-04-018). Patients diagnosed with cervical cancer were eligible for enrollment, whereas those who were cured, pregnant, and diagnosed with multiple cancers were excluded from this study subject. Plasma samples were collected from the patients with cervical cancer on two occasions: approximately one week before commencing any anticancer treatment (P1) and after completing either cisplatin and radiation therapy or radiotherapy alone (P2). A total of 13 patients whose samples were collected at both P1 and P2 were included in metabolomics analysis to evaluate the changes in metabolites induced by cisplatin and radiotherapy or radiotherapy alone (Fig. 1). Of 13 patients, 12 received a combination of cisplatin and radiation therapy, whereas one patient received radiation therapy alone (Table 1).

Figure 1. Flow chart of participants included in the study.

Table 1 Characteristics of cervical cancer patients used in comparison P1 and P2

VariableCervical cancer (n=13)Fold change (P2/P1)p-value
P1P2MeanSD
No. of patients1313--
Age (years)61.7 (47-85)--
Height (cm)153.2 (141.1-162.8)--
Weight (kg)59.8 (46.1-81.8)--
Treatment, n (%)
Chemotherapy-12 (92.3)--
Radiotherapy-13 (100)--
AST (IU/L)37 (17-112)28.5 (13-100)0.90.30.069a
ALT (IU/L)22.1 (13-42)20.6 (12-42)0.90.30.118a
Creatinine (mg/dL)0.8 (0.49-1.19)0.8 (0.5-1.19)1.10.40.509b
Hb (g/dL)12.9 (11.1-14.5)11.1 (8.4-12.4)0.90.10.001a*
HCT (%)38.6 (33.5-43)32.0 (25.2-37.1)0.80.1<0.001b*
HPV infection, n (%)
Positive5 (38.5)
Negative8 (61.5)
Menopause, n (%)
Yes10 (76.9)
No3 (23.1)
Stage, n (%)
I0 (0.0)
II9 (69.2)
III2 (15.4)
IV0 (0.0)
Unknown2 (15.4)

Data are presented as mean (range) in age, height, weight, AST, ALT, creatinine, Hb, and HCT.

P1: before treatment, P2: after treatment, n: number of patients, AST: aspartate aminotransferase, ALT: alanine transaminase, Hb: hemoglobin, HCT: hematocrit, HPV: human papillomavirus, SD: standard deviation.

aWilcoxon-rank test, bpaired t-test. *p<0.05, statistically significant.



Following cisplatin and radiation therapy, the therapeutic effects on these patients were assessed using the Response Evaluation Criteria in Solid Tumors (RECIST) from the World Health Organization (WHO). Among the treated patients, four had complete remission (CR), indicating a complete disappearance of all target lesions; two patients had partial remission (PR), indicating a reduction in the size of target lesions by at least 30%; one patient showed stable disease (SD), which means no significant change in the size of target lesions; and one patient was classified as having progressive disease (PD), indicating an increase in the size of target lesions or the appearance of new lesions. The therapeutic effects were not evaluated for five patients (Supplementary Table 1).

Chemical and reagents

Indoxyl sulfate, phenylacetylglutamine, deoxycholic acid, glycodeoxycholic acid, 2-oxindole, pyroglutamic acid, trans-3-indoleacrylic acid, indole-3-acetic acid and lysophosphatidylethanolamine (LysoPE) (18:1) were purchased from Merck Inc. (Darmstadt, Germany). The donepezil-d4 and thioctic acid-d5, the internal standard for quantification of metabolite, were ordered from C/D/N isotopes (Pointe-Claire, Canada). Liquid chromatography-mass spectrometry (LC-MS) grade reagents, including acetonitrile, deionized water, and methanol were purchased from Merck Inc. Formic acid was purchased from Sigma-Aldrich (Steinheim, Germany).

Sample preparation

Plasma was obtained from 7 mL of whole blood using density gradient centrifugation with Ficoll-Paque™ PLUS (GE Healthcare, Piscataway, NJ, USA) and stored at –70°C. To prepare the samples for analysis, the plasma was thawed at room temperature. Subsequently, 50 μL of the thawed plasma was transferred to a 1.5 mL e-tube, and 200 μL of 100% acetonitrile was added. The mixture was thoroughly mixed for 5 min using a vortex mixer. Following this, the sample was centrifuged at 13,000 g for 5 min at 4°C. The resulting supernatant from the pretreated sample was transferred to a vial, and 5 μL of this solution was injected into the liquid chromatography-tandem mass spectrometry (LC-MS/MS) for analysis.

Ultrahigh pressure liquid chromatography-high resolution mass spectrometry (UHPLC-HRMS) analysis condition

The sample analysis was conducted using an Ultimate 3000 (Thermo Fisher Scientific, Waltham, MA, USA) coupled with Q ExactiveTM Focus Hybrid Quadrupole-OrbitrapTM mass spectrometry (Thermo Fisher Scientific). Chromatographic separation was achieved using a Kinetex C18 column 2.1×100 mm (2.6 μm) (Phenomenex, Torrance, CA, USA), with the column temperature maintained at 40°C. The mobile phase A consisted of 0.1% formic acid in water (v/v), whereas the mobile phase B consisted of 0.1% formic acid in acetonitrile (v/v). The flow rate was set at 0.3 mL/min for a total runtime of 20 min, following the gradient: 5% B for 0-1.5 min, 5-95% B for 1.5-8 min, 95% B from 8-15 min, 95-5% B for 15-16 min, 5% B for 16-20 min. All samples were analyzed in both electrospray ionization positive (ESI+) and negative (ESI−) modes. To ensure system suitability, quality control (QC) samples were created by pooling samples from all patients and used for confirmation. The scan ranges were set from 100-1,000 m/z, and the mass resolution was set at 70,000 for full MS scans and 17,500 for MS2 scans. To prevent the influence of drugs on the results during the analysis process, an exclusion list was created for the mass values of the drugs administered to the gynecological patients, which were excluded from the analysis (Supplementary Table 2).

Data processing and multivariate analysis

The raw data obtained from the LC-MS/MS analysis was processed using Compound Discoverer 3.3 (Thermo Fisher Scientific). The workflow was designed to encompass untargeted metabolomics, and it was processed using various online libraries, including mzCloud (https://www.mzcloud.org/), KEGG, ChemSpider (https://www.chemspider.com/), and BioCyc DB (https://biocyc.org/). To confirm the results, the detected spectrum in the sample was compared with the reference spectrum registered in the online libraries.

The SIMCA 14.1 (Umetrics, Umea, Sweden) was employed for orthogonal projection least squares-discriminant analysis (OPLS-DA). The cut-off range for candidate selection was as follows: variable importance in the projection (VIP) value >1, an absolute p-value >0.05 (Chhonker et al., 2021; Shi et al., 2021), an absolute p(correlation) value ≥0.4, and an endogenous metabolite. To assess the risk of model overfitting, the 200-times permutation test was performed (Zhu et al., 2020; Chen et al., 2023). For candidate metabolites selected through multivariate analysis, validation was conducted by confirming their consistency through compound identification.

Metabolite identification (ID) using standard materials

ID was performed after dissolving each of the analytical standard products in an optimal solvent to obtain a concentration of 100 ppm. Standard materials and samples were analyzed using full MS-data dependent MS2 scan mode and parallel reaction monitoring (PRM) mode (Bourmaud et al., 2016). The fragments of the target compound obtained through analysis from the standard material and plasma sample were compared to finally select the matching metabolites.

Quantification of selected endogenous metabolites

Quantitative analysis was performed on the final selected metabolites from ID. Endogenous metabolites were quantified in plasma using substances with isotopes as internal standard. As the internal standard, donepezil-d4 was used in electrospray ionization positive (ESI+) mode, and thioctic acid-d5 was used in electrospray ionization negative (ESI−) mode. The final 10 μg/mL of internal standard concentration was added to pretreated plasma samples. These samples were then subjected to UHPLC-HRMS analysis under the same conditions as the metabolomic analysis previously described. During the analysis, the concentration of each metabolite was determined by calculating the ratio of the peak area of the metabolite to the peak area of the corresponding internal standard.

Statistical and network analyses

The statistical and correlation analyses were performed using GraphPad Prism 6 software (Dotmatics, Boston, MA, USA). The statistical significance of metabolite concentration changes before and after cisplatin and radiation therapy in patients with cervical cancer was analyzed using either the paired t-test or Wilcoxon rank test. For comparison of the metabolite concentration at P1 or P2 between HPV-positive and HPV-negative groups, either the t-test or Mann–Whitney test was used. The correlation between creatinine and metabolite used was determined using the Spearman’s correlation analysis. The receiver operating characteristic curve (ROC) and area under the curve (AUC) for selected metabolites was evaluated using SPSS 14.1 (IBM, NY, USA). The MetaboAnalyst 5.0 platform (https://www.metaboanalyst.ca/) was used for network analysis of the metabolite.

RESULTS

Clinical characteristics of patients

A total of 13 patients with cervical cancer participated in this clinical trial. Among them, 12 underwent treatment with both cisplatin and radiation therapy, whereas one patient received radiotherapy alone. The average age of these patients was 55 years with an average weight of 61 kg. Following cisplatin and radiation therapy, the concentrations of aspartate aminotransferase (AST), alanine transaminase (ALT), and creatinine in plasma remained unchanged. However, the levels of hemoglobin (Hb) and hematocrit (HCT) showed a significant decrease following P2 compared with P1. This decrease in Hb and HCT can be attributed to cisplatin’s effect of reducing erythropoietin, which is a hormone crucial for red blood cell production (Wood and Hrushesky, 1995; Unami et al., 1996). Hence, the reduction in Hb and HCT is a natural outcome following cisplatin therapy. Regarding the patient characteristics, 38.5% of the patients were infected with HPV, and 76.9% of the patients were in menopause. Nine patients were in stage II, two were in stage III, and no information was available for two patients (Table 1).

Untargeted metabolomics analysis between before treatment (P1) and following cisplatin and radiation therapy (P2) in plasma samples of patients

Following the LC-MS/MS analysis, QC-based normalization of the peak areas was performed for a total of 487 compounds that matched using Compound Discoverer. The normalized peak area data was then used as the dependent variable in the multivariate analysis. The OPLS-DA score plots demonstrated clear discrimination between the P1 and P2 groups (Fig. 2A, 2B), indicating distinct metabolic profiles before and after cisplatin and radiation therapy. Based on the defined criteria (|p|≥0.05, |p|(corr)≥0.4, VIP≥1 from the S-plot), compounds that met the criteria were identified as candidates for further analysis (Fig. 2C-2F). The parameters obtained from the OPLS-DA models were R2Y=0.949 and Q2=0.75 in ESI − mode and R2Y=0.956 and Q2=0.475 in ESI+mode, suggesting a robust and well-fitted model. Moreover, the response permutation test results (R2Y=0.797 and Q2=−0.591 in negative mode, R2Y=0.922 and Q2=−0.248 in positive mode) indicated successful modeling and the absence of overfitting (Fig. 2G, 2H). After comparing the endogenous metabolites’ spectra with the reference library, nine metabolites were selected as potential biomarkers: indoxyl sulfate, phenylacetylglutamine, glycodeoxycholic acid, LysoPE (18:1), deoxycholic acid, 2-oxindole, pyroglutamic acid, trans-3-indoleacrylic acid, and indole-3-acetic acid (Supplementary Table 3). A total of nine metabolites were identified using standard materials. Indoxyl sulfate, glycodeoxycholic acid, deoxycholic acid, LysoPE (18:1), phenylacetylglutamine, and indole-3-acetic acid were obtained as final candidates compared with the standard material spectrum of each metabolite. Phenylacetylglutamine was obtained in both negative and positive modes, and the data analyzed in the positive mode was used, where the intensity was high.

Figure 2. Results of the multivariate analysis. (A, B) Score plot of OPLS-DA between before treatment (P1) and after cisplatin and radiotherapy (P2) in patients with cervical cancer. (C, D) S-plot of the OPLS-DA model. Red dots lines and red box area indicate the selection criteria of metabolite candidates (lpl≥0.05 and lp(correlation)l≥0.4). (E, F) VIP bar graph of the OPLS-DA model. (G, H) Validation plot obtained from 200 permutation tests. (A, C, E), and (G) were analyzed in ESI positive mode. (B, D, F), and (H) were analyzed in the ESI negative mode.

Changes in the concentration of candidate metabolites following cisplatin and radiation therapy

The concentrations of indoxyl sulfate significantly increased by approximately 4.9 times; LysoPE (18:1) by 1.5 times; phenylacetylglutamine by 5.5 times; and indole-3-acetic acid by 1.9 times following P2 compared with P1. However, no significant change was observed in glycodeoxycholic acid concentration following P2. As for deoxycholic acid, the mean concentration in patients increased following P2, but it was not statistically significant (Supplementary Table 4, Fig. 3). Changes in the concentrations of metabolites were also analyzed, except those in a patient who received radiation alone. The concentrations of indoxyl sulfate significantly increased by approximately 4.8 times, LysoPE (18:1) by 1.4 times, phenylacetylglutamine by 5.9 times, and indole-3-acetic acid by 2.0 times at P2 compared with those at P1. However, no significant changes were observed in the concentrations of glycodeoxycholic acid and deoxycholic acid. Radiation therapy alone did not significantly affect the outcome.

Figure 3. Changes in the concentration of metabolites after cisplatin and radiation therapy. (A) Indoxyl sulfate, (B) Glycodeoxycholic acid, (C) Deoxycholic a cid, (D) LysoPE (18:1), (E) Phenylacetylglutamine, (F) Indole-3-acetic acid. Horizontal lines of top and bottom indicate the standard deviation, and the middle line refers to the mean value.

Association between HPV infection and metabolite concentration

The concentration of metabolites between P1 and P2 were compared between the HPV infection and HPV noninfection groups. Indoxyl sulfate, LysoPE (18:1) phenylacetylglutamine were significantly increased following P2 compared with P1 in HPV noninfection group (Fig. 4A-4D). However, in the HPV infection group, there were no significant changes in metabolite concentrations between P1 and P2 (Fig. 4A-4D). Furthermore, within the P1 and P2 time points, there were no significant differences in metabolite concentrations based on whether the patients had HPV infection or not.

Figure 4. The association between clinical outcomes and metabolite concentration. (A-D) The association between HPV infection and metabolite concentration. (E-H) Changes in metabolite concentration according to the therapeutic effect of cisplatin and radiotherapy. The y-axis represents the concentration ratio (P2/P1) of each candidate. Horizontal lines of top and bottom indicate the standard deviation, and the middle line refers to the mean value. (I-L) Spearman’s correlation analysis results. The y-axis represents the concentration of metabolite. The x-axis represents creatinine concentration in plasma samples. The bold lines are trended lines, and the dotted lines are 95% confidence intervals. HPV (+): HPV positive, HPV (−): HPV negative, n.s: Not statistically significant, *p<0.05 statistically significant, P1: before treatment, P2: after cisplatin and radiotherapy. CR: clear remission, PR: partial remission, SD: stable disease, PD: progressive disease. r: Spearman’s correlation coefficient.

Changes in metabolite concentration according to effect of cisplatin and radiation therapy

To compare the concentration ratios (P2/P1) of indoxyl sulfate, LysoPE (18:1), indole-3-acetic acid, and phenylacetylglutamine between the CR/PR and PD/SD groups, no statistically significant differences were observed (Fig. 4E-4H). However, the concentration ratio (P2/P1) of indoxyl sulfate, LysoPE (18:1), and phenylacetylglutamine tended to increase more CR/PR group than the PD/SD group. Conversely, the concentration ratio (P2/P1) of indole-3-acetic acid tended to increase more in the PD/SD group than in the CR/PR group. A total of seven patients were evaluated for the treatment effects, and six patients were not evaluated. Therefore, the difference in metabolic concentration according to the treatment effect requires validation using more patient samples.

Correlation between concentration of creatinine and metabolites in plasma samples

The correlation between the metabolites and creatinine concentration in plasma samples was determined using the Spearman’s correlation analysis. Indoxyl sulfate, LysoPE (18:1), phenylacetylglutamine, and indole-3-acetic acid had a correlation coefficient (r) of 0.073 (p=0.085), 0.158 (p=0.663), 0.298 (p=0.400), and 0.359 (p=0.307), respectively, with creatinine. Four metabolites had a positive correlation with creatinine but were not statistically significant (Fig. 4I-4L).

ROC analysis and network analysis of candidates

The ROC curve analysis was utilized to evaluate the discrimination ability of the four metabolites (indoxyl sulfate, LysoPE (18:1), phenylacetylglutamine, and indole-3-acetic acid) in distinguishing before and after cisplatin and radiation therapy. The analysis revealed that all four metabolites, which had shown statistically significant differences had AUC >0.7. Among them, indoxyl sulfate had the highest AUC of 0.909 (Fig. 5A). Moreover, the ROC analysis was employed to evaluate the discriminative ability of the four metabolites to evaluate the therapeutic effect. LysoPE (18:1) exhibited the highest AUC at 0.7 (Fig. 5B). Network analysis using the four metabolites revealed networks with metabolites associated with tryptophan metabolism, nitrogen metabolism, purine metabolism, aminoacyl-tRNA biosynthesis and glyoxylate and dicarboxylate metabolism pathway (Fig. 5C).

Figure 5. Results of ROC and network analysis. (A) ROC analysis of endogenous metabolites with significantly different concentrations before and after cisplatin and radiation therapy in patients with cervical cancer. (B) ROC analysis of metabolites according to cisplatin and radiotherapy therapeutic effects. (C) Network analysis result.
DISCUSSION

To evaluate changes in endogenous metabolites caused by cisplatin and radiotherapy, the untargeted metabolomic analysis of plasma samples from 13 patients with cervical cancer revealed significant changes in the concentrations of four metabolites (indoxyl sulfate, phenylacetylglutamine, LysoPE (18:1), and indole-3-acetic acid) after cisplatin and radiation therapy (P2) compared to before therapy (P1). Notably, in the HPV-negative group, indoxyl sulfate, phenylacetylglutamine, and LysoPE (18:1) exhibited a significant response to cisplatin and radiation therapy in contrast to that in the HPV-positive group.

The HPV-negative patients exhibited higher responsiveness to cisplatin and radiation therapy compared with HPV-positive patients. In this study, 38.5% of the subjects were HPV positive, and 61.5% were HPV negative. Among the four metabolites, indoxyl sulfate, LysoPE (18:1), and phenylacetylglutamine showed statistically significant increases following P2 compared with P1 in the HPV-negative group. However, no significant changes were observed in the concentrations of these metabolites in the HPV-positive group at both time points (P1 and P2). Previous clinical trials have reported that the HPV-negative group achieved 100% CR after receiving platinum-based chemotherapy and radiotherapy, whereas the HPV-positive group had a lower CR rate (40%) and a higher PR rate (60%) (Badaracco et al., 2010). In the presence of HPV infection, major oncogenes E6 and E7 are overexpressed (Yuan et al., 2012). E6 and E7 overexpression inhibits the tumor necrosis factor (TNF)-based apoptosis pathway by reducing the expression of Bak (Bcl-2 homologous antagonist/killer), an apoptosis-associated protein (Yuan et al., 2012). Reduced expression levels of E6 and E7 result in sensitivity to cisplatin (Putral et al., 2005). These results suggest that HPV infection influences the responsiveness to cisplatin therapy.

Following cisplatin and radiation therapy, a significant increase in indoxyl sulfate and indole-3-acetic acid, derived from tryptophan by gut microbiota, has been reported to induce apoptosis and enhance the effectiveness of chemotherapy (Paeslack et al., 2022). Several studies have demonstrated that indoxyl sulfate enhances the cytotoxic effect of cisplatin, thereby increasing cell death (Tan et al., 2021; Yabuuchi et al., 2021). Additionally, the combination treatment of indole-3-acetic acid and platinum-based chemotherapy significantly reduced tumor weight in an in vivo study (Kwon, 2023). A tryptophan metabolite derived from the gut microbiome can affect cancer’s responsiveness to chemotherapy. Gut-dwelling bacteria bring out indole-3-acetic acid (3-IAA) from food-derived tryptophan (Trp). 3-IAA immigrates to cancer through the circulation and may be oxidized to toxic molecules (3-IAAp) by myeloperoxidase (MPO) and cytotoxic anticancer drugs in intratumoral neutrophils. In turn, 3-IAA and platinum-based chemotherapies induce the downregulation of glutathione peroxidase 3/7 (GPX3/7), reactive oxygen species (ROS)-degrading enzymes, and subsequent ROS accumulation in cancer cells. Increased levels of ROS hinder the autophagy pathway, a crucial process in the proliferation of cancer cell (Kwon, 2023). In the current study, indoxyl sulfate concentrations were higher in the CR/PR group than in the PD/SD group, and although this difference was not statistically significant, it aligned with previous studies. Conversely, indole-3-acetic acid tended to have higher concentrations in the PD/SD group, which contradicted earlier literature results.

Indoxyl sulfate and phenylacetylglutamine are metabolites that have been extensively studied in relation to the nephrotoxic side effects of cisplatin therapy. Among 13 patients, 12 were treated with cisplatin, known for its nephrotoxicity as a common side effect (Pabla and Dong, 2008; Ghosh, 2019). Indoxyl sulfate is known to have strong binding to proteins, making it challenging to completely eliminate from the kidneys (Morimoto et al., 2018; Cheng et al., 2020). Phenylacetylglutamine along with indoxyl sulfate is another metabolite found in patients diagnosed with chronic kidney disease (Barrios et al., 2016; O’Brien et al., 2020). Phenylacetylglutamine is metabolized from phenylbutyrate and excreted into urine; therefore, patients with urea cycle disorder experience increased blood concentrations of phenylacetylglutamine (Mokhtarani et al., 2012). However, in this study, no correlation was observed between creatinine concentration and the levels of indoxyl sulfate and phenylacetylglutamine.

The physiological functions and in vivo roles of lysophosphatidylamines (LysoPEs) were not well studied (Kihara et al., 2015; Yamamoto et al., 2021). LysoPE (18:1) is a lysophospholipid generated through the deacylation of phosphatidylethanolamine hydrolysis catalyzed by phospholipase A1/A2 (Yamamoto et al., 2021). Subsequently, lysophospholipase converts lysophospholipid into lysophosphatidic acid (LPA) (Lee et al., 2019). Higher concentrations of LPA were observed in the plasma of patients with cervical cancer compared with those in healthy individuals (Xu et al., 1998; Sui et al., 2015). In an in vitro study, LPA was observed to decrease when treated with cisplatin, and it was reported to play a protective role against cisplatin-induced apoptosis (Sui et al., 2015). When cancer patients were treated with anticancer drugs, there is a research result indicating that the serum concentration of LysoPE (18:1) increased significantly after treatment compared to before treatment (Wei et al., 2022). Since LysoPE (18:1) is a substance involved in the LPA production pathway, its exact mechanism remains unknown, but it is predicted to be related to cisplatin response.

Indole-3-acetic acid (IAA) is a tryptophan-derived metabolite produced by gut microbiota. It is known to mitigate inflammatory reactions and oxidative stress, although the precise mechanisms it is involved in are not fully understood (Ji et al., 2020). IAA significantly ameliorated of interleukin-1β (IL-1β), IL-6, and monocyte chemoattractant protein-1 (MCP-1) levels. Additionally, IAA was found to generate reactive oxidative species (ROS) and nitric oxide (NO) (Ji et al., 2020). In this study, the plasma concentration of IAA increased after cisplatin and radiation therapy. These results suggest an activation of immune response.

For limitation of the study, sample size was limited in 13 patients. There was an exploratory studies with similar sample size indicating its adequacy for metabolomic comparisons before and after cisplatin and radiotherapy. For instance, one study investigated the effects of chemotherapy and radiotherapy in 18 patients with cervical cancer according to the presence or absence of HPV infection (Badaracco et al., 2010), whereas another study identified chemotherapy-related biomarkers in 13 patients (Liu et al., 2011). However, a study on the identification of diagnostic biomarkers in esophageal SCC was provided the biomarkers using samples from 60 patients (Zhu et al., 2020). Moreover, the evaluation of the association of the candidate metabolites with clinical response was limited because only eight patients had available clinical response information. Consequently, to establish a more robust association between treatment outcomes and the four metabolites (indoxyl sulfate, LysoPE (18:1), phenylacetylglutamine, and indole-3-acetic acid), further verification across a larger patients would be necessary.

In conclusion, through plasma metabolomics profiles, this study demonstrates the metabolite changes due to anti-cancer treatment using cisplatin and radiation therapy in cervical cancer patients. This study observed an increase in indoxyl sulfate, phenylacetylglutamine, indole-3-acetic acid, and LysoPE (18:1) levels after cisplatin and radiation therapy. These identified metabolites have the potential to serve as new biomarkers for evaluating cisplatin and radiation therapy response of cervical cancer patients. As a result, the findings of this study could significantly contribute to the precise treatment of cervical cancer patients in the future by presenting potential biomarker candidates for assessing treatment responses.

ACKNOWLEDGMENTS

This paper was supported by Fund of Biomedical Research Institute, Jeonbuk National University Hospital. This study was supported by the National Research Foundation of Korea (NRF) grant funded by the Korea government (MSIT) (No. RS-2023-00251397) and the 4TH BK21 project (Educational Research Group for Platform development of management of emerging infectious disease) funded by the Korean ministry of education (5199990614732).

CONFLICT OF INTEREST

The authors declare that they have no conflict of interest.

AUTHOR CONTRIBUTIONS

KHS, MGK, and SYL conceived, designed, and directed the study. KHS designed the metabolomic analysis study, and MGK and SYL designed the clinical trial. JYJ, MGK, and SYL collected the patient plasma samples and arranged the clinical data. SIK and SYC performed the untargeted metabolomic analysis in plasma samples using the UHPLC-HRMS. Also, SIK and SYC made the results and drafted the manuscript. KHS, MGK, SYL, and JYJ reviewed and edited the manuscript. All authors discussed the results and commented on the manuscript.

References
  1. Badaracco, G., Savarese, A., Micheli, A., Rizzo, C., Paolini, F., Carosi, M., Cutillo, G., Vizza, E., Arcangeli, G. and Venuti, A. (2010) Persistence of HPV after radio-chemotherapy in locally advanced cervical cancer. Oncol. Rep. 23, 1093-1099.
    Pubmed CrossRef
  2. Barrios, C., Spector, T. D. and Menni, C. (2016) Blood, urine and faecal metabolite profiles in the study of adult renal disease. Arch. Biochem. Biophys. 589, 81-92.
    Pubmed CrossRef
  3. Bourmaud, A., Gallien, S. and Domon, B. (2016) Parallel reaction monitoring using quadrupole-Orbitrap mass spectrometer: principle and applications. Proteomics 16, 2146-2159.
    Pubmed CrossRef
  4. Brown, A., Kumar, S. and Tchounwou, P. B. (2019) Cisplatin-based chemotherapy of human cancers. J. Cancer Sci. Ther. 11, 97.
  5. Chen, H., Zhang, J., Zhou, H., Zhu, Y., Liang, Y., Zhu, P. and Zhang, Q. (2022) UHPLC-HRMS-based serum lipisdomics reveals novel biomarkers to assist in the discrimination between colorectal adenoma and cancer. Front. Oncol. 12, 934145.
    Pubmed KoreaMed CrossRef
  6. Chen, L., Ding, H., Zhu, Y., Guo, Y., Tang, Y., Xie, K., Zhang, G., Dai, G., Gao, Y. and Zhang, T. (2023) Untargeted and targeted metabolomics identify metabolite biomarkers for Salmonella enteritidis in chicken meat. Food Chem. 409, 135294.
    Pubmed CrossRef
  7. Chen, W.-L., Wang, J.-H., Zhao, A.-H., Xu, X., Wang, Y.-H., Chen, T.-L., Li, J.-M., Mi, J.-Q., Zhu, Y.-M. and Liu, Y.-F. (2014) A distinct glucose metabolism signature of acute myeloid leukemia with prognostic value. Blood 124, 1645-1654.
    Pubmed KoreaMed CrossRef
  8. Cheng, T.-H., Ma, M.-C., Liao, M.-T., Zheng, C.-M., Lu, K.-C., Liao, C.-H., Hou, Y.-C., Liu, W.-C. and Lu, C.-L. (2020) Indoxyl sulfate, a tubular toxin, contributes to the development of chronic kidney disease. Toxins 12, 684.
    Pubmed KoreaMed CrossRef
  9. Chhonker, Y. S., Kanvinde, S., Ahmad, R., Singh, A. B., Oupický, D. and Murry, D. J. (2021) Simultaneous quantitation of lipid biomarkers for inflammatory bowel disease using LC-MS/MS. Metabolites 11, 106.
    Pubmed KoreaMed CrossRef
  10. Dasari, S., Wudayagiri, R. and Valluru, L. (2015) Cervical cancer: biomarkers for diagnosis and treatment. Clin. Chim. Acta 445, 7-11.
    Pubmed CrossRef
  11. Ezaki, T., Nishiumi, S., Azuma, T. and Yoshida, M. (2017) Metabolomics for the early detection of cisplatin-induced nephrotoxicity. Toxicol. Res. 6, 843-853.
    Pubmed KoreaMed CrossRef
  12. Gadducci, A., Tana, R., Cosio, S. and Genazzani, A. R. (2008) The serum assay of tumour markers in the prognostic evaluation, treatment monitoring and follow-up of patients with cervical cancer: a review of the literature. Crit. Rev. Oncol. Hematol. 66, 10-20.
    Pubmed CrossRef
  13. Ghosh, S. (2019) Cisplatin: the first metal based anticancer drug. Bioorg. Chem. 88, 102925.
    Pubmed CrossRef
  14. Hascitha, J., Priya, R., Jayavelu, S., Dhandapani, H., Selvaluxmy, G., Singh, S. S. and Rajkumar, T. (2016) Analysis of Kynurenine/Tryptophan ratio and expression of IDO1 and 2 mRNA in tumour tissue of cervical cancer patients. Clin. Biochem. 49, 919-924.
    Pubmed CrossRef
  15. Holditch, S. J., Brown, C. N., Lombardi, A. M., Nguyen, K. N. and Edelstein, C. L. (2019) Recent advances in models, mechanisms, biomarkers, and interventions in cisplatin-induced acute kidney injury. Int. J. Mol. Sci. 20, 3011.
    Pubmed KoreaMed CrossRef
  16. Ji, Y., Yin, W., Liang, Y., Sun, L., Yin, Y. and Zhang, W. (2020) Anti-inflammatory and anti-oxidative activity of indole-3-acetic acid involves induction of HO-1 and neutralization of free radicals in RAW264.7 cells. Int. J. Mol. Sci. 21, 1579.
    Pubmed KoreaMed CrossRef
  17. Khan, I., Nam, M., Kwon, M., Seo, S.-S., Jung, S., Han, J. S., Hwang, G.-S. and Kim, M. K. (2019) LC/MS-based polar metabolite profiling identified unique biomarker signatures for cervical cancer and cervical intraepithelial neoplasia using global and targeted metabolomics. Cancers 11, 511.
    Pubmed KoreaMed CrossRef
  18. Kihara, Y., Mizuno, H. and Chun, J. (2015) Lysophospholipid receptors in drug discovery. Exp. Cell Res. 333, 171-177.
    Pubmed KoreaMed CrossRef
  19. Kwon, B. (2023) A metabolite of the gut microbiota: a facilitator of chemotherapy efficacy in cancer. Signal Transduct. Target. Ther. 8, 238.
    Pubmed KoreaMed CrossRef
  20. Lee, J. H., Kim, D., Oh, Y. S. and Jun, H.-S. (2019) Lysophosphatidic acid signaling in diabetic nephropathy. Int. J. Mol. Sci. 20, 2850.
    Pubmed KoreaMed CrossRef
  21. Liu, H., Han, Y., Mi, R., Zhang, Y., Su, G., Wang, H., Zhou, X., Liu, X. and Zhu, B. (2011) Identification of cervical cancer proteins associated with treatment with paclitaxel and cisplatin in patients. Int. J. Gynecol. Cancer 21, 1452-1457.
    Pubmed CrossRef
  22. Mokhtarani, M., Diaz, G., Rhead, W., Lichter-Konecki, U., Bartley, J., Feigenbaum, A., Longo, N., Berquist, W., Berry, S. and Gallagher, R. (2012) Urinary phenylacetylglutamine as dosing biomarker for patients with urea cycle disorders. Mol. Genet. Metab. 107, 308-314.
    Pubmed KoreaMed CrossRef
  23. Morimoto, K., Tominaga, Y., Agatsuma, Y., Miyamoto, M., Kashiwagura, S., Takahashi, A., Sano, Y., Yano, K., Kakinuma, C. and Ogihara, T. (2018) Intestinal secretion of indoxyl sulfate as a possible compensatory excretion pathway in chronic kidney disease. Biopharm. Drug Dispos. 39, 328-334.
    Pubmed CrossRef
  24. O'Brien, F. J., Mair, R. D., Plummer, N. S., Meyer, T. W., Sutherland, S. M. and Sirich, T. L. (2020) Impaired tubular secretion of organic solutes in acute kidney injury. Kidney360 1, 724-730.
    Pubmed KoreaMed CrossRef
  25. Pabla, N. and Dong, Z. (2008) Cisplatin nephrotoxicity: mechanisms and renoprotective strategies. Kidney Int. 73, 994-1007.
    Pubmed CrossRef
  26. Paeslack, N., Mimmler, M., Becker, S., Gao, Z., Khuu, M. P., Mann, A., Malinarich, F., Regen, T. and Reinhardt, C. (2022) Microbiota-derived tryptophan metabolites in vascular inflammation and cardiovascular disease. Amino Acids 54, 1339-1356.
    Pubmed KoreaMed CrossRef
  27. Putral, L. N., Bywater, M. J., Gu, W., Saunders, N. A., Gabrielli, B. G., Leggatt, G. R. and McMillan, N. A. (2005) RNA interference against human papillomavirus oncogenes in cervical cancer cells results in increased sensitivity to cisplatin. Mol. pharmacol. 68, 1311-1319.
    Pubmed CrossRef
  28. Schwickert, G., Walenta, S., Sundfør, K., Rofstad, E. K. and Mueller-Klieser, W. (1995) Correlation of high lactate levels in human cervical cancer with incidence of metastasis. Cancer Res. 55, 4757-4759.
    Pubmed
  29. Shi, W., Yuan, X., Cui, K., Li, H., Fu, P., Rehman, S.-U., Shi, D., Liu, Q. and Li, Z. (2021) LC-MS/MS based metabolomics reveal candidate biomarkers and metabolic changes in different buffalo species. Animals 11, 560.
    Pubmed KoreaMed CrossRef
  30. Sui, Y., Yang, Y., Wang, J., Li, Y., Ma, H., Cai, H., Liu, X., Zhang, Y., Wang, S. and Li, Z. (2015) Lysophosphatidic acid inhibits apoptosis induced by cisplatin in cervical cancer cells. Biomed Res. Int. 2015, 598386.
    Pubmed KoreaMed CrossRef
  31. Tan, B., Chen, J., Qin, S., Liao, C., Zhang, Y., Wang, D., Li, S., Zhang, Z., Zhang, P. and Xu, F. (2021) Tryptophan pathway-targeted metabolomics study on the mechanism and intervention of cisplatin-induced acute kidney injury in rats. Chem. Res. Toxicol. 34, 1759-1768.
    Pubmed CrossRef
  32. Unami, A., Nishina, N., Terai, T., Sato, S., Tamura, T., Noda, K. and Mine, Y. (1996) Effects of cisplatin on erythropoietin production in rats. J. Toxicol. Sci. 21, 157-165.
    Pubmed CrossRef
  33. Velenosi, T. J., Krausz, K. W., Hamada, K., Dorsey, T. H., Ambs, S., Takahashi, S. and Gonzalez, F. J. (2022) Pharmacometabolomics reveals urinary diacetylspermine as a biomarker of doxorubicin effectiveness in triple negative breast cancer. NPJ Precision Oncol. 6, 70.
    Pubmed KoreaMed CrossRef
  34. Wei, Y., Wang, J., Chen, F., Li, X., Zhang, J., Shen, M., Tang, R. and Huang, Z. (2022) Serum abnormal metabolites for evaluating therapeutic response and prognosis of patients with multiple myeloma. Front. Oncol. 12, 808290.
    Pubmed KoreaMed CrossRef
  35. Wood, P. A. and Hrushesky, W. (1995) Cisplatin-associated anemia: an erythropoietin deficiency syndrome. J. Clin. Invest. 95, 1650-1659.
    Pubmed KoreaMed CrossRef
  36. Xu, Y., Shen, Z., Wiper, D. W., Wu, M., Morton, R. E., Elson, P., Kennedy, A. W., Belinson, J., Markman, M. and Casey, G. (1998) Lysophosphatidic acid as a potential biomarker for ovarian and other gynecologic cancers. JAMA 280, 719-723.
    Pubmed CrossRef
  37. Yabuuchi, N., Hou, H., Gunda, N., Narita, Y., Jono, H. and Saito, H. (2021) Suppressed hepatic production of indoxyl sulfate attenuates cisplatin-induced acute kidney injury in sulfotransferase 1a1-deficient mice. Int. J. Mol. Sci. 22, 1764.
    Pubmed KoreaMed CrossRef
  38. Yamamoto, Y., Sakurai, T., Chen, Z., Furukawa, T., Gowda, S. G. B., Wu, Y., Nouso, K., Fujii, Y., Yoshikawa, Y., Chiba, H. and Hui, S. P. (2021) Analysis of serum lysophosphatidylethanolamine levels in patients with non-alcoholic fatty liver disease by liquid chromatography-tandem mass spectrometry. Anal. Bioanal. Chem. 413, 245-254.
    Pubmed CrossRef
  39. Yuan, C.-H., Filippova, M. and Duerksen-Hughes, P. (2012) Modulation of apoptotic pathways by human papillomaviruses (HPV): mechanisms and implications for therapy. Viruses 4, 3831-3850.
    Pubmed KoreaMed CrossRef
  40. Zhu, H., Luo, H., Zhang, W., Shen, Z., Hu, X. and Zhu, X. (2016) Molecular mechanisms of cisplatin resistance in cervical cancer. Drug Des. Devel. Ther. 10, 1885-1895.
    Pubmed KoreaMed CrossRef
  41. Zhu, Z.-J., Qi, Z., Zhang, J., Xue, W.-H., Li, L.-F., Shen, Z.-B., Li, Z.-Y., Yuan, Y.-L., Wang, W.-B. and Zhao, J. (2020) Untargeted metabolomics analysis of esophageal squamous cell carcinoma discovers dysregulated metabolic pathways and potential diagnostic biomarkers. J. Cancer 11, 3944-3954.
    Pubmed KoreaMed CrossRef


This Article


Cited By Articles
  • CrossRef (0)

Funding Information

Services
Social Network Service

e-submission

Archives