Probing myocardial blood oxygenation reserve with controlled hypercapnia using BOLD CMR

Hsin‐Jung Yang, Roya Yumul, Richard Tang, Ivan Cokic, Michael A. Klein, Avinash Kali, Olivia Sobczyk, Behzad Sharif, Jun Tang, Xiaoming Bi, Sotirios A. Tsaftaris, Debiao Li, James K. Min, Daniel S. Berman, Antionio Hernandez Conte, Joseph A. Fisher, Rohan DharmakumarJournal of Cardiovascular Magnetic Resonance 2014 · DOI

Abstract

Traditional cardiac stress testing relies heavily on pharmacological vasodilators like adenosine to induce hyperemia. However, these agents are often contraindicated and can cause adverse side effects. This study investigates the feasibility of using precisely controlled arterial $CO_2$ (hypercapnia) as a safe, non-invasive alternative. By employing Blood-Oxygen-Level-Dependent (BOLD) Cardiovascular Magnetic Resonance (CMR), researchers evaluated the myocardial response to hypercapnia in humans and canine models. The results indicate that a modest, targeted increase in $PaCO_2$ produces a hyperemic response statistically equivalent to adenosine, even in the presence of coronary stenosis.

Introduction: The Search for a Natural Stressor

To diagnose coronary artery disease, clinicians must assess the heart's ability to increase blood flow under stress—a state known as hyperemia. Currently, over half of these assessments utilize drugs like adenosine. While effective, these pharmacological stressors require intravenous access and are associated with side effects such as chest pain and dyspnea. Furthermore, they are unsuitable for patients with specific pulmonary or cardiac conditions.

Carbon dioxide ($CO_2$) is a potent endogenous vasodilator. This study explores whether controlled inhalation of $CO_2$ can serve as a "natural" stressor, replicating the diagnostic utility of pharmacological agents without their associated risks.

Methodology

The researchers utilized BOLD CMR, an imaging technique sensitive to changes in the ratio of oxygenated to deoxygenated hemoglobin. As blood flow increases, the "flushing" of tissue with oxygenated blood leads to a rise in MRI signal intensity.

The experimental protocol compared two hyperemic triggers:

  1. Pharmacological: Intravenous adenosine infusion ($140\ \mu g/kg/min$).
  2. Physiological: Controlled hypercapnia, where a specialized breathing circuit was used to precisely target increases in the partial pressure of arterial $CO_2$ ($PaCO_2$).

The study involved 18 human volunteers to assess tolerability and 18 canine models (including those with surgically induced LAD artery stenosis) to validate diagnostic performance in diseased states.

Results and Visual Evidence

The study identified a "10 mmHg threshold"—an increase in $PaCO_2$ of approximately $10\ mmHg$ above baseline—as the optimal level for inducing significant vasodilation while remaining well-tolerated by subjects.

Statistical Parity with Adenosine

In human subjects, the BOLD CMR response to a $10\ mmHg$ $CO_2$ increase showed no statistically significant difference from the response to adenosine ($p = 0.7$). Similarly, in canine models, an $11\ mmHg$ increase matched the adenosine-induced response ($p = 0.4$). Crucially, in territories affected by arterial stenosis, both methods correctly identified reduced oxygenation reserve with similar accuracy ($p = 0.7$).

Visualizing Hyperemia

The physiological impact of hypercapnia is clearly visible in CMR imaging. As shown in Figure 1, there is a marked increase in signal intensity across the myocardium during hypercapnia compared to baseline.

BOLD CMR signal intensity increases during hypercapnia compared to baseline in a canine model.

The color overlays in the short-axis (A) and long-axis (B) views highlight the uniform increase in signal intensity (represented by the shift toward warmer colors), mirroring the effects typically seen with clinical vasodilators.

Discussion and Limitations

These findings suggest that "breathing as a stressor" is a viable physiological equivalent to pharmacological induction for cardiac BOLD CMR. This approach could expand stress-testing eligibility to patients for whom adenosine is contraindicated.

However, several hurdles remain for clinical adoption. The current sample size is small, and larger trials are necessary to confirm safety across broader patient populations. Additionally, the requirement for specialized breathing hardware to precisely target $PaCO_2$ levels must be addressed to ensure the technique is user-friendly and scalable for routine clinical environments. Finally, while generally well-tolerated, the sensation of "air hunger" associated with elevated $CO_2$ may require psychological management in some patients.

Summary updated

Abstract

Background Background More than half of the cardiac stress tests require pharmacologic vasodilators for induction of hyperemia to assess myocardial perfusion, but carry the potential for side effects and are contraindicated in many patients considered for testing. We evaluated the feasibility of a non-invasive and safe stress-testing paradigm using a precisely targeted partial pressure of arterial CO2 (PaCO2) to induce myocardial hyperemia, and compared this response to intravenous adenosine. Methods Dose-response studies were performed on spontaneously breathing humans (n = 18), and canines (n = 18) with and without surgically implemented coronary stenosis to determine the optimal increase in PaCO2 required to
BibTeX
@article{yang2014probing,
  author = {Hsin‐Jung Yang and Roya Yumul and Richard Tang and Ivan Cokic and Michael A. Klein and Avinash Kali and Olivia Sobczyk and Behzad Sharif and Jun Tang and Xiaoming Bi and Sotirios A. Tsaftaris and Debiao Li and James K. Min and Daniel S. Berman and Antionio Hernandez Conte and Joseph A. Fisher and Rohan Dharmakumar},
  title = {{Probing myocardial blood oxygenation reserve with controlled hypercapnia using BOLD CMR}},
  journal = {Journal of Cardiovascular Magnetic Resonance},
  year = {2014},
  doi = {10.1186/1532-429x-16-s1-o14},
}
Figure