(Circulation. 1995;92:2266-2275.)
© 1995 American Heart Association, Inc.
Articles |
Pharmacological Evidence for the Persistent Activation of ATP-Sensitive K+ Channels in Early Phase of Reperfusion and Its Protective Role Against Myocardial Stunning
From the Department of Physiology (S.S., T.S., T.A., M.A.) and the First Department of Internal Medicine (T.S., T.S.), Oita Medical University, Japan.
Correspondence to Sakuji Shigematsu, MD, Department of Physiology, Oita Medical University, 1-1, Hasama-machi, Oita 879-55, Japan.
 |
Abstract |
|---|
 Top Abstract IntroductionMethods Results Discussion References |
|---|
Background The activation of cardiac ATP-sensitive potassium channels is reported to protect myocardium during ischemia. However, the behavior and role of this channel during reperfusion remain uncertain.
Methods and Results Guinea pig right ventricular walls were studied by use of microelectrodes and a force transducer. Each preparation was perfused via the coronary artery at a constant flow rate and was stimulated at 3 Hz. In the first protocol, the preparation was subjected to 10 minutes of no-flow ischemia, which was followed by 60 minutes of reperfusion. Introduction of ischemia shortened the action potential duration (APD) to 58.7±3.1% of the preischemic values, in association with a decrease in the resting membrane potential (by 12±0.8 mV) and action potential amplitude (by 34.6±1.8 mV). On reperfusion, although the APD was restored, it remained shortened for up to approximately 30 minutes of reperfusion. In the presence of glibenclamide (10 µmol/L), the shortening of the APD during ischemia was significantly attenuated and the restoration of APD after reperfusion was significantly facilitated. When glibenclamide was applied from the onset of reperfusion, the persistent APD shortening was significantly suppressed. The developed tension decreased during ischemia and recovered after 60 minutes of reperfusion (up to 92.0±6.4% of preischemic values) in the untreated preparations. The application of glibenclamide that was started before ischemia or from the onset of reperfusion significantly suppressed the recovery of contractility (P<.05 versus untreated preparations). In the second series of experiments, 20 minutes of no-flow ischemia and 60 minutes of reperfusion were applied. This protocol produced a sustained contractile dysfunction after reperfusion (to 34.0±3.2% of preischemic values). In the presence of cromakalim (2 µmol/L), the APD shortening was enhanced during both ischemia and the early reperfusion period. Cromakalim significantly improved the contractile recovery (to 79.3±4.1% of preischemic values, P<.05 versus untreated preparations). The application of cromakalim that was started from the onset of reperfusion also improved the contractile recovery during this phase and this effect was associated with enhanced APD shortening. However, the cromakalim-treated preparations demonstrated a higher incidence of ventricular fibrillation during reperfusion.
Conclusions Cardiac ATP-sensitive potassium channels are activated by ischemia, and a fraction of these channels remains activated during the early reperfusion phase. The resulting shortening of the APD prevents the heart from developing myocardial stunning.
Key Words: ischemia • reperfusion • action potentials • stunning • myocardial
|
Introduction |
|---|
|
TopAbstractIntroductionMethods Results Discussion References |
|---|
Cardiac K+ATP channels1 are activated as a result of a decrease in the level of intracellular ATP and shorten the duration of the action potential.2 3 Opening of the channel is also enhanced by increases in intracellular ADP4 and proton concentrations.5 6 All of these changes in the intracellular milieu occur in the setting of ischemia; the physiological role of K+ATP channels under conditions of ischemia is considered important.
Cole et al7 have shown that the activation of this channel plays a role in preserving cardiac function during ischemia. This observation was based on the finding that blockade of K+ATP channels by glibenclamide worsened the recovery of contractility after ischemia, whereas opening of this channel with pinacidil facilitated recovery.7 These findings suggest that the shortening of APD during ischemia, because of the activation of K+ATP channels, is cardioprotective. In line with these results, aprikalim, a new K+ATP channel–opening drug, prevented in vivo myocardial stunning.8 However, little is known about the behavior and pathophysiological role of this channel after reperfusion.
Recently an interesting phenomenon termed ischemic preconditioning has been reported9 in which a brief episode or episodes of ischemia result in an increased tolerance of the myocardium to a subsequent severe ischemia. However, this protective effect is lost when the interval between the initial brief ischemia and the subsequent severe ischemia is longer than 1 hour.10 Gross and Auchampach11 have suggested that activation of the K+ATP channels contributes to this phenomenon. All of these findings suggest that a "short-term memory effect" accounts for ischemic preconditioning. This effect may relate to persistent activation of K+ATP channels that is caused by the preceding brief ischemia.
We investigated the behavior and pathophysiological roles of K+ATP channels and focused our attention particularly on the early phase of reperfusion. We demonstrate here that the activation of K+ATP channels, elicited by preceding ischemia, persists after reperfusion and that this opening of K+ATP channels during the early reperfusion phase makes a major contribution to postischemic contractile recovery.
|
Methods |
|---|
|
TopAbstractIntroductionMethods Results Discussion References |
|---|
Preparation of Hearts
Male guinea pigs (300 to 400 g) were killed by cervical dislocation, and after tracheotomy each animal's chest was opened during artificial respiration. A polyethylene cannula (OD 1.7 mm, ID 1.3 mm) was inserted in a retrograde manner into the ascending aorta. After ligation of the distal part of the aorta, the heart was perfused with oxygenated Tyrode's solution under positive pressure (approximately 100 mm H2O) via the cannula, and the heart was excised en bloc with the proximal aorta. The proximal left coronary artery was ligated; both atria, the left ventricular free wall, and the ventricular septum were discarded, leaving only the right ventricular free wall and a small basal portion of the left ventricle surrounding the aortic root. All of these procedures were performed without interruption of coronary perfusion. The isolated right ventricular free wall preparation, the coronary artery of which was connected to a cannula by the aortic root, was mounted in the recording chamber and pinned at the base of the ventricle; the coronary perfusate was then delivered through a roller pump (MP-3, Tokyo Rikakikai Co). The flow rate was monitored by a custom-made flowmeter consisting of a droplet sensor (PG-602, PG-610, Keyence) and a pulse counter (AT-600G, Nihon Khoden) that was mounted between polyethylene tubes connecting the roller pump and the coronary artery. The flow rate was maintained at 1.0±0.2 mL · min-1 · g wet wt-1, with an intra-aortic perfusion pressure ranging from 40 to 50 mm H2O. To avoid leaks of the perfusate from the aortic valve, the ascending aorta was slightly stretched (2 to 3 mN) to expand the aortic valve leaflets. The preparation was superfused with substrate-free hypoxic Tyrode's solution (10 mL/min) gassed with nitrogen to minimize direct oxygen diffusion into the muscles from the epicardial and endocardial surfaces. The temperatures of these solutions were maintained at 37±0.5°C. These procedures ensured a sufficient supply of oxygen and substrate to the tissue, because normal action potentials could be recorded even from very marginal portions of the preparation, and a constant amplitude of twitch contractions was recorded for more than 2 hours after an equilibration period of 90 minutes. At the end of the experiments, 10 mL of indocyanine green was injected into the coronary artery via the cannula to confirm that all preparations had been well perfused without leaks from the aortic valve. All procedures met the guidelines stipulated by the Physiological Society of Japan and the Ethical Committee of Oita Medical University for Animal Experiments.
Electromechanical Recordings
The heart was stimulated at 3 Hz
throughout the experiment by
use of a pair of platinum electrodes. They were attached to the basal
portion of the preparation and connected to the isolation unit of an
electric stimulator (SEN-3201, SS-302J, Nihon Kohden). Square pulses of
5 milliseconds' duration, with a pulse strength 1.5-fold greater than
the threshold, were used to drive the preparation. Action potentials
were recorded from a cell that was located deep (usually six to
eight cell layers) in the subepicardial surface by use of a flexibly
mounted microelectrode that was suspended with a fine silver wire (200
µm OD). Microelectrodes (tip resistance, 20 to 30 M
) were made by
pulling of filamented capillary tubes (1.5 mm OD, Narishige) with a
pipette puller (PE-2, Narishige), and were filled with 3 mol/L KCl. A
direct current preamplifier (MEZ-7101, Nihon Kohden) with capacitance
compensation was used to record the transmembrane potential.
Contractile tension was recorded with a force transducer (TB-612T,
Nihon Kohden) connected to the apical end of the preparation. rTension
was adjusted to obtain the optimal developed tension (usually 1.0 to
1.5 mN). The membrane potential and contractile tension were monitored
on a multibeam oscilloscope (VC-9A, Nihon Kohden) and recorded
on a thermal arraycorder (WT-645G, Nihon Kohden). All data were stored
on magnetic tapes by use of a PCM data recording system
(PCM-501ES, Sony).
Solutions
The composition of oxygenated Tyrode's
solution was
(in mmol/L) NaCl 136.7, NaHCO3 11.9, KCl 5.4,
NaH2PO4 0.42, MgCl2 0.5,
CaCl2 1.8, and glucose 11 with a pH of 7.35 to 7.40 when
gassed with 97% O2 and 3% CO2.
PO2 of the solution was measured by an
O2 monitor (POG-200BA, Unique Medical) and found to be
>400 mm Hg. Hypoxic Tyrode's solution
(PO2<10 mm Hg) had the same composition as
above, except it contained no glucose (pH 7.35 to 7.40 gassed with 97%
N2 and 3% CO2). Stock solutions of
glibenclamide (1 mmol/L; a kind gift from Hoechst Japan Co) and
cromakalim (1 mmol/L; Sigma Chemical Co) were made by dissolving these
drugs in 5% dimethyl sulfoxide (Sigma Chemical Co). E-4031 (a kind
gift from Eizai Pharmaceutical Co) was dissolved in distilled water and
kept as a stock solution (10 mmol/L). An appropriate volume of each
stock solution was added to the oxygenated Tyrode's
solution immediately before use to make the various final
concentrations of each drug described below.
Experimental Protocol and Data Analysis
After equilibration
(>90 minutes) the coronary flow was
completely stopped by closing of an electromagnetic valve that was
placed at the very end of the tubing, resulting in no-flow
ischemia of the entire preparation (global
ischemia).
In protocol 1, the preparations were subjected to 10 minutes of no-flow ischemia that was followed by 60 minutes of reperfusion, in the absence or presence of glibenclamide, a K+ATP channel blocker. Glibenclamide (10 µmol/L) was applied (1) from 20 minutes before the introduction of no-flow ischemia until the end of the observation period (preischemic treatment, or the pretreatment group) or (2) from the beginning of reperfusion until the end of the observation period (postischemic treatment, or the posttreatment group). In some experiments, E-4031 (1 µmol/L) was used instead of glibenclamide (posttreatment only).
In protocol 2, the preparations were subjected to 20 minutes of no-flow ischemia that was followed by 60 minutes of reperfusion, in the absence or presence of cromakalim, a K+ATP channel opener. Cromakalim (2 µmol/L) was applied (1) from 20 minutes before the introduction of no-flow ischemia until the end of the observation period or (2) from the beginning of reperfusion until the end of the observation period. In some experiments, different concentrations of cromakalim (1 and 5 µmol/L) were applied from 20 minutes before ischemia until the end of the observation period.
All electrical and mechanical parameters that were stored on magnetic tapes were replayed and processed by use of a personal computer (PC-9801, NEC) equipped with an analog-to-digital converter (ADX-98, Canopus). The preparations that developed severe arrhythmias (sustained VT or Vf persisting for more than 30 seconds) were excluded from the electrical and mechanical analyses. Statistical significance was evaluated by two-way ANOVA and paired or unpaired t tests. A P value of less than .05 was regarded as statistically significant.
|
Results |
|---|
|
TopAbstractIntroductionMethods Results Discussion References |
|---|
Under conditions of a normal perfusion, characteristics of the action potentials that were recorded from the right ventricular muscle preparation were comparable to those that have been reported by other investigators.7 12 13 14 After the period of equilibration (>90 minutes), the preparation maintained stable electrical and contractile activities with no evidence of arrhythmias, for example, VPC, VT, or Vf, during a 3-hour period of perfusion (n=3). However, when the perfusion of the preparation was suddenly halted, several important time-dependent changes occurred in both electrical and contractile activities. Representative recordings of electrical and contractile activities before, during, and after 10 minutes of no-flow ischemia are shown in Fig 1
. Table 1
shows statistical analyses of the difference
in the mean values of APD90, APA, RMP, dTension, and
rTension.
|
|
Ten minutes of no-flow ischemia produced fully reversible
changes in electrical and contractile activities, and no severe
arrhythmias (VT or Vf) developed during the ischemia or
reperfusion period. The APD90 decreased after the onset of
ischemia and reached 59% of preischemic values in
10 minutes. On reperfusion, the APD90 was rapidly restored,
although it remained shortened until about 10 minutes of reperfusion
(Fig 1
and Table 1). The RMP depolarized by 12
mV during 10 minutes of
ischemia, and this depolarization persisted for approximately
10 minutes after the initiation of reperfusion. The APA was also
reduced by ischemia, to 71% of its preischemic
value, and this reduction persisted for 30 minutes after reperfusion
(Table 1).
To determine the role of K+ATP channels in the shortening of APD during ischemia and the early reperfusion period, which was associated with decreases in the RMP and APA, we examined the effect of glibenclamide, a potent blocker of K+ATP channels, on these action potential changes.
Effect of Glibenclamide on Action Potential
Fig
2 shows the effects of glibenclamide on changes
in APD90 before, during, and after 10 minutes of
no-flow ischemia. Application of glibenclamide (10
µmol/L) was started either (1) 20 minutes before introduction of
ischemia (pretreatment group, Fig 2A) or (2) at the onset of
reperfusion (posttreatment group, Fig 2B). Statistical analyses
of the change in electrical and contractile parameters are
summarized in Table 1.
|
, n=5) and absence (
,
n=7) of glibenclamide (10 µmol/L). The drug did not alter
APD90 under preischemic conditions, attenuated
the shortening of APD90 during ischemia, and
facilitated the recovery of APD90 after reperfusion. B,
Comparison of the changes of APD90 in the absence (
, n=5) of glibenclamide;
glibenclamide was applied from the onset of reperfusion.
APD90 in the early reperfusion phase was obviously
lengthened by glibenclamide. Results indicate that the
K+ATP channels were activated during
ischemia and remained in the activated state for some
time (
Glibenclamide had no effect on
electrical activity
(APD90, RMP, or APA) or mechanical activity
(rTension or dTension) under normal (preischemic)
conditions. However, in the presence of glibenclamide (pretreatment
group), the extent of APD shortening caused by ischemia was
markedly attenuated, and the recovery of APD90 after
reperfusion was faster than in the untreated group (Fig 2A).
The
APD90 shortened only by 28.1±2.8% during 10 minutes of
ischemia in treated preparations versus 41.3±3.1% in
untreated preparations (P<.05). The APD90 was
restored to up to 98.6±2.6% of the preischemic values
within 5 minutes of reperfusion in treated preparations versus
90.5±2.7% in untreated preparations (P<.05). Pretreatment
with glibenclamide slightly facilitated the depolarization of RMP after
10 minutes of ischemia but had no effect on APA. After
reperfusion the drug slightly accelerated the rate of recovery of
APA but diminished the recovery of RMP.
To gain further insight into
the contribution of
K+ATP channels to the persistent shortening of
APD in the early reperfusion phase, we administrated glibenclamide (10
µmol/L) starting at the onset of reperfusion (Fig 2B). The
time
course of APD shortening during ischemia was identical to that
previously seen in untreated preparations; however, during reperfusion,
the APD90 returned to preischemic values more
rapidly than in the untreated group (98.0±1.7% of the
preischemic values at 5 minutes of reperfusion versus
90.5±2.7% in untreated preparations, P<.05). The APA
returned to preischemic values slightly faster in treated
than in untreated preparations, although the recovery of RMP was
significantly delayed in the former (Table 1). These
observations lend
support to the notion that an outward K+ current through
K+ATP channels, which is activated
during ischemia, remains activated in the early
reperfusion period, accounting for sustained APD shortening in this
phase of reperfusion.
Effects of Glibenclamide on Contraction
Application of
glibenclamide attenuated the shortening of APD not
only during ischemia but also in the early phase of
reperfusion. Therefore, this drug is expected to indirectly affect the
contractile function in addition to APD shortening during
ischemia/reperfusion. Fig 3 shows
representative recordings of contractile
tension before, during, and after 10 minutes of no-flow
ischemia with or without glibenclamide. Percent changes of
dTension are shown in Fig 4. In untreated preparations,
dTension declined rapidly after the initiation of ischemia and
was lost within 10 minutes; this occurred without significant changes
in rTension (Fig 3A). After reperfusion, dTension gradually
recovered,
reaching 92.0±6.4% of preischemic values in 60 minutes
(Figs 3A and 4). rTension was slightly but
significantly elevated on
reperfusion and eventually returned to the preischemic
level (Table 1).
|
|
Pretreatment with glibenclamide (10
µmol/L) caused a significant
elevation of rTension during ischemia that persisted for 60
minutes of reperfusion (Fig 3B and Table 1).
dTension was restored
quickly on reperfusion and reached a peak value within 7 to 8 minutes
of reperfusion, although this dTension decreased again later in
reperfusion (Fig 4). The net recovery of dTension estimated at
30 and
60 minutes of reperfusion was significantly lower in the glibenclamide
pretreatment preparations than in untreated preparations (Table
1).
In contrast, when drug was applied from the start of
reperfusion, ie,
posttreatment, a rapid and transient increase of dTension appeared on
reperfusion (Figs 3C and 4) and was followed by
a subsequent decrease,
identical to the findings in the pretreatment group. The dTension
measured at 10, 30, and 60 minutes of reperfusion was significantly
lower than in untreated preparations (Table 1). The rTension of
posttreated preparations, measured at 30 and 60 minutes after
reperfusion, was significantly greater than that in untreated
preparations (Table 1).
To further examine the effect of rapid prolongation of APD on the recovery process of contraction, we applied a new class III antiarrhythmic agent, E-4031,15 instead of glibenclamide. E-4031 (1 µmol/L) significantly prolonged the APD (from 156±2 to 180±3 milliseconds) without affecting contractility under control (preischemic) perfusion. Application of this agent starting from the onset of reperfusion that followed 10 minutes of ischemia significantly suppressed the recovery of developed tension measured at 60 minutes after reperfusion (72.6±1.8% of preischemic values in treated preparation (n=5) versus 92.0±6.4% in untreated preparation (n=7); P<.01). This effect was associated with a faster recovery of the APD after reperfusion.
Effects of Glibenclamide on Reperfusion
Arrhythmias
There was no evidence of severe arrhythmias before or
during ischemia in the presence or absence of glibenclamide. In
untreated preparations, during reperfusion, we observed sporadic VPCs
in four of seven preparations and short runs of VT (persisting for less
than 3 seconds) in one preparation. In the presence of glibenclamide
(in both pretreatment and posttreatment groups), VPCs were observed in
all preparations tested (n=5 for both), and short runs of VT developed
in three of five pretreatment preparations and in two of five
posttreatment preparations. Sustained VT or Vf was not seen in either
group. Thus, glibenclamide increased the incidence of VPCs and short
runs of VT in the reperfusion phase but did not produce prolonged VT or
Vf.
Effects of Cromakalim on Electrical and Mechanical Activities
During Ischemia and Reperfusion
In the second series of experiments,
we examined the effect of
cromakalim, a K+ATP channel opener, by use of a
protocol entailing 20 minutes of no-flow ischemia that was
followed by 60 minutes of reperfusion. We used a 20-minute
ischemic period (instead of the 10-minute ischemia used
with glibenclamide) because this ischemic period produced a
sustained contractile depression after 60 minutes of reperfusion.
The
alterations of electrical and contractile parameters
during 20 minutes of no-flow ischemia and 60 minutes of
reperfusion in the presence or absence of cromakalim are summarized in
Table 2, and the representative changes
in action potential configuration and the time course of changes in
APD90 are shown in Fig 5. In untreated
preparations ischemia markedly shortened the
APD90, and in two of four preparations the
electrical excitability was completely lost in 17 minutes. In these
preparations, action potentials could not be elicited even when the
intensity of stimulation was increased 10-fold. This electrical
inexcitability was probably not the result of a decrease in the RMP,
ie, by means of inactivation of the inward sodium current, because the
RMP, at -69.5±0.2 mV, remained more negative than -60 mV at
the end
of 20 minutes of ischemia (n=4). Consequently, this electrical
inexcitability may be due to a large increase in the outward current
mediated by K+ATP channels. The amplitude of
the action potential declined progressively during the ischemic
period. The ischemia significantly depolarized the RMP (Table
2). On reperfusion, all electrical parameters
(APD90, APA, and RMP) returned essentially to
preischemic levels within 60 minutes; however, contractile
parameters (rTension and dTension) were not fully restored.
The contractile tension remained depressed (
34% of
preischemic contraction) even at 60 minutes of reperfusion,
and resting tension remained elevated (Table 2). These impaired
contractile parameters did not return to normal even when
the reperfusion period was extended to 3 hours (n=2), indicating that
20 minutes of no-flow ischemia produced myocardial
stunning, a sustained postischemic depression of
contractile function.
|
|
, n=5) or without (
Under control (preischemic) conditions,
cromakalim (2
µmol/L) slightly but significantly shortened the APD90
(by 8.8±0.7%) with no apparent effects on APA, RMP, or contractile
parameters (Table 2). During ischemia and
reperfusion, however, this agent markedly modified the action
potentials and contractility (Figs 5 and 6,
Table 2).
Cromakalim shortened the APD90 more quickly during
ischemia and led to electrical inexcitability in all
preparations tested (n=5) at a much earlier time (15 minutes) than
in untreated preparations (Fig 5). The shortening of
APD90
in the early reperfusion phase was greater and was sustained much
longer in the presence of cromakalim. However, changes in RMP during
ischemia were not altered by cromakalim, and the recovery of
RMP during reperfusion was facilitated (Table 2). The dTension
in
cromakalim-treated preparations tended to decline slightly faster
during ischemia, and exhibited much greater recovery after
reperfusion than in untreated preparations (Fig 6). At
60 minutes after reperfusion, dTension was restored to 79.3±4.1% of
the preischemic values in the cromakalim-treated
preparations versus only 34.0±3.2% in untreated preparations
(P<.05). The elevation of rTension during ischemia
was not affected by the presence of cromakalim (Table 2),
although the
reperfusion-induced increase of rTension was significantly
suppressed in the treated preparations.
|
We examined the effect of
cromakalim at higher (5 µmol/L) or lower (1
µmol/L) concentrations. Cromakalim at 5 µmol/L also improved
contractile functions during reperfusion; however, this concentration
exerted a significant negative inotropic effect (by 30%) in
preischemic conditions (data not shown). In contrast, 1
µmol/L cromakalim hardly improved the recovery of contraction during
reperfusion (n=3, data not shown).
Because the application of cromakalim before ischemia remarkably improved the contractile recovery, we examined the effect of cromakalim applied from the onset of reperfusion. The application of cromakalim (2 µmol/L) enhanced the shortening of APD during reperfusion compared with untreated preparations; ie, the APD90 was 112.0±2.4 milliseconds at 10 minutes and 136.0±1.8 milliseconds at 30 minutes of reperfusion (n=5, P<.05 versus preparations). In contrast, the dTension measured at 60 minutes of reperfusion was significantly greater (48.7±3.2% of preischemic values, n=7; P<.05) than that in untreated preparations (34.0±3.2% of preischemic values, n=4).
Effects of Cromakalim on Reperfusion Arrhythmias and
Modification by Glibenclamide
In untreated preparations, reperfusion
of ventricular
muscles after 20 minutes of no-flow ischemia resulted in
arrhythmias more frequent and more severe than those seen with
reperfusion after 10 minutes of ischemia. VPCs and/or short
runs of VT were observed in all preparations tested (n=6). Sustained VT
(with a duration of 108 seconds) and Vf (not spontaneously
defibrillated) were each observed in one preparation. Treatment with
cromakalim (2 µmol/L) before ischemia did not increase the
number of VPCs or short runs of VT on reperfusion; however, sustained
VT (with an average duration of 53 seconds) occurred in two (20%) and
Vf in four (40%) of the preparations tested (n=10). Application of
cromakalim (2 µmol/L) from the beginning of reperfusion also
increased the incidence of VT (29%) and Vf (57%); however, these
arrhythmias disappeared spontaneously within 3 minutes in all
cases (n=7). These findings suggested that the presence of cromakalim
tends to increase the risk of VT and Vf.
Fig 7 shows a
typical example of reperfusion-induced
arrhythmias observed in the presence of a relatively high
concentration of cromakalim (5 µmol/L). Reperfusion after 20 minutes
of no-flow ischemia immediately produced sustained VT (Fig 7E),
which deteriorated to Vf (Fig 7F). Because Vf persisted for
more
than 5 minutes, glibenclamide (10 µmol/L) was applied in the
continued presence of cromakalim to examine the contribution of
K+ATP current to this arrhythmia. With
the introduction of glibenclamide, Vf terminated; this termination was
preceded by a rapid prolongation of APD (Fig 7G). The same
effect of
glibenclamide was seen in two of three cases of Vf after treatment with
2 µmol/L cromakalim. In our experience, Vf with a duration of more
than 3 minutes never terminated spontaneously in the absence of
glibenclamide. Glibenclamide might have terminated Vf by lengthening
the APD or the refractory period by blockade of
K+ATP channels. It must be noted here that the
recovery of contraction in cromakalim-treated tissues was fairly
good despite the development of Vf (Fig 7H). This finding
implies that
the development of Vf did not result from the
ischemia/reperfusion-induced calcium overload (which
usually leads to triggered activity arising from delayed
afterdepolarization).
|
|
Discussion |
|---|
|
TopAbstractIntroductionMethods Results Discussion References |
|---|
Recently the cardioprotective effect of K+ATP channel activation during ischemia has attracted much attention. Numerous attempts have been made to clarify the mechanisms of cardioprotection of this channel. However, there has been no report describing the role of this channel after reperfusion. We first demonstrated the possibility that the activation of K+ATP channels elicited by preceding ischemia persists for a fairly long time (
30
minutes) after reperfusion and contributes to the contractile recovery
during this period.
The shortening of the APD produced by no-flow ischemia
persists during the early phase of reperfusion, and this APD shortening
is preventable by glibenclamide applied from the onset of reperfusion.
Therefore, the persistent APD shortening was mostly attributed to the
residual activation of K+ATP channels. This
persistent APD shortening, although relatively small compared with the
APD shortening observed during ischemia, is nonetheless
important for the recovery of contractile function and the evolution of
arrhythmias during reperfusion. In theory, shorter APDs would
lead to a reduction in the time for Ca2+ influx via
voltage-gated Ca2+ channels and to an increase
in the time during which the Na+-Ca2+
exchanger may operate to extrude
Ca2+.16 17 Resultant decreases in
transsarcolemmal Ca2+ influx would help maintain
[Ca2+]i at
physiological levels when other
Ca2+ extrusion mechanisms such as the
Ca2+ pump are impaired because of a reduction of
[ATP]i.18 Accordingly, in the very early
reperfusion phase when the Ca2+ extrusion mechanisms
have not been fully restored, the application of glibenclamide, started
even at the onset of reperfusion, might increase the
[Ca2+]i compared with that seen in the
absence of glibenclamide. The rise in
[Ca2+]i and the concomitant increase
in ATP consumption (due to enhanced contraction) might have inhibited
the subsequent recovery of contractility (Fig 4). This
notion is supported by the finding that glibenclamide produced a rapid
but only transient increase in contractions in the very early phase of
reperfusion, which led to a secondary sustained decline of
contractility later during the reperfusion phase (Figs 3 and
4). In preliminary experiments, we applied glibenclamide (10
µmol/L) at 15 minutes after reperfusion that followed 10 minutes of
no-flow ischemia. On application, the action potential
quickly prolonged, and this was associated with an increase in
contraction. However, this increase persisted only for less than 2
minutes and was followed by a severe decline of the contraction in the
later phase of reperfusion.19 We propose that the rapid
prolongation of APD caused by glibenclamide applied during the
reperfusion period increases [Ca2+]i
and produces a transient increase in contraction. This is, however,
followed by a subsequent severe decline in
contractility, probably because of the occurrence of a
calcium overload.
A similar reperfusion-induced transient increase in contraction (followed by a secondary decline) has been noted in pigs in vivo and is termed postischemic hypercontraction.20 The mechanism of this phenomenon has been attributed to a rapid and excessive increase of [Ca2+]i that occurred in the early phase of reperfusion. These findings lend support to the notion that most of the calcium overload develops during reperfusion rather than during ischemia, resulting in depressed contractile recovery associated with an elevation of resting tension. Cascio et al21 have shown that electrical cell-to-cell uncoupling occurs simultaneously with an increase in rTension. This indicates that Ca2+-induced electrical uncoupling is also a consequence of application of glibenclamide.
On the other hand, short APDs that are caused by activation of
K+ATP channels may decrease the contraction of
the heart muscle and consequently help preserve [ATP]i by
reducing energy consumption during ischemia.1 22
However, it is unlikely that inhibition of the decline of contractions
during ischemia is a prime cause of the deleterious effect of
glibenclamide on the contractile recovery seen in the late reperfusion
period. This is because glibenclamide impaired contractile recovery,
even when it was applied from the onset of reperfusion (Fig 4).
Endogenous activation of K+ATP channels during ischemia has recently been reported to improve contractile recovery; however, this cardioprotective effect did not correlate with the preservation of high-energy phosphate.23 Moreover, verapamil, a Ca2+ antagonist, produced dose-dependent protection against ischemic injury in the globally ischemic rat heart. However, this effect could not be entirely attributed to the preservation of [ATP]i and creatine phosphate.24 These findings are consistent with our speculation that the major protective effect of K+ATP channels (activated by ischemia) is inhibition of the development of calcium overload (by shortening of APD).
To support our hypothesis, we examined the effect of E-4031 on the recovery of contractions after ischemia. The application of E-4031 starting from the onset of reperfusion significantly suppressed the recovery of contractility, which was associated with longer APD compared with untreated preparations. These results are in line with our hypothesis that the rapid APD prolongation in the early reperfusion phase is deleterious to the recovery of contraction, even when the APD prolongation is made by means of blockade of K+ channels (eg, delayed-rectifier K+ channels) other than K+ATP channels.
Glibenclamide is reported to inhibit K+ATP channels by binding to a specific site on the channel protein.25 However, these sulfonylurea drugs also affect cardiac metabolism. Glibenclamide stimulated glycolytic ATP synthesis without changes in oxygen consumption and did not elicit a positive inotropic effect in rat hearts.26 Impaired postischemic recovery of contractility found in glibenclamide-treated preparations may not be attributed to this metabolic effect, because an enhancement of ATP production is more likely to improve (and not impair) the recovery of contractility.
K+ATP channels are also involved in the
regulation of coronary blood flow,27 as
demonstrated by the prevention by glibenclamide of
hypoxia-induced coronary vasodilatation in isolated
guinea pig hearts; moreover, these channels regulate the basal
coronary flow in anesthetized dogs.28
However, this effect of glibenclamide cannot account for the impairment
of postischemic contractile recovery, because the rate of
coronary flow was constant in our experiments. Furthermore, the
decrease of coronary flow by glibenclamide cannot explain the
transient increase of contraction seen immediately after introduction
of this drug (Fig 3C).
Cromakalim improved mechanical recovery and ameliorated the
development of myocardial stunning. It has been postulated that some
K+ATP channel openers prevent irreversible cell
injury during
ischemia/reperfusion.7 29 30
Cromakalim increases coronary flow by means of its vasodilating
effect.31 However, in our study the rate of perfusion was
constant throughout the experiments. The protective effects of
cromakalim could be attributable, at least in part, to a reduction of
energy consumption due to the decline in
contractility.32 Such an
energy-sparing effect may facilitate contractile recovery during
reperfusion. However, in our experiments this mechanism did not appear
to play a prime role, because the rate of decline in contractions seen
during ischemia in the presence of cromakalim did not differ
statistically from that found in the absence of this drug (Table
2),
although it tended to occur slightly faster (Fig 6).
Accordingly, we
attribute the protective effect of cromakalim mainly to the attenuation
of the increase in [Ca2+]i during
ischemia and reperfusion. This hypothesis is supported by a
finding that the rise in rTension after reperfusion was significantly
less in the presence of cromakalim compared with untreated preparations
(Table 2). The reduction of rTension may reflect a decrease in
the
resting level of [Ca2+]i.
It should be mentioned that the recovery of RMP was significantly
faster and greater in the cromakalim-treated group (Table 2).
This
faster and greater recovery of RMP is probably mediated by the increase
of outward current via cromakalim-activated
K+ATP channels and seems to be beneficial for
the contractile recovery. The extrusion of cytosolic calcium via the
(electrogenic) Na+-Ca2+ exchanger may be
facilitated in such deeper RMPs, thereby attenuating the development of
calcium overload.
Cromakalim improved contractile recovery even when it was applied from the onset of reperfusion. This improvement was associated with an enhanced shortening of the APD during the reperfusion period. However, the degree of improvement in contractility was considerably lesser than that seen with pretreatment. This finding implies that the cell injury developing during prolonged ischemia (20 minutes) was attenuated by cromakalim in pretreated preparations.
Cole et al7 have reported that glibenclamide (10 µmol/L) increased the incidence of reperfusion-induced triggered activity, and sustained VT, originating from delayed afterdepolarization. In our observation this agent increased the incidence of VPCs and short runs of VT in the reperfusion phase but did not produce sustained VT or Vf. Differences between results obtained with similar experimental methods may be due to differences in the stimulating frequencies and ischemic periods that were applied. Cole et al examined the effects of glibenclamide by using 20 minutes of no-flow ischemia with a stimulation frequency of 2 Hz, whereas we used only 10 minutes of ischemia and a stimulation frequency of 3 Hz. Therefore, their ischemic insult might have been greater than ours.
Alternatively, the application of cromakalim increased the incidence of
Vf during reperfusion in our study. We could not find any difference in
action potential parameters (APD, RMP, and APA) between the
groups that developed or did not develop Vf. However, because the Vf
that developed in cromakalim-treated preparations was readily
terminated by the application of glibenclamide, we speculate that
enhanced activation of K+ATP channels by
cromakalim provides a substrate for reentry, ie, shortening of the
refractory period. The arrhythmogenic actions of
K+ATP channel openers in acute ischemia
have been reported by other
investigators.33 34 35 Cole et
al7 have claimed that application of pinacidil decreases
the incidence of reperfusion arrhythmias, the opposite of our
findings. The difference between the results may, again, arise from
different lengths of the ischemic periods: they used 30 minutes
of ischemia whereas we used 20 minutes. Prolonged
ischemia results in much greater increases in
[Ca2+]i and produces triggered
arrhythmias arising from delayed afterdepolarization.
Shortening of APD by a K+ATP channel opener
might diminish this type of arrhythmia by decreasing
[Ca2+]i. Alternatively, the
tachyarrhythmias we observed after 20 minutes of
ischemia (Fig 7E through 7G) might have been provoked by a
reentry mechanism; these arrhythmias were readily terminated by
glibenclamide, which prolonged the APD and hence the effective
refractory period.
Comparison of the effects of K+ATP channel blockers and openers may shed some light on the problem of a straightforward antiarrhythmic strategy in reperfusion arrhythmias. This is because both potentiation of action potential shortening (reentrant arrhythmia) and action potential lengthening (increase in [Ca2+]i and resulting delayed afterdepolarizations) may precipitate arrhythmias at a critical time of reperfusion. Furthermore, the effects of both interventions on electrical cell-to-cell uncoupling remain to be determined.
Brief episodes of ischemia have been reported to increase the tolerance of the heart muscle to a subsequent severe ischemic insult; this effect has been termed ischemic preconditioning.9 In addition, Gross and Auchampach11 have reported that glibenclamide prevented this preconditioning effect in dogs. Furthermore, the ischemic preconditioning effect was lost when the interval between the foregoing short-term ischemia and the subsequent long and severe ischemia was prolonged to more than 1 hour.10 The mechanism of this short-term memory effect in preconditioning is uncertain. In the present study we observed that a substantial number of K+ATP channels remained activated over the early reperfusion phase. From these observations, it is tempting to speculate that persistent activation of K+ATP channels in early reperfusion is responsible for the memory effect of ischemic preconditioning. Under such circumstances, the activation threshold of the K+ATP channels to subsequent ischemia may be lowered such that these channels could be more readily activated by the subsequent ischemia, resulting in better tolerance of the myocardium to this second ischemic insult.
|
Selected Abbreviations and Acronyms |
|---|
|
|
Acknowledgments |
|---|
The authors gratefully acknowledge Dr A.G. Kléber, Department of Physiology, University of Bern, Switzerland, for his critical readings and valuable suggestions on the manuscript and M. Ohara for comments on the manuscript. We thank K. Moriyama for secretarial assistance.
Received October 6, 1994; revision received March 9, 1995; accepted May 4, 1995.
|
References |
|---|
|
TopAbstractIntroductionMethods Results Discussion References |
|---|
1. Noma A. ATP-regulated K+ channels in cardiac muscle. Nature. 1983;305:147-148. [Medline] [Order article via Infotrieve]
2. Nakamura S, Kiyosue T, Arita M. Glucose reverses 2,4-dinitrophenol induced changes in action potentials and membrane currents of guinea pig ventricular cells via enhanced glycolysis. Cardiovasc Res. 1989;23:286-294. [Medline] [Order article via Infotrieve]
3.
Deutsch N, Klitzner TS, Lamp ST, Weiss JN.
Activation of cardiac ATP-sensitive K+ current
during hypoxia: correlation with tissue ATP levels.
Am J Physiol. 1991;261:H671-H676.
4.
Shen WK, Tung RT, Machulda MM, Kurachi Y.
Essential role of nucleotide diphosphates in
nicorandil-mediated activation of cardiac ATP-sensitive
K+ channel: a comparison with pinacidil and
lemakalim. Circ Res. 1991;69:1152-1158.
5.
Koyano T, Kakei M, Nakashima H, Yoshinaga M, Matsuoka
T, Tanaka H. ATP-regulated K+ channels are modulated
by intracellular H+ in guinea-pig
ventricular cells. J Physiol
(Lond). 1993;463:747-766.
6.
Lederer WJ, Nichols CG. Nucleotide
modulation of the activity of rat heart ATP-sensitive K+
channels in isolated membrane patches. J
Physiol (Lond). 1989;419:193-211.
7.
Cole WC, McPherson CD, Sontag D. ATP-regulated
K+ channels protect the myocardium against
ischemia/reperfusion damage. Circ
Res. 1991;69:571-581.
8.
Auchampach JA, Maruyama M, Cavero I, Gross GJ.
Pharmacological evidence for a role of ATP-dependent potassium
channels in myocardial stunning.
Circulation. 1992;86:311-319.
9.
Murry CE, Jennings RB, Reimer KA.
Preconditioning with ischemia: a delay of lethal cell
injury in ischemic myocardium.
Circulation. 1986;74:1124-1136.
10.
Murry CE, Richard VJ, Jennings RB, Reimer KA.
Myocardial protection is lost before contractile function
recovers from ischemic preconditioning. Am J
Physiol. 1991;260:H796-H804.
11.
Gross GJ, Auchampach JA. Blockade of
ATP-sensitive potassium channels prevents myocardial preconditioning in
dogs. Circ Res. 1992;70:223-233.
12. Cole WC, Picone JB, Sperelakis N. Gap junction uncoupling and discontinuous propagation in the heart: a comparison of experimental data with computer simulations. Biophys J. 1988;53:809-818. [Medline] [Order article via Infotrieve]
13.
Kléber AG. Resting membrane potential,
extracellular potassium activity and intracellular sodium activity
during acute global ischemia in isolated perfused guinea pig
hearts. Circ Res. 1983;52:442-450.
14. Penny WJ, Sheridan DJ. Arrhythmias and cellular electrophysiological changes during myocardial "ischaemia" and reperfusion. Cardiovasc Res. 1983;17:363-372. [Medline] [Order article via Infotrieve]
15. Wettwer E, Scholtysik G, Schaad A, Himmel H, Ravens U. Effect of the new class III antiarrhythmic drug E-4031 on myocardial contractility and electrophysiological parameters. J Cardiovasc Pharmacol. 1991;17:480-487. [Medline] [Order article via Infotrieve]
16.
Kimura J, Miyamae S, Noma A. Identification of
sodium-calcium exchange current in single ventricular
cells of guinea-pig. J Physiol (Lond). 1987;384:199-222.
17.
Sheu S, Sharma VK, Uglesity A.
Na+-Ca2+ exchange contributes to
increase of cytosolic Ca2+ concentration during
depolarization in heart muscle. Am J Physiol. 1986;250:C651-C656.
18.
Krause SM, Gess ML. Characterization of cardiac
sarcoplasmic reticulum dysfunction during short-term
normothermic global ischemia.
Circ Res. 1985;55:176-184.
19. Shigematsu S, Toshimitsu T, Sato T, Arita M, Saikawa T, Sakata T. Persistent activation of ATP-sensitive K+ channel underlies the occurrence of myocardial stunning. Circulation. 1993;88(suppl I):I-131. Abstract.
20.
Aksnes G, Kirkebøen KA, Christensen G, Ilebekk A.
Characteristics and development of myocardial stunning in the
pig. Am J Physiol. 1992;263:H544-H551.
21.
Cascio WE, Yan GX, Kléber AG. Passive
electrical properties, mechanical activity, and extracellular potassium
in arterially perfused and ischemic rabbit
ventricular muscle. Circ Res. 1990;66:1461-1473.
22.
Lederer WJ, Nichols CG, Smith GL. The
mechanism of early contractile failure of isolated rat
ventricular myocytes subjected to complete
metabolic inhibition. J Physiol
(Lond). 1989;413:329-349.
23.
McPherson CD, Pierce GN, Cole WC.
Ischemic cardioprotection by ATP-sensitive
K+ channels involves high-energy phosphate
preservation. Am J Physiol. 1993;265:H1809-H1818.
24. Watts JA, Maiorano L, Maiorano PC. Protection by verapamil of globally ischemic rat hearts: energy preservation, a partial explanation. J Mol Cell Cardiol. 1985;17:797-804. [Medline] [Order article via Infotrieve]
25.
Fosset M, De Weille JR, Green RD, Schmid-Antomarchi H,
Lazdunski M. Antidiabetic sulfonylureas control action potential
properties in heart cells via high affinity receptors that are linked
to ATP-dependent K+ channels. J
Biol Chem. 1988;263:7933-7936.
26. Schaffer SW, Tan BH, Mozaffari MS. Effect of glyburide on myocardial metabolism and function. Am J Med. 1985;79(suppl):48-52.
27.
Daut JW, Maier-Rudolph W, Von Beckerath N, Meherke G,
Günther K, Goedel-Meinen L. Hypoxic dilation of
coronary arteries is mediated by ATP-sensitive potassium
channels. Science. 1990;247:1341-1344.
28.
Imamura Y, Tomoike H, Narishige T, Takahashi T, Kasuya
H, Takeshita A. Glibenclamide decreases basal coronary
blood flow in anesthetized dogs. Am J
Physiol. 1992;263:H399-H404.
29. Kempsford RD, Hawgood BJ. Assessment of the antiarrhythmic activity of nicorandil during myocardial ischemia and reperfusion. Eur J Pharmacol. 1989;163:61-68. [Medline] [Order article via Infotrieve]
30.
McCullough JR, Normandin DE, Conder ML, Sleph PG,
Dzwonczyk S, Grover GJ. Specific block of the
anti-ischemic actions of cromakalim by
sodium-5-hydroxy-decanoate. Circ Res. 1991;69:949-958.
31. Cook NS, Quast U. Potassium channels: structure, classification, function and therapeutic potential. In: Cook NS, ed. Potassium Channel Pharmacology. Chester, UK: Ellis Horwood; 1990:181-255.
32.
Nichols CG, Ripoll C, Lederer WJ. ATP-sensitive
potassium channel modulation of the guinea pig ventricular
action potential and contraction. Circ
Res. 1991;68:280-287.
33. Wollenben CD, Sanguinetti MC, Siegl PKS. Influence of ATP-sensitive potassium channel modulators on ischemia-induced fibrillation in isolated rat hearts. J Mol Cell Cardiol. 1989;21:783-788. [Medline] [Order article via Infotrieve]
34.
D'Alonzo AJ, Darbenzio RB, Parham CS, Grover GJ.
Effects of intracoronary cromakalim on postischaemic
contractile function and action potential duration.
Cardiovasc Res. 1992;26:1046-1053.
35. Grover GJ, Sleph PG, Dzwonczyk S. Pharmacologic profile of cromakalim in the treatment of myocardial ischemia in isolated rat hearts and anesthetized dogs. J Cardiovasc Pharmacol. 1990;16:853-864.[Medline] [Order article via Infotrieve]
This article has been cited by other articles:
 |
|
 |
|
  T. Sato, A. D. T. Costa, T. Saito, T. Ogura, H. Ishida, K. D. Garlid, and H. Nakaya Bepridil, an Antiarrhythmic Drug, Opens Mitochondrial KATP Channels, Blocks Sarcolemmal KATP Channels, and Confers Cardioprotection J. Pharmacol. Exp. Ther., January 1, 2006; 316(1): 182 - 188. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. o. Cosar and C. J. O'Connor Hibernation, Stunning, and Preconditioning: Historical Perspective, Current Concepts, Clinical Applications, and Future Implications Seminars in Cardiothoracic and Vascular Anesthesia, June 1, 2003; 7(2): 115 - 140. [Abstract] [PDF]
|
|
|
|
 |
|
 H. F del Valle, E. C Lascano, and J. A Negroni Ischemic preconditioning protection against stunning in conscious diabetic sheep: role of glucose, insulin, sarcolemmal and mitochondrial KATP channels Cardiovasc Res, August 15, 2002; 55(3): 642 - 659. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. F. del Valle, E. C. Lascano, J. A. Negroni, and A. J. Crottogini Glibenclamide effects on reperfusion-induced malignant arrhythmias and left ventricular mechanical recovery from stunning in conscious sheep Cardiovasc Res, June 1, 2001; 50(3): 474 - 485. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K.-i. Ito, T. Sato, and M. Arita Protein kinase C isoform-dependent modulation of ATP-sensitive K+ channels during reoxygenation in guinea-pig ventricular myocytes J. Physiol., April 1, 2001; 532(1): 165 - 174. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 W. Tang, M. H. Weil, S. Sun, A. Pernat, and E. Mason KATP channel activation reduces the severity of postresuscitation myocardial dysfunction Am J Physiol Heart Circ Physiol, October 1, 2000; 279(4): H1609 - H1615. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Z. Papp, J. van der Velden, and G.J.M Stienen Calpain-I induced alterations in the cytoskeletal structure and impaired mechanical properties of single myocytes of rat heart Cardiovasc Res, March 1, 2000; 45(4): 981 - 993. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y. Li, T. Sato, and M. Arita Bepridil Blunts the Shortening of Action Potential Duration Caused by Metabolic Inhibition via Blockade of ATP-Sensitive K+ Channels and Na+-Activated K+ Channels J. Pharmacol. Exp. Ther., November 1, 1999; 291(2): 562 - 568. [Abstract] [Full Text]
|
|
|
|
 |
|
 L. P. Perrault and P. Menasche Preconditioning: can nature’s shield be raised against surgical ischemic-reperfusion injury? Ann. Thorac. Surg., November 1, 1999; 68(5): 1988 - 1994. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. R.C. Dekker, R. Coronel, E. VanBavel, J. A.E. Spaan, and T. Opthof Intracellular Ca2+ and delay of ischemia-induced electrical uncoupling in preconditioned rabbit ventricular myocardium Cardiovasc Res, October 1, 1999; 44(1): 101 - 112. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
V. H. Thourani, M. Nakamura, R. S. Ronson, J. E. Jordan, Z.-Q. Zhao, J. H. Levy, F. Szlam, R. A. Guyton, and J. Vinten-Johansen Adenosine A3-receptor stimulation attenuates postischemic dysfunction through KATP channels Am J Physiol Heart Circ Physiol, July 1, 1999; 277(1): H228 - H235. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. B. Hoag, Y.-Z. Qian, M. A. Nayeem, M. D'Angelo, and R. C. Kukreja ATP-sensitive potassium channel mediates delayed ischemic protection by heat stress in rabbit heart Am J Physiol Heart Circ Physiol, November 1, 1997; 273(5): H2458 - H2464. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Shigematsu and M. Arita Anoxia-induced activation of ATP-sensitive K+ channels in guinea pig ventricular cells and its modulation by glycolysis Cardiovasc Res, August 1, 1997; 35(2): 273 - 282. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. Menasche, C. Mouas, and C. Grousset Is Potassium Channel Opening an Effective Form of Preconditioning Before Cardioplegia? Ann. Thorac. Surg., June 1, 1996; 61(6): 1764 - 1768. [Abstract] [Full Text]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||