Prognostic Importance of Coronary Anatomy and LVEF
Prognostic Importance of Coronary Anatomy and LVEF
Anatomical assessment of coronary disease and left ventricular systolic function retains prognostic power in spite of OMT and the provision of either initial or symptom-driven PCI. Because this analysis does not include a placebo group, the observed risk for secondary ischemic events during follow-up is a measure of residual risk in spite of modern management. Although traditionally felt to impart residual risk, pLAD was not shown to be predictive of any outcome in the context of modern, evidence-based therapy. These findings are not confounded by periprocedural MIs or PCI events, are based on adjudicated end points, and are adjusted for nonangiographic determinants of risk.
We found no subgroup in which initial PCI improved patient prognosis, thereby confirming the overall COURAGE results. However, the current analysis also underscores, for example, that a patient with single- or double-vessel disease does not "shift" to a lower risk of zero or single-vessel disease merely by successful relief of an obstruction (Figure 1, Figure 3, Figure 5). Based on the most frequent end point in this analysis (death, MI, or NSTE-ACS as shown in Figure 1), a general bedside assessment of residual risk, showing the interplay and relative importance of both LVEF and coronary burden at baseline, is shown in Figure 6.
(Enlarge Image)
Figure 6.
COURAGE "Rule of Thumb" for estimating residual risk on OMT and either elective or symptom-driven PCI. Rates are for death, MI, or NSTE-ACS expressed as percentages per 4.6 years of follow-up and adjusted as described in "Methods" section.
There was, however, a nonsignificant trend toward a lower event rate emerging after >2 years of follow-up in patients with the combination of 3 VD and low LVEF treated with OMT + PCI compared with those treated with OMT alone (Figure 3). A conservative sample size calculation (see "Methods" section) based on rates of death, MI, or NSTE-ACS and the observed risk reduction of 16% (Figure 1) suggests that a trial of 7,008 patients would be required to show this degree of risk reduction with OMT + PCI to be significant. To our knowledge, there are no ongoing trials specifically targeting this highest risk anatomical subgroup of stable CAD patients. Of relevance is the ongoing ISCHEMIA trial, which has an enrollment goal similar to our calculations, but the target population is patients with stable CAD, LVEF ≥35%, and at least moderate ischemia with stress imaging. Randomization to conservative or invasive strategies in that trial occurs before catheterization and is not based on coronary burden of disease.
The presence of pLAD as a rationale for favoring OMT + PCI was not borne out by any of our analyses, a finding that is both provocative and perhaps instructive. Whether this is a function of OMT or lack of power or both is unclear. However, unlike all the other angiographic features that appear to confer risk in both untreated patients and in this aggressively treated population, it may well be that this is the only feature of high risk in an untreated population that is effectively abrogated by OMT. Figure 4, Figure 5 show that pLAD needs to be considered mainly with respect to LVEF and concomitant angiographic burden of disease, compatible with the concept that further MI in patients with already compromised LVEF would be particularly deleterious if caused by pLAD. Current perceptions, practice patterns, and existing guidelines are not conducive to execution of adequately powered clinical trials focusing solely on this particular anatomical subset but may, nevertheless, be warranted, particularly if refined by the additional assessment of the functional significance of the pLAD lesion. This type of patient will be enrolled in the ISCHEMIA trial. The dominant importance of baseline low LVEF as a predictor of mortality in our population of aggressively treated, stable CAD patients is concordant with a recent analysis of patients with ST-segment elevation MI treated in a large scale, contemporary international study of patients who underwent primary PCI with contemporary antithrombotic regimens.
This study has several limitations. It is a post hoc analysis, and some of the subgroup analyses are impacted by small sample sizes and relatively few events. Results may have been different if a ≥70% DS threshold had been used. However, our detailed analysis of the severity of pLAD lesions failed to show any reason to expect highly disparate conclusions using a higher cutoff in the entire cohort. Moreover, it is hard to predict the effect of the interplay between using a higher threshold that might be associated with a higher risk, counterbalanced by smaller cohort sizes and reduced power. The COURAGE study protocol did not include more detailed angiographic analyses such as Syntax scoring, detailed lesion description, or assessment of angiograms and collateralization at the time of secondary events. By design, the study cannot address LM disease or the additional impact of LVEF <30%. Even so, the current descriptors based on simple measures of angiographic disease burden and LVEF remain applicable and of relevance to general cardiology practice. It is highly likely that many of the lesions used to define VD in this study may not have been physiologically significant if assessed by fractional flow reserve, although all patients had clinically accepted documentation of ischemia by exercise electrocardiography or stress imaging or had convincing symptoms of ongoing ischemia. Thus, the results might be considered to be "diluted" by inclusion of physiologically nonsignificant lesions. In consideration of this possibility, it is even more remarkable that the simple parameters of LVEF and coronary disease burden retain prognostic impact. Finally, PCI during the COURAGE trial was predominantly performed using bare-metal stents with a higher restenosis rate than seen with more current stents. Even so, the end points analyzed in this study (death, MI, NSTE-ACS) have not been shown to be altered by restenosis rates.
In conclusion, angiographic burden of coronary disease and LVEF retain prognostic importance even during contemporary therapy and reflect residual risk for secondary ischemic events. These simple variables help identify patients most in need of newer therapeutic approaches.
Discussion
Anatomical assessment of coronary disease and left ventricular systolic function retains prognostic power in spite of OMT and the provision of either initial or symptom-driven PCI. Because this analysis does not include a placebo group, the observed risk for secondary ischemic events during follow-up is a measure of residual risk in spite of modern management. Although traditionally felt to impart residual risk, pLAD was not shown to be predictive of any outcome in the context of modern, evidence-based therapy. These findings are not confounded by periprocedural MIs or PCI events, are based on adjudicated end points, and are adjusted for nonangiographic determinants of risk.
We found no subgroup in which initial PCI improved patient prognosis, thereby confirming the overall COURAGE results. However, the current analysis also underscores, for example, that a patient with single- or double-vessel disease does not "shift" to a lower risk of zero or single-vessel disease merely by successful relief of an obstruction (Figure 1, Figure 3, Figure 5). Based on the most frequent end point in this analysis (death, MI, or NSTE-ACS as shown in Figure 1), a general bedside assessment of residual risk, showing the interplay and relative importance of both LVEF and coronary burden at baseline, is shown in Figure 6.
(Enlarge Image)
Figure 6.
COURAGE "Rule of Thumb" for estimating residual risk on OMT and either elective or symptom-driven PCI. Rates are for death, MI, or NSTE-ACS expressed as percentages per 4.6 years of follow-up and adjusted as described in "Methods" section.
There was, however, a nonsignificant trend toward a lower event rate emerging after >2 years of follow-up in patients with the combination of 3 VD and low LVEF treated with OMT + PCI compared with those treated with OMT alone (Figure 3). A conservative sample size calculation (see "Methods" section) based on rates of death, MI, or NSTE-ACS and the observed risk reduction of 16% (Figure 1) suggests that a trial of 7,008 patients would be required to show this degree of risk reduction with OMT + PCI to be significant. To our knowledge, there are no ongoing trials specifically targeting this highest risk anatomical subgroup of stable CAD patients. Of relevance is the ongoing ISCHEMIA trial, which has an enrollment goal similar to our calculations, but the target population is patients with stable CAD, LVEF ≥35%, and at least moderate ischemia with stress imaging. Randomization to conservative or invasive strategies in that trial occurs before catheterization and is not based on coronary burden of disease.
The presence of pLAD as a rationale for favoring OMT + PCI was not borne out by any of our analyses, a finding that is both provocative and perhaps instructive. Whether this is a function of OMT or lack of power or both is unclear. However, unlike all the other angiographic features that appear to confer risk in both untreated patients and in this aggressively treated population, it may well be that this is the only feature of high risk in an untreated population that is effectively abrogated by OMT. Figure 4, Figure 5 show that pLAD needs to be considered mainly with respect to LVEF and concomitant angiographic burden of disease, compatible with the concept that further MI in patients with already compromised LVEF would be particularly deleterious if caused by pLAD. Current perceptions, practice patterns, and existing guidelines are not conducive to execution of adequately powered clinical trials focusing solely on this particular anatomical subset but may, nevertheless, be warranted, particularly if refined by the additional assessment of the functional significance of the pLAD lesion. This type of patient will be enrolled in the ISCHEMIA trial. The dominant importance of baseline low LVEF as a predictor of mortality in our population of aggressively treated, stable CAD patients is concordant with a recent analysis of patients with ST-segment elevation MI treated in a large scale, contemporary international study of patients who underwent primary PCI with contemporary antithrombotic regimens.
This study has several limitations. It is a post hoc analysis, and some of the subgroup analyses are impacted by small sample sizes and relatively few events. Results may have been different if a ≥70% DS threshold had been used. However, our detailed analysis of the severity of pLAD lesions failed to show any reason to expect highly disparate conclusions using a higher cutoff in the entire cohort. Moreover, it is hard to predict the effect of the interplay between using a higher threshold that might be associated with a higher risk, counterbalanced by smaller cohort sizes and reduced power. The COURAGE study protocol did not include more detailed angiographic analyses such as Syntax scoring, detailed lesion description, or assessment of angiograms and collateralization at the time of secondary events. By design, the study cannot address LM disease or the additional impact of LVEF <30%. Even so, the current descriptors based on simple measures of angiographic disease burden and LVEF remain applicable and of relevance to general cardiology practice. It is highly likely that many of the lesions used to define VD in this study may not have been physiologically significant if assessed by fractional flow reserve, although all patients had clinically accepted documentation of ischemia by exercise electrocardiography or stress imaging or had convincing symptoms of ongoing ischemia. Thus, the results might be considered to be "diluted" by inclusion of physiologically nonsignificant lesions. In consideration of this possibility, it is even more remarkable that the simple parameters of LVEF and coronary disease burden retain prognostic impact. Finally, PCI during the COURAGE trial was predominantly performed using bare-metal stents with a higher restenosis rate than seen with more current stents. Even so, the end points analyzed in this study (death, MI, NSTE-ACS) have not been shown to be altered by restenosis rates.
In conclusion, angiographic burden of coronary disease and LVEF retain prognostic importance even during contemporary therapy and reflect residual risk for secondary ischemic events. These simple variables help identify patients most in need of newer therapeutic approaches.
Source...