Houssam Younes

Background/Objectives: Peripheral artery disease (PAD) impacts more than 200 million individuals globally and leads to mortality and morbidity secondary to progressive limb dysfunction and amputation. However, clinical management of PAD remains suboptimal, in part because of the lack of standardized biomarkers to predict patient outcomes. Growth differentiation factor 15 (GDF15) is a stress-responsive cytokine that has been studied extensively in cardiovascular disease, but its investigation in PAD remains limited. This study aimed to use explainable statistical and machine learning methods to assess the prognostic value of GDF15 for limb outcomes in patients with PAD. Methods: This prognostic investigation was carried out using a prospectively enrolled cohort comprising 454 patients diagnosed with PAD. At baseline, plasma GDF15 levels were measured using a validated multiplex immunoassay. Participants were monitored over a two-year period to assess the occurrence of major adverse limb events (MALE), a composite outcome encompassing major lower extremity amputation, need for open/endovascular revascularization, or acute limb ischemia. An Extreme Gradient Boosting (XGBoost) model was trained to predict 2-year MALE using 10-fold cross-validation, incorporating GDF15 levels along with baseline variables. Model performance was primarily evaluated using the area under the receiver operating characteristic curve (AUROC). Secondary model evaluation metrics were accuracy, sensitivity, specificity, negative predictive value (NPV), and positive predictive value (PPV). Prediction histogram plots were generated to assess the ability of the model to discriminate between patients who develop vs. do not develop 2-year MALE. For model interpretability, SHapley Additive exPlanations (SHAP) analysis was performed to evaluate the relative contribution of each predictor to model outputs. Results: The mean age of the cohort was 71 (SD 10) years, with 31% (n = 139) being female. Over the two-year follow-up period, 157 patients (34.6%) experienced MALE. The XGBoost model incorporating plasma GDF15 levels and demographic/clinical features achieved excellent performance for predicting 2-year MALE in PAD patients: AUROC 0.84, accuracy 83.5%, sensitivity 83.6%, specificity 83.7%, PPV 87.3%, and NPV 86.2%. The prediction probability histogram for the XGBoost model demonstrated clear separation for patients who developed vs. did not develop 2-year MALE, indicating strong discrimination ability. SHAP analysis showed that GDF15 was the strongest predictive feature for 2-year MALE, followed by age, smoking status, and other cardiovascular comorbidities, highlighting its clinical relevance. Conclusions: Using explainable statistical and machine learning methods, we demonstrated that plasma GDF15 levels have important prognostic value for 2-year MALE in patients with PAD. By integrating clinical variables with GDF15 levels, our machine learning model can support early identification of PAD patients at elevated risk for adverse limb events, facilitating timely referral to vascular specialists and aiding in decisions regarding the aggressiveness of medical/surgical treatment. This precision medicine approach based on a biomarker-guided prognostication algorithm offers a promising strategy for improving limb outcomes in individuals with PAD.

Peripheral artery disease (PAD) is associated with an elevated risk of major adverse cardiovascular events (MACE). Despite this, few reliable biomarkers exist to identify patients at heightened risk of MACE. Growth differentiation factor 15 (GDF15), a stress-responsive cytokine implicated in inflammation, atherosclerosis, and thrombosis, has been broadly studied in cardiovascular disease but remains underexplored in PAD. This study aimed to evaluate the prognostic utility of GDF15 for predicting 2-year MACE in PAD patients using explainable statistical and machine learning approaches. We conducted a prospective analysis of 1192 individuals (454 with PAD and 738 without PAD). At study entry, patient plasma GDF15 concentrations were measured using a validated multiplex immunoassay. The cohort was followed for two years to monitor the occurrence of MACE, defined as stroke, myocardial infarction, or death. Baseline GDF15 levels were compared between PAD and non-PAD participants using the Mann–Whitney U test. A machine learning model based on extreme gradient boosting (XGBoost) was trained to predict 2-year MACE using 10-fold cross-validation, incorporating GDF15 and clinical variables including age, sex, comorbidities (hypertension, diabetes, dyslipidemia, congestive heart failure, coronary artery disease, and previous stroke or transient ischemic attack), smoking history, and cardioprotective medication use. The model’s primary evaluation metric was the F1 score, a validated measurement of the harmonic mean of the precision and recall values of the prediction model. Secondary model performance metrics included precision, recall, positive likelihood ratio (LR+), and negative likelihood ratio (LR-). A prediction probability histogram and Shapley additive explanations (SHAP) analysis were used to assess model discrimination and interpretability. The mean participant age was 70 ± SD 11 years, with 32% (n = 386) female representation. Median plasma GDF15 levels were significantly higher in PAD patients compared to the levels in non-PAD patients (1.29 [IQR 0.77–2.22] vs. 0.99 [IQR 0.61–1.63] pg/mL; p < 0.001). During the 2-year follow-up period, 219 individuals (18.4%) experienced MACE. The XGBoost model demonstrated strong predictive performance for 2-year MACE (F1 score = 0.83; precision = 82.0%; recall = 83.7%; LR+ = 1.88; LR− = 0.83). The prediction histogram revealed distinct stratification between those who did vs. did not experience 2-year MACE. SHAP analysis identified GDF15 as the most influential predictive feature, surpassing traditional clinical predictors such as age, cardiovascular history, and smoking status. This study highlights GDF15 as a strong prognostic biomarker for 2-year MACE in patients with PAD. When combined with clinical variables in an interpretable machine learning model, GDF15 supports the early identification of patients at high risk for systemic cardiovascular events, facilitating personalized treatment strategies including multidisciplinary specialist referrals and aggressive cardiovascular risk reduction therapy. This biomarker-guided approach offers a promising pathway for improving cardiovascular outcomes in the PAD population through precision risk stratification.

Background/Objectives: Patients with peripheral artery disease (PAD) have a heightened risk of major adverse cardiovascular events (MACE), including myocardial infarction, stroke, and death. Despite this, limited progress has been made in identifying reliable biomarkers to prognosticate such outcomes. Circulating growth factors, known to influence endothelial function and the progression of atherosclerosis, may hold prognostic value in this context. The objective of this research was to evaluate a broad range of blood-based growth factors to investigate their potential as predictors of MACE in patients diagnosed with PAD. Methods: A total of 465 patients with PAD were enrolled in a prospective cohort study. Baseline plasma levels of five different growth factors were measured, and participants were monitored over a two-year period. The primary outcome was the occurrence of MACE within those two years. Comparative analysis of protein levels between patients who did and did not experience MACE was performed using the Mann–Whitney U test. To assess the individual prognostic significance of each protein for predicting MACE within two years, Cox proportional hazards regression was performed, adjusting for clinical and demographic factors including a history of coronary and cerebrovascular disease. Subgroup analysis was performed to assess the prognostic value of these proteins in females, who may be at higher risk of PAD-related adverse events. Net reclassification improvement (NRI), integrated discrimination improvement (IDI), and area under the receiver operating characteristic curve (AUROC) were calculated to assess the added value of significant biomarkers to model performance for predicting 2-year MACE when compared to using demographic/clinical features alone. Kaplan–Meier curves stratified by IGFBP-1 tertiles compared using log-rank tests and Cox proportional hazards analysis were used to assess 2-year MACE risk trajectory based on plasma protein levels. Results: The average participant age was 71 years (SD 10); 31.1% were female and 47.2% had diabetes. By the end of the two-year follow-up, 18.1% (n = 84) had experienced MACE. Of all proteins studied, only insulin-like growth factor-binding protein 1 (IGFBP-1) showed a significant elevation among patients who suffered MACE versus those who remained event-free (20.66 [SD 3.91] vs. 13.94 [SD 3.80] pg/mL; p = 0.012). IGFBP-1 remained a significant independent predictor of 2-year MACE occurrence in the multivariable Cox analysis (adjusted hazard ratio [HR] 1.57, 95% CI 1.21–1.97; p = 0.012). Subgroup analyses revealed that IGFBP-1 was significantly associated with 2-year MACE occurrence in both females (adjusted HR 1.52, 95% CI 1.16–1.97; p = 0.015) and males (adjusted HR 1.04, 95% CI 1.02–1.22; p = 0.045). Incorporating IGFBP-1 into the clinical risk prediction model significantly enhanced its predictive performance, with an increase in the AUROC from 0.73 (95% CI 0.71–0.75) to 0.79 (95% CI 0.77–0.81; p = 0.01), an NRI of 0.21 (95% CI 0.07–0.36; p = 0.014), and an IDI of 0.041 (95% CI 0.015–0.066; p = 0.008), highlighting the prognostic value of IGFBP-1. Kaplan–Meier analysis showed an increase in the cumulative incidence of 2-year MACE across IGFBP-1 tertiles. Patients in the highest IGFBP-1 tertile experienced a significantly higher event rate compared to those in the lowest tertile (log-rank p = 0.008). In the Cox proportional hazards analysis, the highest tertile of IGFBP-1 was associated with increased 2-year MACE risk compared to the lowest tertile (adjusted HR 1.81; 95% CI: 1.31–2.65; p = 0.001). Conclusions: Among the growth factors analyzed, IGFBP-1 emerged as the sole biomarker independently linked to the development of MACE over a two-year span in both female and male PAD patients. The addition of IGFBP-1 to clinical features significantly improved model predictive performance for 2-year MACE. Measuring IGFBP-1 levels may enhance risk stratification and guide the intensity of therapeutic interventions and referrals to cardiovascular specialists, ultimately supporting more personalized and effective management strategies for patients with PAD to reduce systemic vascular risk.

Background/Objectives: Major adverse cardiovascular events (MACE)—including heart attacks and strokes—are the leading cause of death in patients with peripheral artery disease (PAD), yet biomarker research for MACE prediction in PAD patients remains limited. Inflammatory proteins play a key role in the progression of atherosclerosis and may serve as useful prognostic indicators for systemic cardiovascular risk in PAD. The objective of this study was to evaluate a broad panel of circulating inflammatory proteins to identify those independently associated with 2-year MACE in patients with PAD. Methods: We conducted a prospective cohort study involving 465 patients with PAD. Plasma concentrations of 15 inflammatory proteins were measured at baseline using validated immunoassays. Patients were followed over a two-year period for the development of MACE, defined as a composite endpoint of myocardial infarction, stroke, or mortality. Protein levels were compared between patients with and without MACE using the Mann–Whitney U test. Cox proportional hazards regression was used to determine the independent association of each protein with MACE after adjusting for baseline demographic and clinical variables, including existing coronary and cerebrovascular disease. To validate the findings, a random forest machine learning model was developed to assess the relative importance of each protein for predicting 2-year MACE. Results: The mean age of the cohort was 71 years (SD 10), and 145 participants (31.1%) were female. Over the two-year follow-up, 84 patients (18.1%) experienced MACE. Six proteins were significantly elevated in PAD patients who developed MACE: interferon gamma (IFN-γ; 42.55 [SD 15.11] vs. 33.85 [SD 12.46] pg/mL, p < 0.001), tumor necrosis factor alpha (TNF-α; 9.00 [SD 5.00] vs. 4.65 [SD 4.29] pg/mL, p < 0.001), chemokine (C-X-C motif) ligand 9 (CXCL9; 75.99 [SD 65.14] vs. 5.38 [SD 64.18] pg/mL, p = 0.002), macrophage inflammatory protein-1 beta (MIP-1β; 20.88 [SD 18.10] vs. 15.67 [SD 16.93] pg/mL, p = 0.009), MIP-1δ (25.29 [SD 4.22] vs. 17.98 [SD 4.01] pg/mL, p = 0.026), and interleukin-6 (IL-6; 12.50 [SD 40.00] vs. 6.72 [SD 38.98] pg/mL, p = 0.035). After adjusting for all baseline covariates, only two proteins—TNF-α (adjusted HR 1.66, 95% CI 1.28–2.33, p = 0.001) and IFN-γ (adjusted HR 1.25, 95% CI 1.12–2.29, p = 0.033)—remained significantly and independently associated with 2-year MACE. These findings were corroborated by the random forest model, where TNF-α and IFN-γ received the highest importance scores for predicting 2-year MACE: (TNF-α: 0.15 [95% CI 0.13–0.18], p = 0.002; IFN-γ: 0.19 [95% CI 0.17–0.21], p = 0.001). Conclusions: From a panel of 15 proteins, TNF-α and IFN-γ emerged as inflammatory biomarkers associated with 2-year MACE in PAD patients. Their measurement may aid in cardiovascular risk stratification, helping to identify high-risk individuals who could benefit from early multidisciplinary referrals to cardiology, neurology, and/or vascular medicine specialists to provide intensified medical therapy. Incorporating these biomarkers into PAD management may improve systemic cardiovascular outcomes through more personalized and targeted treatment approaches.

Background/Objectives: Peripheral artery disease (PAD) is associated with an increased risk of major adverse cardiovascular events (MACE), such as myocardial infarction and stroke, which are the top mortality causes in the PAD population. However, the identification of reliable biomarkers for predicting MACE in PAD patients remains limited. Proteins involved in extracellular matrix (ECM) remodeling have been implicated in atherosclerosis and may serve as potential indicators of cardiovascular risk. This study aimed to evaluate a panel of circulating proteins involved in ECM remodeling to identify those predictive of 2-year MACE in individuals with PAD. Methods: A prospective cohort of 465 PAD patients was enrolled and followed for 24 months. At baseline, plasma levels of nine ECM-related proteins were quantified. The outcome of interest was a 2-year MACE, defined as a composite of myocardial infarction, stroke, or mortality. Protein level differences between MACE vs. non-MACE patients were analyzed using Mann–Whitney U tests. Cox proportional hazards models, adjusted for baseline variables (including known cerebrovascular and coronary disease), were used to determine the independent associations between each protein and 2-year MACE. Subgroup analyses were conducted for diabetic and female patients, who are known to be at high risk for adverse events. Results: The mean age of the participants was 71 (SD 10) years, with 31.1% identifying as female and 47.2% having diabetes. Over two years, 84 patients (18.1%) experienced MACE. Among the proteins analyzed, matrix metalloproteinase-10 (MMP-10) and matrix metalloproteinase-7 (MMP-7) were significantly elevated in those who developed MACE compared to those who did not: MMP-10 (710.60 pg/mL [SD 46.09] vs. 672.40 pg/mL [SD 45.04], p = 0.032) and MMP-7 (5.20 pg/mL [SD 4.11] vs. 4.76 pg/mL [SD 3.86], p = 0.048). Both independently correlated with 2-year MACE after adjustment for all baseline factors: MMP-10 (HR 1.32, 95% CI 1.16–1.51, p = 0.023) and MMP-7 (HR 1.17, 95% CI 1.05–2.68, p = 0.026). Subgroup analyses revealed that MMP-10 was associated with MACE in diabetic patients (HR 1.18, 95% CI 1.13–1.53, p = 0.019), while MMP-7 was associated with MACE among females (HR 1.31, 95% CI 1.15–1.69, p = 0.009). Conclusions: MMP-10 and MMP-7 emerged as independent biomarkers for prognosticating 2-year MACE in PAD patients, suggesting their utility in systemic cardiovascular risk stratification. Measuring these proteins could enhance clinical decision-making by identifying high-risk individuals with PAD who may benefit from multidisciplinary vascular evaluation and intensified treatment strategies, ultimately aiming to reduce cardiovascular complications in the PAD population.

Background/Objectives: Major adverse cardiovascular events (MACE) are the primary cause of mortality among individuals with peripheral artery disease (PAD). Despite this, there is limited research on biomarkers that can predict MACE risk in this population. Proteins involved in angiogenesis are integral to both systemic circulation and the development of atherosclerosis, indicating their potential as prognostic markers. This study aimed to identify angiogenesis-related proteins associated with MACE risk in PAD patients. Methods: We conducted a prospective cohort study involving 250 patients diagnosed with PAD. At baseline, plasma levels of 17 angiogenesis-related proteins were measured. Participants were followed for two years, with the primary outcome being the incidence of MACE—a composite of stroke, myocardial infarction, or death. Protein concentrations were compared between those who experienced 2-year MACE and those who did not using the Mann–Whitney U test. Proteins showing significant differences were further analyzed using Cox proportional hazards modeling to assess their independent associations with MACE, adjusting for baseline demographic and clinical variables, including prior coronary and cerebrovascular disease. Kaplan–Meier survival analysis was also employed to compare MACE-free survival based on protein concentration levels. Results: The average age of participants was 69 years (SD 9), with 32% (n = 80) being female. Over the two-year follow-up, 48 patients (19.8%) experienced MACE. Among the proteins assessed, only hepatocyte growth factor (HGF) and angiopoietin-2 were significantly elevated in patients who developed MACE (HGF: 390.83 [SD 319.16] vs. 300.55 [SD 177.53] pg/mL, p < 0.001; angiopoietin-2: 23.67 [SD 17.60] vs. 19.36 [SD 12.06] pg/mL, p = 0.020). Multivariable Cox analysis confirmed that elevated levels of both HGF (adjusted HR 1.37; 95% CI 1.14–1.64; p = 0.001) and angiopoietin-2 (adjusted HR 1.27; 95% CI 1.04–1.55; p = 0.016) were independently associated with increased 2-year MACE risk. Kaplan–Meier curves demonstrated significantly reduced MACE-free survival in patients with higher levels of HGF and angiopoietin-2. Conclusions: HGF and angiopoietin-2 emerged as significant, independent predictors of 2-year MACE in patients with PAD. Measuring plasma levels of these proteins may enhance risk stratification, guiding referrals to appropriate cardiovascular specialists and informing the intensity of medical management. This biomarker-based precision medicine approach holds potential for improving cardiovascular outcomes in the PAD population.

Background: Peripheral arterial disease (PAD) is known to be strongly linked to major adverse limb events, ultimately leading to an increased risk of limb-threatening conditions. We developed a predictive model using five identified biomarkers to predict major adverse limb events, limb loss, diabetic (DM) foot ulcers, and vascular intervention in patients with underlying PAD and DM over 2 years. Methods: A single-center prospective case control study with was conducted with 2 years’ follow up. In the discovery phase the cohort was randomly split into a 70:30 ratio, and proteins with a higher mean level of expression in the DM PAD group compared to the DM non-PAD group were identified. Next, a random forest model was trained using (1) clinical characteristics, (2) a five-protein panel, and (3) clinical characteristics combined with the five-protein panel. Demographic data were analyzed by independent t-test and chi-square test. The importance of predictive features was calculated using the variable importance (gain) score. The model was used and assessed for its ability to diagnose PAD, predict limb loss, predict major adverse limb events (MALEs), predict diabetic foot ulcers, and predict the need for vascular surgery. The model was evaluated using area under the receiver operating characteristic curve and net reclassification index. Results: The cohort of 392 patients was matched for age, sex, and comorbidities. Five proteins were identified (TNFa: tumor necrosis factor alpha, BMP-10: bone morphogenic protein 10, CCL15/MIP1 delta: chemokine (c-c motif) ligand 15/macrophage inflammatory protein 1 delta, MMP-10: matrix metalloprotease 10, and HTRA2/Omi: HTRA2, also known as Omi) as having a significantly higher level of expression in the DM PAD group. HTRA/Omi had the highest contribution to the model’s ability to diagnose PAD in diabetic patients. Model performance was best when combined with clinical characteristics to predict limb loss (AUROC 0.86, 0.76, 0.80), foot ulcer (AUROC 0.87, 0.82, 0.67), MALE (AUROC 0.81, 0.78, 0.67), and the need for vascular surgery (AUROC 0.82, 0.81, 0.61). Conclusions: In this study, we describe a biomarker panel that can be used in combination with clinical characteristics to create an accurate prediction model for diagnosis and prognostication of PAD in the setting of DM.

Background/Objectives: The leading cause of death for people with peripheral artery disease (PAD) is major adverse cardiovascular events (MACE), including heart attacks and strokes. However, research into biomarkers that could help predict MACE in patients with PAD has been limited. Immunomodulatory proteins are known to significantly influence systemic atherosclerosis, suggesting they could be useful prognostic indicators for MACE in patients with PAD. In this study, we evaluated a broad panel of immunomodulatory proteins to identify those linked to MACE in individuals with PAD. Methods: We conducted a prognostic study involving a prospectively recruited cohort of 406 patients consisting of 254 with PAD and 152 without PAD. At the baseline, we measured the plasma concentrations of 17 circulating immunomodulatory proteins and followed the cohort for two years. The primary outcome was 2-year MACE, a composite of myocardial infarction, stroke, or death. Plasma protein concentrations were compared between patients with PAD with and without 2-year MACE using Mann–Whitney U tests. We further examined the prognostic potential of differentially expressed proteins through a Cox proportional hazards analysis, determining their independent associations with 2-year MACE while controlling for all the baseline demographic and clinical characteristics, including the existing coronary artery and cerebrovascular diseases. Additionally, A Kaplan–Meier analysis was performed to evaluate the 2-year freedom from MACE in patients with low versus high levels of the differentially expressed proteins based on the median plasma concentrations. Results: The mean age of the cohort was 68.8 years (SD 11.1), with 134 patients (33%) being female. During the two-year follow-up, 63 individuals (16%) developed MACE. The following proteins were significantly elevated in patients with PAD who experienced 2-year MACE compared to those who did not: galectin-1 (0.17 [SD 0.06] vs. 0.10 [SD 0.07] pg/mL, p = 0.012), alpha-1-microglobulin (16.68 [SD 7.48] vs. 14.74 [SD 6.71] pg/mL, p = 0.019), and galectin-9 (0.14 [SD 0.09] vs. 0.09 [SD 0.05] pg/mL, p = 0.033). The Cox proportional hazards analysis indicated that these three proteins were independently associated with 2-year MACE after adjusting for all the baseline demographic and clinical factors: galectin-1 (HR 1.45 [95% CI 1.09–1.92], p = 0.019), alpha-1-microglobulin (HR 1.31 [95% CI 1.06–1.63], p = 0.013), and galectin-9 (HR 1.35 [95% CI 1.02–1.78], p = 0.028). Over the two-year follow-up, patients with higher levels of galectin-1, galectin-9, and alpha-1-microglobulin had a lower freedom from MACE. Additional analysis showed that these three proteins were not significantly associated with 2-year MACE in patients without PAD. Conclusions: Among the 17 immunomodulatory proteins evaluated, galectin-1, galectin-9, and alpha-1-microglobulin were found to be independently and specifically associated with 2-year MACE in patients with PAD. Assessing the plasma concentrations of these proteins can aid in risk stratification for MACE in patients with PAD, helping to inform clinical decisions regarding multidisciplinary referrals to cardiologists, neurologists, and vascular medicine specialists. This information can also guide the aggressiveness of medical management, ultimately improving cardiovascular outcomes for patients with PAD.