The diagnostic value of serum insulin-like growth factor binding protein 7 in gastric cancer

TCGA database exploration

In order to find the expression pattern of IGFBP-rPs in gastric cancer, we downloaded the related mRNAs expression in gastric cancers from TCGA. Their expressions were presented with heatmaps using the R package “pheatmap”. The differentially expressed IGFBP-rPs between gastric cancer tissues and adjacent normal tissues was defined using Mann-Whitney U test. Gene Expression Profiling Interactive Analysis (GEPIA) is a web service which integrates the TCGA and GTEx data to provide the quick analysis of cancers (Tang et al., 2017). We used GEPIA to apply the Kaplan–Meier survival analysis to assess the prognostic value of IGFBP-rPs. The median value of each IGFBP-rP was used to distinguish high level and low level. Log-rank test was applied to assess the significant differences.

Study subjects

To further explore the expression of serum IGFBP7 in gastric cancers, we collected serum samples of gastric cancer patients and health volunteers from the Cancer Hospital of Shantou University Medical College and the Sun Yat-Sen University Cancer Center. Gastric cancer patients who were enrolled from May 2019 to January 2020 in Sun Yat-Sen University Cancer Center were involved into the training cohort, while those enrolled from November 2018 to April 2019 were in the independent validation cohort. Gastric cancer patients were confirmed based on endoscopy and pathological biopsy. Patients with any cancer history or cancer-related treatment history before diagnosis were excluded for analysis. Their serum samples used in this experiment were collected before surgery or any other anti-cancer therapy. Health volunteers were those with a negative upper endoscopy. After venous sampling, blood was centrifugated at 1,250 g for 10 min, and sera were separated and stored at −80 °C until examination analysis. As for normal control groups, contemporaneous physical examination population from two hospitals were respectively divided into training and independent validation cohorts. Informed consents were signed by both patients and healthy volunteers. This study was approved by the institutional review board of the Cancer Hospital of Shantou University Medical College and the Sun Yat-Sen University Cancer Center (IRB approval number: B2022-329-01), and followed the requirements of the Declaration of Helsinki. Tumor stages were defined according to the eighth edition of the American Joint Committee on Cancer Staging Manual (Amin et al., 2017). In this study, stage 0+I+II was defined as early-stage gastric cancer.

Enzyme-linked immunosorbent assay

ELISA was carried out to examinate the serum levels of IGFBP-7. The experiment protocol was performed based on the manual from Cusabio (Catalog number: CSB-E17249h). The ELISA kit and serum samples were balanced to room temperature before the examination began. The IGFBP7 standard concentrations for standard curves construction were 10, 5, 2.5, 1.25, 0.625, 0.312, 0.156, and 0 ng/ml respectively. The serum samples were diluted by 1:3 according to our previous studies (Huang et al., 2019; Liu et al., 2020). A total of 100 μl serum sample or standard was added to each well and incubated at 37 °C for 2 h. The liquids were removed without washing the wells and 100 μl biotin-antibody were added and incubated at 37 °C for 1 h. After liquids removal and wells washing for three times, 100 μl horseradish peroxidase-avidin was added and incubated at 37 °C for another 1 h. The each well was washed for 5 times, and 90 μl 3,3′,5,5′-tetramethylbenzidine substrate for color development was added protected from light for 20 min followed with 50 μl stop solution for termination. The measurement of optical density (OD) values were executed using Microplate Reader (BioTek® Instruments, Winooski, VT, USA) at 450 nm with 570 nm reference within 5 min.

Statistical analysis

In this study, p < 0.05 (two-tailed) was defined as statistical significance. Most statistical analyses were carried out using Microsoft Excel 2019, SPSS 24.0 and GraphPad Prism 8.0. R package pheatmap was used to plot the expression heatmap. The OD values were changed to real concentrations according to the standard curves using SigmaPlot 10.0. Then unpaired Student’s t test was used to evaluate the difference of serum IGFBP7 levels between gastric cancer patients and normal controls. The receiver operating characteristic (ROC) curve was performed to evaluate the diagnostic efficacy, and the area under the ROC curve (AUC) was calculated with the 95% confidence interval (CI). The cutoff value for distinguishing gastric cancer patients and normal controls was acquired by achieving the maximum sensitivity when the specificity was greater than 90% in the training cohort (Boyle et al., 2011). The Pearson’s correlation analysis was used to evaluate the correlation between serum IGFBP7 and age. The unpaired Student’s t test was applied to evaluate the association between serum IGFBP7 level and other clinical characteristics of gastric cancer patients.

IGFBP-rPs expression in TCGA

Figure S1 exhibited the expression heatmap of IGFBP-rPs. Differential expression analysis showed that IGFBP7, CCN2, ESM1, CCN4 and CCN6 mRNA were over-expressed in gastric cancer tissues while CCN5 mRNA was down-expressed (all p < 0.05, Figs. S2A–S2J). From overall survival analysis, gastric cancer patients with high expression of IGFBP7, CCN2, CCN3, CCN1, CCN5 or CCN4 would have a poor survival time (all p < 0.05, Figs. S3A–S3J). However, only IGFBP7 and CCN5 was associated with disease-free survival (both p < 0.05, Figs. S4A–S4J). Combining both differential expression analysis and survival analysis, we found that only IGFBP7 and CCN5 were both differently-expressed and associated with prognosis. Tumor research mainly focused on those tumor related genes with overexpression, therefore, we will further explore the diagnostic value of IGFBP7 in this following study.

Serum IGFBP7 levels

At this stage, we collected serum samples of gastric cancers patients and normal volunteers from the Cancer Hospital of Shantou University Medical College and Sun Yat-Sen University Cancer Center. As shown in Fig. 1 and Table 1, there were 169 gastric cancer patients and 100 normal volunteers in the training cohort, while there were 55 gastric cancer patients and 55 normal volunteers in the independent validation cohort. There were no significant differences between both cohorts.

Then ELISA was used to test the serum IGFBP7 protein level. ELISA test showed that in the training cohort, the mean concentration ± standard deviation of serum IGFBP7 level in gastric cancer patients was 1.580 ± 0.198 ng/ml, which was lower than in the normal control group (1.850 ± 0.328 ng/ml, p < 0.0001, Fig. 2A). In the independent validation cohort, the serum IGFBP7 level in gastric cancer group was 1.592 ± 0.253 ng/ml, lower than in the normal control group (2.011 ± 0.563 ng/ml, p < 0.0001, Fig. 2B). In early-stage gastric cancer patients, the serum levels of IGFBP7 were also lower when compared with normal control groups in both cohorts (both p < 0.0001). There were no significant differences between early-stage gastric cancer patients and all gastric cancer patients (both p < 0.05).

Diagnostic value of IGFBP7

Using the ROC curve, we acquired an AUC of 0.774 (95% CI [0.713–0.836]) for distinguishing gastric cancer patients from normal volunteers (Fig. 3A) in training cohort. When the specificity was 90.0% (95% CI [82.0–94.8]), the cutoff value was 1.515 ng/ml with the sensitivity of 36.7% (95% CI [29.5–44.5]). In the early-stage gastric cancer (Fig. 3B), with the same cutoff value and specificity, the AUC of diagnostic efficacy was 0.773 (95% CI [0.701–0.845]) with the sensitivity of 36.1% (95% CI [25.4–48.3]). As shown in Fig. 3C, using the same cutoff value of 1.515 ng/ml, in the independent validation cohort, IGFBP7 could identify gastric cancers with an AUC value of 0.758 (95% CI [0.664–0.852]), a sensitivity of 34.5% (95% CI [22.6–48.7]), and a specificity of 85.5% (95% CI [72.8–93.1]). Similarly, for early-stage gastric cancers diagnosis in the independent validation cohort, the AUC value was 0.778 (95% CI [0.673–0.882]) and the sensitivity was 33.3% (95% CI [14.4–58.8]) with the same specificity (Fig. 3D). Table 2 presented the diagnostic efficacy of IGFBP7 based on positive predictive value (PPV), negative predictive value (NPV), positive likelihood ratio (PLR), and negative likelihood ratio (NLR).

Correlation between IGFBP7 and clinical characteristics

From Table 1, we found that the control group was younger than the gastric cancer group, especially in the independent validation cohorts. Then we used Pearson correlation analysis to further evaluate the correlation between serum IGFBP7 and age. As shown in Fig. 4, there was no lineal correlation between serum IGFBP7 and age (both p > 0.05). As shown in the Fig. S5, there were no correlation between IGFBP7 and any clinical characteristics, including depth of invasion, lymph node metastasis status, distant metastasis and Lauren type in both training and independent validation cohorts (all p > 0.05).

As for gender, as shown in the Fig. S6, regardless in male or female, the serum IGFBP7 levels of gastric cancer patients were lower than those of normal volunteers (all p < 0.05). In the training cohort, serum IGFBP7 levels of male gastric cancer patients were lower than those of female patients (p < 0.0001). There were no difference of IGFBP7 levels between male and female normal volunteers.

IGFBP7 in pan-cancers

In order to further explore the expression of IGFBP7 in other cancers, we conducted a pan-cancers analysis using the TCGA and GTEx database. As shown in Fig. S7, IGFBP7 mRNA levels were found to be upregulated in lymphoid neoplasm diffuse large B-cell lymphoma, esophageal carcinoma, glioblastoma multiforme, head and neck squamous cell carcinoma, acute myeloid leukemia, brain lower grade glioma, pancreatic adenocarcinoma, stomach adenocarcinoma, and thymoma. Downregulations were observed mainly in tumors of the urogenital system, lung cancer, and thyroid carcinoma.