Limits...
Non-invasive assessment of functionally significant coronary stenoses through mathematical analysis of spectral ECG components.

Amano T, Shinoda N, Kunimura A, Harada K, Uetani T, Takashima H, Ando H, Kumagai S, Gosho M, Murohara T - Open Heart (2014)

Bottom Line: A consecutive 112 participants suspected to have CAD who were scheduled for elective coronary angiography (CAG) from October 2012 to December 2013 were examined.The FFR values for individual MCG scores with low, borderline and high were 0.77 (0.70 to 0.86), 0.78 (0.71 to 0.82) and 0.69 (0.65 to 0.77), respectively, p=0.042.UMIN ID: 000009992.

View Article: PubMed Central - PubMed

Affiliation: Department of Cardiology , Aichi Medical University , Nagakute , Japan.

ABSTRACT

Objectives: The aim of this study was to evaluate the accuracy of the Multifunction CardioGram (MCG) in detecting the presence of functionally significant coronary ischaemia.

Methods and results: This prospective study evaluated the accuracy of the MCG, a new ECG analysis device used to diagnose ischaemic coronary artery disease (CAD). A consecutive 112 participants suspected to have CAD who were scheduled for elective coronary angiography (CAG) from October 2012 to December 2013 were examined. Their predictive values of relevant ischaemia were measured by MCG, standard ECG and Framingham Risk Score (FRS) and compared. Five levels of ischaemia based on CAG findings adjusted by fractional flow reserve (FFR) values and three levels of MCG score of high, borderline or low were used. The MCG (OR=2.67 (1.60 to 4.44), p<0.001) was the only test significantly associated with ischaemia level. The FFR values for individual MCG scores with low, borderline and high were 0.77 (0.70 to 0.86), 0.78 (0.71 to 0.82) and 0.69 (0.65 to 0.77), respectively, p=0.042. A high MCG score had a specificity of 90.4% (87.0% to 93.9%) in model 1 adjusted by FFR≤0.8 threshold and of 87.0% (83.2% to 90.8%) in model 2 adjusted by FFR≤0.75 threshold, and a negative predictive value of 82.5% (78.3% to 86.7%) in model 1 and of 83.8% (79.6% to 87.9%) in model 2 for the prediction of severe ischaemia.

Conclusions: The MCG showed high specificity with a high negative predictive value, suggesting that the MCG could be used not only to identify functionally significant ischaemia but to reduce unnecessary CAGs.

Trial registration number: UMIN ID: 000009992.

No MeSH data available.


Related in: MedlinePlus

Representative images of coronary angiography and fractional flow reserve (FFR) from a patient who showed no resting ECG abnormality. The average Multifunction CardioGram value of this patient was 6.0, and two vessel disease in the proximal left anterior descending artery (FFR=0.70) in panel A, and the mid right coronary artery (FFR=0.72) in panel B were found. Arrows show the jump up points of FFR.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4248141&req=5

OPENHRT2014000144F4: Representative images of coronary angiography and fractional flow reserve (FFR) from a patient who showed no resting ECG abnormality. The average Multifunction CardioGram value of this patient was 6.0, and two vessel disease in the proximal left anterior descending artery (FFR=0.70) in panel A, and the mid right coronary artery (FFR=0.72) in panel B were found. Arrows show the jump up points of FFR.

Mentions: Table 3 shows the cumulative logit model analysis of the three tests for prediction of ischaemia levels. The MCG (OR=2.67 (1.60 to 4.44), p<0.001 in model 1 and OR=2.39 (1.44 to 3.94), p<0.001 in model 2) was the only test significantly associated with ischaemia level. As shown in figure 1, the MCG level (low, borderline and high) was significantly associated with an increasing level of ischaemia (level 1–5). Figure 2 shows the FFR values for individual MCG scores with low, borderline and high. The FFR values were significantly associated with an increasing risk of the MCG score (p ANOVA=0.042). Table 4 shows the predictive values of severe ischaemia (level 4 and 5) by MCG (high vs borderline/low), ECG (positive vs negative) and FRS (high vs intermediate/low). A high MCG score (≥4.0) had a specificity of 90.4% (87.0% to 93.9%) in model 1 adjusted by FFR≤0.8 (table 4A), and of 87.0% (83.2% to 90.8%) in model 2 adjusted by FFR≤0.75 (table 4B), and a NPV of 82.5% (78.3% to 86.7%) in model 1 and of 83.8% (79.6% to 87.9%) in model 2 for the prediction of severe ischaemia. In examining the κ coefficient in all patients, the MCG (0.42 (0.22 to 0.62)) was significantly superior to the ECG (0.12 (−0.08 to 0.33), p=0.020) and FRS (0.16 (0.02 to 0.31), p=0.024) in model 1. The MCG showed a relatively high predictive accuracy of approximately 80%, irrespective of gender. The area under the ROC (AUC) for the MCG score in prediction of severe ischaemia was 0.66 (0.55 to 0.75) in model 1 and 0.63 (0.53 to 0.73) in model 2. The optimal cut-off value identified by ROC analysis was 4.8 in model 1 and model 2, with a sensitivity of 48.1% and a specificity of 87.7% in model 1, and a sensitivity of 43.5% and a specificity of 84.4% in model 2 (figure 3A). In our study, there were 14 patients having a well-established collateral circulation. When these patients were excluded from analysis, the AUC for the MCG score increased to 0.74 (0.64 to 0.83) in model 1 and 0.76 (0.66 to 0.85) in model 2. The optimal cut-off value, sensitivity and specificity for the prediction of severe ischaemia were 4.8, 60.0% and 87.3% in model 1 and 3.0, 91.7% and 58.1% in model 2, respectively (figure 3B). Figure 4 shows the representative images of CAG and FFR from a patient who showed no resting ECG abnormality. The average MCG value of this patient was 6.0, and two vessel disease in the proximal LAD (FFR=0.70) and the mid RCA (FFR=0.72) were found.


Non-invasive assessment of functionally significant coronary stenoses through mathematical analysis of spectral ECG components.

Amano T, Shinoda N, Kunimura A, Harada K, Uetani T, Takashima H, Ando H, Kumagai S, Gosho M, Murohara T - Open Heart (2014)

Representative images of coronary angiography and fractional flow reserve (FFR) from a patient who showed no resting ECG abnormality. The average Multifunction CardioGram value of this patient was 6.0, and two vessel disease in the proximal left anterior descending artery (FFR=0.70) in panel A, and the mid right coronary artery (FFR=0.72) in panel B were found. Arrows show the jump up points of FFR.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4248141&req=5

OPENHRT2014000144F4: Representative images of coronary angiography and fractional flow reserve (FFR) from a patient who showed no resting ECG abnormality. The average Multifunction CardioGram value of this patient was 6.0, and two vessel disease in the proximal left anterior descending artery (FFR=0.70) in panel A, and the mid right coronary artery (FFR=0.72) in panel B were found. Arrows show the jump up points of FFR.
Mentions: Table 3 shows the cumulative logit model analysis of the three tests for prediction of ischaemia levels. The MCG (OR=2.67 (1.60 to 4.44), p<0.001 in model 1 and OR=2.39 (1.44 to 3.94), p<0.001 in model 2) was the only test significantly associated with ischaemia level. As shown in figure 1, the MCG level (low, borderline and high) was significantly associated with an increasing level of ischaemia (level 1–5). Figure 2 shows the FFR values for individual MCG scores with low, borderline and high. The FFR values were significantly associated with an increasing risk of the MCG score (p ANOVA=0.042). Table 4 shows the predictive values of severe ischaemia (level 4 and 5) by MCG (high vs borderline/low), ECG (positive vs negative) and FRS (high vs intermediate/low). A high MCG score (≥4.0) had a specificity of 90.4% (87.0% to 93.9%) in model 1 adjusted by FFR≤0.8 (table 4A), and of 87.0% (83.2% to 90.8%) in model 2 adjusted by FFR≤0.75 (table 4B), and a NPV of 82.5% (78.3% to 86.7%) in model 1 and of 83.8% (79.6% to 87.9%) in model 2 for the prediction of severe ischaemia. In examining the κ coefficient in all patients, the MCG (0.42 (0.22 to 0.62)) was significantly superior to the ECG (0.12 (−0.08 to 0.33), p=0.020) and FRS (0.16 (0.02 to 0.31), p=0.024) in model 1. The MCG showed a relatively high predictive accuracy of approximately 80%, irrespective of gender. The area under the ROC (AUC) for the MCG score in prediction of severe ischaemia was 0.66 (0.55 to 0.75) in model 1 and 0.63 (0.53 to 0.73) in model 2. The optimal cut-off value identified by ROC analysis was 4.8 in model 1 and model 2, with a sensitivity of 48.1% and a specificity of 87.7% in model 1, and a sensitivity of 43.5% and a specificity of 84.4% in model 2 (figure 3A). In our study, there were 14 patients having a well-established collateral circulation. When these patients were excluded from analysis, the AUC for the MCG score increased to 0.74 (0.64 to 0.83) in model 1 and 0.76 (0.66 to 0.85) in model 2. The optimal cut-off value, sensitivity and specificity for the prediction of severe ischaemia were 4.8, 60.0% and 87.3% in model 1 and 3.0, 91.7% and 58.1% in model 2, respectively (figure 3B). Figure 4 shows the representative images of CAG and FFR from a patient who showed no resting ECG abnormality. The average MCG value of this patient was 6.0, and two vessel disease in the proximal LAD (FFR=0.70) and the mid RCA (FFR=0.72) were found.

Bottom Line: A consecutive 112 participants suspected to have CAD who were scheduled for elective coronary angiography (CAG) from October 2012 to December 2013 were examined.The FFR values for individual MCG scores with low, borderline and high were 0.77 (0.70 to 0.86), 0.78 (0.71 to 0.82) and 0.69 (0.65 to 0.77), respectively, p=0.042.UMIN ID: 000009992.

View Article: PubMed Central - PubMed

Affiliation: Department of Cardiology , Aichi Medical University , Nagakute , Japan.

ABSTRACT

Objectives: The aim of this study was to evaluate the accuracy of the Multifunction CardioGram (MCG) in detecting the presence of functionally significant coronary ischaemia.

Methods and results: This prospective study evaluated the accuracy of the MCG, a new ECG analysis device used to diagnose ischaemic coronary artery disease (CAD). A consecutive 112 participants suspected to have CAD who were scheduled for elective coronary angiography (CAG) from October 2012 to December 2013 were examined. Their predictive values of relevant ischaemia were measured by MCG, standard ECG and Framingham Risk Score (FRS) and compared. Five levels of ischaemia based on CAG findings adjusted by fractional flow reserve (FFR) values and three levels of MCG score of high, borderline or low were used. The MCG (OR=2.67 (1.60 to 4.44), p<0.001) was the only test significantly associated with ischaemia level. The FFR values for individual MCG scores with low, borderline and high were 0.77 (0.70 to 0.86), 0.78 (0.71 to 0.82) and 0.69 (0.65 to 0.77), respectively, p=0.042. A high MCG score had a specificity of 90.4% (87.0% to 93.9%) in model 1 adjusted by FFR≤0.8 threshold and of 87.0% (83.2% to 90.8%) in model 2 adjusted by FFR≤0.75 threshold, and a negative predictive value of 82.5% (78.3% to 86.7%) in model 1 and of 83.8% (79.6% to 87.9%) in model 2 for the prediction of severe ischaemia.

Conclusions: The MCG showed high specificity with a high negative predictive value, suggesting that the MCG could be used not only to identify functionally significant ischaemia but to reduce unnecessary CAGs.

Trial registration number: UMIN ID: 000009992.

No MeSH data available.


Related in: MedlinePlus