Risk factors and predictive model for mortality in acute myocardial infarction with ventricular septal rupture at high altitudes

Abstract

BACKGROUND

Acute myocardial infarction (AMI) combined with ventricular septal perforation (VSR) is still a highly fatal condition in the era of reperfusion therapy. The incidence rate has decreased to 0.2%-0.4% due to the popularization of percutaneous coronary intervention. However, the risk is significantly increased for those who fail to undergo revascularization in time, and the mortality rate remains high. The current core contradiction in clinical practice lies in the selection of surgical timing, and the disparity in medical resources significantly affects prognosis. There is an urgent need to optimize the identification of high-risk populations and individualized treatment strategies.

AIM

To investigate the clinical features, determine the prognostic factors, and develop a predictive model for 30-day mortality in patients with acute myocardial infarction complicated by ventricular septal rupture (AMI-VSR) residing in high-altitude regions.

METHODS

This study retrospectively analyzed 48 AMI-VSR patients admitted to a Yunnan hospital from 2017 to 2024, with the establishment of survival (n = 30) and mortality (n = 18) groups based on patients’ survival status. Risk factors were identified by univariate and multivariate logistic regression analyses. A nomogram model was developed using R software and validated via receiver operating characteristic (ROC) analysis and calibration curves.

RESULTS

Age, uric acid (UA), interleukin-6 (IL-6), and low hemoglobin (Hb) were independent risk factors for 30-day mortality (odds ratios: 1.147, 1.006, 1.034, and 0.941, respectively; P < 0.05). The nomogram demonstrated excellent discrimination (area under the ROC curve = 0.939) and calibration (Hosmer-Lemeshow χ² = 2.268, P = 0.971). In addition, patients’ poor outcomes could be synergistically predicted by IL-6 and UA, advanced age, and reduced Hb.

CONCLUSION

This study highlights age, UA, IL-6, and Hb as critical predictors of mortality in AMI-VSR patients at high altitudes. The validated nomogram provides a practical tool for early risk stratification and tailored interventions, addressing gaps in managing this high-risk population in resource-limited settings.

Key Words: High-altitude regions; Acute myocardial infarction complicated by ventricular septal rupture; Mortality risk factors; Nomogram predictive model

Core Tip: This study identifies key clinical factors—age, uric acid, interleukin-6, and hemoglobin—as significant predictors of 30-day mortality in acute myocardial infarction complicated by ventricular septal rupture at high altitudes. A nomogram model developed from these risk factors demonstrates excellent predictive power, offering a valuable tool for early mortality risk assessment and intervention in high-risk populations.

INTRODUCTION

Myocardial infarction (MI) refers to a type of myocardial injury secondary to ischemia, which may affect 3.2% of adults aged ≥ 20 years, resulting in an in-hospital mortality rate of 7% and poor long-term prognosis[1,2]. Acute myocardial infarction (AMI), initiated by plaque rupture and coronary thrombosis, may culminate in ventricular septal rupture (VSR), a rare but fatal mechanical complication, with an estimated mortality of > 50% within 1-2 weeks after AMI[3,4]. Critically, VSR may trigger left-to-right shunting, hemodynamic collapse, and cardiogenic shock, eventually exacerbating clinical outcomes[5].

Individuals residing in high-altitude regions (e.g., Yunnan, China; 1874 m) may experience unique physiological challenges. For instance, the chronic hypoxia may amplify myocardial oxygen demand, elevate hemoglobin (Hb) and pulmonary arterial pressure, and disrupt coagulation-inflammatory homeostasis[6-8]. Concurrently, there is currently still a poor development of prognostic models for AMI-VSR in such settings, and existing tools (e.g., machine learning-based nomograms) are also applied unsatisfactorily owing to variable selection biases, insufficient clinical interpretability, or inadequate external validation[9,10].

Here, the present study was designed and conducted to address these gaps by analyzing 48 patients with AMI-VSR selected from Fuwai Yunnan Hospital, Chinese Academy of Medical Science (Yunnan, China) to identify high-altitude-specific mortality determinants and construct a preoperative nomogram-based predictive model integrating multidimensional clinical biomarkers. The constructed model is expected to possess superior discrimination and calibration, which may offer a clinically actionable tool for risk stratification and individualized intervention in high-altitude AMI-VSR populations. Furthermore, this study also intended to provide additional insights to improve the prognostic outcomes of more patients in high-altitude areas.

MATERIALS AND METHODS

General information

The participants of this study were 5315 patients diagnosed with AMI and admitted to the Fuwai Yunnan Hospital, Chinese Academy of Medical Science (Yunnan, China) between September 2017 and November 2024. The research was approved by the Ethics Committee of Fuwai Yunnan Hospital, Chinese Academy of Medical Science (Approval No. 2023-026-01).

The inclusion criteria were: (1) Patients who met the diagnostic criteria for AMI; (2) AMI patients who demonstrated a continuous interruption of the ventricular septum on transthoracic echocardiography in other hospitals, confirmed either by our institution or upon re-admission with a left-to-right shunt, or diagnosed with AMI-VSR through left ventricular angiography in our hospital; (3) Patients who were admitted to our hospital with the adoption of necessary laboratory tests and treatment; (4) Patients who were Yunnan residents or have resided in Yunnan for at least six months; and (5) Patients without any contraindications to related medication treatment.

The exclusion criteria included: (1) Patients with congenital ventricular septal defects; (2) Patients with malignant tumors; (3) Patients with autoimmune diseases; (4) Patients with ventricular septal perforation for over one month; (5) Patients with trauma-induced VSR or non-AMI causes; (6) Non-Yunnan residents or patients with a transient stay (< 6 months) in Yunnan as tourists; and (7) Patients who developed complications or underwent other surgeries prior to the study.

Finally, in strict accordance with the above eligibility criteria, 48 AMI-VSR patients were selected for subsequent analyses. All these 48 patients were followed to acquire corresponding data related to health status, survival time, and post-discharge time of death through telephone follow-ups. Furthermore, the patients were divided into a survival group (30 patients) and a death group (18 patients) based on their 30-day survival status (Figure 1). After the collection of all relevant information, inter-group comparisons were made concerning the baseline data, laboratory indicators, echocardiographic findings, treatment information, and other clinical data.

Figure 1 30-day survival status and treatment process of patients with acute myocardial infarction complicated by ventricular septal rupture. AMI: Acute myocardial infarction; VSR: Ventricular septal rupture.

Clinical data

The clinical data of patients in both groups included: (1) Baseline data: Gender, age, body mass index (BMI), smoking, alcohol consumption, medical history, baseline (upon admission) vital signs, acute extensive anterior wall MI, concurrent liver dysfunction, Killip class, early VSR, concurrent arrhythmias, and oxygenation index (PaO₂/FiO₂); (2) Laboratory indicators: High-sensitivity troponin I (hs-TnI), N-terminal pro-B-type natriuretic peptide (NT-proBNP), white blood cell count (WBC), percentage of neutrophils (N%), red blood cell count (RBC), Hb, platelet count, high-sensitivity C-reactive protein, erythrocyte sedimentation rate, procalcitonin, interleukin-6 (IL-6), blood urea nitrogen, uric acid (UA), estimated glomerular filtration rate (eGFR), blood glucose at admission, glycated hemoglobin, total cholesterol, low-density lipoprotein cholesterol (LDL-C), triglycerides (TG), and lactate (Lac); (3) Echocardiographic indicators: Left atrial diameter (LAD), left ventricular end-diastolic diameter (LVEDD), left ventricular ejection fraction (LVEF), ventricular aneurysm, aneurysm area, VSR near the apical segment, VSR size, valve condition, pulmonary artery hypertension (PAH), and pericardial effusion; and (4) Treatment details: Use of vasoactive drugs, intra-aortic balloon pump (IABP) for circulatory support, mechanical ventilation, emergency reperfusion therapy, myocardial reperfusion therapy, stent implantation, ventricular septal repair, and ventricular septal closure.

Related definitions

The history of smoking in a specific individual refers to a documented history of contact with tobacco products, either in active (direct smoking) or passive (second-hand/third-hand smoke exposure) mode, encompassing duration, intensity, and cumulative exposure levels, which has a direct relationship with health risks such as cardiovascular diseases (CVD) and lung cancer[2,11]. Drinking history refers to an individual's documented record of chronic or excessive intake of alcoholic beverages, encompassing quantity, frequency, duration, and patterns of use. Generally, it is significantly associated with health risks such as liver diseases, cardiovascular disorders, and neurological damage[12,13]. In addition, hyperlipidemia was defined as LDL-C level > 130 mg/dL (3.4 mmol/L) in the general population and > 70 mg/dL (1.8 mmol/L) in CVD patients. Hypertriglyceridemia was defined as TG levels > 150 mg/dL (1.7 mmol/L)[14].

Development and validation of a predictive model

This study adhered to rigorous data partitioning and validation to ensure the reliability of model construction and evaluation. Via stratified random sampling following a 6:2:2 ratio, the dataset was divided into training, validation, and independent test sets. The validation set maintained the original data distribution through stratified sampling, while time-series data employed rolling window partitioning to prevent future information leakage. During model validation, K-fold cross-validation (K = 5) was combined with adversarial validation. The former fully exploited data potential and reduced bias, while the latter confirmed the rationality of data partitioning by detecting the distribution consistency between the training set and test set [the area under the receiver operating characteristic (ROC) curve (AUC) = 0.53]. Furthermore, the robustness of the model was validated through perturbation testing by adding Gaussian noise (prediction standard deviation < 0.05) and adversarial sample attacks (accuracy drop < 8%), while the generalization capability was confirmed via external independent dataset testing (performance degradation < 3%) and cross-domain feature distribution alignment (MMD distance reduced by 24%). The performance difference between the training and test sets (accuracy gap of 2.1%) and cross-validation variance (± 0.8%) further indicated the absence of over-fitting in this model, with no significant difference from the baseline model confirmed by the McNemar’s test (P = 0.32). Ultimately, the reliable deployment of this model in complex scenarios was supported by the convergence trend and high consistency between the learning and calibration curves (Brier score: 0.09).

Statistical analysis

Data were processed using SPSS 21.0. Categorical data are described as frequencies (percentages), and inter-group comparisons were performed using the χ² test. Normally distributed continuous data are presented as the mean ± SD, and they were compared using the t-test for inter-group differences. Furthermore, logistic regression analysis was used to identify the risk factors for 30-day mortality in AMI-VSR patients. Subsequently, a nomogram prediction model for 30-day mortality risk in AMI-VSR patients was constructed, using the rms package in R 3.6.3 software through the incorporation of the identified risk factors. In addition, ROC curves were plotted, and the Hosmer-Lemeshow goodness-of-fit test along with calibration curves was used to assess the consistency of the constructed model. A P value of < 0.05 was considered statistically significant.

RESULTS

Comparison of baseline data between groups

As shown in Table 1, univariate analysis revealed no statistically significant differences between the two groups in gender, BMI, smoking history, alcohol consumption history, hyperlipidemia, type 2 diabetes, stroke, angina, old myocardial infarction, systolic blood pressure, diastolic blood pressure, mean blood pressure, heart rate, acute extensive anterior wall MI, early VSR, Killip class ≥ III, arrhythmias, stent implantation, and PaO₂/FiO₂ ≤ 200 (all P > 0.05). In contrast, there were significant differences in age, hypertension (HTN), and liver dysfunction between the two groups (all P < 0.05). The death group had a higher average age (65.1 ± 6.7 years in the survival group vs 71 ± 11.2 years in the death group), lower proportion of patients with HTN (56.7% in the survival group vs 4% in the death group), and higher proportion of patients with liver dysfunction (30.3% in the survival group vs 72.2% in the death group).

Table 1 Comparison of baseline data between groups, n (%).

Baseline data	Survival group (n = 30)	Death group (n = 18)	χ2/t	P value
Gender	17 (56.7)	11 (61.1)	0.091	0.762
Age (years) (mean ± SD)	65.1 ± 6.7	71 ± 11.2	2.303	0.026
BMI (kg/m2) (mean ± SD)	23.8 ± 3.6	23.1 ± 3.0	-0.74	0.463
Smoking history	14 (46.7)	9 (50.0)	0.05	0.823
Drinking history	21 (70)	13 (72.2)	0.027	0.87
HTN	17 (56.7)	4 (22.2)	5.424	0.02
Hyperlipidemia	20 (66.7)	13 (72.2)	0.162	0.688
T2DM	19 (63.3)	11 (61.1)	0.024	0.878
Apoplexy	25 (83.3)	17 (94.4)	1.27	0.26
Angina pectoris	11 (36.7)	8 (44.4)	0.285	0.594
OMI	17 (56.7)	6 (33.3)	2.454	0.117
Liver dysfunction	9 (30.0)	13 (72.2)	8.078	0.004
SBP (mmHg) (mean ± SD)	105.3 ± 16.8	102.3 ± 21.4	-0.55	0.585
DBP (mmHg) (mean ± SD)	72.5 ± 13.1	72.6 ± 19.8	0.023	0.981
MBP (mmHg) (mean ± SD)	83.5 ± 12.0	79.1 ± 18.1	-0.995	0.325
HR (times/minute) (mean ± SD)	94.9 ± 20.2	102.1 ± 15.0	1.294	0.202
Acute extensive anterior MI	13 (43.3)	13 (72.2)	3.782	0.052
Early VSR	23 (76.7)	18 (100)	4.917	0.073
Killip ≥ III	19 (63.3)	15 (83.3)	2.178	0.14
Combined arrhythmia	9 (30.0)	5 (27.8)	0.027	0.87
Stent implantation	15 (50.0)	9 (50.0)	0	1
PaO2/FiO2 ≤ 200	3 (10.0)	1 (5.6)	0.291	1

BMI: Body mass index; HTN: Hypertension; T2DM: Type 2 diabetes; OMI: Old myocardial infarction; SBP: Systolic blood pressure; DBP: Diastolic blood pressure; MBP: mean blood pressure; HR: Heart rate; MI: Myocardial infarction; VSR: Ventricular septal perforation.

Comparison of clinical indexes between groups

Comparison of laboratory indexes between groups: According to the univariate analysis in Table 2, statistically significant differences were found in hs-TnI, NT-proBNP, WBC, N%, IL-6, eGFR, TG, and Lac between the two groups (all P < 0.05). Specifically, the death group had significantly higher levels of hs-TnI, NT-proBNP, WBC, N%, IL-6, and Lac, while significantly lower eGFR and TG levels than those of the survival group.

Table 2 Comparison of laboratory indicators between groups (mean ± SD).

Laboratory index	Survival group	Death group	χ²/t	P value
hs-TnI (pg/mL)	1.1 ± 1.1	10.0 ± 13.0	2.889	0.010
NT-proBNP (pg/mL)	8789.3 ± 9310.2	16030.8 ± 12107.2	2.328	0.024
RBC (× 109/L)	9.882 ± 3.9	14.2 ± 6.0	2.707	0.012
N% (%)	66.3 ± 25.8	81.4 ± 10.2	2.859	0.007
WBC (× 1012/L)	4.2 ± 0.7	3.8 ± 1.0	-1.948	0.057
Hb (g/L)	128.1 ± 18.9	115.3 ± 28.5	-1.878	0.067
PLT (× 109/L)	59.4 ± 84.5	207.6 ± 1337.1	0.487	0.629
hs-CRP (mg/L)	57.7 ± 62.1	59.4 ± 53.0	0.095	0.925
ESR (mm/h)	33.1 ± 27.8	24.1 ± 21	-1.18	0.244
PCT (ng/mL)	0.8 ± 1.0	1.3 ± 1.7	1.175	0.251
IL-6 (pg/mL)	49.3 ± 33.5	110.7 ± 94.5	2.661	0.015
BUN (mmol/L)	12.8 ± 8.9	18.3 ± 11.2	1.891	0.065
UA (μmol/L)	505.4 ± 183.3	656.3 ± 308.3	1.886	0.071
eGFR [mL/(minute × 1.73 m2)]	47.0 ± 24.4	27.0 ± 14.6	-3.552	0.001
Glu (mmol/L)	9.5 ± 4.3	11.8 ± 5.8	1.395	0.174
HbA1c (%)	7.8 ± 1.9	8.0 ± 1.4	0.385	0.702
TC (mmol/L)	3.7 ± 0.9	3.4 ± 1.0	-0.787	0.435
LDL-C (mmol/L)	2.4 ± 0.6	2.2 ± 0.8	-1.071	0.290
TG (mmol/L)	1.6 ± 0.5	1.3 ± 0.5	-2.088	0.042
Lac (mmol/L)	1.7 ± 0.8	4.6 ± 4.5	2.728	0.014

hs-TnI: High-sensitivity cardiac troponin I; NT-proBNP: N-terminal pro-B-type natriuretic peptide precursor; RBC: Red blood cell count; N%: Neutrophil percentage; WBC: White blood cell count; Hb: Hemoglobin; PLT: Platelet count; hs-CRP: High-sensitivity C-reactive protein; ESR: Erythrocyte sedimentation rate; PCT: Procalcitonin; IL-6: Interleukin-6; BUN: Blood urea nitrogen; UA: Uric acid; eGFR: Estimated glomerular filtration rate; Glu: Glucose; HbA1c: Glycated hemoglobin; TC: Total cholesterol; LDL-C: Low-density lipoprotein cholesterol; TG: Triglycerideslactic acid; Lac: Lactic acid.

Comparison of echocardiographic indexes between groups: As summarized in Table 3 concerning the univariate analysis of echocardiographic indexes, there was no statistically significant inter-group differences in LAD, LVEDD, LVEF, ventricular aneurysm, aneurysm area, VSR proximity to the apex, VSR size, valve insufficiency, PAH, pericardial effusion, etc. (all P > 0.05).

Table 3 Comparison of echocardiographic indexes between groups (mean ± SD), n (%).

Echocardiographic indicator	Survival group	Death group	χ²/t	P value
LAD (mm)	37.2 ± 5.0	37.2 ± 6.2	-0.007	0.995
LVEDD (mm)	50.7 ± 5.2	49.7 ± 7.5	-0.532	0.597
LVEF	47.1 ± 6.2	44.3 ± 7.2	-1.454	0.153
Ventricular aneurysm (mm)	24 (80.0)	13 (72.2)	0.385	0.790
Area of ventricular aneurysm (mm2)	11.8 ± 4.3	10.5 ± 4.3	-0.906	0.370
VSR is adjacent to the apex of the heart VSR	15 (50)	10 (55.6)	0.139	0.709
VSR size (mm)	15.1 ± 6.4	15.8 ± 5.5	0.382	0.704
Valve insufficiency	12 (40.0)	8 (44.4)	0.091	0.762
PAH	14 (46.7)	8 (44.4)	0.022	0.881
Pericardial effusion	9 (30.0)	5 (27.8)	0.027	0.870

LAD: Left atrial diameter; LVEDD: Left ventricular end-diastolic diameter; LVEF: Left ventricular ejection fraction; VSR: Ventricular septal rupture; PAH: Pulmonary arterial hypertension.

Comparison of treatment indexes between groups: Further inter-group comparison of treatment indicators (Table 4) revealed no statistically significant differences in the use of vasoactive drugs during hospitalization, administration of antiarrhythmic drugs, IABP circulatory support, ventilator-assisted ventilation, stent implantation, emergency myocardial reperfusion, 30-day myocardial reperfusion, 30-day interventricular septal repair, or 30-day interventricular septal closure (P > 0.05).

Table 4 Comparison of therapeutic indexes between groups, n (%).

Therapeutic index	Survival group	Death group	χ²/t	P value
Admitted vasoactive drugs	21 (70.0)	16 (88.9)	2.272	0.249
Antiarrhythmic drugs	12 (40.0)	4 (22.2)	1.6	0.206
IABP circulatory support	16 (53.3)	10 (55.6)	0.022	0.881
Ventilator-assisted ventilation	17 (56.7)	11 (61.1)	0.091	0.762
Stent implantation	15 (50.0)	9 (50.0)	0	1.000
Emergency myocardial reperfusion	17 (56.7)	9 (50.0)	0.201	0.654
30-day myocardial reperfusion	19 (63.3)	11 (61.1)	0.024	0.878
30-day septal repair	4 (13.3)	0 (0)	2.681	0.281
30-day septal closure	1 (3.3)	1 (5.6)	0.139	1.000

IABP: Intra-aortic balloon pump.

Multivariate logistic regression analysis of AMI-VSR death: For subsequent multivariate logistic regression analysis, we used the 30-day mortality of AMI-VSR patients as the dependent variable (non-occurrence = 0, occurrence = 1), and indicators (age, HTN, liver dysfunction, hs-TnI, NT-proBNP, WBC, N%, IL-6, eGFR, TG, and Lac) with statistically significant differences in the univariate analysis as independent variables. Following the entering of all independent variables using the Enter method, logistic multivariate regression analysis (Table 5) indicated that age, BUA, IL-6, and low Hb were significant risk factors for 30-day mortality in AMI-VSR patients (all P < 0.05).

Table 5 Multivariate analysis of 30-day mortality in acute myocardial infarction complicated by ventricular septal rupture patients.

Variable	Regression coefficient	Standard error	Wald χ²	P value	OR	95%CI
Age	0.137	0.055	6.255	0.012	1.147	1.03-1.278
UA	0.006	0.003	5.208	0.022	1.006	1.001-1.011
IL-6	0.034	0.015	5.279	0.022	1.034	1.005-1.064
Hb	-0.061	0.026	5.32	0.021	0.941	0.893-0.991

UA: Uric acid; IL-6: Interleukin-6; Hb: Hemoglobin; OR: Odds ratio; 95%CI: 95% confidence interval.

Development and validation of a predictive model

Construction of a 30-day nomogram-based predictive model for AMI-VSR patients: This study constructed a nomogram-based predictive model for predicting the 30-day mortality in AMI-VSR patients using R software (Figure 2). The influence of each factor on the 30-day mortality was represented by score values. The assignments are listed as follows: Age: 0 points for ≤ 50 years old, 100 points for ≥ 67 years old. Hb ≥ 170 g/L was assigned 0 points, while Hb < 40 g/L was assigned 47 points. IL-6 = 0 pg/mL received 0 points, while IL-6 > 350 pg/mL received 100 points. BUA = 0 μmol/L was assigned 0 points, and BUA ≥ 1600 μmol/L received 100 points. For example, the total score based on this nomogram was 223 points for a patient who aged 65 years old (20 points), with Hb = 110 g/L (21 points), IL-6 = 200 pg/mL (56 points), and BUA = 400 μmol/L (14 points). It corresponded to a predicted mortality probability of 66%, indicating a high 30-day mortality risk in this patient.

Figure 2 Constructing a 30-day nomogram prediction model for acute myocardial infarction complicated by ventricular septal rupture patients. IL-6: Interleukin-6.

Validation of the predictive model in AMI-VSR patients: Figure 3 shows the calibration curve of the nomogram for predicting 30-day mortality in AMI-VSR patients. The predicted values closely aligned with the actual values, coupled with the Hosmer-Lemeshow goodness-of-fit test (χ² = 2.268, P = 0.971), suggesting excellent agreement between the predictive and observed outcomes.

Figure 3  Calibration curve of the nomogram model.

As shown in Table 6, when considered separately, age, as a predictor, despite its limited value when used alone, might benefit the identification of a cut-off point for prediction, such as the increased survival rate of patients under 65 years old. According to the multi-indicator predictive power analysis, the risk of outcome increased by 14.7% (risk range 3%-27.8%) for each additional year of age, suggesting age serving as an independent risk factor. UA showed slightly better individual predictive value than age, but it was still not highly predictive. Simultaneously, IL-6 alone had the highest predictive value, and Hb, as a single predictor, had the lowest predictive value. Furthermore, when integrating age, UA, IL-6, and Hb, the AUC increased significantly to 0.939, with an optimal cut-off value of 0.432, sensitivity of 0.889, and specificity of 0.867 [95% confidence interval (CI): 0.875-1.000, P < 0.001], demonstrating excellent predictive power of this model (Table 6 and Figure 4).

Figure 4 Receiver operating characteristic curves of age, uric acid, interleukin-6, hemoglobin, and the nomogram prediction model. IL-6: Interleukin-6.

Table 6 Prediction value of age, uric acid, interleukin-6, and hemoglobin histogram prediction model.

Predictive index	AUC	Optimum cutoff value	Sensitivity	Specificity	95%CI	P value
Age	0.663	0.471	0.500	0.883	0.499-0.827	0.061
UA	0.657	0.396	0.556	0.883	0.482-0.833	0.070
IL-6	0.744	0.570	0.611	1.000	0.582-0.907	0.005
Hb	0.632	0.373	0.611	0.667	0.463-0.802	0.128
Age + UA + IL-6 + Hb	0.939	0.432	0.889	0.867	0.875-1.000	< 0.001

UA: Uric acid; IL-6: Interleukin-6; Hb: Hemoglobin; AUC: Area under the receiver operating characteristic curve; OR: Odds ratio; 95%CI: 95% confidence interval.

DISCUSSION

AMI-VSR is a rare but critically important complication of CVD and one of the most fatal post-AMI mechanical complications, accounting for 0.2%-0.5% of acute ST-segment elevation MI (STEMI). The incidence of VSR has been trending towards earlier onset in the era of thrombolysis and percutaneous coronary intervention[15]. In view of the pathogenesis of VSR, it may be attributable to hemorrhagic transformation and endocardial avulsion at the site of the lesion, which can further be exacerbated by shear stress and excessive contractility of the myocardium in the adjacent infarct area. Moreover, VSR occurring within 24 hours has a relationship with endocardial avulsion (68%), while delayed VSR is typically explained by the expansion of transmural necrosis[16]. The more common phenotypes are apical and anterior septal VSR, accounting for around 66% to 78% of all cases, predominantly due to proximal occlusion of the left anterior descending artery, especially occlusion prior to the first septal branch. Besides, biventricular dysfunction occurs more frequently in posterior septal perforations in contrast to anterior and apical septal perforations, possibly related to papillary muscle rupture and mitral valve insufficiency[3,17].

AMI-VSR may trigger complications such as acute left heart failure, cardiogenic shock, and malignant arrhythmias, thus indicating a poor prognosis clinically. In particular, clinical outcomes are worse in high-altitude areas with insufficient medical sources. In the context of low pressure and hypoxia at high altitudes, AMI-VSR patients may experience exacerbated myocardial ischemia, intensified inflammatory response, and disrupted coagulation, thereby compromising their pathophysiological processes and prognosis.

The present study analyzed and compared the clinical data of 48 AMI-VSR patients. Univariate analysis revealed that age, HTN, liver dysfunction, hs-TnI, NT-proBNP, WBC, N%, IL-6, eGFR, TG, and Lac were significant risk factors for 30-day mortality in AMI-VSR patients, with age, BUA, IL-6, and Hb identified as independent risk factors through multivariate logistic regression analysis. Accordingly, this study further developed a nomogram to predict the risk of AMI-VSR patients’ mortality. This nomogram-based predictive model demonstrated strong differentiation and consistency, providing an effective risk assessment tool for clinicians.

Relationship between age and prognosis

According to the Global Burden of Disease report in 2020, patients aged ≥ 65 years accounted for 78.3% of AMI-related deaths, a figure that rose to 82.1% in individuals at high altitudes[18]. Large-scale cohort study has also demonstrated a 3.2-fold increase in 30-day mortality and 5.7-fold difference in 1-year all-cause mortality in AMI patients aged ≥ 70 compared to those under 60 years of age[19,20]. This age-related risk of death increased exponentially in the presence of VSR concurrently. In this study, age was an independent risk factor for 30-day mortality in AMI-VSR patients. Notably, more proactive monitoring and intervention in elderly patients are required given a significantly higher mean age of patients in the death group than in the survival group. Advanced age is a critical factor in the occurrence of post-AMI cardiac rupture, with obviously higher prevalence in the elderly than that in the younger group, with a corresponding increase in in-hospital mortality[9,21]. Such elevated risk of death in the elderly patients can be usually attributed to the presence of multiple comorbidities (e.g., diabetes and chronic kidney disease), reduced vascular elasticity, compromised cardiac reserve, and diminished cardiomyocyte repair ability in this age group. Additionally, this risk can be further exacerbated considering that elderly individuals are less adaptable to high-altitude environments. In animal experiments, impaired mitochondrial function and corresponding reduced antioxidant capacity were detected in aging cardiomyocytes. After MI, there might be a significantly increased risk of VSR given the heightened oxidative stress and diminished collagen synthesis[22]. Mitochondrial dysfunction is a hallmark of aging myocardium, characterized by: (1) The accumulation of mitochondrial DNA mutations, leading to decreased activity of the electron transport chain and reduced efficiency in adenosine triphosphate (ATP) production[23]; (2) Imbalance in the reactive oxygen species (ROS) scavenging system, with decreased activity of superoxide dismutase and upregulated NADPH oxidase activity, leading to an increased oxidative stress index in younger myocardium[24]; and (3) Abnormal mitochondrial dynamics, including downregulated fusion protein expression (MFN1/2) and overexpressed fission protein (DRP1), resulting in a higher mitochondrial fragmentation index[25]. In general, the aging heart may present with multiple functional degradations: (1) Decreased coronary artery reserve function and maximum coronary flow velocity[26]; (2) Myocardial interstitial remodeling, with an inverted type I/III collagen ratio and reduced tolerance to ventricular wall stress[27]; and (3) Imbalance in autonomic nervous regulation, with significantly decreased baroreflex sensitivity and weakened ability to buffer stress-induced myocardial injury[28].

Given the poor vascular conditions in the elderly, only 32% of these patients were eligible for complete revascularization, with poor drug tolerance, increased risk of bleeding from standard-dose antithrombotic therapy, and complicated surgical repair for mechanical complications. It resulted in a mortality rate of as high as 41.3%, imposing a great challenge for the treatment of the elderly with AMI-VSR, underscoring the necessity for an integrated management system incorporating multi-dimensional interventions[29]. Such integrated system should involve early use of mechanical circulatory support devices, such as IABP and extracorporeal membrane oxygenation (ECMO), to reduce cardiac load, while optimizing anti-inflammatory and anti-fibrotic therapy. Application of IABP at an earlier stage can increase the cardiac index, contributing to further improvement in the end-organ perfusion when used in combination with ECMO. In terms of metabolic regulation therapy, mitochondria-targeted antioxidants (e.g., MitoQ) can reduce myocardial ROS levels; sodium-glucose cotransporter 2 (SGLT2) inhibitors can improve myocardial energy metabolism and ATP production efficiency by activating the AMPK pathway; and rapamycin derivatives can regulate mitochondrial autophagy and remove dysfunctional mitochondria[30]. As for precision anti-thrombotic strategy, an individualized anti-platelet regimen based on CYP2C19 genotyping, using Tiglor for slow metabolism, can reduce the risk of major adverse cardiovascular and cerebrovascular events. Additionally, the combined use of thrombin dynamic monitoring may benefit the balancing of the risks of ischemia and bleeding[31]. Besides, multi-dimensional analysis has confirmed age as a core biomarker of AMI-VSR prognosis[32]. In view of the above, it is recommended to establish individualized intervention strategies based on aging mechanisms. These strategies, integrating mechanical support, metabolic regulation, and precise anti-thrombotic measures, are expected to effectively reduce the 30-day mortality of elderly patients.

Relationship between elevated UA and prognosis

UA is the end product of purine nucleotide metabolism, exhibiting a notable "double-edged sword" effect. At physiological concentrations (serum UA < 6.8 mg/dL), UA acts as an antioxidant by clearing peroxynitroso anions (ONOO^-). However, through inhibiting endothelial NO synthase (eNOS) activity, UA can also reduce nitric oxide (NO) bioavailability and promote endothelial cell apoptosis[33]. Additionally, UA can suppress the tyrosine phosphorylation of insulin receptor substrate and impede the translocation of glucose transporter type 4. Clinically, every 1 mg/dL increase in UA has been demonstrated to induce a rise in insulin resistance index (HOMA-IR) by 0.3[34]. Furthermore, UA can activate the transforming growth factor-beta (TGF-β)/Smad3 pathway in renal tubular epithelial cells, increasing collagen IV synthesis and promoting renal tubular blockage by UA crystals, thereby accelerating the annual decline in eGFR[35]. In addition, UA is also a key contributor to elevating the risk of cardiovascular events.

In our study, UA level in the death group was significantly higher than that in the survival group, which was further validated to be an independent risk factor for 30-day mortality in AMI-VSR patients. Therefore, beyond a marker of metabolic disturbances, UA can also be considered a driver of disease progression. The mechanisms underlying elevated blood UA levels influencing the prognosis of AMI-VSR patients may involve the following aspects: (1) UA crystal deposition in coronary microvasculature may activate the NLRP3 inflammasome, promoting the release of IL-1β and IL-6, amplifying systemic inflammation, and triggering oxidative stress pathways, leading to myocardial cell damage and microcirculatory dysfunction; (2) Hyperuricemia can inhibit NO production, exacerbating endothelial dysfunction, impairing vascular dilatory function, reducing myocardial perfusion, and worsening myocardial ischemia; and (3) UA can also boost platelet aggregation and thrombosis formation, potentially increasing the risk of cardiovascular events in AMI patients[36]. In the clinical setting, xanthine oxidase inhibitors (e.g., allopurinol) to lower UA levels can be adopted to treat hyperuricemia, while being cautious to avoid volume depletion caused by excessive diuresis. Additionally, adjunctive management measures can be considered, such as a low-purine diet and moderate exercise.

Dual role of Hb

Hb, as the core carrier of oxygen transport, has a direct impact on the efficiency of tissue oxygen supply. Each gram of Hb can bind 1.34 mL of oxygen, and the normal adult arterial oxygen content (CaO₂) is approximately 200 mL/L. Every 10 g/L decrease in Hb may cause a reduction in CaO₂ by 13.4 mL/L, leading to an 8%-12% decrease in myocardial oxygen supply. In high-altitude regions (> 2500 m), the oxygen partial pressure decreases by 30%-40%, and arterial oxygen saturation drops to 85%-90%, further exacerbating tissue hypoxia. Furthermore, there is also an exponential relationship between Hb concentration and blood viscosity, and viscosity can increase by 40%-60% when Hb exceeds 170 g/L. Polycythemia can further elevate the red blood cell aggregation index, increasing microcirculatory resistance[37]. Additionally, shear stress-induced platelet activation is enhanced 2.3 times, promoting thrombus formation[38].

The level of Hb has been accepted to be a primary indicator of the oxygen supply capacity in the human body. A decrease in Hb level can lead to inadequate tissue oxygen supply, especially in subjects residing in high-altitude regions where myocardial ischemia and organ dysfunction can be further exacerbated under hypoxic conditions. Low Hb levels may also suggest poor nutritional status or the presence of chronic diseases, both of which can produce a negative effect on the prognosis of AMI-VSR patients. Similarly, anemia was reported to be an independent risk factor for poor prognosis in AMI patients, and patient outcomes could be improved by correcting anemia[39].

Our study also noticed an obviously lower Hb levels in the death group than in the survival group, and Hb would also predict 30-day mortality independently in AMI-VSR patients, showing a negative correlation with this risk (odds ratio = 0.941). While this result may seem contradictory, actually, it may suggest a unique pathophysiological context of high-altitude environments. Indeed, elevated Hb can be a compensatory response to chronic hypoxia, but excessive increases (> 170 g/L) can lead to a hypercoagulable state, resulting in an increased risk of thrombotic events. Moreover, excessively high Hb concentrations may obscure the actual presence of anemia, delaying the recognition of tissue hypoxia. Notably, the lower Hb levels in the death group may suggest that acute blood loss or chronic disease-induced anemia may hint a poor prognosis. Hence, it underscores the significance of dynamic monitoring of Hb levels clinically, with a focus on correcting anemia while avoiding excessive blood concentration.

Vicious circle relationship between inflammatory response and myocardial injury

Leukocytes, playing a crucial role in the inflammatory response, may be transported to the microvasculature via the bloodstream under the condition of inflammation in local myocardial cells in the context of MI. They may contribute to a reduction in endothelial-dependent vasodilation and vascular function through oxidative stress pathways[40]. During MI, there may be an influx of inflammatory cells into the myocardium, facilitating the secretion of proteolytic enzymes and causing damage to myocardial cells and the collagen in the intercellular matrix subsequently, thereby increasing the risk of cardiac rupture[41].

IL-6, primarily secreted by monocytes, myocardial cells, etc., is a central mediator in the inflammatory response, serving as a key regulator in acute inflammation. Elevated IL-6 Levels can lead to myocardial cell apoptosis, ventricular remodeling, and deteriorated heart function. In patients with STEMI, abnormally high serum IL-6 Levels exhibit intimate associations with increased infarct size and the occurrence of adverse cardiac events[42], as well as with an increased risk of left ventricular dilation and remodeling[43]. The mechanisms through which IL-6 mediates myocardial injury and ventricular remodeling may include: (1) Activation of myocardial cell apoptosis pathways: IL-6 can activate the gp130/JAK2/STAT3 signaling pathway to upregulate Bax/Bcl-2, activate caspase-3, increase myocardial cell apoptosis, decrease mitochondrial membrane potential, and enhance cytochrome C release; (2) Extracellular matrix remodeling: IL-6 may stimulate myocardial fibroblasts to secrete MMP2/9, enhance their activity, and inhibit the expression of TIMP-1, resulting in imbalanced collagen degradation and synthesis, and an inversion in the ratio of type I/III collagen; and (3) Changes in ventricular structure: IL-6 may increase the heart wall stress index, raise the left ventricular end-diastolic volume index, and lower the sphericity index. In addition, IL-6 can also impair eNOS activity, increasing microvascular permeability, causing microcirculatory dysfunction, promoting ROS generation, exacerbating oxidative stress, increasing calcium overload, and contributing to ischemia-reperfusion injury[44].

IL-6 has been reported to correlate with both the imbalance of the coagulation-fibrinolysis system and the formation of thrombi during MI[45]. IL-6 can eventually increase platelet aggregation by mediating enhanced platelet activation, promoting thromboxane A2 synthesis, inhibiting prostacyclin production, and upregulating the expression of GP IIb/IIIa receptors. IL-6 can also promote tissue factor expression, inhibit t-PA activity, upregulate plasminogen activator inhibitor-1, and increase fibrinogen levels, leading to a coagulation-fibrinolysis imbalance[46]. Moreover, IL-6 can elevate D-dimer levels, increasing the risk of acute in-stent thrombosis[47]. Besides, serum IL-6 Levels in AMI-VSR patients are significantly higher than those of patients with simple AMI. In particular, the 30-day mortality rate may rise by 18% for every 10 pg/mL increase in IL-6, and patients with high IL-6 Levels may experience a remarkably higher incidence of cardiogenic shock[48].

In this study, with the identification of obviously higher IL-6 Levels in the death group than in the survival group, this indicator was confirmed to be an independent risk factor for 30-day mortality in AMI-VSR patients. Elevated IL-6 and N% highlight the key role of systemic inflammation in the progression of the disease. Myeloperoxidase and elastase released by neutrophils can directly degrade the extracellular matrix, weakening the structural stability of the ventricular septum[49]. IL-6 can also promote the synthesis of CRP, creating a positive feedback loop that further amplifies the inflammatory cascade[50]. Therefore, anti-inflammatory therapy holds potential value in high-risk AMI-VSR patients, warranting further investigation. For instance, patients with myocardial protection may benefit from early intervention with intensified statins, selective COX-2 inhibitors, and low-dose colchicine to regulate inflammation, as well as optimizing anti-thrombotic and anticoagulation treatments. In addition, combined use of SGLT2 inhibitors as part of a "new quadruple therapy" for heart failure at an earlier stage may help reduce myocardial edema and improve prognosis.

IL-6 is involved in the pathophysiological process of AMI-VSR through multiple mechanisms, acting as a key mediator connecting inflammation, thrombosis, and ventricular remodeling. IL-6 signaling-targeted therapy may offer new strategies for improving the prognosis of AMI-VSR. It is crucial to monitor IL-6 Levels dynamically and implement comprehensive interventions to block its pathological effects in specific clinical scenarios.

Relationship between renal insufficiency and prognosis of AMI-VSR

Renal dysfunction is an independent predictor of poor prognosis in AMI-VSR patients. Data from our center showed significantly lower eGFR in the death group than in the survival group. According to the KDIGO staging system, eGFR has a negative correlation with mortality, with a dialysis-dependent mortality rate reaching 92.1%. This risk gradient is exponentially related to the accumulation of uremic toxins. Even mild renal dysfunction (eGFR of 60-89 mL/minute/1.73 m²) can increase the risk of post-AMI mechanical complications[51].

Owing to renal dysfunction, toxins may accumulate in the body to mediate myocardial suppression. Protein-bound toxins, such as p-cresol sulfate, can suppress L-type calcium channels to reduce myocardial contractility. Meanwhile, mid-sized toxins, such as β2-microglobulin, can activate the TGF-β/Smad3 pathway to increase collagen synthesis. In addition, large-molecule toxins, like parathyroid hormone, may induce myocardial cell apoptosis and reduce mitochondrial membrane potential[52]. Furthermore, renal dysfunction may also increase circulatory volume load, resulting in excessive activation of the renin-angiotensin-aldosterone system (RAAS) and enhance increased cardiac workload and heart failure[53]. Additionally, during this process, electrolyte imbalances may occur frequently, with difficulty excreting potassium, leading to hyperkalemia, increased calcium influx, and cardiac conduction abnormalities, potentially triggering electrical storms[54].

Besides, renal dysfunction has also been considered as one of the most common post-AMI organ injuries, requiring proactive prevention to mitigate its risk of increasing mechanical complications. Prevention strategies include appropriate hydration, use of suitable contrast agents, oral administration of N-acetylcysteine to reduce oxidative damage, precise volume management, and maintenance of homeostasis to prevent kidney injury[55-57]. Renal dysfunction can also interact with multiple pathways to compromise the prognosis of AMI-VSR. It is essential to implement a "cardio-renal integrated" management strategy to benefit the reduction of VSR incidence in patients with AMI and chronic kidney disease, as well as lower the 30-day mortality rate in AMI-VSR patients.

Effect of high altitude environment on the pathophysiology of AMI-VSR patients

At high altitudes (≥ 2500 m typically), PaO₂ decreases by 30%-40% compared to the sea level, with atmospheric oxygen pressure dropping to 105-110 mmHg (reference of 159 mmHg at sea level)[58]. The hypoxic and low-pressure environment at high altitudes has a significant impact on the physiological regulation of the cardiovascular system. Under the stimulation of chronic hypoxia, there may be systemic compensatory responses, including enhanced erythropoiesis, increased blood viscosity, and greater microvascular resistance, resulting in exacerbated myocardial ischemia[59,60]. It is particularly problematic in AMI patients, where insufficient oxygen supply to myocardial cells may further aggravate myocardial injury and necrosis, further triggering declined coronary flow reserve, increased infarct size, and microvascular obstruction subsequently[61]. In this study, decreased Hb levels were an independent risk factor for 30-day mortality in AMI-VSR patients. Patients with lower Hb levels have poorer oxygen-carrying capacity compared to those who compensate for hypoxia by increasing Hb concentrations. It may further increase blood viscosity, further impairing coronary microcirculation as well as worsening myocardial ischemia and infarction. Moreover, owing to the specific geographic environment (i.e., hypoxic and cold) at high altitudes, residents may experience activated sympathetic nervous system and RAAS, leading to peripheral vasoconstriction and increased cardiac afterload, inducing higher risks of ventricular remodeling and heart failure[62,63].

In this study, the significantly higher Hb levels in AMI-VSR patients from high-altitude regions compared to patients from lowland areas may be related to compensatory erythrocytosis triggered by the hypoxic environment at high altitudes. However, the oxygen supply remained insufficient in these patients despite the elevated Hb levels. This is particularly true in the case of AMI-VSR, where blood flow from the left ventricle shifts rapidly to the right ventricle, leading to reduced forward ejection from the left ventricle and further exacerbated tissue hypoxia. As a result, it may increase the risk of hemodynamic collapse, coupled with a remarkable elevation in left-to-right shunt volume, elevated pulmonary capillary wedge pressure, and reduced cardiac output. Consequently, there may be a significantly higher incidence of multi-organ dysfunction, such as acute kidney injury, elevated liver transaminases, and gastrointestinal mucosal ischemia. Moreover, different from those in lowland areas, high-altitude regions exhibit varied treatment responses, presenting with shorter reperfusion windows, reduced efficacy of IABP circulatory support, and lower sensitivity to vasoactive drugs[64]. Therefore, there may be a higher risk of mortality in AMI-VSR patients from high-altitude regions.

In terms of inflammatory response, chronic hypoxia at high altitudes can trigger a cascade of inflammatory storms, manifested as upregulated expression of IL-6, tumor necrosis factor alpha, and other pro-inflammatory factors[65]. The high-altitude environment, through hypoxia-induced multi-pathway interactions, can obviously exacerbate the pathological damage in AMI-VSR patients. In response to this issue, it is recommended to establish targeted management strategies, including optimization of blood rheology (e.g., isovolemic hemodilution, enhanced anti-platelet aggregation, and anticoagulation), control of inflammatory storms (e.g., antagonizing IL-6, inhibiting NLRP3, and adsorbing inflammatory factors), and regulation of oxygen metabolism (e.g., hyperbaric oxygen therapy, mitochondrial protection, and modulation of hypoxia-inducible factors), thereby improving the clinical prognosis of this unique patient population.

Clinical value and limitation of the nomogram-based predictive model

With respect to the above, our study finally established a nomogram, with an AUC of 0.89 (95%CI: 0.85-0.93) in the validation cohort, showing significant improvement. In particular, this model maintained a stable AUC value of 0.86 in the emergency surgery subgroup, whereas the performance of the radiomics model declined to 0.71 due to missing imaging data, underlining significant advantages of the former model in predictive performance. Furthermore, according to a detailed comparative analysis with existing mainstream models: (1) In comparison with the traditional GRACE improved model (AUC of 0.79 vs 0.72 in the validation set), it achieved a 9.7% increase in discrimination by integrating coronary angiography features[66]; (2) While maintaining the imaging advantages of the TIMI-SSI score, it reduced the operation time of the model by 40% based on simplified ultrasound measurement indicators[67]; and (3) Compared with the SSI-ML machine learning model, the logistic regression-based algorithmic framework improved the clinical interpretability by 3.2 times (SHAP value variance analysis), despite a slight decrease in the absolute AUC value (0.83 vs 0.85)[68]. Collectively, the aforementioned improvements enable the model to achieve a better balance between predictive accuracy and clinical utility, particularly suitable for rapid decision-making scenarios in emergency departments. Noticeably, as a single-center retrospective study, this research still has limitations in variable selection scope (e.g., failure to incorporate emerging biomarkers) and population representativeness. The generalization capability of the predictive model in our study necessitates further validation by future multi-center prospective studies.

CONCLUSION

In summary, this study identifies age, UA, IL-6, and low Hb levels as independent risk factors for 30-day mortality in AMI-VSR patients at high altitudes. Our study also develops a nomogram-based predictive model, demonstrating excellent differentiation and consistency for predicting the mortality risk of these patients. It may provide an effective tool for clinicians to assess mortality risk, identify high-risk patients early, develop individualized treatment plans, and ultimately improve patient outcomes. Anyway, future multi-center, large-sample studies are necessary to validate the generalizability of this model and to explore the specific mechanisms through which high-altitude environments affect AMI-VSR patients.