PTX3 as a diagnostic and prognostic biomarker in lung adenocarcinoma: a comprehensive analysis.

We started by investigating PTX3 mRNA expression in pan-cancer. According to the findings, PTX3 was expressed at low levels in the following cancers: skin cutaneous melanoma (SKCM), stomach adenocarcinoma (STAD), uterine corpus endometrial carcinoma (UCEC), thyroid cancer (THCA), lung squamous cell carcinoma (LUSC), lung adenocarcinoma (LUAD), esophageal carcinoma (ESCA), bladder urothelial carcinoma (BLCA), head and neck squamous cell carcinoma (HNSC), breast invasive carcinoma (BRCA), cervical endometrial adenocarcinoma (CESC), colon adenocarcinoma (COAD), kidney chromophobe (KICH), kidney renal clear cell carcinoma (KIRC), prostate adenocarcinoma (PRAD), kidney renal papillary cell carcinoma (KIRP), stomach adenocarcinoma (STAD), and rectal adenocarcinoma (READ), but cholangiocarcinoma (CHOL), glioblastoma multiforme (GBM) were highly expressed (Fig.  1 A). In addition, data from TCGA and GEO showed that PTX3 mRNA expression was significantly downregulated in LUAD tissues compared to adjacent normal controls ( P  < 0.01) (Fig.  1 B–D). Therefore, PTX3 is lowly expressed in almost all tumor types except cholangiocarcinoma (CHOL) and glioblastoma multiforme (GBM). The correlation between PTX3 mRNA expression and clinical features was analyzed in the TCGA cohort. Detailed parameters are shown in Table  2 . The expression of PTX3 mRNA was high in females and low in males ( P  < 0.05). The expression of PTX3 mRNA was high in the N2 stage and low in the N0 stage ( P  < 0.05). However, There was no correlation between PTX3 mRNA expression and clinical stage, metastasis (M), tumor classification (T), or age (Fig.  1 E–J). Table 2 Clinical features of LUAD patients in TCGA cohort Patient features Number of patients (%) Patient features Number of patients (%) Age Metastasis ≤ 65 210(44.8) M0 333(93.3) > 65 259(55.2) M1 24(6.7) Gender Lymph nodes Female 265(54.3) N0 312(65.8) Male Survival status Alive Dead 223(45.7) 291(62.3) 176(37.7) N1 N2 N3 Stage 90(19.0) 70(14.8) 2(0.4) T classification T1 T2 T3 T4 163(33.7) 260(53.8) 41(8.4) 19(4.1) I II III IV 262(54.8) 112(23.4) 79(16.5) 25(5.3) Clinical features of LUAD patients in TCGA cohort Male Survival status Alive Dead 223(45.7) 291(62.3) 176(37.7) N1 N2 N3 Stage 90(19.0) 70(14.8) 2(0.4) T classification T1 T2 T3 T4 163(33.7) 260(53.8) 41(8.4) 19(4.1) I II III IV 262(54.8) 112(23.4) 79(16.5) 25(5.3) Fig. 1 PTX3 mRNA expression in pan-cancer and LUAD. A PTX3 mRNA expression in pan-cancer. B PTX3 mRNA expression in TCGA. C , D PTX3 mRNA expression in GSE31210 ( C ) and GSE68571 ( D ). ( E – J ) Correlations between PTX3 mRNA expression and clinical features in LUAD patients according to Gender ( E ), Age ( G ), Clinical stage ( H ), N stage ( I ), M stage ( F ), and T classification ( J ). Statistical significance was determined by Mann-Whitney U test. Significance thresholds: ns, P  > 0.05, * P  < 0.05, ** P  < 0.01, *** P  < 0.001 PTX3 mRNA expression in pan-cancer and LUAD. A PTX3 mRNA expression in pan-cancer. B PTX3 mRNA expression in TCGA. C , D PTX3 mRNA expression in GSE31210 ( C ) and GSE68571 ( D ). ( E – J ) Correlations between PTX3 mRNA expression and clinical features in LUAD patients according to Gender ( E ), Age ( G ), Clinical stage ( H ), N stage ( I ), M stage ( F ), and T classification ( J ). Statistical significance was determined by Mann-Whitney U test. Significance thresholds: ns, P  > 0.05, * P  < 0.05, ** P  < 0.01, *** P  < 0.001 In our study, we used the ROC curve to assess the diagnostic ability of PTX3 in the TCGA and GEO databases for LUAD. The combined predictive efficacy of PTX3 and CEACAM5 was evaluated by constructing a binary logistic regression model. ROC curve analysis was performed to calculate the AUC values for the combined diagnostic performance, with comparisons made to individual biomarkers alone. The results indicated that in the TCGA cohort, the AUC of PTX3 was 0.816 (95% CI: 0.772–0.861, P  < 0.001), that of CEACAM5 was 0.817 (95% CI: 0.775–0.858, P  < 0.001), and that of combined detection was 0.881 (95% CI: 0.847–0.915, P  < 0.001). The AUC of the combined model was significantly higher than that of either CEACAM5 or PTX3 alone (DeLong’s test, P  < 0.001) (Fig.  2 A). In the GSE68571 cohort, the AUC was 0.781 (95% CI: 0.663–0.899, P  < 0.001) for PTX3 , 0.870 (95% CI: 0.778–0.961, P  < 0.001) for CEACAM5 , and 0.874 (95% CI: 0.788–0.960, P   0.05) (Fig.  2 B). In the GSE31210 cohort, the AUC was 0.720 (95% CI: 0.637–0.802, P  < 0.001) for PTX3 , 0.815 (95% CI: 0.750–0.880, P  < 0.001) for CEACAM5 , and 0.844 (95% CI: 0.779–0.908, P  < 0.001) for combined detection. The AUC of the combined model was significantly higher than that of either CEACAM5 or PTX3 alone (DeLong’s test, P  < 0.05) (Fig.  2 C). Our results show that PTX3 has a high diagnostic value, and PTX3 and CEACAM5 detecting together can increase diagnostic accuracy. Fig. 2 PTX3 mRNA expression ROC curve in TCGA and GEO LUAD patients. A PTX3 , CEACAM5 , and PTX3 combined CEACAM5 mRNA expression ROC curve in TCGA tumor and normal tissues. B PTX3 , CEACAM5 , and PTX3 combined CEACAM5 mRNA expression ROC curve in GSE68571 tumor and normal tissues. C PTX3 , CEACAM5 , and PTX3 combined CEACAM5 mRNA expression ROC curve in GSE31210 tumor and normal tissues. Statistical Analysis: AUC comparisons were performed using DeLong’s test PTX3 mRNA expression ROC curve in TCGA and GEO LUAD patients. A PTX3 , CEACAM5 , and PTX3 combined CEACAM5 mRNA expression ROC curve in TCGA tumor and normal tissues. B PTX3 , CEACAM5 , and PTX3 combined CEACAM5 mRNA expression ROC curve in GSE68571 tumor and normal tissues. C PTX3 , CEACAM5 , and PTX3 combined CEACAM5 mRNA expression ROC curve in GSE31210 tumor and normal tissues. Statistical Analysis: AUC comparisons were performed using DeLong’s test According to our findings, lower overall survival (OS) was associated with increased PTX3 expression in both clinical samples (Supplemental Fig. 1A) and the TCGA cohort (Fig.  3 A). To confirm these results, we conducted ROC curve on PTX3 expression data from clinical samples (Supplemental Fig. 1B) and the TCGA cohort (Fig.  3 B). Regarding the TCGA cohort, the areas under the ROC curve were 0.578, 0.582, and 0.619. The areas under the ROC curve for clinical samples were 0.720, 0.756, and 0.935. Subsequently, based on PTX3 and CEACAM5 mRNA expression, we separated LUAD patients into four groups and then carried out a Kaplan-Meier survival analysis. The patients with high expression of PTX3 and CEACAM5 had the worst prognosis ( P  < 0.05), according to the results (Fig.  3 C). Finally, we discovered that in both the clinical samples (Supplemental Fig. 1C) and the TCGA cohort (Fig.  3 D), high PTX3 expression was significantly linked to a higher risk of overall survival using univariate Cox analysis. In the TCGA cohort (Fig.  3 E) and clinical samples (Supplemental Fig. 1D), we also conducted multivariate Cox analysis and discovered that high PTX3 expression was strongly linked to a higher risk of OS. Fig. 3 Prognostic correlation analysis of PTX3 in LUAD in TCGA cohort. A Kaplan-Meier analysis for LUAD patients’ overall survival (Log-Rank test). B Time-dependent ROC curves at 3, 5- and 10- years. C Kaplan-Meier analysis of overall survival in four different subgroups of LUAD patients (Log-Rank test). D A forest plot of univariate Cox analysis risk factors for LUAD patients (Wald test). E A forest plot of multivariate Cox analysis risk factors for LUAD patients (Wald test) Prognostic correlation analysis of PTX3 in LUAD in TCGA cohort. A Kaplan-Meier analysis for LUAD patients’ overall survival (Log-Rank test). B Time-dependent ROC curves at 3, 5- and 10- years. C Kaplan-Meier analysis of overall survival in four different subgroups of LUAD patients (Log-Rank test). D A forest plot of univariate Cox analysis risk factors for LUAD patients (Wald test). E A forest plot of multivariate Cox analysis risk factors for LUAD patients (Wald test) The screening criteria were set at P  < 0.05 and |log2FC| ≥ 1 in order to analyze the DEGs between the PTX3 high and low expression groups using the limma package. The size of the |log2FC| value (Fig.  4 A) was used to filter a total of 60 differentially expressed genes (30 up-regulated and 30 down-regulated). STRING ( http://string-db.org ) was used for PPI analysis, and the igraph package was used to create the PPI network model (Fig.  4 B). Pearson correlation analysis was then used to conduct the coexpression analysis of PTX3 . Eleven genes connected with PTX3 expression were chosen based on the P -value and correlation coefficient values, and the R package corrplot was used to show the results (Fig.  4 C). Fig. 4 Identification of PTX3 -related genes. A DEGs with high and low PTX3 expression are displayed in a heatmap. B These differentially expressed genes’ PPI network. C Correlation between PTX3 expression and 11 genes Identification of PTX3 -related genes. A DEGs with high and low PTX3 expression are displayed in a heatmap. B These differentially expressed genes’ PPI network. C Correlation between PTX3 expression and 11 genes For the TCGA cohort, we employed GO terms (Supplemental Fig. 2A), the KEGG pathway (Supplemental Fig. 2B), and GSEA analysis (Supplemental Fig. 2C) to investigate potential biological functions of common DEGs. GO annotation results include cellular component (CC), molecular function (MF), and biological process (BP). The results showed that BP was mainly related to wound healing, including the regulation of wound healing, negative regulation of wound healing, plasminogen activation, and fibrinolysis. For MF analysis, DEGs were mainly enriched in signal receptor activator activity, receptor ligand activity, metal ion transmembrane transport activity, extracellular matrix structural components, and growth factor activity. DEGs were found to be involved in collagen-containing extracellular matrix, cation channel complex, Golgi lumen, platelet alpha granules, and platelet alpha granule lumen, according to the results of CC enrichment analysis (Supplemental Fig. 2A). The KEGG revealed that DEGs were enriched in the following areas: malaria, maturity-onset diabetes of the young, complement and coagulation cascades, cytoskeleton, neuroactive ligand-receptor interaction, PI3K-AKT signaling pathway, cytokine-cytokine receptor interaction, focal adhesion, and ECM-receptor interaction (Supplemental Fig. 2B). Additionally, we evaluated the potential PTX3 pathway in LUAD using GSEA. We discovered a few metabolic pathways that might be connected to LUAD, such as ribosomes, oxidative phosphorylation, and cytochrome P450’s metabolism of xenobiotics. Other pathways include KEGG Olfactory Transduction and KEGG Maturity Onset Diabetes of the Young. Our findings both explain the potential biological pathways of PTX3 in LUAD and show the intricate mechanism of PTX3 in LUAD (Supplemental Fig. 2C). We investigated the potential relationship between the level of PTX3 expression and the infiltration of stromal and immune cells using the ESTIMATE algorithm. We discovered a strong correlation between PTX3 expression and the LUAD stromal, immune, and ESTIMATE scores (Supplemental Fig. 3A). The immune checkpoint correlation analysis indicated that PTX3 had a high correlation with CD80, CD86, CD44, and PDCD1LG2 (Supplemental Fig. 3B), which may provide potential immunotherapeutic targets for patients with LUAD. Subsequently, we used ssGSEA to evaluate the enrichment scores of different immune cells, pathways, and functions to investigate the correlation between PTX3 and immune status. Patients with high expression of PTX3 had higher scores of co-stimulation and co-inhibition of antigen-presenting cells, co-stimulation and co-inhibition of T cells, CCR, checkpoint, HLA, MHC class I, parainflammation and type II IFN response activity than patients with low expression (Supplemental Fig. 3D). Furthermore, we assessed the antigen presentation process in high PTX3 patients that involved ADC, IDC, and DC. Moreover, compared to individuals with low expression of PTX3 , those with high expression had larger proportions of T helper cells, Th1, TILs, neutrophils, macrophages, mast cells, and Treg cells (Supplemental Fig. 3C). Molecular-targeted therapy and chemotherapy are common treatments for patients with advanced lung adenocarcinoma. The IC50 of common chemotherapy drugs and targeted drugs for each patient was estimated in the TCGA cohort. Our data showed the PTX3 high expression group exhibited higher sensitivity to the chemotherapy drug Oxaliplatin (Fig.  5 A). In addition, we also found that the PTX3 high expression group showed higher sensitivity to the molecular-targeted drugs KRAS inhibitor(G12C)-12, Nilotinib, Gefitinib, and Lapatinib (Fig.  5 C–F). In contrast, the PTX3 low-expression group indicated higher sensitivity to Dasatinib than the high-expression group (Fig.  5 B). Fig. 5 Correlations between PTX3 expression and sensitivity to drugs. Comparison of IC50 value for Oxaliplatin ( A ), Dasatinib ( B ), KRAS inhibitor(G12C)-12 ( C ), Nilotinib ( D ), Gefitinib ( E ), and Lapatinib ( F ) respectively. Statistical significance was determined by Mann-Whitney U test. Significance thresholds: ns, P  > 0.05, * P  < 0.05, ** P  < 0.01, *** P  < 0.001 Correlations between PTX3 expression and sensitivity to drugs. Comparison of IC50 value for Oxaliplatin ( A ), Dasatinib ( B ), KRAS inhibitor(G12C)-12 ( C ), Nilotinib ( D ), Gefitinib ( E ), and Lapatinib ( F ) respectively. Statistical significance was determined by Mann-Whitney U test. Significance thresholds: ns, P  > 0.05, * P  < 0.05, ** P  < 0.01, *** P  < 0.001 According to our results, PTX3 was highly expressed in the healthy control group and low expressed in the LUAD group (Fig.  6 A). However, CEA expression was significantly upregulated in the LUAD group (Fig.  6 B). Detailed parameters are shown in Table  3 . We further analyzed the ROC curve of the LUAD group and healthy control group, and the results showed that the AUC of PTX3 was 0.684 (95% CI: 0.591–0.776, P  < 0.001), the AUC of CEA was 0.745 (95% CI: 0.664–0.826, P  < 0.001), and the AUC of combined diagnosis of PTX3 and CEA was 0.765 (95% CI: 0.686–0.844, P  < 0.001) (Fig.  6 C). Subsequently, we analyzed the correlation between PTX3 and the clinical features of the LUAD group. The results showed that PTX3 was highly expressed in the T1 stage and low in the T4 stage ( P  < 0.05). However, PTX3 was not associated with age, gender, metastasis (M), lymph node (N), and clinical stages (Fig.  6 D–I). Table 3 Comparison of serum PTX3 and CEA expression levels between LUAD group and healthy control group Testing Indicators LUAD Healthy control P PTX3(ng/mL) 3.10(2.14,4.68) 4.46(3.24,9.55) < 0.001 CEA(ng/mL) 3.51(1.88,14.41) 1.78(1.28,2.51) < 0.0001 Comparison of serum PTX3 and CEA expression levels between LUAD group and healthy control group Fig. 6 PTX3 and CEA serum protein level in clinical samples. A PTX3 protein level in the serum of tumor patients and healthy controls. B CEA level in the serum of tumor patients and healthy controls. C ROC curves of PTX3, CEA, and their combination for distinguishing tumor patients from healthy controls. D – I Correlation between PTX3 protein level and clinical features in LUAD patients according to T classification ( H ), M stage ( F ), N stage ( G ), clinical stage ( I ), Age ( D ), and Gender ( E ). Statistical significance was determined by Mann-Whitney U test. Significance thresholds: ns, P  > 0.05, * P  < 0.05, ** P  < 0.01, *** P  < 0.001, **** P  < 0.0001 PTX3 and CEA serum protein level in clinical samples. A PTX3 protein level in the serum of tumor patients and healthy controls. B CEA level in the serum of tumor patients and healthy controls. C ROC curves of PTX3, CEA, and their combination for distinguishing tumor patients from healthy controls. D – I Correlation between PTX3 protein level and clinical features in LUAD patients according to T classification ( H ), M stage ( F ), N stage ( G ), clinical stage ( I ), Age ( D ), and Gender ( E ). Statistical significance was determined by Mann-Whitney U test. Significance thresholds: ns, P  > 0.05, * P  < 0.05, ** P  < 0.01, *** P  < 0.001, **** P  < 0.0001 The decreased PTX3 expression observed at the transcript level (Fig.  1 B–D) was concordant with proteomic data from the CPTAC, where PTX3 protein levels in LUAD tissues showed a significant reduction compared to adjacent normal controls (Mann-Whitney U test, P  < 0.0001; Supplemental Fig. 4A). CEA protein levels were significantly increased in LUAD tissues compared to adjacent normal controls (Mann-Whitney U test, P  < 0.0001; Supplemental Fig. 4B). We further analyzed the ROC curve of the LUAD tissues and adjacent normal controls, and the results showed that the AUC of PTX3 was 0.800 (95% CI: 0.738–0.863, P  < 0.001), the AUC of CEA was 0.694 (95% CI: 0.621–0.767, P  < 0.001), and the AUC of combined diagnosis of PTX3 and CEA was 0.865 (95% CI: 0.811–0.919, P  < 0.001). PTX3 demonstrated a significantly higher AUC compared to CEA (DeLong’s test, P  < 0.05), and the combined model exhibited a markedly greater AUC than either CEA or PTX3 alone (DeLong’s test, P  < 0.001) (Supplemental Fig. 4C). The results show that PTX3 is a superior biomarker, and PTX3 and CEA detecting together can increase diagnostic accuracy.

The transcriptome expression data and related clinical data of LUAD patients were obtained from the TCGA database. GSE31210 and GSE68571 were obtained from the GEO database (accession number: GSE31210 , GSE68571 ). Table  1 shows the total number of samples used in this study for all datasets. Table 1 Number of samples incorporated in this study Data Set Number of normal samples Number of LUAD samples Number of survival analysis TCGA 54 503 467 GSE31210 20 226 — GSE68571 Clinical samples CPTAC 9 40 102 87 97 109 — 97 — Number of samples incorporated in this study GSE68571 Clinical samples CPTAC 9 40 102 87 97 109 — 97 — A total of 137 clinical serum samples (tumor samples, 97 cases; normal samples, 40 cases) were collected in this study. Clinical information was also collected, such as TNM stages, age, and gender. All patients did not receive preoperative treatment. The collected serum samples were stored in the refrigerator at -80 ° C. All participants’ informed consent was obtained before the study. Using the TIMER2.0 online tool to analyze the expression of PTX3 in 33 kinds of tumor tissues and adjacent tissues [ 12 ]. The R software was used to process the data downloaded by TCGA and GEO, eliminate the missing and duplicate samples, use the limma package [ 13 ] and ggplot2 package to statistically analyze and visualize the results. The receiver operating characteristic (ROC) curve was created with R Package pROC, and the mRNA data of LUAD from the TCGA and GEO databases was analyzed. For visualization, the ggplot2 package was used. Survival analysis was performed using the Survminer and survival packages from the R software. Samples with incomplete clinical data should be discarded. The Kaplan–Meier method was used to create the survival curve. The log rank was used to evaluate statistical significance, with a P -value of 0.05 as the threshold. Then, univariate and multivariate Cox regression analyses were performed on PTX3 and clinical features to evaluate the relationship between these variables and prognosis. Finally, R software was used to create the ROC curve and confirm the prediction. The Pearson correlation coefficient was used to determine the genes associated with PTX3 using coexpression analysis (coefficient > 0.60, P  < 0.001). Then, using the R package limma, differentiation analysis was used to compare the high-expression and low-expression of PTX3 in order to identify the DEGs. DEGs that satisfied the following requirements were deemed significant: P -value  1. The STRING database [ 14 ] was used to analyze the protein-protein interaction (PPI) of differential genes, and the R package igraph was used for creating the PPI network model. Gene Ontology (GO) [ 15 ], Kyoto Encyclopedia of Genes and Genomes (KEGG) [ 16 ], and Gene Set Enrichment Analysis (GSEA) [ 17 ] on DEGs were performed using the R package cluster Profiler. The signaling pathways and biological functions were then investigated using an enriched plot, and the results were plotted using ggplot2. The stromal score and immune score of LUAD patients in the TCGA cohort were calculated by the ESTIMATE algorithm, the relations between these two scores and PTX3 expression were analyzed by Mann-Whitney U test [ 18 ]. The relationship between PTX3 and immune checkpoint expression was analyzed using the Spearman correlation test. The ssGSEA algorithm of GSVA package was used to investigate immune cell infiltration [ 19 ]. The R package OncoPredict [ 20 ] was used to investigate drug sensitivity in the TCGA cohort. Genomics of Drug Sensitivity in Cancer (GDSC) information was downloaded from oncoPredict ( https://osf.io/c6tfx/ ). The calcPhenotype function in the OncoPredict package was used to estimate the half maximum inhibitory concentration (IC50) for each sample. The IC50 values of drugs often used to treat LUAD were then compared between the two groups. In all serum samples, PTX3 was quantified using the ELISA method. PTX3 ELISA kit was purchased from Bioswamp (China) and used according to the manufacturer’s suggestions. The detection limit of the kit is 0.1ng/ml, and the precision in this range is < 10%. CEA was detected using ARCHITECT i2000sr automatic chemiluminescence immunoanalyzer (Abbott, USA), and the chemiluminescent microparticle immunoassay CEA assay kit was purchased from Abbott (USA). Publicly available proteomic data from the Clinical Proteomic Tumor Analysis Consortium (CPTAC, https://proteomics.cancer.gov ) were interrogated to assess PTX3 protein expression in LUAD ( n  = 109) versus normal adjacent tissue ( n  = 102). Normalized spectral counts were analyzed using the cProSite [ 21 ] ( https://cprosite.ccr.cancer.gov ). R software (version 4.4.0) was used to analyze and plot the data from the TCGA and GEO databases. GraphPad Prism (version 10.3) was used to analyze and plot clinical samples and data. The two groups were compared using the Mann-Whitney U test, and the normal distribution of continuous variables was confirmed using the Kolmogorov Smirnov test. P values < 0.05 were deemed statistically significant, and all statistical P values were bilateral.

Globally, lung cancer continues to be one of the main causes of cancer-related mortality [ 22 ]. LUAD, the main histological subtype of lung cancer, has seen an increasing incidence every year [ 23 ]. Nevertheless, the prognosis for LUAD patients remains poor, primarily due to the difficulty in early detection of lung adenocarcinoma [ 24 ]. Even though the survival rate for LUAD patients has increased dramatically because to developments in targeted therapy, chemotherapy, radiotherapy, and surgery, the disease still presents a serious public health concern worldwide [ 25 , 26 ]. Therefore, finding new treatment targets and diagnostic biomarkers for LUAD is crucial. As a member of the pentameric protein superfamily, PTX3 is a homomultimeric protein [ 27 ]. PTX3 plays an important role in inflammation, innate immunity, tissue repair, and cancer. PTX3 has been shown to have different expressions among different types of cancer, indicating that it may be different in different types of cancer. PTX3 is overexpressed in soft tissue sarcoma [ 28 ], myeloproliferative neoplasms [ 29 ], pancreatic cancer [ 30 ], gliomas [ 31 ] and hepatocellular carcinoma [ 32 ]. The PTX3 gene is lowly expressed in colorectal cancer and leiomyosarcoma [ 33 ]. Research has shown that PTX3 stimulates cancer cell proliferation, apoptosis, and metastasis in various malignant tumors through the PI3K/AKT/mTOR signaling pathway [ 34 , 35 ]. Interestingly, PTX3 also inhibits cell proliferation and tumor metastasis by reducing the production of cell cycle-related proteins during the G2/M phase [ 36 ]. In addition, PTX3 interacts with various fibroblast growth factors (FGFs), such as FGF2 and FGF8b, and inhibits FGF-dependent tumor angiogenesis, thereby inhibiting tumor growth [ 37 ]. So, this protein can be thought to have both anti-cancer and pro-cancer effects. The dual roles of PTX3 may reflect tissue-specific contexts: while it promotes metastasis in breast cancer via FGF signaling, its epigenetic silencing in LUAD (e.g., hypermethylation) could lead to loss of tumor-suppressive functions. While PTX3 is implicated in tumor progression across multiple cancers (e.g., breast and glioblastoma), its role in LUAD remains controversial, with conflicting reports on its expression and function. In a previous study, Infante et al. obtained overexpression of PTX3 at local and systemic levels in lung cancer patients [ 38 ]. Hu et al. found that PTX3 levels in bronchoalveolar lavage fluid of patients with small cell lung tumors were highly elevated. In addition, their study showed no statistical correlation between PTX3 levels and age or gender, which was also confirmed by our experimental results. However, our experimental results indicated that PTX3 was lowly expressed in patients with LUAD, consistent with the results of bioinformatics analysis in TCGA and GEO databases. Improving the prognosis of LUAD patients requires early diagnosis and treatment. Tumor biomarker detection has the advantages of minimally invasive, safe, and rapid detection. CEA, as a broad-spectrum tumor marker, is also up-regulated in LUAD. Our results also prove this point. However, CEA lacks specificity in the diagnosis of LUAD. In our study, we investigated the diagnostic value of PTX3 in LUAD. Consistent with transcriptomic findings, PTX3 protein levels were significantly reduced in LUAD tissues compared to adjacent normal controls (CPTAC dataset). Moreover, the diagnostic AUC of serum PTX3 (ELISA) was comparable to its protein-level AUC in the CPTAC cohort, further supporting its translational potential. The results indicate that PTX3 may be a promising biomarker. In addition, we jointly tested PTX3 and CEA and found that combined testing can significantly improve diagnostic accuracy. Based on the results, we expect to improve the accuracy of diagnosis in future clinical practice by jointly detecting the expression of PTX3 and CEA. To further investigate the association between PTX3 and prognosis, we performed a survival analysis. According to the findings, patients with higher PTX3 expression had a comparatively worse overall survival rate. The PTX3 high-expression group was primarily related to metabolic pathways, according to the GSEA enrichment analysis results. This may imply that the higher the PTX3 expression, the more vigorous the tumor metabolism, and the worse the patient’s prognosis. Previous studies have shown that PTX3 has both pro-cancer and anti-cancer effects. Therefore, we think that the possible reason is that PTX3 has anti-cancer effects in healthy populations while promoting cancer progression in LUAD patients. Of course, this hypothesis requires further research to confirm. Subsequently, the risk factors influencing the prognosis of LUAD were analyzed using univariate and multivariate Cox analysis. As shown in Fig.  3 D–E and Supplemental Fig. 1C–D, PTX3 levels and clinical stage are closely related to overall survival. This further indicates that PTX3 is a potential independent risk factor for LUAD. The expression of DNA damage markers, oxidative DNA damage, and higher Trp53 mutations are evidence of increased DNA damage associated with PTX3 deficiency [ 33 ]. Finally, the PTX3 promoter and regulatory regions are highly methylated in certain types of human mesenchymal and epithelial tumors. This epigenetic modification results in transcriptional inactivation and suppression of PTX3 expression [ 33 , 39 , 40 ]. In colorectal cancer, PTX3 gene methylation and silencing have been detected in adenomas and stage 1 neoplastic lesions [ 33 ]. However, the mechanism by which PTX3 participates in the development of LUAD is still unclear. Therefore, we explored the potential mechanisms of PTX3 -related DEGs. KEGG analysis indicated that these DEGs are mainly involved in the PI3K-AKT signaling pathway and cytokine-cytokine receptors interaction. GO analysis shows that DEGs mainly enrich signal receptor activator activity, receptor ligand activity, extracellular matrix structural components, and growth factor activity. GSEA analysis showed that DEGs mainly involve metabolic pathways, including oxidative phosphorylation and ribosomes. Previous studies have demonstrated that PTX3 can stimulate cancer cell proliferation, apoptosis, and metastasis in various malignant tumors through the PI3K-AKT signaling pathway [ 34 ]. Therefore, by regulating the PI3K-AKT signaling pathway, cellular metabolism, and immune-related response, we speculate that PTX3 -related DEGs could promote the growth and progression of LUAD. An increasing amount of research indicates the TME is crucial in the development and incidence of tumors. In TME, Th2 and Foxp3 + regulatory T (Treg) cells are typically related to tumor growth and a poor prognosis, whereas CD4 + Th1 cells and activated CD8 + T cells are typically related to good prognosis in tumor patients [ 41 , 42 ]. Furthermore, tumor-infiltrating dendritic cells can promote immune suppression and tolerance. Neutrophils tend to transform into pro-tumor M2 type cells [ 43 ]. According to this study, patients with elevated PTX3 expression demonstrated higher immune cell infiltration related to immune suppression, including neutrophils, TIL, Treg, and dendritic cells, which indicates that high PTX3 expression correlated with immunosuppressive TME features, including increased M2 macrophage infiltration and elevated PDCD1LG2 levels, suggesting PTX3 -driven immune evasion. In addition to immunotherapy, chemotherapy and targeted therapy are important treatment modalities for patients with advanced tumors. Previous clinical studies have indicated that oxaliplatin and gemcitabine with bevacizumab have good efficacy in advanced NSCLC [ 44 ]. Mir et al. showed that pemetrexed, oxaliplatin, and bevacizumab could be a first-line treatment option for patients with stage IV NSCLC [ 45 ]. Winegarden et al. showed that oxaliplatin and paclitaxel can be used to treat patients with advanced NSCLC [ 46 ]. Angiogenesis, chemotherapy resistance, tumor development, tumor metastasis, and prognosis are highly related to EGFR mutations [ 47 ]. The first-line treatment for advanced NSCLC patients who have EGFR gene-sensitive mutations has always been targeted drugs like gefitinib [ 48 ]. KRAS mutations are genetic drivers of multiple cancer types. KRAS-G12 mutations are predominant in human cancers, with G12C being the most common subtype of mutation in NSCLC [ 49 ]. Although RAS has long been considered “undruggable,” clinical trials to inhibit oncogenic RAS directly, have recently made promising progress with the development of drugs that specifically bind KRAS-G12C mutant proteins. In particular, patients with advanced NSCLC who have KRAS-G12C mutations are treated with the covalent KRAS-G12C inhibitors sotorasib and garsorasib [ 50 , 51 ]. Lapatinib is a tyrosine kinase inhibitor (TKI) that targets EGFR and HER2. Lapatinib inhibits lung cancer by blocking the polarization of M2 polarized macrophages, reducing the density of M2 polarized macrophages in the tumor region [ 52 ]. In this study, our results similarly showed that patients with high PTX3 expression were sensitive to oxaliplatin, gefitinib, lapatinib, and KARS inhibitor(G12C)-12. However, our research has several limitations that need to be noted. Initially, this study used data from the public databases, and additional clinical validation is needed to assess the prognostic significance of PTX3 in LUAD. Secondly, this study lacks in vitro experiments to elucidate the mechanism of PTX3 involvement in LUAD. Thirdly, to verify PTX3’s ability to predict immune treatment, chemotherapy, and targeted therapy responses, prospective research is required. PTX3 has important predictive value for the response of LUAD immunotherapy, chemotherapy, and targeted therapy. In conclusion, our study confirmed the clinical significance of PTX3 in LUAD. PTX3 has important diagnostic value, and combining PTX3 with CEA detection can significantly improve diagnostic accuracy. These findings could clarify how PTX3 contributes to the occurrence and progression of LUAD.

At present, lung cancer is one of the cancers with the highest incidence and mortality in the world [ 1 ]. Due to its high incidence and mortality, it has attracted countless researchers. Lung cancer has traditionally been divided into two major histological subtypes: small cell lung cancer and non-small cell lung cancer (NSCLC). The most prevalent subtype is NSCLC. Large cell carcinoma, squamous cell carcinoma, and adenocarcinoma are the three main histological subtypes of NSCLC. Adenosquamous cell carcinoma and sarcomatoid cell carcinoma are very rare. Among NSCLC subtypes, lung adenocarcinoma (LUAD) is the most prevalent [ 2 ]. The pathogenesis of lung cancer is still unclear, but there are several widely accepted theories. Genetic and epigenetic changes, gene mutations, smoking, chronic inflammation, etc., can promote the transformation of lung epithelial cells into cancer cells [ 3 ]. There is already evidence that a patient’s prognosis can be greatly enhanced if lung cancer is discovered early [ 4 ]. However, current diagnostic methods, such as bronchoscopy and lung CT, are expensive and invasive [ 5 , 6 ]. Fortunately, plasma protein biomarkers can already be used to diagnose early lung cancer [ 7 ]. Therefore, finding biomarkers that can make early diagnosis of lung cancer is necessary in the future. PTX3, a homo-multimeric protein, belongs to the pentraxin superfamily. It has previously been observed that PTX3 is a fluid-phase pattern recognition molecule conserved in evolution that plays a role in innate humoral immunity and the regulation of inflammatory response [ 8 ]. Current serum biomarkers like carcinoembryonic antigen (CEA) lack specificity, necessitating novel diagnostic targets. CEA refers to the protein detected in serum and tissues, whereas CEACAM5 denotes the encoding gene and its transcriptomic expression. Multiple studies have demonstrated that PTX3 plays a key role in the occurrence and development of tumors. For example, Melanoma cell invasion can be promoted by PTX3 via inflammatory pathways [ 9 ]. Increased expression of PTX3 accelerates the metastasis of breast cancer [ 10 ]. Previous research has established that the concentration of PTX3 in bronchoalveolar Lavage Fluid in lung cancer patients is elevated by ELISA [ 11 ]. However, whether it is elevated in the serum of lung cancer patients has not been fully clarified. We identified PTX3’s expression level, diagnostic value, and prognostic value in LUAD in our research. We further performed gene difference analysis, protein-protein interaction (PPI) network, and functional enrichment analysis to determine the function of PTX3 in LUAD. Moreover, we also correlated PTX3 expression with immune cell infiltration, immune checkpoints, and drug sensitivity. Finally, we also performed basic in vitro experiments to confirm its expression in LUAD and correlation with clinical features. As a result, our study indicates that PTX3 can be used as a novel prognostic and diagnostic biomarker for LUAD.

Below is the link to the electronic supplementary material. Supplementary Material 1 Supplementary Material 1