Previous studies describing the no-reflow phenomenon in patients with acute myocardial infarction (AMI) undergoing percutaneous coronary intervention (PCI) were largely confined to single-center studies or small registries. To better characterize the incidence, predictors, and outcomes of the no-reflow phenomenon in a large contemporary population, we analyzed patients with AMI who were undergoing PCI of native coronary artery stenoses in the CathPCI Registry from January 1, 2004 through September 5, 2008 (n = 291,380). The angiographic no-reflow phenomenon was site reported using a standardized definition. No-reflow developed in 2.3% of the patients with AMI (n = 6,553) during PCI. Older age, ST-segment elevation AMI, prolonged interval from symptom onset to admission, and cardiogenic shock were clinical variables independently associated with the development of no-reflow (p <0.001). The angiographic factors independently associated with no-reflow included longer lesion length, higher risk class C lesions, bifurcation lesions, and impaired preprocedure Thrombolysis In Myocardial Infarction flow (p <0.001). No-reflow was associated with greater in-hospital mortality (12.6% vs 3.8%, adjusted odds ratio 2.20, 95% confidence interval 1.97 to 2.47, p <0.001) and unsuccessful lesion outcome (29.7% vs 6.6%, adjusted odds ratio 4.70, 95% confidence interval 4.28 to 5.17, p <0.001) compared to patients without no-reflow. In conclusion, the development of no-reflow, although relatively uncommon during PCI for AMI, is associated with adverse clinical outcomes. Upfront strategies to reduce the incidence of no-reflow could be considered for high-risk patients to improve outcomes.
No-reflow, a phenomenon in which coronary blood flow remains impaired despite restoration of epicardial coronary artery patency, is a known complication of percutaneous coronary intervention (PCI). In experimental animal models, no-reflow is explained by microvascular dysfunction as a result of capillary injury, endothelial swelling, changes in blood viscosity, oxidative injury, myocardial edema, and thrombus embolization. Current clinical knowledge about the no-reflow phenomenon has been based primarily on single-center or small registry studies. No-reflow is encountered most frequently among patients undergoing PCI for acute myocardial infarction (AMI) or PCI of saphenous vein grafts. In early studies of patients with AMI, the incidence of no-reflow during PCI ranged widely from 11% to 41% of patients and has been associated with increased mortality. However, little is known about its incidence or the factors that predispose to its development among contemporary patients with AMI undergoing PCI. The National Cardiovascular Data Registry (NCDR) offers a unique opportunity to examine the no-reflow phenomenon in a large, nationwide sample of patients undergoing PCI for AMI. In the present study, we analyzed the (1) incidence of no-reflow during PCI, (2) clinical and procedural characteristics associated with no-reflow, and (3) in-hospital outcomes associated with no-reflow among patients undergoing PCI of native coronary arteries for ST-segment elevation myocardial infarction (STEMI) and non–STEMI.
Methods
Data collection and analysis for the NCDR has been previously described. The NCDR CathPCI Registry is an initiative of the American College of Cardiology Foundation and the Society for Cardiovascular Angiography and Interventions that collects clinical data for PCI procedures from >1,000 cardiac catheterization facilities across the United States. Data are collected by trained data abstractors at participating institutions with standardized data elements and definitions. Only submissions meeting predetermined criteria for data completeness and accuracy are included, with data quality assurance including automatic system validation, random on-site auditing of participating centers, and continued training for site data managers.
Patients undergoing PCI for acute STEMI and non–STEMI included in the NCDR from January 1, 2004 to September 5, 2008 were evaluated in the present analysis (n = 305,694 across 886 hospitals). Patients with saphenous vein graft interventions (n = 14,562) and those with missing data regarding no-reflow status (n = 22) were excluded, yielding a final study population of 291,380 patients from 886 sites.
The no-reflow phenomenon was defined during angiography as the transient cessation of blood flow distal to a treated segment after vessel patency has been established by PCI. The occurrence of no-reflow was ascertained and reported by the individual sites without central angiographic core laboratory confirmation. The definition of no-reflow phenomenon associated with PCI used by CathPCI Registry has remained standardized since the inception of the registry and has been reinforced by continued training of site data abstractors. Because the lack of central angiographic adjudication could allow for some drift in the definition of “no-reflow,” we conducted a sensitivity analysis, excluding patients with preprocedural Thrombolysis In Myocardial Infarction (TIMI) 0 or 1 flow.
Lesion success was defined as postprocedural TIMI 3 flow with 1 of the following: residual stenosis <25% with a stent or <50% without a stent. The definitions of all other variables were prospectively defined by a committee of the American College of Cardiology and are available at the Web site ( http://www.ncdr.com/WebNCDR/ELEMENTS.ASPX ).
The baseline clinical characteristics, lesion characteristics, and clinical outcomes were compared among patients with and without the no-reflow phenomenon. Continuous variables are presented as the median and interquartile percentiles and categorical variables as percentages. Univariate analyses were performed using the Wilcoxon rank-sum test for continuous variables and the Pearson chi-square test for categorical variables.
To determine the factors independently associated with the no-reflow phenomenon, a comprehensive list of patient characteristics adapted from a validated mortality risk prediction model developed in this registry was entered into a multivariate logistics model. These variables included age, male gender, white race, a history of hypertension, diabetes mellitus, dyslipidemia, current tobacco use, a family history of coronary artery disease, previous MI, previous PCI, previous coronary bypass surgery, peripheral vascular disease, known cerebrovascular disease, chronic lung disease, estimated glomerular filtration rate, cardiogenic shock at presentation, congestive heart failure at presentation, STEMI, interval from symptom onset to presentation, lesion length, high-risk lesion, preprocedural TIMI flow, preprocedural creatine kinase-MB, bifurcation lesion, and previously treated lesion. In addition, because patients within a hospital are more likely to be treated in a similar fashion, generalized estimating equation models with an exchangeable working correlation structure were used to adjust for correlations among clustered responses (e.g., within-hospital correlations).
Patients who were discharged to extended care units, transitional care units, or other hospitals were excluded from the mortality outcome analyses, because mortality could not be ascertained after transfer owing to current United States privacy laws. To examine the association between no-reflow phenomenon and outcomes, we also used multivariate logistic regression analysis with the generalized estimating equation. Variables entered into the generalized estimating equation model were adapted from a validated mortality risk model for patients undergoing PCI in this registry that included age, body mass index, diabetes, previous congestive heart failure, New York Heart Association class IV heart failure, previous valve surgery, a history of cerebrovascular disease, peripheral vascular disease, chronic lung disease, previous PCI, cardiogenic shock at admission, preprocedure intra-aortic balloon pump use, left ventricular ejection fraction, glomerular filtration rate, coronary stenosis severity ≥50%, preprocedure TIMI 0 flow, high-risk (class C) lesion, lesion location, and PCI status (urgent/emergency/salvage or elective).
A p value of <0.05 was established as the level of statistical significance for all tests. All statistical analyses were performed by the Duke Clinical Research Institute using SAS software, version 9.2 (SAS Institute, Cary, North Carolina).
Results
Of 291,380 patients with AMI undergoing PCI, 6,553 (2.3%) developed no-reflow. This population included 182,467 patients with STEMI, of whom 4,895 (2.7%) developed no-reflow, and 108,913 patients with non–STEMI, of whom 1,058 (1.0%) developed no-reflow.
In the univariate comparisons, patients who developed no-reflow during PCI were more likely to be older and less likely to have major risk factors for coronary artery disease such as tobacco use and dyslipidemia compared to patients without no-reflow ( Table 1 ). Patients with no-reflow were less likely to have previous MI or previous PCI. Patients who developed no-reflow were more likely to present with signs and symptoms of advanced heart failure, cardiogenic shock, and STEMI. Among the patients with STEMI, the rate of fibrinolytic agent usage was greater among the patients who developed no-reflow than among those without no-reflow (12% vs 7%, p <0.0001).
| Variable | No-reflow | p Value | |
|---|---|---|---|
| Yes (n = 6,553) | No (n = 284,827) | ||
| Age (yrs) | <0.0001 | ||
| Median | 64 | 60 | |
| Interquartile range | 54–75 | 51–71 | |
| Men | 69% | 69% | 0.61 |
| White | 83.1% | 83.4% | <0.0001 |
| Body mass index (kg/m 2 ) | <0.0001 | ||
| Median | 28 | 28.3 | |
| Interquartile range | 25–32 | 25–32 | |
| Diabetes mellitus | 23% | 24% | 0.09 |
| Hypertension ∗ | 62% | 64% | 0.001 |
| Current tobacco use | 33% | 41% | <0.0001 |
| Dyslipidemia † | 58% | 61% | <0.0001 |
| Previous myocardial infarction | 16% | 19% | <0.0001 |
| Previous heart failure | 6% | 6% | 0.12 |
| Previous percutaneous coronary intervention | 15% | 20% | <0.0001 |
| Cerebrovascular disease | 8% | 8% | 0.55 |
| Peripheral vascular disease | 7% | 7% | 0.01 |
| Creatinine clearance (ml/min) | <0.0001 | ||
| Median | 63 | 68 | |
| Interquartile range | 44–81 | 50–87 | |
| Dialysis-dependent renal failure | 1% | 1.3% | <0.0001 |
| Interval from symptom onset to admission >12 h | 22.2% | 22.8% | 0.046 |
| ST-segment elevation myocardial infarction | 75% | 62% | <0.0001 |
| Signs of heart failure | 15% | 11% | <0.0001 |
| New York Heart Association class IV | 55% | 46% | <0.0001 |
| Cardiogenic shock on admission | 17% | 7% | <0.0001 |
∗ Defined as a history of hypertension treated with medication, diet, and/or exercise; blood pressure >140/90 mm Hg on ≥2 occasions, or currently taking antihypertensive medication.
† History of a diagnosis and/or treatment of dyslipidemia by a physician; criteria included documentation of total cholesterol >200 mg/dl, low-density lipoprotein ≥130 mg/dl, high-density lipoprotein <30 mg/dl, admission cholesterol >200 mg/dl, or triglycerides >150 mg/dl.
Analysis of the angiographic variables showed that patients who developed no-reflow were more likely to have greater risk (American College of Cardiology class C) lesions, longer lesions, bifurcation lesions, proximal left anterior descending coronary artery lesions, and impaired preprocedural epicardial flow than patients without no-reflow ( Table 2 ).
| Variable | No-reflow | p Value | |
|---|---|---|---|
| Yes (n = 6,553) | No (n = 284,827) | ||
| Urgent, ∗ emergent, † or salvage ‡ procedure | 97% | 94% | <0.0001 |
| Multivessel coronary disease present | 54% | 53% | 0.05 |
| Lesion morphology | |||
| High-risk class C lesion § | 67% | 52% | <0.0001 |
| Lesion length (mm) | <0.0001 | ||
| Median | 20 | 16 | |
| Interquartile range | 14–25 | 12–23 | |
| Bifurcation lesion | 15% | 10% | <0.0001 |
| Left main/proximal left anterior descending stenosis | 22% | 19% | <0.0001 |
| Previously intervened stenoses | 6% | 7% | 0.05 |
| Preprocedural Thrombolysis In Myocardial Infarction coronary flow grade | <0.0001 | ||
| 0 | 58% | 44% | |
| 1 | 14% | 13% | |
| 2 | 15% | 18% | |
| 3 | 13% | 25% | |
∗ Defined as stabilized patient who required procedure during same hospitalization to minimize chance of additional clinical deterioration, worsening chest pain, congestive heart failure, requirement for intra-aortic balloon pump, unstable angina with intravenous nitroglycerin, or angina at rest.
† Defined as ischemic dysfunction (ongoing ischemia, including angina at rest, despite maximal medical therapy [medical and/or intra-aortic balloon pump]), acute evolving myocardial infarction within 24 hours before catheterization laboratory procedure, or pulmonary edema requiring intubation; and/or mechanical dysfunction resulting in shock with or without circulatory support.
‡ Defined as patient undergoing cardiopulmonary resuscitation en route to the cardiac catheterization laboratory or prior to procedure.
§ Defined as lesion length >2 cm, excess proximal tortuosity, angulation >90°, occlusion >3 months, inability to protect major side branch, and degenerated vein grafts.
The factors independently associated with no-reflow after multivariate analysis are presented in Table 3 . The presence of cardiogenic shock at presentation, older age, STEMI, and longer interval from symptom onset to presentation were the clinical variables most strongly associated with the development of no-reflow. Lesion length, high-risk type C lesions, impaired preprocedural epicardial coronary blood flow, and bifurcation lesions were the angiographic features associated with the development of no-reflow.
| OR | 95% CI | Chi-square | p Value | |
|---|---|---|---|---|
| Cardiogenic shock at presentation | 1.83 | 1.69–1.98 | 213 | <0.001 |
| Lesion length (per 10-mm increase) | 1.17 | 1.14–1.20 | 143 | <0.001 |
| Age (per 10-year increase) | 1.14 | 1.12–1.17 | 134 | <0.001 |
| American College of Cardiology/American Heart Association class C lesion | 1.47 | 1.36–1.59 | 102 | <0.001 |
| ST-segment elevation myocardial infarction vs none | 1.39 | 1.30–1.48 | 100 | <0.001 |
| Current smoker | 0.78 | 0.74–0.83 | 72 | <0.001 |
| Preprocedure Thrombolysis In Myocardial Infarction flow | 64 | <0.001 | ||
| 0 (vs 3) | 2.12 | 1.83–2.45 | ||
| 1 or 2 (vs 3) | 1.84 | 1.60–2.12 | ||
| Bifurcation lesion | 1.29 | 1.19–1.40 | 36 | <0.001 |
| Interval from symptom onset to admission >12 h | 1.18 | 1.10–1.26 | 23 | <0.001 |
| Previous percutaneous coronary intervention | 0.82 | 0.75–0.90 | 18 | <0.001 |
| Signs of heart failure at presentation | 1.17 | 1.07–1.28 | 12 | <0.001 |
The use of intra-aortic balloon pump support was greater among patients who developed no-reflow ( Table 4 ). Glycoprotein IIb/IIIa antagonists were used more frequently in patients who developed no-reflow than in those without no-reflow. Patients who developed no-reflow were less likely to be stented or to receive a drug-eluting stent. Coronary complication rates (dissection or perforation) were greater among patients who developed no-reflow. Patients with intraprocedural no-reflow were significantly less likely to achieve normal epicardial flow at the end of the case. As such, the lesion success rates were lower in patients who developed no-reflow compared to patients without no-reflow. After multivariate adjustment, no-reflow remained significantly associated with an unsuccessful lesion outcome (adjusted odds ratio 4.70, 95% confidence interval 4.28 to 5.17, p <0.001).
| Variable | No-reflow | p Value | |
|---|---|---|---|
| Yes (n = 6,553) | No (n = 284,827) | ||
| Treatment | |||
| Intra-aortic balloon pump use ∗ | 23% | 8% | <0.0001 |
| Glycoprotein IIb/IIIa inhibitor use | 76% | 72% | <0.0001 |
| Stent used | 88% | 91% | <0.0001 |
| Drug-eluting stent used | 54% | 61% | <0.0001 |
| Outcomes | |||
| Postprocedural Thrombolysis In Myocardial Infarction coronary flow | <0.0001 | ||
| 0 | 8% | 2% | |
| 1 | 6% | 1% | |
| 2 | 14% | 3% | |
| 3 | 72% | 94% | |
| Dissection | 9% | 2% | <0.0001 |
| Perforation | 1% | 0.3% | <0.0001 |
| Lesion success † | 70% | 93% | <0.0001 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
