Bioinformatics analysis and preliminary clinical validation of ITGA9 and NPTX2 as prognosis-related genes in acute myeloid leukemia Authors: Lili Chen, Na Yang, Fei Zhang, Fang Lv, Ningping Tan, Yanan Zhang, Guangqi Wu Published: 2026 ABSTRACT AbstractObjectiveTo investigate the expression and clinical significance of integrin subunit alpha 9 (ITGA9) and neuronal pentraxin 2 (NPTX2) in acute myeloid leukemia (AML), and to evaluate their potential diagnostic and prognostic value.MethodsDifferentially expressed genes (DEGs) associated with AML prognosis were identified from the Gene Expression Omnibus (GEO) dataset (GSE12417), followed by Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analyses. The expression and prognostic relevance of candidate genes were further validated using The Cancer Genome Atlas (TCGA) dataset and the Xiantao Academic online platform. Correlations of ITGA9 and NPTX2 with immune-regulatory genes, immune checkpoint molecules, and immune cell infiltration were analyzed to explore their potential immune associations. Peripheral blood samples from 60 newly diagnosed AML patients and 50 healthy controls were collected for quantitative real-time PCR validation. Diagnostic and prognostic performances were assessed using receiver operating characteristic (ROC) curve and Kaplan–Meier survival analyses.ResultsSeven prognosis-related genes were identified, among which ITGA9 and NPTX2 were significantly upregulated in AML. ITGA9 and NPTX2 were predominantly involved in modulating signaling receptor functions and were primarily associated with the phosphatidylinositol 3-kinase–AKT (PI3K-Akt) signaling. Immune correlation analysis revealed that ITGA9 and NPTX2 were associated with multiple immune-regulatory and checkpoint molecules, indicating potential involvement in AML immune microenvironment modulation. Clinical validation confirmed that ITGA9 and NPTX2 were markedly overexpressed in the AML patients. ROC curve analysis demonstrated strong diagnostic efficacy for ITGA9 and NPTX2. Among the 60 AML patients followed up, 29 died, and the deceased group exhibited significantly higher ITGA9 and NPTX2 expression. High expression of ITGA9 and NPTX2 was associated with reduced relapse-free survival.ConclusionITGA9 and NPTX2 are significantly overexpressed in AML and are closely associated with poor prognosis. Their strong diagnostic performance suggests that these two genes may serve as potential biomarkers and therapeutic targets for AML.Supplementary InformationThe online version contains supplementary material available at 10.1007/s12672-026-05307-3.Keywords: AML, Integrin subunit alpha 9, Neuronal pentraxin 2, Biomarker, Prognosis Introduction Cancer remains a leading cause of morbidity and mortality worldwide, and its treatment has evolved from conventional chemotherapy to targeted therapies and immunotherapies, ushering in an era of precision medicine [1]. AML, originated from the clonal abnormal proliferation of hematopoietic stem or progenitor cells, is a malignant hematologic neoplasm. It is characterized by impaired myeloid differentiation and excessive accumulation of immature blasts in the bone marrow and peripheral blood, resulting in severe hematopoietic dysfunction [2]. Despite considerable advances in AML management in recent decades, including targeted therapies, immunotherapeutic approaches, and hematopoietic stem cell transplantation, relapse remains common, and many patients continue to face unfavorable clinical outcomes [3]. Given the heterogeneity of AML, identifying novel molecular biomarkers is essential to guide prognostic prediction and personalized therapeutic strategies within the modern precision-medicine framework. AML is highly heterogeneous, and different molecular subtypes exhibit distinct patterns of disease progression, drug responsiveness, and survival outcomes [4, 5]. Therefore, identifying novel molecular biomarkers for prognostic prediction and personalized therapeutic guidance is of great clinical importance. With the advancement of high-throughput sequencing technologies and bioinformatics, analyzing gene expression patterns has offered fresh perspectives on the molecular processes driving AML [6]. Increasing evidence suggests that dysregulated gene expression and aberrant pathway activity play crucial roles in AML initiation, drug resistance, and relapse [7]. Systematic transcriptomic analyses provide a valuable approach for discovering prognostic biomarkers and understanding disease mechanisms, particularly in AML, where patient stratification remains challenging. Comprehensive examination of publicly available resources, including the GEO and The Cancer Genome Atlas (TCGA), enables efficient detection of key genes associated with disease development and prognosis, thereby offering a robust foundation for subsequent functional validation [8, 9]. However, systematic screening and clinical validation of prognostic biomarkers specifically associated with poor outcomes in AML remain insufficient. ITGA9, which regulates cell adhesion and migration, can modulate immune cell behavior and signaling in the tumor microenvironment, influencing antitumor immunity and immune cell infiltration [10]. NPTX2, a pentraxin family member, has been implicated in immune regulation and inflammatory processes in various disease contexts, suggesting a potential link to immune-related pathways [11, 12]. Both ITGA9 and NPTX2 have been reported to be associated with cancer and other diseases [13–16]. Recent transcriptomic analyses in AML and pan-cancer cohorts have highlighted integrin and pentraxin family members, such as ITGA4, ITGAM, and NPTX1, as potential prognostic biomarkers in myeloid malignancies [17, 18]. Nevertheless, the expression patterns, clinical significance, and prognostic value of ITGA9 and NPTX2 in AML remain largely largely unclear, which motivated the present study to systematically investigate these genes in AML patients. In this study, DEGs associated with AML prognosis were identified based on the GEO database, and their potential biological functions were explored through functional enrichment analyses. The expression levels of key candidate genes, ITGA9 and NPTX2, were further validated in clinical samples, and their diagnostic and prognostic value in AML was assessed. Overall, this study aims to situate AML biomarker discovery within the broader landscape of precision oncology and provide novel theoretical evidence for molecular subtyping and individualized therapy. Materials and methods Data acquisition AML expression profiles were retrieved from the GEO database (GSE12417). Raw data were preprocessed and normalized using R software (version 4.2.2) and Bioconductor packages. Identification of DEGs Based on the GPL570 platform, data analyses were conducted using the GEO dataset GSE12417. This dataset included bone marrow mononuclear cell samples from 79 AML patients, comprising 32 patients with favorable outcomes and 47 with poor outcomes. Outcome classification was strictly based on the survival status provided in the original GEO clinical annotations. Specifically, survival status at the last recorded follow-up was used, with patients annotated as “dead” defined as having a poor outcome and those annotated as “alive” defined as having a favorable outcome. No additional time-based cutoff or follow-up duration information was available in the dataset, and no further reclassification was performed. DEGs between the two groups were identified using the “limma” R package, with thresholds of |log₂FoldChange| > 1 and P < 0.05. To control for multiple testing, P-values were adjusted using the false discovery rate (FDR) method, and genes with FDR < 0.05 were considered significant. The identified DEGs were used for subsequent functional enrichment and prognostic analyses. A heatmap was generated to visualize the expression patterns of DEGs. Gene ontology (GO) and Kyoto encyclopedia of genes and genomes (KEGG) Based on the R package clusterProfiler (version 4.4.4), GO and KEGG analyses were carried out. GO analysis was conducted across three domains: biological process (BP), molecular function (MF), and cellular component (CC), with the significance threshold set at P < 0.05. KEGG analysis was performed to identify the major signaling pathways and biological processes involving the DEGs. Gene expression and prognostic analysis Expression data and clinical follow-up information of AML patients were obtained from the GEO and TCGA databases. Candidate gene expression and prognostic relevance in AML were analyzed based on these datasets. Prognostic analyses were performed utilizing the R libraries survival (v3.3-1) and survminer (v0.4.9). Kaplan–Meier survival curves were generated, and hazard ratios (HRs) and log-rank P values were calculated to evaluate the relationship between gene expression (high vs. low) and overall survival or relapse-free survival (RFS). A pan-cancer analysis was performed using the Xiantao Academic online platform, and visualization was completed using R package ggplot2 (version 3.4.4) to assess the expression profiles of the candidate genes across multiple tumor types. Immune correlation analysis Spearman’s correlation analysis was applied to evaluate the relationships between the candidate genes and immune-related genes, immune checkpoint molecules, and immune cell infiltration levels. Immune gene information was obtained from the TCGA database, and immune cell infiltration was estimated using the ESTIMATE algorithm. To reduce false-positive findings, correlation analyses were performed with false-discovery-rate (FDR) correction, and adjusted P values (FDR < 0.05) were considered statistically significant. Clinical sample collection A total of 60 patients newly diagnosed with AML between March 2022 and March 2024 and 50 age-matched healthy individuals undergoing routine health examinations were enrolled in this study. The number of samples was pre-calculated with G*Power software, applying a one-tailed test with an effect size of 0.5, significance level α = 0.05, and statistical power of 0.95. Inclusion criteria: (1) confirmed diagnosis of AML according to established diagnostic criteria; (2) complete baseline clinical data available; (3) availability of cryopreserved pre-treatment bone marrow or peripheral blood samples; and (4) waiver of informed consent. Exclusion criteria: (1) secondary AML, defined as AML evolving from a prior hematologic disorder (such as myelodysplastic syndrome or myeloproliferative neoplasm) or therapy-related AML following prior chemotherapy or radiotherapy; (2) concomitant hematologic malignancies; (3) pregnancy or lactation; (4) loss to follow-up or unclear survival status; and (5) severe non-hematologic comorbidities that may independently affect survival outcomes, such as heart failure, end-stage hepatic or renal failure, or active systemic infection. The research protocol received approval from the institutional ethics committee of The Affiliated Suqian Hospital of Xuzhou Medical University (No. 2025-042-01), and with waiver of informed consent due to the retrospective observational nature of the study. The study was carried out following the ethical guidelines outlined in the Declaration of Helsinki. All deaths in the validation cohort were confirmed to be AML-related. Detection of ITGA9 and NPTX2 expression Blood samples were collected from all participants using heparin as an anticoagulant. Peripheral blood mononuclear cells (PBMCs) were then separated via Ficoll density gradient centrifugation, briefly by diluting whole blood 1:1 with PBS, layering it over Ficoll, and centrifuging at 400 ×g for 30 min at room temperature without brake. The mononuclear cell layer at the interface was collected and washed twice with PBS. Total RNA was isolated with TRIzol reagent, and its concentration and purity were measured using a NanoDrop 2000 spectrophotometer. Complementary DNA (cDNA) was synthesized according to the instructions provided with the reverse transcription kit. Quantitative polymerase chain reaction (qPCR) was carried out using a qPCR detection kit on a LightCycler 480 real-time PCR system with the following cycling parameters: initial denaturation at 95 °C for 30 s, followed by 40 cycles of 95 °C for 3 s and 60 °C for 20 s. β-actin (ACTB) served as the internal control. The expression of ITGA9 and NPTX2 were calculated using the 2−ΔΔCt method (Table 1). Table 1Primer sequencesGenePrimer sequence (5’−3’)ITGA9F: TCAGGTCACAGAGAAGCTGCAGR: CACATGCTCACTGAGGCTGTAGNPTX2F: AGGCATTCCAGGACGGAGAGAAR: CTGAGTGGCATCAAACCTACCCACTBF: CACCATTGGCAATGAGCGGTTCR: AGGTCTTTGCGGATGTCCACGT Primer sequences Prognostic follow-up All AML patients were followed for at least one year to assess clinical outcomes. Follow-up was conducted through regular clinical visits, and survival status was recorded for each patient. ROC curve analysis and diagnostic evaluation The ROC analyses were conducted using an independent clinical cohort. Based on ITGA9 and NPTX2 expression in clinical samples, ROC curves were generated using the R package pROC (version 1.18.0) to evaluate their diagnostic and prognostic performance in AML. The AUC and 95% confidence interval (CI) were calculated to assess model accuracy. The optimal cut-off values were determined using the Youden index, and patients were stratified into high- and low-expression groups for subsequent survival analyses. Statistics All data analyses and visualizations were conducted using the Xiantao Academic platform (https://www.xiantao.love/) and validated with SPSS 27.0. Categorical data were presented as counts and percentages [n (%)], with differences between groups assessed using the chi-square test. The Shapiro–Wilk test was employed to examine the normality of continuous variables. Variables with a normal distribution were reported as mean ± standard deviation (x̄ ± s) and compared between groups using the independent-samples t test, while non-normally distributed variables were expressed as median (interquartile range) and analyzed using the Mann–Whitney U test. ROC curves were constructed to assess the diagnostic performance of genes in AML. Kaplan–Meier survival curves were plotted. Survival differences between groups were analyzed using the log-rank test, with a P value of less than 0.05 indicating statistical significance. Results Prognosis-related genes in AML identification Based on the GSE12417 dataset, DEGs associated with AML prognosis were identified. Using the criteria of |logFC| > 1 and P < 0.05, seven potential prognostic genes were obtained: ITGA9, ABCC13, NPTX2, F13A1, EREG, HBG2///HBG1, and ALAS2. Among these, ITGA9, NPTX2, F13A1, and EREG were significantly upregulated, while ABCC13, HBG2///HBG1, and ALAS2 were markedly downregulated (Fig. 1A–B). These genes were primarily associated with biological processes including wound healing, regulation of signaling receptor activity, and acyltransferase activity. DEGs were enriched in pathways such as PI3K–Akt signaling and AML-relevant pathways, including complement and coagulation cascades, providing preliminary mechanistic insights while acknowledging that these analyses remain exploratory (Fig. 1C–D). Screening of prognosis-related genes and functional enrichment analysis in AML. (A) Volcano plot of DEGs in the GSE12417 dataset. (B) Heatmap showing the expression profiles of the seven prognosis-related genes. (C) GO enrichment analysis. (D) KEGG pathway enrichment analysis Expression patterns and prognostic significance of ITGA9 and NPTX2 in AML To further identify potential prognostic biomarkers for AML, we analyzed the expression and prognostic relevance of the seven candidate genes in AML patients. ITGA9 and NPTX2 exhibited relatively prominent expression differences (Fig. 2A). ITGA9 overexpression can boost the maintenance of leukemia stem cells and may serve as a potential therapeutic target in AML [19]. Therefore, we focused subsequent analyses on ITGA9 and NPTX2. Expression profiles and prognostic significance of ITGA9 and NPTX2 in AML. (A) Expression statistics of seven candidate genes in AML. (B) Pan-cancer expression analysis of ITGA9 and NPTX2. (C) Survival analysis of ITGA9 and NPTX2 in AML. Cancer type abbreviations are shown in the figure. Full names of all cancer types are provided in Supplementary Table 1. *P < 0.05, **P < 0.01, ***P < 0.001 Pan-cancer analysis indicated elevated expression of ITGA9 and NPTX2 in AML (Fig. 2B). Patients with high ITGA9 expression were accompanied with a slightly increased survival risk (HR = 1.41, P = 0.326), though the difference was not statistically significant. In contrast, patients with high NPTX2 expression exhibited a lower overall survival rate than those with low expression (HR = 1.48, P = 0.07), suggesting a trend toward worse prognosis, albeit without reaching statistical significance (Fig. 2C). Immune correlation analysis of ITGA9 and NPTX2 in AML To explore the potential immune-related mechanisms of ITGA9 and NPTX2 in AML, we assessed their correlations with immune-regulatory genes, immune checkpoint molecules, and immune cell infiltration. The expression of ITGA9 showed positive correlations with immune-regulatory genes such as CD47, STAT3, and IFNG, as well as with CD274 (PD-L1), LAG3, and TIGIT (P < 0.05). Immune infiltration analysis indicated a mild positive correlation between ITGA9 expression and overall immune infiltration level (R = 0.057, P = 0.485). In contrast, NPTX2 expression was correlated with immune-regulatory genes CXCL12, CD47, and TGFB1, as well as with immune checkpoint molecules IL10RA, TNFRSF18, and ICOSLG (P < 0.05). However, its correlation with immune infiltration was weaker (R = 0.045, P = 0.582), implying that NPTX2 may be linked to immune-related signaling at the expression level but does not strongly associate with immune cell infiltration (Fig. 3A-C). Immune correlation analysis of ITGA9 and NPTX2 in AML. (A) Correlation of ITGA9 and NPTX2 with immune regulatory genes in AML. (B) Correlation of ITGA9 and NPTX2 with immune checkpoint molecules in AML. (C) Correlation of ITGA9 and NPTX2 expression with immune cell infiltration in AML. *P < 0.05 Collectively, both ITGA9 and NPTX2 showed associations with multiple immune-related molecules, implying that they may participate in AML immune microenvironment regulation through distinct mechanisms. Validation of ITGA9 and NPTX2 expression and diagnostic value in AML To further evaluate the clinical significance of ITGA9 and NPTX2, their expression levels and diagnostic performance were validated in AML patient samples. Both ITGA9 and NPTX2 were markedly upregulated in AML tissues compared with healthy controls (P < 0.05, Fig. 4A). ROC curve displayed excellent diagnostic efficiency, with an AUC of 0.856 (95% CI: 0.783–0.928) for ITGA9 and 0.890 (95% CI: 0.827–0.952) for NPTX2, indicating strong diagnostic potential for AML (Fig. 4B). Validation of ITGA9 and NPTX2 expression and their diagnostic value in AML. (A) qPCR-based expression analysis of ITGA9 and NPTX2 in bone marrow samples from AML patients and healthy controls. (B) ROC curve analysis of ITGA9 and NPTX2 for AML diagnosis. ***P < 0.001 High expression of ITGA9 and NPTX2 and their prognostic value in AML During a one-year follow-up period, 29 AML patients died. Compared with survivors, deceased patients exhibited significantly higher expression levels of ITGA9 and NPTX2 (P < 0.05, Fig. 5A). Correlation analysis using Spearman method showed a weak but statistically significant positive correlation between ITGA9 and NPTX2 expression in AML patients (Spearman r = 0.314, P = 0.015), suggesting a potential co-regulatory relationship (Supplement Fig. 1). ROC curve analysis for prognosis revealed that ITGA9 (AUC = 0.755, 95% CI: 0.631–0.879, cut-off = 4.095) and NPTX2 (AUC = 0.893, 95% CI: 0.806–0.981, cut-off = 4.535) had good predictive accuracy (Fig. 5B). High expression of ITGA9 and NPTX2 and their prognostic value in AML. (A) Expression levels of ITGA9 and NPTX2 in AML patients stratified by survival outcomes. (B) ROC curve analysis of ITGA9 and NPTX2 for predicting survival status at one-year follow-up in AML. (C) Kaplan-Meier analysis of relapse-free survival (RFS) of AML patients grouped by ITGA9 and NPTX2 expression. ***P < 0.001 Patients were subsequently stratified based on the cut-off values into high- and low-expression groups: ITGA9 high expression (> 4.095, n = 18) and low expression (< 4.095, n = 42); NPTX2 high expression (> 4.535, n = 27) and low expression (< 4.535, n = 33). Patients stratified by ITGA9 or NPTX2 expression showed no significant differences in age, sex, WBC, or FAB subtype, t(8;21)/inv(16), FLT3-ITD, while ELN risk was higher in the high-expression groups (P < 0.05), indicating overall comparability between groups (Tables 2 and 3). Multivariate Cox regression analysis revealed that high ITGA9 expression was significantly associated with poorer relapse-free survival and served as an independent adverse prognostic factor (HR = 5.08, 95% CI: 2.22–11.64, P < 0.001), whereas sex, FAB subtype, ELN risk, white blood cell count (WBC), and age showed no significant impact (P > 0.05). Similarly, high NPTX2 expression was significantly associated with reduced relapse-free survival (HR = 8.57, 95% CI: 3.77–19.47, P < 0.001), with ELN risk also demonstrating a modest independent prognostic effect (HR = 2.18, 95% CI: 1.01–4.72, P = 0.048). Other clinical variables, including age, sex, FAB subtype, and WBC, were not statistically significant. Overall, these results indicate that high expression of ITGA9 and NPTX2 may serve as independent predictors of poor prognosis in AML patients (Tables 4 and 5). Table 2Baseline demographic and clinical characteristics of AML patients stratified by ITGA9 expressionVariableITGA9-low (n = 42)ITGA9-high (n = 18)t/χ²P valueAge (years)47.31 ± 2.3147.83 ± 1.92−0.9100.369Sex, n (%)Male25 (59.5%)10 (55.6%)0.0820.775Female17 (40.5%)8 (44.4%)WBC (×10⁹/L)42.26 ± 4.8844.06 ± 5.20−1.2470.222FAB subtype, n (%)M1–224 (57.1%)10 (55.6%)0.0130.909M4–518 (42.9%)8 (44.4%)ELN risk, n (%)Non-adverse30 (71.4%)8 (44.4%)3.9510.047Adverse12 (28.6%)10 (55.6%)t(8;21)/inv(16), n (%)6 (14.3%)2 (11.1%)-1.000FLT3-ITD, n (%)5 (11.9%)3 (16.7%)-0.686 Baseline demographic and clinical characteristics of AML patients stratified by ITGA9 expression Table 3Baseline demographic and clinical characteristics of AML patients stratified by NPTX2 expressionVariableNPTX2-low (n = 33)NPTX2-high (n = 27)t/χ²P valueAge (years)47.64 ± 2.9647.56 ± 3.150.1020.919Sex, n (%)Male19 (57.6%)16 (59.3%)0.0170.895Female14 (42.4%)11 (40.7%)WBC (×10⁹/L)42.84 ± 5.1744.56 ± 4.93−1.3050.197FAB subtype, n (%)M1–219 (57.6%)15 (55.6%)0.0250.875M4–514 (42.4%)12 (44.4%)ELN risk, n (%)Non-adverse25 (75.8%)13 (48.1%)4.8750.027Adverse8 (24.2%)14 (51.9%)t(8;21)/inv(16), n (%)5 (15.2%)3 (11.1%)-0.719FLT3-ITD, n (%)4 (12.1%)5 (18.5%)-0.718 Baseline demographic and clinical characteristics of AML patients stratified by NPTX2 expression Table 4Multivariate Cox proportional hazards regression analysis of relapse-free survival for ITGA9 expressionVariablesHR95% CI P ITGA9 (high vs. low)5.082.22 ~ 11.64< 0.001Age1.060.88 ~ 1.290.532Sex1.930.87 ~ 4.310.108WBC0.920.85 ~ 1.000.062FAB subtype1.540.70 ~ 3.350.281ELN risk0.570.25 ~ 1.280.173 Multivariate Cox proportional hazards regression analysis of relapse-free survival for ITGA9 expression Table 5Multivariate Cox proportional hazards regression analysis of relapse-free survival for NPTX2 expressionVariablesHR95% CI P NPTX2 (high vs. low)8.573.77 ~ 19.47< 0.001Age1.010.94 ~ 1.090.795Sex0.780.35 ~ 1.740.545WBC1.020.97 ~ 1.060.494FAB subtype0.910.41 ~ 2.010.813ELN risk2.181.01 ~ 4.720.048 Multivariate Cox proportional hazards regression analysis of relapse-free survival for NPTX2 expression Patients with high ITGA9 or NPTX2 expression had significantly lower relapse-free survival (RFS) rates (P < 0.001), suggesting that elevated levels of both genes are strongly correlated with unfavorable outcomes in AML (Fig. 5C). Discussion AML is a highly heterogeneous hematologic malignancy characterized by complex genetic and molecular abnormalities. Although recent advances in targeted therapies and immunotherapy have yielded clinical benefits for subsets of patients, the overall survival rate remains unsatisfactory. Particularly in adverse-risk subtypes, relapse and drug resistance continue to pose major therapeutic challenges [20, 21]. Therefore, identifying novel molecular biomarkers to improve AML diagnosis, risk stratification, and prognostic prediction remains an important focus of current research. In this study, analysis of the GEO dataset GSE12417 identified seven prognosis-related genes in AML, among which ITGA9 and NPTX2 were highlighted as potential important candidates. Their expression patterns and prognostic value were further validated using clinical bone marrow samples. Differential expression analysis revealed significant upregulation of ITGA9 and NPTX2 in AML, both of which were associated with poor prognosis. GO and KEGG enrichment analyses indicated that these genes were mainly involved in biological processes such as regulation of cell signaling, lipid metabolism, and the PI3K-Akt pathway, suggesting that they may be involved in AML progression through regulating cell survival and metabolic homeostasis. ITGA9 is important for cell adhesion, migration, and extracellular matrix (ECM)-mediated signaling [22, 23]. Previous studies have demonstrated that ITGA9 overexpression contributes to tumor invasion, metastasis, and therapy resistance in multiple malignancies [24]. Mechanistically, ITGA9 can activate oncogenic cascades and PI3K-Akt pathways, thereby maintaining cancer stemness and promoting malignant progression [25, 26]. Notably, a recent report identified ITGA9 as a key factor supporting leukemic stem cell self-renewal and survival, suggesting its potential as a therapeutic target in AML [19]. In our study, ITGA9 overexpression was observed in AMLand correlated with reduced relapse-free survival, indicating a possible association with disease progression rather than definitive causation. NPTX2, a secreted member of the pentraxin family, is primarily expressed in neuronal tissues but has recently been implicated in several malignant tumors [27]. NPTX2 can influence tumor progression by modulating intercellular signaling and immune-related pathways within the tumor microenvironment [28]. In the present study, NPTX2 was markedly upregulated in AML patients, and its high expression was associated with poor prognosis. Immune correlation analysis further revealed that NPTX2 expression positively correlated with multiple immune regulatory factors, including CXCL12, TGFB1, and IL10RA, suggesting a possible role in AML immune evasion, although causality cannot be confirmed without functional experiments. Importantly, these immune-related findings are derived from bulk transcriptomic analyses and should therefore be interpreted with caution. This aligns with previous reports indicating that NPTX2 is involved in immune suppression and inflammatory regulation in the tumor microenvironment [29]. Given that the observed correlations were relatively weak and based on computational inference, they may not fully reflect actual immune cell composition or function in the AML microenvironment. Despite relatively weak correlations with immune cell infiltration, NPTX2 may indirectly influence AML immune homeostasis through modulation of immune signaling molecules, which warrants further validation using orthogonal approaches such as flow cytometry or immunophenotyping in future studies. The expression validation in clinical samples provided further support for the bioinformatic predictions. Both ITGA9 and NPTX2 were significantly upregulated in bone marrow samples from AML patients compared with healthy controls. However, as the control group consisted of healthy individuals rather than patients with other hematologic disorders that may clinically mimic AML, the reported diagnostic performance may be overestimated. Therefore, the ROC results should be interpreted as preliminary biomarker exploration rather than definitive clinical diagnostic evidence. ITGA9 (AUC = 0.856) and NPTX2 (AUC = 0.890) show potential diagnostic value, but external validation using independent AML cohorts (e.g., TCGA-LAML or other GEO datasets) or comparison with related hematologic diseases is required to confirm reproducibility and clinical applicability. Moreover, patients with high expression of ITGA9 or NPTX2 had significantly shorter relapse-free survival compared with the low-expression groups, highlighting their potential prognostic relevance rather than proven causative effect. Collectively, these findings indicate that ITGA9 and NPTX2 may serve as diagnostic and prognostic biomarkers and could be explored as therapeutic targets, pending functional validation. Nevertheless, several limitations should be acknowledged in this study. First, the relatively small, single-center sample size may have introduced potential selection bias. Second, the analysis was primarily based on transcriptomic data, lacking protein-level validation and mechanistic investigation of ITGA9 and NPTX2. In particular, bulk transcriptomic approaches are subject to both technical and biological biases, which may affect the accuracy of immune inference and gene–function interpretation [30]. Third, the immune correlation analysis was conducted using bioinformatic inference and thus requires experimental confirmation. Finally, the exploratory nature of both diagnostic and prognostic analyses should be noted, consistent with translational biomarker frameworks, which emphasize that clinical applicability requires validation in larger multi-center and multi-omic studies [31, 32]. Further research is needed to confirm these results in larger, multicenter populations and to perform functional studies, such as gene knockdown or CRISPR interference in AML cell models, to clarify whether ITGA9 and NPTX2 truly influence leukemic cell fitness and the immune microenvironment. Conclusion In summary, this study identified and validated ITGA9 and NPTX2 as significantly upregulated genes in AML that are strongly associated with poor prognosis. Both genes are likely to be key contributors to AML development and the regulation of its immune microenvironment, highlighting their potential as diagnostic and prognostic biomarkers. These findings provide new insights into the molecular mechanisms underlying AML and offer promising directions for individualized therapeutic strategies. Supplementary Information Supplementary Material 1. Supplementary Material 2. Acknowledgements Not applicable. Author contributions All authors including LLC, NY, FZ, FL, NPT, YNZ and GQW contributed to the conceptualization, analysis, and writing of the manuscript. All authors approved the manuscript for publication. Funding None. Data availability Data supporting the findings of this study were obtained from publicly accessible databases. Gene expression data were retrieved from the Gene Expression Omnibus database (accession number: GSE12417; platform: GPL570, https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi? acc=GSE12417). Data from The Cancer Genome Atlas (TCGA) were downloaded from the TCGA data portal (https://portal.gdc.cancer.gov). All datasets are publicly available and can be accessed using the provided accession information and links. No new data were generated in this study. Declarations Ethics approval and consent to participate The research protocol received approval from the institutional ethics committee of The Affiliated Suqian Hospital of Xuzhou Medical University (No. 2025-042-01). The study was carried out following the ethical guidelines outlined in the Declaration of Helsinki. Consent for publication Not applicable. Informed consent Informed consent was waived due to the retrospective observational nature of 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. Lili Chen and Na Yang are regarded as co-first author. References Associated Data Supplementary Materials Supplementary Material 1. Supplementary Material 2. Data Availability Statement Data supporting the findings of this study were obtained from publicly accessible databases. Gene expression data were retrieved from the Gene Expression Omnibus database (accession number: GSE12417; platform: GPL570, https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi? acc=GSE12417). Data from The Cancer Genome Atlas (TCGA) were downloaded from the TCGA data portal (https://portal.gdc.cancer.gov). All datasets are publicly available and can be accessed using the provided accession information and links. No new data were generated in this study.