Prognostic Value of Shock Index Creatinine in Patients with Severe Aortic Stenosis Undergoing Transcatheter Aortic Valve Replacement

Severe aortic stenosis (AS) is not uncommon in the elderly population and often accompanies congestive heart failure (HF) [1]. Without intervention, left ventricular HF secondary to AS may progress to biventricular HF and subsequently to cardiogenic shock [2, 3]. Both biventricular HF and cardiogenic shock are associated with increased mortality, emphasizing the importance of early intervention [4, 5]. Transcatheter aortic valve replacement (TAVR) is increasingly being performed in patients with severe AS across risk strata [6‒8]. Accurate risk stratification is essential for appropriate clinical management and prognostic assessment. However, traditional risk prediction models such as the Society of Thoracic Surgery (STS) risk score, EuroSCORE-2, etc. are not only complex and difficult to obtain at the bedside but also suboptimal in predicting adverse outcomes after TAVR [9].

Hypotension and tachycardia are commonly attributed to the exacerbation of HF due to volume overload or inadequate effective circulating blood volume, reflecting a poorer cardiac status. It is estimated that 1%–4% of patients undergoing TAVR may develop cardiogenic shock, which is characterized by congestion and inadequate tissue or end-organ perfusion [10]. Higher resting heart rate (HR) has also been reported to be associated with the development of AS and the risk of cardiovascular death [11]. Shock index (SI), defined as the ratio of HR to systolic blood pressure (SBP), and its derivatives, including age shock index (ASI), modified shock index (MSI), and so on, have also been validated for various cardiovascular diseases such as coronary artery disease [12] and HF [13], demonstrating a good predictive value. Recently, shock index creatinine (SIC), a novel simple risk stratification tool based on SI and renal function, has been proposed and externally validated in patients with acute coronary syndrome [14, 15]. SIC has demonstrated superior discriminative ability for in-hospital mortality in ST-elevation myocardial infarction compared to other SI-based risk indices [16]. Given that renal dysfunction [17, 18] are also significant prognostic factors in patients undergoing TAVR, SIC holds promise as a valuable prognostic tool for patients with severe AS undergoing TAVR. Nevertheless, the clinical relevance and clinical significance of SIC in patients undergoing TAVR for severe AS remain unclear and require further investigation. Therefore, the objective of this study is to evaluate the prognostic value of SIC in predicting clinical outcomes after TAVR in patients with severe AS.

This retrospective study included patients with severe symptomatic AS who consecutively underwent TAVR at Guangdong Provincial People’s Hospital between April 2016 and March 2023. The indications for TAVR were determined by a consensus of a multidisciplinary cardiac team. Inclusion criteria were as follows: (1) age ≥18 years; (2) undergoing TAVR procedure. We excluded 30 patients who underwent TAVR for pure severe aortic valve regurgitation. A total of 555 patients were included in the final analysis of the study.

All patients undergoing TAVR were enrolled in our hospital’s computerized database for TAVR patients, where clinical data, laboratory data, echocardiographic data, operative variables, and follow-up information were prospectively collected. Vital signs on admission, including SBP and HR, were determined from data first recorded on admission in a supine or sitting position. Mean arterial pressure was calculated as (SBP + 2 × DBP)/3. SI and MSI were calculated as a ratio of HR to SBP and HR to mean arterial pressure, respectively. ASI was calculated as age × SI. The estimated creatinine clearance (CCr) was calculated using the published Cockcroft-Gault formula: for males, (140 − age)/serum creatinine; for females, (140 − age)/serum creatinine × 0.85. The calculation formula for SIC was as follows: (SI × 100) − CCr [14]. The Geriatric Nutritional Risk Index (GNRI) was calculated using the following formula: GNRI = 1.489 × serum albumin (g/L) + 41.7 × weight/ideal weight. Ideal weight was calculated as follows: for males, height − 100 − [(height − 150)/4]; for females, height − 100 − [(height − 150)/2.5] [19]. Frailty was defined as GNRI <98, in accordance with previously reported criteria [20]. The STS score was calculated for each participant using online calculators. Anemia was defined as hemoglobin concentration <120 g/L in females or <130 g/L in males. Hypoalbuminemia was defined as serum albumin concentration <35 g/L. All patients underwent pre- and postoperative transthoracic echocardiography evaluation. All parameters were assessed in accordance with the current guidelines of the American Society of Echocardiography and the European Society of Echocardiography [21, 22]. The incidence of mortality, stroke, myocardial infarction, pericardial tamponade, acute kidney injury (AKI), permanent pacemaker implantation, major vascular complication, life-threatening bleeding, new-onset atrial fibrillation (or flutter), and aortic valve re-intervention were recorded in accordance with the Valve Academic Research Consortium-3 (VARC-3) definition [23]. For each patient who survived to discharge, follow-up was conducted by reviewing medical records and/or via telephone interviews with patients or their family members. All patients were followed up. The primary endpoint was all-cause mortality during the follow-up period, with a median of 17 months (interquartile range: 8–31 months). Secondary endpoints included complications during hospitalization as defined by the VARC-3 criteria.

The Kolmogorov-Smirnov test was employed to assess the normality of variables. For normally distributed variables, continuous variables were expressed as mean ± standard deviation, and comparisons between groups were made using the independent samples t test. For nonnormally distributed variables, continuous variables were presented as medians and interquartile ranges (25th–75th), and between-group comparisons were made using the Mann-Whitney U test. Categorical data were presented as frequencies and percentages, and comparisons were made using χ² tests. The Spearman correlation coefficient was employed to assess relationships between variables. The Kaplan-Meier method was used to plot unadjusted cumulative death curves, with statistical significance determined using the log-rank test. Cox regression analysis was conducted to determine the association between variables and mid-term mortality. All significant variables (p < 0.05) identified in the univariate analysis were included in the multivariate Cox regression model to identify independent predictors. In order to avoid multicollinearity, variables such as HR, SBP, SI, MSI, ASI, creatinine, CCr, peak velocity, and chronic kidney disease were excluded from the multivariate model. The results were expressed as odds ratios with 95% confidence intervals (CI). The relationship between the SIC and the primary outcome was described using four-knot restricted cubic splines, with the knots set at the 5th, 35th, 65th, and 95th percentiles of the SIC. Receiver-operating characteristic curve analysis was employed to evaluate the performance of the diagnostic indicators, and Youden’s index was applied to determine the optimal cutoff value. The area under the receiver-operating characteristic curve (AUC) was compared between different models using the method described by DeLong et al. [24]. Two-sided tests were performed with a significance level of α = 0.05. Statistical analyses were completed using IBM SPSS Statistics version 25.0 (IBM Corporation) and R software 4.3.0.

A total of 555 patients with severe AS underwent TAVR. The study population was consisted of 235 women (42.3%) and 320 men (57.7%), with a median age of 72 years, as shown in Table 1. The optimal cutoff value of SIC for predicting 1-year mortality was 16.5 (AUC: 0.711 [95% CI 0.622–0.800], p < 0.001), with a sensitivity of 67.60% and a specificity of 72.20%. Consequently, 386 patients were classified as belonging to the low SIC group, while 169 patients were assigned to the high SIC group. In comparison to patients with a low SIC, those with a high SIC were older, had a lower body mass index, a faster HR, lower SBP, were more likely to have a New York Heart Association Functional classification (NYHA) ≥ III, and had a history of percutaneous coronary intervention, chronic kidney disease, and hemodialysis. Spearman correlation analysis revealed a negative correlation between SIC and GNRI (r = −0.263, p < 0.001), and frailty was more common in the high SIC group. Additionally, there was a positive correlation between SIC and the Society of Thoracic Surgery (STS) risk score (r = 0.458, p < 0.001). In terms of laboratory findings, patients in the high SIC group exhibited higher levels of N-terminal pro-B-type natriuretic peptide (NT-proBNP) and were more susceptible to anemia and hypoalbuminemia. Apart from whether pre-dilatation was performed, procedure characteristics were comparable between the two groups, as demonstrated in Table 1. In relation to echocardiographic findings, patients with a high SIC displayed more extensive cardiac damage than those with a low SIC, characterized by reduced left ventricular ejection fraction, larger left ventricular end-diastolic diameter, a higher prevalence of moderate or greater degrees of mitral and tricuspid regurgitation, and a higher prevalence of pulmonary hypertension (Table 1).

At a mean follow-up of 21.5 months, 51 (9.2%) patients died from any cause, including 35 (6.3%) from cardiovascular causes and 16 (2.9%) from noncardiovascular causes. The all-cause mortality rate was significantly higher among patients with a high SIC level compared to those with a low SIC level (18.3% vs. 5.2%, p < 0.001) (Table 2). The Kaplan-Meier analysis for 1-year all-cause mortality is shown in Figure 1. Furthermore, patients with a high SIC had higher rates of cardiovascular mortality (13% vs. 3.4%, p < 0.001) and noncardiovascular mortality (5.3% vs. 1.8%, p = 0.046) over the follow-up time, as shown in Table 2.

The in-hospital and follow-up outcomes are shown in Table 2. Among 555 patients, 9 (3.2%) patients died (all-cause deaths), 6 (1.1%) underwent re-interventions, 3 (0.5%) had myocardial infarction, 10 (1.8%) had stroke, 32 (5.8%) had AKI, 49 (8.8%) experienced VARC type ≥2 bleeding, 32 (5.8%) had major vascular complications, 32 (5.8%) had permanent pacemaker implantation, 49 (8.8%) had new-onset atrial fibrillation or flutter, 16 (2.9%) had pericardial tamponade, and 12 (2.2%) had coronary obstruction during their hospital stay. Of note, the in-hospital mortality rate in the high-SIC group was significantly higher than that in the low-SIC group, with a more than tenfold higher incidence (4.7% vs. 0.3%; p < 0.001). Patients with a high SIC were more likely to experience VARC type ≥2 bleeding (13.6% vs. 6.7%, p = 0.014) and AKI (10.1% vs. 3.9%, p = 0.008) compared to those with a low SIC. No other significant differences were found in terms of other adverse outcomes (Table 2).

In univariate Cox regression analysis, several factors were identified as significant predictors of mid-term mortality, including SIC, STS score, frailty, NYHA ≥ III, aspartate aminotransferase, NT-proBNP, left ventricular ejection fraction, aortic mean gradient, tricuspid regurgitation ≥ moderate, and pulmonary hypertension. However, multivariable analysis revealed that only SIC and STS score remained as independent predictors of all-cause mortality. The unadjusted HR for high SIC and all-cause mortality is 3.96 (95% CI: 2.21–7.11; p < 0.001), while the adjusted HR for high SIC and all-cause mortality is 2.188 (95% CI: 1.103–4.341; p < 0.001) (Table 3). The dose-response curve of SIC with all-cause mortality is depicted in Figure 2. When SIC is less than 0, the risk of death does not change significantly with SIC variation. However, when SIC exceeds 0, the risk of death sharply increases with higher SIC (Fig. 2).

The performance of SIC was comparable to that of the STS score in predicting 1-year mortality (AUC: 0.711 vs. 0.649, p = 0.16). The incorporation of SIC into the STS score resulted in a notable enhancement in predictive efficacy when compared to the use of the STS score in isolation (AUC: 0.731 vs. 0.649, p = 0.01) (Fig. 3). Furthermore, the results demonstrated that SIC exhibited the highest AUC compared to other SI-based risk indices, including SI, ASI, and MSI (AUC = 0.64, 0.648, and 0.626, respectively) (Fig. 4). Risk stratification based on these indices can accurately identify high-risk patients, and they performed similarly in independently predicting 1-year mortality (all p > 0.05).

This study was the first to evaluate the clinical significance of SIC in patients with severe AS undergoing TAVR, and the findings were as follows: (1) After adjusting for covariates, SIC was independently associated with mid-term mortality following TAVR. These associations were nonlinear. (2) A high SIC was associated with poorer cardiac status, frailty, and malnutrition. (3) The prognostic performance of SIC for 1-year mortality was comparable to that of the STS score. Furthermore, the incorporation of SIC into the STS score significantly enhanced the prediction of 1-year all-cause mortality. Our study demonstrated that SIC, as a straightforward index, can be employed to identify severe AS patients at high risk of adverse outcomes, even after TAVR.

Several previous studies have confirmed that higher resting HR is associated with the progression of AS [11, 25]. However, the exact pathophysiological mechanisms behind this association remain unclear. The development of AS is an active inflammatory process that is thought to be initiated and propagated, at least in part, by mechanical shear stress. It is plausible that higher resting HR may have a direct impact on AS pathology. However, it may also reflect concomitant clinical conditions or serve as a marker of some degree of sympathetic-vagal imbalance. However, a retrospective analysis of 349 patients undergoing TAVR found that neither baseline nor discharge resting HR was associated with all-cause mortality or cardiovascular mortality after TAVR [26]. This study used an HR cutoff of 77 bpm, based on the results of the randomized Systolic Heart Failure Treatment with the If Inhibitor Ivabradine Trial (SHIFT). This threshold may not be the optimal HR threshold for distinguishing TAVR outcomes, potentially leading to the negative conclusion of the study. The decline in SBP observed in TAVR patients may be associated with a reduction in cardiac output due to severe AS. In severe cases, this reduction can contribute to cardiogenic shock, which is correlated with increased mortality rates following TAVR [4, 5].

SI was initially proposed in 1967 by Allgöwer and Burri [27] as a means of assessing the severity of hypovolemic shock in clinical settings. SI integrates information on HR and SBP and provides a comprehensive evaluation of cardiovascular status and shock risk. An increase in HR and a reduction in SBP can result in an elevated SI. An elevated SI is typically indicative of inadequate cardiac compensation and may reflect underlying cardiac damage or severe circulatory load. Généreux et al. [2] established a staging classification of cardiac damage based on conventional echocardiographic parameters through an analysis of severe AS patients in the PARTNER 2 trials. The findings revealed that the degree of cardiac injury was independently associated with increased mortality rates following aortic valve replacement, with a hazard ratio of 1.46 for each stage increase (95% CI 1.27–1.67, p < 0.0001). In our study, patients with elevated admission SIC levels exhibited more pronounced cardiac injury, characterized by increased left ventricular end-diastolic volume, exacerbated mitral and tricuspid regurgitation, as well as heightened pulmonary arterial hypertension. BNP and NT-proBNP are classic markers released during ventricular and atrial stretching, reflecting myocardial volume load. Several studies have demonstrated a close association between BNP and NT-proBNP and the short-term and long-term outcomes of AS patients undergoing TAVR [28, 29]. Our observations indicated a correlation between elevated SIC on admission and higher levels of NT-proBNP. Pathophysiologically, an increase in SI may reflect a greater hyperdynamic state and/or activation of the sympathetic nervous system [30]. Nevertheless, the actual usefulness of this index in cardiovascular diseases has only been evaluated in recent years. AS enhances cardiac sympathetic nervous system activity [31]. Although the exact role of SI in severe AS remains unclear, the potential relationship between excessive sympathetic autonomic nervous system stimulation in severe AS and changes in left ventricular stroke volume may play a role in this regard, possibly reflecting interactions between the nervous and cardiovascular systems.

Low-flow, low-gradient (LFLG) AS represents a distinct patient cohort with an exceptionally elevated mortality risk. This group can be further subdivided into patients with severe paradoxical low-flow, low-gradient (P-LFLG) AS and patients with classical low-flow, low-gradient (C-LFLG) AS, with the latter having an even higher mortality risk [32]. The present study demonstrated that patients with elevated SIC levels were more likely to exhibit low gradients. However, our hospital’s echocardiographic protocol does not routinely measure stroke volume index, making it challenging to differentiate LFLG-AS patients. Future research should focus on exploring the relationships between SIC, flow, transvalvular pressure gradients, and clinical outcomes to gain a deeper understanding of the associated hemodynamic changes. This could potentially provide more insight into the mechanisms by which SIC influences prognosis.

Renal insufficiency has been widely established as a robust predictor of short- and medium-term mortality following TAVR [17, 18]. Additionally, the Cockcroft-Gault CCr has been employed for preoperative renal function classification and has demonstrated advantages in mortality prediction. This association has been corroborated across diverse populations, including those with HF [33] and those undergoing cardiac valve surgery [34]. Furthermore, renal impairment assessed based on the Cockcroft-Gault formula is a significant indicator in the EuroSCORE II, a recognized cardiac surgery scoring system. Consequently, incorporating CCr into SI may enhance the predictive accuracy for severe AS patients undergoing TAVR. Further research could investigate the comparative predictive efficacy of various equations for eGFR, such as other formulas based on serum creatinine, those based on cystatin C alone, and those utilizing a combination of serum creatinine and cystatin C. Compared to risk assessment tools like the STS score or EuroSCORE II, which encompass a greater number of variables, SIC comprises only three easily collectible and computable factors, facilitating rapid bedside calculation. AKI and adverse outcomes following TAVR were associated in both short-term and long-term scenarios [17, 18]. In our cohort, elevated SIC is correlated not only with higher short-term and medium-term mortality rates but also with a higher incidence of in-hospital AKI occurrence. Various mechanisms underlying the occurrence of AKI following TAVR have been proposed, with the most commonly cited being contrast agent exposure leading to acute tubular necrosis, hemodynamic instability, rapid pacing, and thrombotic phenomena [35]. However, research has also indicated that the increased staging of extravalvular cardiac damage correlated with a higher incidence of AKI following TAVR. AKI has been shown to exert a negative impact on long-term clinical outcomes, but this appears to be confined to patients with advanced cardiac injury [36]. In patients with AS, potential mechanisms of cardiorenal interaction may involve late-stage ventricular pressure overload and systemic venous congestion resulting from severe AS, which subsequently leads to reduced renal perfusion and congestion, resulting in renal injury. This condition is classified as type 2 cardiorenal syndrome. Conversely, renal insufficiency induces hyperphosphatemia, which may exacerbate vascular and valvular calcification, resulting in increased pressure overload, chronic inflammation, and dyslipidemia in patients with AS. This may, in turn, exacerbate the symptoms and disease progression of AS and HF [37]. In the event that correctable factors causing renal injury, such as hypovolemia or hypotension, are present prior to TAVR, it is of the utmost importance to address these underlying causes proactively in order to optimize renal function improvement. In the case of patients requiring dialysis, it may be advisable to perform the dialysis within 24 h prior to the planned procedure.

The majority of patients undergoing TAVR are elderly, and comorbidities such as frailty and malnutrition are associated with an increased risk of adverse outcomes following TAVR [20, 38, 39]. Therefore, it is of paramount importance to assess the adequacy of nutritional status in order to risk-stratify patients undergoing TAVR. Our study showed that patients in the high SIC group had lower body mass index and triglyceride levels [39], as well as higher rates of anemia and hypoalbuminemia [20, 38], all of which have been reported as major risk factors for frailty and malnutrition. The GNRI is a simple and well-established nutritional screening tool. Seoudy et al. [20] demonstrated that a low GNRI was associated with an elevated risk of all-cause mortality in TAVR patients. In our study, a low GNRI was associated with all-cause mortality in univariate Cox regression analyses. Furthermore, we observed a negative correlation between SIC and GNRI, and additionally, patients with high SIC showed a higher prevalence of frailty. It is regrettable that the precise etiology of hypoalbuminemia and anemia in this study remains elusive. In addition to advanced age and comorbidities, it is plausible that blood dilution resulting from volume overload might also exert a significant influence. Further research is needed to confirm whether early correction of hypoalbuminemia and anemia can improve clinical outcomes.

In this study, SIC exhibited the highest AUC value for predicting 1-year mortality in AS patients undergoing TAVR, compared to SI, MSI, and ASI. Furthermore, SIC demonstrated a higher AUC value compared to the STS score, though the difference was not statistically significant. Therefore, it cannot be asserted that SIC is more effective than the STS score in predicting post-TAVR mortality. However, our study was not designed to compare these two scores. Nevertheless, SIC remains a valuable tool as it can be calculated using only three simple variables (HR, SBP, and CCr). The SIC scoring system does not require patient medical history, clinical presentation, or echocardiographic results, nor does it necessitate digital tools for score computation, unlike the STS score. Hence, SIC is cost-effective, user-friendly, and provides rapid results. On the other hand, traditional surgical risk scores lack the capacity to assess hemodynamic stability and quantifiable parameters of frailty. However, when used in conjunction, they can offer a more comprehensive assessment and risk prediction. The AUC indicates that incorporating SIC enhances the prognostic capability of the STS score. Additionally, SIC is significantly associated with AKI and bleeding. Consequently, SIC serves as both a standalone, straightforward, and efficient tool for rapid and accurate risk stratification of various adverse events and an adjunctive tool to enhance the predictive ability of traditional scores. Early identification of high-risk patients using SIC allows for more intensive perioperative monitoring and postoperative follow-up. For these high-risk patients, the prophylactic use of miniaturized venoarterial extracorporeal membrane oxygenation may be both safe and advisable [40]. Future research should focus on identifying the risk factors associated with elevated SIC and exploring early intervention. Such efforts could potentially improve outcomes for patients with AS by mitigating the risk associated with high SIC levels.

This study has several limitations. Primarily, it was a single-center retrospective observational analysis, which requires further validation in large-scale, multicenter studies. Future research should consider external validation to enhance the generalizability of the findings. Furthermore, there may be unmeasured confounding factors affecting the relationship between SIC and mortality. For instance, data on the use of calcium channel blockers and beta-blockers at admission were not collected, which could influence SBP and HR and potentially impact the results. Another limitation is related to soluble suppression of tumorigenicity 2, a member of the interleukin-1 receptor family and a prominent biomarker for predicting outcomes after TAVR [41]. However, the role of soluble suppression of tumorigenicity 2 in this context remains uncertain due to the infrequent measurement of this biomarker in our cohort. Additionally, this study was based on a database collected over more than 8 years, during which time there have been changes in technology, devices, and participant selection. Moreover, the predictive value of SIC has only been validated in patients with severe AS undergoing TAVR. Consequently, caution should be exercised when using this index to assess patients who did not undergo TAVR or underwent TAVR for other reasons. Despite these limitations, this study was the first to evaluate the relationship between SIC, an easily quantifiable mortality risk prediction index, and clinical outcomes subsequent to TAVR.

In conclusion, SIC is a simple and efficient risk prediction system that may provide valuable information for predicting in-hospital and mid-term mortality in patients undergoing TAVR. A high SIC was significantly associated with cardiac injury and malnutrition. However, the current results warrant further confirmation through additional studies or validation efforts.

This study was conducted in accordance with the principles set forth in the Declaration of Helsinki and was reviewed and approved by the Ethics Board of Guangdong Provincial People’s Hospital (Approval No. GDREC2019384H).

The authors have no conflicts of interest to declare.

This research was supported by Guangdong Provincial Clinical Research Center for Cardiovascular Disease (2020B1111170011), Guangdong Provincial Key Laboratory of Coronary Heart Disease Prevention (Y0120220151), Clinical Major Technology Project of Guangzhou (2023FTJCZ0017), and Science and Technology Program of Guangzhou, China (2023B03J1256). The funders played no role in the study’s design, data collection, data analysis, or reporting.

Bangyuan Yang and Changjin Wang contributed equally to the conceptualization, study design, data analysis and interpretation, drafting, and manuscript revision. Songyuan Luo, Jie Li, and Jianfang Luo significantly contributed to the study design, supervision, and manuscript revisions. Ting Zhou, Yinghao Sun, Shengneng Zheng, and Jiaohua Chen made substantial contributions to data collection, interpretation, and critical revisions of important intellectual content. All authors reviewed and approved the final manuscript submission.

Bangyuan Yang and Changjin Wang contributed equally to this work.Similarly, Songyuan Luo, Jianfang Luo, and Jie Li also made equal contributions.

The data supporting the results of this study are not publicly available due to concerns regarding the privacy of the research participants. However, they can be obtained from the corresponding author, J.F. Luo (email: [email protected]), upon reasonable request.