The latest Guidelines for the Diagnosis and Treatment of Hypertrophic Cardiomyopathy in Chinese Adults 2023 [1] indicate that hypertrophic cardiomyopathy (HCM) is a disease primarily characterized by myocardial hypertrophy caused by pathogenic mutations in genes encoding sarcomeric and/or related proteins. It remains one of the leading causes of sudden cardiac death (SCD) in young individuals. The incidence of HCM in the general population has gradually increased from 1 in 500 to 1 in 200 in recent years, showing a persistent upward trend [2, 3]. The 2014 European Society of Cardiology (ESC) Guidelines [4] and the 2020 American Heart Association/American College of Cardiology (AHA/ACC) Guidelines [5] explicitly state that stress echocardiography can evaluate and provide abundant prognostic information (COR IIa), which can serve as a reference for the clinical treatment and follow-up in HCM patients. The SCD risk assessment model for HCM, recommended by the 2020 AHA/ACC [5], includes seven factors: “maximum wall thickness, ventricular tachycardia, syncope, family history of SCD, ventricular aneurysm, left ventricular ejection fraction (LVEF), and late gadolinium enhancement on cardiac magnetic resonance imaging”, all of which involve at least three different types of examinations. However, some HCM patients are unable to complete these examinations due to economic conditions or contraindications. Currently, “left atrial strain” has been verified in several studies [6, 7, 8, 9, 10] to have predictive value for adverse cardiovascular events in diseases such as heart failure with preserved ejection fraction, chronic kidney disease, moderate aortic valve stenosis, and diastolic dysfunction. Therefore, in this study, we hypothesized that “left atrial reservoir strain” has clinical significance in predicting adverse cardiovascular events in HCM patients. The aim of this study was to identify and assess methods and parameters in high-risk HCM patients using exercise stress echocardiography (ESE) that could be easily and reliably used in everyday clinical practice.

In this study, ESE was used to conduct a comprehensive analysis of exercise-induced cardiac function and reserve in HCM patients, with a 5-year follow-up for positive cardiovascular events. We observed that in the HCM positive event group, both resting and peak left atrial reservoir strain and GLS were significantly impaired, accompanied by marked deterioration or even loss of reserve. R_4D_LASr and P_2D_GLS had independent predictive value for positive cardiovascular events.

Huang et al. [14] found that in HCM, whether the ventricular wall is thickened or not, there are abnormalities in GLS, and mechanical abnormalities may precede the occurrence of hypertrophy [15]. At the same time, through multivariate analysis, it was found that GLS reflects impaired myocardial contraction and is related to the severity of HCM [16]. Wu et al. [17] conducted further research on HCM and found that the GLS of HCM decreased both before and after exercise, the systolic reserve during exercise was also significantly reduced, and the decreased GLS at peak was significantly related to exercise intolerance. Further research by Badran et al. [18] found that the systolic function reserve of HCM patients decreased by 23% after exercise. Based on previous studies, this study combined the influencing factors of GLS. Not only did it select the most physiologically appropriate treadmill exercise stress mode for HCM patients, but according to the 5-year follow-up of HCM, it was found that even if the GLS of HCM patients at rest is at the edge of the critical value or slightly decreased, the GLS after exercise stress will immediately show a significant decrease, and the degree of reduction in systolic strain reserve is about 22%. In the positive events group, the GLS at rest and peak exercise is significantly decreased, and the reduction in systolic reserve ($\mathrm{\Delta }$2D_GLS %) is up to 33%, and the reserve capacity is in a deteriorated state. At the same time, we found that the peak GLS has predictive value for positive events. Because GLS is a parameter obtained by digital processing of sound images after detecting myocardial deformation by ultrasonic speckle-tracking imaging (STI), it reflects the quantitative and localized analysis of myocardial mechanics. The 2D-GLS obtained under 2D-STI is not affected by angle dependence and the motion interference of adjacent myocardial segments, so as to realize the continuous observation of myocardial motion, which has a better predictive value for the prognosis of cardiovascular diseases [19, 20] and has been regarded as one of the main tools for evaluating systolic function by the ASE/ESC [21]. However, 2D-GLS also has its own limitations and is easily affected by load factors. When the cardiac afterload significantly increases, its accuracy will be diminished. Therefore, in this study, the exercise stress mode was selected to make up for the shortcomings of speckle tracking technology, and a comprehensive evaluation was carried out on the resting and peak strains of HCM and the reserve capacity after load strain to reflect cardiac systolic function and reserve from a more objective perspective. The final result that “peak GLS has a more significant predictive efficacy for positive events” also supported this observation.

In the 2014 and 2022 ESC guidelines [4, 22] for the clinical risk factors for SCD in HCM, the atrial index is “left atrial internal diameter/size”, and the LA_Vlmax $>$34 mL/m² is also considered a key parameter for determining heart failure with preserved ejection fraction. In this study, there was a statistically significant difference in the resting inner diameter of LA(R_LAd) between the HCM group and the control group. The resting left atrial volume index (R_LA_Vlmax) of HCM was significantly higher than that of the control group. However, usually clinical measurements are combined with the patient’s BSA. Therefore, LA_Vlmax should be more accurate, after all, the abnormal diastolic function of HCM leads to changes in the structure and function of the left atrium, the remodeling pattern in terms of structure is manifested as an increase in three-dimensional volume, which can be verified in the myocardial magnetic resonance 3D imaging study of HCM patients by Muresan et al. [23]. All the HCM patients enrolled in this study were capable of undergoing the treadmill exercise stress test, and were not in the subclinical heart failure or clinical heart failure stage. According to the occurrence and development of pathophysiological diseases and previous research results [7], the “left atrial size” is also affected by many factors such as the duration of the disease, the degree of disease development, and mitral regurgitation. Therefore, simply detecting “left atrial size” may not be sensitive and rigorous enough in terms of a pathological mechanism, which affects the accurate judgment of diastolic function and the risk of HCM in clinical practice, which is one of the reasons why the left atrial strain series parameters were introduced in this study.

Left atrial reservoir strain has been shown by previous studies to not only sensitively and objectively reflect diastolic dysfunction in HCM at an earlier stage, but also that the sensitivity and accuracy in predicting adverse cardiovascular events are significantly improved compared to resting GLS and E/e’ [24, 25, 26, 27, 28, 29, 30]. Left atrial storage strain has been proven to be closely related to pulmonary capillary wedge pressure and the early response to earlier treatment, and is a strong predictor of long-term prognosis in HCM patients [14, 31]. In this study, for the analysis of left atrial strain, regardless of 2D or 3D-STI, ventricular end-diastole was selected as the zero baseline of the left atrial strain curve, and longitudinal strain was generated from each atrial segment. It was found that the left atrial storage strain, conduit strain, and contraction strain in the positive events group were significantly worse than those in the non-events group. After exercise, the left atrial peak storage strain, conduit strain, and contraction strain in the positive events group were further reduced, and the reserve ($\mathrm{\Delta }$LASr_ED, $\mathrm{\Delta }$LASr_ED %) was significantly deteriorated. Both the regression analysis and the ROC found that “R_4D_LASr” had excellent predictive efficacy for positive events, since “R_4D_LASr” is combined with full volume real time dynamic images. It reflects the mechanical changes in both the longitudinal and circumferential directions, and it can reflect the mechanical changes during the diastolic process of the entire cardiac cycle without being limited by heart rate. It has less dependence on angles and loads, and it can sensitively distinguish the active and passive movements of myocardial tissue, thus objectively and comprehensively evaluating the storage, conduit, and contraction functions of the left atrium.

In all HCM patients, both resting obstruction and obstruction which occurred during exercise were labeled as obstruction patients. As shown in Table 4, in the non-event group (102 HCM patients), there were 21 cases of obstruction, accounting for 20.6%, and in the positive event group (40 HCM patients), there were 18 cases of obstruction, accounting for 45.0%, which was statistically significant (p = 0.006). In the obstruction group, the R_4D_GLS, R_2D_GLS, P_2D_GLS, R_LASr_ED, R_4D_LASr, P_LASr_ED and P_LASct_ED values were significantly worse, both the $\mathrm{\Delta }$2D_GLS, $\mathrm{\Delta }$2D_GLS % and $\mathrm{\Delta }$LASr_ED, $\mathrm{\Delta }$LASr_ED % were in a deteriorated state. Norrish et al. [32] also found an inverted U-shaped relationship between left ventricular hypertrophy and the risk of SCD. The influence of more factors on the risk of SCD needs to be further evaluated. In this study, we conducted correlation and logistic regression analyses on whether obstruction affects cardiovascular events and found that the positive events were weakly correlated with R_LVOT-PG (r = 0.35), and P_LVOT-PG had an “inverted U” shape relationship with positive events, which suggested that the probability of cardiovascular events does not simply increase with increasing outflow tract obstruction. Moreover, in the further multivariate progressive logistic regression analysis, it was also found that the predictive value of R_LVOT-PG and P_LVOT-PG for cardiovascular events in HCM is not independent, and does not play a decisive role independently. LVOT-PG can be affected by cardiac function. Maurizi et al. [33], in a multidisciplinary center study observed that even if outflow tract obstruction was successfully relieved, some patients would still progress to heart failure. Elliott et al. [34] found that the 5-year survival rate of all-cause mortality in patients with obstruction was significantly correlated with decreased cardiac function (New York Heart Association (NYHA) Class I: 91.0%; NYHA Class II: 83.3%; NYHA Class III/IV: 82.6%, p = 0.002), and that the sudden death mortality rate of asymptomatic left ventricular outflow tract obstruction (LVOTO) patients was relatively low. Maron et al. [35] found that although the overall mortality rate of patients with obstruction was significantly higher than that of patients without obstruction (relative risk, 2.0; p = 0.001), the likelihood of severe symptoms and death associated with obstruction did not increase with the increased threshold ($\ge$30 mmHg) of the outflow gradient. Based on the fact that all the HCM patients enrolled in this study were able to complete the treadmill exercise stress ultrasound test, their exercise capacity and cardiac function were both at NYHA Class I, suggesting that the factor of “outflow tract obstruction” that we have been continuously concerned about for a long time may only be a manifestation. On the one hand, obstruction causes sharp and abnormal changes in short-term cardiac function and hemodynamics, as well as abnormal myocardial electrical activity. On the other hand, it triggers myocardial structure remodeling and reshaping of myocardial mechanical function and electrical activity, all of which may have a greater impact on cardiovascular events in HCM. This will require additional in-depth multimodal research at the molecular level to further elucidate these issues.

A retrospective study involving more than 3000 HCM patients showed that abnormal resting LV-GLS was related to adverse cardiovascular events [36]. In addition, a large study by Yang et al. [37] also confirmed that the impairment of resting GLS was significantly related to adverse cardiovascular events in HCM, and the GLS can assist in identifying and determining the risk degree of HCM, which also partially explains the high rate of SCD in young HCM patients. At the same time, left atrial strain has been proven to have predictive value for adverse cardiovascular events in clinical practice [16, 38]. The quantitative analysis of left atrial strain can also provide a basis for the diagnosis, classification, prediction of new atrial fibrillation, and identification of adverse cardiovascular events in subclinical left atrial dysfunction [39]. In this 5-year follow-up study, it was found that the GLS and its reserve, as well as the LASr and its reserve in the positive event group were worse than those in the non-event group. Logistic regression and collinearity analysis showed that R_4D_LASr and P_2D_GLS had independent predictive value for positive cardiovascular events. In Fig. 5 and Table 7, the R_4D_LASr and P_2D_GLS showed strong predictive ability for cardiovascular events in HCM. According to classical pathophysiology, abnormal diastolic function is not only an early pathophysiological feature of HCM but also one of the important reasons for the deterioration and progression of HCM. It strongly implies that positive events may be simultaneously influenced by both the systolic and diastolic functions of the ventricles and atria, with more emphasis on the diastolic function, and that these positive clinical events may be the result of a combination of multiple factors.

Speckle tracking echocardiography suffers from a number of relevant technical limitations, such as the intervendor variability, dependency on optimal image quality, sufficient frame rates (typically $\ge$40 fps), operator expertise, loading conditions and chest wall conformation [40, 41, 42]. In future studies, we plan to expand the sample size, extend the follow-up duration, and incorporate a broader range of ultrasound systems with advanced strain analysis algorithms.

In the study, the endpoint events such as heart failure, ventricular tachycardia, atrial fibrillation, and syncope were direct manifestations of the natural progression or deterioration of the disease. Implantable cardioverter-defibrillator (ICD) implantation was a decision made by clinicians through comprehensive judgment, which indirectly reflects the deterioration of the disease to a certain extent. The main purpose of this study at this stage is to screen risk factors and prediction models. Therefore, “ICD implantation” is set as the “Primary Prevention”, which still inevitably leads to “bias” and “endpoint heterogeneity” at the current stage. In the subsequent stage, on the one hand, classical survival analysis or competitive risk analysis will be conducted, and on the other hand, more sensitive multimodal potential new markers will be added to continuously improve and enrich the results and conclusion.

At this stage of the study, no additional biological markers were included, such as B-type natriuretic peptide (BNP). This study considers that the pathophysiological mechanism of HCM shows that abnormal diastolic function is an early pathophysiological feature and also one of the important reasons for the deterioration and development of HCM. When designing the research protocol, the non-invasive method of stress ultrasound was given more priority. At the current stage, we hope to explore new non-invasive imaging parameters to predict and assess the risk of HCM. BNP is a “molecular marker” reflecting cardiac load and cardiac function. Its pathophysiological core is the response to increased ventricular pressure/volume load. Although it is widely used in the diagnosis, assessment and management of heart failure in clinical practice, it has certain limitations in the specific risk assessment of HCM: Elevated BNP is not only seen in HCM, but is also related to multiple factors such as age, renal insufficiency, atrial fibrillation, and hypertension. BNP has certain correlations with some parameters. After inclusion, it may reduce the model efficiency due to collinearity. At this stage of the study, all the enrolled patients were able to perform the treadmill exercise test, and the cardiac function was NYHA Class I. The clinical endpoint events were Syncope, Heart failure, ICD or pacemaker implantation, ventricular tachycardia, and atrial fibrillation. BNP is more likely to reflect events related to cardiac insufficiency (such as heart failure), has a relatively low predictive sensitivity for arrhythmia events such as SCD and atrial fibrillation, and has a weak correlation with the “composite cardiovascular events” of concern in this study. Although BNP was not included in this present study, its clinical value is not denied. Instead, it is a phased choice in the research design. Our aim is to integrate multi-omics and multimodal imaging technologies, further optimize the research protocol, and explore deeper mechanisms to improve the assessment of cardiac function and stratify the prognosis of HCM patients, ultimately enhancing their diagnosis, treatment and long-term management.

ESE provides critical information to predict risk factors in HCM patients: R_4D_LASr and P_2D_GLS had independent predictive value for positive cardiovascular events, which can assist in the clinical assessment, and identification of high-risk HCM patients, promote the individualized and precise risk stratification of HCM in clinical practice, and improve the long-term prognosis.

Since this study is still ongoing, the data are not currently fully publicly available. Upon completion of the subsequent research, the data can be made accessible upon reasonable request.

YS, LY, and CL: conception and design. LY: administrative support. CL and YS: provision of study materials or patients. YS, QP and BC: collection and assembly of data, and data analysis and interpretation. All authors contributed to manuscript writing and final approval of the manuscript. All authors have participated sufficiently in the work and agreed to be accountable for all aspects of the work.

The studies involving human participants were reviewed and approved by the Human Body Research Institution Committee of Sichuan Provincial People’s Hospital (approval No.185) and were performed in accordance with the Declaration of Helsinki (as revised in 2013). All patients signed the informed consent. The study procedures strictly followed the rules for the protection of patient privacy, and all data were anonymized.

Special thanks to “Bo Chen” for his assistance during the analysis and revision process.

This study was supported by Sichuan Science and Technology Program (Grant No. 2025ZNSFSC1699).

The authors declare no conflict of interest.