Prognostic value of left ventricular structure and strain in chronic kidney disease patients by cardiovascular magnetic resonance imaging: a retrospective study

Introduction

Patients with chronic kidney disease (CKD) often face a higher incidence of cardiovascular issues like left ventricular hypertrophy, heart failure, and cardiac mortality (Go et al., 2004). CKD contributes to fluid imbalance, electrolyte changes, and increased vascular pressure, leading to sodium and potassium elevation and hypertension, which in turn can cause left ventricular hypertrophy and diastolic dysfunction (Furuland et al., 2018; Luo et al., 2016). Additionally, CKD is associated with ongoing inflammation and oxidative stress, plus an overactive renin-angiotensin-aldosterone system, all contributing to left ventricular systolic dysfunction (Herzog et al., 2011). Echocardiography is crucial for assessing the cardiac structure and function in CKD patients, serving as a key diagnostic tool in clinical settings (Pressman et al., 2015; Hensen et al., 2018), despite its limitations such as operator dependence and poor reproducibility. Speckle-tracking echocardiography studies have consistently shown that CKD patients exhibit increased left ventricular mass and reduced myocardial strain, factors significantly linked to adverse cardiovascular outcomes such as cardiac death, heart failure hospitalization, and myocardial infarction (Pressman et al., 2015; Hensen et al., 2018; Sági et al., 2022). Cardiovascular magnetic resonance (CMR) is considered the “gold standard” for assessing ventricular structure and function (Lu et al., 2011). CMR is the top method for evaluating ventricular health but is underused in CKD due to its time and cost demands. CMR-feature tracking (CMR-FT) offers a feasible alternative for assessing myocardial function in CKD without needing contrast agents. Recent studies indicate that CMR-FT can provide critical prognostic data for heart diseases, such as dilated cardiomyopathy and cardiac amyloidosis (Gao et al., 2021; Geers et al., 2023). This research seeks to understand the connection between left ventricular structure, myocardial strain via CMR, and the combined outcomes of heart failure hospitalization and mortality in CKD patients.

Methods

Results

The study included 77 patients diagnosed with end-stage CKD. The average age of the patients was 54 ± 14 years, and 45% of the participants were female. Of these, 73 patients were classified as stage 5 chronic kidney disease, while four patients were categorized as stage 4. Among the entire cohort, 71 individuals were non-smokers (92%), 31 had diabetes (40%), and 65 had hypertension (with no distinction made between primary and renal hypertension) (84%). Based on the grouping, there were 28 patients (36%) with LVGLS<−15.2%, while 49 patients (64%) had LVGLS ≥−15.2% (Fig. 1). Comparative analyses between groups indicated no statistically significant disparities in hypertension, diabetes, or smoking status. Laboratory assessments encompassing uric acid, serum calcium, high-density lipoprotein, albumin, and hemoglobin levels revealed no discernible differences between the groups (Table 1).

Table 2 illustrates those patients with LVGLS ≥ −15.2% demonstrate augmented LVESD and LVESV on CMR compared to those with LVGLS <−15.2%. Furthermore, individuals with LVGLS ≥ −15.2% exhibit heightened LVM and LVMI (80.94 g/m² vs. 62.33 g/m²), alongside significantly diminished LVEF (42.63% vs. 56.42%) and lower RVEF. Moreover, patients with LVGLS ≥−15.2% manifest decreased overall GRS (20.04% vs. 31.15%) and GCS in the left ventricular myocardium. Notably, no significant difference is observed in LAD and LVEDVI between the two groups.

The findings of the univariable and multivariable analyses are delineated in Table 3. LVESD and LVM were excluded from the multivariable model to mitigate multicollinearity. The univariable analysis encompassed variables known to impact LVGLS, with the majority demonstrating significant disparities between patients with LVGLS ≥ −15.2% and LVGLS <−15.2%. The multivariable analysis incorporated variables significantly linked to LVGLS ≥−15.2%. Ultimately, the outcomes revealed that solely an increased LAD was independently associated with LVGLS ≥ −15.2%.

During a mean follow-up period of 22 ± 9 months, a total of 33 (43%) patients experienced the composite endpoint of HHF and all-cause mortality, comprising 11 (14%) deaths and 30 (39%) instances of heart failure-related hospitalizations, with 8 patients experiencing both events concurrently. The Kaplan–Meier curves depicted in Fig. 2 illustrate the outcomes of the composite endpoint for patients with LVGLS ≥ −15.2% compared to those with LVGLS <−15.2%. Patients exhibiting LVGLS ≥ −15.2% demonstrated notably lower cumulative event-free survival rates for HHF and all-cause mortality compared to their counterparts with LVGLS <−15.2% (0.62 vs. 0.84 at 20 months of follow-up; log-rank P = 0.014). Additionally, the Kaplan–Meier curves in Fig. 3 depict the results of all-cause mortality among patients with LVGLS ≥ −15.2% versus LVGLS <−15.2%. Notably, patients with LVGLS ≥ −15.2% exhibited substantially inferior prognosis relative to those with LVGLS <−15.2%, with cumulative survival rates at 20 months of 0.76 vs. 1.00, respectively (log-rank P = 0.001).

In the univariate regression analysis, variables such as age, LAD, LVMI, LVGLS, and diabetes were correlated with a heightened risk for the combined endpoint of HHF and all-cause mortality (Table 4). Upon conducting a Cox proportional hazards multivariate analysis, age, diabetes, and LVGLS were significantly associated with the risk of endpoint events, whereas LVMI showed no significant association. After adjustments for age, diabetes, LAD, and LVMI, a compromised LVGLS was independently predictive of an elevated risk for the composite endpoint (HR: 3.40, 95% CI [1.35–8.56], P = 0.009) as delineated in Table 4.

Discussion

This investigation demonstrates that the predominant fraction of individuals with advanced CKD exhibit compromised left ventricular myocardial systolic function, notwithstanding the preservation of left ventricular ejection fraction in certain cases. An enlargement in LAD independently correlates with deteriorated LVGLS. Additionally, CKD subjects exhibiting impaired LVGLS encounter an elevated incidence of HHF and all-cause mortality relative to their counterparts within this cohort.

This study noted that nearly two-thirds of individuals with CKD demonstrated left ventricular systolic dysfunction. An augmented LAD independently correlated with LVGLS of ≥ −15.2%. Various factors, such as myocardial fluid retention, inflammation from CKD, toxin accumulation, and myocardial fibrosis, may contribute to the reduction in LVGLS observed in CKD patients (Arcari et al., 2021). Furthermore, the study by Zhang et al. (2021) posited that individuals with concurrent diabetes and CKD are prone to more severe cardiovascular impairment, including compromised LVGLS, alongside an escalated risk of heart failure and mortality. In our research, while a direct comparison of LVGLS between CKD patients with and without diabetes was not conducted, a diminished LVGLS notably elevated the risk of heart failure hospitalization and all-cause mortality, even after adjustment for pertinent variables (HR: 3.40, 95%CI [1.35–8.56], P = 0.009), aligning with certain speckle-tracking echocardiography studies. This finding is consistent with some speckle-tracking echocardiography studies. Pressman et al. (2015) uncovered a marked relationship between adjusted LVGLS and mortality, utilizing an LVGLS ≥-16% threshold (HR: 1.30, 95%CI [1.10–1.56], P = 0.002). Similarly, Hensen et al. (2018) reported analogous findings in CKD patients with preserved left ventricular ejection fraction, applying an LVGLS ≥−15.2% benchmark (HR: 0.91, 95%CI [0.83–0.999], P = 0.047). Future research may need to explore whether the extent of LVGLS damage is more severe in patients with CKD and diabetes and investigate its impact on prognosis.

Our findings suggest no significant link between elevated LVMI and the risk of HHF or mortality in CKD patients. Additionally, evidence points to lower LVMI levels in the Chinese population compared to Western cohorts (Lu et al., 2011). The current association between changes in LVMI and increased risk of heart failure admission and mortality remains uncertain. Early studies using CMR by (Bluemke et al., 2008) suggested that LVMI is the best predictor of heart failure . Meanwhile, (Sági et al., 2022) conducted a study using echocardiography and proposed that higher LVMI may be an independent prognostic factor for end-stage kidney disease and cardiovascular events. In contrast, the study involving over 12,000 patients with hypertension, showed that the reduction in left ventricular hypertrophy was not correlated with cardiovascular events (Costanzo et al., 2013). Agarwal et al. (2020) posited that the relationship between LVMI and the progression of CKD is bidirectional. Another meta-analysis concluded that the clinical significance of reduced LVMI in patients with CKD post-treatment remains uncertain, indicating that LVMI is not suitable for predicting adverse outcomes in patients with CKD (Badve et al., 2016). The elevation of LVMI in CKD patients stems from interrelated factors. Reduced glomerular filtration rate exacerbates cardiac remodeling, myocardial fibrosis, and hypertrophy, partly due to uremic toxin accumulation in the myocardial interstitium and their direct impact on myocardial cells (Hayer et al., 2020). Furthermore, traditional cardiovascular risk factors like hypertension and diabetes, prevalent among CKD patients, contribute to this increase (Kidney Disease: Improving Global Outcomes (KDIGO) Diabetes Work Group, 2022). These findings suggest LVMI elevation is a consequence, not a cause, of advancing CKD. Therefore, we believe that the use of LVMI for prognostic assessment in Asian CKD patients should be approached with caution, and further multicenter studies with large sample sizes are essential.

Conclusion

In summary, this research offers valuable insights into the correlation between LVGLS and HHF as well as all-cause mortality in CKD patients. The utility of LVMI in prognostication for Asian CKD patients necessitates careful consideration. These findings carry significant clinical relevance for optimizing CKD patient management and enhancing their prognostic outlook.

Supplemental Information