This study strictly adheres to the STROBE Statement. Our investigation entails a retrospective analysis of individuals who underwent gastrectomy at the Affiliated Hospital of Qingdao University from July 2010 to October 2012. The study population included gastric cancer patients who underwent surgical resection as their initial treatment, experienced no significant perioperative complications, and had complete clinicopathological data and adequate tissue files. We collected tumor tissues and paired adjacent tissues from 198 patients (132 males and 66 females), aged 25-90 years, with a median age of 68 years. Follow-up for all patients ended in April 2022, and the median follow-up duration was 52.00 months.

We collected pertinent information regarding demographic and clinical parameters from the study participants, encompassing age, gender, time of operation, pathological type, pathological grading, tumor size, tumor site, lymphatic invasion, lymph node positivity, and the pathological staging of tumors was conducted following the 8th edition of the American Joint Committee on Cancer (AJCC) criteria, which classifies stages I-IV based on tumor-node-metastasis (TNM) characteristics. Overall survival (OS) was defined as the interval starting from the initial surgical resection to either 60 months or the occurrence of patient demise. Resected tumor biopsy samples were subjected to fixation in 10% formalin at 4°C for 24 hours, followed by embedding in paraffin for subsequent histological and immunohistochemical examination. This study followed the principles outlined in the Declaration of Helsinki and was granted approval by the Ethics Committee of the Affiliated Hospital of Qingdao University (No.QDFY27398). Prior to participation, written informed consent was obtained from every individual patient.

Immunohistochemistry (IHC) was executed utilizing the streptavidin-biotin technique (Universal DakoCytomation LSAB2 system; DAKO, Glostrup, Denmark). Following deparaffinization of the tissue sections, they underwent autoclave heating at 120°C for 5 minutes, followed by antigen retrieval in a citrate buffer solution (2mM citric acid and 9mM sodium citrate dihydrate, pH 6.0). Rabbit polyclonal antibody against ITGB6 (dilution, 1:500; #28378-1-AP; Proteintech; USA) and Rabbit polyclonal antibody against Rac1 (dilution, 1:500; #24072-1-AP; Proteintech; USA) were utilized. The primary antibody was applied to the slides and incubated in a humid chamber at 4°C for 12-18 hours. Detection of bound antibody was achieved using a modified labeled avidin-biotin reagent for 20 minutes, followed by phosphate-buffered saline washing. A 0.1% diaminobenzidine solution served as the chromogen and was applied for 5 minutes. Slides were subsequently counterstained with Mayer’s haematoxylin for 5-10 minutes.

To perform a semiquantitative analysis of the immunoreactivity of ITGB6 and Rac1 receptors, the H-score was utilized in this study (16). More than 500 tumor cells were counted in each case, and the H-score was calculated by summing the percentages of strongly stained nuclei (3×), moderately stained nuclei (2×), and weakly stained nuclei (1×), resulting in a potential range of 0-300. The score was independently determined by two pathologists who identified the immunostained areas on slides using a double-headed light microscope. In this study, the interobserver differences were found to be less than 5%, and the average of the two values was obtained.

Human gastric carcinoma cell line SCG7901 was obtained from the Cell Bank of the Chinese Academy of Sciences. SCG7901 cells were cultured in RPMI-1640 medium (Gibco, CAT#C11875500BT), supplemented with 10% fetal bovine serum (Gibco, CAT#C11995500BT) and 1% Penicillin/Streptomycin, in a controlled environment with 5% CO₂ at a temperature of 37°C.

ITGB6-specific small interfering RNAs (siRNAs, Ribobio, Guangzhou, China) were employed to selectively suppress ITGB6 expression in SCG7901 cells. The cells were transfected with siRNA (si-NC; si-ITGB6, 5’-GAAAGAUUGUGUUAGUUAAGU-3’, 5’-UUAACUAACACAAUCUUUCUA-3’) using the transfection reagent Lipofectamine 2000 (Invitrogen). Upon achieving a confluence rate of 70% in the medium, 5 μL of siRNA and Lipofectamine 2000 reagents were gently mixed and the culture was continued for 4 hours within a humid environment containing 5% CO₂. On the second day subsequent to transfection of SCG7901 cells, a portion of the si-NC and si-ITGB6 cells were incubated in medium supplemented with NSC233766 (50μmol), designated as the si-NC+NSC233766 group and si-ITGB6+NSC233766 group, respectively. NSC23766 (Cat#: 733767-34-5) was obtained from Med Chem Express (NJ, U.S.A.) and dissolved in double distilled water at a concentration of 10 mM for storage purposes. Following an incubation period of at least 48 hours, the cells were utilized for subsequent experiments.

The Cell Counting Kit-8 (CCK-8, Fude Biological, CAT#FD3788) was employed to assess cellular proliferation. More specifically, cells were seeded onto 96-well plates at an initial density of 2000 cells per well. Following incubation at 37°C for durations of indicated time points, 10 μl of CCK8 solution was added to each well, followed by a coincubation of 2 hours. Subsequently, the absorbance was measured at 450 nm using spectrophotometry.

The experiment employed a 24 well plate with transwell chamber (Corning Costar, CAT#3422) to conduct cell migration and invasion assays. For the migration assay, 600 μl of RPMI-1640 medium with 10% FBS and 600 μl of serum-free RPMI-1640 medium were added to the upper and lower chambers, respectively. Subsequently, 5×10⁴ cells, with or without an inhibitor, were introduced into the upper chamber. Following a 24-hour incubation period, non-migrated cells were eliminated using cotton swabs. The infiltrated cells were fixed with 4% formaldehyde for 15 minutes and stained with 0.5% crystal violet for 10 minutes at room temperature.

Cell invasion assay followed a similar procedure, with the exception of Matrigel (BD Biosciences, CAT#354277) pre-coating the upper chamber. Microscopic images of the cells were captured and cell counting was performed using both a microscope and ImageJ software. All experiments were repeated three times.

The association between ITGB6 and Rac1 expression and clinicopathological variables was evaluated using either the chi-square test or Fisher’s exact test. For categorical variables, frequency and percentage were used. Quantitative data were presented as the mean ± SD. Student’s t test was used to determine the differences between two groups. Differences among multiple groups were determined by one-way analysis of variance (ANOVA). Survival outcomes were examined through the utilization of the Kaplan-Meier method and log-rank test. Using the Receiver Operating Characteristic (ROC) curve, we evaluated the prognostic value of ITGB6 and Rac1 expression in predicting the prognosis and lymph node metastasis in gastric cancer patients. The correlation between ITGB6 and Rac1 expression levels was assessed using the Spearman correlation analysis. Univariate and multivariate analyses of cancer-specific mortality were conducted by implementing the Cox proportional hazards model. Graphics were created using GraphPad Prism 7 (GraphPad Software, Inc, La Jolla, CA, USA) and Photoshop software (Adobe, Version CS5.1). The statistical analysis was carried out by employing the SPSS version 26.0 (IBM, Armonk, NY, USA). In order to establish statistical significance, a significance level of P<0.05 was defined.

The nomogram was developed and validated following the established guidelines for constructing nomograms (17, 18). A nomogram was developed using the independent prognostic factors to predict survival outcomes. Furthermore, the nomograms were utilized for prognostic prediction using the RMS package in R software, version 3.1.3 (https://www.r-project.org/). Model performance was assessed through measures of discrimination and calibration (19). The discrimination ability of the nomogram was assessed using Harrell’s concordance index (C-index) (20). The range of the C-index is between 0.5 and 1.0, where a value of 0.5 represents no discrimination, and a value of 1.0 indicates perfect discrimination. A calibration plot was employed to visually assess the congruence between the predicted prognosis from the nomogram and the actual observed prognosis.

Table 1 illustrates the clinical and pathological characteristics of 198 patients diagnosed with gastric cancer. Among these individuals, 132 (64.7%) were of the male gender while 66 (32.4%) were female, with an average age of 65.86 ± 11.57 years (ranging from 25 to 90 years). The optimal threshold for age was determined based on the survival duration. Likewise, the optimal thresholds for tumor size were defined in relation to survival time, specifically 1-6.5, 7-9, and 9.5-19, respectively. The critical milestone for the positive lymph node rate was fixed at 42.9% taking into account survival time. The gastric cancer tissues were classified in accordance with the international classification system first proposed by the World Health Organization (WHO) back in 1979. These classifications include adenocarcinoma, mucinous adenocarcinoma, sigma-ring cell carcinoma, and undifferentiated carcinoma. Moreover, the anatomical localization of the tumor served as a basis for categorizing gastric cancer into the upper third (comprising the preventriculus and fundus of the stomach), middle third (corresponding to the body of the stomach), and lower third (including the antrum of the stomach and pylorus), constituting 34.3%, 34.3%, and 31.3%, respectively. The pathologic tumor-node-metastasis (TNM) classification and cancer stage were determined in accordance with the eighth edition of the American Joint Committee on Cancer (AJCC) stage groupings. After undergoing surgery, all patients were diligently monitored, and the median survival period was found to be 52 months (ranging from 0.03 to 71 months).

Immunostaining was conducted to ascertain the expression of ITGB6 and Rac1 in gastric carcinoma specimens in comparison to neighboring normal tissues ( Figures 1A–H ). A semiquantitative H-score, based on the overall staining intensity and extent of positive cells, as previously described, was assigned. By employing the H-score, the expression levels of ITGB6 and Rac1 were determined and compared between gastric carcinoma tissues and adjacent normal tissues within the diagnostic cohort. The results showed that ITGB6 and Rac1 were highly expressed in gastric cancer tissues (H-score, 33.87 ± 22.15, 46.99 ± 24.23, respectively) and lymph node metastases (H-score, 36.73 ± 23.69, 50.29 ± 25.96, respectively) compared with adjacent normal tissues (H-score, 22.52 ± 5.61, P<0.001; 40.13 ± 15.37, P<0.001, respectively) and non-metastatic tissues (H-score, 25.60 ± 14.19, 37.44 ± 14.91, respectively, Figures 2A–D ).

The X-tile plot was employed to ascertain the most favorable threshold value of the H-score for evaluating the statistical significance pertaining to the overall survival (OS) of patients (H-score=33.2). Classifying the diagnostic cohort based on this threshold, 198 individuals suffering from gastric cancer were partitioned into categories of high and low ITGB6 expression. Among these patients, 119 out of 198 (60.1%) exhibited low levels of ITGB6 expression, while 79 out of 198 (39.9%) showed high levels. The fundamental attributes of these two groups are comprehensively outlined in Table 1 . ITGB6 expression was significantly associated with pathological grading (P=0.013), tumor size (P=0.030), location (P=0.031), N stage (P=0.002), TNM stage (P=0.002), positive lymph node rate (P=0.002), and survival status (P<0.001) ( Table 1 ).

The examination of Kaplan-Meier survival analysis unveiled that individuals demonstrating elevated ITGB6 expression showcased notably inferior overall survival outcomes in comparison to those with diminished ITGB6 expression(P<0.0001; log-rank test, χ² = 30.845). Figure 3A illustrates the correlation between the expression of ITGB6 and the survival rate of patients over time. Subsequently, a time-dependent receiver operating characteristic (ROC) analysis was performed to evaluate the prognostic significance of ITGB6 expression in individuals diagnosed with gastric cancer. This analysis resulted in an area under the curve (AUC) of 0.708, which indicates a moderate predictive value.(95% CI: 0.636-0.780, sensitivity: 52.8%, specificity: 80.4%) ( Figure 3C ). A ROC curve was also constructed to appraise the potential of ITGB6 as a biomarker in prognosticating lymph node metastasis. The area under the curve (AUC) for ITGB6 in discerning patients with lymph node metastasis stood at 0.604 (95%CI:0.516-0.692, sensitivity:41.7%, specificity:76.6%) ( Figure 3D ).

To assess the statistical significance of H-score, the X-tile plot was employed to ascertain the optimal threshold value, derived from its association with the overall survival (OS) of patients, which was identified as 47.1. Utilizing the optimal threshold of Rac1 expression within the diagnostic cohort, a segregation was performed on the 198 patients afflicted with gastric cancer, resulting in the formation of two distinct groups: one characterized by elevated Rac1 expression, and the other marked by diminished Rac1 expression. Of the total sample size, a majority of 63.6% (126 out of 198) exhibited a diminished level of Rac1 expression, while the remaining 36.4% (72 out of 198) displayed an elevated Rac1 expression. Table 1 showcases the fundamental attributes of these distinct groups. The findings demonstrated a notable association between the expression of Rac1 and various factors, including tumor size (P=0.013), location (P=0.031), N stage (P=0.002), TNM stage (P=0.035), rate of positive lymph nodes (P<0.001), as well as survival status (P<0.001). The prevalence of Rac1 in TNM stage III-IV specimens was determined to be 41.9%, a statistically significant increase compared to the prevalence in TNM stage I-II specimens (27.0%). Nevertheless, no significant associations were observed between Rac1 expression and age, gender, pathological type, pathological grade, lymphatic invasion, neural invasion, T stage, or M stage when employing a significance level of P<0.05 ( Table 1 ).

Patients exhibiting elevated levels of Rac1 expression demonstrated markedly lower overall survival rates compared to patients with negative Rac1 expression (P<0.0001, The log-rank test, χ² = 27.060). Patient survival according to Rac1 expression over time is illustrated in Figure 3B . The predictive prognostic performance of Rac1 was evaluated using ROC analysis, which yielded an AUC of 0.708 (95% CI: 0.636-0.780, sensitivity: 50.9%, specificity: 79.3%) ( Figure 3C ). We additionally crafted ROC curves and computed AUC scores to investigate the potential of Rac1 as a biomarker for prognosticating lymph node metastasis, yielding an AUC value of 0.638. (95% CI: 0.551-0.725, sensitivity: 40.4%, specificity: 74.5%) ( Figure 3D ).

The elevated expression level of Rac1 was observed to be 53.2% in tissues exhibiting high ITGB6 expression, while it was 25.2% in tissues demonstrating low ITGB6 expression. Spearman correlation analysis unveiled a positive association between Rac1 and ITGB6 expression (r=0.285, P<0.001, Table 2 ).

Based on the expression levels of ITGB6 and Rac1, 198 patients were categorized into four distinct groups: group 1 comprised individuals with low levels of both Rac1 and ITGB6 (n=89); group 2 consisted of patients with low Rac1 but high ITGB6 levels (n=37); group 3 included individuals with high Rac1 but low ITGB6 levels (n=30); and group 4 comprised patients with high levels of both Rac1 and ITGB6 (n=42). It was observed that patients in the high Rac1/high integrin αvβ6 group exhibited a significantly poorer overall survival rate compared to the other groups(P<0.0001, The log-rank test, χ² = 35.712). The chart in Figure 4A exhibits the long-term survival outcomes of individuals who possess Rac1 expressing ITGB6. Furthermore, the area under the receiver operating characteristic (ROC) curve (AUC) for the amalgamated two biomarkers escalated to 0.658 (95% CI, 0.582-0.733), accompanied by estimations of sensitivity and specificity amounting to 35.8% and 95.7%, correspondingly ( Figure 4B ).

In this investigation, we conducted univariate and multivariate analyses of the data to ascertain the prognostic significance of ITGB6 and Rac1 expression through the utilization of Cox proportional hazards regression models. The age surpasses the designated threshold (P=0.001), Pathological grading (P=0.016), tumor size (P<0.001), Lymphatic invasion (P=0.001), T stage (P=0.003), N stage (P<0.001), M stage (P<0.001), clinical stage (P<0.001), high αvβ6 expression (P<0.001) and high Rac1 expression (P<0.001) were identified as determinants indicating an unfavorable prognosis in the univariate analysis (as indicated in Table 3 ). Then a significance level of P<0.10 was utilized as a variable in the multivariate analysis. Through this analysis, it was discovered that elevated αvβ6 expression and heightened Rac1 expression served as unfavorable independent prognostic factors (relative risk (RR): 2.212 and 2.073; P<0.001 and P=0.001, respectively). Notably, age, tumor size, and TNM stage also emerged as independent prognostic factors (RR: 2.977, 2.553, and 1.760; P<0.001, P=0.003, and P=0.035, respectively) (As presented in Table 3 ).

The final model for constructing a prognostic nomogram predicting overall survival ( Figure 5A ) incorporated age, tumor size, TNM stage, ITGB6, and Rac1 expression based on the aforementioned findings. Table 4 shows the nomogram prognostic factor scores. In order to assess the discernment of the Nomogram, we employed the C-index. The prognostication model demonstrated remarkable precision, exhibiting a C-index of 0.751(95% CI: 0.704-0.798). Meticulously developed calibration curves were implemented to evaluate the congruity between the prognosticated probabilities derived from the nomogram and the veritable observed survival rates. The calibration plot of postoperative OS showed that the predicted combined expression of ITGB6 and Rac1 based on nomogram was basically consistent with the actual observation ( Figure 5B ). In order to assess the practical value of the OS nomogram, the methodology of decision curve analysis (DCA) was employed to quantify the overall advantage at various threshold probabilities. The clinical applicability and advantages of the OS nomogram were contrasted with those of the TNM using DCA. The graphical representation of DCA demonstrated that the nomogram exhibited superior prediction and clinical relevance compared to TNM ( Figure 5C ).

The gastric cancer cell line 7901 was cultured and the expression of ITGB6 was downregulated by transfection with ITGB6 interfering RNA. The siRNA effect was confirmed by RT-PCR ( Figure 6A ). Subsequently, CCK8 experiments were used to investigate changes in cell activity after ITGB6 interference, and it was found that downregulation of ITGB6 can significantly inhibit gastric cancer cell activity. In addition, the Rac1 activity inhibitor NSC23766 had a similar effect to si-ITGB6 ( Figure 6B ). Transwell migration and invasion experiments (with Matri-gel) were conducted to explore the effects of si-ITGB6 and Rac1 inhibitor on the migration and invasion ability of gastric cancer cells. It was found that downregulation of ITGB6 expression and Rac1 activity inhibition could significantly inhibit the migration and invasion ability of gastric cancer cells ( Figure 6C ). To further validate this effect, we treated gastric cancer cell lines transfected with NC, si-ITGB6, and NSC23766. Simultaneously, experimental results showed that after inhibiting Rac1 activity, the role of ITGB6 in gastric cancer cell activity, migration, and invasion disappeared, suggesting that ITGB6 might play a role in the proliferation, migration, and invasion of gastric cancer cells through Rac1 ( Figures 6D, E ).