Temporary leukocyte depletion reduces ventricular dysfunction during prolonged postischemic reperfusion

Cardiopulmonary Bypass, Myocardial Management, and Support Techniques

Temporary leukocyte depletion reduces ventricular dysfunction during prolonged postischemic reperfusion Leukocyte depletion improves early postischemic ventricular performance in neonatal models of global myocardial ischemia. However, the rate of leukocyte reaccumulation after cardiopulmonary bypass and its subsequent impact on myocardial function is not known. This laboratory study examined the effect of leukocyte depletion on myocardial performance during the initial 6-hour period after bypass in an in situ, in vivo porcine model of neonatal cardiac surgery. Fifteen 3- to 5-day-old piglets (eight control and seven leukocyte depleted animals) were instrumented by placement of left ventricular short-axis sonomicrometry crystals and an intraventricular micromanometer catheter. Mechanical leukocyte depletion was achieved with Pall RCI00 filters (Pall Biomedical, Inc., Fajardo, Puerto Rico) in the cardiopulmonary bypass circuit. Neonatal hearts were subjected to 90 minutes of hypothermic ischemia after a single dose of cold crystalloid cardioplegia. Two control animals died after the operation and were excluded from data analysis. Leukocyte filtration reduced the granulocyte count during initial myocardial reperfusion to 0.8 % of control values. However, circulating granulocyte counts increased in leukocyte depleted animals throughout the postoperative period, reaching 68 % of control values by 6 hours. Despite this rapid return of circulating granulocytes, animals subjected to leukocyte. depletion had significantly better preservation of left ventricular performance (measured by preload recruitable stroke work, p ::5 0.02), left ventricular systolic function (measured by end-systolic pressure-volume relationship, p ::5 0.05~ and ventricular compliance (p ::5 0.04) during the experiment. These changes in ventricular function were associated with a significant increase in left ventricular water content (p ::5 0.02) and tissue myeloperoxidase activity (p ::5 0.005) in control animals compared with leukocyte depleted animals. This study demonstrates that leukocyte depletion during initial reperfusion results in sustained improvement in postischemic left ventricular function despite the rapid return of granulocytes to the circulation. (J THoRAc CARDIOVASC SURG 1993;106:805-10)

Ian C. Wilson, MB, ChB (by invitation), Timothy J. Gardner, MD, Joseph M. DiNatale, BS (by invitation), A. Marc Gillinov, MD (by invitation), William E. Curtis, MD (by invitation), and Duke E. Cameron, MD (by invitation), Baltimore, Md.

From the Department of Cardiac Surgery, Johns Hopkins Medical Institution, Baltimore, Md. Read at the Seventy-second Annual Meeting of The American Association for Thoracic Surgery, Los Angeles, Calif., April 26-28, 1993. Address for reprints: D. E. Cameron, MD, Department of Cardiac Surgery, Johns Hopkins Medical Institution, 600, North Wolfe St., Baltimore, MD 21205. Copyright w 1993 by Mosby-Year Book, Inc. 0022-5223/93 $1.00 +.10

12/6/48808

LeUkocyte depletion during myocardial ischemia and reperfusion reduces postischemic myocardial leukocyte accumulation 1 and improves early postischernic ventricular function in adult and neonatal hearts. 2, 3 However, the leukocytes are known to return rapidly to the circulation; the effect of their reappearance on myocardial function is not known. The present laboratory study examined the result of mechanical leukocyte depletion on

805

806

The Journal of Thoracic and Cardiovascular Surgery November1993

Wilson et al.

-

3,000

~

Control

-cr-

LD

::€ 2,500 ~

..

~

'ai
!. en

w

1,500

§ ...J

c

P B

1,000

::::>

z ~ (!J

500 __ -:0:-

A

0

-I-~ 1

234

5

6

TIME POST CPS (hours)

Fig. 1. Granulocyte counts measured during initial myocardial reperfusion (R) and after discontinuation of CPB.

myocardial performance during a 6-hour period immediately after cardiopulmonary bypass (CPB) in an in situ, in vivo porcine model of neonatal cardiac surgery.

Methods and materials Fifteen 3- to 5-day-old piglets were premedicated with ketamine (44 mg/kg) and anesthetized by inhalation of 2% isoflurane. Mechanical volume ventilation was initiated with an inspired oxygen concentration of 1.0. Maintenance anesthesia was provided by isoflurane inhalation (0.5% to 1.0%). The ductus arteriosus was ligated after median sternotomy and both vagal nerves were divided to reduce autonomic reflex influence on left ventricular function. Heparin (3 rug/kg) was administered intravenously. Left ventricular pressure was recorded by a micromanometer catheter (model PC-350, Millar Instruments, Inc., Houston, Tex.) inserted into the ventricle through its apex, and left ventricular short-axis dimension was recorded by subendocardially placed pulse-transit ultrasonic dimension transducers (Physiological Monitoring Systems, Duke Medical Center, Durham, N.C.) connected to a sonomicrometer (model 120, Triton Technology, Inc., San Diego, Calif.). Tapes were positioned around the inferior and superior venae cavae to allow manipulation of ventricular preload during evaluation of ventricular function. The 15 piglets were divided into two groupseight in the control group and seven in the leukocyte depletion (LD) group. CPB. The CPB circuit was constructed with a bubble oxygenator (Bentley-5 Pediatric Blood Oxygenator; Bentley American, Irvine, Calif.) and a pediatric extracorporeal blood filter (Pall LPE-1440; Pall Biomedical, Inc., Fajardo, Puerto Rico). The circuit was primed with heparinized whole blood. In the LD group, leukocyte depletion was achieved by three leukocyte filters (Pall RCI 00; Pall Biomedical Inc.) positioned in parallel on the arterial limb of the bypass circuit. Piglets were cooled to 22 0 C (rectal) before crossclamping of the aorta. Cold crystalloid cardioplegic solution (Johns Hopkins cardioplegic solution, 20 ml/kg) was administered in antegrade fashion. The patent foramen ovale was closed during a brief period of circulatory arrest in all animals. Myocardial temperatures were maintained

below 150 C by topical application of 4 0 C saline solution. Normothermic myocardial reperfusion was instituted after 90 minutes of ischemia and was continued for 30 minutes with the heart in a beating, nonworking state before weaning from CPB. Atrial pacing (150 beats/min) was used for any heart with an intrinsic rate below 140 beats/min. Measurements and calculations I. Systemic arterial leukocyte counts were measured with a Coulter Counter (Coulter Electronics, Inc., Hialeah, Fla.) during CPB, every 30 minutes for 3 hours after CPB, and hourly thereafter. 2. Left ventricular function was analyzed by assessment of preload recruitable stroke work (representing overall ventricular performance), end-systolic pressure-volume relationship (representing ventricular systolic function), and ventricular stiffness. Measurements were performed before ischemia and every 30 minutes after discontinuation of CPB. 3. Myocardial water content was calculated at completion of the experiment by incubating a section of the left ventricle at 1000 C for 24 hours. 4. Myocardial myeloperoxidase activity was measured on left ventricular biopsy specimens taken at the conclusion of the experiment by standard spectrophotometric functional assay. Results are presented as micromoles per 10 mg wet weight of tissue. Experimental protocol. Hematologic and left ventricular functional analyses were performed serially for 6 hours after discontinuation of CPB. Ventricular biopsy specimens were taken at the completion of the experiment. Statistical analysis. Data are represented as the mean ± standard error of the mean. Differences between the two groups were tested by Student's t test for unpaired samples or repeated-measures analysis of variance, as appropriate. Animal care. 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 Institute of Laboratory Animal Resources and published by the National Institutes of Health (NIH Publication No. 86-23, revised 1985).

The Journal of Thoracic and Cardiovascular Surgery Volume 106, Number 5

Wilson et al.

807

110

--I-of - I- +l f -I- -!-f -1--1

~ 100 a: a.. w 90

Z :::::;

w 80 (/)

~ > 70 a: w > 0

C p

60

U

w

a:

*'

B

-- Control

50

.... LD

40 Baseline

a

1

234

5

6

TIME POST CPB (hours)

psO.02 Fig. 2. Percent recovery of preload recruitable stroke work (PRSW) measured after discontinuation of CPB.

120

a: 110 ~ (/) w 100

w Z :::::; 90 w

i

80

~ w

70

§a: *'

60 50

. , , ... .. " " .. ...

~

.

·""f-r~

c P B

--Controls

a

--- LD 1

234

5

6

TIME POST CPB (hours)

psO.05 Fig. 3. Percent recovery of end-systolic pressure-volume relationship (ESPVR) after discontinuation of CPB.

Results All piglets were weaned from CPB satisfactorily. However, two piglets in the control group died of progressive left ventricular failure 3V2 hours and 4V2 hours after CPB. These animals were excluded from the analysis. In the LD animals mechanicalfiltration of leukocytes

resultedin reductionof the granulocytecount to 0.8% of controlgroupvaluesduringinitialmyocardialreperfusion (2 ± 1 . 103/ml in LD group and 417 ± 174· 103/ml in control group). After the operation, the granulocyte count increased at a similar rate in both groups until completion of the experiment, at whichpoint circulating granulocyte countsin the LD group had reached 68%of

The Journal of Thoracic and Cardiovascular Surgery November 1993

8 0 8 Wilson et al.

70

en en w

z

lt~

60 50

-"

~" en Q:' 40 a: "

:5 ~30 ::::>0 c.>cn 0: ...... z~

c

-I- -1--1-I -1- -I -I

P 8

'/

20

"

w

>

--- Control

10

....... LD

0'--------------------------

Baseline

o

123

4

5

6

TIME POST CPB (hours)

pSO.04 Fig. 4. Ventricular stiffness before ischemia and after discontinuation of CPR Stiffness calculated as the slope dP /dd at end-diastole (units = mm Hg/cm), The x-intercept remained constant between animals and between the experimental groups.

the count in the control group (1374 ± 355 . 103 jml in the LD group compared with 2013 ± 495 . 103jml in the control group). The percentage of immature granulocytes within the circulating granulocyte pool fell in both groups during initial coronary reperfusion (45% ± 7% and 44% ± 4% before CPB in the control and LD animals, respectively, compared with 20% ± 7% and 7% ± 7%, respectively, after removal of the aortic crossclamp). However, there was no difference in granulocyte maturity between the two groups throughout the period of coronary reperfusion. Immature granulocyte counts increased to 29% ± 3% of the circulating granulocyte pool in the control group at 6 hours after CPB compared with 43% ± 6% in the LD group. There was also no statistically significant difference in circulating platelet counts between the two groups during coronary reperfusion. Despite rapid return of granulocytes to the circulation in both groups, the pattern of postischemic ventricular dysfunction was different in LD and control animals. Control animals demonstrated an immediate reduction in preload recruitable stroke work to 87% ± 6% of baseline values 30 minutes after CPR This reduction was followed by progressive deterioration of preload recruitable stroke work during the early postoperative period, reaching a nadir of 72% ± 13% ofbaseline values 4 hours after CPR Preload recruitable stroke work then improved to 86% ± 6% of baseline by the sixth postoperative hour. In

contrast, LD animals demonstrated a less severe reduction in preload recruitable stroke work to 94% ± 4% of baseline values at 30 minutes of reperfusion and no further reduction in preload recruitable stroke work throughout the remainder of the experiment (p ::::; 0.02). These differences in left ventricular performance (preload recruitable stroke work) were associated with alterations in ventricular systolic and diastolic function, each with its own temporal course. End-systolic pressure-volume relationship was reduced to 71% ± 5% of baseline values in control animals 3Qminutes after CPR Over the next 5V2 hours end-systolic pressure-volume relationship improved to reach 96% ± 8% of systolic ventricular function by 6 hours after the operation. LD animals demonstrated a similar pattern of early deterioration in end-systolic pressure-volume relationship (82% ± 4% 30 minutes after CPB) followed by progressive improvement during the remainder of the experiment (101 % ± 6% by 6 hours). However, the overall reduction in end-systolic pressure-volume relationship was significantly less, and there was more rapid recovery in the LD group than in the control group (p ::::; 0.05). The pattern of changes in left ventricular compliance was different from that of other measured indexes of ventricular function. Increased ventricular stiffness was documented in both groups at 30 minutes of reperfusion (control group, 32.0 ± 7.7 mm Hgjcm; LD group, 26.6 ± 3.8 mm Hgjcm). However, during prolonged

The Journal of Thoracic and Cardiovascular Surgery Volume 106, Number 5

reperfusion, control animals demonstrated a progressive increase in ventricular stiffness throughout the postoperative period, reaching 49.5 ± 8.8 mm Hg/cm by 6 hours. In contrast, ventricular stiffnessin LD animals did not increase further during the period of observation (28.2 ± 3.4 mm Hg/cm at 6 hours) and was significantly different from that of the control animals during the entire period of reperfusion (p -< 0.04). These differences in ventricular stiffness were associated with reduction of left ventricular water content in LD animals (79.6% ± 0.3%) compared with that in control animals (80.9% ± 0.3%, p :'5 0.02). In control animals, left ventricular myeloperoxidase activity was increased 210% compared with the activity in LD animals (102 ± 8 ~mol/lO mg left ventricle versus 48 ± 12 ~mol/ 10 mg left ventricle, respectively, p z: 0.005). Normal piglet myocardial myeloperoxidase activity was 32 ± 4 ~mol/ 10 mg left ventricle. Discussion Previousstudies of the influenceof neutrophil depletion during ischemia and reperfusion on myocardial function havedemonstrated an early improvement in postischemic ventricular function in reversibly injured myocardium.2, 3 However, the effect of the reappearance of granulocytes on late ventricular function is unknown. This study demonstrates that granulocytes return rapidly to thecirculation after discontinuation of mechanical filtration despite considerable perioperative depletion. The similarrate of leukocyte reappearance in LD and control animalssuggests that a large mass of leukocytes is available for mobilization from the bone marrow pool. The increased proportion of immature granulocytes documentedin this study also suggests a bone marrow rather than endothelial source. Demargination of leukocytes may occur later after CPB, and the effect of the circulationof these leukocytes may not have been examined by this study. Analysis of ventricular function in this model shows that the temporal courses of diastolic and systolicpostischemic ventricular dysfunction differ but that both are altered by leukocyte depletion. Previous clinical studies have shown a progressive reduction in left ventricular ejection fraction over the first 6 postoperative hours, with subsequent improvement to preoperative values by 48 hours." However, in these clinical series ventricular functionwas analyzed by means of afterload-dependent measurementswhile changes in systemic vascular resistance were occurring. The present study confirms progressive deterioration in global ventricular performance during earlymyocardial reperfusion, but uses an afterload-independent indicator of ventricular function (preload recruitable stroke work). Reduced global ventricular

Wilson et al. 8 0 9

performance appears to be a reflectionof both systolicand diastolic ventricular dysfunction. Systolic function (endsystolic pressure-volume relationship) reaches its nadir early during reperfusion and improves to normal values by 6 hours after the operation. However, ventricular compliance, measured as the inverse of ventricular stiffness, deteriorates over the time course of the experiment and reaches a plateau at 6 hours, a pattern also seen in clinical series.' Leukocyte depletion results in more rapid normalization of systolicfunction and lessdeterioration of diastolic function. The severe depression of contractile protein function seen in control animals may be due to the release of cytotoxic agents from the granulocytes within the coronary vascular bed6, 7 or to a reduction in coronary blood flow (the no-reflewphenomenon) as a consequence of capillary plugging.f The increased ventricular stiffness seen in control animals was associated with an increase of myocardial water content and may reflect greater endothelial injury and interstitial fluid accumulation during prolonged reperfusion. Previous work in this laboratory demonstrated that myocardial myeloperoxidase activity was increased by 255% after 90 minutes of reperfusion in control animals compared with normal piglet myocardium. This present study shows that extending the reperfusion period from 90 minutes to 390 minutes resulted in only a further 47% increase in myeloperoxidase activity. This observation suggests that myocardial leukocyte accumulation occurs predominantly early during reperfusion. LD animals also demonstrated an increase in myocardial leukocyte accumulation during prolonged reperfusion (27 ~mol/lO mg left ventricle at 90 minutes compared with 48 ~mol/ 10 mg left ventricle at 390 minutes), although the increase in myeloperoxidase activity was significantly less than in control animals. This reduced tendency to accumulate granulocytes during late reperfusion, despite the increase in circulating granulocytes, may be due to a reduction in myocardial chemotactic factors? during initial reperfusion or to a reduction in the surface expression of neutrophil adherence receptors after discontinuation of CPB. Regardless of the cause, this pattern of early leukocyte accumulation after ischemia may explain the prolonged benefits achieved by temporary leukocyte depletion. The results of this study suggest that granulocyte accumulation during early myocardial reperfusion significantly influences late ventricular function. Mechanical depletion of circulating granulocytes during ischemia and early reperfusion reduces postischemic ventricular dysfunction over the first 6 hours after CPB despite the rapid reappearance of granulocytes in the circulation. These findingsalso suggest important new pathways to explore in the continuing effort to improve myocardial protection in cardiac surgery.

8 10

The Journal of Thoracic and Cardiovascular Surgery November 1993

Wilson et al.

REFERENCES 1. Litt MR, Jeremy RW, Weisman HF, Winkelstein JA, Becker LC. Neutrophil depletion limited to reperfusion reduces myocardial infarct size after 90 minutes of ischemia: evidence for neutrophil-mediated reperfusion injury. Circulation 1989;80:1816-27. 2. Westlin W, Mullane KM. Alleviation of myocardial stunning by leukocyte and platelet depletion. Circulation 1989;80:1828-36. 3. Breda MA, Drinkwater DC, Laks H, et al. Prevention of reperfusion injury in the neonatal heart with leukocyte-depleted blood. J THORAC CARDIOVASC SURG 1989;97:65465. 4. Breisblatt WM, Stein KL, Wolfe CJ, et al. Acute myocardial dysfunction and recovery: a common occurrence after coronary bypass surgery. J Am Coll CardioI1990;15: 12619. 5. Rinder CS, Wheeler LR, Alpern WD. Pulsed Doppler assessment of diastolic function before and after cardiopulmonary bypass [Abstract]. Anesthesiology 1988;69:A2. 6. Lucchesi BR. Role of neutrophils in ischemic heart disease: pathophysiologic role in myocardial ischemia and coronary artery reperfusion. Cardiovasc Clin 1987;18:35-48. 7. Engler RL. Free radical and granulocyte-mediated injury during myocardial ischemia and reperfusion. Am J Cardi01 1989;63:19E-23E.

8. Schmid-Schonbein G. Capillary plugging by granulocytes and the no-reflow phenomenon in the microcirculation. Fed Proc 1987;46:2397-401. 9. Dreyer W J, Smith CW, Michael LH, et al. Canine neutrophil activation by cardiac lymph obtained during reperfusion of ischemic myocardium. Circ Res 1989;65:1751-62.

Discussion Dr. Waiter E. Pae (Hershey, Pa.). The leukocyte depletion filters that are commercially available are designed to accommodate flows of 6 L/min or so, which probably affects their ability deplete leukocytes. Do you have any data to indicate to what level you must deplete the circulating leukocyte count to see an effect? Does the level of depletion correlate with the effects that are seen? Dr. Wilson. I think those questions are still unanswered. The work published by Engler and Covell in 1987, investigating the role of neutrophil-mediated tissue injury in a model of regional myocardial ischemia, showed that 10% of the normal circulating granulocyte count was sufficient to cause significant postischemic ventricular dysfunction. In this model of cardiac surgery we depleted the circulating granulocyte count to less than 1% of that in control animals during initial coronary reperfusion and demonstrated a reduction in postischemic ventricular dysfunction. However, whether a lesser degree of leukocyte depletion would achieve similar preservation of ventricular function is unknown.