Many patients discontinue statin after acute myocardial infarction (AMI) despite its necessity. However, limited data are available describing the clinical impact of statin withdrawal after AMI. This study enrolled 3,807 patients in the Korean multicenter registry who survived for 1 year after AMI. All patients were prescribed statin at discharge and were divided into 2 groups on the basis of statin withdrawal history; 603 patients had a history of statin discontinuation and 3,204 patients continued statin therapy. The primary outcome was mortality from any cause. We also analyzed the incidence of cardiac death, nonfatal myocardial infarction, any revascularization, and stroke. The duration of follow-up was 4 years after AMI. Statin withdrawal was associated with higher mortality than continued statin treatment (hazard ratio 3.45, 95% confidence interval 2.81 to 4.24, p <0.001), primarily as the result of increased cardiac mortality (hazard ratio 4.65, 95% confidence interval 3.14 to 6.87, p <0.001). However, the incidences of nonfatal myocardial infarction, any revascularization, and stroke were not different between the groups. Analysis by propensity score matching did not affect the results. In conclusion, many patients experienced statin withdrawal after AMI, which significantly increased long-term mortality in the present study. Careful education and monitoring are needed to reduce adverse cardiac outcomes in patients after AMI.
Few data are available describing the long-term clinical outcomes in patients who discontinue statin use compared with those who continue use, and no studies have described whether even a short period of withdrawal affects clinical outcomes in patients who were prescribed a statin at discharge after acute myocardial infarction (AMI). In the present study, therefore, we investigated through a Korean multicenter registry the long-term clinical outcomes of statin withdrawal at 1 year after AMI in surviving patients who were prescribed a statin at discharge.
Methods
The Convergent Registry of Catholic and Chonnam University for AMI (COREA-AMI) is a Korean prospective, multicenter, and observational registry that was designed to reflect real-world practice in Korean patients with AMI at 9 centers with facilities for primary percutaneous coronary intervention (PCI) representing 2 universities from January 2004 to December 2009. Of the 4,748 patients in the COREA-AMI registry, a total of 3,807 patients were analyzed in the present study. The patients with a history of dyslipidemia or who had taken a statin before hospitalization were included in the analysis. We excluded 941 patients owing to in-hospital death (116 patients), patients who were not prescribed statin at discharge (623 patients), patients who did not survive 1 year after AMI (145 patients), and patients with insufficient data (57 patients). Enrolled patients were divided into 2 groups on the basis of statin withdrawal history during the 1 year after AMI. The statin withdrawal group was defined as at least one incidence of statin discontinuation at an outpatient department or during rehospitalization (n = 603); the nonwithdrawal group was defined as having no history of statin disruption at 1 year after AMI (n = 3204). We matched 577 patients in both groups on the basis of the propensity score to balance the limitations of a nonrandomized trial. The ethics committee of each participating hospital approved the study protocol and all patients provided written informed consent.
The diagnosis of ST-segment elevation myocardial infarction (MI) was based on ST-segment elevation >2 mm in at least 2 precordial leads, ST-segment elevation >1 mm in at least 2 limb leads, or new left bundle branch block on a 12-lead electrocardiogram in the infarct-related artery distribution, as determined by coronary angiography with increased cardiac-specific biomarkers. All laboratory variables were measured upon admission, except for lipid profiles, which were obtained after at least 9 hours of fasting within 24 hours of hospitalization. A history of renal insufficiency consisted of a history of chronic kidney disease and patients receiving chronic hemodialysis and peritoneal dialysis. Baseline left ventricular ejection fraction was determined by 2-dimensional echocardiography performed before or immediately after PCI. Coronary blood flow before and after PCI was classified by the Thrombolysis In Myocardial Infarction score, and coronary lesion complexity was based on the American College of Cardiology/American Heart Association definitions. Patients who underwent PCI received 300 mg aspirin and 600 mg clopidogrel as a loading dose before PCI. Doses of 50 to 70 U/kg of unfractionated heparin were used before or during PCI to maintain an activated clotting time at 250 to 300 seconds. After PCI, 100 to 300 mg aspirin and 75 mg clopidogrel were prescribed daily.
The duration of follow-up was 4 years after AMI. The primary study outcome was mortality from 1 year after MI, although we also evaluated the incidence of cardiac mortality, nonfatal MI, any revascularization, and stroke. Nonfatal recurrent MI was defined as the development of recurrent angina symptoms with new 12-lead electrocardiographic changes or increased cardiac-specific biomarkers. Any revascularization procedure consisted of target lesion, target vessel, or nontarget vessel revascularization. Stroke was defined as ischemic, hemorrhagic, or undetermined stroke.
Continuous variables are presented as means ± SDs and were compared by the Student t test or Mann-Whitney U test. Categorical variables were analyzed by Pearson’s chi-square test or Fisher’s exact test to determine the significance of differences. Kaplan-Meier analysis was performed for patients in statin withdrawal or nonwithdrawal group to compare primary study outcome, and difference between the groups was assessed by log-rank test. Cox regression analysis was used to estimate the clinical impact of statin withdrawal for each end point with adjustment for covariates that had a p <0.1 in the univariate analysis and variables correlated with outcomes.
To balance the limitations of an observational registry, we used propensity score matching by nearest neighbor method to adjust for potentially confounding factors and selection biases. Propensity scores for statin withdrawal were calculated by use of a logistic regression model with the demographic, baseline clinical, angiographic, and procedural variables listed in Tables 1 and 2 . The C-statistic value for the logistic model was 0.69 by receiver operating curve, and we performed 1:1 matching using estimated propensity scores. We successfully matched 577 pairs in both groups and compared baseline, angiographic, and clinical outcomes in the propensity score matched group.
| Variables | Crude population | P | Propensity-matched population | P | ||
|---|---|---|---|---|---|---|
| Statin withdrawal(+) (n = 603) | Statin withdrawal(-) (n = 3204) | Statin withdrawal(+) (n = 577) | Statin withdrawal(-) (n = 577) | |||
| Age (years) | 64.4±12.4 | 60.9±12.3 | <0.001 | 64.3±12.3 | 64.2±12.3 | 0.899 |
| Men | 414 (68.7%) | 2395 (74.8%) | 0.002 | 401 (69.5%) | 395 (68.5%) | 0.750 |
| Body mass index (kg/m 2 ) | 23.9±3.2 | 24.4±3.2 | <0.001 | 23.7±3.3 | 23.9±3.3 | 0.356 |
| Systolic BP (mmHg) | 128.9±28.6 | 130.7±28.2 | 0.163 | 129.3±27.3 | 130.9±28.9 | 0.305 |
| Heart rate (/min) | 75.4±18.9 | 76.1±18.5 | 0.444 | 75.3±17.9 | 76.1±17.9 | 0.416 |
| CPR at presentation | 13 (2.2%) | 33 (1.0%) | 0.039 | 10 (1.7%) | 10 (1.7%) | 1.000 |
| Current or ex-smoke | 354 (58.7%) | 1875 (58.5%) | 0.964 | 342 (59.3%) | 335 (58.1%) | 0.720 |
| Hypertension | 295 (48.9%) | 1559 (48.7%) | 0.929 | 280 (48.5%) | 299 (51.8%) | 0.289 |
| Diabetes mellitus | 190 (31.5%) | 953 (29.7%) | 0.384 | 176 (30.5%) | 185 (32.1%) | 0.612 |
| Familial history of CAD | 30 (5.0%) | 189 (5.9%) | 0.445 | 28 (4.9%) | 22 (3.8%) | 0.470 |
| Renal insufficiency | 20 (3.3%) | 95 (3.0%) | 0.605 | 20 (3.5%) | 16 (2.8%) | 0.612 |
| Cerebrovascular accident | 14 (2.3%) | 153 (4.8%) | 0.006 | 12 (2.1%) | 14 (2.4%) | 0.843 |
| Previous myocardial infarction | 24 (4.0%) | 113 (3.5%) | 0.553 | 24 (4.2%) | 26 (4.5%) | 0.885 |
| Previous PCI | 29 (4.8%) | 117 (3.7%) | 0.202 | 28 (4.9%) | 21 (3.6%) | 0.381 |
| STEMI | 371 (61.5%) | 1974 (61.6%) | 0.964 | 352 (61.0%) | 337 (58.4%) | 0.401 |
| Killip class ≥ 3 | 78 (12.9%) | 247 (7.7%) | <0.001 | 65 (11.3%) | 68 (11.8%) | 0.854 |
| Serum creatinine (mg/dL) | 1.2±1.2 | 1.1±0.8 | 0.064 | 1.2±1.1 | 1.2±1.1 | 0.657 |
| Peak troponin-I (mg/dL) | 56.5±124.3 | 38.0±59.5 | <0.001 | 49.5±66.4 | 47.4±74.8 | 0.618 |
| Peak CK-MB (mg/dL) | 91.8±111.8 | 105.5±133.1 | 0.019 | 89.8±108.9 | 85.3±105.2 | 0.474 |
| Total cholesterol (mg/dL) | 181.4±41.6 | 183.9±40.3 | 0.184 | 181.8±41.5 | 181.0±40.9 | 0.754 |
| Triglyceride (mg/dL) | 124.0±88.0 | 126.4±95.5 | 0.556 | 124.1±88.5 | 127.3±98.5 | 0.552 |
| HDL-cholesterol (mg/dL) | 43.9±11.6 | 42.8±10.9 | 0.024 | 43.9±11.3 | 43.5±12.1 | 0.580 |
| LDL-cholesterol (mg/dL) | 116.5±37.2 | 118.5±35.1 | 0.237 | 116.9±37.1 | 116.3±34.8 | 0.785 |
| Serum glucose (mg/dL) | 173.5±80.7 | 167.8±77.1 | 0.120 | 170.7±76.0 | 170.1±74.9 | 0.889 |
| N-terminal pro BNP (pg/mL) | 3176.5±6571.1 | 1778.5±4294.6 | <0.001 | 2750.9±5662.3 | 3083.9±6113.3 | 0.337 |
| High-sensitivity CRP (mg/L) | 27.2±45.5 | 19.6±34.8 | <0.001 | 24.7±40.8 | 26.8±43.8 | 0.394 |
| Left ventricular EF (%) | 54.2±11.5 | 54.4±11.1 | 0.650 | 54.4±11.4 | 54.2±11.6 | 0.731 |
| Medications at discharge | ||||||
| Aspirin | 602 (99.8%) | 3199 (99.8%) | 1.000 | 577 (100%) | 576 (99.8%) | 1.000 |
| Clopidogrel | 600 (99.5%) | 3194 (99.7%) | 0.446 | 574 (99.5%) | 574 (99.5%) | 1.000 |
| Beta-blocker | 479 (79.4%) | 2511 (78.4%) | 0.589 | 459 (79.5%) | 472 (81.8%) | 0.371 |
| ACE inhibitor or ARB | 481 (79.8%) | 2588 (80.8%) | 0.575 | 461 (79.9%) | 463 (80.2%) | 0.941 |
| Variables | Crude population | P | Propensity-matched population | P | ||
|---|---|---|---|---|---|---|
| Statin withdrawal(+) (n = 603) | Statin withdrawal(-) (n = 3204) | Statin withdrawal(+) (n = 577) | Statin withdrawal(-) (n = 577) | |||
| Infarct-related coronary artery | ||||||
| Left-anterior descending | 287 (47.6%) | 1531 (47.8%) | 0.965 | 270 (46.8%) | 279 (48.4%) | 0.637 |
| Right | 207 (34.3%) | 1081 (33.7%) | 0.779 | 203 (35.2%) | 190 (32.9%) | 0.456 |
| Left circumflex | 101 (16.7%) | 520 (16.2%) | 0.764 | 97 (16.8%) | 98 (17.0%) | 1.000 |
| Left main | 8 (1.3%) | 61 (1.9%) | 0.406 | 7 (1.2%) | 10 (1.7%) | 0.626 |
| Multivessel disease | 320 (53.1%) | 1629 (50.8%) | 0.329 | 306 (53.0%) | 298 (51.6%) | 0.680 |
| ACC/AHA B2/C lesion | 475 (78.8%) | 2494 (77.9%) | 0.668 | 452 (78.3%) | 457 (79.2%) | 0.773 |
| Pre-PCI TIMI flow grade 0 | 252 (42.1%) | 1344 (44.2%) | 0.368 | 238 (41.2%) | 228 (39.5%) | 0.589 |
| PCI with stent placement | 570 (94.5%) | 3160 (98.7%) | <0.001 | 547 (94.8%) | 546 (94.6%) | 1.000 |
| Drug-eluting stent | 503 (84.1%) | 2966 (93.1%) | <0.001 | 487 (84.4%) | 489 (84.7%) | 0.935 |
| Total no. of stents | 1.7±0.9 | 1.7±0.9 | 0.249 | 1.7±0.9 | 1.7±0.9 | 0.517 |
| Total stent length (mm) | 39.3±24.4 | 38.2±23.9 | 0.347 | 39.6±24.3 | 38.5±25.0 | 0.474 |
| Mean stent diameter (mm) | 3.1±0.4 | 3.2±0.4 | <0.001 | 3.2±0.4 | 3.1±0.4 | 0.538 |
| Use of IVUS | 179 (29.7%) | 849 (26.5%) | 0.110 | 175 (30.3%) | 179 (31.0%) | 0.848 |
| Post-PCI TIMI flow grade 3 | 579 (96.5%) | 2878 (92.2%) | <0.001 | 553 (95.8%) | 556 (96.4%) | 0.652 |
| Procedural complications | ||||||
| Periprocedural shock | 40 (6.6%) | 58 (1.8%) | <0.001 | 32 (5.5%) | 33 (5.7%) | 1.000 |
| No-reflow phenomenon | 35 (5.8%) | 164 (5.1%) | 0.485 | 34 (5.9%) | 36 (6.2%) | 0.902 |
| VT or VF | 15 (2.5%) | 40 (1.2%) | 0.025 | 14 (2.4%) | 15 (2.6%) | 1.000 |
| Thrombus aspiration | 30 (5.0%) | 146 (4.6%) | 0.672 | 27 (4.7%) | 25 (4.3%) | 0.887 |
| IABP insertion | 34 (5.6%) | 113 (3.5%) | 0.020 | 28 (4.9%) | 30 (5.2%) | 0.893 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
