JACC Vol . 22, No. 3 758 September 1993 :758-67

Long-Term Follow-Up (12 to 35 Weeks) After Dynamic Cardlomyoplasty CAROLINE M. H. B. LUCAS, MD, FREDERIK H . VAN DER VEEN, PHD, EMILE C . CHERIEX, MD, ROBERTO LORUSSO, MD, MICHAEL HAVENITH, MD, OLAF C. K. M. PENN, MD, HEIN J . J. WELLENS, MD, FACC Maastricht, The Netherlands

Objectives. To obtain Information on the long-term effects of nearly complete transformation to type I muscle fibers. The dynamic cardiomyoplasty on hemodynamics and muscle histol- remaining 14 goats showed extensive li mat is in the latissimus ogy, this surgical. method was evaluated In goats . dorsl muscle, with severe intlmal hyperplasia and proliferation of lilac rid. Dynamic c lomyoplasty has been introduced as smooth muscle cells in the walls of the thoracodorsal artery and Its a new method to treat patients with severe cardiac failure . branches . An Increase in endoneural and endomysial connective Methods. In 24 pats, the left iatissimus dorsi muscle was tissue was observed, with some goats showing destroy nerve wrapped around the . The muscle was then subjected to branches near the electrodes . These findings differed from those progressive electric stimulation . In 16 goats, invasive transesoph- observed after long-term electrical stimulation of goat latissimus Doppler echocardi phic measurements and histol is dorsi muscle in situ . evaluation of the lat us dorsi muscle were performed at a 12 Conclusions. Dynamic cardlomyoplasty is of use in the treat- weeks after the wrapping. ment of severe he failure if the histol is structure of the Res+dis. Only two `oats showed an increase in aortic and left wrapped latissimus dorsi muscle remains Intact . Long-term re and right ventricular pressures concomitant with increased aortic suits in goats suggest that the current approach used in dynamic flow during latissimus dorsi muscle stimulation both before and cardiomyoplasty may lead to deterioration of the wrapped muscle . after induction of cardiac failure using Imipramine. This was (J Ant Coll Cardiol 1993 ;22 :758-b7) accompanied by a preserved latissimus dorsi muscle structure and

In the last decade, dynamic cardiomyoplasty has been Dynamic ca iomyoplesty procedure . Twenty-four goats introduced as a new surgical technique to treat patients with with a body weight varying from 28 to 55 kg were studied . severe heart failure . During this procedure, the latissimus Anesthesia was induced intravenously with sodium thiop, n- dorsi muscle is wrapped around the heart and electrically tal (Nesdonal, 15 mg(kg body weight). After endotracheal stimulated synchronously with cardiac contraction. intubation, goats were maintained with oxygen/nitrous oxide Since the first clinical application in 1985 (1), a number of (1:2) and 1.5% fluothane . During the experiments, the lungs reports (2-11) have examined acute and long-term effects of were ventilated with a positive pressure respirator (Pulmo- dynamic cardiomyoplasty in animal experiments and pa- mat). The entire procedure was performed under sterile tients. This report describes our experience during long-term conditions, with antibiotic coverage using ampicillin (Am- fellow-up in a large number of goats . fipen) (1,000 mg intravenously) before operation . Pedicle preparation . A left-sided midaxdlary incision was performed and the collateral blood vessels to the distal Methods part of the latissimus dorsi muscle coagulated . All attach- The experiments were performed according to the "Na. ments of the muscle were disconnected, with the exception tional Academy of Sciences Guide for the Care and Use of of the axillary pedicle to preserve the thoracodorsal artery, Laboratory Animals" (12) . vein and nerve . Two intramuscular electrodes (Medtronic SP5528) were implanted in the upper part of the latissimus dorsi muscle flap perpendicular to the main branches of the From the Departments of Cardiology, Pathology and Cardiothoracic thoracodorsal nerve . Threshold value of stimulation was Surgery, University of Limburg, Maastricht, The Netherlands . This study was supported in part by a grant from the Netherlands Heart Foundation, The measured during temporary connection of these electrodes Hague, The Netherlands . to a programmer (Medtronic SP3028) . Manuscript received July 7, 1992; revised manuscript received January Wrapping procedure . A partial resection of the anterior 19, 1993, accepted February 12, 1993 . Add= f coresn xk=: Caroline M. H. B. Lucas, MD, Department arch of the second rib was performed to transfer the muscle Of Cudiokw, Academic Hospital Maastricht, P .O. Box 5800, 6202 AZ and pacing leads into the thorax . The thoracic cavity was Maastricht, The Netherlands . opened at the fifth left intercostal space and the pericardiurr

®1993 by the American College of Cardiology 0735-1097!9316.00

JACC Vol . 22, No . 3 LUCAS ET AL . 759 September 1993 :758-67 LATISSIMLS DORSIMUSCLE FOR CARDIAC ASSIST

Table t . Stimulation Protocol of to Situ Stimulated Latissimus visible contraction of the entire latissimus dorsi muscle Dorsi Muscles* without causing apparent discomfort to the arimal . The laaterpulse Interval muscle was stimulated continuously, 24 h/day, for a period No. of (ams)/Burst Bursts of 12 weeks . Week Pulses Frequency (11Z) Min After this period, biopsy specimens were taken from the Protocol A (n = 3) midportion of stimulated and nonstiroku'Lled latissimuss dorsi 1 to 2 1 50 muscles . One biopsy specimen from both sides was quickly 3W4 2 100/10 50 frozen in isopentane. A second specimen was stored in 4% 5 to 6 3 67115 50 paraformaldehyde . Small biopsy samples for electron micro- 7WS 6 33/30 50 9 to 10 6 33130 60 scopic evaluation were also taken from both muscles . After 11102 6 33/30 50 the 12-week period, all six goats were killed using an Protocol B (aa 3) overdose of thiopental . 1 to 2 3 33/30 30 Hemodynamic evaluation . The hemodynainic effects of 3w4 3 33/30 40 dynamic cardiomyoplasty were investigated using both in- 3 33/30 5 to 6 50 vasive and Doppler echocardiographic measurements . Mea- 7 to 8 6 33/30 50 surements were made both in the normal heart and after the 9 to 10 6 33B0 60 induction of cardiac failure by imipramine 11 to 12 6 33/30 80 (13). Anesthesia, ventilation, infusion and electrocardiographic "'AY0 slightly diflercat protocols were usck, ivc indicaled by A and B . recordings were performed as described for the wrapping procedure. Aortic blood pressure was measured with a 7F removed. A sensing electrode (Medtronic SP5548) was im- Millar catheter tip micromanometer (Millar FC470) posi- planted in the left ventricular wall . Subsequently, the left tioned in the aortic arch through the femoral artery . Left latissimus dorsi muscle was wrapped in a counterclockwise ventricular pressure measurements were obtained with a 7F fashion around both ventricles as described previously by Millar catheter tip micromanometer (Millar PC470) inserted Anderson et al . (8). All three electrodes were connected to a through the left brachial artery into the left . Left cardiomyostimulator (SP1005 Medtronic) that was subse- ventricular end-diastolic pressure was recorded at a sensi- quently positioned in the interscapular region . tivity of 20 nu i Hg. The first derivative of left ventricular All goats received ampicillin (1,000 mg/day intramuscu- pressure (pos:tzve and negative dP/dt max) was determined larly for 5 days) as antibiotic treatment and buprenorfine as described by Schaper et al . (14). Right ventricular pres- (Temgcsic, 0 .6 mg/day intramuscularly) for analgesia. Two sure was measured using a Millar catheter tip micromanom- weeks after the wrapping procedure, electrical stimulation eter inserted into the right ventricle through the left jugular was started using the protocol described by Chachques et al . vein. All catheters were flushed repeatedly with saline (6). One pacing burst was synchronized to every heartbeat, solution . with an upper limit of 100 latissimus do rsi muscle contrac- Aortic blood flow velocity recordings were made using a tions/min . Hemodynamic studies were performed after a transesophageal approach. A 64-element 5-MHz phased follow-up period of 2:12 weeks, that ig,, after the end of the array transducer was positioned at the level of the ascending conditioning period of the wrapped latissimus dorsi muscle . perpendicular to the aortic valve plane, allowing In situ latisshnus dorsi stimulation. Six healthy goats with registration of the blood flow just above the valve using a body weight ranging from 29 to 54 kg were studied . The pulsed Doppler . The flow integrals were animals were anesthetized initially with sodium thiopental recorded using a strip chart recorder with a paper speed of (15 mg/kg) into the jugular vein, intubated and ventilated 50 inm/s and were traced offline on a digitizing tablet with oxygen/nitrous oxide (1:2) and 1 .5% halothane . A (Summasketch II) using a scientific measurement program 15-cm longitudinal skin incision was made in the left mid- (Sigmascan) . axillary line under sterile conditions, with preoperative To stimulate the wrapped muscle, different modes of antibiotic coverage using ampicillin (1,000 m j) intravenous- stimulation were used in the order given in Table 2 . Con- ly). Subsequently, two intramuscular electrodes (Medtronic traction of the latissimus dorsi muscle was synchronized to SP5528) were attached approximately 6 err, apart in the every 3rd heartbeat as long as the cardiac frequency re- proximal part of the latissimus dorsi muscle perpendicular to mained >100 beats/min . When the heart rate was less than the main branches of the left thoracodorsal nerve and this value, the cardiomyostimulator switched automatically connected to an Itrell pulse generator (Medtronic SP7420). to a synchronization ratio of one latissimus dorsfi muscle Stimulation of the left latissimus dorsi muscle was started contraction to every other cardiac contraction . Voltage of 2 weeks after this operation according to the protocol shown stimulation was programmed at 5 V and subsequently in- in Table 1, which is comparable to that commonly used after creased to S V when no effect was seen on hewn ynamic a dynamic cardiumyoplasty procedure . All individual elec- variables. All pacing modes were applied for a 2 Min . trical pulses had a duration of 0 .21 ms . The voltage was Stimulation-free intervals of approximately 30 s were pro- adjusted to 2 .6 ± 0.4 V to produce an easily palpable and grammed between the different protocols . After this part of

JACC Voi . 22, No . 3 760 LUCAS ET AL. LATISSIMUS DORSI MUSCLE FOR CARDIAC ASSIST September 1993 :759-67

Table 2 . Successive Stages of the Stimulation Protocol removed. Four biopsy specimens from the latissimus dorsi Sensing Delay muscle were taken at -lifferent locations, starting at the No. of Pulses Interpulse Interval (ms)/ to the R pedicle and descending t.) a part of the muscle next to the Stage in a Burst Burst Frequency (Hz) Wave (ms) apex of the heart (corrcfqronding to locations A . B, C and D . These specimens were quick frozen in I I 4 as described later) 2 100/10 4 isopentane and stored at -70°C for further analysis . All III 3 67/15 4 pieces of tissue were carefully positioned for transverse IV 6 3350 4 sectioning. V 10' 2060 4 Small samples (I x I X I mm) from these biopsy 12/85 4 VI 16 specimens were taken and fixed in 2 .5% glutaraldehyde in VII I 100 0.01M phosphate buffer at pH 7 .3 . The osmolality of the VIII 2 100/10 100 IX 3 67/15 100 solution was adjusted to 330 mOsm by the addition of X 6 030 10A sucrose for subsequent electron microscopic investigation. X1 to 20/50 100 The remaining part of the heart and the wrapped latissimus X11 16 12/85 100 dorsi muscle were stored in 4% paraformaidehyde for sub- sequent analysis . Histologic analysis . All specimens frozen in isopentane the protocol was completed, depression of cardiac function were used for fiber typing, capillary counts and the measure- was induced by continuous infusion of imiprambe (Tofranil) ment of fiber diameter of type I and type 11 fibers . This was in a dosage of 7.5 mg/kg body weight per hour (13). There- done by combining immunohistochemical fiber typing with after, the same stimulation protocol of the latissimus dorsi the monoclonal antibody R I IMO and enzyme histochemical muscle was repeated . capillary staining using alkaline phosphatase as has been Data analysis. In each goat, 10 nonassisted beats were described previously (15) . Rl ID10 is an antibody against, measured immediately before the start of every mode of the human left ventricular heavy chain myosin (16) and demon- stimulation protocol during the stimulation-free intervals; 5 strates immunoreactivity to type I muscle fibers in human assisted beats were measured thereafter . This was done (it, canine (17) and goat (unpublished observations) skele- during continuous artificial respiration because respiratory tal muscle. arrest frequently resulted in ventricular extrasystoles . The Three hundred muscle fibers were identified immuno- same approach was used after the induction of cardiac histochemically in randomly distributed fascicles in all spec- depression with imipramine . imens. Furthermore, the cross-sectional area of these indi- Statistical analysis . The 10 nonstimulated and 5 stimu- vidual muscle fibers was measured and expressed as the lated beats during every mode of stimulation were averaged diameter of the circle of an equivalent area . The number of for every goat . The mean values for the nonassisted beats capillaries that were adjacent to the muscle fibers used for were subsequently averaged for every stimulation mode (12 fiber typing and fiber area measurements were counted (15) . in total) and compared using repeated measures one-way Slides were also stained for aeetylcholinesterase enzyme analysis of variance. The mean value for all control beats activity to identify the mote r endplates . was compared with the mean values for all stimulated beats All with wrapped latissimus dorsi muscles tempo- using the Student t test for unpaired samples . The same rarily stored in 4% paraformaldehyde were cut into four procedure was performed after depression of cardiac func- slices perpendicular to the basoapical axis, starting at the tion by imipramine. The effect of imipramine was evaluated base of the heart near the atrioventricular groove and ending by comparing the pooled control beats from all stimulation at the apex. From every slice, at least three specimens protocols before and after administration of this cardio- composed of myocardium with adjacent latissimus dorsi depressive drug using the Student t test for unpaired sam- muscle were taken at different locations . Specimens were ples. Differences were considered significant at p < 0.05. All also taken from the tissue around both intramuscular elec- data are expressed as the mean value ± SD . trodes in all 24 goats . All specimens were embedded in Histopthologic evaluation. After dynamic cardioinyo- paraffin and cut in 3-Am sections. For histologic analysis, plasty . After completion of the last hemodynamic measure- hematoxylin-eosin and elastin-van Gieson were used . For ments, a transmural biopsy specimen was taken from the immunohistochemical staining, the monoclonal antibody midportion of the right nonstimulated latissimus dorsi mus . RUM and sibs red were used. In addition, some speci- cle for histopathologic investigation . This specimen was mens were stained using periodic acid-Schiff . immediately frozen in isopentane, quenched in dry ice and The picrosirius-polarization method, staining collagen stored at --70°C. types I and III, was used to demonstrate the amount of Subsequently, the go; is were killed with an overdose of endomysial connective tissue within the left and right latis- sodium thiopental. After the thoracic cavity was opened, the simds dorsi muscles (18). Quantification of the amount of heart with the wrapped left latissimus dorsi muscle was endomysial connective tissue was performed as described by carefully excised. Adherent pulmonary and fatty tissue were Van Krimpen et al . (19). Five skeletal muscle fascicles were

JACC Vol . 22, No. 3 LUCAS ET AL. September 1993:758-67 LATISSIMUS DORM MUSCLE FOR CARDIAC ASSIST 761

Table 3. Hemodynamic Res Its of Assisted Versus Nonassisted Beats Both Before and After the Administration of Ir-tipramine (n = 0) TVI RVP LVEDP LVP AoP LV dP/di "cm) (mm Hg) (mgt Hg) (mm Hg) (mm Hg) M ni Hg/s) C 7.66 'Ji 27±8 14= 6 110 ± 27 111 ± 26 1,300 411 5.33 ; .07` 29 ± 9* 14 t 7 112 ± 27 112 ± 26 1 .326 411 C Imi 6.50 1 .08t 19 ± 6f 16 ± 3t 62 ± 21t 63 ± 21 554 124t S Imi 7.19 2.44y 21±7* 1;,±3 63 ± 21 64 ± 22 576 146 *Significant difference between nonassisted and assisted beats (p < in . 6jilicant difference between nonassisted beats before and after the infusi m of inupramine (p < 0.05). .11 data are expressed as mean value ± SD . AoP = aortic pressure; C = nonassisted beats ; C Imi = nonassisted beats after imipramine ; LV dP/dt = first derivative of left ventricular pressure ; LVEDP "- left ventricular end-diastoli,: pressure ; LVP = left ventricular pressure ; RVP = right ventricular pressure; S = assistec beats ; S fini = assisted beats after irnipramine ; TVI = time-velocity integral of flow in the ascending aorta .

chosen at random from one specimen at all four slices . The 12 to 35 weeks was reached in 16 goats, Eight goats either amount of endomysial connective tissue was measured at died spontaneously or were killed because of infection five different sites within each selected fascicle . before the end of week 12, which was the end of the Evaluation of hisiologic slides . All specimens embedded latissimus dorsi muscle conditioning period . Two goats from in paraffin were evaluated by two observers. First, the this group died before the start of the stimulation protocol . percent of type I and type 11 fibers was calculated, counting emodynamic data . A complete hemodynamic evalua- 300 fibers/specimen . The amount of connective tissue lion was obtained in 13 of the 16 goats that were killed after around the electrodes was checked by measuring the width 12 weeks . Three goats died during the transvenous insertion of the collagen sheath at 10 different sites a-.,aund every of the right ventricular catheter before the stimulation pro- intramuscular electrode in all goats using a MOP Videoplan tocol for the wrapped latissimus dorsi muscle could be Kontron system . Subsequently, the amount of fatty tissue started . As demonstrated in Table 3, pooling of the data from within the separate muscle fascicles was scored by compar- all stimulation protocols resulted in a small but significant ing the specimens stained with elastin-van Gieson from the increase in right ventricular pressure both before and after wrapped latissimus dorsi muscle with those from the non- the infusion of imipramine . The time-velocity integral of stimulated right latissitrius dorsi muscle . The area occupied aortic flow showed a small but significant increase only by fat cells in the stimulated latissimus dorsi muscle was before the induction of cardiac depression by imiprarnine . scored semiquantitatively as follows: I + (0% to 25%), 2 + No significant effect was seen for all the other measured (25% to 75%) and 3 + (>75%) . hernodynat ic variables . Imipramine itself caused severe Electron microscopic studies . Transverse sections of depression of left ventricular pressure, left ventricular dP/di both left and right latissimus dorsi muscles were stained with and right ventricular pressure, suggesting biventricular fail- lead citrate and uranyl acetate and examined using a Philips ure. Furthermore, a decrease in aortic pressure was seen 300 electron microscope . To assess the structure of individ- concomitant with a decrease in cardiac frequency frogs ual latissimus dorsi muscle fibers, six areas were selected at 117 ± 22 to 102 ± 17 beats/min . The decrease in the the peripheral and six at the central part of each muscle fiber time-velocity integral of aortic flow appeared less r .a- and photographed at a final magnification of x 14,000 . nounced. Two goats from the group of 13 goats, both with a Ira situ latissimus dorsi muscle stimulation . As mentioned follow-up period of 18 weeks, demonstrated significant in- previously, in all six goats biopsy specimens were taken creases in pressures and aortic flow before and after the from the midportion of left and right latissimus dorsi muscles administration of imipramine . Representative examples from distal to the intramuscular pacing electrodes. These speci- one of these goats are given in Figures I and 2 . As will be mens were handled and evaluated as described for the discussed ,n the next section, these excellent hemodynamic cardiomyoplasty group. results were accompanied by good preservation of skeletal Statistical analysis . Results from left and right latissimus muscle structure . dorsi muscles from wrapped goats and in situ stimulated Histopathologic data . Dynamic cardiomyoplasty group . goats are compared using repeated measures analysis of In Table 4 all histologic data are presented from the 16 goats variance followed by the Student t test for unpaired samples . having a follow-up period of 2t 12 weeks. Electrical stimula- Differences were considered significant at p < 0.05. Values tion resulted in pronounced transformation to the fatigue- are presented as mean value ± SD . resista,it type I fibers in all examined latissimus dorsi mus- cles. This was accompanied not only by a significant decrease in fiber diameter of type I fibers but also by a Large Results variability in fiber size . According to the protocol, all goats with the in situ The percent of endomysial connective tissue increased to stimulated latissimus dorsi muscles were killed after 12 a maximum of 17 .5 ± 9.0% at the part of the latissimus dorsi weeks . In the cardiomyoplasty group, a follow-up period of muscle adjacent to the apex of the heart (location D).

762 LUCAS ET AL. JACC Vol. 22, No 3 LATISSIMUS DORSI MUSCLE FOR CARDIAC ASSIST Septen ber 1993 :758-67

120 1-1 1-1 ri hyperplasia and proliferation of smooth muscle cells were ri rl LVP seen in the larger arteries in a large number of specimens U U %_j U k liompl-i V (Fig. 3).

LVEdP As demonstrated by the fatty tissue score, wrapping and fill" electrical stimulation resulted in a marked increase in fat W u Lj U cells both in the interfascicular space and in particular within 04

2000 the individual muscle fascicles in 14 of the 16 goats (Fig . 4). wows N ~r', LVdP/dt This increase in adipocytes appeared to be the result of both progressive destruction of myofibrillar bundles, followed by 125 replacement with fat and the invasion of fat cells from the AOP MAM outside of the muscle fascicles . This was observed in partic- 0 ular in those latissimus dorsi muscies that mainly consisted of type I muscle fibers. In most of the specimens, a great RYP variability in fiber size was seen together with necrotic fibers, whorled fibers, coiled fibers and signs of phagocyto- Ece sis. Only two goats, both having a follow-tip period of 18 weeks, slowed excellent preservation of latissimus dorsi Figure 1. Recording showing the effect of synchronous burst pacing of the wrapped latissimus dorsi muscle on different variables of muscle structure and complete transformation to type I cardiac function. Pacing spikes of the stimulator are visualized in the fibers at all biopsy locations without a significant increase in tracing at the bottom left of the figure, demonstrating burst stimula- fatty tissue . tion of the wrapped latissimus dorsi muscle synchronized to every In six animals, electrical stimulation was started after other heartbeat . AOP - aortic pressure; ECG = electrocardiogram; cardiomyoplasty, but death occurred at a mean of 6 LVdP/dt - positive and negative first derivative of left ventricular .3 ± 2.0 pressure; LVEdP = left ventricular end-diastolic pressure ; UP weeks after operation. Biopsy specimens from these animals left ventricular pressure ; RVP = right ventricular pressure . demonstrated less impressive signs of muscle cell destruc- tion and an increase in adipocytes as judged by a maximal fatty tissue score of 1 .0 ± 0 .8 at location C. A specimen from The width of the connective tissue sheath around the one goat in this group is shown in Figure 5 and demonstrates intramuscular electrodes measured <0.5 mm if no signs of an an ongoing process of fiber transformation with a mixture of inflammatory reaction were visible (0.37 ± 0.12 mm around type I, type 11 and intermediate fibers . In the two goats that the proximal [n = 21] and 0.34 t 0.09 mm around the distal died within 2 weeks after the cardiomyoplasty procedure, no [n = IQ electrodes). When inflammation occurred, the width important signs of damage were visible in specimens from of the surrounding collagen increased to 1 .53 ± 0.76 mm the wrapped muscle . around the proximal (n = 3) and 2 .67 ± 0.2 mm around the Periodic acid-Schiff staining showed isolated individual distal (n = 8) electrode . fibers with a strong staining reaction as a sign of muscle fiber Contrary to the results for the in situ stimulated latissimus regeneration in a number of biopsy specimens. dorsi muscles (Table 5), no increase in capillary/fiber ratio In the nonstimulated latissimus dorsi muscle, acetylcho- was found in the wrapped muscles . Moreover, severe intimal linesterase enzyme staining demonstrated clearly defined

flail 1411144141104 141114 111111441"141141111 oil 1411111#141141111111111111 off lot oil too 1111411111146141111111111 Ito I $I mol 0 .0 oils .. Figure 2. Transesophageal Doppler 4 AoFlow echocardiogram of aortic flow (Ao-

so Flow) demonstrating on the left the increase in aortic flow during stim- ulation of the wrapped latissimus dorsi muscle using a burst of six Im/s pulses at a synchronization ratio of one latissimus dorsi muscle con- traction to every other cardiac con- traction . On the right, the basal nonassisted aortic flow is dis- played . A 'Wt ventricular pressure curve is shown concomitant with V, the flow recording. Respiration is seen as a third curve within the $/ j' u yyp rrri rr-rrr-.vrrv, . tracing . The electrocardiogram shows the individua! pacing spines 9 TLI!U_ 11111 "I during electrical stimulation of the latissimus dorsi muscle. Is

JACC Vol . 22. No . 3 LUCAS ET AL, 763 September 1993 :758-67 L,01SSIMUS DORM MUSCLE FOR CARDIAC ASSIST

Table 4 . Histologic Data Concerning Changes in Wrapped Ladssimus Dorsi Muscles Stimulated foi >12 Weeks

Right LD Left A Left B Left C Left D

%type l 27 ± 7' 89 :t 16* 92 13 , 92 (3* A :!: 5* • D& I ( gy m) 575-t62 41 .8 1010 42 .1 914 417 W 45 .8 7 .1 * D& II Win) 5t .6 5.2 50 .3 i 15 .1 46 .4 18 .7 50.8 12 .9 411 40 % area endwriphl can tissue 3 .7 13 (3 .3 ± 12 .8 15 .0 5j" 17 .5 %0* CVMkr ratio 1 .2 ¢ 0.2t 1 .3 ± 0.6 1j ± 0 .4t 0 .9 0"t L I M2 1, Fatty tissue score 0+0 1 .8±1 .0'8 13 1 .6 IJ4 ( .6 U"

All data are expressed as mean ± SD (n = 16) . *Significant difference between right and left latissimus dorsi muscle (LD) (p < 0 .05) . -; Significant difference between the in situ stimulated latissimus dorsi group and the ca%myoplasty group of goats (p < 0 .05) . Ock I = cl°ameter of a circle equivalent to the measured area of type I muscle libem Dcir II = diameter of a circle equivalent to the measured area of type 11 muscle fibers ; Capffiber ratio = capillary to fiber ratio ; Left A, B, C and D = measurements from biopsy specimens taken at different locations in the wrapped latissimus dorsi muscle, starting from the base to the apex of the heart, respectively ; % area endomysial con tissue = percent of the latissimus dorsi muscle area occupied by endomysial connective tissue ; % type percent of type I muscle fibers.

motor endplates restricted to a small area of the muscle cell a iiihmi e of atrophic fibers next to giant whorled fibers was membrane, However, in the wrapped and stimulated latissi- seen, All stimulated latissimus dors .:, muscles showed a small mus dorsi muscles . a number of slides demonstrated muscle increase in fatty tissue, which was confined to the interfas- cells with a diffuse spread of the enzyme acetyleholinester- cicular space . ase around the entire cell membrane . In the majority of the Stimulation resalted in a significant decrease in fiber larger nerves, a pronounced increase in endoneural connec- , diameter of type 1 fibers and an increase in intrafascicular Live tissue was also apparent compared with that in the nerves in the right nonstimulated latissimus dorsi muscle . In several biopsy specimens, signs of damage were seen in the Figure 3. Photograph of a wrapped and long-term stimulated latis- large nerve branches in the vicinity of the intramuscular simus dorsi muscle demonstrating severe intimal hyperplasia and electrodes . proliferation of smooth muscle cells in the arterial wall (elastin-van Gieson stain) . The contact between myocardium and wrapped iatissi- mus dorsi muscle appeared very heterogeneous . At several sites, the muscle was firmly attached to the heart with a small rim of collagen, containing a number of small vessels . At other biopsy sites, however, the latissimus dorsi muscle was connected very loosely to the heart . The latissimus dorsi muscles stimulated for >12 weeks demonstrated severe destruction of myofibrillar bundles with concomitant loss of their original orientation in the majority of muscle fibers (Fig . 6) . The amount and distribu- tion of mitochondria were very variable and a large number of these mitochondria . appeared to be severely damaged . In situ latissimus dorsi stimulation, Five of the six in situ-stimulated latissimus dorsi muscles demonstrated a histologically normal, vital latissimus dorsi muscle structure mainly composed of type I fibers (Table 5). In the sixth goat,

Table S. Histopathologic Data Concerning Changes in the In Situ- Stimulated Left Latissimus Dorsi (LD) Muscle Compared With the Right Nonstimulated Latissimus Dorsi Muscle

Right LD Left LD

% type 1 19 ± 4 98 ± 3* Dcir I (Am) 64 .1 ± 7 .9 52 .6 ± 9 .5 Dcir II (jAm) 63 .3 t 10 .8 % area endomysial con tissue 3 .2 ± 2 9.0 ± 4 .3* Capifiber ratio 1 .0 ± 0 .1 1 .8 ± 0 .4"

*Significant difference between right and left latissimus dorsi muscles (p < 0 .05). All data are expressed as mean value ± SD (n = 6 for right latissimus dorsi muscle ; n = 5 Fir left LD muscle). Abbreviations as in Table 4 . JACC Vol . 22, No . 3 704 LUCAS ET AL . LATISSIMUS DORSI MUSCLE FOR CARDIAC ASSIST September 1993:758-67

FIgure 4. Photograph of a wrapped and long-term stimulated latis- 5. Tissue specimen from a wrapped latissimus dorsi muscle simus dorsi muscle demonstrating severe infiltration with and re- stimulated for 4 weeks demonstrating an ongoing process of muscle placement of muscle cells (dark ) by adipocytes (RIIDOO fiber transformation from white-staining type II fibers to dark- monoclonal antibody stain). staining type I fibers !RI IDIO monoclonal antibody stain) . No signs of skeletal muscle cell destruction are seen. connective tissue (Table 5) . Furthermore, a significant in- crease in capillary/fiber ratio was noticed. No abnormalities were seen in the larger arteries of these muscles . Acetylcho- data on the long-term effects of this procedure . In particular, linesterrase enzyme staining showed intact motor endplates histologic observations are scarce (2,3,6,8). The positive in both muscles . hemodynamic results in two of our goats demonstrate that Ultrastructura ly, well preserved, well oriented myofibril- improvement of cardiac function in both the normal and the lar bundles with a significant increase in the number of failing heart is possible as long as latissimus dorsi muscle normal mitochondria per muscle fiber were present. structure is preserved . Unfortunately, however, our experiments demonstrate severe degenerative changes in latissimus dorsi muscle Discussion structure in the majority of the goats after wrapping and Dynamic cardiornyoplasty as a new surgical method to subsequent continuous electrical stimulation . The muscle increase depressed cardiac output has been used to treat changes in these animals look similar to those in patients patients with severe cardiac failure . However, as demon- with muscular dystrophy and animal models simulating this strated by our combined hemodynamic and histopathologic disease (25,26). These changes were not seen in the in data, satisfactory results are obtained only when the situ-stimulated latissimus dorsi muscles . Although both the wrapped latissimus dorsi muscle remains vital despite dis- wrapped and in situ-stimulated latissimus dorsi muscles section and continuous electrical stimulation. showed almost complete transformation to fatigue-resistant A positive effect of dynamic cardiomyoplasty on hemo- type I fibers that are capable of l wig-term continuous con- dynamic ftmction of both healthy and failing hearts has been traction, these fibers were progressively replaced by fat cells reported in acute experiments (20-23) . However, only a in the majority of the wrapped latissimus dorsi muscles . limited number of research groups (1-11,24) have published Moreover, the increase in endomysial connective tissue

JACC Val . 22 . No . 3 LUCAS ET AL. 765 September 1993 :758-67 LATISSIMUS DORSI MUSCLE FOR CARDIAC ASSIST

tendons . If direct trauma is responsible for damage to the transposed latissimus dorsi muscle, one would suspect it to be more pronounced in the distal part of the muscle because, compared with the pedicle, this part is more intensely handled and distorted during the operation . This was not the case in our goats. In addition, no damage was seen in the in situ latissimus dorsi muscle at the site of the inserted intramuscular electrodes even though this part of the muscle is handled during electrode implantation . Ischemin of the latissimus dorsi muscle is another possi- ble cause of deterioration of muscle structure . As demon- strated, no increase in capillary/fiber raim was seer, in the wrapped and stimulated latissimus dorsi rc.dscles in contrast to a significant increase it, the in situ stimulated muscles . liudlicka and Price (30) also demonstrated a significant increase in capillary/fiber ratio in normally perfused in situ stimulated tibillis anterior and extensor digitorum muscle in rats. Muscles with limited blood supply, however, failed to demonstrate an increme in capillary/fiber ratio together with a significantly lesser increase in blood flow during stimula- tion . Severe loss of muscle mass and atrophy of the muscle FiL SC 6. Electron micrograph from a wrapped and long-term stim- fibers together with partial necrosis of the central part of the ulated latissimus dorsi muscle fiber demonstrating severe disorgani- muscles were seen in their animals . On the basis of these zation and destruction of the individual myofibrillar bundles . (x 14,000 reduced by 30% .) observations. they considered the increased blood flow in chronically stimulated muscles as the main factor responsi- ble for the increase in capillary/fiber ratio . Thus, blood How tended to be more pronounced in the wrapped latissimus in the, majority of the wrapped latissimus dorsi muscles dorsi muscle compared with the in situ-stimulated muscle . seemed to become impaired . This is supported by the As reviewed by Salmons and Henriksson (27), in situ observation of severe abnormalities in a large number of stimulation of the latissimus dorsi muscle can be performed larger arteries in the latissimus dorsi muscles . without detrimental effect on muscle structure . This finding Again, one can only speculate about possible causes . is supported by our data, which suggest that deterioration of Although precaution was taken to create a window in the left goat latissimus dorsi muscle structure after cardiomyoplasty thoracic wall by resecting a part of the seccad rib seemingly probably results from the distal dissection followed by the sufficiently large for the muscle pedicle, kiroking of the wrapping around the heart . Subsequent electrical stimula- thoracodorsal artery at the entrance of the left latissimus tion may have played an additive role in the progressive dorsi muscle in the thorax may have occurred, leading to a destruction of these latissimus dorsi muscle fibers . Compa- reduction in blood supply . rable negative effects on skeletal muscle stricture have also A direct effect from the electrical current on the vessel been described (28) in transposed and electrically stimulated wall of the larger arteries cannot be excluded . In 1979, Betz canine sartorius muscle used as a neosphincter around an and Schlote (31) described the proliferation of smooth mus- intestinal loop . cle cells in the arterial wall after electrical stimulation. That Possible causes of destruction of latissimus dorsi muscle these vessel abnormalities were not seen in the in situ structure. One can only speculate ors possible causes for the stimulated group could have been due to the lower voltage observed destruction . Direct injury to the muscle during the used. surgical procedure is one possible cause . In that case, Another possibility for the negative effect on blood sup- however, one would expect to see signs of destruction of ply could be the stimulation protocol that is used in fully muscle fibers at an earlier stage during the conditioning transformed latissimus dorsi muscles after a conditioning period, possibly followed by regeneration of muscle cells . period of 12 weeks. According to this protocol, pacing of the Our histologic data from a small number of goats killed soon muscles is finally done with a burst of 185 ms and a burst after the wrapping procedure do not support that mecha- frequency of 30 Hz, w6ich is synchronized to every heart- nism. There were signs of regeneration of individual muscle beat. The maximal number of latissimus dorsi muscle con- fibers, its shown by the periodic acid-Schiff staining in our tractions is programmed at 100/min . We (32) have found that goats, but this did not appear to be sufficient to keep a larger this rate of latissimus dorsi contractions is too hig) to allow part of the muscle vital . Guelinckx et al. (29) demonstrated sutfi4.c::t %jilt. . -r relaxation, thus possibly hampering 'atis- that handling of a transplanted rabbit rectos femoris muscle simus dorsi muscle blood flow . This in partiWar holds true was the cause of damage to muscle adjacent to the repaired for the fully transformed latissimus dorsi muscle, which is LUCAS ET AL . JACC Vol . 22, No . 3 766 LATISS[MUS DORSI MUSCLE FOR CARDIAC ASSIST September 1993 :758-.67 almost entirely composed of slowly contracting and relaxing infiltration of histiocytic cells and fat substitution of myo- type I muscle fibers. Geddes et al . (33) also described an fibrillar bundles . Stretch also produces a positive effect on optimal stimulation rate for maximal muscle blood flow in muscle blood flow (42). However, "too much" stretch of the paced c nine latissimus dorsi and rectus abdominis muscles . wrapped latissimus dorsi muscle could lead to impairment of Pacing the muscle above this rate resulted in a decrease in cardiac filling . The optimal amount of stretch still remains muscle blood flow . one of the many problems that has to be solved to obtain The studies of Barclay (34) and Gorman et al . (35) stress more satisfying results in cardiomyoplasty . the need to pay attention to maximal blood supply to the Conclusions. Our observations on dynamic cardiomyo- latissimus dorsi muscle because of the effect of blood flow on plasty in goats with a follow-up period of ~ 12 weeks show fatigue and force and the occurrence of severe skeletal that support of both normal and depressed cardiac function muscle "stunning" after 3D min of hypoperfusion . The can be obtained as long as the wrapped and electrically possibility that electrical stimulation of the wrapped latissi- stimulated latissimus dorsi muscle remains vital . These mus dorsi muscle may add to muscle ischemia is supported observations stress the need for further research to elucidate by a report by Hanxlikova and Gutmann (36). They de- and possibly eliminate the factors responsible for the pro- scribed complete restoration of contractile and histochemi- gressive deterioration of the transposed muscle . In our cal properties of rat soleus muscle after 30 days of complete opinion, such research should concentrate on the optimal interruption of main arterial and collateral blood supply . surgical technique with respect to stretch and fiber orien€-- These muscles were not subjected to electrical stimulation as tion of the transposed muscle and the best stimulation is the case in dynamic cardiomyoplasty . Guelinckx (37) also protocol after wrapping. Finally, damage to the nerve demonstrated that transposition without electrical stimula- branches should be avoided . tion of the rabbit latissimus dorsi muscle to the thoracic wall with an intact neurovascular pedicle was possible without detrimental effects to the latissimus dorsi muscle structure . References ischemia could also have been responsible for the de- struction of nerve branches, resulting in a diffuse spread of I. Carpentier A, Chachques JC . Myocardial substitution with a stimulated skeletal muscle : first successful clinical case (letter) Lancet 1985 ;1 :1267 . acetylcholinesterase enzyme activity along the entire cell 2 . Rossi MA, [haile DM, Souaa DRS, et al. Dynaunic cardiomyoplasty in membrane . Subsequent denervation could have led to the chronic Chagas° heart disease: clinicopathological data. Ann Torac Surg decay of a large number of muscle fibers, with subsequent 1991 ;51 :649-51 . replacement by fat cells as described by Schmalbruch (26). 3, Magovern GJ, Heckler FR . Park SB, et al . Paced skeletal muscle for dynamic cardiomyoplasty, Ann Thorac Surg 1988 ;45 :614-9. Direct trauma to the large nerve branches in the pedicle 4 . Molteni L, Almada H, Ferrern R . Synchronously stirnulatcd skeletal near the intramuscular electrodes concomitant with an in- muscle grail for left ventricular assistance . .1 'rhorac Carrdiovasc Surg crease in endoneural connective tissue is another possible 1989 ;97 :439-46. explanation for the appearance of signs of denervation . 5, Moreira LFP, Stc1f NAG, Bocchi EA, et al . Latissimus dorsi cardio . myoplasty in the treatment of patients with dilated cardiomyolrathy . Again, the absence of these signs of damage in the in situ- Circulation 1990,82(suppl IN):IV-257®63. stimulated latissimus dorsi muscles could be attributed to the 6 . Chachques JC, Grandjean P, Schwartz K, ct al . Effect of latissimus dorsi lower pacing voltage . dynamic cardiomyoplasty on ventricular function, Circulation 1988 ; 78(suppl 110:111.203--16. Transposition ofthe distal part of the muscle around the 7 . Carpentier A, Chachques JC . Latissimus dorsi cardiomyoplasty to in- heart is another possible cause. This maneuver results in an crease cardiac output . In: Rabago G, Cooley DA, eds. orientation of the individual muscle fibers that is quite Replacement: Current Status and Future Trends. Mount Kisco, NY : Future, 1987 :473-86. different from the original one . Our electron microscopic 8 . Anderson WA . Andersen JS, Acker MA, et al. Skeletal muscle grafts photographs support this point as they demonstrate com- applied to the heart : a word of caution . Circulation 1988 ;78(suppl [II): []I- plete loss of orientation, even of the individual myofibrils 180-90. within the separate muscle cells . This loss of orientation 9 . Chagas ACP, Moreira LFP, da Luz PL, et al . Stimulated preconditioned skeletal muscle cardiomyoplasty : an effective means of cardiac assist . could also account for the histologic differences between our Circulation 1989 ;80(suppl Il1) :III-202-8 . latissimus dorsi muscles wrapped around the heart and 10 . Moreira LFP, Seferian P, Edimar A. Survival improvement with dynamic latissimus dorsi muscles used in the so-called skeletal muscle cardiomyoplasty in patients with dilated cardlomyopathy . Circulation ventricle (38). In the latter application, the original direction 1991 ;84(suppl III) :[I[-296-302 . 11 . Almada H, Molteni L, Ferreira R, et al . The value of the echo-Dopp4;: in of the muscle fibers is preserved in a more optimal way . cardiomyoplasty procedures . J Cardiac Surg 1991 ;6(suppl):113-8 . Loss of original stretch of the muscle when transposed 12. National Academy of Sciences . Guide for the Care and Use of Laboratory around the heart may play a role in the loss of muscle Animals . Washington, DC : DHHS publication no. NIH 85-23, revised 1985 . vitality. This results in an increase in connective tissue (39), 13. Lucas CMHB, Cheriex EC, vat der Veers FH, et al . Itttiprarninu induced which can be prevented by electrical stimulation according heart failure in the dog; a model to study the effect of cardiac assist to Williams et al. (40). Unfortunately, this preventive effect devices . Cardiovasc Res 1992 ;28:8-9 . was not seen in our wrapped latissimus dorsi muscles . 14. Schaper W, Lewi P, Jagerieau AHM . The determination of the rate of Radermecker et al change of the left ventricular pressure (dPfdt) . Arch Kreislaufforsch . (41) demonstrated that loss of stretch of 1965 ;46 :27-41 . in situ-stimulated goat latissimus dorsi muscle results in 15. Lucas CMHB, Havenith M, van der Veen FH, et al . Changes in canine !ACC Vul . 22, No . 3 LUCAS ET AL . September 1993 :758-67 LATISSINIUS DORSI MUSCLE FOR CARDIAC ASS' ST

latissimus dorsi muscle during 24 weeks of continuous electrical stimula- 30 . Fludlicka 0, Price S . The role of blood flow and/or inns tion. J Appl Physiol 1992 ;72 :828-35 . capillary growth in chronically stimulated fast muscles . a 16. Khaw BA, Mattis JA, Melincoff 0, Straus HW, Gold HK, Haber E . 1990 ;417 :67-72 . Monoclonal antibody to cardiac myosin : imaging of experimental myo- 31 . Betz E, Schlote W. Responses of vessel walls to zhrm_ically applied i ardial infarction. Hybridoma 199d;3 :11-23 . electrical stimuli . Basic Res Cardiol 1979;74:10-20. 17 . Havenith MG, Visser R, Schrijvers-van Schendel JM, Rosman FT . 32 . Lucas CMHB, Van der Veen Fl'!, Cheriex hC, Van Oasmen V, Penn Muscle fiber typing in routinely processed sheelai muscle with monoclo- OCKM, Wellens HJJ . The importance of muscle relaxation in dynamic . Histochemistry 1990-,93 :497-9, nal antibodies cardiomyoplasty . PACE 1,992 ;15 :1430-6 . 18 . Junquira LCU, Montes GS, Sanchez EM . The influence of tissue section 33 . Geddes LA, Badylak SF, Wessale IL, et a] . The use of electrically thickness on the study of collagen by the picrosirius-polariwion method. stimulated skeletal muscle to pump blood . PACE 1990 ;13 :344-62. Histochemistry 1982 ;74:153-6. 34 . Barclay JK . A delivery-independent blood flow effect on skeletal muscle 19 . Van Krimpen C, Smits JFM, Cleu1jens JPM, et al . DNA synthesis in the fatigue . J Appi Physiol 1986;61 :1084-90. nou-ini~a,cied cardiac inici stiiituu after left coronary artery ligatioa in the rat : effects of captopril . J Mol Cell Cardiol 1991 ;23 :1245-53. 35 . Gorman MW, Barclay JK, Sparks HV . Effects of ischemia on V0, tension, and vascular resistance in contracting canine skeletal muscle . 20. Bioniochanical Cardiac Asy ;st : OlAinmyoplasty and Muscle-Powered J Appi Fhysiol 1988*,65:1075-81 . Devices . Chin RC-J, ed . Mount Kisco, NY : Futura, 1996 :1-18. 36. Hanzlikova V, Gutmann E . Effect of ischerma on contractile and hislo- 21 . Transformed Muscle For Cardiac Assist and Repair. Chin RC-1, Bourgeois chemical properties of the rat soleus muscle . PlIagers Arch IT/9 ;3'79:2(P)- IM, eds . Mount Kisco, NY: Futura, 1990:31140 . 14, 22 . Cardiomyoplasty . Carpentier A, Chacliques JC, GrandJican P, eds . Mount 37. Guelinckx PJ . Free microneurovascular muscle grafts : Deficits resulting Kisco, NY: Futura, 1991 :195-8 . from transplantation and speculations for myocardial repair . I Cardiac 23 . Stephenson LW, ed . 4th World Symposium on Transformed Skeletal Surg 1 0,91 ;6(suppl) :190-4. Muscle for Cardiac Assist . Mount Kisco, NY : Futura, J Cardiac Sarg 199 1 ;6(suppl): 129 -3 1 . 38 . Anderson DR, Pozhettino A, Hammond RL, et al . Autogenously lined skeletal muscle ventricles in circulation . J 'fhorac Cardiovasc Surg 24 . Jatene AD, Moreira U11 P, Stolf NAG, et al, left ventricular function changes after caydiomyoplqsty in patients with dilated cardiomyoplasty . 1991 ;101 :661-'10, J Thorne Cardiovasc Sura, 1991 ;W2:!32-9 . 19, Williams PE, Goldspink G . Connective tissue changes in pmmobbilhzctl 25 . Dubowitz V . Muscle Biopsy : A Practical Approach . 2nd ed . Suffolk, muscle . 3 Anat 1934 ;08:3143-50, England: Baillicre -a indall (W13 Saunders), 1985 :289-404 . 40 . Williams FE, Caumcw T, Ltwey EG . Goldspink C . Te importance of 26 . Scbmalbruch H . The effect of peripheral nerve injury on immature motor stretch and contractile activity in ?hce prevention of connective tissue and sensory neurons and oi : muscle fibres . Rev Neurol 1988 ;144.°.721-9 . accumulation . J Anat 1988 ;158:109-14. 27 . Salmons S, Henriksson J . !he adaptive response of skeletal muscle to 41 . Raderitiecker MA, Sluse FE, Focant B, Rezfu*k H, Fourny J, Linnet R . increased use . Muscle Nerve 1981 ;4:94-105 . Influence of tension reduction and peripheral dissection on histologic, 28. Hallan RI, Williams NS, Hutton MILE, et al . Electrically stimulated biochemical and bioenergetic profiles, and kinetics of skeletal muscle sartorius neosphincter : canine model of activation and skeletal muscle fast-to-slow transformation .,; Cardiac Surg 1991 ;6(suppl):195-203 . transformation . Br J Surg 1990;77 :208-13 . 42. Leighton WD, Russell RC, Feller AM, Eriksson E, Mathur A, Zook EG . 29. Guelinckx PJ, Faulkner JA, Essig DA . Neurovascular-aoastonioF!~d mus- Experimental pretransfer expansion of free-flap donor sites : IL PhYsir+ cle grafts in rabbits : functional deficits result from tendon ;epais - Muscle ogy, histology and clinical correlation . PLst Recoostr Surg 1985 ;82 :76- Nerve 1988 ;11 :745-51 . 82 .