Death Following Creatine Kinase-MB Elevation After Coronary Intervention
Identification of an Early Risk Period: Importance of Creatine Kinase-MB Level, Completeness of Revascularization, Ventricular Function, and Probable Benefit of Statin Therapy
Background— Creatine kinase (CK)-MB elevation after percutaneous coronary intervention (PCI) has been associated with subsequent cardiac death. The patients at risk, the timing of risk, and potential treatment implications are uncertain.
Methods and Results— Eight thousand, four hundred nine consecutive non– acute myocardial infarction patients with successful PCI and no emergency surgery or Q-wave myocardial infarction were followed for 38±25 months; 1446 (17.2%) had post-PCI CK-MB above normal on routine ascertainment. Patients were prospectively stratified into those with CK-MB 1 to 5× or CK-MB >5× normal. No patient with CK-MB 1 to 5× normal died during the first week after PCI, and excess risk of early death for patients with CK-MB elevation occurred primarily in the first 3 to 4 months. The actuarial 4-month risk of death was 8.9%, 1.9%, and 1.2% for patients with CK-MB >5×, CK-MB 1 to 5×, and CK-MB ≤1× normal (P<0.001). Death within 4 months was independently correlated with the degree of CK-MB elevation, creatinine ≥2 mg%, post-PCI C-reactive protein, low ejection fraction, age, and congestive heart failure class (P<0.01 for all). In a matched subset analysis, incomplete revascularization (P<0.001), congestive heart failure class (P=0.005), and no statin treatment at hospital discharge (P=0.009) were associated with death.
Conclusions— Patients with CK-MB elevation after PCI are at excess risk of death for 3 to 4 months, although prolonging hospitalization for CK-MB 1 to 5× is unlikely to modify risk. CK-MB >5× normal, incomplete revascularization, elevated C-reactive protein, heart failure, the elderly, and hospital discharge without on statin therapy increases risk. Several of these factors suggest that inflammation may play a part in the excess risk of death.
Received March 27, 2002; revision received June 18, 2002; accepted June 18, 2002.
Relatively low-level release of cardiac enzymes (myonecrosis) after percutaneous coronary intervention (PCI) occurs in 10% to 40% of the ≈1.7 million patients treated annually.1 In 1994, Kugelmass et al2 reported that PCI leading to a > 5× elevation in creatine kinase (CK) was associated with increased mortality rates 2 years after intervention. Shortly thereafter, Abdelmeguid et al3 reported that even low-level elevations of CK (1 to 2× upper limit of normal [ULN] with positive MB fraction) were associated with an increased risk of death over a mean follow-up of 3 years. This observation has been confirmed by numerous investigators4–7⇓⇓⇓ but disputed by some.8–9⇓ The underlying cause of this apparent risk of death has remained obscure, with some attributing it to the deleterious consequences of myonecrosis on left ventricular function or electrophysiologic stability; others attribute it to the association between risk of CK-MB elevation and coronary atheroma burden.10 These uncertainties have made it difficult for the clinician to know how to treat these patients—specifically, which, if any, patients should have hospitalization prolonged for observation and whether or not their postdischarge medical regimen should be altered.
Baseline, procedural, and outcome data are prospectively recorded on all patients undergoing PCI at the Cleveland Clinic by trained personnel on dedicated case report forms. Consecutive patients are prospectively identified for long-term follow-up.11 Patients, family, and, if necessary, referring physicians, involved hospitals, and Social Security Death Index, were routinely contacted at 30 days and 1, 3, and 5 years after PCI to ascertain the occurrence of death and other adverse events. This study cohort consisted of all patients undergoing PCI from January 1995 to June 2001, identified for follow-up. Patients with acute or recent myocardial infarction with still-elevated CK-MB (n=542), technical failure (lesion >50% or TIMI flow ≤2) (n=424), emergency bypass surgery or periprocedural Q-wave infarction (n=101), or death <24 hours after PCI (in which Q-wave infarction could not be excluded) (n=36) were excluded. Follow-up was 97.7% complete. Discharge medications were confirmed by chart review for all patients in the matched cohort (see below).
Measurement of Creatine Kinase Isoenzymes
Creatine kinase-MB levels were routinely obtained 6 to 8 hours after PCI, the morning after PCI, and in the event of the occurrence of symptoms suggesting ischemia. Measurements were performed with the use of a chemiluminescence immunoassay and Elecsys 2010 analyzer from Roche Diagnostics. Laboratory ULNs were 220 U for CK and 8.8 U for CK-MB.
Baseline Characteristics, Potential Confounders, and End Points
Patient demographics and potential confounders are enumerated in Table 1. The primary study end point was death 2 days to 4 months after PCI (for rationale of time window, see below).
“Hospital discharge medications” refers to medications at the time of hospital discharge, or, for those patients who died before discharge, those medications received 24 to 48 hours after PCI. “Complete revascularization” refers to successful treatment (final diameter stenosis <50% and TIMI flow 3) of all lesions in vessels ≥2.0 mm in diameter supplying viable myocardium (myocardium assumed to be viable unless akinetic on ventriculography or scar by noninvasive imaging).
Data are presented as mean±SD, median and interquartile range, or as a percentage, as appropriate. Propensity analyses were performed to adjust for the likelihood of prescribing medications at hospital discharge (eg, statins). Kaplan-Meier curves were used to present the occurrence of endpoint data over time. Univariate between-group analyses were performed with the use of the χ2 test, t test, or Kolmogorov-Smirnov analyses. Cox proportional hazards regression techniques were used to evaluate independent correlates of the study endpoints. A probability value of <0.05 was considered statistically significant.
Patients with CK-MB elevation were prospectively divided into a group with 1 to 5× ULN and those with >5× ULN elevation. To ascertain risk factors for and modes of cardiac death, patients treated from January 1995 to June 2000 with CK-MB elevation who died within the first 4 months after PCI were formally compared with 3 different patient groups: (1) all other patients meeting study entry criteria, (2) patients with CK-MB elevation matched by CK-MB within 15% and the procedure date, who did not die during the specified time period, (3) all patients without CK-MB elevation who died within the time window.
Timing of Death After Coronary Intervention
As can be seen in Figures 1 and 2⇓, for those patients with CK-MB >5× ULN, there was a clear excess of deaths that began almost immediately and appeared to extend ≈3 to 4 months after PCI. Thereafter, the survival curves appear to be nearly parallel for 12 months. For the group with CK-MB elevation 1 to 5×, the excess risk became apparent during the second week after PCI, although it was of considerably less magnitude than that for the CK-MB >5× group. It also appeared to abate after the first 3 to 4 months. By 4 years, an additional 2.0% and 4.5% excess mortality rate relative to the CK-MB <1× group was observed for the CK-MB 1 to 5× and >5× groups (P=0.024 and P<0.001, respectively). On this basis, we defined the time window of early risk to be 4 months after PCI.
Clinical Presentation, Management, and Outcome
The characteristics of the 4 groups of interest (CK-MB elevation with death <4 months, matched CK-MB elevation surviving 4 months, no CK-MB elevation with death within 4 months, and all patients) are enumerated in Tables 1 and 2⇓.
Correlates of Death
As shown in Table 3, the independent correlates of death within 4 months of PCI for the entire patient cohort were degree of CK-MB elevation (P=0.001), baseline creatinine ≥2 mg% (P=0.001), C-reactive protein (CRP) measured within 24 hours after PCI (P=0.001), low ejection fraction (P=0.004), age (P=0.006), and New York Heart Association congestive heart failure (CHF) class (P=0.006). When a similar analysis was performed by using only those patients with CK-MB elevation who died and the cohort matched on CK elevation who survived, the independent correlates of death were incomplete revascularization (P<0.001), CHF class (P=0.005), and hospital discharge without a statin (P=0.009). Kaplan-Meier survival curves for selected risk groups are provided (Figure 3).
Deaths on Days 2 to 7 After PCI
Pertinent details regarding presentation of the 9 patients dying within the week after PCI are provided in Table 4. Six of the 9 patients had not yet been discharged from the hospital. Of the remainder, one had a definite stent thrombosis after a maximum in hospital post-PCI CK-MB=45, another had sudden death after CK-MB=103 (possible stent thrombosis), and the third had an intracerebral hemorrhage after CK-MB of 2.
Principal conclusions from this, one of the largest evaluations of the prognostic implications of CK-MB elevation after PCI to date are (1) CK-MB elevation to any level above normal appears to heighten the risk of death after PCI, although the magnitude of the risk for patients with CK-MB elevation 1 to 5× normal is substantially less than that for those patients with CK-MB elevation ≥5× normal, (2) there is an early risk period apparently limited to 3 to 4 months after PCI, regardless of the extent of CK-MB rise above normal, (3) for patients with CK-MB elevation 1 to 5× the ULN, the risk of death during the first week after PCI is extraordinarily low; hence, prolongation of hospitalization would not be expected to decrease mortality rates, (4) the risk of death is greatest for patients with the largest CK-MB elevations, kidney failure, elevated post-PCI CRP, poor left ventricular function or CHF, incomplete revascularization, and advanced age, (5) the limited period of risk, the relation of CRP, statin therapy, and possibly nonsteroidal anti-inflammatory agents to subsequent outcome have implications regarding the pathophysiology and management of this problem.
These results shed light on the apparently contradictory results of previous smaller studies.5– 10⇓⇓⇓⇓⇓ Few debate the importance of CK-MB elevations >5 to 8× the ULN, but excess risk with elevations less than that has not been detected in all studies. The modest magnitude of the risk with CK-MB elevation 1 to 5× normal in this study is evident. The fact that it follows a time course of risk similar to that seen with higher levels of CK-MB elevation provides consistency and further supports the argument that the finding of a modest increase in risk is real.
Risk factors for early death identified in this study have previously been identified as risk factors after PCI in general or for patients with coronary disease.12–15⇓⇓⇓ The primary difference, however, is the magnitude of risk. Thus, for example, in the absence of CHF, the presence of CK-MB elevation >5× normal increased the 1-month mortality rate from 0.2% to 0.4% and the 4-month mortality rate from 0.9% to 6.0%. However, with preexistent heart failure and increase in CK-MB >5× normal increased the 1-month mortality rate from 1.4% to 4.5% and the 4-month mortality rate from 1.9% to 14.0% (Figure 3). Thus the high-risk patient has a CK-MB elevation >5× normal and at least one of the other risk factors identified.
The finite period of principal risk strongly argues that there is something about the myocardial necrosis, above and beyond whatever risk may be associated with an increased atheroma burden, that increases the likelihood of a fatal outcome. The amount of myonecrosis relative to many patients presenting with spontaneous non–Q-wave myocardial infarction seems modest. In this setting, the findings that patients who are destined to die within 4 months have higher post-PCI CRP (P=0.001) and less frequently use statins (P=0.006) or nonsteroidal anti-inflammatory agents (P=0.029) argues that perhaps in some patients the relatively modest myonecrosis engenders an inflammatory process that leads either to CHF or further myocardial infarction, the two most common causes of death in this population (Table 1). This hypothesis could be tested by the utilization of anti-inflammatory therapy targeted to such patients in a randomized clinical setting.
The delayed risk period beyond 12 to 24 months (Figure 2) would be less likely to derive from the effects of necrosis and inflammation but perhaps more likely to derive from the observed relation between CK-MB risk and atheroma burden.10
Despite its origin from >8000 treated patients, this study is limited by the relatively small numbers of patients dying—the focus of the analysis. In addition, the cause of death could not be assessed in all patients. CRP was only obtained routinely late during the study period (1337 patients), and some parameters (eg, white blood cell count) were only obtained for clinical indication. Lastly, although we attempted to compensate for baseline patient differences by using a propensity score analysis, the analysis of statin effect must be viewed with discretion because randomized allocation to treatment was not performed.
The implications of these findings to the practicing cardiologist appear to be 2-fold: (1) that while the patients with elevated CK-MB after PCI do appear to be at higher risk of death in the ensuing 4 months, it is probably only reasonable to prolong hospitalization for patients with CK-MB elevation >5× normal in conjunction with other risk factors, particularly those relating to CHF, inability to revascularize large portions of myocardium, advanced patient age, or kidney failure, (2) patients at highest risk for death as the result of CK-MB elevation and the presence of other risk factors should be strongly considered for statin therapy. Furthermore, formal testing of the hypothesis that anti-inflammatory medications might reduce the risk of early death in this patient population might well be considered.
- ↵Abdelmeguid AE, Topol EJ, Whitlow PL, et al. Significance of mild transient release of creatine kinase-MB fraction after percutaneous coronary interventions. Circulation. 1996; 94: 1528–1536.
- ↵Hong MK, Mehran R, Dangas G, et al. Creatine kinase-MB enzyme elevation following successful saphenous vein graft intervention is associated with late mortality. Circulation. 1999; 100: 2400–2405.
- ↵Kini A, Marmur JD, Kini S. Creatine kinase-MB elevation after coronary intervention correlates with diffuse atherosclerosis, and low-to-medium level elevation has a benign clinical course: implications for early discharge after coronary intervention. J Am Coll Cardiol. 1999; 34: 663–671.
- ↵Stone GW, Mehran R, Dangas G, et al. Differential impact on survival of electrocardiographic Q-wave versus enzymatic myocardial infarction after percutaneous intervention: a device-specific analysis of 7147 patients. Circulation. 2001; 104: 642–647.
- ↵Mehran R, Dangas G, Mintz GS, et al. Atherosclerotic plaque burden and CK-MB enzymes elevation after coronary interventions. Intravascular Ultrasound study of 2256 patients. Circulation. 2000; 101: 604–610.
- ↵Califf RM, Armstrong PW, Carver JR, et al. 27th Bethesda Conference: matching the intensity of risk factor management with the hazard for coronary disease events. Task Force 5. Stratification of patients into high, medium and low risk subgroups for purposes of risk factor management. J Am Coll Cardiol. 1996; 27: 1007–1019.
- ↵Ellis SG, Cowley MJ, DiSciascio G, et al. Determinants of 2-year outcome after coronary angioplasty in patients with multivessel disease on the basis of comprehensive preprocedural evaluation. Implications for patient selection. Circulation. 1991; 83: 1905–1914.