RT-3DE TEE combined with STI to evaluate the correlation between left atrial and left auricular function and stroke in patients with non-valvular atrial fibrillation: a clinical investigation

Xinyi Li, Shiyi Huang, Chang Zhou, Xingyu Liu, Runyu Zhu, Wenshu Hu, Haotian Hu, Dian Shen, Ao-Yi ZhangResearch Square 2025 · DOI

This clinical investigation explores how advanced cardiac imaging can better predict stroke risk in patients with non-valvular atrial fibrillation (NVAF). By combining 3D ultrasound of the heart's interior with "speckle tracking" to measure muscle elasticity, the researchers identified specific mechanical markers that outperform standard clinical risk scores.

TL;DR

Researchers found that combining real-time 3D imaging with speckle tracking significantly improves our ability to predict strokes in patients with atrial fibrillation. By measuring the specific volume and "stretchiness" (strain) of the left atrium and its appendage, they achieved a predictive accuracy (AUC of 0.854) that far exceeds current clinical scoring methods.

What Problem Does This Paper Address?

Atrial fibrillation (AF) is a leading cause of cardiogenic embolic stroke (CES). In AF, the heart's upper chambers don't beat effectively, allowing blood to pool and form clots—most often in a small pouch called the left atrial appendage (LAA).

Currently, doctors use the CHA₂DS₂-VASc score to estimate stroke risk. While useful, this score relies entirely on clinical history (like age, hypertension, and diabetes) and ignores the actual physical state of the heart. Many patients with "low" clinical scores still suffer strokes because their heart's anatomy has remodeled in ways the score can't see. This paper seeks to fill that gap by using advanced imaging to quantify the heart's structural and functional decline.

Methods

The study utilized two sophisticated imaging techniques:

  • RT-3D TEE (Real-time 3D Transesophageal Echocardiography): An ultrasound probe is placed in the esophagus (right behind the heart) to create high-resolution 3D maps of the left atrium (LA) and LAA. This allows for precise volume measurements that 2D imaging often misses.
  • 2D-STI (2D Speckle Tracking Imaging): This technique tracks the movement of natural "speckles" in the heart tissue during the cardiac cycle. It measures strain—the degree of deformation or "stretchiness" of the heart muscle—which is a sensitive indicator of early dysfunction.

The researchers integrated these tools to evaluate both the "reservoir" function of the atrium and the specific emptying/filling dynamics of the appendage.

Datasets

The study was a clinical investigation involving 221 participants:

  • 125 NVAF patients: Subdivided into 55 who had experienced a stroke (CES) and 70 who had not.
  • 96 healthy controls: Individuals in normal sinus rhythm.
  • Follow-up: Patients were monitored for a mean duration of 9.1 months to track clinical outcomes.

Key Insights and Ideas

  • Mechanical Markers > Clinical History: The study highlights that physical remodeling—specifically the dilation of the LAA and the loss of elasticity in the LA—are more direct predictors of clot formation than traditional risk factors like age or blood pressure.
  • The "Reservoir" Matters: Left atrial reservoir strain (LASr) emerged as a critical metric. When the atrium loses its ability to stretch and store blood effectively, the risk of stasis and subsequent clotting rises sharply.
  • Synergistic Prediction: The "aha moment" is that imaging doesn't have to replace clinical scores; it augments them. The highest predictive power came from a model that combined the CHA₂DS₂-VASc score with specific imaging parameters (LASr and LAAVImax).

Key Results

  • Independent Risk Factors: Multivariate analysis identified three key independent predictors for stroke: a CHA₂DS₂-VASc score ≥ 2, impaired LA reservoir strain (LASr), and an increased LAA maximum volume index (LAAVImax).
  • Superior Accuracy:
    • The standard clinical score (CHA₂DS₂-VASc) had an AUC of 0.685.
  • The LAA volume index (LAAVImax) alone was better, with an AUC of 0.803.
  • The combined model (Clinical score + LASr + LAAVImax) reached an AUC of 0.854, with a high sensitivity of 90.9%.
  • Significant Dysfunction: Patients in the stroke group showed significantly higher volumes and lower strain rates compared to both the healthy controls and the AF patients who had not suffered a stroke.

Limitations and Future Work

  • Single-Center Study: The data comes from one hospital, which may limit the generalizability of the findings to broader, more diverse populations.
  • Sample Size: While sufficient for statistical significance, a larger cohort would strengthen the predictive model.
  • Confounders: The authors note that residual confounders from other potential causes of stroke might still be present.
  • Next Steps: Future research should involve multi-center, large-scale prospective trials to validate these imaging markers as a standard part of stroke prevention protocols in AF patients.

Summary updated

Abstract

<title>Abstract</title> <bold>Aim</bold> : Non-valvular atrial fibrillation (NVAF) serves as the primary cause of cardiogenic embolic stroke (CES), and the recurrence rate and mortality of CES are on an annual upward trend. The objective is to explore the structural and functional changes of the left atrial (LA) and left atrial appendage (LAA) in patients with NVAF by means of real-time three-dimensional transesophageal echocardiography (RT-3D TEE) in combination with two-dimensional speckle tracking imaging (2D-STI), and to identify the risk factors for CES. <bold>Methods</bold> : A total of 125 patients with NVAF were recruited from our hospital, among whom 55 had CES and 70 did not. Additionally, 96 controls with sinus rhythm were included. General clinical data, as well as transthoracic and transesophageal echocardiographic data, were collected. RT-3D TEE and 2D-STI were employed to assess the structural and functional characteristics of the LA and LAA. Multivariate logistic regression and receiver operating characteristic (ROC) curve analyses were conducted to evaluate the independent risk factors and their predictive value for CES. <bold>Result</bold> : In comparison with the control group, both NVAF groups exhibited elevated LA and LAA volumes as well as impaired functional parameters, with more pronounced alterations observed in the CES subgroup. Multivariate logistic regression analysis determined that CHA₂DS₂-VASc≥2, impaired LA reservoir strain (LASr), and an increased LAA maximum volume index (LAAVImax) were independent risk factors for CES. ROC curve analysis indicated that LAAVImax (AUC= 0.803) and LASr (AUC=0.720) had greater predictive value than CHA₂DS₂-VASc≥2 alone (AUC=0.685). The combination of all three factors resulted in the highest predictive accuracy (AUC=0.854). <bold>Conclusion</bold> : Patients with NVAF and CES demonstrate more significant dysfunction of the LA and LAA. The LASr and LAAVImax are robust predictors of CES and enhance risk stratification when integrated with clinical scores. <bold>Clinical trial numbe</bold> r: Not applicable.
BibTeX
@inproceedings{li2025rt,
  author = {Xinyi Li and Shiyi Huang and Chang Zhou and Xingyu Liu and Runyu Zhu and Wenshu Hu and Haotian Hu and Dian Shen and Ao-Yi Zhang},
  title = {{RT-3DE TEE combined with STI to evaluate the correlation between left atrial and left auricular function and stroke in patients with non-valvular atrial fibrillation: a clinical investigation}},
  booktitle = {Research Square},
  year = {2025},
  doi = {10.21203/rs.3.rs-7888856/v1},
}
Figure