Dynamic cardiomyoplasty at seven years

Dynamic cardiomyoplasty at seven years Since January 1985, the date of the first dynamic cardiomyoplasty, until April 1992, 52 patients with end-stage heart disease were operated on in our institution. Mean preoperative New York Heart Association functional class was 3.3 and ventricular ejection fraction 16 % ± 3 %. Associated procedures in 23 patients comprised ventricular aneurysm resection (10), valve surgery (9), coronary artery bypass (8), and tumor resection (3). Thirty-eight patients had a ventricular reinforcement, 13 a ventricular substitution, and 1 an atrial reinforcement using the left latissimus dorsi muscle. Preassist mortality rate before full latissimus dorsi muscle stimulation was 7 of 13 patients (54 %) in the 1985 to 1987 period and 5 of 39 (12 %) in the 1988 to 1992 period. The causes of death were heart failure (4), multiorgan failure (4), septicemia (2~ ventricular fibrillation (1), and sudden death (1). Multivariate analysis of factors influencing hospital mortality showed that age, cardiac suture technique, associated surgical procedures, biventricular heart failure, and hemodynamic instability plus inotropic drug support were predictors of unfavorable outcome. All patients were followed up for from 2 months to 7 years (mean 21 months). Postassist mortality rate was 8 of 40 (20%). Causes of death included heart failure (5), ventricular fibrillation (1), myocardial infarction (1), and gastric bleeding (1). Preoperative risk factors influencing long-term mortality were permanent New York Heart Association functional class IV, biventricular heart failure, atrial fibrillation, cardiothoracic ratio greater than 60 %, and ejection fraction less than 15%. Actuarial survival at 7 years was 70.4% (preassist mortality excluded). Surviving patients were in a mean New York Heart Association functional class of 1.8 (preoperatively 3.3, p < 0.05). The average ejection fractions (rest/stress) were 25% /28% at 1 year, 26% / 30% at 2 years, and 23%/28% at 3 years. Average cardiothoracic ratios were 57% ± 3% at 1 year, 56 % ± 2 % at 2 years, and 57 % ± 2.5 % at 3 years. Catheterization obtained in 20 patients showed no significant changes at rest in capillary wedge pressure, pulmonary artery pressure, and diastolic left ventricular pressure when compared with preoperative pressures. Average ejection fractions increased from 24 % to 30.6 %. Maximal oxygen consumption increased from 12.8 ± 3.5 to 18.6 ± 4 ml/min per kilogram. The number of rehospitalizations resulting from congestive heart failure was reduced to 0.4 hospitalizations per patient per year (preoperatively 2.4, p < 0.05). In 62 % of the patients, pharmacologic therapy was diminished after the operation. Three patients required orthotopic heart transplantation 6 months, 4 years, and 5 years after cardiomyoplasty. All are alive. (J THORAC CARDIOVASC SURG 1993;106:42-54)

Alain Carpentier, MD, PhD, Juan Carlos Chachques, MD (by invitation), Christophe Acar, MD (by invitation), John Reiland, MD (by invitation), Serban Mihaileanu, MD (by invitation), Danielle Bensasson, MD (by invitation), Jean-Philippe Kieffer, MD (by invitation), Pierre Guibourt, MD (by invitation), Dominique Tournay, MD (by invitation), Isabelle Roussin, MD (by invitation), and Pierre A. Grandjean, MS (by invitation), Paris, France

From Hopital Broussais, Paris, France. Read at the Seventy-second Annual Meeting of The American Association for Thoracic Surgery, Los Angeles,Calif., April 26-29, 1992. Address for reprints: Alain Carpentier, MD, PhD, Hopital Broussais, Clinique Rene Leriche, 96, rue Didot, 75764 Paris Cedex 14, France. Copyright

1993 by Mosby-Year Book, Inc.

0022-5223/93 $1.00 + .10

42

12/6/45059

During the past 7 years, a new operation was developed in our institution, which consists in the reinforcement of cardiac contraction by means of an electrically stimulated skeletal muscle for patients with end-stage cardiac insufficiency. The concept itself was not new and numerous experimental works on cardiac reinforcement and muscular stimulation had been done before, but its prac-

The Journal of Thoracic and Cardiovascular Surgery Volume 106, Number 1

Carpentier et at.

Table I. Clinical status ofpatients before

Table II. Associated disease (contraindications to

cardio~yoplasty

heart transplantation)

Functional class IV III

Left heart failure Right heart failure Biventricular failure Cardiothoracic ratio (%) Indexed left ventricular end-diastolic volume

34

5 13 56 ± 6 178 ± 31

(ml/rrr')

Maximal oxygen consumption (rnl/rnin per kilogram) Left ventricular ejection fraction (at rest) (%)

Insulin-dependent diabetes Pulmonaryartery hypertension Respiratory insufficiency Gastric ulcer Renal insufficiency Cancer

23 29

12.8 ± 3.5

Patients From January 1985 to April 1992, 52 patients with severe heart failure were operated on at Hopital Broussais in Paris. Their ages ranged from 15 years to 68 years (mean 50 years). There were 44 male and 8 female patients. Etiology. The two main causes of heart failure were chronic myocardial ischemia and dilated cardiomyopathy. Myocardial ischemia. The 33 patients with myocardial ischemia had had several myocardial infarctions and 4 had had previous coronary artery revascularization. All were considered to be beyond the possibility of efficient myocardial revascularization at the time of cardiomyoplasty. Ten patients, however, had one severely obstructed coronary artery suitable for bypass which, in the absence of pain, raised the question of the need for revascularization. Twenty-four patients had a global dilatation of the heart and 10 had an extensive anteroapical left ventricular aneurysm. Three patients had mitral valve insufficiency and I had mitral and tricuspid valve insufficiency. Dilated cardiomyopathy. Fifteen patients had a dilated cardiomyopathy, idiopathic in 13 and of valvular origin in 2. Among the 13 patients with a dilated cardiomyopathy, 12 had a predominant left ventricular cardiomyopathy and I patient aged 15 had an isolated right ventricular cardiomyopathy. Cardiactumor. Three patients had a cardiac tumor. The first patient had an extensive fibroma involving the posterior wall of the right ventricle and the diaphragm. The second had a lymphoma of the right side of the heart involving the right atrium,

10

10 7 5 3

2

Table III. Surgical modalities

16 ± 3

tical realization had been postponed because of fatigue of the muscle whenever stimulated continuously at heart rate. This drawback was overcome in our laboratory by the development of an original protocol of progressive sequential stimulation, I which, after experimental evaluation;' led to the first clinical application in January 1985. 3 As of April 1992, a total of 52 cardiomyoplasties have been performed at Hopital Broussais and an additional 130 worldwide. The purpose of this paper is to report our 7-year experience with particular attention to surgical technique, midterm results, and clinical indications.

43

Reinforcement technique Substitution technique Anteriorwrap Posterior wrap Atrial wrap Extracorporeal circulation Intraaortic balloon counterpulsation

No.

%

39 13

75 25 9.5 88.5

5 46 I

2

20

38.5 34.5

18

the tricuspid valve, and the anterior wall of the right ventricle. The third patient had a fibroma involving the anterior wall and the infundibulum of the right ventricle, a portion of the tricuspid valve, and one cusp of the pulmonary valve. Congenital malformation. One patient, aged 27, had elevated pulmonary resistances and severe venous and liver congestion, which had developed 6 years after a Fontan operation done to correct single ventricle with pulmonary artery stenosis. Clinical status. All patients but I were in New York Heart Association (NYHA) functional class III or IV before the operation (mean 3.3) (Table I). Most of the patients had been hospitalized repeatedly for congestive heart failure with an average of 2.4 episodes per patient per year. Twelve of these patients had one or several episodes of pulmonary edema, reversible with maximal medical therapy, in the year preceding the operation. Forty-eight patients were in sinus rhythm and the remaining 4 were in atrial fibrillation. Ventricular extrasystoles were present in 24 patients among whom 5 had several episodes of ventricular tachycardia. Other clinical or functional data are listed in Table I. Mitral valve incompetence was present in six patients, severe in two and mild to moderate in four. The functional mechanism was mitral valve incompetence with normal leaflet motion (type I) in one patient, leaflet prolapse (type II) in two patients, and restricted leaflet motion (type III) in three. Tricuspid valve insufficiency was present in five patients. The functional mechanism involved was normal leaflet motion (type I) in one patient and restricted leaflet motion (type III) in two patients. In the remaining two patients, tricuspid insufficiency resulted from the involvement of the leaflet tissue by a cardiac tumor. Associated disease and contraindications to heart transplantation. Associated noncardiac diseases were present in the majority of these patients and precluded heart transplantation (Table II). The two patients with malignant disease (Hodgkin's disease and renal carcinoma) had been treated in the years preceding cardiomyoplasty and were considered cured or in remis-

sron.

44

The Journal of Thoracic and Cardiovascular Surgery July 1993

Carpentier et al.

J

Fig. 1. Posterior wrap-the "flap-sliding technique." Left, Two long clamps hold two stay sutures placed on the leading edge of the flap. Posterior and medial displacement of these clamps allows the free edge of the muscle to slide below the apex and then behind the left ventricle without the heart being mobilized. Right, The two stay sutures are secured to the pericardial sac, the wrap is completed, and the pericardial flap (tailored at the beginning of the operation) is sutured to the free edge of the muscle flap so as to maintain its proper position.

Thirty-six patients (69.2%) had one or several contraindications to cardiac transplantation: medical in 22, tumors in 5, age over 65 years in 4, and socioeconomic or geographic in 6. Among the 16 patients who had no contraindication to cardiac transplantation, II refused transplantation and 5 were judged to have a lower risk associated with a cardiomyoplasty than with cardiac transplantation (ejection fraction> 20% in 4 patients and isolated right ventricular failure in I). None of these 5 patients died.

Surgical technique Cardiomyoplasty procedure. Surgical modalities are summarized in Table III. The technical procedure described earlier" has benefited from the following modifications to reduce the operative complications and to increase the efficiency of the operation. I. Extracorporeal circulation was avoided whenever possible to decrease the amount of bleeding, the risk of infection, and the difficulty of weaning the patient from bypass. Of 52 patients, only 20 were operated on with the aid of extracorporeal circulation, most of them in the beginning of our experience.

2. The resection of the second rib was extended anteriorly and the window enlarged to four fingerbreaths so as to avoid compression injury of the muscle flap. 3. The superficial surface of the muscle was placed in contact with the epicardium in 38 patients and the deep fascia in contact with epicardium in 14 patients. The choice was guided by the shape of the flap; we selected the position that provided the largest covering without distortion of the neurovascular pedicle. 4. Posterior wrap or clockwise rotation (i.e., the muscle flap placed posteriorly in contact with the left ventricle) was used in 46 cases. The positioning of the flap has been improved by the following "flap-sliding technique" so as to further reduce heart mobilization (Fig. I). The two stay sutures placed on the leading edge of the muscle flap are used to slide the flap behind the heart. This sliding is facilitated by holding the two sutures at their origin on the muscle with two long clamps. Posterior and medial displacement of these clamps allows positioning of the free edge of the muscle, first below the apex and then behind the left ventricle. The distal stay suture is secured to the pericardium in front of the inferior vena cava and the proximal suture is secured to the remnant ofthe pericardium at the levelof the left atrial appendage.

The Journ al of Thoraci c and Cardiovascular Surgery Volume 106, Number 1

5. .A nteri?r wrap was prefer red when ever predominant right ventricular insufficiency was present (t hree cases) or in right hea rt reconst ruction after tu mor resect ion (two cases). T he techn ique of a nterio r wrapping! has bee n recentl y modi fied as follows (Fig. 2) : the muscle is folded longitudinally so th at two layers of mu scle are applied to the right ventricle an d th e extremity of t he mu scle is sutured to the perica rdial sac as far posteriorly as possible, that is, at th e j unction between th e diaphrag mat ic a nd t he poste rior par ts of th e pericardium. Th is orienta tion a llows coverage of th e entire righ t ventricle a nd the diap hra gmatic wall of the left ventricle. 6. A lria~ wrap was used in a pa tient in whom elevated pulmonar y resistanc es developed after a Fontan opera tion. Th e two atria were d issected and separated, a homograft valve was impla nted in the inferior vena cava, a nd the left la tissimus dorsi was used to wrap the right a trium according to a technique previously described.> 6 7. Th e second sensing electrode is no longer used whenever the first electrode provides adequate signal collection, tha t is, a t least 5 mY amplitude and 0.5 Y/sec slew rate. 8. In cases of left ventricular a neurysm resect ion or reconstruction, our current policy is to prefer the technique proposed by Magovern a nd associates," 8 to recon struct th e ventricl e with a fla p of fibrous tissue ta ken from the a neurysm itself provided that the und erlying end oth elium was intact with no th rombus forma tion. In ca ses of tumor resect ion, the myocardial loss of substa nce was cor rected by with a patch of glutaraldehyde treated a utologous pericard ium reinfor ced by th e m uscle flap.? 9. Int raaortic balloon coun terpulsa tion was used in 18 pa tients. It is now used systematica lly immediately a fter a nesthesia whenever one or several risk facto rs were present or ca rdiotonic drugs were needed. 10. The subcutaneous tissue was applied to the chest wall with multiple tacking sutures to decrease the amo unt of fluid in the dead space left by the tra nsferred lat issimu s dorsi mu scle. In addition, a compressive dressing was left in place for 10 days aft er the operation. Associated surgical procedure and coronary artery revas-

cu/arization. Coronety artery revascularizationwas carried out in eleven patients. In twopatients with a tumor ofthe right ventricle. the posterior descending artery wasrevascu!arized after resection of t he right coronary artery and the right ventricular wall affec ted by t he tum or. In pa tients with ischemic hea rt d iseases, at the beginning of our expe rience, revasculari zat ion was thought to be unnecessar y because the pati ents had no residual angina a nd becau se the visible sten ot ic coronary a rteries were in infarcted a reas . Thi s policy proved to be wrong beca use two pat ients died of recurrent myocardial infar ction 6 months and 2 years a fter the ope ration. Revasculari zat ion was systemati ca lly und ertak en thereafter in nine patients with accessible obstructed coronary a rteries. R evascular ization was perform ed either before the ca rdio myoplasty procedure in three patients or during this procedure in six patients. Th e techniques used were tran scutaneous balloon dilation (t hree procedures in two patients) a nd/or corona ry a rte ry bypass gra fting (eight procedures in eight pat ients). Co ronary a rtery bypass gra fting without extracorporeal circulat ion was preferred whenever possible because of th e frequent difficulty of wea ning these very sick pat ients from ca rdiopulmona ry bypass. Th is was possible in one patient with a gra ft to th e right coron ary a rtery that could be easily dissected a nd imm obilized a nd in a nother patie nt with a graft to the anterior descend ing artery becau se of the dec reas ed

Carpentier et al. 4 5

/

/ I

(

\

I

Fig. 2. A nterior wra p-the "t wo-layer technique." The m us-

cle flap is foldedlongitudinally and its distal end is secured to the pericardia! sac as far posterior as possible at the junction between the diaphragmatic and the posterior parts of the perica rdium. anterior wall mot ion. In thi s case, th e coronary artery was widely dissected and surro unded by two ta pes. Aft er heparin was given, a U-s ha ped arterial clamp was ap plied th at occluded th e artery and stabil ized the a nastomotic site. Under these conditions, a bypass gra ft could be constructed eas ily. Muscle stimulation. Th e protocol of progressive seq uent ial stim ulation, as described in previous papers, 1, 5 was used in all pa tient s. The stim ulato r used was th e M edtronic SP 1005 device ( Medtronic, Inc., M inneapolis, Minn.). After a 2-week period of no stimul ation (the time necessar y for the mu scle to adhere to th e heart), progressive sequential stimulation was und erta ken which after 6 weeks culminated in full stimula tion with tr ain s of impulses a nd maximum pu lses a mplitude (4 Y ± 0.6). The choice of the mod e (I : I or 1:2 mod e) depended on each indiv idua l pat ient and was based on pa tient comfor t and hemodynamic respon se as evalu ated by Doppler ec hoca rdiogra phy. Doppler echoca rdiogra phy also served to assess th e proper tim ing of stim ulation with regard to mitral va lve closur e. It was found

/

4 6 Carpentier et ai.

Table IV. Preoperative risk factors influencing hospital and late mortality (multivariate analysis) Influence hospital mortality Clinical factors Age >65 years Filling pressures >45 mm Hg Patients receiving preoperative inotropic drug support Biventricular heart failure Pulmonary vascular hypertension Surgical factors Cardiac suture technique Cardiopulmonary bypass Associated valve procedures Influence late mortality Permanent NYHA functional class IV Cardiothoracic ratio >0.70 Left ventricular ejection fraction < 15% Biventricular heart failure Pulmonary vascular hypertension Atrial fibrillation

essential that the muscle contract only after mitral valve closure. 10. I I Results Results were divided into two periods: (I) the preassist period (first 2 months), which was strongly influenced by the patient's preoperative condition and his capacity to recover from the operation without muscle assistance, and (2) the assist period, which reflected the efficiency of the procedure and the evolution of the underlying heart disease. Preassist period Mortality. Twelve patients (23%) died during the preassist period despite pharmacologic support in all and intraaortic balloon counterpulsation in nine. The causes of death were cardiac failure in three, myocardial infarction in one, ventricular tachycardia in two, multiorgan failure in four, and sepsis in two. Multivariate analysis of preoperative risk factors for hospital mortality showed that age over 65 years, cardiac suture technique, associated mitral valve procedure, pulmonary artery hypertension, and severe hemodynamic instability were predictors of unfavorable outcome (Table IV). All of the patients with a cardiac index of less than 1.3 L/min per square meter, a capillary wedge pressure over 40 mm Hg, a left ventricular end-diastolic pressure over 50 mm Hg, or a combination of these conditions died during the preassist period. Taking into consideration these risk factors led to an improved selection of the patients and a reduced preassist mortality in the past 4 years: 12.8% (5/39) in the 1988 to 1992 period versus 54% (7/13) in the 1985 to 1987 period.

The Journal of Thoracic and Cardiovascular Surgery July 1993

Complications. Most of the patients had a difficult postoperative course because of their severe preoperative condition. Only 10 patients remained in the intensive care unit for fewer than 3 days and 25 for fewer than 6 days. Two patients remained in the intensive care unit for 20 and 45 days, respectively. Low cardiac output, present in 16 patients (30%), was by far the most frequent complication. In 10 patients prolonged pharmacologic support and balloon counterpulsation proved effective. Eight patients died of either cardiac failure (4 patients) or multiorgan failure (4 patients). Among the 4 patients who died of cardiac failure, I had myocardial infarction and 2 had necrosis of the latissimus dorsi at autopsy. All patients displayed some episodes of arrhythmia during the operation. Ventricular ectopic beats were the most commonly encountered arrhythmia. Severe ventricular arrhythmias were usually associated with hypoxemia. In one case, irreversible ventricular fibrillation led to patient death in the operating room. The development of the "noncardiac suture" technique in 19885 and the "flap sliding maneuver" more recently have been associated with a reduced risk of arrhythmia. Renal failure, encountered in I0 patients, was caused by poor preoperative condition, low cardiac output, or infection. Four patients required renal dialysis. Six other patients had elevated creatinine concentrations and were given dopamine, 3 rug/kg per minute. Infection was present in 5 of 52 patients (9.5%). Two patients died of septicemia. Both were in functional class IV; one had an extensive tumor of the heart and the other had renal insufficiency before the operation. In the 3 remaining patients, infection was localized to the area of dissection of the latissimus dorsi (I case) or to the pocket of the stimulator (2 cases). All were successfully treated by debridement, and long-term antibiotics. One patient with infection of the pocket of the stimulator had the stimulator removed and subsequently received a heart transplant. The other patient did require removal of the stimulator and its subsequent replacement. Four patients had partial necrosis of the skin in the area of dissection of the latissimus dorsi muscle, for which they were successfully treated by skin resection and suture. Three patients treated by intraaortic balloon counterpulsation had an ischemia ofone leg. One of these patients eventually died. Postassist period. Forty patients were followed up for from 2 months to 7 years (mean 3 years, 6 months). Late mortality. During this period of follow-up, 8 of 40 patients (20%) died. The causes of death were cardiac failure in 5, ventricular fibrillation in I, gastric bleeding in I, and myocardial infarction in I. Among the 5 patients

The Journal of Thoracic and Cardiovascular Surgery Volume 106, Number 1

Carpentier et al.

47

% ALIVE

1()()~\'-. 80

60

40

'.

,

,

,

....--_---tlt---.... - . - .. - - - -.. 70.4%

,, ,,

,

,

20 40 21 11 3 7 0 + - - - - , - -.....- - , - - - , - - . . , . - - - - , - -.....---1 2 4 o 3 5 6 7 8 YEARS

Fig. 3. Actuarial survival in 40 patients.Hospitalmortalitywasexcluded becauseit variedconsiderably during the time frame of this study. Taking into account the 12.8% hospital mortality of the 1988 to 1992 period would still make a very significant difference with medically treated patients (dotted line)." who died of cardiac failure, 4 had ischemic heart disease and I had a dilated cardiomyopathy with diabetes and infection. All 5 had an ejection fraction below 14% and 2 had biventricular failure. The patient who died of ventricular fibrillation had a dilated cardiomyopathy. He was in functional class IV at the time of the operation with an ejection fraction of9% and showed only limited improvement during the 2 years after cardiomyoplasty. Multivariate analysis of preoperative risk factors showed that functional class IV, cardiothoracic ratio greater than 0.70, left ventricular ejection fraction less than 15%, biventricular failure, and atrial fibrillation were predictors of unfavorable outcome (Table IV). Excluding hospital mortality, the actuarial survival at 5 years was 70.4% (Fig. 3). Complications. In the 7-year period of follow-up ofthis series, the following complications were encountered. Cardiac failure was encountered in 10 patients (25.6%). The causes were onset of atrial fibrillation in 2, premature battery depletion in 2, battery removal for infection in 1, muscular wasting after hepatitis in 1, and pulmonary insufficiency and infection in I. In the remaining 3 patients, cardiac failure was linked to the continuous evolution of the underlying disease: two patients were in class IV at the time of operation with an ejection fraction below 14% and the third patient had a recurrent myocardial infarction. Among the 10 patients with cardiac failure, 5 died and 3 underwent successful transplantation at 6 months, 4 years, and 5 years after cardiomyoplasty. The first patient

Table V. Functional improvement in 29 patients surviving cardiomyoplasty and not undergoing heart transplantation Preop.

Functional class (mean) Exercise capacity (W) Hospitalizations (No. per patient peryear)

p Value

Postop. 1.6

3.3

p<0.05 96 ± 18

60 ± 24

p
Oxygen consumption (rnl/min 12.8 ± 3.5 per kilogram)

0.4

p<0.05 18.6 ± 4

p<0.05

had severe postoperative hepatitis with a 20% loss of body weight and severe muscular wasting. The second patient had the cardiomyostimulator removed for infection, and the third had atrial fibrillation that further aggravated his condition. Finally, the 2 patients with premature battery depletion were rehospitalized for severe cardiac failure and pulmonary edema. The condition of both improved dramatically after the stimulator had been changed. Except for the 2 patients described earlier, battery depletion was usually progressive and marked by shortness of breath, fatigue, and peripheral edema. The battery was changed in 7 patients after an average duration of function of 3 years and 2 months. Finally, 4 patients had cholecystitis and underwent successful cholecystectomy.

The Journal of Thoracic and Cardiovascular Surgery July 1993

4 8 Carpentier et af.

80 70 60 50 40 30

20 10 0

1 YEAR

PRE·OP

2 YEARS

3 YEARS

Fig . 4. Cardiothoracic ratio over the years of evolution shows stabilization of process of ventricular dilation. LVEF 40

(%)

~---------------, 30

28

30

20

10

TESTTIME

o PRE-oP (23)

•

1YEAR(23)

2 YEARS (12)

3 YEARS (7)

At re st At exercise

Fig. 5. Left ventricular ejection fraction (L VEF) (radioisotopic) at 1,2, and 3 years.

Functional improvement. Among the 32 survivors, the 29 who had not undergone transplantation displayed a significant functio nal improvement (Table V). All but 3 improved from functional class III and IV to class lor II . T he other 3 were in class III . T he mean NYHA class improved from 3.3 before the operation to 1.6 afterward (p < 0.05) . In most patients, improvement continued over a period of 6 months after the operation and up to 1 year in 2 patients. All but 3 patients returned toan "enjoyable" social life. Functional improvement was associated with increased exercise capacity, significant reduction in the need for drugs in 62% of the patients, and reduction of the number of episodes of hospitalizations for congestive

heart failure (Table V). With the exception of the 3 patients who required a cardiac tra nsplantation, the functional improvement persisted throughout the years within the time frame of this stud y. Noninvasive investigations. Cardiothoracic ratio at 1, 2, and 3 years showed stabi lization of heart size (Fig. 4). The mea n postoperative ejection fraction (multiple gated acquisition scan) showed an improvement over the preoperati ve ejection fraction (25% versus 16%) and was sta ble over years. Ejection fraction showed some improvement during exercise (Fig . 5). Hemodynamic evaluation by aortic flow Doppler study with the cardiomyostimulator turned off and on was performed in the 21 patients

The Journal of Thoracic and Cardiovascular Surgery Volume 106, Number 1

Carpentier et at.

Table VI. Hemodynamic evaluation by aortic flow Doppler study (cardiomyostimulator off/on comparison) Follow-up

No. of patients No. of patients with Doppler improvements Left ventricular stroke volume (% increase) Cardiac output (% increase)

2 yr

3 yr

21 15(71%)

11 7 (63%)

+27 ± 8

+22 ± 4

+12 ± 3

+10 ± 2

p< 0.05

p<0.05

survivmg more than 2 years after the operation and showed a significant improvement in 71% of them (Table VI). Catheterization. Catheterization obtained in 22 patients followed up for more than 2 years showed no significant changes in capillary wedge pressure, pulmonary artery pressure, and left ventricular end-diastolic pressure when compared with preoperative values. Postoperative left ventricular end-diastolic volume showed no significant change. However, the cardiac index and ejection fraction increased significantly (Table VII). Discussion End-stage heart failure is a major concern in cardiology because of the severity of its prognosis and its prevalence in Western countries. Actuarial survival at 3 years isonly 20%.12 Heart transplantation is efficient in patients with end-stage heart failure. This technique, however, is limited by shortage of donors, contraindications, and complications. Cardiomyoplasty has the advantage of using the patient's own living tissues and thereby avoiding rejection and related problems. The hemodynamic effect of this technique has been demonstrated in shortterm experiments: Skeletal muscle wrapped around the heart was able to enhance heart performance'
49

Table VII. Hemodynamic data (n == 20) Preop. MPAP (mm Hg) MPWP (mm Hg) L VEDP (mm Hg) CI (Lymiri/rrr')

LVEF (%) L VEDV (ml/m 2)

24A ± 6.2 15.8 ± 2.8 20J 2.38 24 178

± 4.2 ± Al ± 6.2

± 31

Postop. (2 yr) 26.2 17.6 18.7 2.87 30.6 186

Significance

NS NS NS

± 4 ± 3.1

± 2.6 ± .63 ± 5 ± 36

p

= 0.049

P = 0.043

NS

indexed MPAP. Mean pulmonary artery pressure; MPWP, mean pulmonary wedge pressure; LVEDP, left ventricular end-diastolicpressure; CI, cardiac index; LVEF, left ventricular ejection fraction; LVEDV, left ventricular end-diastolic volume; NS, not significant.

patients with various cardiac diseases: tumor, aneurysm, ischemic cardiomyopathy, dilated cardiomyopathy, and Chagas disease.v ?-11, 19,20 Results have been encouraging so far, although extreme variations were found because of the heterogeneity ofthe patient population and of the techniques used. Therefore numerous questions remain to be addressed. Surgical technique. The surgical technique has evolved throughout the years with the aim of reducing preassist mortality and increasing the efficiency of the procedure. In the past 4 years, the use of the "noncardiac suture" technique' and the "flap-sliding maneuver" described in this article effectively reduced the operative mortality to 12%. There are, however, still controversial technical points that must be discussed. Muscle orientation. Recent experimental studies have shown that an anterior positioning with a counterclockwise wrapping of the flap can be more efficient," but in patients who have an enlarged heart the anterior positioning of the muscle flap does not allow the left ventricle to be completely covered. A pragmatic approach must be preferred in clinical practice, taking into consideration the type of ventricular failure, the size of the heart, the size of the flap, and the patient's condition. In predominant left heart failure, the flap must be positioned posteriorly in contact with the left ventricle. In predominant right heart failure the flap is positioned anteriorly. Enlarged heart size is a serious limitation to muscle wrapping, particularly when the muscle flap is small. Partial excision ofthe ventricle to reduce its size has not produced good results in our experience. Poor results were also observed whenever we tried to reduce the size of the heart by excess tension on the muscle flap. It seems that severely dilated ventricles have adapted themselves to a given wall tension that cannot be reduced without impairing the residual contractility. Finally, the patient's condition may also playa role in the choice of type of

5 0 Carpentier et al.

wrap: an anterior wrap is better tolerated than a posterior wrapping. It is unfortunate that, because of the short neurovascular pedicle, only the distal half of the flap, the less efficient, can be used for wrapping. The use of a free flap with revascularization by the internal mammary vessels would be an ideal solution to this problem, but experimental section of the nerve has always led to progressive dystrophy of the muscle. Muscle stimulation and ventricular compliance. Although some hemodynamic benefits have been obtained with single pulse muscle stimulation," the protocol of progressive sequential stimulation" and pulse trains'? are preferable for muscle transformation and cardiac assistance. The interval between the sensed QRS complex and each pulse train is of paramount importance. Both experimental and clinical data have shown that muscle contraction may affect ventricular filling when no delay was present between the QRS complex and the pulse train.P Echocardiography must be used to adjust this delay to provide optimal synchronization between muscle contraction, ventricular systole, and mitral valve closure. Tachycardia is another factor that may impair ventricular filling because in the current model of cardiomyostimulator (SP 1005) the pulse train has a fixed duration (185 msec). For this reason, the stimulator automatically increases the heart muscle contraction ratio (2: 1,3: 1,4: 1) with increasing heart rates, which results in assisting only one systole over two or three or four during exercise. The heart muscle contraction ratio must be adapted to each patient according to functional improvement. Most of the time the optimal ratio is 2:1.4 Whenever synchronization between heart and muscle contraction is adequate, the muscle does not impair the ventricular compliance because of the greater wall tension of the dilated ventricles. Survival. This study demonstrates that cardiomyoplasty improves long-term survival. The 70% actuarial survival at 5 years contrasts with the 20% survival at 3 years in medically treated patients.U Improved survival in patients with cardiomyoplasty was also demonstrated by Moreira and colleagues.P comparing medically treated patients with cardiomyoplasty-treated patients. The 2-year survival was 80% in the cardiomyoplasty group and 30% in the medically treated group. Survival can be improved further by taking into consideration the risk factors of the preassist and the postassist periods. Whether the postoperative mortality can be reduced by conditioning the muscle some weeks before the cardiomyoplasty procedure so as to benefit from full muscle contraction immediately after the operation was open to question. However, a 2-week period of nonstimulation is

The Journal of Thoracic and Cardiovascular Surgery July 1993

necessary after the operation, the time necessary for the muscle to adhere to the heart and to resorb the edema after muscle dissection and transfer into the thorax. During this 2-week period, most of the transformed slowtwitch fibers unfortunately return to their original fasttwitch condition and glycolytic metabolism. Functional benefit. In addition to improved survival, a significant functional improvement was found in the great majority of the patients. Functional class decreased from an average 3.3 before the operation to 1.8 afterward. Exercise capacity was generally improved, the number of hospitalizations per patient per year was reduced from 2.4 to 0.4, and in 62% of the patients the need for drugs was significantly reduced. Similar improvements have been described byothers.l ': 23, 24 Also of interest is the fact that functional improvement was progressive over a period of up to 1 year, which shows the capacity of the muscle to adapt to its new function. Functional improvement varied, however, from patient to patient, the least improvement being seen in patients with preoperative biventricular failure, excessive cardiac enlargement, and ejection fraction below 15%. Hemodynamics. Most disappointing was the fact that functional improvement did not always imply hemodynamic improvement. Ejection fraction did increase in the majority of the patients (16% before the operation versus 25% afterward), but no significant changes in filling pressures could be detected in the majority of the patients in this series. In a recent report, however, Jatene and associates!' reported a significant reduction of the mean pulmonary artery pressure (from 37.4 to 26.6 mm Hg) and of the mean pulmonary wedge pressure (from 24.8 to 16.6 mm Hg) 12 months after the operation. Patients with greater left ventricular dimensions had lesser improvements in left ventricular ejection fraction. More precisely, significant improvement in left ventricular function was observed in patients whose left ventricular diastolic diameter was below 70 mm whereas no change was detected in patients with a ventricular diastolic diameter greater than 75 mm. Both in Jatene's experience and in our experience, left ventricular dimensions throughout the years showed a stabilization in the process of ventricular dilatation in the majority of the patients. Mechanism of enhanced heart function. The improved ejection fraction observed in most cases, although limited, demonstrates that the functional improvement results from enhanced heart function. A placebo effect can therefore be ruled out. This is confirmed by the functional impairment after battery depletion observed in two patients and its complete reversibility after a new battery was implanted. Enhanced heart function involves several mechanisms.

The Journal of Thoracic and Cardiovascular Surgery Volume 106, Number 1

Active reinforcement. The heart and the wrapped skeletal muscle behave like a cybernetic system with two interactive components: the heart governing the muscle and the muscle reinforcing heart contraction. The efficiency of the system depends on the mechanical characteristics of these two components and the synchronization of their action. According to Laplace's law, the larger the ventricles the greater the tension that must be generated to produce a given pressure. Indeed, greater left ventricular dimension presented lesser improvements in left ventricular ejection fraction. I I Active reinforcement depends also on the size of the muscle and the amount of wrapping. The power of the transformed muscle has been calculated to be 40 W jkg, identical to the power developed by cardiac muscle during systolic contraction.P It is unfortunate that only a portion of this force is used because only a part of the muscle can be wrapped around the heart so as to prevent excessive tension on the neurovascular pedicle. Even though the free portion of the graft may also participate to some extent in cardiac assistance by traction and elevation of the heart during contraction, more than 50% of the force of the muscle is lost. A technique enabling the neurovascular pedicle to be elongated needs to be developed. Passive reinforcement. By its intrinsic tension, the muscle prevents further dilation of the ventricles and can restore more orderly cardiac contraction. In our experience, a stabilization of the process of ventricular dilation could be documented as well as reduction in the paradoxic motion of the ventricular septum. Even in the case of passive reinforcement, the stimulation of the muscle is important because it increases the tension of the muscle at the appropriate time. Biologic effect. By assuming part of the workload of the heart without increasing the myocardial oxygen requirement, the muscle actually increases the oxygen supply to the heart so that some recovery of the cardiac function may result from this mechanism. This biologic effect is suggested by two patients of our series who kept the full functional benefit of the operation for several days after the battery had been removed for infection. Enhanced coronary revascularization. Although scattered coronary capillaries have been stained after injection of dye into the latissimus dorsi artery in experimental animals after 1 year, this has not been observed in the human being thus far.

Indications and conclusion Although some positiveconclusions can be given at this stage, numerous points remain obscure and serious drawbacks must be underlined. The two most important are the variability of the results and the limited improvement

Carpentier et at.

5I

Table VIII. Optimal indications and contraindications for cardiomyoplasty Optimal indications Patients below 70 years of age Dilated cardiomyopathy (idiopathic or ischemic) Functional class III or intermittent class IV Sinus rhythm Mitral valve regurgitation <++ Left ventricular ejection fraction> 15% Left ventricular end-diastolic volume <250 ml Contraindications Hypertrophied cardiomyopathy Permanent functional class IV Severe, intractable ventricular arrhythmia Pulmonary artery hypertension Biventricular failure Mitral valve incompetence >++ (temporary contraindication) Neurovascular disease or cachexia Left pleural calcifications Severe respiratory insufficiency ( <55% predicted) Ejection fraction <10% Cardiac index <1.5 L/min per square meter Left ventricular end-diastolic volume> 300 ml Left ventricular end-diastolic pressure >45 mm Hg

observed after this operation. Numerous factors may explain these drawbacks: heterogeneous patient population, heart size, suboptimal indications, variable muscle anatomy, and vascularization. Some of these factors have been improved. They may be further improved on in the future, especially the selection of the patients. Table VIII summarizes the indications and contraindications as we consider them at present. In contrast, some positive points must be recognized that justify and stimulate our continuous effort to develop this operation. The fact that continuous fatigueless contraction of a skeletal muscle at the frequency of the heart has been obtained for periods of more than 7 years in the human being is in itself remarkable. The fact that patient survival was increased and that the functional condition of most patients was improved, with a significant increase in cardiac output and ejection fraction in many of them, is most encouraging. Finally, the fact that similar results have been obtained by various groups throughout the world, provided that they had access to burst-train generators, demonstrates that the operation is reproducible. All new operations present uncertainties and shortcomings in their developmental phases. As experience has accumulated, some of them have already been solved. Further progress is necessary, however, for this operation to face the challenge of the growing lack of donors for cardiac transplantation.

52

The Journal of Thoracic and Cardiovascular Surgery July 1993

Carpentier et al.

REFERENCES I. Carpentier A, Chachques JC, Grandjean PA, Perier P,

2.

3.

4.

5.

6.

7.

8.

9. 10.

11.

12.

13.

14.

15. 16.

17.

Mitz V, Bourgeois I. Transformation d'un muscle squelettique par stimulation sequentielle progressive en vue de son utilisation comme substitut myocardique. CR Acad Sci Paris 1985;301:581-6. Chachques JC, Grandjean PA, Schwartz K, et al. Effect of latissimus dorsi dynamic cardiomyoplasty on ventricular function. Circulation I 988;78(Suppl):III203-1 6. Carpentier A, Chachques Jc. Myocardial substitution with a stimulated skeletal muscle: first successful clinical case. Lancet 1985; I: 1267. Carpentier A, Chachques J'C, Surgical technique. In: Carpentier A, Chachques JC, Grandjean PA, eds. Cardiomyoplasty. Mount Kisco, New York: Futura, 1991:105-22. Chachques JC, Carpentier A, Grandjean PA. Atrial or ventricular assistance using the cardiomyoplasty procedure. In: Chiu RC-J, Bourgeois I, eds. Transformed muscle for cardiac assist and repair. Mount Kisco, New York: Futura,1990;161-78. Loulmet D, Carpentier A, Beaudet RL. Atriomyoplasty. In: Carpentier A, Chachques JC, Grandjean PA, eds. Cardiomyoplasty. Mount Kisco, New York: Futura, 1991:1958. Magovern GJ, Heckler FR, Park SB, et al. Paced skeletal muscle for dynamic cardiomyoplasty. Ann Thorac Surg 1988;45:614-9. Magovern GJ, Park SB, Kao RL, Christlieb IY, Magovern GJ Jr. Dynamic cardiomyoplasty in patients. J Heart Transplant 1990;9:258-63. Carpentier A. The use of autologous pericardium for tissue patching. Club Mitrale Newsletter 1992;2:7. Moreira LF, Stolf NA, Bocchi EA, et al. Latissimus dorsi cardiomyoplasty in the treatment of patients with dilated cardiomyopathy. Circulation I990;82(Suppl):IV257-63. Jatene AD, Moreira LFP, Stolf NAG, et al. Left ventricular function changes after cardiomyoplasty in patients with dilated cardiomyopathy. J THORAC CARDlOYASC SURG 1991;102:132-9. Franciosa JA, Wilen M, Ziesche S, Cohn IN. Survival in men with severe chronic left ventricular failure due to either coronary heart disease or idiopathic dilated cardiomyopathy. Am J CardioI1983;51:831-6. Kantrowitz A, McKinnon W. The experimental use of the diaphragm as an auxiliary myocardium. Surg Forum 1959;9:266. Termet H, Chalencon JL, Estour E, et al. Transplantation sur Ie myocarde d'un muscle strie excite par pace-maker. Ann Chir Thorac Cardiovasc 1966;5:270. Nakamura K, Glenn WL. Graft of diaphragm as a functioning substitute for myocardium. J Surg Res 1964;4:435. Macoviak J, Stephenson LW, Spielman S, et al. Replacement of ventricular myocardium with diaphragmatic skeletal muscle: short-term studies. J THORAC CARDIOyASC SURG 1981;81:519. Dewar ML, Drinkwater DC, Chiu RC-J. Synchronously

18.

19.

20.

21.

22.

23.

24.

25.

stimulated skeletal muscle graft for myocardial repair. J THORAC CARDIOyASC SURG 1984;87:325. Moreira LFP, Chagas ACP, Camarano GP. Cardiomyoplasty benefits in experimental myocardial dysfunction. J Cardiac Surg 1989;4:164-70. Carpentier A, Chachques J C. The use of stimulated skeletal muscle to replace diseased human heart muscle. In: Chiu RC-J, ed. Biomechanical cardiac assist. Mount Kisco, New York: Futura, 1986:85-102. Carpentier A, Chachques Jc. Latissimus dorsi cardiomyoplasty to increase cardiac output. In: Rabago G, Cooley DA, eds. Heart valve replacement and future trends in cardiac surgery. Mount Kisco, New York: Futura, 1987:47386. Kao RL, Christlieb IY, Magovern GJ, Park SB, Magovern GJ Jr. The importance of skeletal muscle fiber orientation for dynamic cardiomyoplasty. J THORAC CARDIOyASC SURG 1990;99:134-40. Molteni L, Almada HE, Ferreira RP, Ortega D. Assessment of the optimal time interval between QRS and simple pulse stimulation in dynamic cardiomyoplasty. In: Chiu RC-J, Bourgeois I, eds. Transformed muscle for cardiac assist and repair. Mount Kisco, New York: Futura, 1990;189-96. Moreira LF, Seferian P, Bocchi EA, et al. Survival improvement with dynamic cardiomyoplasty in patients with dilated cardiomyopathy. Circulation 1991;84(Suppl):III296-302. Almada H, Molteni L, Ferreira R, Ortega D. Clinical experience with dynamic cardiomyoplasty. J Cardiac Surg 1990;5:193-8. Salmons S, Jarvis JC: Cardiomyoplasty: a look at the fundamentals. In: Carpentier A, Chachques JC, Grandjean PA, eds. Cardiomyoplasty. Mount Kisco, New York: Futura, 1991:3-17.

Discussion Dr. Albert Starr (Portland, Ore.). I would like to compliment Dr. Carpentier and his team for a landmark achievement in using skeletal muscle for cardiac augmentation. We in Portland, Oregon, have had a chance to learn from his experience, visiting him often, and have been part of the phase 2 chmcal study in the United States in performing this procedure. I'd like to present our early results with this operation and then ask Professor Carpentier a few questions. With Jeff Swanson and Dr. Trehan, we have performed 12 myoplasties over the past 8 months. These were done mainly in elderly patients with a mean age of 65 years. It was a ra~her pure series with no previous operations, and 11 of the 12 patients had a dilated cardiomyopathy. Only I patient had an ischemic cardiomyopathy. Like Professor Carpentier's patients, the patients had markedly reduced ventricular function. The mean left ventricular ejection fraction was 19% and the mean right ventricular ejection fraction was 21%. In this group of patients there have been two early deaths and one late death. The late death was due to arrhythmia about 3 months after the operation. The early deaths were due to multiorgan failure as a result of poor

The Journal of Thoracic and Cardiovascular Surgery Volume 106, Number 1

cardiac performance after the operation. Our first five patients survived easily. Then we selected patients who were extremely ill (class IV); two of them died early and one late. We have observed some complications from the procedure, for example, one early lead detachment from the myocardium and one generator failure. We have also had a few minor problems with the chest wound-perhaps Dr. Carpentier could address the issue of managing this large dead space in the back of the chest-and one major infection. Only four of our patients have thus far reached 6 months of follow-up; however, the condition of all of the patients improved dramatically after discharge from the hospital. Three of the patients with 6 months' follow-up are in NYHA class I and one is in class II. All these patients were in late class III or early class IV before the operation. Ejection fraction improved dramatically in one patient, it decreased slightly in another, and in a third the ejection fraction remained the same after operation, despite the fact that the functional class of these patients improved enormously. There was a 30% increase in exercise performance in two of our patients. Looking at the risk factors for mortality, we found a different level of significance for various factors. The factors that appear to be the most significant are (I) a markedly diminished exercise oxygen use before the operation, (2) a very low cardiac index at the end of the operation, (3) a very poor right ventricular or left ventricular ejection fraction, and (4) a markedly dilated heart in systole and diastole before the operation. In light of this experience thus far, Professor Carpentier, would you consider performing this operation in patients who would otherwise be completely acceptable for cardiac transplantation? All of our patients were elderly, and of the two patients who were potential transplant candidates, one had carcinoma of the breast and the other refused transplantation. As a result, we have not operated on any patient who is on a transplant list. Would you do so? Second, have you an explanation for the remarkably improved functional result, even in the absence of profound hemodynamic improvement after the operation? Dr. Radu C. Deac (Philadelphia, Pa.). Professor Carpentier deserves credit for the first human clinical application of cardiomyoplasty, which opened a new era in cardiac assistance. This is an impressive series. Professor Carpentier, are there any data in this group of patients about the preoperative condition of the skeletal muscle (such as magnetic resonance imaging spectroscopy) used in the cardiomyoplasty procedure that could be correlated with late hemodynamic performance? Also, are there any histologic data about the skeletal muscle used in patients with failed cardiomyoplasty who required heart transplantation? One of the most important factors in the efficacy of cardiomyoplasty is the condition of the skeletal muscle, which can be influenced by the preoperative structure and function, the geometry of the cardiomyoplasty complex, and electrical conditioning and stimulation. Last, the vascularization of the skeletal muscle plays a crucial role in the remodeling process, essential for adequate hemodynamic performance in the long run. Dr. Chu-Jeng Chiu (Montreal. Quebec, Canada). I would like to bring up one point. In doing cardiomyoplasty we have to match each patient's heart with his own latissimus dorsi muscle. Not only the sizes of the patients' hearts vary, and often they are very large, but also the latissimus dorsi muscles are of dif-

Carpentier et al.

53

ferent sizes, lengths, and quality. In our smaller series we have noted that the best results, both clinical and hemodynamic, seem to be obtained in patients in whom we were able to completely wrap the heart with a good quality muscle. I think this may be an important variable. In how many of your patients were you able to wrap the heart completely without the need for pericardial patches? Did you try to analyze your clinical results and hemodynamic results, relating them specifically to the completeness and the quality of each muscle wrap? Dr. James A. Magovern (Pittsburgh, Pa.). At Allegheny General we have now done 26 cardiomyoplasty procedures. Our overall results are quite similar to yours, but we have not been able to demonstrate consistent changes in cardiac function in patients who have undergone the procedure without a concomitant operation. Because of this finding we began work in the research laboratory to develop a procedure that would have greater effects on left ventricular function. At the 1991 Surgical Forum we presented a study that showed cardiomyoplasty to be more effective when the right latissimus dorsi muscle was used rather than the left. The right muscle appears to have a better orientation to improve left ventricular function because it pulls the left ventricle toward the septum rather than toward the left shoulder. This work has now been extended in long-term studies. Pressure-volume loops were measured in dogs 8 weeks after right cardiomyoplasty, when the latissimus dorsi was fully conditioned. With muscle stimulation there are large increases in stroke work and stroke volume, but no impairment of diastolic filling. This difference is more dramatic if the animal is given esmolol to depress left ventricular function. The stroke work with stimulated beats is nearly three times that of the unstimulated beat. Because of these promising experimental findings we have begun using the right latissimus dorsi muscle in our clinical series. As of April 1992, we have used the right latissimus dorsi muscle in II patients. There have been no operative deaths, and all the patients have had some improvement in ventricular function. In the first eight patients at 6 weeks after the procedure, there was an increase in left ventricular ejection fraction from 27% to 36% and a small increase in right ventricular ejection fraction. Also, the left ventricular end-diastolic volume decreased from a mean of 366 ml to a mean of 307 ml. I have two questions for Dr. Carpentier. Have you seen any improvements in the patients who have had cardiomyoplasty without a concomitant operation-s-changes in myocardial function or in exercise capacity or functional improvement? Have the patients who have had a concomitant procedure been any different from the ones who have had an isolated cardiomyoplasty? If not, do you have any proposed changes or improvements in operative technique that might lead to improved cardiac function? Dr. Quentin R. Stiles (Los Angeles. Calif). Vincent Dor has achieved improved ventricular function in patients with ventricular aneurysms by resecting the endocardial segment, then replacing it with a Dacron patch, and thereby getting improved ejection fractions. One thing that has always concerned me about this myoplasty is this: Is the end-diastolic pressure within the left ventricle sufficient to stretch the latissimus dorsi muscle to get real contraction? Are ejection fractions improved because of the muscle contraction or because of the support (similar to

54

Carpentier et al.

the way the Dacron patch supports), just by decreasing the ventricular chamber size? A decrease in chamber size means that less ventricular wall tension is required to generate the same pressure. I think that is an important physiologic concept for us to learn about. Dr. Carpentier. I would like to thank the discussants. Dr. Starr raised the problem of how to prevent seroma and infection at the site of the mobilized latissimus dorsi. Seroma and infection are facilitated by the important dead space left by the muscle after mobilization. To reduce this dead space, we secure the subcutaneous tissue to the costal plane by multiple sutures and leave three Hemovac drains (Zimmer Patient Care Division, Dover, Ohio) in place. In addition, a compressive thoracic dressing is applied for 10 days after the operation. The question of whether we would propose this operation to candidates for transplantation has two answers. Up to now, we have reserved this operation for patients who had a contraindication to heart transplantation or who refused heart transplantation. In the future, there will be three reasons to modify this attitude: (I) There are not enough donors for cardiac transplantation, (2) the mortality of patients awaiting heart transplantation is higher than the mortality of patients undergoing cardiomyoplasty, and (3) cardiomyoplasty does not preclude cardiac transplantation. Regarding the question of how to explain the difference between significant functional improvement and limited hemodynamic improvement, several hypotheses have been put forth in our paper. Ejection fraction is improved in the majority of the patients because of an improved force of contraction, but the filling pressures are not modified because they are mainly dependent on the speed of contraction, which is not increased by the skeletal muscle.

The Journal of Thoracic and Cardiovascular Surgery July 1993

Dr. Deac, we have had three patients in whom the muscle was examined histologically after more than 6 months of function. These patients required a heart transplantation. In two cases, the histologic studies showed normal structures. In one case, there was a distal necrosis and fibrosisof the muscle because of poor vascularization. The latter finding emphasizes the comments made by Dr. Chiu. I fully agree with him that many variables may explain the variability in the results, in particular the size of the heart, the size of the muscle flap, and the vascularization of the muscle. The left ventricle was wrapped completely in all the patients (provided that the posterior position was used), whereas the right ventricle was wrapped incompletely. This probably explains why the results have not been as good in patients with biventricular failure. Dr. Stiles, I have already answered your question regarding the size of the ventricle and the absence of change in pressures. With regard to possible impairment of the ventricular compliance as a result of muscle wrapping, we have not found this to be a problem except when there was not a good synchronization between muscle contraction and heart contraction. The wall tension is so important in the dilated ventricle that there is no difficulty for the muscle to be distended during relaxation. To Dr. Magovern, who first introduced this operation in this country, I must confess that I still do not understand how he can wrap the left ventricle with the right latissimus dorsi muscle in the human. We use the anterior position only when there is predominant right heart failure. Additional procedures carry an additional risk. Improvements in your results resulted from better patient selection and technical modifications. Our results have not been as good when valve surgery or aneurysm resection was necessary.