Prognostic Value of Computed Tomography-Derived Extracellular Volume in TAVR Patients With Low-Flow Low-Gradient Aortic Stenosis

Abstract

Objectives: The association between extracellular volume (ECV) measured by computed tomography angiography (CTA) and clinical outcomes was evaluated in low-flow low-gradient (LFLG) aortic stenosis (AS) patients undergoing transcatheter aortic valve replacement (TAVR).

Background: Patients with LFLG AS comprise a high-risk group with respect to clinical outcomes. Although ECV, a marker of myocardial fibrosis, is traditionally measured with cardiac magnetic resonance, it can also be measured using cardiac CTA. The authors hypothesized that in LFLG AS, increased ECV may be associated with adverse clinical outcomes.

Methods: In 150 LFLG patients with AS who underwent TAVR, ECV was quantified using pre-TAVR CTA. Echocardiographic and clinical information including all-cause death and heart failure rehospitalization (HFH) was obtained from electronic medical records. A Cox proportional hazards model was used to evaluate the association between ECV and death+HFH.

Results: During a median follow-up of 13.9 months (range 0.07 to 28.9 months), there were 31 death+HFH events (21%). Patients who experienced death+HFH had a greater median Society of Thoracic Surgery score (9.9 vs. 4.7; p < 0.01), lower left ventricular ejection fraction (42.3 ± 20.2% vs. 52.7 ± 17.2%; p < 0.01), lower mean transvalvular gradient (24.9 ± 8.9 mm Hg vs. 28.1 ± 7.3 mm Hg; p = 0.04) and increased mean ECV (35.5 ± 9.6% vs. 29.9 ± 8.2%; p < 0.01) compared with patients who did not experience death+HFH. In a multivariable Cox proportional hazards model, increase in ECV was associated with increase in death+HFH, (hazard ratio per 1% increase: 1.04, 95% confidence interval: 1.01 to 1.09; p < 0.01).

Conclusions: In patients with LFLG AS, CTA measured increase in ECV is associated with increased risk of adverse clinical outcomes post-TAVR and may thus serve as a useful noninvasive marker for prognostication.

Keywords: aortic stenosis; computed tomography; extracellular volume; low-flow low-gradient; transcatheter aortic valve replacement.

Figure 1.. Patient Selection

A flow diagram of patient selection for retrospective analysis is shown. From 208 patients who underwent TAVR from Dec 2016 to Dec 2017 and with low-flow, low-gradient (LG) severe AS at Cedars Sinai Medical Center, Los Angeles, CA, 150 underwent pre-contrast and delayed post-contrast CT at the time of their CTA for procedural planning of TAVR of whom 150 patients survived to discharge from the hospital and were included in the study.

Figure 2.. Case example of a patient with severe AS

Regions of interest were drawn in the septum and lateral wall in the mid-left ventricle in pre-contrast (panel A) and delayed post-contrast (panel B) CT images on axial slices. The yellow region of interest represents blood pool in the left ventricle. Values for CT attenuation in the regions of interest in Hounsfield units are shown. Hematocrit was 33%. Calculated ECV for the septum was (1–0.33) x (95–44) ÷ (135–52) =41% and for the lateral wall was (1–0.33) x (97–45) ÷ (135–52) =42%.

Figure 3.. Mean aortic valve gradient compared to ECV

A scatter plot of ECV against mean gradient across the aortic valve illustrates the distribution of ECV in relation to the distribution of mean aortic valve gradients. Patients with ECV ≤33% are represented in orange and patients with ECV >33% are in blue.

Figure 4.. Survival Analysis of TAVR patients stratified by ECV

A plot of survival probability shows that with progression of time from TAVR, patients with ECV>33% (green) had more events (death and heart failure hospitalization) compared to patients with ECV≤33% (red), p<0.01.

Central Illustration.

ECV Quantification from ECG-Gated Cardiac CT Regions of interest were drawn in the septum and lateral wall in the mid-left ventricle in pre-contrast and delayed post-contrast CT images on axial slices. The yellow region of interest represents blood pool in the left ventricle. Values for CT attenuation in the regions of interest in Hounsfield units are shown. Myocardial ECV, expressed as a percentage, was calculated as ECV_CT = (1–hematocrit) × (ΔHU_myo/ΔHU_blood), where ΔHU_myo and ΔHU_blood are given by (HU_delayed – HU_early) of the myocardium and blood pool) Increased ECV is Associated with Death and Heart failure hospitalization A plot of survival probability shows that with progression of time from TAVR, patients with ECV>33% (green) had more events compared to patients with ECV≤33% (red), p<0.01