Author + information
- Received November 22, 2002
- Revision received March 6, 2003
- Accepted March 20, 2003
- Published online August 6, 2003.
- Massoud A Leesar, MD*,
- Marcus F Stoddard, MD*,
- Yu-Ting Xuan, PhD*,
- Xian-Liang Tang, MD* and
- Roberto Bolli, MD*,* ()
- ↵*Reprint requests and correspondence:
Dr. Roberto Bolli, Division of Cardiology, ACB, Third Floor, 550 South Jackson Street, University of Louisville, Louisville, Kentucky 40292, USA.
Objectives The objective of this study was to use electrocardiogram (ECG)-independent parameters to determine whether preconditioning (PC) exists in humans during percutaneous transluminal coronary angioplasty (PTCA).
Background Several studies suggest that both ischemia and adenosine induce PC in the human heart during PTCA. However, because almost all of these studies relied on ST-segment shifts as indicators of the severity of ischemia, their conclusions continue to be questioned, and the very existence of ischemic or adenosine PC in humans remains controversial.
Methods Eighteen patients received either intracoronary adenosine (n = 9) or normal saline (n = 9); 10 min later, they underwent PTCA (three 2-min balloon inflations 5 min apart).
Results Compared with the first inflation, in untreated patients the second and third inflations were associated with less systolic dysfunction (two-dimensional echocardiography), less diastolic dysfunction (color M-mode echocardiography), less lactate production, and less H+release into the great cardiac venous blood. In adenosine-treated patients, the extent of all of these abnormalities during the first inflation was less than in untreated patients and did not change with subsequent inflations.
Conclusions Previous exposure to a brief episode of ischemia (first balloon inflation) or to adenosine produces concordant decreases in ECG, subjective, mechanical, and metabolic manifestations of ischemia during PTCA. These data support the concept that both ischemic PC and pharmacologic PC exist in humans and that PTCA is a useful clinical setting in which to discern their mechanism.
In experimental animals, brief episodes of ischemia render the myocardium more resistant to subsequent ischemic episodes, a phenomenon known as ischemic preconditioning (PC) (1–3). Because ischemic PC is the most powerful cardioprotective intervention discovered to date, there is considerable interest in exploiting it to protect the ischemic myocardium in patients with coronary artery disease (1–3). Among all of the clinical situations potentially associated with ischemic PC, percutaneous transluminal coronary angioplasty (PTCA) has been studied most extensively. Numerous studies have shown that, in the course of PTCA, the severity of myocardial ischemia (as assessed by ST-segment elevation and chest pain) decreases with subsequent inflations, suggesting the development of ischemic PC (4–10). Percutaneous transluminal coronary angioplasty is an attractive setting for investigating the mechanism of ischemic PC in humans because: the duration of ischemia is well defined; measurements can be made before, during, and after ischemia at precise time points; pretreatment is possible; and potentially confounding variables are minimized. Thus, if the existence of ischemic PC during PTCA can be definitively proven, this could be a useful model for gaining insights into the mechanisms underlying cardioprotection in humans.
A considerable body of evidence in experimental animals suggests that adenosine plays a central role as an endogenous mediator of ischemic PC (1). We have previously found that, judging from the ST-segment shifts on the electrocardiogram (ECG) and the severity of chest pain, pretreatment with adenosine induces an early PC effect against ischemia in patients undergoing PTCA (5). These results suggest that adenosine may be useful for PC human myocardium, a concept that may have significant therapeutic implications.
Despite the data reviewed here, however, both the use of PTCA for the study of ischemic PC and the ability of adenosine to mimic ischemic PC in humans continue to be questioned, based on the fact that almost all of the studies performed to date have relied on ST-segment shifts as indicators of the severity of ischemia (9). These changes in cardiac repolarization are often viewed as “soft evidence” of protection or lack thereof, and as a result, the very existence of ischemic PC during PTCA continues to be controversial (9). A recent study has challenged the concept that changes in the ST-segment shift are predictive of changes in infarct size in experimental animals subjected to ischemic PC (11). Thus, it remains unclear whether the attenuation of ST-segment elevation during subsequent inflations in the course of PTCA indicates a decrease in the severity of ischemia.
One of the earliest and most fundamental consequences of ischemia is the development of anaerobic glycolysis, resulting in production of lactate and H+(12). Furthermore, one of the characteristic features of ischemic PC in experimental animals is that the preconditioned myocardium exhibits a reduced accumulation of lactate and H+(13–16). This appears to be a consequence of reduced myocardial energy demands, leading to preservation of high-energy phosphates and decreased glycolysis (1,13,14). Accordingly, to verify the existence of PC during PTCA, we examined whether ischemic PC and adenosine PC affect coronary sinus blood pH and lactate levels.
The overall objective of this study was to use ECG-independent parameters to determine whether PC exists in humans. To this end, we compared the end points classically used in the setting of PTCA (ST-segment shift and chest pain) with changes in left ventricular (LV) systolic (segmental wall motion) and diastolic (LV inflow propagation rate [Vp]) properties assessed by echocardiography and with changes in lactate production and cardiac venous blood pH during or after three consecutive balloon inflations in patients undergoing PTCA. These analyses were performed both in untreated patients and in patients pretreated with intracoronary adenosine, which enabled us to examine both ischemic PC and adenosine-induced (pharmacologic) PC. We used the same protocol and the same dosage of adenosine that were previously shown by ECG criteria to induce PC (5).
The patient population for this study consisted of 23 subjects referred for PTCA of an isolated obstructive lesion (internal diameter reduction >70% by visual assessment) in the proximal two-thirds of the left anterior descending coronary artery. Patients were prospectively selected on the basis of the following criteria: 1) no angiographically visible collateral vessels; 2) no history or ECG evidence of prior myocardial infarction in the territory supplied by the vessel undergoing PTCA; 3) no conduction defects on the ECG; 4) no evidence of LV hypertrophy on the echocardiogram; 5) no baseline ST-segment abnormalities on the surface or intracoronary ECG; and 6) no use of sulfonylureas. The protocol was approved by the Institutional Review Board, and all patients gave written informed consent.
Percutaneous transluminal coronary angioplasty was performed by a standard technique as described (5,7,10). In this single-blind study, patients were randomly allocated to a control or an adenosine-treated group. Adenosine (Adenocard, Fujisawa Pharmaceutical Co., Deerfield, Illinois) was dissolved in sterile normal saline (20 mg in 50 ml) and infused at a rate of 2 mg/min over 10 min through a 2.2F Tracker coronary-infusion catheter (Boston Scientific Inc., Minneapolis, Minnesota) into the proximal coronary artery. The control group received an equivalent volume of vehicle (normal saline). A 6F sheath was placed in the right internal jugular vein or right femoral vein and a 6F multipurpose catheter was advanced and positioned in the great cardiac vein to withdraw blood samples. The stability of the catheter position was verified throughout the procedure by comparison of fluoroscopic images and by contrast injection 1 to 2 min before each blood sampling. Paired blood samples were obtained simultaneously from the aorta and the great cardiac vein immediately before each balloon inflation, immediately after each balloon deflation, and 2 min after the third balloon deflation for determination of blood pH and plasma levels of lactate. The lactate extraction ratio (LER) was calculated as: (A − GCV/A) × 100, where A = plasma lactate in the aorta and GCV = plasma lactate in the great cardiac vein (11). After infusion of adenosine or vehicle, the Tracker catheter was removed. After a 10-min drug-free period, patients underwent three balloon inflations, each lasting 120 s, interspersed with 5-min periods of reperfusion.
Assessment of ST-segment shift and chest pain
The ECG recordings were taken as previously described (5,7,10)and analyzed by a cardiologist who had no knowledge of the study protocol. At the end of each inflation, the intensity of the cardiac pain was assessed using a visual-analog scale (17)as described (5,7,10).
The methods for quantitating regional wall motion on the two-dimensional echocardiograms have been described in detail (5,18). To assess the consequences of ischemia on LV relaxation, the Vp was measured at the end of each inflation as described (19). Specifically, the slope of the color-chamber interface movement in early diastole was assessed from color M-mode echocardiography performed along a line that connects the center of the mitral valve annulus to the LV apex in a four-chamber view (19). The echocardiographic studies were analyzed by an echocardiographer (M.F.S.) who had no knowledge of the treatment.
Lactate and pH measurements
Blood samples for lactate analysis were collected in pre-cooled tubes and centrifuged immediately, and the plasma was stored at −20°C until analyzed. Samples were analyzed using a commercial kit (Boehringer, Mannheim, GMBH Diagnostica, Germany). Blood gases and pH were determined in the hospital laboratory immediately after the procedure.
All data are reported as means ± SEM. ST-segment shifts, chest pain score, chordal shortening, LV ejection fraction, lactate extraction, and coronary sinus blood pH were analyzed with two-way repeated-measures analysis of variance. Post hoc contrasts between groups at various time points or between time points within one group were performed with Student ttests for unpaired or paired data, as appropriate, using the Bonferroni correction (20).
Nine patients in the control group and nine in the adenosine-treated group met the criteria detailed under Methods and had technically adequate intracoronary and surface ECGs associated with complete resolution of ischemia between balloon inflations. Complete resolution of ischemia was defined as chest pain resolution and return of the ST-segment on the intracoronary and surface ECGs to within 1 mm of baseline during the 5 min that elapsed between the first, second, and third balloon inflations. Five patients were excluded from the study because the intracoronary ECG signals became inadequate during balloon inflations (three patients) or because the echocardiographic window was not adequate to delineate the endocardial border (two patients).
The clinical features of the control and adenosine-treated patients are outlined in Table 1. There were no significant differences between the two groups.
The anatomic and hemodynamic features of the study population are summarized in Table 2. Heart rate and arterial blood pressure did not differ between the two groups during the three inflations (data not shown). The rate-pressure product was also similar (Table 2). There was no ECG or enzymatic evidence of myocardial injury in any patient.
ECG manifestations of myocardial ischemia
All patients exhibited ST-segment elevation. The values reported herein were measured at 120 s into each inflation. In the control group, the ST-segment shift was significantly greater during the first balloon inflation than during the second and third inflations on the intracoronary ECG (27 ± 5 mm vs. 16 ± 3 and 13 ± 2 mm, respectively) (Fig. 1). In contrast, in the adenosine-treated group there were no differences in the ST-segment shift during the first, second, and third balloon inflations on the intracoronary ECG (7 ± 2 mm, 9 ± 2 mm, and 6 ± 2 mm, respectively) (Fig. 1) or on the surface ECG (8 ± 2 mm, 7 ± 2 mm, and 8 ± 2 mm, respectively) (Fig. 1).
On the intracoronary ECG, the ST-segment shift was significantly smaller in the adenosine-treated group than in the control group during each of the three inflations (7 ± 2 mm vs. 27 ± 5 mm [−74%], p < 0.05, during the first inflation; 9 ± 2 mm vs. 16 ± 2 mm [−43%], p < 0.05, during the second; and 6 ± 2 mm vs. 13 ± 2 mm [−53%], p < 0.05, during the third inflation) (Fig. 1). The intracoronary ST-segment shift recorded during the first inflation in the adenosine group was significantly (p < 0.05) less than that recorded during the third inflation in the control group (Fig. 1); as a result, the reduction in ST-segment shift afforded by adenosine during the first inflation (−74% vs. the first inflation in control patients) was greater than that afforded by ischemic PC during the third inflation in the control group (−53% vs. the first inflation in this group).
On the surface ECG, the ST-segment shift was significantly smaller in the adenosine-treated group than in the control group during the first inflation (8 ± 2 mm vs. 16 ± 5 mm, respectively; p < 0.05) but did not differ significantly between the two groups during the second and third inflations (7 ± 2 mm vs. 12 ± 2 mm and 8 ± 2 mm vs. 8 ± 2 mm, respectively; p = NS) (Fig. 1). The fact that the effects of adenosine on the ST-segment shifts were less striking on the surface than on the intracoronary ECG probably reflects the greater sensitivity of the latter for detecting ischemia in the perfusion bed of the PTCA artery (5,21).
In the control group, the severity of chest pain was significantly greater during the first inflation than during the second and third inflations (74 ± 6 mm vs. 58 ± 6 mm and 37 ± 7 mm, respectively) (Fig. 1). In contrast, in the adenosine-treated group the chest pain score did not differ significantly during the first, second, and third inflations (41 ± 8 mm, 37 ± 8 mm, and 35 ± 8 mm, respectively) (Fig. 1). The chest pain score was significantly smaller in the adenosine-treated group than in the control group during the first inflation (−44% [p < 0.05]) and during the second inflation (−36% [p < 0.05] (Fig. 1).
Echocardiographic and Doppler data
In the control and adenosine-treated groups, chordal shortening in the segments that received the infusion of normal saline or adenosine averaged 7.9 ± 0.3 mm and 7.8 ± 0.3 mm before the infusion and 8.1 ± 0.3 mm and 8.0 ± 0.3 mm after the infusion, respectively. Thus, administration of adenosine had no appreciable effect on regional LV wall motion. In the control group, chordal shortening decreased markedly (by 66 ± 5%) during the first balloon inflation and recovered 5 min after deflation (Fig. 2). During the second and third inflations, the decrease in chordal shortening was significantly less than during the first inflation (49 ± 6% and 41 ± 7%, respectively; p < 0.05 vs. first inflation) (Fig. 2). In the adenosine-treated group, the reduction in chordal shortening during the first inflation was significantly smaller than in the control group (−48 ± 4% vs. −65 ± 5%, p < 0.05) (Fig. 2). Furthermore, in contrast to the control group, in the adenosine-treated group there were no significant differences in chordal shortening during the first, second, and third balloon inflations (−48 ± 4%, −49 ± 4%, and −37 ± 7%, respectively) (Fig. 2).
In the control group, the Vp decreased by 55 ± 5% and 50 ± 5% during the first and second balloon inflations, respectively (Fig. 2). During the third inflation, the decrease in Vp was significantly less (−37 ± 4%; p < 0.05 vs. first and second inflations) (Fig. 2). In the adenosine-treated group, the reduction in Vp during the first inflation was significantly smaller than in the control group (−30 ± 4% vs. −55 ± 5%; p < 0.05) (Fig. 2). Furthermore, in contrast to the control group, in the adenosine-treated group there were no significant differences in the Vp during the first, second, and third balloon inflations (−30 ± 4%, −35 ± 4%, and −37 ± 4%, respectively) (Fig. 2).
Metabolic parameters (LER and pH)
In all patients, the LER was positive at baseline, reflecting lactate extraction, and became negative immediately after balloon inflations, reflecting lactate production (Table 3, Fig. 3). In the control group, the LER was significantly lower after the first balloon inflation than after the second and third inflations (−136 ± 30% vs. −77 ± 23% and −80 ± 32%, respectively [p < 0.05]) (Fig. 3). In contrast, in the adenosine-treated group, the LER did not change appreciably after the first, second, and third balloon inflations (−68 ± 22%, −63 ± 15%, and −83 ± 28%, respectively) (Fig. 3). The LER was significantly higher in the adenosine-treated group than in the control group after the first balloon inflation (−68 ± 22% vs. −136 ± 30%, p < 0.05) (Fig. 2) and did not differ significantly between the two groups after the second and third inflations (−63 ± 15% vs. −77 ± 23% and −83 ± 28% vs. −80 ± 32%, respectively (Fig. 3).
In all patients, the transcardiac gradient in blood pH ([A-GCV] pH) was positive at baseline and increased immediately after balloon inflations, reflecting a decrease in GCV pH (Table 3, Fig. 4). In the control group, the (A-GCV) pH was significantly greater after the first balloon inflation than after the second and third inflations (0.15 ± 0.02 vs. 0.09 ± 0.02 and 0.10 ± 0.02, respectively [p < 0.05]) (Fig. 4). In contrast, in the adenosine-treated group the (A-GCV) pH remained essentially unchanged after the first, second, and third balloon inflations (0.08 ± 0.01, 0.07 ± 0.01, and 0.07 ± 0.01, respectively) (Fig. 4). The (A-GCV) pH was significantly smaller in the adenosine-treated group than in the control group after the first balloon inflation (0.08 ± 0.01 vs. 0.15 ± 0.02, p < 0.05) (Fig. 4) and did not differ significantly between the two groups after the second and third inflations (0.07 ± 0.01 vs. 0.09 ± 0.02 and 0.07 ± 0.01 vs. 0.10 ± 0.02, respectively) (Fig. 4).
This study provides a comprehensive comparison of ECG, subjective, echocardiographic, and metabolic parameters of ischemia during PTCA. The results demonstrate that all of these parameters are shifted in parallel during ischemic or adenosine PC, indicating true alleviation of ischemia.
Over the past decade, the setting of PTCA has been used extensively to investigate the mechanisms underlying the phenomenon of PC in humans (4–10). The results of these studies have provided numerous insights regarding the occurrence of ischemia-induced and pharmacologically induced PC. However, widespread acceptance of these concepts has been hindered by persisting concern that the major end point used in virtually all of the studies to date, namely, the ST-segment shift on the ECG, may not necessarily be a reliable marker of the severity of ischemia (9). This concern has been rekindled by a recent study in which changes in ST-segment elevation did not correlate with changes in infarct size in rabbits (11).
In order to address this fundamental uncertainty that undermines the use of PTCA as a system to study the mechanism of PC in humans, we examined two metabolic parameters of ischemia, namely, lactate production and acidosis. As expected, we found that in control patients the ischemic myocardium switched from lactate extraction to lactate production during the first balloon inflation (Fig. 3). Five minutes after release of the first inflation, the myocardium had reverted to its baseline status of lactate extraction. Lactate production was observed again during the second and third balloon inflations, but the magnitude was less than during the first inflation. Importantly, the myocardium reverted to lactate extraction at each subsequent baseline (before the second and third inflations), and the baseline values of lactate extraction were similar between control and adenosine-treated patients (Table 3). Our data are consistent with those reported by Deutsch et al. (4)and Eltchaninoff et al. (22); the major difference is that we measured lactate production/extraction in the same patients in which we measured ECG, subjective, and echocardiographic parameters of ischemia, whereas those previous reports (4,22)used separate subsets of patients.
In accordance with the changes in lactate extraction, in control patients the release of the first balloon inflation was associated with a significant fall in great cardiac vein blood pH, which became attenuated upon release of the second and third inflations (Table 3, Fig. 4). To the best of our knowledge, this is the first report in humans to document that the fall in coronary venous blood pH decreases during consecutive balloon inflations, suggesting that ischemic PC reduces myocardial H+production. Taken together, the measurements of lactate production and pH demonstrate that the first balloon inflation modified the metabolic response of the myocardium to the second and third inflations in a manner that is consistent with the development of ischemic PC. Studies in experimental animals have shown that preconditioned myocardium exhibits less lactate production and less acidosis than non-preconditioned myocardium (13–16). The mechanism responsible for this is unclear but is thought to involve a decrease in the rate of glycolysis as a result of downregulated energy consumption (12–16).
We also studied a group of patients preconditioned with adenosine. We have previously shown that this dose of adenosine induces an early PC effect, judging from the changes in ST-segment elevation on the intracoronary and surface ECG and from the severity of chest pain (5). The present study demonstrates that these salubrious actions of adenosine are associated with parallel effects on lactate extraction (Fig. 3) and acidosis (Table 3, Fig. 4), further confirming the ability of adenosine to mimic ischemic PC in the clinical setting. The data reported herein represent the first evidence that adenosine induces the metabolic features of the preconditioned phenotype (less accumulation of lactate and H+) in humans. In addition, this study demonstrates that adenosine PC improves systolic function as well as diastolic filling during ischemia.
In contrast to previous studies (22,23)in which a qualitative echocardiographic method was used to assess the improvement of segmental wall motion abnormalities during successive balloon inflations, in the present study regional wall motion during successive balloon inflations was analyzed using a quantitative method (the centerline method) that corrects for ventricular translation (5,18). Previous investigations have also used color M-mode Doppler echocardiography to assess the diastolic properties of the myocardium during PTCA (19,24). These studies (19,24)have shown that during the acute ischemia associated with balloon inflation, LV filling is delayed and Vp (an index of LV relaxation) decreases; however, color M-mode Doppler echocardiography has not been studied previously in the setting of ischemic PC. To the best of our knowledge, the present study is the first evidence that ischemic PC attenuates the LV relaxation abnormalities during acute myocardial ischemia in humans.
Taken together, the results presented herein demonstrate that two different forms of PC (ischemic PC and adenosine PC) alleviate the severity of myocardial ischemia, as determined by six separate end points (ST-segment elevation, chest pain severity, regional wall motion abnormalities, diastolic abnormalities, and two metabolic indices of ischemia). Unlike previous investigations, all of these six variables were measured in each patient, thereby enabling us to compare their changes in the same subjects. By using parameters independent of the ECG (i.e., LV chordal shortening, Vp, myocardial lactate extraction, and blood pH), we provide incontrovertible evidence that the severity of ischemia is indeed attenuated on subsequent balloon inflations during PTCA and that adenosine alleviates the severity of ischemia during the first balloon inflation, supporting the notion that both ischemic PC and pharmacologic PC do indeed exist in humans and that the changes in ST-segment elevation previously described (4–10)are truly indicative of alleviation of myocardial ischemia. The present results have important conceptual implications. The results have also significant methodologic implications because they support the utility of PTCA as a clinical setting to dissect the mechanisms responsible for PC in the human heart, although it is unlikely that the more distal events in the signaling cascade of PC can be assessed in this model.
☆ Supported in part by NIH grants R01 HL-43151, HL-55757, HL-68088, and HL-70897 (Dr. Bolli), and HL-65660 (Dr. Xuan).
- great cardiac vein
- lactate extraction ratio
- left ventricular
- percutaneous transluminal coronary angioplasty
- left ventricular inflow propagation rate
- Received November 22, 2002.
- Revision received March 6, 2003.
- Accepted March 20, 2003.
- American College of Cardiology Foundation
- Bolli R.
- Yellon D.M.,
- Dana A.
- Deutsch E.,
- Berger M.,
- Kussmaul W.G.,
- Hirshfeld J.W.,
- Herrmann H.C.,
- Laskey W.K.
- Leesar M.A.,
- Stoddard M.F.,
- Ahmed M.,
- Broadbent J.,
- Bolli R.
- Tomai F.,
- Crea F.,
- Gaspardone A.,
- et al.
- Leesar M.A.,
- Stoddard M.F.,
- Manchikalapudi S.,
- Bolli R.
- Tomai F.,
- Crea F.,
- Gaspardone A.,
- et al.
- Tomai F.,
- Crea F.,
- Chiariello L.,
- Gioffre P.A.
- Leesar M.A.,
- Stoddard M.F.,
- Dawn B.,
- Jasti V.G.,
- Masden R.,
- Bolli R.
- Birincioglu M.,
- Yang X.M.,
- Critz S.D.,
- Cohen M.V.,
- Downey J.M.
- Murry C.E.,
- Richard V.J.,
- Reimer K.A.,
- Jennings R.B.
- Tatsumi T.,
- Matoba S.,
- Kobara M.,
- et al.
- Rehring T.F.,
- Shapiro J.I.,
- Cain B.S.,
- et al.
- Duval-Moulin A.M.,
- Dupouy P.,
- Burn P.,
- et al.
- Wallenstein S.,
- Zucker C.L.,
- Fleiss J.L.
- Friedman P.L.,
- Shook T.L.,
- Kirshenbaum J.M.,
- Selwyn A.P.,
- Ganz P.
- Hauser A.M.,
- Gangadharan V.,
- Ramos R.G.,
- Gordon S.,
- Timmis G.C.
- Stugaard M.,
- Smiseth O.A.,
- Risoe C.,
- Ihlen H.