Effects of triiodothyronine supplementation after myocardial ischemia

Effects of Triiodothyronine Supplementation After Myocardial Ischemia Cornelius M. Dyke, MD, Mai Ding, MD, Anwar S. Abd-Elfattah, PhD, Kathy Loesser, PhD, Rebecca J. Dignan, MD, Andrew S. Wechsler, MD, and David R. Salter, MD Department of Surgery, Medical College of Virginia, Virginia Commonwealth University, Richmond, Virginia

Cardiopulmonary bypass causes a "euthyroid-sick" state characterized by low levels of circulating triiodothyronine. Triiodothyronine supplementation in this setting has been postulated to improve postischemic left ventricular function by increasing the availability of myocardial high-energy phosphates. These postulates have not been substantiated, however, using load-independent parameters of left ventricular function and analysis of highenergy phosphate metabolism. To test these hypotheses, 14 healthy pigs (30 to 40 kg) were placed on cardiopulmonary bypass and instrumented with left ventricular minor-axis ultrasonic crystals and micromanometertipped pressure catheters. Hearts were subjected to 30 minutes of global, normothermic ischemia. Triiodothyronine (0.1 mg/kg; n = 7) or placebo (n = 7) was administered in a random, investigator-blinded fashion at the removal of the aortic cross-clamp and after 60

minutes of reperfusion. Hemodynamic, metabolic, and ultrastructural data were obtained before ischemia and after 30, 60, 90, and 120 minutes of reperfusion. By 90 minutes of reperfusion left ventricular contractility had returned to preischemic levels in hearts supplemented with triiodothyronine, despite postischemic myocardial adenosine triphosphate levels of 50% to 60% of baseline in both groups. Ultrastructurally, the sarcoplasmic reticulum and mitochondria were significantly better preserved in the group treated with triiodothyronine. This study suggests that triiodothyronine supplementation significantly enhances postischemic left ventricular functional recovery and that this recovery is due to mechanisms other than enhanced availability of myocardial high-energy phosphates.

D

myocardial dysfunction and that T, supplementation in this setting improves ventricular systolic performance in the immediate postoperative period. However, studies describing the effects of acute T, supplementation on cardiac function have been limited by the use of highly load-dependent indices of contractility, making assessment of ventricular function difficult. Using an isolated rabbit heart model, we have previously demonstrated that T, has no intrinsic inotropic properties when given in either physiologic or pharmacologic doses to normal hearts [171. However, T3 supplementation significantly enhanced left ventricular functional recovery after ischemic injury [17]. The mechanisms by which acute T, supplementation enhances postischemic function are not clear. Novitsky and associates [15] suggested that T, supplementation increases the aerobic oxidative capacity of the myocardium, thereby leading to an increased availability of high-energy phosphates for contraction. The purpose of the present study was to evaluate left ventricular functional recovery after ischemia in an in vivo, large animal model using a load-independent index of left ventricular function. Additionally, we wished to test the hypothesis that T, supplementation enhances postischemic ventricular function by increasing the availability of myocardial high-energy phosphates available for contraction.

espite an operative mortality of less than 2% for elective coronary artery bypass grafting, postoperative left ventricular dysfunction remains problematic [11. Transient decreases in left ventricular function in the immediate postoperative period have been reported in patients with normal preoperative ventricular ejection fractions [2-61, as well as in patients with impaired ventricular function. The cause of this transient, postoperative left ventricular mechanical derangement is usually attributed to either inadequate intraoperative myocardial protection or reperfusion injury. Alternatively, however, it has been suggested that alterations in thyroid hormone metabolism may contribute to cardiac dysfunction after open heart operations. Cardiopulmonary bypass leads to a "euthyroid-sick" state, defined by depressed levels of free, unbound triiodothyronine (T,) and elevated levels of reverse T3, with normal or slightly elevated circulating free thyroxine levels [7-121. Novitzky and associates [13-161 have suggested that this transient depression in free T, levels after cardiopulmonary bypass contributes to postoperative Presented at the Twenty-ninth Annual Meeting of The Society of Thoracic Surgeons, San Antonio, TX, Jan 2527, 1993. Address reprint requests to Dr Dyke, Medical College of Virginia, Virginia Commonwealth University, Box 645MCV Station, Richmond, VA 23298.

0 1993 by The Society of Thoracic Surgeons

(Ann Thorac Surg 1993;56:215-22)

0003-4975/93/$6.00

216

DYKEETAL T, SUPPLEMENTATION AFTER ISCHEMIA

Material and Methods lnstrumenta tion Fourteen healthy pigs weighing between 30 and 40 kg were studied. All animals received humane care in compliance with the ”Principles of Laboratory Animal Care” formulated by the National Society for Medical Research and the “Guide for the Care and Use of Laboratory Animals” prepared by the National Academy of Sciences and published by the National Institutes of Health (NIH publication 85-23, revised 1985). Before instrumentation, pigs were sedated with an intramuscular injection of atropine sulfate (400 pg), acepromazine maleate (50 mg), and ketamine HCl (500 mg), and then anesthetized with pentobarbital (13 mg/kg intravenously). Subsequent doses of pentobarbital were administered every 1 to 2 hours as needed. Tracheostomy was performed to secure an airway, and animals were ventilated with a large animal respirator. The external jugular vein was isolated and cannulated for administration of intravenous fluids. Lidocaine HC1 (1 mg/kg intravenously) and bretylium tosylate (8 mg/kg intravenously) were administered before median sternotomy for prophylaxis against ventricular arrhythmias. The left femoral artery was isolated, and a micromanometer-tipped catheter (Millar Instruments, Houston, TX) was introduced and positioned in the thoracic aorta for continuous monitoring of systemic blood pressure. After median sternotomy, the right and left hemiazygos veins were ligated and the heart was suspended in a pericardial cradle. The right and left phrenic nerves were sectioned to eliminate diaphragmatic contraction. Left ventricular pressure was monitored with an intracavitary micromanometer-tipped catheter introduced through the left ventricular apex. A pair of hemispherical, ultrasonic, piezoelectric crystals was positioned across the left ventricular anteroposterior minor axis to record minor-axis dimension throughout the cardiac cycle. After systemic heparinization, cardiopulmonary bypass was initiated using the right common carotid artery for arterial cannulation and a two-stage venous cannula introduced through the right atrial appendage for venous drainage. A membrane oxygenator primed with non-cross-matched porcine blood was used. Arterial blood gases were monitored before and during cardiopulmonary bypass. A mean perfusion pressure of 60 mm Hg was maintained throughout the experiment.

Experimental Protocol 3,5,3‘-triiodo-~~-thyronine (Sigma Chemical Co, St. Louis, MO) was solubilized in a 0.9% NaC1, 0.1 moVL NaOH vehicle solution for infusion. Pigs were randomized into two groups, receiving either T, (n = 7) or vehicle solution (n = 7) after ischemia. Triiodothyronine or vehicle was given in a randomized, investigator-blinded manner. Baseline data were obtained after approximately 15 minutes of cardiopulmonary bypass. After baseline data had been obtained, hearts from both groups were subjected to 30 minutes of normothermic, global ischemia by aortic cross-clamping. A temperature probe positioned in

Ann Thorac Surg 1993:56215-22

the ventricular septum documented maintenance of myocardial temperatures at 37°C. Group I received T, supplementation (0.1 pg/kg) at the initiation of reperfusion, and a second dose after 60 minutes of reperfusion (0.1 pg/kg). Group I1 served as controls and received the vehicle solution at identical timepoints. Triiodothyronine and vehicle were administered through a port in the arterial perfusion cannula. Hemodynamic and metabolic data were obtained after 30, 60, 90, and 120 minutes of reperfusion.

Assessment of Ventricular Performance Hemodynamic and dimensional data were acquired online from the pressure catheters and ultrasonic crystals at 200 Hz on a Zenith 386 personal computer with interactive software developed in our laboratory. Analog data were digitized and passed through a recursive, low-pass Butterworth filter with a cut-off frequency of 30 Hz. The derivative of left ventricular pressure with respect to time (dP/dt) was calculated and the cardiac cycle defined such that maximally negative dP/dt defined end-systole. Left ventricular pressure-dimension work loops were produced over a physiologic range of end-diastolic volumes by emptying venous blood into the cardiopulmonary reservoir. The slope of the linear relationship between left ventricular stroke work and end-diastolic length has been demonstrated to be a load-insensitive index of contractility and was used to quantify ventricular performance [MI.

Assessment of Myocardial Adenine Nucleotide Pool Transmural myocardial biopsy specimens were obtained with a Tru-Cut needle (Travenol Laboratories, Deerfield, IL) before ischemia, at the end of ischemia, and after 30, 60, 90, and 120 minutes of reperfusion. Biopsy specimens were obtained from the anterior and posterior left ventricular walls randomly, with care to avoid injury to the coronary vasculature. Biopsy specimens were blotted and immediately frozen (within 3 seconds) in liquid nitrogen for storage until final analysis. Frozen tissue was homogenized in 0.4 mL 12% trichloroacetic acid for 30 minutes and centrifuged at 8,000 rpm for 10 minutes to separate denatured protein. The supernatant contained the soluble acid extract, from which adenine nucleotides and metabolites were measured using high-pressure liquid chromatography (model M-490; Waters Associates, Milford, MA) [19]. Protein content of the sediment was determined as previously described [20]. Final values of adenine nucleotides are expressed as nanomoles per milligram of protein.

Electron Microscopy Myocardial ultrastructure was evaluated in five hearts from each group. Left ventricular transmural myocardial biopsy specimens were obtained before ischemia and after 120 minutes of reperfusion using a Tru-Cut needle. Biopsy specimens were immediately placed in fresh 3% glutaraldehyde, 1.5% paraformaldehyde in 0.1 moVL sodium cacodylate buffer (pH 7.4) and stored until processing for transmission electron microscopy. Before fixation, biopsy specimens were rimed three

DYKEETAL T, SUPPLEMENTATION AFTER ISCHEMIA

Ann Thorac Surg 1993;56:21522

Table 1. Grading Scale for Myocardial Ultrastructure" Grade Nuclei Mitochondria 0 1 2 3

Normal; no margination Normal; no swelling or membrane or dense bodies disruption Slight margination and/ Slight swelling, no or perinuclear edema dense bodies Margination, edema, and/or nuclear clearing Dense margination and nuclear clearing

Myofibrils

Interstitium

217

Sarcolemma

Normal; no edema or

Mild edema

Normal; continuous membrane, glycocalyx still intact

I-band-Z-line

Mild to moderate edema Moderate to severe edema

No membrane discontinuity, but glycocalyx disrupted

disruption fuzziness

Swollen, dense bodies Band fuzziness, Some membrane discontinuity, and/or clumping of hypercontraction no glycocalyx cristae and/or edema All or most of above; Edema, Severe edema Severely disrupted membrane membrane fragmentation, disruption contraction banding

Organelles were scored on a 0 (no damage) to 3 (complete disruption) scale. All specimens were scored by an independent and blinded microanatomist (K.L.)

a

unpaired t test. Significance was accepted at the p less than 0.05 level. Unless otherwise stated, all values are expressed as mean f standard error of the mean.

times in 0.1 moYL sodium cacodylate buffer, pH 7.4 with 4% sucrose added, and postfixed for 2 hours with 2% osmium tetroxide, 0.8% potassium ferricyanide, in 0.1 moYL sodium cacodylate on ice. Samples were rinsed with water and en bloc stained with saturated aqueous uranyl acetate for 2 hours at 60°C, followed by dehydration in a graded series of ethanols, then propylene oxide. After embedding in Spurr resin, silver sections were cut from each of the samples and photographed on a JEOL EM 1200 electron microscope. Micrographs were quantitatively evaluated for myocyte damage using a 0 (indicating no damage) to 3 (indicating complete disruption) grading system for the nucleus, myofibrils, sarcolemma, mitochondria, and interstitium (Table 1).Samples were coded, prepared, and graded in a blinded fashion; the code was broken and the results were tabulated only after completion of ultrastructural grading.

Results Hemodynamic Data No significant differences in preischemic, baseline hemodynamic data were evident between groups (Table 2). Left ventricular systolic pressure, end-diastolic pressure, dP/ dt, end-diastolic length and heart rate were comparable between groups. After ischemia and 120 minutes of reperfusion, left ventricular peak pressure, end-diastolic pressure, dP/dt, and end-diastolic length were not significantly different from baseline in the group treated with T,. Hearts treated with T, did have a significantly higher heart rate compared with baseline, however (see Table 2). After ischemia and 30 minutes of reperfusion, the slope of the left ventricular stroke worWend-diastolic relationship, a load-independent index of contractility, was significantly depressed compared with preischemic values in both groups (Fig 1).The group of animals treated with T,, however, demonstrated a gradual improvement in ventricular function over the ensuing 2-hour reperfusion interval, whereas ventricular function in the placebo group remained depressed (see Fig 1).This improvement in functional recovery approached significance after 60

Statistical Analysis All statistical analysis was performed using personal computer-based software from the SAS Institute (Cary, NC). Data obtained at multiple time points were compared using a one-way analysis of variance. The difference in ability to be weaned from cardiopulmonary bypass between groups was assessed using ,$ analysis with Fisher's exact test. Electron microscopic ultrastructural scores between the two groups were compared using an

Table 2. Load-Devendent Hemodunamic Data" ~

Time Baseline ( -)T3 ( +)T3 120 min R ( -)T3 (+IT3

~~

LV PkP

*

LVEDP

93.1 4.6 86.8 2 4.2

6.6 ? 1.3 6.5 f 1.1

70.8 81.2

7.5 7.2

?

f

7.1 3.4

5 f

dP/dt

2310 2504

? f

487 334

1360 2 335 2142 f 276

1.8 1.4 ~~~~

LV EDL 48.6 47.8

2

f

2.6 0.8

51.0 ? 2.2 50.5 .+- 1.3

HR 140

143

f

Wean 8

*9

148 ? 8 167 2 6b

... ... 217 717'

~

Load-dependent hemodynamic data in the control group [(-)T3] and the group treated with T, [(+)T3]before ischemia and after 120 minutes of p < 0.05 ,$ analysis. reperfusion are detailed. p < 0.05 versus baseline by analysis of variance. a

dPldt = first time derivitive of pressure; LVEDP = left ventricular end-diastolic HR = heart rate; LVEDL = left ventricular end-diastoliclength; pressure; LV PkP = left ventricular peak pressure; R = reperfusion; Wean = ability to be weaned successfully off cardiopulmonary bypass after 120 minutes of reperfusion.

218

DYKEETAL T3 SUPPLEMENTATION AFTER ISCHEMIA

Ann Thorac Surg 1993;56215-22

phate, adenosine monophosphate, adenosine, hypoxanthine, xanthine, or total adenine nucleotide content between the group treated with T, and the group treated with placebo, either after ischemia or during reperfusion.

Myocardial Ultrastrucfure

2ot '

*p<0.05 ANOVA t p<0.05 ANOVA

PiE

g0

$0

i0

I;O

'

Time after Ischemia (minutes) Fig 1. The slope of the stroke worklend-diastolic length (SWEDL) relationship was used to quantify left ventricular contractility. Contractile function decreased significantly after ischemia and 30 minutes of reperfusion in both groups. Hearts treated with triiodothyronine fld, however, demonstrated a significant improvement in systolic function compared with controls, such that after 120 minutes T3treated hearts had returned to baseline function, whereas systolic function in hearts treated with placebo remained approximately 50% of baseline throughout the reperfusion period. (ANOVA = analysis of variance.)

minutes of reperfusion ( p = 0.10) and was statistically significant after 90 and 120 minutes of reperfusion ( p < 0.05). Left ventricular systolic function in the vehicletreated group did not improve and remained significantly depressed compared with baseline and the T,-treated group. After 2 hours of reperfusion, all animals treated with T, were weaned and maintained off cardiopulmonary bypass without the use of inotropic support. In comparison, only 2 of 7 animals in the placebo group were able to be sustained off cardiopulmonary bypass at the end of 2 hours of reperfusion ( p = 0.02 by 2 analysis) (see Table 2). Hemodynamic data from the animals treated with placebo and unable to be fully weaned from bypass were obtained after optimization of preload and with partial support from the cardiopulmonary bypass pump.

Metabolic Data At baseline, there was no difference in left ventricular adenosine triphosphate (ATP) levels or levels of ATP metabolites between groups (Fig 2; Table 3). After 30 minutes of normothermic ischemia, myocardial ATP levels were significantly depressed in both groups. In both groups, ATP levels remained depressed compared with baseline, preischemic levels throughout reperfusion. There was no difference in ATP levels in the group treated with T, compared with control. Similarly, T, supplementation did not affect ATP metabolism after ischemia (see Table 3). There was no significant difference in myocardial adenosine diphos-

There were no ultrastructural differences between groups before ischemia. Figures 3 and 4 are representative micrographs after ischemia and 120 minutes of reperfusion from the group treated with T, and the group treated with placebo, respectively. In the postischemic hearts treated with placebo, the sarcolemma is disrupted and incongruent, whereas in the T,-treated group the integrity of the sarcolemmal membrane is preserved. Similarly, the mitochondria in the hearts treated with T, are less swollen, with less disruption of the christae. These qualitative differences were significant when myocardial ultrastructure was assessed quantitatively using the schema outline above (see Table 1). The sarcoplasmic reticulum and mitochondria were significantly better preserved when treated with T, (Table 4).No differences in other cellular organelles were observed.

Comment It is increasingly evident that cardiopulmonary bypass affects thyroid hormone metabolism, leading to a euthyroid-sick state characterized by low levels of circulating free T,. Although the pathophysiologic significance of this acute stress-induced hypothyroidism is not clear, Novitsky and associates [ll, 13-16] have suggested that acute correction of the low T, state has significant hemodynamic benefit. In extensive laboratory and clinical work, they have suggested that T, is a positive inotrope,

-g c .-

a,

40

35 . 30

.

n o 25.

:G -a, c -

-

20. 15. *p<0.05 ANOVA

l5 or U

PRE POST 30

60

90

120

Time after Ischemia (minutes) Fig 2. Myocardial adenosine triphosphate (ATP) levels in both groups decreased approximately 50% from baseline after ischemia and remained depressed throughout the reperfusion period. There was no difference in ATP levels in hearts treated with triiodothyronine CrJ or placebo. (ANOVA = analysis of variance.)

219

DYKEETAL T, SUPPLEMENTATION AFTER ISCHEMIA

Ann Thorac Surg 1993;5621522

Table 3. Effect of Triiodothyronine Supplementation on Adenine Nucleotides, Purine Bases, and NAD+ After Ischemia" Before Ischemia

After Ischemia

5.8 6.7

7.5 6.5

30 min Reperfusion

60 min Reperfusion

90 min

Reperfusion

120 min

Reperfusion

ADP

(-V3 (fF3 AMP ( -IT3

f

0.4

f

1.0

f f

0.6 0.6

4.0 5.4

f 0.4 f 2.0

4.0 3.9

f f

0.6 0.5

3.3 4.3

f f

0.7 0.3

3.2 3.6

f

0.2 0.3

f

0.1 0.08

f

?

0.7 f 0.3 0.2 f 0.07

1.5 f 0.6' 2.1 t 0.7b

0.1 0.2

0.2 f 0.07 0.04 t 0.02

1.9 f 0.5b 0.8 f 0.3'

0.14 0.1

0 0

1.2 f 0.4' 1.6 f 0.4'

0

(+)T3

0

0 0.4 t 0.3

0 0.06 t 0.05

0 0

( -)T3 ( +)T3

0 0

0 0

0 0.6 t 0.7

0 0

0 0

0 0

(+F3

Adenosine (-)T3 ( +)T3

Hypoxanthine ( -)T3

Xanthine NAD (-IT3 ( +)T3 EadN (-F3 (+IT3

4.8 5.1

f

34.3 36.9

f

f

f

4.2 t 0.2 3.1 f 0.6

0.4 0.6

25.1 23.8

2.2 3.9

f

3.1b

f 4.4'

4.8 4.2

f 0.08 f 0.07

f 0.07 f 0.05

f 0.4

Ifr

0.3

20.6 f 2.4' 22.1 2 4.7b

0.2 f 0.08 0.4 t 0.2 0.14 0.1

4.7 5.2

f

f

f

f

0.06 0.13

0.6 0.8

22.7 f 3.4' 19.1 f 4.1b

Of0 0.2 f 0.06 0.10 0.13

f f

0.06 0.08

4.7 2 1.0 5.3 f 0.6

3.9 4.4

19.9 f 4.8b

20.9

f

3.8b

f

?

0.7 0.4

0.1 0.09 0.8 0.06

f

0.8

f

1.5

20.6 2 4.3b 16.4 f 3.Zb

Myocardial adenine nucleotide and metabolite levels were measured at baseline, immediately before reperfusion (after ischemia), and every 30 minutes p < 0.05 versus baseline by analysis of variance. during reperfusion. Values are expressed as nanomoles per milligram of protein.

a

ADP = adenosine diphosphate; AMP dinucleotide; T, = triiodothyronine.

=

adenosine monophosphate;

perhaps acting by improving the "aerobic capacity" of the myocardium after ischemia and thereby leading to an enhanced availability of myocardial high-energy phosphates available for contractile work [15]. However, their studies have lacked a load-independent demonstration of enhanced contractility. Thyroid hormone has multiple systemic effects and is a potent mediator of cellular thermogenesis and local arterial vasodilatation [21-241. Chronic hyperthyroidism is well recognized to lower systemic vascular resistance. Acute T, administration may also profoundly alter ventricular loading conditions. Load-dependent estimates of ventricular function such as cardiac output or stroke work, therefore, may be significantly altered on this basis alone, without any change in myocardial contractility [25]. Additionally, cardiopulmonary bypass induces large fluid shifts, which alter ventricular loading conditions. In an earlier study using an isolated rabbit heart model to rigorously control ventricular loading conditions, we were unable to demonstrate an inotropic effect of acute T, administration in normal hearts, although acute T, administration significantly and rapidly improved left ventricular function after ischemia ii7i. This study was not designed to determine the mechanism Of action Of T, activity after ischemia. The present study confirms these earlier findings of a rapid enhancement of myocardial mechanical perfor-

EadN = total adenine nucleotides;

NAD

=

nicotinamide adenine

Fig 3, A representative electron micrograph fYom the left ventricle of a heart treated with placebo after 30 minutes of ischemia and 120 minutes of reperfusion. The arrow points to a break in the sarcolemma. Mitochondria (m) are swollen and edematous, whereas the contractile filaments are well preserved.

220

DYKEETAL T, SUPPLEMENTATION AFTER ISCHEMIA

Ann Thorac Surg 1993;56:21522

Fig 4 . A representative electron micrograph from the left ventricle of a heart treated with triiodothyronine after 30 minutes of ischemia and 220 minutes of reperfusion. Arrowhead points to calcium deposits within the mitochondria1 cristae. arrow demonstrate a lipid deposit. The sarcolemmas in the photomicrograph are particularly well preserved. (m = mitochondria; N = nucleus.)

mance in an in vivo, large animal model using a loadindependent index of ventricular function. No differences in left ventricular systolic pressure, end-diastolic pressure, end-diastolic length, dP/dt, or heart rate were present between groups before ischemia (see Table 2). After 30 minutes of normothermic ischemia and 30 minutes of reperfusion, left ventricular systolic performance, as measured by the stroke work/end-diastolic length relationship, was significantly depressed in both groups, signifying significant injury. Left ventricular systolic performance gradually improved in the group supplemented with T,, whereas ventricular function in the placebo group remained depressed throughout the reperfusion interval (see Fig 1). The improvement in ventricular systolic function after ischemia in the group treated with T, was rapid but not immediate; after 60 minutes of reperfusion a trend was apparent, but a clear improvement in

systolic function was evident after 90 minutes of reperfusion. This enhancement in functional recovery after ischemia was not only statistically significant but clinically relevant, as each animal treated with T, was able to be sustained off cardiopulmonary bypass after 120 minutes of reperfusion without inotropic support. In contrast, only 2 of 7 animals were able to be successfully weaned from cardiopulmonary bypass when administered placebo. Although supporting the hypothesis that acute T, supplementation improves left ventricular mechanical recovery after ischemia and reperfusion, however, we did not find any effect of T, administration on ATP metabolism. Novitsky and associates have speculated that T, improves ventricular function after ischemia by improving the aerobic capacity of the myocardium and increasing the levels of high-energy phosphates available for contraction. In an earlier study, they reported a significant increase in myocardial ATP levels after ischemia in a group of baboons treated with T, compared with controls. Data detailing ATP degradation products were not presented in their report. It was hypothesized that the reported increase in myocardial ATP stores was mediated through the activity of T, on the adenine nucleotide translocase moiety located on the inner mitochondrial membrane. This enzyme, a regulator of adenosine diphosphate/ATP exchange across the inner mitochondrial membrane, binds T, and has been demonstrated to be rapidly up-regulated by exogenous administration of T, [26]. Our data, however, do not support the hypothesis that T, improves myocardial performance by enhancing the availability of myocardial high-energy phosphate stores. In contrast to Novitsky and associates' work, we did not demonstrate an affect of T, on ATP metabolism after ischemia. In both placebo- and T,-treated groups, myocardial ATP levels decreased 50% to 60% from baseline after ischemia and remained depressed throughout 2 hours of reperfusion. Similarly, there was no difference in ATP degradation products between groups. Stores of diffusable metabolites (adenosine monophosphate, hypoxanthine, and xanthine) were increased in both groups after ischemia but before reperfusion, and were washed out with reperfusion. Whether this is consistent with Novitsky and associates' study is unknown because ATP metabolites were not reported in their report [15]. Although total myocardial adenine nucleotide stores can be measured quite accurately, the intracellular distri-

Table 4. Myocardial Ultrastructure" Time

Mitochondria

Nucleus

Myofibril

Interstitium

Sarcolemma

Before ischemia

1.24 2 0.14

0.32 2 0.16

1.62 2 0.13

0.34 2 0.10

0.44 2 0.09

After ischemia ( -)T3 ( +)T3

1.74 ? 0.13 1.05 2 0.03b

0.7 ? 0.23 0.73 2 0.25

1.64 2 0.15 1.35 2 0.05

0.77 5 0.14 0.42 2 0.14

1.34 2 0.08 0.77 2 0.05b

~~~~~

a Using the ultrastructural gradlng scale detailed in Table 1, myocardial ultrastructure before ischemia and after 120 minutes of reperfusion was assessed. Mitochondria and the sarcoplasma reticulum from the group of animals treated with triiodothyronine (T3) showed a significant preservation of normal p < 0.05 versus controls architecture after ischemia and reperfusion compared with controls. Data are shown as mean t standard error of the mean. by unpaired t test.

DYKEETAL T, SUPPLEMENTATION AFTER ISCHEMIA

Ann Thorac Surg 1993;56215-22

bution of ATP stores is difficult to quantify. Possibly, changes in intracellular compartmentalization of ATP may be affected by T,, perhaps through the adenine nucleotide translocase moiety, and thereby affect contractile performance. These data do not address this hypothesis. Although thyroid hormone exerts powerful nuclearmediated effects on protein synthesis, the enhanced recovery of ventricular function is rapid enough to preclude a nuclear-mediated, transcriptional-based mechanism for enhanced contractility within this early reperfusion period [27]. It is doubtful that any significant change in myosin isoform expression could account for the functional improvement seen in this acute setting, although a posttranscriptional mechanism affecting protein synthesis could have a rapid affect on contractility. Triiodothyronine has several sites of action at the plasma membrane that may affect ventricular function. Of particular interest are the nonnuclear sites of action that affect calcium cycling within the cell. Triiodothyronine has been demonstrated to increase intracellular calcium content in isolated myocytes by upregulating the activity of the sarcoplasmic reticulum Caz+ adenosine triphosphatase [28, 291. Also, acute T, administration has been demonstrated to affect sodium channels on the plasma membrane, resulting in an increased inward sodium flux, which may affect contractility [30]. Both of these effects of T, administration on ion transport are rapid in onset and may enhance myofibrillar shortening. Whether the anatomical observation of superior sarcolemmal preservation with T, infusion is the ultrastructural correlate to the functional changes observed is not clear. However, both the sarcolemmal membrane and the mitochondria have been postulated to be mechanistically related to improved cardiac function with T, infusion after ischemic injury. Further work will be required to test this hypothesis. In summary, T, supplementation significantly improved left ventricular functional recovery after ischemia but did not affect ATP metabolism. Levels of ATP and its metabolites were depressed after ischemia in both the group treated with T, and placebo. This suggests that the enhancement of left ventricular systolic performance after ischemia with T, supplementation is not mediated by an increased availability of myocardial high-energy phosphates. Although not immediate, the improvement in function was rapid enough to likely preclude a nuclearmediated mechanism involving myosin isoform switches. The improvement in recovery of the sarcoplasmic reticulum and mitochondria on transmission electron micrographs suggests that these organelles may be related t o the improvement in postischemic performance. Changes in ion flux across the plasma membrane and sarcoplasma reticulum remain an attractive hypothesis for the enhancement of left ventricular functional recovery after ischemia that occurs with acute T, supplementation. This study was supported in part by grant 5 R 0 1 H12 26302-11 from the National Institutes of Health. We would like to acknowledge the expert technical assistance of Ms Nancy Lee and Ms Cynthia Lanier, and the steadfast support and administrative assistance of Mr Steve Crute.

221

References 1. Rahimtoola SH, Nunley D, Grunkemeir G, Tepley J, Lambert L, Starr A. Ten year survival after coronary bypass surgery. N Engl J Med 1983;308:676. 2. Reduto LA, Lawrie GM, Reid JW, et al. Sequential postoperative assessment of left ventricular performance with gated cardiac blood pool imaging following aortocoronary bypass surgery. Am Heart J 1981;101:59. 3. Gray R, Maddahi J, Berman D, et al. Scintigraphic and hemodynamic demonstration of transient left ventricular dysfunction immediately after uncomplicated coronary artery bypass grafting. J Thorac Cardiovasc Surg 1979;77504. 4. Maddahi J, Gray R, Berman D, et al. Hemodynamic and scintigraphic demonstration of transient right ventricular dysfunction immediately after uncomplicated coronary artery bypass surgery [Abstract]. Am J Cardiol 1979;43:423. 5. Roberts AJ, Spies SM, Meyers SN, et al. Early and long-term improvement in left ventricular performance following coronary bypass surgery. Surgery 1980;88:467. 6. Phillips HR, Carter JE, Okada RD, et al. Serial changes in left ventricular ejection fraction in the early hours after aortocoronary bypass grafting. Chest 1983;88:28. 7. Bremner WF, Taylor KM, Baird 5, et al. Hypothalamopituitary-thyroid axis function during cardiopulmonary bypass. J Thorac Cardiovasc Surg 1978;75:392-9. 8. Zaloga GR, Chernow 8, Smallridge RC, et al. A longitudinal evaluation of thyroid function in critically ill surgical patients. Ann Surg 1985;201:456-64. 9. Holland FW 11, Brown PS Jr, Weintraub BD, Clark RE. Cardiopulmonary bypass and thyroid function: a “euthyroid sick syndrome.” Ann Thorac Surg 1991;52:4&50. 10. Robushi G, Medeci D, Fesani F, et al. Cardiopulmonary bypass: a low T3 syndrome with blunted thyrotropin (TSH) response to thyrotropin-releasing hormone (TRH). Horm Res 1986;23:151-8. 11. Novitzky D, Cooper DKC, Swanepoel A. Inotropic effect of triiodothyronine (T3) following myocardial ischemia and cardiopulmonary bypass: initial experience in patients undergoing open-heart surgery. Eur J Cardiothorac Surg 1989; 3:140-5. 12. Chu 5-H, Huang T-5, Hsu R-8, Wang S-S, Wang C-J. Thyroid hormone changes after cardiovascular surgery and clinical implications. Ann Thorac Surg 1991;52:7914. 13. Novitzky D, Cooper DKC, Barton CI, et al. Triiodothyronine as an inotropic agent after open heart surgery. J Thorac Cardiovasc Surg 1989;98:972-8. 14. Novitzky D, Human PA, Cooper DKC. Inotropic effect of triiodothyronine following myocardial ischemia and cardiopulmonary bypass: an experimental study in pigs. Ann Thorac Surg 1988;45:50-5. 15. Novitzky D, Human PA, Cooper DKC. Effect of triiodothyronine (T3) on myocardial high energy phosphates and lactate after ischemia and cardiopulmonary bypass. J Thorac Cardiovasc Surg 1988;96:600-7. 16. Novitzky D, Matthews N, Shawley D, Cooper DKC, Zuhdi N. Triiodothyronine in the recovery of stunned myocardium in dogs. Ann Thorac Surg 1991;51:10-7. 17. Dyke CM, Yeh T Jr, Lehman JD, et al. Triiodothyronineenhanced left ventricular function after ischemic injury. Ann Thorac Surg 1991;5214-9. 18. Salter DR, Goldwtein JP, Abd-Elfattah A, Murphy CE, Brunsting LA, Wechsler AS. Ventricular function after atrial cardioplegia. Circulation 1987;76:12940. 19. Hull-Ryde EA, Lewis WR, Veronne CD, Lowe JE. Simple step gradient elution of the major high-energy compounds and their catabolites in cardiac muscle using high performance liquid chromatography. J Chromatogr Biomed Appl 1986;377:16574. 20. Lowry OH, Rosenbrough NJ, Farr AL, Randall RJ. Protein measurement with the Folin phenol reagent. J Biol Chem 1951;193:265-75. 21. Levey GS, Klein I. Catecholamine-thyroid interactions and

222

22. 23. 24. 25.

DYKEETAL T, SUPPLEMENTATlONAITER ISCHEMlA

the cardiovascular manifestations of hyperthyroidism. Am J Med 1990;88:6424. Klein I. Thyroid hormone and the cardiovascular system. Am J Med 1990;88:631-7. DeNayer P. Thyroid hormone action at the cellular level. Horm Res 1987;26:48-57. Salter DR, Dyke CM, Wechsler AS. Triiodothyronine (T3) and cardiovascular therapeutics: a review. J Cardiac Surg 1992;7:1-12. Feldman T, Borow KM, Sarne DH, et al. Myocardial mechanics in hyperthyroidism: importance of left ventricular loading conditions, heart rate and contractile state. J Am Coll Cardiol 1986;7967-74.

Ann Thorac Surg 1993;56:21522

26. Sterling K. Direct thyroid hormone activation of mitochondria: the role of adenine nucleotide translocase. Endocrinology 1986;119:292-5. 27. Dillman WH. Biochemical basis of thyroid hormone action on the heart. Am J Med 1990;88:626-30. 28. Segal J. Action of the thyroid hormone at the level of the plasma membrane. Endocrine Res 1989;15:619-49. 29. Lomax RB, Cobbold PH, Allshire AP, et al. Triiodothyronine increases intracellular calcium levels in single rat myocytes. J Mol Endocrinol 1991;7:77-9. 30. Dudley SC, Baumgarten CM. Physiologic concentrations of triiodo-L-thyronine increase bursting of cardiac sodium channels. Biophys J 1991;59:98a.

DISCUSSION D R JULIE A. S W A I N (Las Vegas, NV): Your conclusion is that you do not know the mechanism, but I am sure you have speculated on the mechanism. What are your speculations? My second question is that your study shows that giving T, during an operation appears to be beneficial. Have you considered preoperative treatment? Clinically we would think that this might help the heart failure and valve replacement patients to try to effect a V1-V3 switch of the myosin heavy chain isoforms as suggested by the work from Mahdavi and Nadal-Ginard [l]. D R DYKE: Thank you for your comments. The mechanism, as we alluded to, has not been well described or proved; however, we think that one particular area, namely, the cycling of intracellular calcium or sodium ions, either through the calcium adenosine triphosphatase or the sodium-potassium adenosine triphosphatase, respectively, may contribute to the improvement in function that we see after ischemia. We believe that the rapid response, within an hour, rules out a nuclear-based, transcriptional-mediated mechanism. Regarding your second question, we have not looked at preoperative T, supplementation in this model, although similar studies in an isolated heart model are ongoing. However, one of the rationales for looking at T, supplementation perioperatively is that most patients are euthyroid preoperatively, and the initiation of the euthyroid-sick syndrome seems to occur with cardiopulmonary bypass and continues through the early postoperative period. D R DIMITRI NOVITZKY (Tampa, FL): I would like to congratulate you for this excellent presentation, which further supports the work that was initiated in Cape Town using T, for nonthyroid conditions. Our first observation of a low free T3 level state was inducing experimental brain death in baboons. This free T, reduction followed an initial catecholamine storm; the thyroidstimulating hormone level remained unchanged. A similar response was observed during and after cardiopulmonary bypass. Triiodothyronine was administered to pigs subjected to cardiopulmonary bypass for 3 hours of aortic cross-clamping and cardioplegic arrest; these animals were easily weaned from the heart-lung machine, sustaining good hemodynamics for 2 hours. However, animals not receiving T, remained bypass-dependent and died by the end of the experiment. Further studies performed in baboons showed that the administration of T, definitely prevented the cardiac tissue lactic acidosis and normalization of ATP levels. Approximately 8 years ago, I administered T, to a patient who underwent emergent mitral valve replacement. She was septic, had bacterial endocarditis and pneumonia, and was intubated for pulmonary edema. After the valve replacement, in spite of

multiple defibrillation attempts, maximum inotropic support, and boluses of calcium, she remained dependent on cardiopulmonary bypass. The anesthesiologist in charge of the case suggested that we give T, to this patient. We already had gained significant experience administering T, and rescuing hemodynamically unstable brain-dead organ donors dependent on highdose inotropic support, observing excellent hemodynamics in the recipients after cardiac transplantation. The patient received 6 pg of T3 as a bolus; within few minutes, the fibrillation became more pronounced and the heart was defibrillated successfully, starting in sinus rhythm and allowing us to discontinue cardiopulmonary bypass within 20 minutes with only minimal inotropic support. The low free T, state is part of the “euthyroid sick syndrome” observed in acute shock states and chronic conditions such as cancer or starvation. I have found T3 replacement beneficial after acute insults, and I believe this condition should be totally differentiated from the chronic states. In the acute phase such as sepsis, hemorrhagic shock, in brain-dead organ donors, and in patients on cardiopulmonary bypass, the administration of T, was beneficial in the recovery of the myocardium, allowing rapid reduction of inotropic support. Triiodothyronine also has systemic effects and may also have a beneficial effect on other organs. At the present stage, two mechanisms of action are apparent: the first seems to be extranuclear, affecting various calciumdependent adenosine triphosphatases acting on the contractile apparatus and also on the mitochondria-stimulating aerobic metabolism, and the second is mediated via DNA-RNA-protein synthesis. Have you measured free T, levels and the impact of heparin on these? Have you measured levels of other hormones besides T3 such as reverse T,, total thyroxine, free thyroxine, thyroidstimulating hormone, and thyrotropin-releasing hormone? I congratulate you for this excellent piece of work. DR DYKE: Thank you, Dr Novitzky. We are well aware of and

appreciate your work beginning in the early 1980s. A stimulus for this study was the increase in ATP with T, supplementation that you described in one of your early studies. Although we demonstrated an improvement in cardiac performance, we were unable to show a change in ATP levels with T, supplementation as your earlier work did, and that has led us away from an increase in oxidative phosphorylation capacity of the myocardium as a primary mechanism of action. We have not looked at the effect of heparin on T, levels.

Reference 1. Mahdavi V, Nadal-Ginard 8 . Molecular basis of cardiac contractility. In: Yacoub M, Pepper J, eds. Annual of cardiac surgery. London: Current Science Ltd, 1991.