Coronary Artery Occlusions Diagnosed by Transthoracic Doppler
There were 35 patients with and 73 patients without unstable coronary artery disease (Table 1). For stable patients, the mean time from echocardiographic examination to angiography was 24.7 ± 31.7 days. In the group with unstable coronary artery disease two patients had a postponed angiography due to unrelated comorbid conditions. The mean time from echocardiographic examination to angiography was 4.3 ± 3.4 days for the remaining 33 patients.
Eight patients had dominant Cx with PDA originating from the distal Cx, and in one patient, co-dominance was demonstrated by angiography. Table 2 lists findings on QCA for stenoses in the three main coronary arteries (segments 1–8, 11, 13, and 15 in the American Heart Association's 16-segment model) in stenosis groups 2 – 4. At angiography a total occlusion was demonstrated in 28 coronary artery segments. Three patients had two occluded arteries, while the rest had occlusion of a single artery. Twenty arteries showed retrograde flow at angiography and seven arteries showed anterograde flow downstream to the occlusion (Table 2), all with Rentrop grade ≥ 2 collateral circulation. One patient with a proximal LAD occlusion demonstrated ambiguous flow distal to the occlusion, with Rentrop grade 1 collateral circulation.
TTE showed retrograde flow in two LADs and six RCA/PDAs. Coronary angiography confirmed retrograde flow in all these arteries, except in two PDAs where anterograde flow was seen. This discrepancy was probably due to misinterpretation of LAD running around the apex. The sensitivity and specificity for identifying occlusion in arteries with angiographic downstream retrograde flow were 30% and 99%, respectively. Analyses showed no statistical differences in TTE detection of retrograde flow in a coronary artery when adjusted for baseline characteristics of the study cohort, with the exception of reduced feasibility in patients with acute coronary syndromes (ACS) (p = 0.02).
Retrograde flow was demonstrated by colour Doppler in Sb-LADs in five patients. Four of these patients had an upstream LAD occlusion, while one patient had an upstream high-grade LAD stenosis. Angiography showed, downstream to the four LAD occlusions, anterograde flow in three patients and retrograde flow in one patient. TTE showed retrograde flow in Sb-PDAs in 11 patients. RCA was occluded in proximal or mid segments in seven of these patients, while the remaining four patients showed upstream high-grade RCA stenoses. Retrograde flow in CxMb was found in one patient (Figures 2C and D), with angiography showing retrograde flow downstream to a mid Cx occlusion.
In patients with visible anterograde flow in septal perforating vessels, pDVs were measured in Sb-LADs in 47 of 49 patients (96%) and in Sb-PDAs in 2 of 3 patients. Coronary connections through the septum are mainly between the LAD and PDA, while collateral connections to the Cx territory mostly are epicardial. Both borderline and high-grade stenoses in the ipsilateral and/or contralateral coronary artery might influence pDV. Four patient groups with pDV measurements were defined: group A = patients without significant coronary disease in the three main coronary arteries, group B = patients with borderline stenoses in the ipsilateral and/or contralateral coronary artery, group C = patients with high-grade stenoses in the ipsilateral and/or contralateral coronary artery, and group D = patients with occlusion in the contralateral coronary artery. Peak DVs measured in groups A - D are listed in Table 3. No statistically significant differences were found when comparing pDVs in groups A, B, and C, although there was a trend toward increasing pDV with increasing lesion severity in the contralateral artery. There was, however, statistically significant higher pDV in group D compared with pDV in group A (p < 0.001). Moreover, pDV was measured in Sb-LAD in two patients (pDV 1.63 m/sec and 0.68 m/sec, respectively) with occluded LAD without other significant coronary lesions. Both patients had intraseptal collateral bypass supplying blood through Sb-LADs originating upstream to the occlusion to Sb-LADs downstream to the occlusion, illustrating the diversity and complexity of coronary pathophysiology in the course of occluded coronary arteries. The ROC curve for pDV for the detection of coronary occlusion in the contralateral, collateral-receiving artery is shown in Figure 6. ROC curve analysis demonstrated that the optimal pDV cutoff value of 0.57 m/sec had a sensitivity of 79% and a specificity of 69% for detection of occlusion in the contralateral artery. Excluding Sb-PDA, the same pDV cutoff value for Sb-LAD had a sensitivity of 85% and a specificity of 70% for detection of occlusion in RCA. Two patients showed septal collateral flow with pDV ≥ 0.57 m/sec to a contralateral artery with upstream high-grade stenosis. The sensitivity and specificity of pDV above the cutoff value in the detection of diameter stenoses in the range 76% - 100% in the contralateral artery were 68% and 69%, respectively.
(Enlarge Image)
Figure 6.
ROC curve for anterograde peak diastolic flow velocities in septal perforators in diagnosing contralateral coronary occlusion. Receiver operating characteristic (ROC) curve for anterograde peak diastolic flow velocities (pDV) in septal perforating branches in the diagnosis of occlusion in the contralateral, collateral-receiving main coronary artery, with optimal pDV ≥ 0.57 m/sec (sensitivity 79%, specificity 69%). AUC = area under the curve, CI = confidence interval, DS = diameter stenosis.
Excluding the septal collateral pathway, seven patients were identified by TTE having other intramyocardial or epicardial collaterals to occluded main coronary arteries. The most common finding was apically located epicardial and intramyocardial collaterals between the distal LAD and PDA (Figure 5A). Other findings were intramyocardial and epicardial free wall collaterals to myocardial territories originally supported by the LAD or Cx. These findings correctly identified the occluded coronary artery in all seven patients.
Findings of anterograde pDV ≥ 0.57 m/sec in septal branches, retrograde flow in a coronary artery or septal branch, or demonstration of other collaterals were all significantly related to angiographic occlusion (p < 0.001). The individual coronary occlusion could in our study be identified by at least one of the above mentioned TTE findings. Using these criteria, 25 of 28 coronary occlusions (89%) were correctly identified. Seven of eight LAD occlusions, two of four Cx occlusions, and all 16 RCA/PDA occlusions were correctly identified. Detecting at least one positive TTE parameter indicating occlusion had a sensitivity of 89%, specificity of 94%, PPV of 63%, and NPV of 99% for detection of coronary occlusion, as defined by angiography. Analyses showed no statistical differences in the degree of demonstrating collateral flow when adjusted for baseline characteristics of the study cohort, left or right dominance, Rentrop collateral flow, or clinical presentation. Seven arteries with high-grade stenosis demonstrated collateral flow to the artery downstream to the lesion (pDV ≥ 0.57 m/sec or retrograde flow in septal perforating branches). With the combined use of several TTE parameters indicating a coronary occlusion, the sensitivity, specificity, PPV, and NPV for detecting either an occlusion or a high-grade stenosis were 52%, 97%, 82%, and 88%, respectively.
There was no significant difference in mean pDV value between the observers. The biases for pDV were 2.0% for intraobserver (p = 0.14) and 2.3% for interobserver measurements (p = 0.11). The intraobserver and interobserver coefficients of variation for pDV were 3.9% and 4.3%, respectively.
Results
Coronary Angiography
There were 35 patients with and 73 patients without unstable coronary artery disease (Table 1). For stable patients, the mean time from echocardiographic examination to angiography was 24.7 ± 31.7 days. In the group with unstable coronary artery disease two patients had a postponed angiography due to unrelated comorbid conditions. The mean time from echocardiographic examination to angiography was 4.3 ± 3.4 days for the remaining 33 patients.
Eight patients had dominant Cx with PDA originating from the distal Cx, and in one patient, co-dominance was demonstrated by angiography. Table 2 lists findings on QCA for stenoses in the three main coronary arteries (segments 1–8, 11, 13, and 15 in the American Heart Association's 16-segment model) in stenosis groups 2 – 4. At angiography a total occlusion was demonstrated in 28 coronary artery segments. Three patients had two occluded arteries, while the rest had occlusion of a single artery. Twenty arteries showed retrograde flow at angiography and seven arteries showed anterograde flow downstream to the occlusion (Table 2), all with Rentrop grade ≥ 2 collateral circulation. One patient with a proximal LAD occlusion demonstrated ambiguous flow distal to the occlusion, with Rentrop grade 1 collateral circulation.
TTE in Demonstrating Occluded Main Arteries
TTE showed retrograde flow in two LADs and six RCA/PDAs. Coronary angiography confirmed retrograde flow in all these arteries, except in two PDAs where anterograde flow was seen. This discrepancy was probably due to misinterpretation of LAD running around the apex. The sensitivity and specificity for identifying occlusion in arteries with angiographic downstream retrograde flow were 30% and 99%, respectively. Analyses showed no statistical differences in TTE detection of retrograde flow in a coronary artery when adjusted for baseline characteristics of the study cohort, with the exception of reduced feasibility in patients with acute coronary syndromes (ACS) (p = 0.02).
Retrograde flow was demonstrated by colour Doppler in Sb-LADs in five patients. Four of these patients had an upstream LAD occlusion, while one patient had an upstream high-grade LAD stenosis. Angiography showed, downstream to the four LAD occlusions, anterograde flow in three patients and retrograde flow in one patient. TTE showed retrograde flow in Sb-PDAs in 11 patients. RCA was occluded in proximal or mid segments in seven of these patients, while the remaining four patients showed upstream high-grade RCA stenoses. Retrograde flow in CxMb was found in one patient (Figures 2C and D), with angiography showing retrograde flow downstream to a mid Cx occlusion.
In patients with visible anterograde flow in septal perforating vessels, pDVs were measured in Sb-LADs in 47 of 49 patients (96%) and in Sb-PDAs in 2 of 3 patients. Coronary connections through the septum are mainly between the LAD and PDA, while collateral connections to the Cx territory mostly are epicardial. Both borderline and high-grade stenoses in the ipsilateral and/or contralateral coronary artery might influence pDV. Four patient groups with pDV measurements were defined: group A = patients without significant coronary disease in the three main coronary arteries, group B = patients with borderline stenoses in the ipsilateral and/or contralateral coronary artery, group C = patients with high-grade stenoses in the ipsilateral and/or contralateral coronary artery, and group D = patients with occlusion in the contralateral coronary artery. Peak DVs measured in groups A - D are listed in Table 3. No statistically significant differences were found when comparing pDVs in groups A, B, and C, although there was a trend toward increasing pDV with increasing lesion severity in the contralateral artery. There was, however, statistically significant higher pDV in group D compared with pDV in group A (p < 0.001). Moreover, pDV was measured in Sb-LAD in two patients (pDV 1.63 m/sec and 0.68 m/sec, respectively) with occluded LAD without other significant coronary lesions. Both patients had intraseptal collateral bypass supplying blood through Sb-LADs originating upstream to the occlusion to Sb-LADs downstream to the occlusion, illustrating the diversity and complexity of coronary pathophysiology in the course of occluded coronary arteries. The ROC curve for pDV for the detection of coronary occlusion in the contralateral, collateral-receiving artery is shown in Figure 6. ROC curve analysis demonstrated that the optimal pDV cutoff value of 0.57 m/sec had a sensitivity of 79% and a specificity of 69% for detection of occlusion in the contralateral artery. Excluding Sb-PDA, the same pDV cutoff value for Sb-LAD had a sensitivity of 85% and a specificity of 70% for detection of occlusion in RCA. Two patients showed septal collateral flow with pDV ≥ 0.57 m/sec to a contralateral artery with upstream high-grade stenosis. The sensitivity and specificity of pDV above the cutoff value in the detection of diameter stenoses in the range 76% - 100% in the contralateral artery were 68% and 69%, respectively.
(Enlarge Image)
Figure 6.
ROC curve for anterograde peak diastolic flow velocities in septal perforators in diagnosing contralateral coronary occlusion. Receiver operating characteristic (ROC) curve for anterograde peak diastolic flow velocities (pDV) in septal perforating branches in the diagnosis of occlusion in the contralateral, collateral-receiving main coronary artery, with optimal pDV ≥ 0.57 m/sec (sensitivity 79%, specificity 69%). AUC = area under the curve, CI = confidence interval, DS = diameter stenosis.
Excluding the septal collateral pathway, seven patients were identified by TTE having other intramyocardial or epicardial collaterals to occluded main coronary arteries. The most common finding was apically located epicardial and intramyocardial collaterals between the distal LAD and PDA (Figure 5A). Other findings were intramyocardial and epicardial free wall collaterals to myocardial territories originally supported by the LAD or Cx. These findings correctly identified the occluded coronary artery in all seven patients.
Findings of anterograde pDV ≥ 0.57 m/sec in septal branches, retrograde flow in a coronary artery or septal branch, or demonstration of other collaterals were all significantly related to angiographic occlusion (p < 0.001). The individual coronary occlusion could in our study be identified by at least one of the above mentioned TTE findings. Using these criteria, 25 of 28 coronary occlusions (89%) were correctly identified. Seven of eight LAD occlusions, two of four Cx occlusions, and all 16 RCA/PDA occlusions were correctly identified. Detecting at least one positive TTE parameter indicating occlusion had a sensitivity of 89%, specificity of 94%, PPV of 63%, and NPV of 99% for detection of coronary occlusion, as defined by angiography. Analyses showed no statistical differences in the degree of demonstrating collateral flow when adjusted for baseline characteristics of the study cohort, left or right dominance, Rentrop collateral flow, or clinical presentation. Seven arteries with high-grade stenosis demonstrated collateral flow to the artery downstream to the lesion (pDV ≥ 0.57 m/sec or retrograde flow in septal perforating branches). With the combined use of several TTE parameters indicating a coronary occlusion, the sensitivity, specificity, PPV, and NPV for detecting either an occlusion or a high-grade stenosis were 52%, 97%, 82%, and 88%, respectively.
Observer Variability
There was no significant difference in mean pDV value between the observers. The biases for pDV were 2.0% for intraobserver (p = 0.14) and 2.3% for interobserver measurements (p = 0.11). The intraobserver and interobserver coefficients of variation for pDV were 3.9% and 4.3%, respectively.
SHARE
