Hysteretic device characteristics indicate cardiac contractile state for guiding mechanical circulatory support device use

Brian Y. Chang, Zhengyang Zhang, Kimberly Feng, Noam Josephy, Steven P. Keller, Elazer R. EdelmanIntensive 2021 · DOI

Abstract

Mechanical circulatory support hysteretic behavior can be well modeled using device signals and used to estimate contractility changes. Contractility estimate is correlated with other accepted metrics, captures temporal trends that elucidate changing cardiac state, and is able to accurately indicate changes in inotropy. Inherently available during MCS deployment, this measure will guide titration and inform need for further intervention.

Introduction

In the management of cardiogenic shock or acute heart failure, mechanical circulatory support (MCS) devices like the Impella micro-axial pump are frequently deployed to maintain systemic perfusion. A primary clinical challenge in the intensive care unit is the real-time assessment of native cardiac recovery to determine the optimal timing for "weaning" a patient from the device. Current methods for assessing cardiac contractility—such as intermittent echocardiography or invasive pressure-volume catheters—are either operator-dependent or too invasive for routine continuous monitoring. This study addresses the need for a continuous, automated, and non-invasive measure of cardiac state derived directly from existing MCS device signals.

Methods and Modeling

The researchers developed a method to estimate cardiac contractility by analyzing the hysteresis inherent in the interaction between the heart and the MCS device. In a static environment, a pump's pressure-power relationship is predictable. However, within a beating heart, the cyclical contraction and relaxation create a time-lagged relationship between motor current and pressure head, forming a hysteretic loop.

Figure 1: Overview of device-heart interactions and the quantification of hysteresis as a contractility metric.

Using Euler fluid equations, the team modeled the pressure head ($\Delta P$) as a function of motor current ($i$), its rate of change ($di/dt$), and motor speed ($\omega$): $$\Delta P = a \cdot i + b \cdot \frac{di}{dt} + c\omega^2$$

By gating these signals to the cardiac cycle, they isolated the systolic portion of the hysteresis loop. The area enclosed by this systolic loop was hypothesized to serve as a proxy for native cardiac contractility. This "Systolic Area" metric was validated against invasive "gold standard" measures, including the End-Systolic Pressure-Volume Relationship (ESPVR) and the maximum rate of pressure change ($dP/dt_{max}$).

Results and Validation

The proposed metric was validated using two distinct datasets:

  • Animal Model: Four Yorkshire swine underwent acute interventions, including pharmacological challenges (Epinephrine to increase contractility, Esmolol to decrease it) and a heart failure model. This provided 113 high-fidelity data points.
  • Clinical Data: Retrospective data from eight human patients undergoing high-risk Percutaneous Coronary Intervention (PCI) with Impella support, totaling over 10,940 data points.

Key findings include:

  • High Correlation: The "Systolic Area" metric showed excellent correlation with $dP/dt_{max}$ in both animal ($r = 0.95$) and human ($r = 0.88$) models. It also correlated strongly with ESPVR ($r = 0.84$).
  • Diagnostic Accuracy: ROC analysis demonstrated a high ability to detect directional changes in contractility, with an Area Under the Curve (AUC) of 0.92 in animals and 0.90 in patients.
  • Trend Tracking: The metric successfully captured both rapid physiological changes (e.g., drug boluses) and gradual declines in function during progressive heart failure.

Discussion

The most significant insight from this work is the transformation of the MCS device from a purely therapeutic tool into a diagnostic sensor. Traditionally, the non-linear relationship between pump parameters and physiological variables—known as hysteresis—was viewed as a technical hurdle to overcome. By embracing this complexity, the authors have shown that the degree of hysteresis is a direct reflection of the heart's contractile strength.

Specifically, isolating the systolic area allows for a measurement that is relatively independent of preload, a common confounding factor in other cardiac metrics. This "sensor-less" approach requires no hardware modifications to existing clinical setups, as it leverages signals already recorded by the device controller. This paves the way for "smart" circulatory support systems that can dynamically respond to a patient's physiological state, potentially automating weaning protocols or alerting clinicians to early signs of heart failure progression.

Limitations and Future Work

While the current model is tuned for axial-flow pumps like the Impella, the underlying concept of hysteresis is likely applicable to other MCS designs, such as centrifugal pumps, though specific coefficients would need recalibration. Future work will focus on determining if absolute contractility values can be derived without baseline normalization and conducting larger prospective studies to validate the metric's predictive value for clinical outcomes.

Conclusion

Quantifying the systolic portion of the heart-device hysteresis loop provides a continuous, reliable marker of native contractility. This inherently available measure can guide clinical titration of pump support and prompt timely therapeutic or diagnostic interventions, filling a critical gap in ICU monitoring.

Summary updated

Abstract

Mechanical circulatory support hysteretic behavior can be well modeled using device signals and used to estimate contractility changes. Contractility estimate is correlated with other accepted metrics, captures temporal trends that elucidate changing cardiac state, and is able to accurately indicate changes in inotropy. Inherently available during MCS deployment, this measure will guide titration and inform need for further intervention.
BibTeX
@inproceedings{chang2021hysteretic,
  author = {Brian Y. Chang and Zhengyang Zhang and Kimberly Feng and Noam Josephy and Steven P. Keller and Elazer R. Edelman},
  title = {{Hysteretic device characteristics indicate cardiac contractile state for guiding mechanical circulatory support device use}},
  booktitle = {Intensive},
  year = {2021},
  doi = {10.1186/s40635-021-00426-3},
}
Figure