CT analysis of left ventricular function predicts short term survival in patients following transcatheter aortic valve implantation: 1-year outcome data

Introduction

Contrast-enhanced cardiac CT routinely performed prior to transcatheter aortic valve implantation (TAVI), allows assessment of cardiac morphology and function. We assessed left ventricular function in CT as a predictor of survival following TAVI.

Methods

500 consecutive patients referred for CT assessment of aortic root anatomy prior to TAVI were screened for inclusion in this analysis. All CT data sets were acquired using a third-generation dual source system. For assessment of aortic root anatomy, acquisitions were acquired using ECG-gated retrospective spiral acquisition and multiphase reconstructions in 10% increments of the cardiac cycle were rendered (slice thickness 0.75, increment 0.4 mm). left ventricular endocardial contours were automatically traced by a dedicated software (syngo. via, Siemens Healthineers, Forchheim, Germany) throughout the cardiac cycle and manually adjusted if required. Global left ventricular function parameters (end-diastolic and end-systolic volumes, stroke volume, cardiac output and ejection fraction) were derived by volumetric assessment.

Results

Out of 500 patients, 439 patients (mean age 80±6 years, 56% males, and Log EuroScore 23±14%) were included in this analysis (61 patients were excluded due to poor CT image quality or missing outcome data). Previous cardiac surgery had been performed in 15% of the patients, 50% had obstructive CAD with previous interventional or surgical revascularisation and 18% had a previous acute coronary syndrome. One-year survival was 83% (366/439 patients). Parameters of left ventricular function were as follows: mean LVEDV 172±56 ml, mean LVESV 78±62 ml, mean LV ejection fraction 59±18%, mean LV stroke volume index 51±22 ml/m2, mean LV cardiac output 6.6±3.3 L/min and LV cardiac index 3.5±1.7 l/min/m2. Cluster analysis of multiple LV-function surrogate parameters identified a group of patients with higher 1-year mortality, with LVEF identified as a predictor of 1-year survival with a cut-0ff of ≥37% associated with an OR 0f 0.52 (95% CI 0.27 to 0.98).

Conclusion

Assessment of left ventricular function using functional CT data sets is feasible and allows risk stratification of patients following TAVI. Among LV functional parameters, CT derived LV-ejection fraction with a cut-off ≥37% identifies patients with better short-term outcome.

Funding Acknowledgement

Type of funding sources: None.

Membranous septum length as predictor for permanent pacemaker implantation after TAVI

Background

High-degree atrioventricular (AV) block and permanent pacemaker (PPM) implantation represent major complications after transcatheter aortic valve implantation (TAVI). Extension of indication for TAVI towards subjects with lower surgical risk requires to reduce the likelihood for the requirement of permanent pacemaker (PPM) implantation. Data on the role of membranous septum length as potential predictor for AV block after TAVI are scarce.

Purpose

We examined the role of membranous septum length as potential predictor for AV block and the need for PPM implantation in a large cohort of consecutive subjects after TAVI.

Methods

In a cohort of 1365 patients without prior permanent pacemaker who underwent transfemoral TAVI, clinical and procedural characteristics were assessed systematically. Based on cardiac computed tomography performed prior to TAVI, membranous septum length was measured orthogonal to the anulus plane (see figure).

Results

Median age of subjects was 81 (IQR 7) years, 50% were male. Logistic euroSCORE was 12.8 (IQR 15.7), STS score 3 (2.7). 9,8% of subjects had a pre-interventional complete right bundle branch block (RBBB). 71% of patients received a balloon-expandable, 29% a self-expandable valve. In n=153 patients (11.2%), PPM implantation was necessary due to high-degree AV block. Median membranous septum length was 2.9 mm (IQR 2.5mm) in subjects who received a PPM versus 4.3 mm (IQR 3.2 mm) in subjects who did not need a PPM (p=0.061). In univariate regression analysis, pre-interventional complete RBBB (p<0.001, OR 7.8), implantation of a self-expandable prosthesis (p=0002, OR 1.7) and membranous septum length (p=0.027, OR 0.9 per 1 mm) were identified as significant predictors for PPM implantation. In multivariate regression analysis, all parameters remained significant, including membranous septum length (p=0.009, OR 0.9 per 1 mm).

Conclusion

In a large cohort of consecutive patients, we were able to confirm the significant independent predictive value of membranous septum length, in addition to pre-interventional complete RBBB or implantation of a self-expandable prosthesis, regarding the occurrence of post-procedural AV block with the need for PPM implantation. The results may contribute to improved risk stratification for potential PPM implantation after TAVI.

Funding Acknowledgement

Type of funding sources: None.

CT-FFR in predicting future cardiovascular events

Background

Computational fluid dynamics offer a non-invasive way for functional assessment of coronary stenosis in CT-FFR. Currently, minimal FFR thresholds used for determining ischemia were established in invasive coronary angiography. It remains unknown whether minimal absolute CT-FFR values are the most adequate tool for predicting future cardiovascular events.

Methods

This analysis enrolled 605 of 767 consecutive patients with suspected CAD undergoing CCTA from 2005 to 2007. Of the 161 (21%) patients excluded, image artifacts were the most common cause (97), followed by software error (43) and chronic total occlusion in (15). CT-FFR was measured using an on-site prototype (cFFR Version 3.0, Siemens Healthineers, Forchheim, Germany) by an experienced observer. CT-FFR values were measured at the proximal and distal end of each coronary segment and the maximum drop in CT-FFR (ΔFFR) over one segment recorded, respectively. Follow up data included major cardiovascular events (non-fatal stroke, non-fatal myocardial infarction, cardiovascular death) and revascularization (PCI or CABG).

Results

This study included 605 patients (60±11 years, 61.2% men) with minimal CT-FFR 0.81±0.14. Mean ΔFFR was 0.13±0.11. MACE occurred in 24 (4.0%) cases, revascularization in 66 (10.9%) during a median follow up of 2823 days (7–8 years). The area under the Receiver operator characteristic (ROC) curve was 0.887 for revascularization during follow up and 0.608 for MACE during follow up. AUC for the combined end point of MACE and revascularization was 0.806.

Conclusion

Maximum drop in CT-FFR (ΔFFR) is a good predictor for future revascularization but less useful in predicting MACE. This underlines the current use of CT-FFR in guiding treatment decisions.

Funding Acknowledgement

Type of funding sources: None.

Predictors of CT-derived FFR in patients with suspected CAD beyond severity of coronary stenosis

Background

Functional assessment of coronary stenosis using computational fluid dynamics is increasingly used, however other factors besides coronary stenosis may affect the results. We assessed several predictors for CT-derived fractional flow reserve (CT-FFR) in patients with suspected coronary artery disease (CAD) undergoing coronary computed tomographic angiography (CCTA).

Methods

2505 consecutive patients with suspected CAD undergoing CCTA from 2008 to 2016 were screened, 1549 were excluded due to incomplete data (934), image quality (345), software error (147) or other reasons (123). Minimal CT-FFR was measured using an on-site prototype (cFFR Version 3.0, Siemens Healthineers, Forchheim, Germany) in coronaries ≥2mm. Several clinical as well as technical criteria were assessed for predicting the minimal CT-FFR per patient.

Results

956 patients (51±12 years, 51.2% men) were included in this analysis. Mean EF was 59.4±7.4%, heart rate 63±9 bpm, systolic (126.5±20mmHg) and diastolic (70±11 mmHg) blood pressure (BP). Regression analysis and ANOVA showed low but significant impact on minimal CT-FFR (mean 0.85±0.10) by EF, aortic valvular dysfunction, heart rate and systolic blood pressure as well as image quality (esp. blooming and image noise). See Tables 1 and 2.

Conclusion

Coronary stenosis may not be the only relevant predictor for CT-FFR. Several clinical criteria (EF, heart rate, BP, aortic valve dysfunction) as well as image criteria (image quality, artifacts) can affect CT-FFR results.

Funding Acknowledgement

Type of funding sources: Foundation. Main funding source(s): Cleveland Clinic Foundation Table 1. ANOVA analysisTable 2. Regression analysis

Computed tomography-derived characterisation of pericoronary,epicardial and paracardial adipose tissue and its association with myocardial ischemia as assessed by computed and invasive fractional flow

Introduction

Increased pericoronary adipose tissue (PCAT) computed tomography (CT) attenuation derived from coronary CT angiography (CTA) around the right coronary artery (RCA) reflects coronary inflammation and relates to cardiac mortality.

Purpose

We aimed to investigate the yet unclear association between CT-derived characterisation of different cardiac adipose tissue compartments and the presence of myocardial ischemia as assessed by fractional flow reserve (FFR).

Methods

133 stable individuals (64 years, 74% male) with coronary artery disease (CAD) underwent CTA including computed FFR (FFR-CT) measurement followed by invasive angiography with FFR (invasive FFR) assessment. The CT attenuation (HU) and volume (mm3) of PCAT were quantified around the RCA (10 to 50 mm from RCA ostium), the proximal 40 mm of the left anterior descending artery (LAD) and the circumflex artery (LCX) with the help of semi-automated software. The per patient PCAT CT attenuation was calculated as followed: (PCAT CT attenuation of RCA+LAD+LCX)/3. Quantification of epicardial adipose tissue (EAT) and paracardial adipose tissue (PAT; all intrathoracic adipose tissue outside the pericardium) were performed in non-contrast cardiac CT data sets using a fully automated deep-learning based algorithm.

Results

Median FFR-CT was 0.86 [0.79, 0.91] and median invasive FFR was 0.87 [0.81, 0.93]. Patients with presence of myocardial ischemia (n=26) defined by a FFR-CT threshold of ≤0.75 showed a significant higher PCAT CT attenuation of RCA (−75.1 HU vs. −81.1 HU, p=0.011) and per patient (−74.5 HU vs. −77.7 HU, p=0.045) than individuals without myocardial ischemia (n=107). In multivariable analysis adjusted for age, BMI, gender and traditional risk factors, both RCA and per patient PCAT CT attenuation were significant predictors of myocardial ischemia as assessed by FFRCT ≤0.75. Between individuals with myocardial ischemia compared to individuals without myocardial ischemia there was no significant difference neither in the volume and CT attenuation of EAT and PAT nor in the PCAT volume of RCA, LAD, LCX and per patient PCAT volume.

Conclusions

Our observations suggest that PCAT CT attenuation instead of PCAT volume, EAT and PAT measures might be associated with the presence of myocardial ischemia as assessed by FFR.

Funding Acknowledgement

Type of funding sources: Foundation. Main funding source(s): German Heart Foundation e.V.

CT-derived characterization of pericoronary, paracardial and epicardial adipose tissue and its association with myocardial ischemia as assessed by adenosine stress CMR perfusion imaging

Funding Acknowledgements

Type of funding sources: Foundation. Main funding source(s): German Heart Foundation e.V.

Introduction

 Increased attenuation of pericoronary adipose tissue (PCAT) around the right coronary artery (RCA) is a new imaging biomarker to detect coronary inflammation derived from routine coronary CT angiography (CTA) and has been shown to be associated with cardiac mortality. Increased volume of epicardial adipose tissue (EAT) has been reported be associated with myocardial ischemia.

Purpose

 We aimed to investigate a potential association between CTA-derived PCAT measures and myocardial ischemia as assessed by adenosine stress CMR perfusion imaging. 

Methods

In this single-centre study 109 stable individuals (mean age of 62 ± 11 years, 77% males) with coronary artery disease underwent CTA followed by adenosine stress CMR perfusion imaging to detect myocardial ischemia. PCAT CT attenuation (HU) and PCAT volume (cm3) was measured around the RCA (10 to 50 mm from RCA ostium), the proximal 40 mm of the left anterior descending artery (LAD) and the circumflex artery (LCX) using semi-automated software. Per patient PCAT CT attenuation was calculated as followed: (PCAT attenuation of RCA + LAD + LCX)/3). Non-contrast CT data sets were used for coronary calcium scoring and the quantification of EAT (located between the myocardial surface and the pericardium) and paracardial adipose tissue (PAT; intrathoracic and outside of the pericardium). 

Results

 Between patients with evidence of significant myocardial ischemia as assessed by adenosine stress CMR perfusion imaging (n = 35) and patients without myocardial ischemia (n = 74) there was no significant difference in the CT attenuation of RCA (-85.3 HU vs. -85.7  HU, p = 0.87), LAD (-84.8 HU vs. -85.7 HU, p = 0.66) and LCX (-82.8 HU vs. -83.2 HU, p = 0.79) as well as in the per patient PCAT CT attenuation (-84.2 HU vs. -84.9 HU, p = 0.76). Neither did patients with myocardial ischemia within the RCA territory show increased RCA PCAT CT attenuation (-87.7 HU vs. -85.3 HU, p = 0.40); nor was such a relationship found for the territory of the LAD (-80.6 HU vs.  85.8 HU, p = 0.11) or LCX (-83.1 HU vs. -83.0 HU, p = 0.99). The CT attenuation of EAT (-77.9 vs. -78.7 HU, p = 0.65) and PAT (-89.9 HU vs. -90.0 HU, p = 0.93) did not differ between patients with myocardial ischemia compared to patients without myocardial ischemia. Between patients with myocardial ischemia and patients without myocardial ischemia there was no significant difference in the volumes of EAT (118.1 cm3 vs. 110.6 cm3, p = 0.55), PAT (279.5 cm3 vs. 240.9 cm3, p = 0.20) and the per patient PCAT volume (1021.9 mm3 vs. 1015.5 mm3, p = 0.90). In logistic regression analysis the volume and CT attenuation of the different intrathoracic fat compartments PCAT, EAT and PAT were not independently associated with the presence of myocardial ischemia (n.s.). 

Conclusions

  In this single-centre study CTA-derived quantified CT attenuation and volume of PCAT, EAT and PAT were not associated with myocardial ischemia as assessed by adenosine stress CMR perfusion imaging.

[ISCHEMIA trial : Is coronary CT angiography the new gatekeeper?]

BACKGROUND

The ISCHEMIA trial systematically compared two major principles in the therapy of coronary artery disease (CAD): medical therapy versus revascularization in patients with a positive noninvasive test for myocardial ischemia. Specifically, it was designed to answer the question whether in patients with demonstrated ischemia, after ruling out left main stenosis by coronary computed tomography angiography (CTA), a routine interventional strategy in addition to optimal medical therapy would improve clinical outcome over an initial strategy of medical therapy alone.

CONCLUSION

Overall, this hypothesis could not be confirmed. In several ways, the trial yields interesting information in the field of cardiac imaging. First, a positive stress test result was not associated with a prognostic benefit of revascularization. Second, even though the evaluation of coronary CTA was not part of the protocol, the good outcome achieved by using coronary CTA as a "gatekeeper" during randomization supports the potential of coronary CTA as a diagnostic tool-both first- and second-line-when CAD is suspected. However, the trial also raises new questions in the field of cardiac imaging which will need to be addressed in future studies.

Resolution of leaflet thrombosis under anticoagulant therapy in patients after transcatheter aortic valve implantation: influence of prosthesis type and size

Introduction

Leaflet thrombosis can frequently be identified by computed tomography angiography (CTA) in patients after transcatheter aortic valve implantation (TAVI). Oral anticoagulation is assumed to lead to resolution of thrombosis. We analyzed the resolution of leaflet thrombosis after TAVI by anticoagulant therapy in serial CTA and assessed the influence of prosthesis type.

Methods

Consecutive TAVI patients who underwent CTA follow-up were screened and individuals with leaflet thrombosis on CTA (defined by the presence of hypo-attenuated leaflet thickening, HALT) in whom oral anticoagulation was initiated and who underwent follow-up CTA were included. The type of anticoagulation was according to physicians' discretion. We assessed the resolution of HALT and compared patients with and without resolution of HALT regarding prosthesis type, prosthesis diameter and type of anticoagulation.

Results

Out of 395 patients screened for participation, 36 patients (mean age 80±7, 67% men) with leaflet thrombosis underwent follow-up CTA at a medial interval of 3 months (IQR: 3; 5.75 months) after anticoagulation was initiated. 36 patients received either vitamin-K antagonists (n=28, 78%) or Factor-Xa Inhibitors (n=8, 22%). A total of 22 (61%) balloon-expandable and 14 (39%) self-expandable transcatheter aortic valves were implanted. Nominal prosthesis diameter was 23, 25, 26, 27 and 29 mm in 7 (19%), 1 (3%), 10 (28%), 7 (19%) and 11 (31%) patients, respectively. 30 patients (83%) with anticoagulation showed resolution of HALT, whereas persistent HALT was detected in 6 patients (17%), of whom 1 patient with balloon-expandable and 5 patients with self-expandable valve. No difference was seen in duration of anticoagulation between patients with and without resolution of HALT (p=0.984). In univariate analysis, prosthesis type (balloon-expandable vs. self-expandable valves) showed a significant association of self-expandable valves with lack of resolution of leaflet thrombosis (p=0.017). In multivariable logistic regression analysis, this association persisted (p=0.043) and was independent of the type of anticoagulation (p=0.660) and prosthesis diameter (p=0.942).

Conclusion

Persisting leaflet thrombosis despite anticoagulation is not infrequent and seems to be associated with prosthesis-type rather than small valve diameter or type of anticoagulation. Further research is necessary to identify structural aortic valve determinants for this finding.

Funding Acknowledgement

Type of funding source: None

The association of pericoronary adipose tissue attenuation with major adverse cardiac events (MACE) and atherosclerosis-relevant inflammatory mediators

Background

Increased attenuation of pericoronary adipose tissue (PCAT) around the right coronary artery (RCA) is a new imaging biomarker to detect coronary inflammation derived from routine coronary CT angiography (CTA).

Purpose

We aimed to investigate a potential association between RCA PCAT attenuation and i) serum levels of atherosclerosis-relevant cytokines, ii) different grades of coronary calcification iii) future coronary revascularization within the same coronary artery and iV) MACE (defined by revascularization, myocardial infarction (MI) and/or cardiac death).

Methods

In 293 stable individuals (59.0±9.8 years, 69% males) with intermediate likelihood for coronary artery disease (CAD) blood was drawn and subsequently analyzed for different atherosclerosis-relevant cytokines interleukin (IL)-2, IL- 4, IL-6, IL-7, IL-8, IL-10, IL-13, Il-15, IL-17, TNF-a, IP-10, CRP, MCP-1, MIP-1a, Eotaxin and GM-CSF, followed by coronary calcium scoring (CCS) in non-contrast CT followed by CTA. PCAT CT attenuation (HU) was measured around the RCA (10 to 50 mm from RCA ostium) and the proximal 40 mm of the left anterior descending artery (LAD) and the circumflex artery (LCX) using semi-automated software. Increased RCA PCAT attenuation was defined as PCAT attenuation above the highest quartile (>−73.5 HU). A long-term follow-up over 9.6 years was performed.

Results

PCAT attenuation was similar in different grades of coronary calcification (CAC=0,-80.3 HU; CAC 1–99, −79.2 HU; CAC 100–400, −79.5 HU; CAC >400, −81.0 HU; p>0.05). Adipocytokine MCP-1 (r=0.23, p<0.01) and pro-inflammatory mediator IL-7 (r=0.12, p=0.04) correlated positively with RCA PCAT attenuation, whereas anti-inflammatory mediators Il-4, -10 and -13 correlated inversely (each r<−0.12, each p<0.05). In patients with increased RCA PCAT attenuation the serum levels of MCP-1 were increased (2.37 vs. 2.20, p<0.01), whereas anti-inflammatory mediators IL-4 and -13 were reduced (each p<0.05). 40 patients experienced MACE during follow-up. In multivariable Cox regression analysis, when adjusted by age, gender, baseline medications, obstructive coronary stenosis and CCS, the highest quartiles of PCAT attenuation are an independent predictor of MACE (HR 7.9, p=0.035). In patients with percutaneous coronary intervention (PCI) of the RCA during follow-up, RCA PCAT attenuation was increased at baseline CTA (−73.1 vs −80.2 HU, p=0.008). In patients with PCI of the LAD or LCX during follow-up, PCAT attenuation of LAD and LCX were not increased at baseline CTA (p>0.05).

Conclusions

The information captured by PCAT attenuation is independent of coronary calcification and showed a trend towards a weak association with serum levels of atherosclerosis-relevant inflammatory biomarkers. Increased RCA PCAT attenuation is an independent predictor of MACE and could guide future prevention strategies in stable patients.

Funding Acknowledgement

Type of funding source: None

Influence of lesion location on the accuracy of CT derived FFR: head-to-head comparison with invasive FFR

Background

The accuracy of CT-derived FFR (FFRCT) has been repeatedly reported. However, the influence of lesion location on accuracy is unknown. Therefore, we evaluated the diagnostic accuracy of FFRCT to detect lesion-specific ischemia and determined the influence of lesion location (proximal vs. distal vessel segments) compared to invasively measured FFR in patients with suspected CAD.

Methods

A total of 136 vessels in which “Dual-Source”-CT coronary angiography had been performed due to suspected CAD and who were further referred for invasive coronary angiography with invasive FFR measurement within three months of the index CT examination were retrospectively identified and screened for inclusion in this analysis. Patients with either left main coronary artery stenoses, bifurcation or ostial stenoses were excluded. Invasive FFR was measured using a pressure wire (CERTUS®, St. Jude Medical, Minnesota, USA or Verrata®, Volcano, San Diego, USA). FFRCT was calculated using an on-site prototype (cFFR Version 3.0, Siemens Healthineers, Forchheim, Germany). All vessels were analyzed by an experienced observer blinded to the results of invasive FFR. Stenoses with invasively measured FFR ≤0.80 were classified as hemodynamically significant. We evaluated the diagnostic accuracy of FFRCT in proximal vs. non-proximal vessel segments. Proximal lesions included stenoses located in segment one, six, eleven and twelve. All other stenoses were categorized as distal lesions.

Results

Out of 136 coronary stenoses, 47 (35%) were located in proximal segments and 89 (65%) lesions were located in distal segments. Compared to invasive FFR, the sensitivity of FFRCT to correctly identify/exclude hemodynamically significant stenoses in proximal vessel segments was 93% (95% CI: 68–99.8%) and the specificity was 100% (95% CI: 89–100%), compared to a sensitivity of 72% (95% CI: 46.5–90%) and a specificity of 87% (95% CI: 77–94%) for FFRCT in distal lesions. The positive predictive value was 100% and the negative predictive value was 97% (95% CI: 82.8–99.5%) compared to a positive predictive value of 59% (95% CI: 42–93.9%) and a negative predictive value of 93% (95% CI: 85.4–96.3%) for proximal vs. distal vessel segment, respectively. This corresponds to an accuracy of 98% vs. 84%, respectively (p=0.02). ROC-Curve analysis showed a slightly higher – albeit non-significant – area under the curve for FFRCT to detect hemodynamic relevance in proximal lesions compared to distal lesions (AUC 0.95, p<0.001 vs. AUC: 0.86, p<0.001, respectively, p=0.2).

Conclusion

FFRCT obtained using an on-site prototype shows overall a high diagnostic accuracy for detecting lesions causing ischemia as compared to invasive FFR with a trend towards better diagnostic performance in proximal vessel segments.

Funding Acknowledgement

Type of funding source: None