Low septal to lateral wall 18F-FDG ratio is highly associated with mechanical dyssynchrony in non-ischemic CRT candidates

Myocardial perfusion and flow reserve in the asynchronous heart: mechanistic insight from a computational model

The tight coupling between myocardial oxygen demand and supply has been recognized for decades, but it remains controversial whether this coupling persists under asynchronous activation, such as during left bundle branch block (LBBB). Furthermore, it is unclear whether the amount of local cardiac wall growth, following longer-lasting asynchronous activation, can explain differences in myocardial perfusion distribution between subjects. For a better understanding of these matters, we built upon our existing modeling framework for cardiac mechanics-to-perfusion coupling by incorporating coronary autoregulation. Regional coronary flow was regulated with a vasodilator signal based on regional demand, as estimated from regional fiber stress-strain area. Volume of left ventricular wall segments was adapted with chronic asynchronous activation toward a homogeneous distribution of myocardial oxygen demand per tissue weight. Modeling results show that 1) both myocardial oxygen demand and supply are decreased in early activated regions and increased in late-activated regions; 2) but that regional hyperemic flow remains unaffected; while 3) regional myocardial flow reserve (the ratio of hyperemic to resting myocardial flow) decreases with increases in absolute regional myocardial oxygen demand as well as with decreases in wall thickness. These findings suggest that septal hypoperfusion in LBBB represents an autoregulatory response to reduced myocardial oxygen demand. Furthermore, oxygen demand-driven remodeling of wall mass can explain asymmetric hypertrophy and the related homogenization of myocardial perfusion and flow reserve. Finally, the inconsistent observations of myocardial perfusion distribution can primarily be explained by the degree of dyssynchrony, the degree of asymmetric hypertrophy, and the imaging modality used.NEW & NOTEWORTHY This versatile modeling framework couples myocardial oxygen demand to oxygen supply and myocardial growth, enabling simulation of resting and hyperemic myocardial flow during acute and chronic asynchronous ventricular activation. Model-based findings suggest that reported inconsistencies in myocardial perfusion and flow reserve responses with asynchronous ventricular activation between patients can primarily be explained by the degree of dyssynchrony and wall mass remodeling, which together determine the heterogeneity in regional oxygen demand and, hence, supply with autoregulation.

Deep Learning-Based Image Registration in Dynamic Myocardial Perfusion CT Imaging

Registration of dynamic CT image sequences is a crucial preprocessing step for clinical evaluation of multiple physiological determinants in the heart such as global and regional myocardial perfusion. In this work, we present a deformable deep learning-based image registration method for quantitative myocardial perfusion CT examinations, which in contrast to previous approaches, takes into account some unique challenges such as low image quality with less accurate anatomical landmarks, dynamic changes of contrast agent concentration in the heart chambers and tissue, and misalignment caused by cardiac stress, respiration, and patient motion. The introduced method uses a recursive cascade network with a ventricle segmentation module, and a novel loss function that accounts for local contrast changes over time. It was trained and validated on a dataset of n = 118 patients with known or suspected coronary artery disease and/or aortic valve insufficiency. Our results demonstrate that the proposed method is capable of registering dynamic cardiac perfusion sequences by reducing local tissue displacements of the left ventricle (LV), whereas contrast changes do not affect the registration and image quality, in particular the absolute CT (HU) values of the entire CT sequence. In addition, the deep learning-based approach presented reveals a short processing time of a few seconds compared to conventional image registration methods, demonstrating its application potential for quantitative CT myocardial perfusion measurements in daily clinical routine.

What Have We Learned in the Last 20 Years About CRT Non-Responders?

Although cardiac resynchronization therapy (CRT) has become well established in the treatment of heart failure, the management of patients who do not respond after CRT remains a key challenge. This review will summarize what we have learned about non-responders over the last 20 years and discuss methods for optimizing response, including the introduction of novel therapies.

Can nuclear imaging accurately detect scar in ischemic cardiac resynchronization therapy candidates?

BACKGROUND

Accurate scar assessment is crucial in cardiac resynchronization therapy (CRT) candidates, since its presence is a negative predictor for CRT response. Therefore, we assessed the performance of different PET parameters to detect scar in CRT candidates.

METHODS

Twenty-nine CRT candidates underwent 18F-fluorodeoxyglucose (18F-FDG)-PET/computed tomography (CT), resting 13N-NH3-PET/CT and cardiac magnetic resonance (CMR) prior to CRT implantation. Segmental 18F-FDG uptake, late 13N-NH3 uptake and absolute myocardial blood flow (MBF) were evaluated for scar detection using late gadolinium enhancement (LGE) CMR as reference. A receiver operator characteristic (ROC) area under the curve (AUC) ≥0.8 indicated a good accuracy of the methods evaluated.

RESULTS

Scar was present in 111 of 464 segments. None of the approaches could reliably identify segments with nontransmural scar, except for 18F-FDG uptake in the lateral wall (AUC 0.83). Segmental transmural scars could be detected with all methods (AUC ≥ 0.8), except for septal 18F-FDG uptake and MBF in the inferior wall (AUC < 0.8). Late 13N-NH3 uptake was the best parameter for transmural scar detection, independent of its location, with a sensitivity of 80% and specificity of 92% using a cutoff of 66% of the maximum tracer activity.

CONCLUSIONS

Late 13N-NH3 uptake is superior to 13N-NH3 MBF and 18F-FDG in detecting transmural scar, independently of its location. However, none of the tested PET parameters was able to accurately detect nontransmural scar.

Left ventricular regional glucose metabolism in combination with septal scar extent identifies CRT responders

BACKGROUND

Cardiac resynchronization therapy (CRT) is effective in selective heart failure (HF) patients, but non-response rate remains high. Positron emission tomography (PET) may provide a better insight into the pathophysiology of left ventricular (LV) remodeling; however, its role for evaluating and selecting patients for CRT remains uncertain.

PURPOSE

We investigated if regional LV glucose metabolism in combination with myocardial scar could predict response to CRT.

METHODS

Consecutive CRT-eligible HF patients underwent echocardiography, cardiac magnetic resonance (CMR), and ¹⁸F-fluorodeoxyglucose (FDG) PET within 1 week before CRT implantation. Echocardiography was additionally performed 12 months after CRT and end-systolic volume reduction ≥ 15% was defined as CRT response. Septal-to-lateral wall (SLR) FDG uptake ratio was calculated from static FDG images. Late gadolinium enhancement (LGE) CMR was analyzed semi-quantitatively to define scar extent.

RESULTS

We evaluated 88 patients (67 ± 10 years, 72% males). ¹⁸F-FDG SLR showed a linear correlation with volumetric reverse remodeling 12 months after CRT (r = 0.41, p = 0.0001). In non-ischemic HF patients, low FDG SLR alone predicted CRT response with sensitivity and specificity of more than 80%; however, in ischemic HF patients, specificity decreased to 46%, suggesting that in this cohort low SLR can also be caused by the presence of a septal scar. In the multivariate logistic regression model, including low FDG SLR, presence and extent of the scar in each myocardial wall, and current CRT guideline parameters, only low FDG SLR and septal scar remained associated with CRT response. Their combination could predict CRT response with sensitivity, specificity, negative, and positive predictive value of 80%, 83%, 70%, and 90%, respectively.

CONCLUSIONS

FDG SLR can be used as a predictor of CRT response and combined with septal scar extent, CRT responders can be distinguished from non-responders with high diagnostic accuracy. Further studies are needed to verify whether this imaging approach can prospectively be used to optimize patient selection.