Catheterization and Cardiovascular Diagnosis 41:117–119 (1997)

Editorial

Creatine Kinase Release After Catheter-Based

Elevation of total creatine kinase (CK) or myocardial when procedural mechanical complications were entered
isoenzyme (CK-MB) activity has been recorded in up to (P 5 0.055).
20% of successful percutaneous angioplasty (PTCA) The Cleveland Clinic reported that patients who had
procedures (Table I) [1–5]. This observation—in the CK $ 23 normal were a higher risk group with a greater
absence of overt myocardial infarction—has not been incidence of recent myocardial infarction, unstable an-
universally accepted to indicate an adverse prognosis gina, thrombotic or complex lesions, vein grafts, and
[1–3]. Conversely, it has been suggested that transient DCA procedures [5]. Although LVEF ,40% and multives-
in-laboratory vessel closure is associated with adverse sel disease were similarly distributed in all groups, the
long-term outcome only if CK is elevated [6]. Although prevalence of three-vessel disease and inoperable disease
subclinical CK rises after otherwise straightforward coro- was not stated. Most of the cardiac mortality occurred in
nary intervention may differ fundamentally from sponta- the first year. Significant multivariate correlates of car-
neous non-Q wave myocardial infarction, there has been diac death were CK elevation (odds ratio 2.19) and vein
a recent spate of reports raising a flag of caution about the graft procedures (odds ratio 2.09). In a subgroup analysis
prognostic significance of CK-MB release without clini- focusing on raised CK-MB only in those with total CK ,
cally apparent explanation [4,5,7,8]. Just how concerned 23 normal [7], multivessel, vein graft, or DCA proce-
should the interventional community be? dures and thrombus-associated lesions were again more
The Beth Israel group [3] found that angiographic frequent. The strongest correlates with a raised CK-MB
abnormalities (dissection, side-branch occlusion, throm- were DCA followed by the occurrence of at least one
bus, transient no-reflow) could be identified in .40% of in-lab complication: side-branch compromise, transient
patients with raised CK-MB. Only 13 of 558 cases (2.3%) in-lab vessel closure, or a greater residual stenosis. For
had a rise in CK-MB greater than 5 times the upper limit the entire population (n 5 4,461), the annual cardiac
of normal. This was associated with a trend (P 5 0.08) mortality rate was 2–3% in the first year, then 1–2% in
towards decreased survival whereas multivessel and vein years 2–5 and 0–0.2% beyond 5 years [7]. For the group
graft disease, reflecting more advanced atherosclerosis, with raised CK-MB (n 5 708), the absolute increase in
remained independent predictors. Multivariate analysis cardiac mortality risk was of the order of only 1%, with a
confirmed that multivessel disease and depressed ejection multivariate risk ratio of 1.27. It is not reported what
fraction (LVEF) were the strongest predictors of reduced proportion of cardiac deaths were sudden, arrhythmic, or
survival, consistent with early observations of the natural occurred in the context of acute infarction or progressive
history of coronary disease. heart failure. Neither is it clear how follow-up of patients
Subclinical CK-MB release is more common after with small CK rises should differ from usual manage-
directional atherectomy (DCA) than balloon PTCA (19 ment.
vs. 8%) [9], and in vein graft intervention than native
vessel procedures largely because of the potential for Division of Cardiology, Department of Medicine, University of
distal embolization [10]. In the CAVEAT database Maryland School of Medicine, Baltimore, Maryland
(n 5 1,012), of 60 patients with abrupt vessel closure,
55% had raised CK $ 23 or CK-MB $ 33 [4]. *Correspondence to: Dr. Warren K. Laskey, Cardiac Catheterization
Laboratory, University of Maryland Hospital, 22 South Greene Street,
Regression analysis showed these enzyme indices to be Baltimore, MD 21201-1595.
predictive of all-cause 1-year mortality when only base-
line characteristics were considered (P 5 0.038) but not Received 21 January 1997; Accepted 29 January 1997

r 1997 Wiley-Liss, Inc.

TABLE I. Elevation of Total Creatine Kinase or Myocardial Isoenzyme Activity in PTCA Procedures
No. .
Timing threshold Significant
Reference n of CK Threshold (%) Mortality covariates Comments
Oh et al. [1] 128 $33 in 24 h CKMB $2% 25 (20) 0% at mean 10 m Periprocedural chest Retrospective
1979–82 pain
MI within 1 m
Side-branch occlusion
(32%)
Klein et al. [2] 249 q 6 h for 24 h CK $200 IU/L or 38 (15) 0% at hospital dis- Prospective; in-lab
1989–90 CKMB $4% charge event in 63%
Kugelmass 558 23 in 24 h CKMB $10 IU/L 64 (11.5) 12.5% (cardiac) at Advanced age Retrospective; Vein
et al. [3] (DCA 271) mean 2 years Female gender graft intervention in
1988–92 (Stent 287) DCA use 27%; thrombus in
31%; side-branch
occlusion in 16%;
slow reflow in 25%
Harrington 1,012 q 12 h for 24 h CKMB $33 or 112 (11) 4.4% (all-cause) at Retrospective; native
et al. [4] (DCA 512) CK $23 1 year vessels only
1991–92 (PTCA 500)
Abdelmeguid 4,664 $23 in 24 h CK $23 184 (4) 5.9% (cardiac) at Vein graft procedures Retrospective; only
et al. [5] 1 year Thrombotic lesions vein graft proce-
1984–91 Complex lesions dures a multivariate
In-lab vessel closure correlate of cardiac
Side-branch compro- death
mise
Dissection
Distal embolization
Vasopressors for
hypotension

In prognostic evaluation, risk ratio may obscure actual negative predictive value would give her an optimistic
event rates because it is simply a summary measure of the chance of being alive at 2 years.
likelihood of an adverse event. This ratio may therefore These data, in common with earlier analyses, derive
appear significant even if the event rate in the group with from interventional practices and outcomes that are no
a risk factor is low, as long as the event rate in the group longer applicable to the current era. The rapid evolution
without the risk factor is lower. A positive predictive of interventional approaches to patients presenting with
value allows us to quote a more meaningful risk estimate acute coronary syndromes, new devices, more refined
to our patients. In the Cleveland Clinic database [7], techniques, and adjunctive pharmacotherapy have dramati-
patients with raised CK-MB were 27% (95% confidence cally altered the landscape. The reported associations
interval 2 to 60%) more likely than patients without between elevated CK/CK-MB activities, procedural fac-
raised CK-MB to suffer cardiac death over a mean of 36 tors and outcome may lose statistical significance when,
months. However, the vast majority of patients with as in current practice, the prevalence of ‘‘high-risk’’
raised CK-MB survived. No predictive values are re- procedures increases. Furthermore, virtually all studies
ported by the Cleveland Clinic investigators; from the have been retrospective and thus suffer from some bias in
Kaplan-Meier survival curves, however, the positive the sampling of blood for enzyme activities.
predictive value of elevated CK-MB is estimated to be No retrospective analysis can control for all factors
only 2–3% at 1 year and 3–5% at 2 years [7]. At face influencing outcome, nor be sensitive enough to detect all
value, this would imply that a 40-year-old man with a interactions between various factors. Multiple sources of
good left ventricle will have a 3–5% chance of cardiac co-linearity exist; baseline left ventricular function, le-
death within 2 years if his technically successful DCA for sion complexity, previous revascularization, whether there
single vessel disease results in a raised CK-MB. Con- is a stent or surgical option, choice of device, ease of
versely, an 80-year-old woman with diabetes, LVEF 25%, procedure, success and completeness of revasculariza-
diffuse three-vessel disease, previous CABG 3 4 (redo tion, residual disease, CK release, and outcome are not
twice), and a normal CK-MB despite intervention on her unrelated. Other factors that may determine CK release
last remaining 12-year-old vein graft, would still be include the viability of myocardium subtended by target
expected to have a guarded prognosis although a 97% vessel, degree of collateralization, diabetes mellitus,
 Creatine Kinase Release After Catheter-Based Coronary Intervention 119

cardiogenic shock, intra-aortic counterpulsation, and counseled), and the case should be undertaken only by
whether a ‘‘bailout’’ procedure occurred. Three-vessel experienced operators. Some post-procedural cardiac
disease and poor LVEF should be forced into the multivar- deaths, however, are not amenable to preventive mea-
iate analysis to determine whether CK release truly has sures. We may know who is at higher risk of dying
incremental prognostic value. The prognostic impact of but—even as interventionists—we may not always be in
LVEF ,40% and ‘‘multivessel disease’’ is different in a position to do very much about it.
different patient subsets: repeat angioplasty, first-time
angioplasty, redo CABG, and first-time CABG. Intu-
itively, a fivefold elevation in CK-MB activity is worse REFERENCES
than a twofold rise, because CK-MB release implies 1. Oh JK, Shub C, Ilstrup DM, Reeder GS: Creatine kinase release
muscle loss, which may be reflected in reduction of after successful percutaneous transluminal coronary angioplasty.
LVEF. Therefore, the adverse prognosis associated with Am Heart J 109:1225–1231, 1985.
CK-MB release may simply reflect myocardial necrosis 2. Klein LW, Kramer BL, Howard E, Lesch M: Incidence and clinical
whether subclinical or otherwise. Although any degree of significance of transient creatine kinase elevations and the diagno-
sis of non-Q wave myocardial infarction associated with coronary
necrosis is undesirable, a twofold CK-MB rise may be
angioplasty. J Am Coll Cardiol 17:621–626, 1991.
more important—if myocardial reserve is already re- 3. Kugelmass AD, Cohen DJ, Moscucci M, Piana RN, Senerchia C,
duced—than a fivefold rise in someone with well- Kuntz RE, Baim DS: Elevation of the creatine kinase myocardial
preserved LVEF. When the measurement is continuous, isoform following otherwise successful directional coronary ather-
receiver operating characteristic (ROC) analysis avoids ectomy and stenting. Am J Cardiol 74:748–754, 1994.
the need to set an arbitrary dichotomous threshold, and 4. Harrington RA, Lincoff AM, Califf RM, Holmes DR Jr, Berdan
LG, O’Hanesian MA, Keeler GP, Garratt KN, Ohman EM, Mark
may help to determine appropriate ‘‘cut-points.’’ How DB: Characteristics and consequences of myocardial infarction
well a prognostic index discriminates between patients after percutaneous coronary intervention: Insights from the Coro-
with and without adverse events can also be judged by the nary Angioplasty Versus Excisional Atherectomy Trial (CAVEAT).
area under the ROC curve. J Am Coll Cardiol 25:1693–1699, 1995.
Future studies should progress from describing the 5. Abdelmeguid AE, Ellis SG, Sapp SK, Whitlow PL, Topol EJ:
Defining the appropriate threshold of creatine kinase elevation
association in retrospect and attempt to prospectively after percutaneous coronary interventions. Am Heart J 131:1097–
validate the concern while testing whether adjunctive 1105, 1996.
strategies incorporating platelet IIB/IIIA receptor antago- 6. Abdelmeguid AE, Whitlow PL, Sapp SK, Ellis SG, Topol EJ:
nists, magnesium, stenting, and debulking can ameliorate Long-term outcome of transient, uncomplicated in-laboratory
CK-MB release. CK-MB release should certainly be coronary artery closure. Circulation 91:2733–2741, 1995.
7. Abdelmeguid AE, Topol EJ, Whitlow PL, Sapp SK, Ellis SG:
documented prospectively in the evaluation of experimen- Significance of mild transient release of creatine kinase-MB
tal technology to address the scientific skepticism sur- fraction after percutaneous coronary interventions Circulation
rounding its clinical importance. Based on the available 94:1528–1536, 1996.
published data, identifying a subclinical CK-MB rise as a 8. Abdelmeguid AE, Topol EJ: The myth of the myocardial ‘‘infarct-
major complication after intervention [7] is premature let’’ during percutaneous coronary revascularization procedures.
Circulation 94:3369–3375, 1996.
and, as previously cautioned [3], may distort the risk ratio 9. Topol EJ, Leya F, Pinkerton CA, Whitlow PL, Hofling B,
in the evaluation of new devices. In all likelihood, Simonton CA, Masden RR, Serruys PW, Leon MB, Williams DO:
CK-MB release is an epiphenomenon in high-risk pa- A comparison of directional atherectomy with coronary angio-
tients who can already be identified as such before (and plasty in patients with coronary disease. N Engl J Med 329:221–
during) intervention. In such cases, the question should 227, 1993.
10. Hinohara T, Robertson GC, Selmon MR, Vetter JW, McAuley BJ,
be asked: Is there a better or safer alternative? If not, more Sheehan DJ, Simpson JB: Directional coronary atherectomy:
caution will be required, recognizing that the procedural Complications and management. Cathet Cardiovasc Diagn (Suppl)
risk is higher (and the patient and family appropriately 1:61–71, 1993.