Misclassification rate of the angiographically identified culprit lesion in NSTEMI

Background

Correct identification of the culprit lesion in NSTEMI is essential, in particular for patients in whom a culprit-only strategy is attractive (e.g., elderly and frail patients). However, when identifying the culprit lesion in NSTEMI, angiography can be ambiguous and correct culprit identification can therefore be challenging when based on angiography, ECG- and echocardiographic changes alone (standard-of-care). In fact, this challenge remains unresolved and is a continuous limitation in guidelines and in the few clinical trials investigating the revascularization strategy in NSTEMI.

Purpose

We aimed to investigate the agreement between angiography and cardiac magnetic resonance (CMR) and optical coherence tomography (OCT) in identifying the culprit lesion in non-ST segment elevation myocardial infarction (NSTEMI).

Methods

In two centres we prospectively enrolled 104 patients. CMR was performed prior to angiography. Operators, blinded to CMR, identified a culprit lesion based on angiography and standard-of-care. OCT was subsequently performed on operator-suspected culprit lesions and stenoses ≥50% diameter. CMR and OCT were reviewed blinded to angiographic culprit identification. Myocardial oedema on CMR was considered the reference standard for a culprit. In the absence of oedema, OCT was used. In case of multiple suspected OCT-lesions, hierarchical criteria for culprit identification were used: acute thrombus > plaque rupture with a cavity > organising thrombus > dissection > calcific nodule.

Results

The majority of included patients were male (75%) at a mean 63 years of age. Obstructive disease was observed in 85 (82%) patients, of which 53 (51%) had multivessel disease. On a patient-level, angiography identified a culprit lesion in 90 (87%) patients, of which CMR/OCT only identified a culprit in 74 (82%) patients. This constituted a moderate overall positive predictive value of angiography, which was found inferior to CMR/OCT in identifying the culprit lesion. On a lesion-level, CMR/OCT identified a different culprit lesion than angiography in 12 (16%) patients. Of these, only one patient did not receive revascularization of the true culprit lesion. Moreover, in the 14 patients without an angiographic culprit, CMR/OCT identified a culprit in 7 (50%) patients. Thus, angiography including standard-of-care falsely identified the culprit lesion in overall 35 (34%) patients: 7 false negatives, 16 false positive at patient-level, and 12 misclassified angiographic culprits on lesion-level (Figure 1). Specifically, OCT identified 13% misclassified culprit lesions in proximal segments, and provided an added diagnostic value.

Conclusions

Angiography misclassified the culprit lesion in one in three patients with NSTEMI with respect to both presence and location. OCT complemented angiography in ambiguous cases which underscores the value of OCT in aiding treatment and diagnosis in NSTEMI.

Funding Acknowledgement

Type of funding sources: Public hospital(s). Main funding source(s): Research Grant at Rigshospitalet, Copenhagen University Hospital, DenmarkNovo Nordisk Foundation, Denmark

Left atrial reservoir strain predicts ischemic stroke after coronary artery bypass grafting independent of postoperative atrial fibrillation

Background

Patients undergoing coronary artery bypass grafting (CABG) surgery face an elevated long-term risk of ischemic stroke.

Left atrial (LA) mechanics including LA strain are known to predict both stroke and atrial fibrillation in other populations. Evidence is conflicting with regards to the clinical significance of postoperative atrial fibrillation (POAF) and its role in the association between LA mechanics and ischemic stroke is unclear.

Purpose

The aim of the study was to investigate the association between LA reservoir strain and long-term ischemic stroke in patients undergoing CABG surgery and whether the presence of POAF modified this relationship.

Methods

Patients undergoing isolated CABG were included from 2006–2011 as part of a retrospective cohort study. Echocardiography was performed prior to surgery. We included patients with available LA reservoir strain measurements, while patients with known atrial fibrillation and moderate to severe valvular disease were excluded. The primary endpoint was ischemic stroke. The association between LA reservoir strain and ischemic stroke was investigated in uni- and multivariable Cox proportional hazards regression models including adjustment for POAF. Poisson regression was used to create a restricted cubic spline model of the continuous association between LA reservoir strain and risk of ischemic stroke.

Results

We included 542 patients in the analysis. Mean age was 67.3±8.9 years, 89 (16.4%) were female, and median EuroSCORE II was 1.31 (IQR 0.93–1.96). Patients with LA reservoir strain <27.2% (median) were older, more likely to have diabetes, and had a higher EuroSCORE II.

In total, 96 patients (17.7%) developed POAF. During a median follow-up period of 3.7 years (IQR 2.7–4.8 years), 21 (3.9%) patients experienced ischemic stroke. In univariable modeling, LA reservoir strain was significantly associated with ischemic stroke (HR 1.10 (95% CI 1.03–1.18, p=0.003) per 1% absolute decrease). After adjustment for EuroSCORE II, LA volume index (LAVi) and prior stroke, LA reservoir strain remained a significant predictor of ischemic stroke (HR 1.09 (95% CI 1.02–1.17, p=0.011) per 1% absolute decrease and HR 4.24 (95% CI 1.37–13.10, p=0.012) for < vs. >median). Adding POAF as a covariate did not alter the significance of LA reservoir strain in the model. In addition, POAF did not modify the association between LA reservoir strain and POAF (p for interaction = 0.07). The association between LA reservoir strain and ischemic stroke persisted in multiple sensitivity analyses including restricting the analysis to patients with normal left atrial volumes (LAVi <34 ml/m2), patients without POAF, and when excluding patients who developed atrial fibrillation during follow-up.

Conclusion

In patients undergoing CABG, LA reservoir strain was an independent long-term predictor of ischemic stroke. The predictive value of LA reservoir strain was unaffected by the presence of POAF.

Funding Acknowledgement

Type of funding sources: None.

Longitudinal change in cardiac structure and function following acute coronary syndrome stratified by culprit coronary artery lesion site

Funding Acknowledgements

Type of funding sources: None.

Background

Acute coronary syndrome (ACS) has adverse consequences for the myocardium and subsequent cardiac function and structure. No reports exist comparing the differences in impact of culprit coronary artery lesion site on longitudinal remodeling and changes left ventricular structure and function.

Method

A total of 299 ACS patients treated with PCI were included in the present study. All patients had two echocardiographic examinations performed. The first was performed median 2 (IQR: 1; 3) days following PCI, while the second was performed median 240 (IQR: 81; 881) days after the first. Patients were grouped based on culprit coronary artery lesion (left anterior descending artery (LAD), right coronary artery (RCA) and circumflex artery (Cx)). Patients with multiple lesions were excluded from the present study. Univariable linear regression analysis was utilised to assess the association between culprit coronary artery lesion site and longitudinal change in cardiac structure and function.

Results

Mean age was 63 ± 11 years and 77% were male. At follow-up, mean left ventricular ejection fraction was 42 ± 9% and global longitudinal strain (GLS) was -13 ± 4%. Culprit coronary artery lesion was allocated as follows; 168 ACS patients were treated in LAD, 95 patients were treated in RCA, and 36 patients were treated in Cx. In the linear regression analysis, LAD patients displayed a greater improvement in GLS (b =-0.116, p = 0.048) compared to the two other lesion sites. LAD patients had the poorest GLS at both baseline and follow-up echocardiography (Figure). RCA lesions were associated with the largest decrease in left atrial maximum volume (LAVmax) (b = -0.156, p = 0.011) and the largest increase in relative wall thickness (RWT) (b = 0.139, p = 0.030), consequently resulting in an LAVmax smaller and an RWT larger at follow-up than other lesion sites (Figure). Lastly, Cx lesions were significantly associated with the largest decrease in ratio between peak early diastolic transmitral flow velocity and peak early diastolic mitral annular tissue velocity (E/e’) (b = -0.262, P <0.001). Cx lesion patients were observed to have elevated E/e’ at baseline, which generally normalised at follow-up (Figure).

Conclusion

The present study suggests that culprit coronary artery lesion site has a differential impact on cardiac remodeling. This information can potentially aid the clinical understanding of cardiac structure and function following ACS according to coronary artery lesion site.

Abstract Figure