Comparative effectiveness of coronary screening in heart valve surgery: Computed tomography versus conventional coronary angiography

Invasive Coronary Angiography Versus Noninvasive Computed Tomography Coronary Angiography as Preoperative Coronary Imaging for Valve Surgery

Coronary computed tomography angiography (CCTA) has emerged as a noninvasive alternative to invasive coronary angiography (ICA) for diagnosing coronary artery disease (CAD). Hence, the question of CCTA's ability to guide surgical decision-making moves into the center of attention. CCTA is specifically powerful in ruling out CAD. We, therefore, performed a meta-analysis and systematic review to compare the clinical end points between patients who received ICA or CCTA to rule out CAD before valve surgery. A total of 3 databases were assessed. The primary outcome was perioperative mortality. Secondary outcomes were acute kidney injury (AKI), myocardial infarction (MI), stroke, and major adverse cardiovascular events (MACEs). The odds ratio (OR) and the respective confidence interval (CI) was calculated. A random-effects model was performed. A total of 5 studies with 6,654 patients qualified for the analysis. There was no significant difference between the 2 groups regarding the primary end point (OR 1.20, 95% CI 0.67 to 2.15, p = 0.53). The secondary outcomes also did not show any significant differences in AKI (OR 1.14, 95% CI 1.14, 0.88 to 1.49, p = 0.32), MI (OR 0.89, 95% CI 0.65 to 1.22, p = 0.45), stroke (OR 1.12, 95% CI 0.48 to 2.60, p = 0.79), or MACEs (OR 1.17, 95% CI 0.86 to 1.59, p = 0.33) incidences. The analysis suggests that CCTA is a safe and reliable noninvasive alternative to ICA for coronary imaging before valve surgery. Conceivable differences in imaging modalities were not associated with increases in perioperative mortality, AKI, MI, stroke, or MACEs.

Development and evaluation of a radiomics model of resting 13N-ammonia positron emission tomography myocardial perfusion imaging to predict coronary artery stenosis in patients with suspected coronary heart disease

BACKGROUND

Coronary angiography (CAG) is usually performed in patients with coronary heart disease (CHD) to evaluate the coronary artery stenosis. However, patients with iodine allergy and renal dysfunction are not suitable for CAG. We try to develop a radiomics machine learning model based on rest 13N-ammonia (13N-NH3) positron emission tomography (PET) myocardial perfusion imaging (MPI) to predict coronary stenosis.

METHODS

Eighty-four patients were included with the inclusion criteria: adult patients; suspected CHD; resting MPI and CAG were performed; and complete data. Coronary artery stenosis >75% were considered to be significant stenosis. Patients were randomly divided into a training group and a testing group with a ratio of 1:1. Myocardial blood flow (MBF), perfusion defect extent (EXT), total perfusion deficit (TPD), and summed rest score (SRS) were obtained. Myocardial static images of the left ventricular (LV) coronary segments were segmented, and radiomics features were extracted. In the training set, the conventional parameter (MPI model) and radiomics (Rad model) models were constructed using the machine learning method and were combined to construct a nomogram. The models' performance was evaluated by area under the curve (AUC), accuracy, sensitivity, specificity, decision analysis curve (DCA), and calibration curves. Testing and subgroup analysis were performed.

RESULTS

MPI model was composed of MBF and EXT, and Rad model was composed of 12 radiomics features. In the training set, the AUC/accuracy/sensitivity/specificity of the MPI model, Rad model, and the nomogram were 0.795/0.778/0.937/0.511, 0.912/0.825/0.760/0.936 and 0.911/0.865/0.924/0.766 respectively. In the testing set, the AUC/accuracy/sensitivity/specificity of the MPI model, Rad model, and the nomogram were 0.798/0.722/0.659/0.841, 0.887/0.810/0.744/0.932 and 0.900/0.849/0.854/0.841 respectively. The AUC of Rad model and nomogram were significantly higher than that of MPI model. The DCA curve also showed that the clinical net benefit of the Rad model and nomogram was similar but greater than that of MPI model. The calibration curve showed good agreement between the observed and predicted values of the Rad model. In the subgroup analysis of Rad model, there was no significant difference in AUC between subgroups.

CONCLUSIONS

The Rad model is more accurate than the MPI model in predicting coronary stenosis. This noninvasive technique could help improve risk stratification and had good generalization ability.

Incidental findings on routine preoperative noncontrast chest computed tomography and chest radiography prior to cardiac surgery in the multicenter randomized controlled CRICKET study

Objective

To describe the prevalence and consequences of incidental findings when implementing routine noncontrast CT prior to cardiac surgery.

Methods

In the multicenter randomized controlled CRICKET study, 862 adult patients scheduled for cardiac surgery were randomized 1:1 to undergo standard of care (SoC), which included a chest-radiograph, or an additional preoperative noncontrast chest CT-scan (SoC+CT). In this subanalysis, all incidental findings detected on the chest radiograph and CT-scan were analyzed. The influence of smoking status on incidental findings was also evaluated, adjusting for sex, age, and group allocation.

Results

Incidental findings were observed in 11.4% (n = 49) of patients in the SoC+CT group and in 3.7% (n = 16) of patients in the SoC-group (p < 0.001). The largest difference was observed in findings requiring follow-up (SoC+CT 7.7% (n = 33) vs SoC 2.3% (n = 10), p < 0.001). Clinically relevant findings changing the surgical approach or requiring specific treatment were observed in 10 patients (1.2%, SoC+CT: 1.6% SoC: 0.7%), including lung cancer in 0.5% of patients (n = 4) and aortic dilatation requiring replacement in 0.2% of patients (n = 2). Incidental findings were more frequent in patients who stopped smoking (OR 1.91, 1.03–3.63) or who actively smoked (OR 3.91, 1.85–8.23).

Conclusions

Routine CT-screening increases the rate of incidental findings, mainly by identifying more pulmonary findings requiring follow-up. Incidental findings are more prevalent in patients with a history of smoking, and preoperative CT might increase the yield of identifying lung cancer in these patients. Incidental findings, but not specifically the use of routine CT, are associated with delay of surgery.

Key Points

• Clinically relevant incidental findings are identified more often after a routine preoperative CT-scan, when compared to a standard of care workup, with some findings changing patient management.

• Patients with a history of smoking have a higher rate of incidental findings and a lung cancer rate comparable to that of lung cancer screening trials.

• We observed no clear delay in the time to surgery when adding routine CT screening.

A deep learning-based automatic analysis of cardiovascular borders on chest radiographs of valvular heart disease: development/external validation

OBJECTIVES

Cardiovascular border (CB) analysis is the primary method for detecting and quantifying the severity of cardiovascular disease using posterior-anterior chest radiographs (CXRs). This study aimed to develop and validate a deep learning-based automatic CXR CB analysis algorithm (CB_auto) for diagnosing and quantitatively evaluating valvular heart disease (VHD).

METHODS

We developed CB_auto using 816 normal and 798 VHD CXRs. For validation, 640 normal and 542 VHD CXRs from three different hospitals and 132 CXRs from a public dataset were assigned. The reliability of the CB parameters determined by CB_auto was evaluated. To evaluate the differences between parameters determined by CB_auto and manual CB drawing (CB_hand), the absolute percentage measurement error (APE) was calculated. Pearson correlation coefficients were calculated between CB_hand and echocardiographic measurements.

RESULTS

CB parameters determined by CB_auto yielded excellent reliability (intraclass correlation coefficient > 0.98). The 95% limits of agreement for the cardiothoracic ratio were 0.00 ± 0.04% without systemic bias. The differences between parameters determined by CB_auto and CB_hand as defined by the APE were < 10% for all parameters except for carinal angle and left atrial appendage. In the public dataset, all CB parameters were successfully drawn in 124 of 132 CXRs (93.9%). All CB parameters were significantly greater in VHD than in normal controls (all p < 0.05). All CB parameters showed significant correlations (p < 0.05) with echocardiographic measurements.

CONCLUSIONS

The CB_auto system empowered by deep learning algorithm provided highly reliable CB measurements that could be useful not only in daily clinical practice but also for research purposes.

KEY POINTS

• A deep learning-based automatic CB analysis algorithm for diagnosing and quantitatively evaluating VHD using posterior-anterior chest radiographs was developed and validated. • Our algorithm (CB_auto) yielded comparable reliability to manual CB drawing (CB_hand) in terms of various CB measurement variables, as confirmed by external validation with datasets from three different hospitals and a public dataset. • All CB parameters were significantly different between VHD and normal control measurements, and echocardiographic measurements were significantly correlated with CB parameters measured from normal control and VHD CXRs.

Coronary Evaluation Before Heart Valvular Surgery by Using Coronary Computed Tomographic Angiography Versus Invasive Coronary Angiography

Background

Coronary computed tomography angiography (CCTA) is a noninvasive, less expensive, low‐radiation alternative to invasive coronary angiography (ICA). ICA is recommended for coronary evaluation before heart valvular surgery, and the supporting evidence for CCTA is insufficient. Our study is a single‐center, prospective cohort study designed to evaluate the feasibility of CCTA instead of ICA in detection of coronary artery disease before surgery.

Methods and Results

Heart valvular surgery candidates were consecutively enrolled between April 2017 and December 2018. Nine hundred fifty‐eight patients in the CCTA group underwent CCTA primarily, and those with ≥50% coronary stenosis or uncertain diagnosis underwent subsequent ICA. One thousand five hundred twenty‐five patients in the ICA group underwent ICA directly before surgery. Coronary artery bypass grafting decision was made by surgeons according to CCTA or ICA results. Most of the patients (78.8%) in the CCTA group avoided invasive angiography. Thirty‐day mortality (0.7% versus 0.9%, P=0.821), myocardial infarction (6.4% versus 6.9%, P=0.680 ), and low cardiac output syndrome (4.2% versus 2.8%, P=0.085) were similar in the CCTA and ICA groups. Median duration of follow‐up was 19.3 months (interquartile range, 14.2–30.0 months), cumulative rates of mortality (2.6% versus 2.6%, P=0.882) and major adverse cardiac events (9.6% versus 9.0%, P=0.607) showed no difference between the 2 groups. Coronary evaluation expense was lower in the CCTA group ($149.6 versus $636.0, P<0.001).

Conclusions

The strategy of using CCTA as a doorkeeper in coronary evaluation before heart valvular surgery showed noninferiority in identification of candidates for coronary artery bypass grafting and postoperative safety.