Mechanical circulatory support in postoperative cardiogenic shock

Mechanical circulatory support in postoperative cardiogenic shockMechanical circulatory support in postoperative cardiogenic shock
A 38-year-old white woman had cardiogenic shock after elective mitral valve replacement and was unresponsive to pressor drugs and intra-aortic balloon counterpulsation. A left ventricular assist device (left ventricle to ascending aorta) was implanted 16 hours after the initial operation and provided circulatory support for 8 days. Improvement in the patient's own cardiac performance was documented, and there were no complications attributable to the assist device. However, intercurrent medical problems resulted in clinical deterioration on the fifth day after operation, and the patient died 8 days after operation. The findings in this patient suggest a potential role for this left ventricular assist device in future cases of acute, intractable, but potentially reversible myocardial failure.
Paulo Radvany, M.D., Michael Pine, M.D., Ronald Weintraub, M.D., Walter H. Abelmann, M.D., and William F. Bernhard, M.D., Boston, Mass.
Acute myocardial insults, whether from disease or related to cardiac surgery, frequently are associated with severe depression of myocardial function. Al- though the muscle injury may be potentially reversible, occasionally it is impossible to maintain sufficient car- diac function to sustain adequate tissue perfusion, and death occurs before myocardial recovery can take place. Use of intra-aortic balloon counterpulsation (lABP) has been highly effective in the patient with a cardiac index of 0.8 L. per minute per square meter or greater! who requires limited circulatory support. However, when IABP fails to achieve adequate tissue perfusion, a more effective pump assist is needed to satisfy tissue metabolic requirements.v' Such a device has been developed during the past decade and has undergone extensive evaluation in animals by Bernhard and co-workers'r !" prior to commencing clinical trials. The following case report describes use of this device
From the Cardiovascular Surgical Research Laboratory, Children's Hospital Medical Center, and the Departments of Medicine and Surgery, BethIsrael Hospital and HarvardMedical School, Bos- ton, Mass.
Supported in part by Contract No. I-HV-5-3007 and by Grants HL 10539 and HL 5909 from the National Heart, Lung and Blood Institute, National Institutesof Health, Bethesda, Md.
Received for publication May 12, 1977. Accepted for publication July 14, 1977.
Address for reprints: W. H. Abelmann, M.D., Cardiovascular Unit, BethIsraelHospital, 330 Brookline Ave., Boston, Mass. 02215.
in a patient with cardiogenic shock following mitral valve replacement.
Case report
A. B., a 38-year-old white woman, was admitted to the Beth Israel Hospital for elective mitral valve replacement. She had been hospitalized at 3 years of age with "arthritis," but after discharge she was active and totally asymptomatic. Pregnancies when she was 23 and 30 years of age resulted in normal deliveries without medical complications. Two years prior to operation, she noticed mild exertion fatigue and she was evaluated at the Beth Israel Hospital. Physical examina- tion was consistent with isolated mitral regurgitation. She also had mild systemic hypertension (150/100 mm. Hg), which was managed with a diuretic. Four months prior to operation, fatigue and shortness of breath developed upon mild exertion. Examination revealed atrial fibrillation and bibasilar inspiratory rales. She was treated with digoxin after attempts at cardioversion had failed.
Cardiac catheterization was performed 2 months prior to operation. Roentgenographic examination at that time re- vealed moderate left atrial and left ventricular enlargement but no evidence of congestive heart failure (Fig. I). Findings at catheterization included the following pressures (expressed in millimeters of mercury): aorta-l20/75; left ventricle- 120/12 (end diastolic); pulmonary artery wedge (mean)-18; pulmonary artery-34/18; right ventricle-34/l1 (end dia- stolic). Cardiac angiograms revealed moderate enlargement of the left atrium and mild enlargement of the left ventricle. There was 4+ mitral regurgitation. Left ventricular wall mo- tion appeared to be normal with a calculated ejection fraction of 55 percent. Forward flow (cardiac index) was 2.2 L. per minute per square meter by the Fick method, and total left ventricular output was 4.2 L. per minute per square meter by
97
Thoracic and Cardiovascular
Surgery
Fig. 1. Preoperative chest roentgenogram. Posteroanterior (A) and left lateral (B) views prior to operation reveal cardiomegaly and left atrial enlargement. The lungs appear to be normal.
Fig. 2. Hemodynamic evolution. Cardiac index (upper panel), mean left atrial pressure (middle panel), and systemic arterial pressure (lower panel) on and off left ventricular as- sist are shown. Note the fall of cardiac index and arterial pressure immediately after mitral valve replacement in spite of intra-aortic balloon counterpulsation (IABC). Left ventricular assist resulted in improvement of cardiac index and increase of arterial pressure, associated with reduction of the mean left atrial pressure. There was also gradual im- provement of intrinsic left ventricular performance.
20
''It NT1tJ :l:U~ o -61 0 2 4 6 8
OAYS POST MITRAL VALVE REPLACEMENT
cardiac angiography. Pulmonary vascular resistance was normal. The calculated mitral valve area was 1.2 sq. cm., assuming a mitral valve flow of 4.2 L. per minute per square meter. The patient was considered to have moderate mitral regurgitation, mild mitral stenosis, and good left ventricular function.
Two months after catheterization the patient underwent cardiac surgery. She was anesthetized with morphine. The left atrium was approached through the interatrial groove, and the regurgitant, thickened mitral valve was replaced with a Starr-Edwards No. 3M series 6120 prosthesis. The aorta was cross-clamped for 20 minutes. Initial attempts to wean the patient from cardiopulmonary bypass were unsuccessful be- cause the left ventricular developed pressure would not ex- ceed 60 mm. Hg. Upon the administration of both dopamine and isoproterenol, the patient's systolic blood pressure rose to 100 mm. Hg, with a cardiac index of 3.0 L. per minute per square meter and she was separated from bypass after a total of 97 minutes. During closure of the chest, systemic blood pressure fell and left atrial pressure rose. With the aid of a 20 ml. Avco intra-aortic balloon, the patient's systolic blood pressure stabilized at 90 to 100 mm. Hg with a cardiac index of 1.9 L. per minute per square meter. During the next 16 hours with the patient in the intensive care unit, counterpulsa- tion and administration of pressor drugs failed to prevent further deterioration of hemodynamic function. Cardiac out- put and systemic pressure fell, and left atrial pressure in- creased, as shown in Fig. 2.
Eighteen hours after mitral valve replacement, the patient was returned to the operating room in cardiogenic shock. The median sternotomy incision was opened quickly. The ventricles were barely contracting. The electrocardiogram re- vealed regular sinus rhythm. The systolic blood pressure was approximately 50 mm. Hg. Intravenous heparin was adminis- tered, and preparations were made rapidly to begin car- diopulmonary bypass prior to inserting a left ventricular assist
-On Assist DOff Assist
Mechanical circulatory support in cardiogenic shock 99
Fig. 3 . Left, The left ventricular assist device is shown with the curved metal inflow tube and attached conduit and the outflow valved conduit, both attached to the pump housing. Right, The assist device after implantation .
device (L VAD) . Catheters were inserted into the right atrium and the ascend ing aorta . The left ventricle was decompressed with an apical vent.
Components of the left ventricular assist device. The blood pump (Fig . 3) consists of a metal housing containing a cylindrical, polyurethane pump chamber (5 .0 cm . in diameter and 8 .7 em. in length) with a maximum stroke volume of 80 ml. The inflow and outflow conduits consist of two woven Dacron grafts, 18 mm. in diameter and 15 em. in length . Each one of these grafts contains a 15 mm . glutaraldehyde- preserved xenograft (porcine) valve sutured firmly in posi- tion. The proximal position of each conduit is protected by a series of metal rings imbedded in a layer of silicone rubber tubing and covered with an outer Dacron sleeve . The outer covering permits fixation of the tubes to the subcutaneous tissue and skin after positioning between the left ventricular apex and ascend ing aorta. The inflow tube assembl y permits insertion of metal tubes of varying diameters which can be affixed with a screw thread attachment. Polyester fibrils, 25 J.L in diameter and 250 J.L in length , line the polyurethane pump chamber and are held in place by multiple applications of Adiprene L-l 00 polyurethane. Thus a uniform blood-material interface is presented to circulating blood after the pump is assembled.
Both counterpulsation and fixed-rate pumping can be main- tained with a drive console which is connected to the pump housing by a pneumatic tube . For synchronization, an elec- trocardiographi c signal is processed through a QRS detector, which is routed via a time-delay circuit. After the delay inter- val, the pump is actuated if the heart rate is within certain specified limits. When the heart rate is either above or below
these limits , the device is driven by a fixed-rate timer. Loss of electrocardiographic trigger at any time results in the auto- matic onset of fixed-rate pumping.
A pneumatic monitor, consisting of a stroke-limiting trans - ducer, is integrated into the drive console. It is inserted into the pneumatic tube between the drive console and the pump, and it limits the stroke volume of the device to prevent appo- sition of the internal surfaces . In addition, the monitor is used as a transducer to permit an accurate determination of stroke volume. The unit is instrumented with a linear voltage dis- placement transducer and a pressure transducer; electronic calculations are made to determine the stroke volume, rate, and flow in a continuous fashion .
Pump implantation procedure. Following re-estab- lishment of total card iopulmonary bypas s, the outflow Dacron graft containing a xenograft valve was sutured to the ascending aorta proximal to the arterial inflow catheter from the pump oxygenator. This graft , which previously had been preclotted with donor blood, then was brought out through the right anterior chest wall by excising a 2.5 em . segment of the third rib.
Prior to placement of the inflow tube of the assist device, a circular silicone rubber ring with an attached Teflon sewing cuff was sutured to the apex of the left ventricle . Care was taken to avoid injury to small coronary arterial branches . A No. 24 balloon catheter was passed through the apical ring into the body of the left ventricle , and the balloon was inflated with 20 ml. of carbon dioxide gas. A circular knife then was used to exci se a small section of myocardium to allow inser- tion of the inflow tube and attached valved conduit into the sinus of the ventricle. Connection of the conduit to the pump
10 0 Radvany et al. The Journal of
Thoracic and Cardiovascular
.PAP I ./V'f - \. .11.
L I
I I I
Fig. 4. Hemodynamics on the first day and on Day 5 of ventricular assist. A, The recorded electrocardiographic strip shows atrial flutter at the top of the tracing; arterial pressure falls and mean left atrial (LA) pressure rises when the left ventricular assist device is turned off. Pulmonary artery (PA) pressure does not change. Arterial and left atrial pressure returned to previous values with resumption of left ventricular assist. B. Day 5 of left ventricular assist. The arterial pressure off assist is higher here than on Day I. indicating improvement of intrinsic myocardial function. The assist device is functioning asynchronously with the electrocardiogram. Paper speed was slowed in center of tracing.
housing was completed after filling the entire system with Ringer's lactate solution. A vent was inserted into the ascend- ing aorta to permit the escape of trapped air bubbles. The assist pump was activated gradually while cardiopulmonary bypass flow was reduced simultaneously. After several min- utes, the assist pump flow was increased to 4.0 L. per minute, and cardiopulmonary bypass was discontinued. Protamine sulfate was administered and anticoagulants were discon- tinued permanently. The chest wall was closed. However, excessive bleeding from the median sternotomy incision re- quired several additional hours in the operating room, and re-exploration was performed eventually with coagulation of several substernal bleeding points. The chest wall was closed again, and the patient was returned to the surgical intensive care unit.
Hemodynamic performance of the left ventricular assist device. After the patient's condition had stabilized in the intensive care unit, hemodynamic measurements were per- formed with and without left ventricular assist. These mea-
surements are illustrated in Figs. 2 and 4. Systemic arterial pressure was 110/55 mm. Hg, left atrial pressure was 10 mm. Hg, and cardiac index was 2.1 L. per minute per square meter. When the assist device was turned off momentarily, arterial pressure fell to 35/30 mm. Hg, and left atrial pressure rose abruptly to 19 mm. Hg. This small pulse pressure sug- gested minimal pulsatile flow. Representative recordings are shown in Fig. 4. For the next 7 days, a variety of supraven- tricular tachycardias (mainly atrial fibrillation and flutter) made it difficult to synchronize spontaneous myocardial con- traction with the pumping of the assist device. Despite this problem, hemodynamic improvement was impressive. Car- diac index stabilized at 2.0 L. per minute per square meter, and systemic arterial pressure remained improved (Fig. 2).
Fig. 4 illustrates improved left ventricular function on day 5 after insertion of the LVAD. Arterial blood pressure off the assist device rose to 90/40 mm. Hg from 35/30 mm. Hg immediately after implantation of the device. A rise in left atrial pressure from 12 to 16 mm. Hg off assist was smaller
Volume 75
Number 1
January, 1978 Mechanical circulatory support in cardiogenic shock 101
than the 9 mm. Hg rise in left atrial pressure seen in the immediate postinsertion tracing. Card iac index at the time of this recording was 1.8 L. per minute per square meter.
After operation the patient was ventilated mechanically with Ft02 below 40 percent. Arterial blood gases were nor- mal. Daily chest roentgenograms revealed mild pulmonary congestion . The patient required continued dopamine and isoproterenol along with large volumes of intravenous fluids to maintain the left atrial filling pressure in the range of 15 mm. Hg. Renal failure secondary to preoperative card iogenic shock was managed after operation by hemod ialysis . Despite severe azotem ia and jaundice, the patient remained conscious until the fourth day after operation and obeyed verbal com- mands . She lost consciousness after Day 5 and remained comatose until the LV AD was finally discontinued on the eighth day after operation.
Following the first of two periods of hemodialysis, platelet transfusions were administered for thrombocytopenia. He- matologic evaluation revealed a 50 percent prolongation over control of the prothrombin and partial thromboplastin times, but there was no evidence of disseminated intravascu- lar coagulation or microangiopathic anemia . A low-grade fever persisted throughout the hospital course, and repeated urine culture grew out Escherichia coli . Several tracheal cul- tures and two blood cultures also were positive for E. coli toward the end of the first week of circulatory support . On the eighth day after insertion of the LVAD , an attempt to wean the patient off the device failed . Systemic arterial pressure off assist was maintained transiently in the range of 70 to 90 mm . Hg, but gradually dropped to 40 to 50 mm. Hg. It was appar- ent that prolonged mechanical circulatory support would be required for this unconscious patient who, by this time, re- vealed evidence of sepsis along with continued renal fa ilure . Later on Day 8, ventricular fibrillation occurred , followed by electrical standstill, whereupon the LV AD was discontinued.
Autopsy findings. The heart weighed 510 grams with bilateral dilatation and biventricular hypertrophy (Fig . 5). The left ventricular cavity was small . The left ventricular assist device was in place, and suture lines were intact. The Starr-Edwards prosthesis appeared to protrude into the left ventricular outflow tract. Microscopic examination revealed scattered foci of biventricular and right atrial necrosis, esti- mated to be 12 hours to 7 days old (Fig. 6) . These findings are nonspecific and are most compatible with prolonged hypoten- sion. There were small foci of myocardial necrosis without evidence of severe localized damage . There was mild pulmo- nary congestion with patchy bilateral edema and atelectasis . An organizing pneumonitis was found in the left lower lobe. The abdomen contained about 4 L. of serous ascites . The liver showed passive congestion with severe centrilobular ne- crosis and moderate fatty degeneration . The spleen was en- larged, weighing 370 gram s. Examination of the brain re- vealed a massive left-sided intracerebral hemorrhage.
Discussion
The reason for development of cardiogenic shock following uneventful mitral valve replacement for rheumatic mitral regurgitation in this patient is not en - tirely clear. Neither the clinical nor hemodynamic evaluation carried out prior to operation suggested an
unusual operative risk. Congestive heart failure was
Fig. 5. The heart at autopsy . The left ventricle is open , re- vealing left ventricular hypertroph y and a relativel y small left ventricular cavity . There is a Starr-Edwards prosthesis in the mitral position . Arrows indicate the sites of implantation of the left ventricular assist device in the left -ventricular apex and in the ascending aorta.
mild, of recent onset, and probably was related to the development of atrial fibrillation. Left ventricular func- tion at catheterization appeared to be well preserved, with a left ventricular ejection fraction of 55 percent. Therefore, it is unlikely that this patient had the "burned-out" left ventricle occasionally associated with long-standing mitral regurgitation. Cardiopulmo- nary bypass time was relatively short , and no operative
difficulties were encountered. Neither postoperative electrocardiograms nor postmortem examination sug- gested an intraoperative myocardial infarction.
Despite the paucity of histologic findings, severe de- pression of myocardial function clearly was present when the heart was inspected immediately prior to in- sertion of the left ventricular assist device . The cardiac index at this time was less than 0.8 L. per minute per square meter, despite the use of an IABP device im- mediately following mitral valve replacement. The
pathogenesis of this depressed ventricular performance
10 2 Radvany et al. The Journal of
Thoracic and Cardiovascular Surgery
Fig. 6. Representative histologic section fromthe left ventricle. Patchy areasof myocardial necrosis are present. (Original magnification x75.)
is not completely clear, although a possible explanation might be left ventricular outflow tract obstruction by the cage of the Starr-Edwards ball valve. This hy- pothesis was suggested by autopsy findings of a rel- atively small left ventricular cavity and some narrow- ing of the outflow tract produced by secondary myocardial hypertrophy. However, left ventricular pressures were not measured immediately after inser- tion of the mitral prosthesis, so that the possibility of obstruction of left ventricular outflow accounting for the severe cardiogenic shock has been neither estab- lished nor ruled out. The clinical course of the patient following LVAD implantation with gradual improve- ment in myocardial function between the third and fifth day following assist device support is compatible with mechanical obstruction preventing the heart from re- suming satisfactory performance.
The LVAD employed in this clinical resuscitative effort was developed by Bernhard in conjunction with a research engineering group* under the auspices of the National Heart, Lung and Blood Institute, National Institutes of Health, Bethesda, Md. A cooperative clin- ical program has been established, involving five Bos- ton area hospitals, to provide a nucleus of patients un- dergoing cardiac surgery who might be benefited by this type of postoperative mechanical circulatory sup- port. A requirement for use of the device is the failure of all standard methods of cardiac resuscitation, includ- ing the use of pharmacologic agents and IABP. In addi- tion to these prerequisites, selection of patients follows *Thermo Electron Engineering Corp ., Waltham, Mass.
a carefully constructed clinical protocol agreed upon by the hospitals…