RPS 1003b - Transcatheter structural cardiac intervention: TAVI and beyond

Purpose:

To determine the prognostic impact of objective and subjective measures of frailty and their relationship to existing validated assessment methods after transcatheter aortic valve implantation (TAVI).

Methods and materials:

Consecutive patients were identified from the Leeds Teaching Hospital Trust TAVI database. Frailty was quantified objectively by measuring total psoas muscle area (TPMA) on routine CT scans and subjectively using a clinician-based judgement of overall frailty, termed the Leeds-Subjective Assessment of Frailty in the Elderly (L-SAFE) score. Post-intervention morbidity and mortality were examined between these scoring systems and validated scoring systems.

Results:

This study included 420 patients having undergone TAVI between January 2013 and December 2015. Median clinical follow-up was 4.0 years (IQR 2.9-5.0). Standardised measurements of TPMA were not associated with overall all-cause mortality (adjusted HR 1.11, 95% CI 0.94 to 1.31, p=0.206). The L-SAFE score was associated with the duration of hospital stay (adjusted regression co-efficient -0.93, 95% CI -1.71 to -0.15, p=0.019) and overall all-cause mortality (adjusted HR 0.66, 95% CI 0.53 to 0.82, p<0.001). Only the L-SAFE score significantly correlated with traditional methods of frailty assessment.

Conclusion:

We demonstrate a purely subjective measure of frailty to be prognostic in determining post-TAVI morbidity and mortality when compared to the measurement of the TPMA.

Limitations:

The retrospective nature of the study meant that there were elements of data that could not be collected. The biggest weakness is the lack of interobserver measurements for the L-SAFE score, however, this is the start of our team's work using this scoring system with the aim of further research including multiple assessor scores.

Ethics committee approval

n/a

Funding:

No funding was received for this work.

Purpose:

To assess the possibility of estimating the aortic valve calcium score (AVCS) based on the angiographic phase of multidetector computed tomography (MDCT) performed before transcatheter aortic valve implantation (TAVI).

Methods and materials:

The study group consisted of 31 patients who had undergone two-phase MDCT of the heart and large vessels before TAVI: the native phase for AVCS assessment and the angiographic phase for assessment of the aortic valve size and assessment of the access route. Mean age was 78.61±6.04 years. The semi-automatic evaluation of AVCS in slices thickness 3 mm and 2 mm (AVCS_native3.0 and AVCS_native2.0) was made on the basis of the native phase. Using the angiographic phase, AVCS was estimated for slices thickness 0.6 mm, with increased differentiation thresholds for contrasting aortic lumen with calcifications in the aortic valve from a typical 130 HU to 500 HU and 600 HU (AVCS_{CTA0.6 500HU}and AVCS_{CTA0.6 600HU}).

Results:

AVCS measured by various methods were: AVCS_native3.0- 3640.58±2289.03, AVCS_native2.0- 3397.09±2138.25, AVCS_{CTA0.6 500HU}- 1871.94±1304.46, and AVCS_{CTA0.6 600HU}- 1586.06±1723.89. There were statistically significant positive linear correlation between AVCS values evaluated on the basis of the native phase and AVCS values estimated on the basis of the angiographic phase: r AVCS_native3.0 vs. AVCS_{CTA0.6 500HU} - 0.90, r AVCS_native3.0 vs. AVCS_{CTA0.6 600HU} - 0.40, r AVCS_native2.0 vs. AVCS_{CTA0.6 500HU} - 0.89, and r AVCS_native2.0 vs. AVCS_{CTA0.6 600HU} - 0.39. Using regression analysis, mathematical formulas for calculating native AVCS based on AVCS values estimated on the basis of angiographic phase were defined: AVCS_native3.0= 644.7+1.8 AVCS_{CTA0.6 500HU,} AVCS_native3.0= 2922.1+0.5 AVCS_{CTA0.6 600HU,} AVCS_native2.0= 689.2+1.6 AVCS_{CTA0.6 500HU,} and AVCS_native2.0= 2789.4+0.5 AVCS_{CTA0.6 600HU}.

Conclusion:

AVCS can be reliably estimated using only the angiographic phase of MDCT of the heart and large vessels.

Limitations:

Small size of study group

Ethics committee approval

Wroclaw Medical University Bioethical Committee.

Funding:

No funding was received for this work.

Purpose:

Coronary artery disease (CAD) is a common concurrent condition in patients receiving transcatheter aortic valve replacement (TAVR), but consensus on the prognostic implications of CAD and its pre-procedural treatment with percutaneous coronary intervention (PCI) is missing. If we could select the coronary segments in which pre-procedural PCI treatment will have a positive effect on the outcome, we could evaluate these segments for CAD on pre-procedural CTA.

Methods and materials:

An institutional TAVR database was complemented with data extracted from coronary angiography and PCI reports. The extent of CAD, lesion location, lesion severity, and the location of PCI were scored. Survival analysis was performed to investigate the impact on mortality within 3 years after TAVR.

Results:

Among the 1,514 included patients, the mean age was 82 years, 55% were female, 31% had diabetes mellitus, and CAD prevalence was 55%. Survival analysis revealed no significant association of CAD extent on mortality, but significant LAD-lesions (≥70% DS) were significantly associated with increased 3-year and short-term mortality (HR:1.6, 95%CI:1.0-2.4 and HR:3.4, 95%CI:1.4-8.1, respectively). Pre-procedural PCI of unselected lesions was significantly associated with an increased 3-year and short-term mortality (HR:1.7, 95%CI:1.0-2.7 and HR:2.3, 95%CI:1.1-4.9, respectively), but selective PCI of proximal lesions resulted in comparable mortality rates.

Conclusion:

Not the extent of CAD but only significant LAD lesions are independently associated with higher mortality after TAVR. Pre-procedural PCI of unselected lesions is also associated with higher mortality, but there was no significant difference in mortality between patients with and without selective pre-procedural PCI of proximal LAD and LM lesions.

Limitations:

Single-centre, non-randomised cohort and we used CAG and PCI reports from local and referral hospitals.

Ethics committee approval

n/a

Funding:

No funding was received for this work.

Purpose:

To investigate the feasibility and accuracy of cardiac multidetector computed tomography (MDCT) prosthesis sizing of ostium secundum atrial septal defect (ASD).

Methods and materials:

70 consecutive patients were included in this retrospective bicentric study between May 2012 and June 2018. They underwent cardiac MDCT and transoesophageal echocardiography (TEE) before transcatheter closure of ASD; dimensions of the defect and peripheral rims were measured. Defect measurements obtained at TEE and MDCT were compared to prosthesis size. Our primary criterion was the comparison of ASD maximal diameter obtained at MDCT (CT-Dmax) to prosthesis size. Intraclass correlation coefficient (ICC), Bland Altman plots, and linear regression were calculated. Intraobserver and interobserver agreement were calculated for MDCT defect measurements.

Results:

43 patients were finally included for defect measurements: 17 did not undergo transcatheter closure and 10 had incomplete data. For CT-Dmax, ICC was 0.88 (CI95% = [0.78 – 0.93]; p=0.06), mean difference was – 0.82 +/- 5.73 mm, and regression linear equation was 0.9 x + 3.2 (p < 0.001). For maximal diameter at TEE versus prosthesis size, ICC was 0.46 (CI95% = [0.21 – 0.61]; p = 0.003), mean difference was – 6.06 +/- 8.26 mm, and regression linear equation was 0.91 x + 7.6 (p < 0.001). Intraobserver and interobserver agreement for CT-Dmax was 0.97 (CI95% = [0.95 – 0.98]) and 0.86 (CI95% = [0.73 – 0.93]), respectively.

Conclusion:

MDCT is a reliable tool for sizing the defect of ostium secundum ASD, making it an accurate alternative to pre-procedural TEE.

Limitations:

Retrospective and small cohort.

Ethics committee approvalFunding:

No funding was received for this work.

Purpose:

Transcatheter mitral valve interventions (TMVI) are emerging as alternative treatments in patients with severe mitral regurgitation (MR) who are non-eligible for surgery. Cardiac CT plays a crucial role in TMVI planning, guiding the assessment of patient suitability, prosthesis sizing, and access planning. However, the entity of the modifications of mitral valve geometry throughout the cardiac cycle and the best phase for annulus sizing are still unclear.

Methods and materials:

44 patients with severe MR candidate to TMVI (3 type-I, 16 type-II, 10 type-IIIa, 15 type-IIIb according to Carpentier classification) and 21 patients without significative MR (control group) were retrospectively evaluated. Each patient underwent a multiphase (0-90%) retrospective ECG-gated cardiac CT. Mitral valve annulus was manually extracted every 10% steps of the R-R interval, according to the recommended D-shape segmentation model.

Results:

Patients with severe MR had enlarged left ventricle (LV-EDV=248,7±112,2 ml) and atrium (LA-ESV=201,6±71,1 ml) with reduced ejection fraction (EF=40,5±15,9%) compared to control patients (p<0,05). Moreover, MR patients showed larger maximum annular area (16,0±4,0 vs 9,5±1,8cm²; p<0,001), particularly in the case of type-II MR. Only the TT diameter showed no significant difference between the two groups (p=0,252). In MR patients, the largest annular area was found predominantly (48%;21/44) during systolic phases (20-40%), while in control patients (67%;14/21) during proto-diastolic phases (40-60%). In the remaining patients, the maximum area was randomly distributed among the remaining phases.

Conclusion:

We can conclude that a multiphase acquisition with the inclusion of the systole and a personalised approach to the segmentation of the mitral annulus could be important for the correct planning of TMVI.

Limitations:

Limited sample size. No comparative data in a population without MR.

Ethics committee approval

This study has been approved by the istitutional review board.

Funding:

No funding was received for this work.

Purpose:

To evaluate the predictors, occurrence, and distribution of TAVI-related silent ischemic brain lesions using diffusion MRI.

Methods and materials:

We investigated 109 consecutive patients from the prospective arm of the RETORIC study (NCT02826200) who underwent brain MRI one week after percutaneous transfemoral aortic valve implantation (TAVI). To determine the occurrence and distribution of periprocedural cerebral ischemic lesions, averaged diffusion-weighted images (trace), and mean diffusivity (MD) maps from the DTI dataset were used. To evaluate the aortic valve calcium score (AVCS), we assessed the preoperative prospectively ECG-triggered cardiac CT (256-slice MDCT). We also assessed the periprocedural factors such as periprocedural time and pre- and postdilatation. Linear regression analysis was performed to identify the independent predictors of TAVI-related ischemic lesions.

Results:

After TAVI, a total of 918 new cerebral ischemic lesions were detected in 100/109 patients (92%). The median ischemic lesion volume was 237 µl (interquartile range 89.5-650) with a median lesion number of 6 (2-10) per patient. 759/918 lesions (83%) were supratentorial (389 left vs. 370 right). Most lesions (628/918, 68%) were subcortical; the left and right MCA territories were the most affected (left: 190/918, 21% vs. right: 177/918,19%). The vast majority of ischemic lesions were clinically silent (96%); 4% of patients had a stroke which was proven by MRI.
The median AVCS was 2,769 (interquartile range 1,858-4,537). Ballon pre-dilatation during TAVI procedure showed significant correlation with increased total ischemic lesion volume (p<0.001, ß = 0.370) on univariate analysis, AVCS, periprocedural time, or post-dilatation were not associated with ischemic load on MRI (p>0.05).

Conclusion:

Periprocedural ischemic lesions are frequent (92%), however, most of them are clinically silent. Only balloon pre-dilatation during TAVI was a predictor of increased ischemic load.

Limitations:

n/a

Ethics committee approval

Written informed consent.

Funding:

No funding was received for this work.

Purpose:

Recent studies show that right ventricular (RV) function is closely associated with the outcome of different cardiac conditions. This study aims to evaluate the value of RV functional parameters obtained from cardiac magnetic resonance imaging (cMR) for predicting mortality in patients undergoing transcatheter aortic valve implantation (TAVI).

Methods and materials:

Patients with severe aortic stenosis were prospectively recruited to undergo cMR before TAVI. After exclusion of 11 patients due to insufficient image quality, cMR images of 112 patients (mean age 82±6 years; 34% women) were analysed using dedicated tissue tracking software (Circle cvi42). Obtained parameters included RV ejection fraction (RV-EF), RV global longitudinal, circumferential strain (RV-GLS, RV-GCS), RV longitudinal systolic velocity (RV-vel), left ventricular ejection fraction (LV-EF), and left ventricular global circumferential strain (LV-GCS).

Results:

Mean follow-up was 3.5±1.8 years. Mortality after one and three years was 14% (16/112) and 29% (32/112), respectively.

RV-EF, RV-GLS, and RV-vel significantly predicted 3-year-all-cause-mortality in univariate Cox-regression (p=0.012, p=0.033, p=0.007), while RV-GCS, LV-EF, and LV-GCS did not (p=0.25, p=0.32, p=0.33). Results remained significant for RV-EF and RV-vel when adjusted for sex and age (p=0.015, p=0.013).

In ROC-analysis, RV-EF, RV-GLS, and RV-vel (AUC=0.64; 0.62; 0.63) outperformed current clinical risk scores (EuroSCORE-II, AV-Score, STS-Score) in predicting 3-year-all-cause-mortality.

Conclusion:

cMR-derived RV functional parameters, especially parameters of longitudinal contraction, predicted mortality in our cohort. They performed better than left ventricular functional parameters and currently used clinical risk scores, and may improve outcome-prediction after TAVI.

Limitations:

Only patients without contraindications for cMR were included, which may potentially bias the study cohort.

Ethics committee approval

The study was approved by the local ethics committee. Patients provided written informed consent.

Funding:

No funding was received for this work.

Purpose:

Preimplantation cardiac-CT is ideal for assessing left-atrial-appendage (LAA) regarding size for selection of proper device and detection of contraindications such as in situ thrombi. Due to poor contrast-medium distribution in arterial-phase, the distinction between artificial filling-defects versus true thrombi is difficult. A delayed-phase is required for confirmation, which needs a second bolus to maintain adequate contrast. In this study, we investigated if a second bolus can be omitted when virtual-monoenergetic-images (VMI) from dual-energy spectral-detector-CT (SDCT) are used to enhance contrast in the delayed-phase.

Methods and materials:

51 consecutive patients undergoing SDCT-imaging of the LAA were prospectively included. Imaging-protocol comprised dual-phase acquisitions with single-bolus contrast-injection. Conventional images (CI) from both phases and 40keV VMI from the delayed-phase were reconstructed. Mean and standard deviation of attenuation were calculated by placing regions-of-interest in the LAA, left-atrium (LA), and pericardial fat. Additionally, two radiologists evaluated conspicuity, homogeneity of contrast distribution, and diagnostic certainty for the presence of a thrombus.

Results:

The difference of mean attenuation between LAA and LA was significantly lower in the delayed-phase (arterial-phase: 10.8±37.7HU, delayed-phase: 2.1±12.5HU, p<0.05), indicating more homogenous contrast distribution. The contrast of the LAA decreased significantly in the delayed-phase, but was significantly improved by VMI; attenuation and image quality parameters were comparable to CI of the arterial-phase (attenuation/SNR/CNR, CI arterial-phase: 263.9±116.9HU/13.9±7.3/6.4±3.9; CI delayed-phase: 106.2±35.2HU/5.8±3.1/1.0±0.9; VMI delayed-phase 256.6±108.5/17.7±10.7/4.7±3.4). The subjective reading confirmed the objective findings and showed significantly improved diagnostic certainty for the evaluation of the LAA when VMI of the delayed phase were used.

Conclusion:

The investigated single-bolus dual-phase acquisition protocol improved visualisation of the LAA; homogeneity of contrast media was higher in the delayed-phase while VMI could maintain high contrast and image quality comparable to arterial-phase.

Limitations:

No echocardiography for evaluaiton of presence of LAA-thrombus was available.

Ethics committee approval

IRB-approved.

Funding:

No funding was received for this work.