Abstract Objective This study aimed to identify independent prognostic factors for advanced unresectable pancreatic ductal adenocarcinoma (PDAC) and construct a nomogram-based prediction model. The efficacy of different chemotherapy regimens was evaluated based on metabolic risk levels. Methods Clinical data from 276 patients with unresectable PDAC treated between 2020 and 2022 were retrospectively analyzed. Cox proportional hazards regression identified prognostic factors, and survival analysis was performed using Kaplan-Meier methods. A nomogram was developed, and ROC analysis assessed its predictive performance. Two-way ANOVA evaluated chemotherapy efficacy, and TCGA transcriptomic data explored metabolic correlations. Results Metabolic syndrome (MetS) and distant metastasis were independent prognostic factors. Patients with MetS had significantly shorter survival. The nomogram showed good discrimination (AUC: 0.815 training, 0.793 validation). Patients without MetS had better outcomes with FOLFIRINOX or GS regimens. Transcriptomic analysis revealed metabolic pathways linked to PDAC progression. Conclusions MetS and distant metastasis significantly impact PDAC prognosis. Patients without MetS benefit more from specific chemotherapy regimens. Our predictive model may aid personalized treatment strategies. Keywords: Metabolic syndrome, Pancreatic ductal adenocarcinoma, Prediction model, Nomogram Introduction Pancreatic cancer (PC) is the third leading cause of cancer-related mortality worldwide, with a 5-year survival rate of only 8-9% [[32]1]. Due to the insidious onset of PC, early-stage patients often present with no obvious symptoms, and most are diagnosed at advanced stages. Only approximately 20% of patients are candidates for surgical resection [[33]2]. Classic pancreatic ductal adenocarcinoma (PDAC), which accounts for over 80% of all PC cases, primarily affects the pancreatic head and is characterized by its high malignancy and invasiveness [[34]3]. Current research has confirmed that risk factors for PDAC include obesity, diabetes, smoking, alcohol abuse, hyperlipidemia, chronic pancreatitis, and genetic predisposition [[35]4]. Surgical resection is the primary treatment for PDAC, complemented by chemotherapy. For patients with unresectable disease, chemotherapy becomes the main therapeutic approach. The American Society of Clinical Oncology (ASCO) guidelines recommend first-line chemotherapy regimens such as FOLFIRINOX (fluorouracil, leucovorin, irinotecan, and oxaliplatin), the AG regimen (gemcitabine and albumin-bound paclitaxel), and other gemcitabine-based chemotherapy regimens (the GS regimen: gemcitabine and S-1) [[36]5]. Metabolic syndrome (MetS) is defined as the presence of at least three metabolic abnormalities, including abdominal obesity, diabetes, hypertension, hypertriglyceridemia, hypercholesterolemia, and low high-density lipoprotein levels [[37]6]. MetS has been closely linked to the development of liver cancer, breast cancer, colorectal cancer, endometrial cancer, pancreatic cancer, and gastroesophageal junction cancer [[38]7]. Moreover, compared to tumors in patients without MetS, those with MetS have higher mortality and recurrence rates [[39]8]. Many components of MetS are also recognized as risk factors for PDAC. In the context of metabolic dysregulation and the associated chronic inflammatory state, chemotherapy-related hepatic and renal toxicity, along with metabolic side effects, may be exacerbated, potentially influencing treatment outcomes. Therefore, chemotherapy toxicity could be a significant factor affecting the prognosis of cancer patients with MetS [[40]9]. This study aims to construct a prognostic prediction model for PDAC patients by identifying risk factors and explore the most appropriate chemotherapy regimens for patients with different metabolic risk profiles. The goal is to extend survival and improve the prognosis for these patients. Patients and methods Patients A retrospective study was conducted, selecting patients with advanced unresectable PDAC who were treated at the First Affiliated Hospital of Zhengzhou University between January 1, 2020, and June 1, 2022. Clinical data of these patients were collected, following inclusion and exclusion criteria. Patients were followed up for survival status, with outcomes recorded at the observation cutoff time, along with survival time from diagnosis to the observation cutoff and overall survival time from diagnosis to death (up to June 11, 2023, with all enrolled patients having reached the endpoint event). After excluding decompensated liver cirrhosis patients (n = 1) or those who also had other gastrointestinal tumors (n = 3), data for 276 cases with advanced unresectable PDAC were used for subsequent analyses (Fig. [41]1). The clinical baseline data, laboratory test results, and imaging findings of the study subjects were collected, including age, gender, morning blood pressure, abdominal circumference, height, weight, history of hypertension, diabetes, and hyperlipidemia, as well as chemotherapy regimen and frequency. Laboratory test results included morning blood glucose (Glu), white blood cell count (WBC), red blood cell count (RBC), platelet count (PLT), hemoglobin (Hb), albumin (ALB), total bilirubin (TBIL), direct bilirubin (DBIL), alanine aminotransferase (ALT), aspartate aminotransferase (AST), total cholesterol (TC), triglycerides (TG), high-density lipoprotein (HDL), carbohydrate antigen 199 (CA199), and carbohydrate antigen 125 (CA125). Imaging assessments included tumor location, perivascular invasion, lymph node metastasis, and distant metastasis. These data were systematically collected to provide a comprehensive overview of the clinical and laboratory characteristics of the patients, contributing to the analysis of prognostic factors and the evaluation of treatment responses. Data for this study were accessed for research purposes on March 16, 2024. During data collection and analysis, all patient records were anonymized, and the authors did not have access to information that could directly identify individual participants. Fig. 1. [42]Fig. 1 [43]Open in a new tab Flowchart for screening classification. Flowchart illustrating the selection process of patients diagnosed with advanced unresectable PDAC from 2020 to 2022. A total of 280 patients were initially included, with 1 patient excluded due to decompensated liver cirrhosis and 3 patients excluded due to other gastrointestinal tumors. The final cohort consisted of 276 patients, who were randomly divided into training (n = 193) and validation (n = 83) sets at a 7:3 ratio for subsequent analysis Criteria for inclusion and exclusion Inclusion Criteria: Patients with a definitive pathological diagnosis of PDAC. Preoperative imaging diagnosed as advanced unresectable PDAC or patients who underwent palliative surgery following intraoperative confirmation of unresectability. Patients aged over 18 years, regardless of gender. Patients with an ECOG performance status score of 0 or 1. Patients who received at least 4 cycles of chemotherapy with the AG regimen, FOLFIRINOX regimen, or GS regimen. Patients who had not received any antitumor therapy prior to diagnosis. Patients with complete clinical baseline data, laboratory tests, and imaging results required for the study. Exclusion Criteria: Patients diagnosed with tumors in other tissues or organs in addition to PDAC. Patients with other end-stage diseases. Patients with infectious diseases, hematological disorders, or severe infectious conditions. Pregnant or breastfeeding patients at the time of diagnosis. Patients who have used corticosteroids or other medications that may affect laboratory test results within three months prior to diagnosis. Patients with missing follow-up data. Diagnostic criteria for MetS According to the diagnostic criteria of the International Diabetes Federation (IDF), the following conditions define MetS: (1)Abdominal obesity: Male waist circumference ≥ 90 cm, Female waist circumference ≥ 85 cm, and/or Body Mass Index (BMI) ≥ 25 kg/m²; (2)Fasting blood glucose (FBG) ≥ 6.1 mmol/L, and/or a diagnosis of diabetes with ongoing treatment; (3)Blood pressure (BP) ≥ 130/85 mmHg, and/or a diagnosis of hypertension with ongoing treatment; (4)Triglycerides (TG) ≥ 1.7 mmol/L, and/or Total Cholesterol (TC) ≥ 5.6 mmol/L; (5)High-Density Lipoprotein Cholesterol (HDL-C) < 1.04 mmol/L. A diagnosis of MetS is confirmed if the patient meets three or more of the above criteria. The scoring system and diagnostic flowchart constructed based on these criteria are shown in Fig. [44]2. Fig. 2. [45]Fig. 2 [46]Open in a new tab MetS diagnostic scoring system. Flowchart depicting MetS diagnosis in 276 pancreatic cancer patients based on five criteria: central obesity, BMI ≥ 25 kg/m², diabetes (FBG ≥ 6.1 mmol/L), hypertension (BP ≥ 130/85 mmHg), and dyslipidemia (TG ≥ 1.7 mmol/L, TC ≥ 5.6 mmol/L, or HDL-C < 1.04 mmol/L). Patients scoring ≥ 3 points were classified as MetS (n = 84, 30.43%), while 192 patients (69.57%) were non-MetS Statistical analysis A total of 276 patients were randomly divided into a training set and a validation set in a 7:3 ratio. The training set included 193 patients, while the validation set comprised 83 patients. Normally distributed continuous variables were expressed as means and standard deviations, and group differences were analyzed using the t-test. For variables that did not follow a normal distribution, data were presented as medians and interquartile ranges (first quartile, Q1; third quartile, Q3), and differences between groups were analyzed using the Mann-Whitney U test. Categorical data were expressed as frequencies and percentages, and differences between groups were assessed using the chi-square test. Univariate and multivariate Cox proportional hazards regression analyses were performed on the training set data, with collinearity risks excluded. Independent prognostic factors were identified, and results were expressed as hazard ratios (HR) with 95% confidence intervals (CI). Survival analysis of independent risk factors was performed using the Kaplan-Meier method, and differences in survival rates were compared using the Log-rank test. Based on the final set of included risk factors, a nomogram predictive model was constructed using the “rms” package in R software, with the model used to predict patients’ 1-year survival rates. Receiver operating characteristic (ROC) curves were plotted to evaluate the model’s discrimination power, with the area under the curve (AUC) and concordance index (C-index) used as indicators. Calibration curves were constructed to assess the model’s calibration ability, with the proximity of the calibration curve to the reference line reflecting the model’s consistency. Finally, decision curve analysis (DCA) was performed to evaluate the clinical utility of the predictive model. For the evaluation of chemotherapy efficacy, two-way analysis of variance (ANOVA) was used, with Bonferroni correction applied for multiple comparisons. Differential gene expression analysis was performed on mRNA sequencing data from the TCGA pancreatic cancer cohort, which included 183 samples (179 cancer tissues and 4 adjacent non-cancerous tissues). Using the R language environment in RStudio software, the “edgeR” package was applied to perform differential expression analysis of the counts data. Genes with|Log2 fold change (FC)| ≥ 2 and corrected p-values < 0.05 were considered significant for inclusion, and these genes were subjected to KEGG pathway enrichment analysis. All statistical analyses were performed with SPSS (26.0) and the R programming language (4.2.1). Statistical significance was set at p < 0.05. This study was conducted and reported in accordance with the STROBE (Strengthening the Reporting of Observational Studies in Epidemiology) guidelines. Results Demographic and clinical characteristics This study included a total of 16 independent variables, consisting of both continuous and categorical variables. To facilitate subsequent data processing, categorical variables were coded as follows: male gender was assigned a value of 1, and female gender was assigned a value of 0; patients diagnosed with MetS were assigned a value of 1, otherwise a value of 0; patients aged ≥ 60 years were assigned a value of 1, otherwise a value of 0; tumors located in the head of the pancreas were assigned a value of 0, those in the body of the pancreas were assigned a value of 1, and those in the tail of the pancreas were assigned a value of 2; patients with distant metastasis were assigned a value of 1, otherwise a value of 0. The Shapiro-Wilk normality test was performed on the continuous variables in both the training and validation sets, and the results showed that all the continuous data followed a skewed distribution. The test results are presented in Table [47]1. The comparison of clinical baseline data is shown in Table [48]2. There were no statistically significant differences between the clinical data of the training and validation sets regarding gender, age, MetS, tumor location, distant metastasis, CA199, CA125, ALT, AST, ALB, TBIL, DBIL, WBC, RBC, Hb, and PLT. Table 1. Normality test Feature Shapiro-Wilk Normality Test (P-value) Total (n = 276) Training set (n = 193) Validation set (n = 83) CA199 < 0.001 < 0.001 < 0.001 CA125 < 0.001 < 0.001 < 0.001 ALT < 0.001 < 0.001 < 0.001 AST < 0.001 < 0.001 < 0.001 ALB < 0.001 < 0.001 < 0.001 TBIL < 0.001 < 0.001 < 0.001 DBIL < 0.001 < 0.001 < 0.001 WBC < 0.001 < 0.001 < 0.001 RBC < 0.001 < 0.001 < 0.001 Hb 0.039 0.019 0.037 PLT < 0.001 < 0.001 < 0.001 [49]Open in a new tab Table 2. Comparison of clinical data between the training and validation sets Feature Total (n = 276) Training set (n = 193) Validation set (n = 83) Statistical Value p-value Gender, n (%) χ²=0.163 0.687  0 98 (35.51) 70 (36.27) 28 (33.73)  1 178 (64.49) 123 (63.73) 55 (66.27) MetS, n (%) χ²=0.129 0.719  0 192 (69.57) 133 (68.91) 59 (71.08)  1 84 (30.43) 60 (31.09) 24 (28.92) Age, n (%) χ²=0.019 0.891  0 118 (42.75) 82 (42.49) 36 (43.37)  1 158 (57.25) 111 (57.51) 47 (56.63) Location, n (%) χ²=0.266 0.876  0 140 (50.72) 96 (49.74) 44 (53.01)  1 86 (31.16) 61 (31.61) 25 (30.12)  2 50 (18.12) 36 (18.65) 14 (16.87) Metastasis n (%) χ²=0.163 0.687  0 98 (35.51) 70 (36.27) 28 (33.73)  1 178 (64.49) 123 (63.73) 55 (66.27) CA199, M (Q1, Q3) 510.77 (106.80–1510.00) 468.10 (109.60–1440.00) 537.57 (106.20–1868.50) Z = 0.271 0.787 CA125, M (Q1, Q3) 34.52 (16.68–143.10) 30.85 (17.32–121.00) 47.02 (14.64–198.65) Z = 0.621 0.535 ALT, M (Q1, Q3) 21.50 (14.00–53.00) 20.00 (14.00–53.00) 23.00 (16.00–55.50) Z = 0.734 0.464 AST, M (Q1, Q3) 25.50 (18.00–45.00) 24.00 (17.00–47.00) 26.00 (19.00–42.00) Z = 0.696 0.487 ALB, M (Q1, Q3) 40.35 (37.70–42.40) 40.90 (38.50–42.80) 39.90 (36.10–42.30) Z = 1.631 0.103 TBIL, M (Q1, Q3) 13.12 (8.72–23.70) 12.88 (8.72–23.26) 13.70 (8.81–25.95) Z = 0.384 0.701 DBIL, M (Q1, Q3) 5.67 (3.80–11.10) 5.70 (3.71–10.20) 5.60 (4.25–13.90) Z = 0.440 0.660 WBC, M (Q1, Q3) 5.96 (4.77–7.30) 5.90 (4.80–7.27) 6.10 (4.70–7.45) Z = 0.155 0.876 RBC, M (Q1, Q3) 4.16 (3.71–4.48) 4.20 (3.79–4.48) 4.08 (3.68–4.49) Z = 0.606 0.544 Hb, M (Q1, Q3) 127.00 (114.00–137.00) 127.00 (114.20–136.10) 125.50 (113.40–140.75) Z = 0.083 0.934 PLT, M (Q1, Q3) 185.00 (144.00–240.00) 185.00 (144.00–240.00) 197.00 (138.50–239.50) Z = 0.187 0.852 [50]Open in a new tab Cox regression analysis and survival analysis In this study, a total of 276 patients were included. At the observation cutoff time, 186 patients (67.39%) had died, while 90 patients (32.61%) were still alive. In the 193 patients of the training set, 127 patients (65.80%) had died, and 66 patients (34.20%) were alive at the observation cutoff time. In the 83 patients of the validation set, 59 patients (71.08%) had died, and 24 patients (28.92%) were alive at the observation cut-off time. Univariate Cox proportional hazards regression analysis was performed on the feature variables, identifying CA199, MetS, PLT, and distant metastasis as risk factors. A multicollinearity diagnosis was subsequently conducted. The variance inflation factor (VIF) values for all four variable groups were found to be less than 5, indicating no multicollinearity risk. Finally, multivariate Cox proportional hazards regression analysis revealed that MetS and distant metastasis were independent prognostic factors. The results of the Cox regression analysis are shown in Table [51]3, and the multicollinearity diagnosis results are presented in Table [52]4. Table 3. Univariate and multivariate Cox regression analysis Feature Univariate Multivariate HR (95%CI) p-value HR (95%CI) p-value CA199 1.01 (1.01 ~ 1.01) 0.011 1.00 (1.00 ~ 1.00) 0.170 CA125 1.00 (1.00 ~ 1.00) 0.368 ALT 1.00 (1.00 ~ 1.00) 0.243 AST 1.00 (1.00 ~ 1.00) 0.717 TBIL 1.00 (0.99 ~ 1.00) 0.053 DBIL 1.00 (0.99 ~ 1.00) 0.107 ALB 1.01 (0.99 ~ 1.04) 0.350 WBC 0.96 (0.90 ~ 1.02) 0.186 RBC 1.01 (0.79 ~ 1.30) 0.920 Hb 1.00 (0.99 ~ 1.01) 0.503 PLT 0.99 (0.99 ~ 0.99) < 0.001 1.00 (1.00 ~ 1.00) 0.069 Gender  0 Reference  1 0.93 (0.69 ~ 1.25) 0.610 Age  0 Reference  1 0.96 (0.72 ~ 1.29) 0.799 MetS  0 Reference Reference  1 2.33 (1.73 ~ 3.14) < 0.001 2.09 (1.54 ~ 2.83) < 0.001 Location  0 Reference  1 1.24 (0.90 ~ 1.73) 0.191  2 1.14 (0.77 ~ 1.68) 0.514 Metastasis  0 Reference Reference  1 3.44 (2.41 ~ 4.91) < 0.001 3.07 (2.13 ~ 4.43) < 0.001 [53]Open in a new tab Table 4. Multicollinearity diagnosis Feature VIF Tolerance CA199 1.058 0.945 MetS 1.017 0.983 Metastasis 1.109 0.902 PLT 1.345 0.743 [54]Open in a new tab MetS was identified as an independent prognostic factor in the multivariate Cox regression analysis, as shown in Fig. [55]3A and B. Kaplan-Meier survival curves were generated for patients with and without MetS, and the survival rate differences were compared using the Log-rank test. The survival curves for both the training and validation sets indicate that the presence of MetS is significantly associated with a shorter survival time. The median survival time for patients without MetS was 317 days, while the median survival time for patients with MetS was 212 days. Fig. 3. [56]Fig. 3 [57]Open in a new tab MetS survival curves and model evaluation in training and validation sets. Prognostic analysis and model validation for MetS in PDAC patients. (A, B) K-M survival curves comparing OS between MetS (n = 84) and non-MetS (n = 192) patients in the training (A) and validation (B) cohorts. Patients with MetS show significantly shorter OS (Log-rank test, p < 0.05). (C, D) ROC curves evaluating the discriminatory power of the prognostic model in the training (C) and validation (D) cohorts. The AUC indicates good predictive accuracy. (E, F) DCA curve assessing the clinical utility of the model in the training (E) and validation (F) cohorts. The net benefit curves demonstrate the model’s advantage over default strategies. (G, H) Calibration curves comparing predicted vs. actual survival probabilities in the training (G) and validation (H) cohorts. The close alignment with the reference line indicates strong predictive performance Development and validation of the prediction model A Nomogram model was constructed using MetS and distant metastasis as the two independent prognostic factors identified from the multivariate Cox regression analysis. This model allows for the prediction of the survival probability for patients with advanced unresectable PDAC based on these two key factors (Fig. [58]4B). The discriminatory ability of the clinical prediction model was assessed using ROC curves for both the training and validation sets, with the C-index and AUC as key indicators. The results showed that the AUC values for the model were 0.815 (95% CI: 0.752–0.879) in the training set and 0.793 (95% CI: 0.621–0.858) in the validation set, while the C-index values were 0.680 and 0.656, respectively. These findings indicate that the model has good discriminatory power (Fig. [59]3C and D). The calibration curves for both the training and validation sets demonstrated that the model has good predictive accuracy (Fig. [60]3G and H). To further evaluate the model’s effectiveness, Decision Curve Analysis was conducted, and the results indicated that the model has strong clinical applicability (Fig. [61]3E and F). This suggests that the model provides meaningful and practical value in guiding clinical decision-making. Fig. 4. [62]Fig. 4 [63]Open in a new tab Efficacy analysis model, nomogram and bioinformatics analysis. Analysis of chemotherapy efficacy, prognostic modeling, and bioinformatics insights in advanced unresectable PDAC. (A) Bar chart illustrating the relationship between OS and the presence of MetS in the two-way ANOVA analysis of chemotherapy efficacy. (B) Nomogram for predicting OS, incorporating key prognostic factors to facilitate individualized risk assessment. (C) Volcano plot displaying differentially expressed genes identified through bioinformatics analysis. (D) KEGG pathway enrichment analysis highlighting the key molecular pathways associated with PDAC progression Selection and efficacy of chemotherapy regimens In the evaluation of chemotherapy regimens and efficacy in patients with unresectable PDAC, we first collected the baseline data of the patients (Table [64]5). Using two-way analysis of variance under the main effects test, we compared the differences in OS between patients receiving different chemotherapy regimens and with or without MetS (Table [65]6). The results indicated that the presence of MetS significantly affected OS, while the chemotherapy regimen did not have a significant impact on OS (Fig. [66]4A). Furthermore, the interaction between MetS and the chemotherapy regimen was not statistically significant. This finding aligns with the survival analysis, where MetS was identified as an independent prognostic factor for PDAC patients. After controlling for individual variables and performing pairwise post-hoc comparisons using the Bonferroni method, we found that, when controlling for chemotherapy regimens, the OS differences between patients with and without MetS were not statistically significant for those receiving the AG regimen. However, for patients receiving the FOLFIRINOX and GS regimens, the OS differences between those with and without MetS were statistically significant, with the OS in the non-MetS group being significantly higher than that in the MetS group (Tables [67]7 and [68]8). Table 5. Statistical description of chemotherapy regimen application Feature Total (n = 276) OS, M (Q₁, Q₃) 275.50 (174.00, 453.00) MetS, n (%)  0 192 (69.57)  1 84 (30.43) Regimen, n (%)  AG 94 (34.06)  FOLFIRINOX 86 (31.16)  GS 96 (34.78) [69]Open in a new tab Table 6. Between-Subjects effects test Item Sum of Squares DF Mean Square F-value p-value Adjusted Mode 774990.954 5 154998.191 2.831 0.016 Intercept 23594844.620  1 23594844.620 430.959 0.000 MetS 543115.935  1 543115.935 9.920 0.002 Regimen 120566.132  2 60283.066 1.101 0.334 MetS*Regimen 167430.895  2 83715.447 1.529 0.219 Error 14782393.872 270 54749.607 Total 47851762.000 276 Adjusted Total 15557384.826 275 [70]Open in a new tab DF: Degrees of Freedom Table 7. Pairwise comparison results controlling for MetS variable MetS Regimen Mean Difference SE p-value 95%CI Lower Limit Upper Limit 0 AG FOLFIRINOX -47.510 42.047 0.779 -148.801 53.781 GS 7.627 41.444 1.000 -92.212 107.466 FOLFIRINOX AG 47.510 42.047 0.779 -53.781 148.801 GS 55.138 40.751 0.532 -43.030 153.305 GS AG -7.627 41.444 1.000 -107.466 92.212 FOLFIRINOX -55.138 40.751 0.532 -153.305 43.030 1 AG FOLFIRINOX 79.428 64.023 0.647 -74.802 233.658 GS 98.382 59.713 0.302 -45.465 242.230 FOLFIRINOX AG -79.428 64.023 0.647 -233.658 74.802 GS 18.955 66.663 1.000 -141.636 179.545 GS AG -98.382 59.713 0.302 -242.230 45.465 FOLFIRINOX -18.955 66.663 1.000 -179.545 141.636 [71]Open in a new tab SE: Standard Error Table 8. Pairwise comparison results controlling for regimen variable Regimen MetS Mean Difference SE p-value 95% CI Lower Limit Upper Limit AG  0  1 24.951 50.227 0.620 -73.936 123.838  1  0 -24.951 50.227 0.620 -123.838 73.936 FOLFIRINOX  0  1 151.889 57.828 0.009 38.038 265.740  1  0 -151.889 57.828 0.009 -265.740 -38.038 GS  0  1 115.706 52.540 0.028 12.265 219.147  1  0 -115.706 52.540 0.028 -219.147 -12.265 [72]Open in a new tab Bioinformatics analysis A total of 183 transcriptomic clinical samples of pancreatic cancer were retrieved from the TCGA database, comprising 179 tumor tissue samples and 4 adjacent normal tissue samples. Differential gene expression analysis was performed using the R programming language. The identified differentially expressed genes were further analyzed through KEGG enrichment. The results indicated that the pancreatic cancer transcriptomic samples were significantly enriched in pathways associated with lipid metabolism, carbohydrate metabolism, arachidonic acid metabolism, and pancreatic endocrine and exocrine functions (Fig. [73]4C and D). Discussion Global burden of PDAC and relevance with MetS Pancreatic cancer, predominantly PDAC, ranks as the third leading cause of cancer-related mortality worldwide, based on comprehensive global incidence reports encompassing 185 countries [[74]10]. The prognosis of pancreatic cancer remains dismal, with a 5-year relative survival rate of only 8%, primarily attributable to its tendency to present at a locally advanced or metastatic stage at the time of diagnosis [[75]11]. Established risk factors for PDAC include smoking, chronic pancreatitis, advanced age, diabetes, obesity, occupational exposure, and high-fat dietary patterns [[76]12]. Metabolic dysregulation is key to both the onset and progression of PDAC. A recent large-scale study by Miyashita et al. demonstrated a clear association between MetS and the incidence of pancreatic cancer, highlighting MetS as an independent risk factor for PDAC across diverse populations [[77]13]. MetS is a cluster of metabolic abnormalities, including central obesity, hyperglycemia, hypertension, hypertriglyceridemia, and low HDL cholesterol levels [[78]14]. It can be readily diagnosed through standard clinical assessments, providing a cost-effective and widely accessible tool for prognostic evaluation and monitoring of chemotherapy efficacy. In this retrospective study of patients with advanced unresectable PDAC, we analyzed standardized clinical variables obtained from routine examinations to ensure accessibility and consistency. Multivariate analyses using R and SPSS revealed that both MetS and distant metastasis were independent predictors of overall survival. Based on these factors, we constructed a nomogram with strong discriminatory power and clinical applicability. Pathophysiological links between MetS components and PDAC Each component of MetS exerts both individual and combined effects in the development and progression of PDAC. Hyperglycemia and insulin resistance—hallmarks of type 2 diabetes—contribute to tumorigenesis through the activation of insulin-like growth factor-1 (IGF-1) signaling, which promotes tumor cell proliferation, inhibits apoptosis, and enhances angiogenesis within the tumor microenvironment [[79]15]. Additionally, hyperglycemia activates the p38 MAPK pathway and induces the release of inflammatory cytokines such as IL-6 and vascular endothelial growth factor (VEGF), facilitating tumor invasion and distant metastasis [[80]16, [81]17]. Visceral obesity, reflected by increased BMI and central adiposity, results in elevated metabolic activity within adipose tissue and excessive production of lactate dehydrogenase and reactive oxygen species (ROS) [[82]18]. These factors disrupt adipokine balance—particularly leptin and adiponectin—and promote a chronic proinflammatory state that alters the tumor microenvironment and supports PDAC progression [[83]19]. Leptin has been shown to stimulate tumor cell proliferation and angiogenesis via the leptin–Notch axis, whereas reduced adiponectin levels enhance IL-6 and TNF-α signaling, further exacerbating local and systemic inflammation [[84]20–[85]22]. Dyslipidemia, especially fluctuations in total cholesterol levels, has also been implicated in PDAC development. Aberrant lipid metabolism activates sterol regulatory element-binding proteins (SREBPs) and upregulates transforming growth factor-beta (TGF-β), which drives epithelial–mesenchymal transition and increases metastatic potential [[86]23]. Furthermore, chronic hypertension can lead to pancreatic microvascular injury, endocrine dysfunction, and immune impairment, while the long-term use of antihypertensive agents may disturb metabolic homeostasis, increase oxidative stress, and modulate tumor-related signaling pathways [[87]24, [88]25]. Collectively, these findings suggest that the systemic metabolic disturbances observed in MetS not only contribute to tumor initiation but also foster a microenvironment conducive to PDAC progression and therapeutic resistance. MetS severity and cumulative risk for PDAC In this study, both univariate and multivariate Cox regression analyses revealed that MetS was significantly associated with poorer prognosis in patients with PDAC. Survival curves from both the training and validation cohorts consistently demonstrated that patients with MetS had significantly shorter overall and median survival times compared to those without MetS. These findings are supported by previous clinical studies indicating that MetS is an independent risk factor for cancer development in the general population. Notably, individuals with MetS have a 57% higher risk of developing PDAC compared to those without MetS but with other metabolic risk factors (HR = 1.57; 95% CI: 1.39–1.76) [[89]26]. Further research has highlighted the dynamic nature of MetS. Individuals who recovered from MetS still exhibited a higher risk of pancreatic cancer than those who never developed MetS, although their risk remained lower than that of patients with persistent MetS [[90]13]. These observations suggest that the reversibility of MetS through pharmacological or lifestyle interventions may help reduce cancer risk. Additionally, the number of metabolic abnormalities appears to be positively correlated with PDAC risk. The presence of four or five MetS components significantly increased the risk of developing pancreatic cancer by approximately 40% (HR = 1.47; 95% CI: 1.19–1.81) and 60% (HR = 1.64; 95% CI: 1.06–2.51), respectively [[91]27]. These findings underscore the cumulative impact of metabolic dysfunction on PDAC development and highlight the need for early identification and management of MetS in high-risk populations. Impact of chemotherapy regimens on prognosis in MetS vs. Non-MetS patients The treatment of unresectable locally advanced or metastatic pancreatic cancer necessitates a comprehensive and individualized approach to chemotherapy selection. According to ASCO guidelines, recommended first-line regimens include AG (gemcitabine plus albumin-bound paclitaxel), FOLFIRINOX (fluorouracil, leucovorin, irinotecan, and oxaliplatin), and GS (gemcitabine plus S-1) [[92]28]. In this study, patients received one of three ASCO-recommended first-line chemotherapy regimens—AG, FOLFIRINOX, or GS—for at least four cycles. Our findings suggest that the AG regimen may be less influenced by metabolic status, whereas patients without MetS appeared to benefit more from FOLFIRINOX or GS. AG, the standard regimen for Chinese patients, is typically administered on days 1, 8, and 15 of a 4-week cycle and is associated with lower toxicity, making it more tolerable and less likely to exacerbate metabolic disturbances. FOLFIRINOX, a four-drug combination, is considerably more intensive and carries a high risk of severe toxicity, including neutropenia, gastrointestinal side effects, hepatotoxicity, and renal impairment. Its use is generally limited to patients with good performance status and stable systemic metabolic conditions. These characteristics are consistent with our observations, which suggest that patients without MetS tolerate the regimen better and may achieve greater clinical benefit. The GS regimen, combining oral S-1 with intravenous gemcitabine, is relatively well tolerated and suited for patients with poorer functional status. However, its shorter treatment cycle and continuous exposure to S-1 may sustain elevated systemic drug levels, potentially aggravating metabolic disturbances. Therefore, for patients with significant metabolic risk, FOLFIRINOX and GS may be less optimal than AG. Transcriptomic insights into metabolic pathways in PDAC To further substantiate the findings of this study and provide a foundation for future research, we conducted transcriptomic analyses of pancreatic cancer samples using data from TCGA. The results revealed significant enrichment in metabolic pathways closely linked to the pathophysiology of PDAC, including those involved in carbohydrate and lipid metabolism, arachidonic acid metabolism, and both endocrine and exocrine pancreatic functions. These enriched pathways mirror the metabolic dysregulation observed in individuals with MetS, particularly in relation to hyperglycemia and hyperlipidemia, further reinforcing the clinical relevance of MetS in PDAC progression. Among these, arachidonic acid metabolism—particularly via the cyclooxygenase (COX) enzymatic pathway—emerged as a potential contributor to PDAC development [[93]29]. Notably, COX-2, a key enzyme in this pathway, has been reported to be overexpressed in PDAC and is implicated in tumor progression, inflammatory responses, and therapeutic resistance [[94]30]. These transcriptomic insights from public datasets like TCGA not only validate the metabolic alterations observed clinically but also underscore the importance of metabolic reprogramming in PDAC pathogenesis. This integrative approach lays a solid groundwork for further mechanistic studies and the development of metabolism-targeted therapeutic strategies in pancreatic cancer. Limitations This study has several important limitations that should be acknowledged. First, it was conducted using data derived from a single medical center, which may limit the external validity and generalizability of the findings. The prevalence and clinical presentation of MetS are known to vary significantly across different geographic regions, ethnic groups, and healthcare systems. Therefore, multicenter studies involving more diverse populations are needed to validate and extend our results. Second, the retrospective design of the study introduces the potential for recall and selection biases, particularly in the collection and accuracy of baseline clinical data, lifestyle factors, and comorbidities. Although we employed rigorous data collection procedures, the inherent limitations of retrospective analyses may affect the robustness of certain findings. Third, our study cohort was limited to patients with advanced, unresectable PDAC, a population typically associated with poor prognosis and limited overall survival. As such, our predictive model was constrained to estimating 1-year survival, which may not capture the full prognostic trajectory of PDAC across its clinical spectrum. Future studies should incorporate patients with resectable and borderline resectable disease to allow for a more comprehensive survival analysis and broader clinical applicability. Lastly, although we assessed the efficacy of chemotherapy, we did not account for variations in chemotherapy regimens, dose intensity, treatment duration, or patient-specific tolerability, all of which may have had a meaningful impact on treatment response and survival outcomes. Detailed treatment-related variables should be integrated in future models to improve predictive accuracy and clinical utility. Highlights and future directions While the link between MetS and pancreatic cancer—especially in terms of disease susceptibility and early-stage progression—has been well-documented, our study provides several novel insights that advance the current understanding of this relationship. First, unlike many previous studies that focused on early-stage pancreatic cancer or general associations, our research specifically investigates the prognostic role of MetS in advanced unresectable pancreatic cancer, a population with limited treatment options and poor prognosis. Second, we developed and internally validated a novel nomogram-based predictive model incorporating MetS and metastasis status, which demonstrated good discrimination and calibration, providing practical value for individualized prognostication. Third, we analyzed the impact of MetS on chemotherapy efficacy, offering additional insights into the interaction between metabolic status and treatment response—an area rarely addressed in prior studies. Finally, our findings are based on real-world clinical data, enhancing the applicability and translational relevance of the results. These strengths enhance the clinical significance of our study and support its contribution to optimizing treatment strategies in advanced PDAC. Our study highlights the prognostic relevance of MetS in advanced PDAC and supports the integration of metabolic parameters into individualized risk stratification. In line with this, a recent study by Shi et al. proposed a novel metabolic prognostic score based on multiple metabolic indicators to predict survival across various cancer types, demonstrating its broad applicability and clinical utility [[95]31]. These findings collectively underscore the importance of systemic metabolic status in cancer prognosis and suggest that integrating comprehensive metabolic scoring systems may further enhance the predictive performance of existing models in PDAC and beyond. Future studies may consider combining traditional MetS components with additional metabolic biomarkers to construct more robust, tumor-specific prognostic tools. Moreover, our study underscores the importance of increasing public awareness of MetS and encouraging healthier lifestyle modifications, such as the cessation of deleterious habits and the adoption of improved dietary practices. Given the potential reversibility of MetS, targeting this condition may offer a promising avenue for the primary prevention of cancer, highlighting the critical need for early intervention and proactive risk management strategies. Conclusion Through the analysis of clinical data from 276 patients with advanced, unresectable PDAC, this study identified MetS and distant metastasis as independent prognostic risk factors. A Nomogram-based prediction model was developed accordingly. Treatment efficacy analysis revealed that when employing FOLFIRINOX or GS chemotherapy regimens, consideration of the patient’s metabolic status is essential, as patients without MetS demonstrated superior therapeutic outcomes. Furthermore, bioinformatics analyses highlighted the significant association of multiple metabolic pathways with pancreatic cancer. Future investigations will aim to elucidate the underlying molecular mechanisms, expand sample sizes, and validate findings through multi-center cohort studies. Acknowledgements