Exploration of the association between SF3B4 and HMGB1 expression and the clinicopathological features and prognosis of gastric cancer

Abstract

BACKGROUND

Gastric cancer ranks among the leading malignancies worldwide, noted for its high morbidity and mortality, and remains a significant challenge to global public health.

AIM

To investigate the association between the expression of splicing factor 3b subunit 4 (SF3B4) and high mobility group box 1 (HMGB1) with the clinical characteristics and prognostic outcomes of gastric cancer patients.

METHODS

A retrospective cohort study was conducted involving 114 individuals diagnosed with gastric cancer and admitted to our institution from January 2020 to December 2021. A comparison group of 90 patients diagnosed with benign gastric disorders during the same period was also included. Expression levels of SF3B4 and HMGB1 were assessed using real-time quantitative polymerase chain reaction. Expression patterns were analyzed in relation to various clinicopathological features. Receiver operating characteristic curves were constructed to evaluate the ability of SF3B4 and HMGB1, alone and in combination, to predict unfavorable one-year outcomes. Multivariate logistic regression was utilized to identify independent predictors of mortality. Kaplan-Meier survival curves were generated to examine survival differences based on SF3B4 and HMGB1 expression levels.

RESULTS

Both SF3B4 and HMGB1 were markedly upregulated in tumor tissues of gastric cancer patients compared to adjacent normal tissues and to tissues from non-malignant gastric disease patients (^aP < 0.05). Higher expression levels of these two genes were significantly associated with aggressive pathological features, including poor differentiation, tumor size > 5 cm, deep infiltration (T3-T4), lymph node involvement, and advanced clinical stage (III–IV) (^aP < 0.05). Receiver operating characteristic analysis revealed that the combined use of SF3B4 and HMGB1 yielded an area under the curve of 0.914, surpassing the predictive performance of either marker alone (SF3B4: 0.776; HMGB1: 0.757). Multivariate analysis identified SF3B4 ≥ 1.45, HMGB1 ≥ 0.93, poor differentiation, larger tumor size, deeper invasion, lymph node metastasis, and advanced clinical tumor-node-metastasis staging as independent factors contributing to one-year mortality (^aP < 0.05). Survival analysis indicated that patients with elevated SF3B4 and HMGB1 levels had a shorter median survival (25.74 ± 5.46 months) compared to those with lower expression levels (33.29 ± 6.71 months, log-rank = 10.534, ^aP < 0.05).

CONCLUSION

Elevated SF3B4 and HMGB1 expression in gastric cancer tissue is significantly associated with tumor aggressiveness, worse prognosis, and reduced survival. These biomarkers may offer clinical value in stratifying patients by risk and in forecasting outcomes. Their combined assessment improves predictive accuracy for poor prognosis and may serve as a more effective tool than individual evaluation.

Key Words: SF3B4; HMGB1; Gastric cancer; Pathological characteristics; Prognosis

Core Tip: The expression levels of splicing factor 3b subunit 4 and high mobility group box 1 in gastric biopsy tissues of gastric cancer patients are significantly elevated, and their expression is closely related to disease severity, prognosis, and survival time. They can serve as biomarkers for evaluating disease status and prognosis in gastric cancer, and their combined detection has a higher predictive value for poor prognosis than individual diagnosis.

INTRODUCTION

Gastric cancer is a major malignancy with global distribution and remains one of the leading causes of cancer-related mortality, particularly in East Asia, where incidence and death rates are especially high[1,2]. Due to the insidious onset and lack of distinctive early symptoms, a large proportion of cases are diagnosed at advanced stages, contributing to generally poor clinical outcomes[3,4]. This underscores the importance of identifying molecular indicators that can reflect disease progression and assist in predicting prognosis.

Splicing factor 3b subunit 4 (SF3B4), a vital component of the RNA splicing complex, is involved in the regulation of pre-mRNA processing and has been increasingly recognized for its role in oncogenesis[5]. Previous investigations have indicated that dysregulation of SF3B4 is associated with the development and progression of several tumor types, including liver, breast, and lung cancers[6-8]. Nonetheless, data regarding SF3B4 expression in gastric cancer and its link to tumor behavior and clinical outcomes remain limited.

High mobility group box 1 (HMGB1), a non-histone chromatin-binding protein, participates in multiple cellular processes such as gene transcription, DNA damage repair, and stress response regulation[9]. Apart from its nuclear functions, HMGB1 can also be released extracellularly, where it acts as a cytokine-like mediator influencing inflammation and modulating the tumor microenvironment, thereby facilitating cancer progression and metastasis[10]. Elevated HMGB1 levels have been observed in various cancers and are frequently associated with unfavorable prognosis[11-13]. However, the expression status and prognostic significance of HMGB1 in gastric cancer are not yet fully clarified.

In light of this, the present study retrospectively examined gastric tissue samples from 114 patients with gastric cancer and 90 individuals diagnosed with benign gastric conditions. The expression levels of SF3B4 and HMGB1 were measured using real-time quantitative polymerase chain reaction (RT-qPCR), and their associations with clinicopathological parameters and survival outcomes were analyzed. This study aims to explore the potential utility of SF3B4 and HMGB1 as biomarkers for disease assessment and prognostic evaluation in gastric cancer.

MATERIALS AND METHODS

Study subjects

This study retrospectively reviewed clinical information from patients treated at our institution between January and December 2020. A total of 114 individuals diagnosed with gastric cancer formed the gastric cancer cohort. For comparison, 90 patients diagnosed during the same period with non-malignant gastric conditions, including chronic gastritis, gastric ulcers, and gastric polyps, were selected as the control group. Demographic and baseline clinical indicators, including sex, age, body mass index, eastern cooperative oncology group performance score (range: 0-5, with higher scores indicating poorer functional status), body surface area, Helicobacter pylori infection status, and educational background, were statistically comparable between the two groups (^bP > 0.05), indicating baseline equivalence. Detailed baseline characteristics are summarized in Table 1. The study received approval from the hospital’s ethics committee and was carried out in full accordance with the ethical standards outlined in the Declaration of Helsinki.

Table 1 Comparison of baseline characteristics, n (%).

	Gastric cancer group, n = 114	Non-gastric cancer group, n = 90	t/χ²	P value
Gender			0.242	0.622
Male	71 (62.28)	53 (58.89)
Female	43 (37.72)	37 (41.11)
Age, years	57.45 ± 7.96	57.62 ± 7.85	0.152	0.879
BMI, kg/m²	23.37 ± 2.14	23.19 ± 2.26	0.581	0.561
ECOG score, points	1.73 ± 0.64	1.65 ± 0.59	0.917	0.360
Body surface area, m²	2.26 ± 0.77	2.14 ± 0.82	1.074	0.284
Helicobacter pylori infection			0.418	0.517
Present	101 (88.60)	77 (85.56)
Absent	13 (11.40)	13 (14.44)
Education level			0.408	0.522
High school or below	91 (79.82)	75 (83.33)
College or above	23 (20.18)	15 (16.67)

BMI: Body mass index; ECOG: Eastern Cooperative Oncology Group.

Inclusion and exclusion criteria

Inclusion criteria: (1) Gastric cancer group: All patients had histologically confirmed gastric cancer in accordance with the 2019 edition of the World Health Organization classification of digestive system tumors[14]. Only cases with complete pathological data, such as degree of tumor differentiation, tumor-node-metastasis (TNM) staging, and lymph node involvement, were included. None had undergone preoperative chemotherapy, radiotherapy, or targeted treatment. All participants had paired tumor and adjacent normal gastric tissue samples acquired via surgical resection or gastroscopic biopsy, and samples met quality standards for RT-qPCR analysis. Informed consent was obtained from all participants; and (2) Non-gastric cancer group: Included were patients diagnosed with benign gastric conditions and confirmed negative for malignancy via pathology. Participants had not received any anti-cancer therapy (including radiotherapy, chemotherapy, or targeted agents) prior to tissue sampling. Complete clinical records and signed informed consent were required.

Exclusion criteria: Participants were excluded if they met any of the following conditions: (1) History of or coexisting malignant tumors other than gastric cancer; (2) Severe impairment of major organ function (heart, liver, kidney) or diagnosis of autoimmune disease; (3) Active infectious diseases such as hepatitis B or tuberculosis; (4) Pregnancy or lactation; (5) Inability to complete the study protocol or follow-up due to personal reasons, or missing key clinical data affecting outcome analysis; and (6) Tissue specimens not meeting technical requirements for RT-qPCR, such as severe RNA degradation or sample contamination.

Detection of SF3B4 and HMGB1 expression in tissue samples

In this study, tissue specimens were obtained from all enrolled participants. For patients in the gastric cancer group, both tumor tissues and adjacent non-tumorous tissues were collected during gastroscopic biopsy or surgical procedures. For those in the non-gastric cancer group, lesion tissues were retrieved for comparison. All collected samples were pathologically confirmed, immediately immersed in RNA later tissue preservation solution, and subsequently flash-frozen in liquid nitrogen to preserve RNA integrity. Total RNA was isolated using the Trizol reagent protocol, and the quality of the extracted RNA was assessed via a NanoDrop 2000 UV spectrophotometer. Only samples with an A260/A280 ratio between 1.8 and 2.0 were used for downstream applications. High-purity RNA samples were reverse transcribed into complementary DNA using a reverse transcription kit (Takara, Dalian, China), in a 20 μL total reaction volume, as per manufacturer’s instructions.

Quantitative analysis of SF3B4 and HMGB1 mRNA expression was performed using RT-qPCR. Reactions were carried out with reagents from the Titanium One-Step RT-PCR Kit (Takara, Batch No. 639503, Beijing, China), using the MX3000 real-time polymerase chain reaction (PCR) system (Agilent Technologies, CA, United States). Gene-specific primers were designed, synthesized, and purified by Shanghai Sangon Biotech Co., Ltd. Primer sequences are listed in Table 2. Each PCR reaction (25 μL total volume) consisted of: 2 μL complementary DNA template, 12.5 μL 2 × PCR premix, 0.5 μL of both forward and reverse primers (10 μmol/L), 0.5 μL ROX reference dye, and RNase-free water to make up the final volume. Amplification conditions were as follows: Reverse transcription at 52 °C for 20 minutes, initial denaturation at 94 °C for 1 minute, followed by 32 amplification cycles of denaturation at 94 °C for 15 seconds, annealing at 58 °C for 1 minute, and extension at 64 °C for 40 seconds. A final elongation step was carried out at 72 °C for 8 minutes.

Table 2 Primer sequences for real-time quantitative polymerase chain reaction amplification.

Gene	Primer sequence
SF3B4	F: 5’-GGATGAGAAGGTTAGTGAACCGC-3’
	R: 5’-GGCATAGTCAGCATCTTCCTCAC-3’
HMGB1	F: 5’-AATACGAAAAGGATATTGCT-3’
	R: 5’-GCGCTAGAACCAACTTAT-3’

SF3B4: Splicing factor 3b subunit 4; HMGB1: High mobility group box 1.

Glyceraldehyde-3-phosphate dehydrogenase was used as the internal control gene to normalize expression levels. The relative quantification of SF3B4 and HMGB1 was determined by the 2^-∆∆Ct method. All PCR reactions were conducted in triplicate, and the mean of the three runs was used for subsequent statistical analyses. To confirm the specificity of HMGB1 primers, blast sequence alignment was conducted against the National Center for Biotechnology Information nucleotide database, confirming exclusive alignment with the target gene. Additionally, melt curve analysis was performed post-PCR to verify the absence of nonspecific amplification. Selected amplicons were subjected to Sanger sequencing, which further validated the accuracy of the target sequence amplification.

Follow-up procedures and prognostic evaluation

All individuals diagnosed with gastric cancer were systematically monitored through scheduled outpatient visits or telephone interviews every two months until the end of December 2024. During follow-up, patients underwent routine laboratory assessments, including complete blood counts and evaluations of hepatic and renal function, as well as necessary imaging studies to assess disease status. Prognostic endpoints, including survival status at final follow-up and overall survival duration, were documented in detail. A poor prognosis within one year was defined as the occurrence of any of the following events: Patient death, confirmed tumor recurrence, disease progression, or serious clinical complications such as gastrointestinal hemorrhage, malignant cachexia, or intestinal obstruction.

Statistical analysis

All statistical computations were conducted using SPSS version 25.0, and figures were generated with GraphPad Prism 8.0 software. Descriptive statistics for categorical variables were presented as n (%), and group differences were evaluated using the χ² test. For continuous variables, normality was first assessed via the Shapiro-Wilk test. Variables following a normal distribution were expressed as mean ± SD. Comparisons between two groups were performed using independent sample t-tests, while ANOVA was applied when comparing more than two groups.

The diagnostic performance of SF3B4 and HMGB1 in predicting unfavorable one-year outcomes was assessed using receiver operating characteristic (ROC) curve analysis. The area under the curve (AUC), along with sensitivity and specificity values, was calculated for both markers individually and in combination. To determine independent factors associated with one-year mortality in gastric cancer, multivariate logistic regression analysis was employed. Furthermore, Kaplan-Meier survival analysis was conducted to explore differences in overall survival among subgroups defined by SF3B4 and HMGB1 expression levels. Statistical significance was established at a threshold of ^aP < 0.05.

RESULTS

Expression levels of SF3B4 and HMGB1 in gastric and control tissues

Quantitative analysis revealed that both SF3B4 and HMGB1 mRNA expression levels were significantly elevated in tumor tissues derived from gastric cancer patients when compared with their adjacent non-tumorous tissues and with lesion tissues from patients in the non-gastric cancer group (^aP < 0.05). However, no significant differences were observed between paracancerous tissues from gastric cancer cases and lesion tissues from the control group (P > 0.05; Figure 1).

Figure 1 Comparison of splicing factor 3b subunit 4 and high mobility group box 1 expression in tissues. A: Splicing factor 3b subunit 4; B: High mobility group box 1. GCCT: Cancerous tissue in the gastric cancer group; GCPT: Paracancerous tissue in the gastric cancer group; LCLT: Lesion tissue in the lung cancer group; SF3B4: Splicing factor 3b subunit 4; HMGB1: High mobility group box 1. ^aP < 0.05 vs cancerous tissue in the gastric cancer group.

Association between SF3B4 and HMGB1 Expression and clinical variables

No significant variation in SF3B4 or HMGB1 expression was noted with respect to patient sex, age, Helicobacter pylori infection status, tumor site, serum carcinoembryonic antigen levels, or histopathological subtype (P > 0.05). In contrast, patients with more aggressive tumor features, namely poor differentiation, tumor size exceeding 5 cm, deep invasion (T3-T4), presence of lymph node metastasis, and advanced clinical stage (clinical TNM III-IV), showed markedly higher expression of both SF3B4 and HMGB1 (^aP < 0.05; Tables 3 and 4).

Table 3 Comparison of splicing factor 3b subunit 4 expression in tissues with different clinical data, n (%).

Item	Cases, n = 114	SF3B4 expression	t	P value
Gender			0.639	0.523
Male	71 (62.28)	1.53 ± 0.32
Female	43 (37.72)	1.49 ± 0.33
Age, years			0.293	0.769
≥ 60	47 (41.23)	1.53 ± 0.34
< 60	67 (58.77)	1.51 ± 0.37
Helicobacter pylori infection			0.000	1.000
Yes	101 (88.60)	1.50 ± 0.36
No	13 (11.40)	1.50 ± 0.33
Tumor location			0.268	0.789
Fundus	38 (33.33)	1.51 ± 0.34
Body	43 (37.72)	1.53 ± 0.33
Antrum	33 (28.95)	1.52 ± 0.35
CEA			0.444	0.657
Elevated	63 (55.26)	1.54 ± 0.38
Normal	51 (44.74)	1.51 ± 0.33
Pathological type			0.186	0.852
Adenocarcinoma	102 (89.47)	1.51 ± 0.35
Others	12 (10.53)	1.53 ± 0.37
Tumor differentiation			4.387	< 0.001
Moderate-high	68 (59.65)	1.35 ± 0.22
Poor	46 (40.35)	1.58 ± 0.34
Maximum tumor diameter, cm			7.620	< 0.001
< 5	71 (62.28)	1.26 ± 0.16
≥ 5	43 (37.72)	1.65 ± 0.38
Tumor invasion depth			6.980	< 0.001
T1-T2	35 (30.70)	1.23 ± 0.17
T3-T4	79 (69.30)	1.62 ± 0.31
Lymph node metastasis			10.197	< 0.001
Yes	84 (73.68)	1.04 ± 0.23
No	30 (26.32)	1.68 ± 0.43
cTNM stage			5.295	< 0.001
Stage I-II	34 (29.82)	1.25 ± 0.26
Stage III-IV	80 (70.18)	1.62 ± 0.37

SF3B4: Splicing factor 3b subunit 4; CEA: Carcinoembryonic antigen; cTNM: Clinical tumor-node-metastasis.

Table 4 Comparison of high mobility group box 1 expression in tissues with different clinical data, n (%).

Item	Cases, n = 114	HMGB1 expression	t	P value
Gender			0.728	0.467
Male	71 (62.28)	0.86 ± 0.13
Female	43 (37.72)	0.88 ± 0.16
Age, years			1.557	0.122
≥ 60	47 (41.23)	0.85 ± 0.11
< 60	67 (58.77)	0.89 ± 0.15
Helicobacter pylori infection			0.503	0.615
Yes	101 (88.60)	0.85 ± 0.13
No	13 (11.40)	0.87 ± 0.17
Tumor location			0.955	0.342
Fundus	38 (33.33)	0.86 ± 0.13
Body	43 (37.72)	0.89 ± 0.15
Antrum	33 (28.95)	0.87 ± 0.15
CEA			1.080	0.282
Elevated	63 (55.26)	0.88 ± 0.16
Normal	51 (44.74)	0.85 ± 0.13
Pathological type			0.621	0.535
Adenocarcinoma	102 (89.47)	0.89 ± 0.16
Others	12 (10.53)	0.86 ± 0.14
Tumor differentiation			10.158	< 0.001
Moderate-high	68 (59.65)	0.73 ± 0.08
Poor	46 (40.35)	0.95 ± 0.15
Maximum tumor diameter (cm)			10.028	< 0.001
< 5	71 (62.28)	0.69 ± 0.12
≥ 5	43 (37.72)	0.93 ± 0.13
Tumor invasion depth			9.681	< 0.001
T1-T2	35 (30.70)	0.68 ± 0.09
T3-T4	79 (69.30)	0.96 ± 0.16
Lymph node metastasis			17.636	< 0.001
Yes	84 (73.68)	0.63 ± 0.07
No	30 (26.32)	0.98 ± 0.14
cTNM stage			15.391	< 0.001
Stage I-II	34 (29.82)	0.60 ± 0.06
Stage III-IV	80 (70.18)	1.01 ± 0.15

HMGB1: High mobility group box 1; CEA: Carcinoembryonic antigen; cTNM: Clinical tumor-node-metastasis.

Predictive accuracy of SF3B4 and HMGB1 for one-year adverse outcomes

ROC curve analysis demonstrated that the AUC for predicting poor outcomes at one year was 0.776 for SF3B4 and 0.757 for HMGB1. When both biomarkers were assessed in combination, the AUC increased to 0.914, indicating enhanced diagnostic accuracy. The combined detection strategy also yielded superior sensitivity and specificity compared to individual marker analysis. Full details are presented in Table 5 and visualized in Figure 2.

Figure 2 Receiver operating characteristic curve of splicing factor 3b subunit 4 and high mobility group box 1 expression in tissues for predicting poor prognosis of gastric cancer at 1 year. SF3B4: Splicing factor 3b subunit 4; HMGB1: High mobility group box 1.

Table 5 Analysis of the predictive value of splicing factor 3b subunit 4 and high mobility group box 1 expression in tissues for poor prognosis of gastric cancer at 1 year.

Index	Optimal cutoff value	AUC	95%CI	P value	Sensitivity, %	Specificity, %
SF3B4	1.65	0.776	0.692-0.835	< 0.001	78.35	79.18
HMGB1	0.92	0.757	0.661-0.813	< 0.001	80.26	78.74
Combined	-	0.914	0.819-0.947	< 0.001	90.17	91.43

SF3B4: Splicing factor 3b subunit 4; HMGB1: High mobility group box 1; AUC: Area under the curve; CI: Confidence interval.

Independent predictors of one-year mortality in gastric cancer

To identify factors associated with one-year mortality in patients with gastric cancer, a multivariate logistic regression model was constructed. The dependent variable was defined as survival outcome (0 = survived, 1 = deceased). Variables significantly associated with prognosis from prior univariate analyses (Figure 1, Tables 3 and 4) were included as independent predictors, with coding details summarized in Table 6. The analysis identified high SF3B4 expression (≥ 1.45), elevated HMGB1 levels (≥ 0.93), poor histological differentiation, tumor size ≥ 5 cm, advanced invasion depth (T3-T4), lymph node involvement, and clinical TNM stage III-IV as independent risk factors for one-year mortality (^aP < 0.05; Table 7).

Table 6 Variable assignment.

Independent variable	Assignment method
	0, survived	1, deceased
SF3B4	< 1.45	≥ 1.45
HMGB1	< 0.93	≥ 0.93
Tumor differentiation	Moderate to high differentiation	Poor differentiation
Maximum tumor diameter, cm	< 5	≥ 5
Tumor invasion depth	T1-T2	T3-T4
Lymph node metastasis	No	Yes
cTNM stage	I-II	III-IV

cTNM: Clinical tumor-node-metastasis.

Table 7 Multivariate logistic regression analysis of risk factors for 1-year mortality in gastric cancer patients with poor prognosis.

Factor	β	SE	Wald	P value	Odds ratio	95% confidence interval
SF3B4 ≥ 1.45	1.443	0.155	6.456	< 0.001	4.221	1.037-4.625
HMGB1 ≥ 0.93	1.352	0.103	5.794	< 0.001	3.875	1.143-6.192
Poor tumor differentiation	0.819	0.091	6.012	< 0.001	2.264	1.269-5.732
Maximum tumor diameter, ≥ 5 cm	0.914	0.075	5.568	< 0.001	2.493	1.335-7.519
Tumor invasion depth, T3-T4	0.742	0.061	5.937	< 0.001	2.089	1.026-5.368
Lymph node metastasis	1.025	0.094	6.027	< 0.001	2.783	1.165-4.594
cTNM stage III-IV	1.109	0.086	5.685	< 0.001	3.017	1.418-5.798

SF3B4: Splicing factor 3b subunit 4; HMGB1: High mobility group box 1; cTNM: Clinical tumor-node-metastasis.

Correlation between SF3B4/HMGB1 expression and overall survival

Of the 114 patients in the gastric cancer cohort, 43 were still alive, while 71 had died by the conclusion of the follow-up period. Kaplan-Meier survival analysis revealed that patients with high expression of both SF3B4 (> 1.45) and HMGB1 (> 0.93) had a significantly shorter median survival time (25.74 ± 5.46 months) compared to those with lower expression levels (33.29 ± 6.71 months, log-rank = 10.534, ^aP < 0.05; Figure 3).

Figure 3 Relationship between splicing factor 3b subunit 4 and high mobility group box 1 expression and survival time. SF3B4: Splicing factor 3b subunit 4; HMGB1: High mobility group box 1.

DISCUSSION

Gastric cancer ranks among the leading malignant tumors globally, characterized by notably high rates of incidence and mortality[15]. Despite advances in diagnostic techniques such as expanded use of endoscopic screening, and therapeutic progress including targeted agents and immunotherapies, the prognosis for many patients remains poor. This is especially evident in those diagnosed at advanced stages, where the five-year survival rates remain disappointingly low[16,17]. Given this challenge, the discovery and validation of reliable molecular markers for early detection, accurate disease staging, and prognosis prediction have become critical areas of research in gastric oncology. In recent years, emerging evidence has highlighted the crucial involvement of RNA splicing regulators and inflammation-associated proteins in cancer development and progression[18,19]. Among them, SF3B4, a vital component of the spliceosome complex, has attracted attention due to its aberrant expression patterns in several cancer types. SF3B4 modulates alternative splicing events, thereby influencing tumor cell behaviors and promoting malignancy through altered gene expression[20]. Meanwhile, HMGB1, a nuclear non-histone chromatin-binding protein, not only maintains DNA architecture but also acts as a pro-inflammatory mediator once released extracellularly, triggering multiple signaling pathways that facilitate tumor growth and metastasis[21]. However, the exact clinical implications of SF3B4 and HMGB1 expression in gastric cancer remain poorly defined, and the possibility of their combined effect has not been thoroughly explored. To address this gap, our study examined the levels of SF3B4 and HMGB1 in gastric cancer tissue samples and evaluated their associations with patients’ clinicopathological features and survival outcomes.

The role of SF3B4 in gastric cancer initiation and progression

SF3B4 serves as a vital element within the RNA splicing machinery, playing an essential role in post-transcriptional gene regulation. Recent research has increasingly linked aberrant SF3B4 expression to the development of diverse malignancies. For example, in hepatocellular carcinoma, SF3B4 modulates epithelial splicing regulatory protein 1, thereby affecting tumor cell proliferation and motility[22]. Similarly, in breast cancer, SF3B4 enhances tumor cell survival through activation of the protein kinase B signaling pathway[23]. Our current investigation revealed a markedly elevated expression of SF3B4 in gastric cancer tissues compared with adjacent normal gastric mucosa and tissues from benign gastric conditions. Furthermore, increased SF3B4 levels were strongly correlated with more aggressive pathological features, including poor tumor differentiation, tumor size exceeding 5 cm, deeper invasion at T3-T4 stages, presence of lymph node metastases, and advanced clinical TNM stages III-IV. These observations support the hypothesis that SF3B4 facilitates tumor invasion and metastatic spread. Although the detailed molecular mechanisms remain to be fully elucidated, prior studies suggest that SF3B4 may influence the alternative splicing of key oncogenes and tumor suppressor genes, thus modulating malignant phenotypes. Specifically, SF3B4 has been implicated in regulating variant splicing of CD44 (CD44v), which may enhance gastric cancer cell adaptability within their microenvironment and promote invasive capabilities[24]. Additionally, SF3B4 may affect the splicing of apoptosis regulators such as protein 53, B-cell lymphoma 2, and B-cell lymphoma 2-associated X protein, leading to reduced programmed cell death and increased tumor cell survival[25]. Together, these pathways underline SF3B4 as a pivotal factor driving gastric cancer progression.

HMGB1’s contribution to gastric cancer invasion and poor outcomes

HMGB1 is known to interact with receptors including receptor for advanced glycation end products and Toll-like receptor 4, which activate downstream signaling cascades such as nuclear factor-κB, signal transducer and activator of transcription 3, and phosphatidylinositol 3-kinase/protein kinase B pathways. These activations promote cancer cell proliferation, contribute to chemoresistance, support angiogenesis, and facilitate immune evasion[26,27]. In our study, HMGB1 expression was significantly upregulated in gastric cancer tissues compared to adjacent non-tumorous and benign gastric tissues. This elevated expression was also linked with adverse pathological features: Poor differentiation, deeper tumor invasion (T3-T4), lymph node involvement, and advanced TNM stages. These findings echo previous research suggesting HMGB1’s role in remodeling the inflammatory and tumor microenvironment to favor cancer progression[28,29]. Furthermore, HMGB1 has been reported to induce epithelial-mesenchymal transition by downregulating E-cadherin and upregulating vimentin, thereby enhancing the motility and invasiveness of gastric cancer cells[30,31]. Additionally, HMGB1 shapes the immune landscape of tumors by recruiting and activating immune cell subsets, including macrophages, dendritic cells, and T lymphocytes. However, our current work did not assess immune cell infiltration via immunohistochemistry; future studies incorporating immunoprofiling for markers such as CD8+ T cells and CD68+ macrophages would be valuable for understanding HMGB1’s immunomodulatory impact and its potential as a predictor for immunotherapy response.

Prognostic implications of SF3B4 and HMGB1 in gastric cancer

Through ROC curve analysis, both SF3B4 and HMGB1 were found to independently predict poor prognosis at one year in gastric cancer patients, with respective AUC values of 0.776 and 0.757. Remarkably, when combined, these biomarkers achieved a higher AUC of 0.914, indicating superior predictive accuracy compared to either alone. Kaplan-Meier survival curves further demonstrated that patients exhibiting high levels of both SF3B4 and HMGB1 had significantly shorter median survival (25.74 ± 5.46 months) than those with lower expression (33.29 ± 6.71 months; ^aP < 0.05). These results imply that SF3B4 and HMGB1 may synergistically promote gastric tumor aggressiveness and could serve as early indicators of high-risk patients. While the exact interplay between these two molecules is yet to be fully delineated, we hypothesize that SF3B4 could modulate alternative splicing of genes regulating HMGB1, thereby amplifying HMGB1-driven inflammatory and oncogenic signaling. Both factors are also known to engage the nuclear factor-κB pathway, which enhances tumor proliferation, resistance to therapy, and metastasis[32,33]. This underscores the potential of targeting SF3B4 and HMGB1 together for precision therapeutic interventions in gastric cancer. To clarify their mechanistic relationship, further studies employing dual gene knockdown, RNA sequencing, co-immunoprecipitation, and proteomics are necessary to establish whether SF3B4 and HMGB1 cooperate directly or act in parallel pathways during gastric carcinogenesis.

CONCLUSION

Several limitations should be acknowledged in this study. First, the retrospective nature and single-center design, coupled with a relatively modest sample size, may restrict the representativeness and extrapolation of the findings. Future investigations should adopt a prospective, multicenter approach with larger cohorts to enhance external validity. Second, this research primarily concentrated on evaluating the clinical relevance and prognostic utility of SF3B4 and HMGB1, without delving into the underlying molecular pathways. Subsequent studies should incorporate in vitro cell models and in vivo animal experiments to dissect the functional roles and mechanistic contributions of these molecules in gastric cancer development. Third, integrating multi-omics platforms, such as transcriptomic and proteomic profiling, may help clarify the complex signaling cascades and regulatory networks through which SF3B4 and HMGB1 influence tumor biology.

Despite these limitations, the current findings highlight that both SF3B4 and HMGB1 are markedly overexpressed in gastric cancer tissues and are closely linked to more aggressive pathological features and unfavorable clinical outcomes. Elevated expression levels of either biomarker, particularly when evaluated in combination, offer significant predictive value for patient prognosis. Taken together, these results support the potential application of SF3B4 and HMGB1 as dual biomarkers for risk stratification and as promising molecular targets for individualized therapeutic interventions in gastric cancer management.