Abstract
Aim
The purpose of this study is to explore the prognostic evaluation value of Ki67 and pSmad3L (S204) for gastric cancer (GC) patients by jointly analyzing their expression levels in GC tissues.
Methods
This single-center observational study enrolled 98 patients with pathologically confirmed GC, collecting 82 paired samples of tumor and adjacent normal tissues. Immunohistochemistry was utilized to detect the expression levels of CD133, E-cad, Ki67, and pSmad3L (S204). Pearson correlation coefficients were calculated to assess inter-protein relationships, while chi-square tests evaluated associations with clinicopathological parameters. Kaplan-Meier analysis generated survival curves, and univariate and multivariate COX regression analyses were conducted to establish a prognostic prediction model.
Results
Compared with adjacent normal tissues, the expressions of CD133, E-cad, and Ki67 were upregulated in GC tissues. The high expression of CD133 was associated with the tumor type, and the high expression of Ki67 was related to tumor grading and distant metastasis. Multivariate analysis showed that the depth of invasion, lymph node metastasis, Ki67, and pSmad3L (S204) were independent prognostic risk factors for GC patients. The co-expression of high levels of Ki67 and pSmad3L (S204) predicted a poor prognosis.
Conclusion
High expression of Ki67 and high expression of pSmad3L(S204) are independent risk factors for the prognosis of patients with GC, and the combined analysis of Ki67 and pSmad3L(S204) can help to improve the risk prediction.
Keywords: Gastric cancer, CD133, E-cad, Ki67, PSmad3L(S204), Prognosis
Introduction
Cancer remains a major public health problem worldwide, with nearly 20 million new cancer cases and nearly 9.7 million cancer deaths expected in 2022, according to the latest report from the International Agency for Research on Cancer (IARC). Among these, GC ranks fifth among common malignancies in terms of both incidence (4.9%) and mortality (6.8%) [1]. Although the mortality rate of GC has shown a decreasing trend in recent years, it is still higher than that in developed countries such as the United Kingdom and the United States, which is mainly related to its low rate of early diagnosis and the lack of uniformity in clinical cancer diagnosis and treatment strategies [2]. GC is now mainly treated with systemic chemotherapy, radiotherapy, surgery, targeted and immunotherapy and other integrated treatments [3]. Early symptoms of GC are insidious and progress rapidly, and the best time for surgical treatment is usually missed when it is first diagnosed [4]. Although systemic chemotherapy can improve the quality and prolong the survival time of GC patients, the clinical benefit is still very limited [5]. Patients with advanced GC have a very poor prognosis due to the difficulty of screening for targeted drugs and multidrug resistance [6]. Previous studies have shown that radiomics technology based on computed tomography (CT) can achieve accurate identification and staging assessment of GC by extracting tumor imaging features [7]. However, compared with imaging features, molecular markers have higher specificity and sensitivity in the early diagnosis and prognosis prediction of GC [8]. Therefore, the search for molecular markers that can diagnose GC at an early stage and predict the prognosis of GC patients is of great importance in realizing new therapeutic targets and improving prognosis.
The gene encoding Ki67, which is located on human chromosome 10 (10q25-qter) and consists of 15 exons, is a widely used marker of cell proliferation [9, 10]. Several clinical studies have confirmed that in pancreatic neuroendocrine tumors, dynamic elevation of the Ki67 proliferation index is significantly associated with increased tumor aggressiveness and poor patient prognosis, suggesting that the Ki67 index can be used as an important biological indicator to assess the progression and prognosis of pancreatic neuroendocrine tumors [11]. In uterine smooth muscle sarcoma (uLMS), quantitatively elevated Ki67 proliferation index was significantly and positively correlated with an increased risk of death in patients, in which a Ki67 index ≥ 10% was identified as an independent poor prognostic factor in uLMS, providing an important quantitative indicator for clinical prognostic assessment [12].Although the role of Ki67 has been extensively studied in a variety of tumors, its prognostic value in GC has not been fully elucidated.
Samd3 is an important transcriptional regulator in the TGF-β signaling pathway and is a member of the SAMD protein family. In the classical TGF-β/SAMD3 pathway, phosphorylated SAMD2/3 and SAMD4 form a transcriptional complex that translocates to the nucleus and participates in the transcriptional regulation of several target genes [13]. It was shown that in lung epithelial cells, SAMD3 and ATOH8 interact to form a transcriptional complex capable of directly repressing the expression of key genes associated with cell cycle progression. This molecular mechanism not only accelerated the senescence process of lung epithelial cells, but also further promoted the transformation of senescent cells to a malignant phenotype [14]. In addition, single nucleotide polymorphisms (SNPs) in the SMAD3 gene were significantly associated with the prognosis of patients with lung adenocarcinoma (LUAD) treated with gefitinib, which may serve as an important molecular marker for predicting treatment response and survival [15]. In cervical cancer, phosphorylation of SAMD3 has been shown to significantly increase the transcriptional activity of the FOXP3 gene, which in turn induces the expansion of regulatory T lymphocytes. This process is closely associated with immune escape from cervical cancer and the immunosuppressive state of the tumor microenvironment, which ultimately leads to poor patient prognosis [16]. However, the relationship between Samd3 and Ki67 is not well understood.
In summary, Smad3 protein, as a key regulator of the TGF-β signaling pathway, plays an important role in tumorigenesis and development by phosphorylation at its S204 site. Meanwhile, Ki67, as a marker of cell proliferation, has been widely used in the prognostic assessment of various malignant tumors. However, the co-expression of Smad3 S204 phosphorylation and Ki67 in GC and its effect on tumor metastasis and prognosis have not been fully investigated. Therefore, the aim of this study was to investigate the expression characteristics of Smad3 S204 phosphorylation and Ki67 in GC tissues, to analyze their correlation with clinicopathological parameters and prognosis, and to explore the potential value of co-expression of pSmad3 (S204) and Ki67 in GC metastasis and prognosis.
Methods and materials
Patients and specimens
Patients and specimens
The microarrays were obtained from Shanghai Outdo Biotech Company, the diagnosis of the microarray samples was determined according to American Joint Committee on Cancer (AJCC) criteria [17], the histological grading of the tumors followed the World Health Organization (WHO) guidelines for classification of tumors of the gastrointestinal system [18], including 98 cases of cancerous and paraneoplastic tissues, and the use of the samples was approved by the Ethics Committee of Shanghai Outdo Biotech Company (Shanghai, China) (NO.: SHYJS-CP-1801009).
IHC
Immunohistochemical (IHC) staining experiments were performed as previously described [19]. Specifically, paraffin-embedded tissue Sect. (4 μm in thickness) were deparaffinized and fully hydrated, followed by antigenic repair by thermal induction in citrate buffer (pH 6.0) to ensure adequate exposure of antigenic epitopes.
Subsequently, endogenous peroxidase activity was blocked using H₂O₂ to eliminate background interference. Next, the tissue sections were incubated overnight at 4 °C with the following primary antibodies: anti- CD133(item number 18470-1-AP), anti-Ki67 (item number GM724007), anti- E-cad ((item number PA073), anti-PSMAD3L (S204) (item number 2816414), respectively. The next day, the sections were washed three times with phosphate buffer solution (PBS) to remove unbound primary antibodies. The sections were then incubated with anti-rabbit HRP-conjugated IgG secondary antibody (Zsbio, China) for 1 h at 37 °C to ensure specific binding. Meanwhile, sections treated with PBS alone were used as a negative control to exclude non-specific staining. At the end of the incubation, the sections were washed again with PBS and incubated with peroxidase substrate (Zsbio, China) for 20 min at 37 °C to develop the color reaction. Finally, the nuclei were restained with hematoxylin to enhance the visualization of nuclear structures. After completion of staining, the sections were dehydrated and coverslipped for long-term storage and microscopic observation. Staining results were independently assessed by two pathologists who were unaware of the background of the study, based on the German semi-quantitative scoring method [20]. To assess inter-rater reliability, 20% of the Sect. (19 cases of GC tissue and 16 cases of adjacent normal tissue) were randomly selected for double-blind re-scoring. Cohen’s kappa coefficient was used to quantify the agreement between the two pathologists. The kappa values for CD133, E-cad, Ki67, and pSmad3L (S204) were 0.810, 0.939, 0.875, and 0.810, respectively, indicating excellent inter-observer agreement. In cases of scoring discrepancies, the two pathologists would conduct joint discussions to reach a consensus before determining the final score. The scoring criteria took into account the overall intensity of the staining as well as the percentage of positively stained cells as an objective reflection of the expression level of the target proteins. The overall staining intensity was graded into four levels: no staining (0 points), light yellow staining (1 point), yellow staining (2 points), and dark yellow/brown staining (3 points). The percentage of positive staining was then defined as the proportion of positively stained glandular epithelial cells in the tissue on the slide and scored on the following scale: <5% (0 points); 5%−25% (1 point); 25%−50% (2 points); 50%−75% (3 points); >75% (4 points). By combining the total staining intensity score with the positive staining percentage score, protein expression levels were further characterized as: negative (0–2 points), + (3–5 points), ++ (6–8 points), and +++ (9–12 points). Finally, all tissue samples were categorized into low expression groups (- or +) and high expression groups (+ + or ++++) based on the scoring results [19].
Statistical analysis
Data were analyzed using SPSS 26.0 and R language (4.4.0). Measurement data were assessed for normality by the Kolmogorov-Smirnov test; normally distributed data were expressed as mean ± standard deviation (Mean ± SD), and comparisons between groups were made using the independent samples t-test; non-normally distributed data were expressed as median and interquartile range (Median [IQR]), and comparisons between groups were made using the Mann-Whitney U test. Categorical data were expressed as frequencies and percentages (n [%]), and the chi-square test was used for comparisons between groups. Survival analysis was performed by the Kaplan-Meier method, and Log-rank test was used for between-group comparisons. Clinical prediction models were constructed by screening variables with P <.05 by one-way Cox regression, and further by screening independent prognostic factors with P <.05 by multifactorial Cox regression, and forest plots were plotted using the “forestplot” package, and column line plots were plotted using the “rms” package. The “forestplot” package was used to draw forest plots, and the “rms” package was used to draw column plots. The “pROC” package was used to draw the ROC curve and calculate the AUC to assess the differentiation, the “rms” package was used to draw the calibration curve to assess the calibration degree, and the “rmda” package was used to draw the decision curve to analyze the clinical utility of the model. The “rms” package plotted calibration curves to assess calibration, and the “rmda” package plotted decision curves to analyze and assess clinical utility. p <.05 was considered a statistically significant difference.
Results
Clinical data
A total of 98 GC patients were included in this study, and their detailed clinicopathological characteristics are shown in Table 1. The median follow-up period of this study was approximately 1245 days. During the follow-up period, 66 (67.3%) of the GC patients were confirmed dead.
Table 1.
Clinical baseline table of patients with GC [cases (%)]
| Variable | levels | Overall |
|---|---|---|
| Gender, n(%) | Females | 36 (36.7%) |
| Male | 62 (63.3%) | |
| Age, n(%) | ≤ 60 | 33 (33.7%) |
| >60 | 65 (66.3%) | |
| Tumor type, n(%) | Mucinous Adenocarcinoma | 8 (8.2%) |
| Adenocarcinoma | 54 (55.1%) | |
| Tubular Adenocarcinoma | 19 (19.4%) | |
| Undifferentiated Carcinoma | 12 (12.2%) | |
| Signet ring cell carcinoma | 5 (5.1%) | |
| Pathological grade, n(%) | Ⅱ | 14 (14.3%) |
| Ⅲ | 72 (73.5%) | |
| Ⅳ | 12 (12.2%) | |
| Tumor size, n(%) | ≤ 5 cm | 61 (62.2%) |
| >5 cm | 37 (37.8%) | |
| T Stage, n(%) | T1 | 6 (6.1%) |
| T2 | 9 (9.2%) | |
| T3 | 64 (65.3%) | |
| T4 | 19 (19.4%) | |
| N Stage, n(%) | N0 | 21 (21.4%) |
| N1 | 16 (16.3%) | |
| N2 | 26 (26.5%) | |
| N3 | 35 (35.7%) | |
| Metastasis, n(%) | M0 | 89 (90.8%) |
| M1 | 9 (9.2%) | |
| AJCC Stage, n(%) | Ⅰ | 9 (9.2%) |
| Ⅱ | 31 (31.6%) | |
| Ⅲ | 49 (50.0%) | |
| Ⅳ | 9 (9.2%) |
Comparison of the expression of CD133, E-cad, Ki67, pSmad3L (S204) proteins in GC tissues and normal tissues adjacent to the cancer
In order to investigate the roles of CD133, E-cad, Ki67, and pSmad3L (S204) in the metastasis and progression of GC, in this study, the expression of the above indexes was detected in 98 cases of GC tissues and 82 cases of matched normal tissues adjacent to the cancer through IHC staining method, and 82 matched paracancerous normal tissues, as shown Fig. 1. The high expression rates of CD133 (56.1% (55/98)), E-cad (66.3% (65/98)) and Ki67 (64.3% (63/98)) in GC tissues were significantly higher than those in normal tissues adjacent to the cancer (13.4% (11/82), 26.9% (22/82), 7.3% (6/82)) and the differences were statistically significant (all P <.05). The high expression rate of pSmad3L (S204) was 31.7 (31/98), which was significantly lower than that of normal tissues next to cancer (79.3% (65/82)), and the difference was statistically significant (P <.001), as shown in Table 2.
Fig. 1.
Expression of CD133, E-cad, Ki67 and pSmad3L (S204) in GC. This figure demonstrates the typical IHC staining results of CD133, E-cad, Ki67 and pSmad3L (S204) in GC tissues and normal gastric mucosa tissues adjacent to the cancer (×20, 100 μm);
Table 2.
CD133, E-cad, Ki67, pSmad3L (S204) protein expression in GC and normal gastric tissues adjacent to the cancer [cases (%)]
| CD133 | P value | Ecad | P value | Ki67 | P value | pSmad3L（S204） | P value | |||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Tumor | Ad-tissue | Tumor | Ad-tissue | Tumor | Ad-tissue | Tumor | Ad-tissue | |||||
| - | 2(2.0) | 20(24.4) | <0.001 | 10(10.2) | 1（1.2) | <0.001 | 1(1.0) | 23(28.0) | <0.001 | 12(12.2) | 4(4.9) | <0.001 |
| + | 41(41.8) | 51(62.2) | 23(23.5) | 59(72.0) | 34(34.7) | 53(64.6) | 55(56.1) | 13(15.9) | ||||
| ++ | 46(47.0) | 11(13.4) | 35(35.7) | 19(23.2) | 31(31.6) | 5(6.1) | 23(23.5) | 30(36.6) | ||||
| +++ | 9(9.2) | 0（0） | 30(30.6) | 3(3.70) | 32(32.7) | 1(1.2) | 8(8.2) | 35(42.7) | ||||
Relationship between the expression of CD133, E-cad, Ki67, pSmad3L (S204) in GC tissues and the clinicopathological features of patients
In order to investigate the effects of the expression of Ki67 and pSmad3L (S204) and their downstream molecules on the metastasis and progression of GC, we analyzed the expression levels of CD133, E-cad, Ki67, pSmad3L (S204) and their downstream molecules in GC tissues of 98 cases of GC by means of chi-square test and Fisher exact probability. We analyzed the relationship between the expression levels of CD133, E-cad, Ki67, pSmad3L (S204) and the clinicopathological features of GC patients by chi-square test and Fisher’s exact probability method. The results of the correlation of clinicopathological features showed that the high expression of CD133 was correlated with tumor type (P =.044); the high expression of Ki67 was correlated with tumor grade (P =.043) and distant metastasis (P =.007); and the differences of the expression of E-cad and pSmad3L (S204) in different clinicopathological features were not statistically significant, as shown in Table 3.
Table 3.
Relationship between the protein expression levels of CD133, E-cad, Ki67, pSmad3L (S204) and clinicopathologic features of GC patients [cases (%)]
| Clinicopathological factors | CD133 | Ecad | Ki67 | pSmad3L(S204) | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| precedent | Low n(%) | High n(%) | P value | Low n(%) | High n(%) | P value | Low n(%) | High n(%) | P value | Low n(%) | High n(%) | P value | |
| Gender | |||||||||||||
| Female | 36 | 16(44.4%) | 20(55.6%) | 0.928 | 13(36.1%) | 23(63.9%) | 0.708 | 10(27.8%) | 26(72.2%) | 0.262 | 22(61.1%) | 14(38.9%) | 0.352 |
| Male | 62 | 27(43.5%) | 35(56.5%) | 20(32.3%) | 42(67.7%) | 25(40.3%) | 37(59.7%) | 45(72.6%) | 17(27.4%) | ||||
| Age | |||||||||||||
| <60 | 33 | 13(39.4%) | 20(60.6%) | 0.727 | 12(36.4%) | 21(63.6%) | 0.945 | 14(42.4%) | 19(57.6%) | 0.138 | 22(66.7%) | 11(33.3%) | 0.775 |
| ≥ 60 | 65 | 30(46.2%) | 35(53.8%) | 21(32.3%) | 44(67.7%) | 21(32.3%) | 44(67.7%) | 45(69.2%) | 20(30.8%) | ||||
| Tumor type | |||||||||||||
| Mucinous Adenocarcinoma | 8 | 6(75.0%) | 2(25.0%) | 0.044 | 3(37.5%) | 5(62.5%) | 0.55 | 4(50.0%) | 4(50.0%) | 0.229 | 8(100.0%) | 0(0.0%) | 0.082 |
| Adenocarcinoma | 54 | 18(33.3%) | 36(66.7%) | 16(29.6%) | 38(70.4%) | 17(31.5%) | 37(68.5%) | 36(66.7%) | 18(33.3%) | ||||
| Tubular Adenocarcinoma | 19 | 11(57.9%) | 8(42.1%) | 5(26.3%) | 14(73.7%) | 7(36.8%) | 12(63.2%) | 12(63.2%) | 7(36.8%) | ||||
| Undifferentiated Carcinoma | 12 | 7(58.3%) | 5(41.7%) | 5(41.7%) | 7(58.3%) | 4(33.3%) | 8(66.7%) | 8(66.7%) | 4(33.3%) | ||||
| Signet ring cell carcinoma | 5 | 1(20.0%) | 4(80.0%) | 4(80.0%) | 1(20.0%) | 3(60.0%) | 2(40.0%) | 3(60.0%) | 2(40.0%) | ||||
| Pathological grade | |||||||||||||
| Ⅱ | 14 | 8(57.1%) | 6(42.9%) | 0.11 | 5(35.7%) | 9(64.3%) | 0.654 | 4(28.6%) | 10(71.4%) | 0.043 | 9(64.3%) | 5(35.7%) | 0.624 |
| Ⅲ | 72 | 28(38.9%) | 44(61.1%) | 23(31.9%) | 49(68.1%) | 27(37.5%) | 45(62.5%) | 50(69.4%) | 22(30.6%) | ||||
| Ⅳ | 12 | 7(58.3%) | 5(41.7%) | 5(41.7%) | 7(58.3%) | 4(33.3%) | 8(66.7%) | 8(66.7%) | 4(33.3%) | ||||
| Tumor size | |||||||||||||
| ≤ 5 cm | 61 | 26(42.6%) | 35(57.4%) | 0.596 | 21(34.4%) | 40(65.6%) | 0.978 | 21(34.4%) | 40(65.6%) | 0.198 | 46(75.4%) | 15(24.6%) | 0.117 |
| >5 cm | 37 | 17(45.9%) | 20(54.1%) | 12(32.4%) | 25(67.6%) | 14(37.8%) | 23(62.2%) | 21(56.8%) | 16(43.2%) | ||||
| T Stage | |||||||||||||
| T1 | 6 | 4(66.7%) | 2(33.3%) | 0.234 | 1(16.7%) | 5(83.3%) | 0.87 | 4(66.7%) | 2(33.3%) | 0.683 | 4(66.7%) | 2(33.3%) | 0.852 |
| T2 | 9 | 7(77.8%) | 2(22.2%) | 3(33.3%) | 6(66.7%) | 4(44.4%) | 5(55.6%) | 6(66.7%) | 3(33.3%) | ||||
| T3 | 64 | 28(43.8%) | 36(56.3%) | 21(32.8%) | 43(67.2%) | 20(31.3%) | 44(68.8%) | 45(70.3%) | 19(29.7%) | ||||
| T4 | 19 | 4(21.1%) | 15(78.9%) | 8(42.1%) | 11(57.9%) | 7(36.8%) | 12(63.2%) | 12(63.2%) | 7(36.8%) | ||||
| N Stage | |||||||||||||
| N0 | 21 | 11(52.4%) | 10(47.6%) | 0.836 | 9(42.9%) | 12(57.1%) | 0.298 | 9(42.9%) | 12(57.1%) | 0.807 | 17(81.0%) | 4(19.0%) | 0.475 |
| N1 | 16 | 5(31.3%) | 11(68.8%) | 5(31.3%) | 11(68.8%) | 7(43.8%) | 9(56.3%) | 13(81.3%) | 3(18.8%) | ||||
| N2 | 26 | 13(50.0%) | 13(50.0%) | 9(34.6%) | 17(65.4%) | 9(34.6%) | 17(65.4%) | 18(69.2%) | 8(30.8%) | ||||
| N3 | 35 | 14(40.0%) | 21(60.0%) | 10(28.6%) | 25(71.4%) | 10(28.6%) | 25(71.4%) | 19(54.3%) | 16(45.7%) | ||||
| Metastasis | |||||||||||||
| M0 | 89 | 41(46.1%) | 48(53.9%) | 0.578 | 32(36.0%) | 57(64.0%) | 0.297 | 33(37.1%) | 56(62.9%) | 0.007 | 62(69.7%) | 27(30.3%) | 0.630 |
| M1 | 9 | 2(22.2%) | 7(77.8%) | 1(11.1%) | 8(88.9%) | 2(22.2%) | 7(77.8%) | 5(55.6%) | 4(44.4%) | ||||
Effect of CD133, E-cad, Ki67, pSmad3L (S204) in GC tissues on patients’ overall survival
To investigate the prognostic value of the expression of CD133, E-cad, Ki67, pSmad3L (S204) in patients with GC, in this study, we calculated the overall survival rate of patients with GC after operation by Kaplan-Meier analysis and plotted the Survival curves were calculated and plotted by Kaplan-Meier analysis. The results of Kaplan-Meier analysis showed that the 5-year survival rates of patients with high expression of CD133, E-cad, Ki67, and pSmad3L (S204) were 36.4%, 43.1%, 36.5%, and 19.4%, respectively; and those with low expression of CD133, E-cad, Ki67, and pSmad3L (S204) were 46.5%, 36.4%, 48.6%, and 50.7%, respectively. The median survival of those with high expression of pSmad3L (S204) was significantly shorter than that of those with low expression, and the difference was statistically significant (P =.022; 660 days vs. 1890 days); there was no statistically significant difference between the expression of CD133, E-cad, and Ki67, and the overall survival of the patients, as detailed in Fig. 2.
Fig. 2.
Kaplan Meier survival curve analysis: the relationship between the protein expression levels of CD133, E-cad, Ki67, and pSmad3L (S204) and the prognosis of patients with GC. a the relationship between the protein expression level of CD133 and OS in patients with GC; (b) the relationship between the protein expression level of E-cad and OS in patients with GC; (c) the relationship between the protein expression level of Ki67 and OS in patients with GC; (d) the relationship between the protein expression level of pSmad3L (S204) and OS in patients with GC;
Effect of combined analysis of Ki67 and pSmad3L (S204) in GC tissues on patients’ overall survival
To further investigate the value of the combined expression of Ki67 and pSmad3L (S204) on the prognosis of patients with GC, in this study, we calculated the postoperative overall survival rate of patients with GC and plotted the survival curves by Kaplan-Meier analysis. Kaplan-Meier analysis showed that the 5-year survival rates of patients with Ki67(+) pSmad3L(S204)(+), Ki67(-) pSmad3L(S204)(-), Ki67(+) pSmad3L(S204)(-), Ki67(-) pSmad3L(S204)(+), and Ki67(-) pSmad3L(S204)(+) were 9.0%, 50.0%, 51.2%, and 44.4%.The median survival of patients with high expression of both Ki67 and pSmad3L(S204) was significantly shorter than that of the other groups, and the difference was statistically significant (P <.001; 555 days vs. 1845 days), as shown in Fig. 3.
Fig. 3.
Kaplan-Meier survival curve analysis: Prognostic relationship between Ki67 combined pSmad3L (S204) protein expression levels and GC patients; (a) patients with high/low Ki67 expression and/or high/low pSmad3L (S204) expression were analyzed; (b) patients with high expression of both Ki67 and pSmad3L (S204) were analyzed in comparison with the remaining patients;
Prognostic value of Ki67, pSmad3L (S204) expression in patients with GC
In order to investigate the factors affecting postoperative survival of patients with GC, the present study was conducted by the COX proportional risk model to analyze the gender, age, tumor type, tumor grade, maximum tumor diameter, depth of infiltration, lymph node metastasis, distant metastasis, AJCC stage, CD133, E-cad, Ki67, and pSmad3L (S204) expression levels were analyzed. The results of the univariate COX proportional risk regression model showed that tumor type, tumor grade, depth of infiltration, lymph node metastasis, distant metastasis, AJCC stage, Ki67, and pSmad3L (S204) were all predictive factors for the prognosis of patients with GC (HR = 1.32; HR = 1.75; HR = 1.67; HR = 1.52; HR = 2.83; HR = 1.87; HR = 1.39; HR = 1.51; all P <.05). There were no statistically significant differences in gender, age, CD133, and E-cad differences. Variables with statistically significant differences in the univariate analysis were included in the multifactorial COX proportional risk regression model analysis, and the results showed that the depth of infiltration, lymph node metastasis, Ki67, and pSmad3L (S204) were independent risk factors predicting the prognosis of patients with GC (HR = 1.65; HR = 1.44; HR = 1.63; HR = 1.55; all P <.05). See Table 4 for details.
Table 4.
Analysis of unifactorial and multifactorial COX proportional risk regression models affecting the prognosis of patients with GC
| Characteristics | univariate analysis | multivariate analysis | ||||
|---|---|---|---|---|---|---|
| HR | 95%CI | P value | HR | 95%CI | P value | |
| Gender | 0.77 | 0.47–1.26 | p =.301 | |||
| Age | 1.3 | 0.77–2.19 | p =.333 | |||
| Tumor type | 1.32 | 1.04–1.66 | p =.021 | 1.22 | 0.90–1.64 | p =.206 |
| Pathological grade | 1.75 | 1.09–2.80 | p =.021 | 1.39 | 0.79–2.44 | p =.257 |
| Tumor size | 1.52 | 0.93–2.50 | p =.097 | |||
| T Stage | 1.67 | 1.15–2.42 | p =.007 | 1.65 | 1.05–2.59 | p =.030 |
| N Stage | 1.52 | 1.21–1.91 | p <.001 | 1.44 | 1.07–1.94 | p =.017 |
| Metastasis | 2.83 | 1.34–5.98 | p =.007 | 2.32 | 0.76–7.04 | p =.138 |
| AJCC Stage | 1.87 | 1.32–2.64 | p <.001 | 0.88 | 0.46–1.66 | p =.682 |
| CD133 | 1.21 | 0.84–1.73 | p =.300 | |||
| Ecad | 0.91 | 0.71–1.17 | p =.465 | |||
| Ki67 | 1.39 | 1.04–1.87 | p =.026 | 1.63 | 1.20–2.21 | p =.002 |
| pSmad3L(S204) | 1.51 | 1.11–2.06 | p =.008 | 1.55 | 1.10–2.18 | p =.013 |
Establishment of clinical prediction model
The independent risk factors were screened in this study, and forest plots and nomograms were drawn according to the risk prediction model, as shown in Figs. 4 and 5. Evaluation of differentiation: The results of the ROC curve analysis showed that the area under the curve (AUC) of the prediction model for the survival rate of GC patients in the 1 st, 3rd, and 5th years after surgery was 0.74 (95% CI: 0.618), 0.76 (95% CI: 0.666–0.857), 0.79 (95% CI: 0.699–0.877), and 0.79 (95% CI: 0.699–0.877) respectively. ~0.870), 0.76 (95% CI: 0.666–0.857), and 0.79 (95% CI: 0.699–0.877), respectively. This indicates that the prediction model has better predictive efficacy and higher discriminative ability for the prognosis of GC patients, as shown in Fig. 6A.Calibration evaluation: the results of the calibration curve analysis showed that the predicted probability of the prediction model and the actual probability of the prediction model showed a higher degree of agreement in the first year of the postoperative period of the GC patients, but the degree of agreement gradually declined in the third year and the fifth year. This indicates that the prediction model has good calibration ability in the first year after surgery for GC patients. For details, see Fig. 6B, C and D for details. Evaluation of effectiveness: The DCA curve analysis results show that the DCA curves of this prediction model in the 1 st, 3rd, and 5th years after the operation of GC patients are located in the upper right of the intersection point of ALL curve and None curve. See Fig. 6E for details.
Fig. 4.
Forest plot for multifactor COX risk proportionality modeling;
Fig. 5.
Prognostic column line diagram for GC patients;
Fig. 6.
Evaluation of the multivariable Cox proportional hazards model: (A) ROC curves for the column-line diagram model; (B) Calibration curves for GC patients at year 1 (C) Calibration curves for GC patients at year 3 (D) Calibration curves for GC patients at year 5 E shows the decision curve of the line diagram
Effect of Ki67 and pSmad3L (S204) expression on overall survival in GC patients with lymph node metastasis
To further investigate the relationship between Ki67 and pSmad3L (S204) protein expression and survival in GC patients with lymph node metastasis, the present study was conducted to calculate the overall survival rate of GC patients with lymph node metastasis after surgery by Kaplan-Meier analysis and to plot the survival curves. The results showed that the 5-year survival rates of patients with high expression of Ki67 and pSmad3L(S204) were 23.8% and 8.3%, respectively, and the 5-year survival rates of patients with low expression were 36.8% and 40.5%, respectively. High pSmad3L(S204) expressors had a significantly shorter median survival than low expressors, and the difference was statistically significant (P =.015; 600 days vs. 840 days); there was no statistically significant difference between Ki67 expression and overall survival of patients, as shown in Fig. 7.
Fig. 7.
Kaplan Meier survival curve analysis: relationship between Ki67, pSmad3L (S204) protein expression levels in GC patients with lymph node metastasis and the prognosis of GC patients. a Relationship between Ki67 expression levels and OS in GC patients with lymph node metastasis; (b) Relationship between pSmad3L (S204) expression levels and OS in GC patients with lymph node metastasis
Discussion
Ki67, a marker of cell proliferation, is an indicator of the malignancy and aggressiveness of tumor cells [21]. Our results showed that the high expression of Ki67 in GC tissues was significantly higher than that in normal tissues adjacent to the cancer (P =.001), and the high expression of Ki67 was closely correlated with M stage (P =.007) and tumor grade (P =.043). This result preliminarily indicated that the high expression of Ki67 might be highly correlated with the occurrence of GC. Yerushalmi [22] and others have shown that Ki67 can be used as a predictive and prognostic marker for breast cancer, and an expression level higher than 10–14% indicates that the patient has a high prognostic risk. Wei [23] further found that high expression of Ki67 was significantly correlated with poor prognosis and disease progression in lung cancer patients by meta-analysis, suggesting that it can be used as a potential biomarker for lung cancer. La Rosa [24] and others confirmed that the ability of Ki67 to predict the prognosis of tumors is tissue-specific, and that its clinical value varies depending on the type of tumor and the site of origin. To further explore the prognostic value of Ki67, we performed univariate and multivariate Cox regression analysis. The results showed that high expression of Ki67 was an independent risk factor for the prognosis of GC patients (HR = 1.63, 95%CI: 1.20–2.21, P =.002). This finding suggests that Ki67 may be involved in the metastatic process of GC by promoting the proliferation and invasion of tumor cells. Meanwhile, the high expression of Ki67 was closely associated with poorer prognosis of patients, suggesting that it may be a new target for GC treatment. It is worth noting that although the multivariate model identified Ki67 as an independent risk factor, there was no significant statistical difference between the expression level of Ki67 and overall survival (P =.18). This result indicates that the multivariate analysis effectively controlled confounding factors, thereby revealing the true prognostic value of Ki67, and also suggests the complexity of interactions between multiple molecular pathways in the progression of gastric cancer.
The TGF-β signaling pathway exhibits a dual role in cancer progression. In the early stage of cancer, TGF-β exerts its tumor suppressor effects by inducing cell cycle arrest and promoting apoptosis. However, with tumor progression, TGF-β promotes cancer value-addition and infiltration by promoting epithelial-mesenchymal transition (EMT), regulating the tumor microenvironment, invasion, evasion of immune surveillance and metastatic spread [25]. The Smad protein family serves as a key mediator of the TGF-β signaling pathway, helping TGF-β to enter the nucleus to mediate transcriptional activation of multiple target genes during signal transduction to achieve tumor suppression or tumorigenesis [26], thus serving as a potential immunotherapeutic target for a variety of malignancies [27].
In this study, we found that high pSmad3L(S204) expression was significantly associated with shorter overall survival in GC patients (P =.022; 660 days vs. 1890 days). Notably, in the subgroup with lymph node metastasis, the survival of patients with high pSmad3L(S204) expression was further shortened (P =.015; 600 days vs. 840 days), a finding that is highly consistent with previous studies [28, 29]. By univariate and multivariate COX regression analyses, we confirmed that high expression of pSmad3L(S204) was an independent risk factor for the prognosis of GC patients (HR = 1.55, 95%CI: 1.10–2.18, P =.013), suggesting that it has a high predictive value for the prognosis of GC patients. Yoshida [30] et al. found that Smad phosphorylation isoform signaling may serve as a potential biomarker for predicting the effect of drug intervention in hepatic fibrocellular carcinoma. In addition, Hori [31] et al. confirmed the significant prognostic value of pSmad3L in pancreatic intraductal papillary mucinous neoplasia (IPMN) by immunohistochemistry, which provides more reliable evidence for predicting malignant transformation and prognosis of IPMN. Based on the above research findings, we believe that pSmad3L(S204) not only serves as a reliable predictor of GC prognosis, but also may be a potential tumor therapeutic target. This finding provides a new idea for the precision treatment of GC and lays a theoretical foundation for the development of targeted therapeutic strategies based on pSmad3L(S204).
It is worth noting that although this model has good discriminative ability and clinical utility, its calibration performance has an obvious limitation: the predicted survival rate is in good agreement with the actual survival rate only in the first year after surgery, and gradually decreases in the 3rd and 5th years. This limitation may stem from two factors: (1) The medium sample size (98 GC patients) and the reduced number of long-term follow-up events (40 patients completed 5-year follow-up) limit the statistical accuracy of long-term survival rate estimation; (2) Unmeasured confounding factors in the model, such as the high heterogeneity of GC in the progression of the disease and post-progression treatment strategies (such as salvage therapy, targeted therapy, etc.), which may weaken the consistency between the predicted long-term survival rate and the actual long-term survival rate. In the future, we plan to include variables related to the long-term progression of GC to improve the calibration effect of the model on long-term survival rates. In addition, we can also combine multi-center as an external validation cohort to enhance its applicability in different populations.
Samd3, as a non-classical phosphorylated form in the TGF-β/Smad signaling pathway, shows a significant positive correlation with the expression level of vascular endothelial growth factor receptor 1 (VEGFR-1) (r =.201, P =.047). Existing studies have confirmed that VEGFR-1 can provide “nutritional support” and “metastatic soil” for the proliferation activity of tumor cells mediated by Ki67 by promoting tumor angiogenesis and constructing a pre-metastatic microenvironment, thereby indirectly enhancing the proliferation-driven effect of Ki67 [32]. In addition, pSmad3L (S204) can also participate in the TGF-β-driven EMT process, induce the recruitment of immunosuppressive cells in the tumor microenvironment, and construct an immunosuppressive microenvironment [33, 34]. Tumor cells with high Ki67 expression can further escape immune surveillance under this background of immune escape, achieve uncontrolled rapid proliferation, and form a synergistic cycle of “immunosuppression - accelerated proliferation”, which together promote the progression of GC. To further evaluate the synergistic effect of Ki67 and pSmad3L(S204) in the prognosis of GC, we performed co-expression analysis. The results showed that GC patients with high expression of both Ki67 and pSmad3L(S204) had significantly shorter overall survival (P <.001; 555 days vs. 1845 days). This finding not only confirmed the independent prognostic value of Ki67 and pSmad3L(S204) in GC progression, but also revealed the possible synergistic effect of the two in promoting malignant tumor progression, which will provide a more comprehensive molecular basis for individualized treatment of GC.
Conclusion
High expression of Ki67 and high expression of pSmad3L(S204) are independent risk factors for the prognosis of patients with GC, and the combined analysis of Ki67 and pSmad3L(S204) can help to improve the risk prediction.
Acknowledgements
Acknowledgments This work was supported by the Key Laboratory Project of Digestive Diseases in Jiangxi Province (2024SSY06101), and Jiangxi Clinical Research Center for Gastroenterology (20223BCG74011).
Abbreviations
- GC
Gastric Cancer
- IARC
International Agency for Research on Cancer
- uLMS
Uterine Smooth Muscle Sarcoma
- SNPs
Single Nucleotide Polymorphisms
- LUAD
Lung Adenocarcinoma
- AJCC
American Joint Committee on Cancer
- WHO
World Health Organization
- IHC
Immunohistochemical
- PBS
Phosphate Buffer Solution
- HRP
Horseradish Peroxidase
- SD
Standard Deviation
- IQR
Interquartile Range
- ROC
Receiver Operating Characteristic
- AUC
Area Under the Curve
- TNM
Tumor Node Metastasis
- DCA
Decision Curve Analysis
- EMT
Epithelial-Mesenchymal Transition
- IPMN
Intraductal Papillary Mucinous Neoplasia
Authors’ contributions
Huang DQ and Kang YH conceived and designed the study, analyzed and interpreted the data, and wrote the manuscript; Hu Q, Guo P, Lv SL, Sun HT, Luo LY and Chen RS collected clinical samples and completed the related experiments; All authors approved the final manuscript.
Funding
supported by the National Nature Science Foundation of China (Grants 82360517, 82060450, 81460374, 31460304), Nature Science Foundation of Jiangxi province of China (20232BAB206086, 20192BAB205072, 20203BBGL73206, 2017BCB23086, 2017BAB205062, 2018BAB205050).
Data availability
The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.
Declarations
Ethics approval and consent to participate
The study was conducted in accordance with the Declaration of Helsinki. This study involving patient participants was reviewed and approved by the Ethics Committee of Shanghai Outdo Biotech Company (Shanghai, China) (NO.: SHYJS-CP-1801009). The patients provided their written informed consent to participate in this study. Written informed consent was obtained from all participants.
Consent for publication
Informed consent was obtained from all subjects involved in the study.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- 1.Bray F, Laversanne M, Sung H, Ferlay J, Siegel RL, Soerjomataram I, et al. Global cancer statistics 2022: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin. 2024;74:229–63. [DOI] [PubMed] [Google Scholar]
- 2.Feng RM, Zong YN, Cao SM, Xu RH. Current cancer situation in China: good or bad news from the 2018 Global Cancer Statistics? Cancer Commun (Lond). 2019;39:22. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Joshi SS, Badgwell BD. Current treatment and recent progress in gastric cancer. CA Cancer J Clin. 2021;71:264–79. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Li P, Huang C-M, Zheng C-H, Russo A, Kasbekar P, Brennan MF, et al. Comparison of gastric cancer survival after R0 resection in the US and China. J Surg Oncol. 2018;118:975–82. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Glimelius B, Ekström K, Hoffman K, Graf W, Sjödén PO, Haglund U, et al. Randomized comparison between chemotherapy plus best supportive care with best supportive care in advanced gastric cancer. Ann Oncol. 1997;8:163–8. [DOI] [PubMed] [Google Scholar]
- 6.Jin X, Liu Z, Yang D, Yin K, Chang X. Recent progress and future perspectives of immunotherapy in advanced gastric cancer. Front Immunol. 2022;13:948647. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Gu X, Rong J, Zhu L, Dai Z, Ren S, Chen J, et al. Hepatoid adenocarcinoma of the stomach: discrimination from conventional gastric adenocarcinoma with a computed tomography-based radiomics nomogram. J Gastrointest Oncol. 2024;15:2041–52. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Dong C, Wang Y, Gu X, Lv X, Ren S, Wang Z, et al. Differential diagnostic value of tumor markers and contrast-enhanced computed tomography in gastric hepatoid adenocarcinoma and gastric adenocarcinoma. Front Oncol. 2023;13:1222853. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Duchrow M, Schlüter C, Key G, Kubbutat MH, Wohlenberg C, Flad HD, et al. Cell proliferation-associated nuclear antigen defined by antibody Ki-67: a new kind of cell cycle-maintaining proteins. Arch Immunol Ther Exp (Warsz). 1995;43:117–21. [PubMed] [Google Scholar]
- 10.Scholzen T, Gerdes J. The Ki-67 protein: from the known and the unknown. J Cell Physiol. 2000;182:311–22. [DOI] [PubMed] [Google Scholar]
- 11.Bourdeleau P, Pokossy-Epée J, Hentic O, Tihy M, Awan ZA, Couvelard A, et al. Temporal increase in Ki-67 index in patients with pancreatic neuroendocrine tumours. Endocr Relat Cancer. 2025;32(3):e240321. [DOI] [PubMed]
- 12.Travaglino A, Raffone A, Catena U, De Luca M, Toscano P, Del Prete E, et al. Ki67 as a prognostic marker in uterine leiomyosarcoma: a quantitative systematic review. Eur J Obstet Gynecol Reprod Biol. 2021;266:119–24. [DOI] [PubMed] [Google Scholar]
- 13.Suwanabol PA, Seedial SM, Zhang F, Shi X, Si Y, Liu B, et al. TGF-β and Smad3 modulate PI3K/Akt signaling pathway in vascular smooth muscle cells. Am J Physiol Heart Circ Physiol. 2012;302:H2211–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Liu X, Li X, Wang S, Liu Q, Feng X, Wang W, et al. ATOH8 binds SMAD3 to induce cellular senescence and prevent Ras-driven malignant transformation. Proc Natl Acad Sci U S A. 2023;120:e2208927120. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Zhang L, Li QX, Wu HL, Lu X, Yang M, Yu SY, et al. SNPs in the transforming growth factor-β pathway as predictors of outcome in advanced lung adenocarcinoma with EGFR mutations treated with gefitinib. Ann Oncol. 2014;25:1584–90. [DOI] [PubMed] [Google Scholar]
- 16.Ni H, Zhang H, Li L, Huang H, Guo H, Zhang L, et al. T cell-intrinsic STING signaling promotes regulatory T cell induction and immunosuppression by upregulating FOXP3 transcription in cervical cancer. J Immunother Cancer. 2022;10(9):e005151. [DOI] [PMC free article] [PubMed]
- 17.Edge SB, Compton CC. The American joint committee on cancer: the 7th edition of the AJCC cancer staging manual and the future of TNM. Ann Surg Oncol. 2010;17:1471–4. [DOI] [PubMed] [Google Scholar]
- 18.Fenoglio-Preiser C, Carneiro F, Correa P, Guilford P, Lambert R, Megraud F, et al. Gastric Carcinoma. 2000;1:35–52. [Google Scholar]
- 19.Zou J, Li C, Jiang S, Luo L, Yan X, Huang D, et al. AMPK inhibits Smad3-mediated autoinduction of TGF-β1 in gastric cancer cells. J Cell Mol Med. 2021;25:2806–15. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Cheng X, Li D, Qi T, Sun J, Zhou T, Zheng WV. Objective to identify and verify the regulatory mechanism of DTNBP1 as a prognostic marker for hepatocellular carcinoma. Sci Rep. 2022;12:211. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Li K, Cao YQ, Xiao JJ, Chen H. Expression and clinical significance of TTF-1, NapsinA and Ki-67 in non-small cell lung cancer. Hebei Med J. 2018;40:735–8. [Google Scholar]
- 22.Yerushalmi R, Woods R, Ravdin PM, Hayes MM, Gelmon KA. Ki67 in breast cancer: prognostic and predictive potential. Lancet Oncol. 2010;11:174–83. [DOI] [PubMed] [Google Scholar]
- 23.Wei DM, Chen WJ, Meng RM, Zhao N, Zhang XY, Liao DY, et al. Augmented expression of Ki-67 is correlated with clinicopathological characteristics and prognosis for lung cancer patients: an up-dated systematic review and meta-analysis with 108 studies and 14,732 patients. Respir Res. 2018;19:150. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.La Diagnostic RS. Prognostic, and predictive role of Ki67 proliferative index in neuroendocrine and endocrine neoplasms: Past, Present, and future. Endocr Pathol. 2023;34:79–97. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Huynh LK, Hipolito CJ, Ten Dijke P. A Perspective on the Development of TGF-β Inhibitors for Cancer Treatment. Biomolecules. 2019;9(11):743. [DOI] [PMC free article] [PubMed]
- 26.Sekimoto G, Matsuzaki K, Yoshida K, Mori S, Murata M, Seki T, et al. Reversible Smad-dependent signaling between tumor suppression and oncogenesis. Cancer Res. 2007;67:5090–6. [DOI] [PubMed] [Google Scholar]
- 27.Matsuzaki K, Kitano C, Murata M, Sekimoto G, Yoshida K, Uemura Y, et al. Smad2 and Smad3 phosphorylated at both linker and COOH-terminal regions transmit malignant TGF-β signal in later stages of human colorectal cancer. Cancer Res. 2009;69:5321–30. [DOI] [PubMed] [Google Scholar]
- 28.Majumder S, Bhowal A, Basu S, Mukherjee P, Chatterji U, Sengupta S. Deregulated E2F5/p38/SMAD3 circuitry reinforces the Pro-Tumorigenic switch of TGF beta signaling in prostate cancer %J. J Cell Physiol %J. 2016;231:2482–92. [DOI] [PubMed] [Google Scholar]
- 29.Cho SY, Ha SY, Huang S-M, Kim JH, Kang MS, Yoo H-y, et al. The prognostic significance of Smad3, Smad4, Smad3 phosphoisoform expression in esophageal squamous cell carcinoma %J Medical oncology %J. 2014;31(11):236. [DOI] [PubMed]
- 30.Yoshida K, Matsuzaki K, Murata M, Yamaguchi T, Suwa K, Okazaki K. Clinico-Pathological Importance of TGF-β/Phospho-Smad Signaling during Human Hepatic Fibrocarcinogenesis. Cancers (Basel). 2018;10(6):183. [DOI] [PMC free article] [PubMed]
- 31.Hori Y, Ikeura T, Yamaguchi T, Yoshida K, Matsuzaki K, Ishida M, et al. Role of phosphorylated Smad3 signal components in intraductal papillary mucinous neoplasm of pancreas. Hepatobiliary & Pancreatic Diseases International. 2020;19:581–9. [DOI] [PubMed] [Google Scholar]
- 32.Lv SL, Guo P, Zou JR, Chen RS, Luo LY, Huang DQ. Prognostic significance and relationship of SMAD3 phospho-isoforms and VEGFR-1 in gastric cancer: a clinicopathological study. World J Gastrointest Oncol. 2024;16:118–32. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Wu X, Jin Z, Li B, Lu Y, Hou J, Yao L, et al. Deciphering of intra-tumoural heterogeneity and the interplay between metastasis-associated meta-program and myofibroblasts in gastric cancer. Clin Transl Med. 2025;15:e70319. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Shoji H, Hirano H, Nojima Y, Gunji D, Shinkura A, Muraoka S, et al. Phosphoproteomic subtyping of gastric cancer reveals dynamic transformation with chemotherapy and guides targeted cancer treatment. Cell Rep. 2024;43:114774. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.