Evaluation of the pulsatility of a new pulsatile left ventricular assist device—the integrated cardioassist catheter—in dogs

Evaluation of the pulsatility of a new pulsatile left ventricular assist device-the integrated cardioassist catheter-in dogs A new pulsatile left ventricle-femoral artery bypass system (integrated cardioassist catheter system) has been developed for rapid, percutaneous insertion as a left ventricular assist device. Previous experiments revealed its superiority over the intraaortic balloon pump system in maintaining the peripheral circulation and in improving myocardial blood flow and afterload. Our objective was to determine whether the pulsatility of left ventricular bypass of this system would be preferable for maintaining the peripheral circulation and managing the ischemic myocardium as compared with nonpulsatile left ventricular bypass. Ten dogs with profound heart failure were supported by this system. Their hemodynamic status and myocardial blood flow were measured under control, nonpulsatile left ventricular bypass, or synchronous pulsatile left ventricular bypass. Significant differences between the nonpulsatile bypass group and the pulsatile bypass group were observed in the mean increase in aortic pressure (3.5% versus 22.2 %, respectively; p < 0.001), total cardiac output (13.0% versus 21.7%; p = 0.004), and myocardial blood flow (9.5% versus 21.8%; p < 0.001). No differences were found between groups in the decrease in left atrial pressure (-20.2 % versus -20.2 %; p > 0.05). The ratio of diastolic time index/tension time index was shown to be improved significantly in the pulsatile bypass group compared with that of control and nonpulsatile bypass groups (p < 0.001). Thus, the pulsatility of the integrated cardioassist catheter system may support the peripheral circulation and improve the myocardial blood flow and oxygen supply/demand ratio. (J THoRAc CARDIOVASC SURG 1994;107:569-75)

Hirofumi Ide, MD,a Atsushi Yamaguchi, MD,b Takashi Ino, MD,b Hideo Adachi, MD,b Akihiro Mizuhara, MD,b Kohji Kawahito, MD,b Hiroshi Matsumoto, MD,c and Iwao Fujimasa, MD,c Tokyo and Saitama, Japan

We

developed a new pulsatile left ventricle (L V)femoral artery (FA) bypass system that used a transaortic valve-LV drainage catheter with an intraaortic balloon situated in the descending aorta (integrated cardioassist catheter [ICAC]), a centrifugal pump and an From the Department of Thoracic and Cardiovascular Surgery, Kyorin Medical School, Tokyo; the Department of Cardiovascular Surgery, Omiya Medical Center, Jichi Medical School, Saitarna," and the Research Center for Advanced Sciences and Technology, University of Tokyo, Tokyo," Japan. Received for publication Jan. 13, 1993. Accepted for publication May 24, 1993. Address for reprints: Hirofumi Ide, MD, Department of Thoracic and Cardiovascular Surgery, Kyorin Medical School, 6-20-2, Shinkawa, Mitaka-shi, Tokyo, 181, Japan. Copyright © 1994 by Mosby-Year Book, Inc. 0022-5223/94 $1.00 +.10

12/1/49788

intraaortic balloon pump (IABP) console (Fig. I). This new device system can be rapidly placed in an emergency, unlike the conventional LV assist techniques, such as left atrium (LA)-FA bypass 1-4 or even other LV-aorta bypass techniques.l which can be difficult to establish. This system would be useful in treating patients with not only profound LV dysfunction after cardiac operations as do other LV assist devices.v4, 6-8 but also with an evolving myocardial infarction or with cardiogenic shock, which necessitate the rapid insertion of an LV assist device and coronary intervention. Particularly because of recent advances in interventional therapy, such as thrombolytic therapy, angioplasty, and the surgical revascularization of the patient with ischemic heart disease, our interests have centered on the limitation of myocardial infarct size. Recent studies- 7, 9, 10 have shown the superiority of pulsatile LV

569

570

The Journal of Thoracic and Cardiovascular Surgery February 1994

Ideetal.

bypass and pulsatile venoarterial (VA) bypass I I, 12 in unloading the heart and providing myocardial salvage during early reperfusion. The ease of use and the possibility of percutaneous implementation with this new ICAC system and the initiation of early stage reperfusion with pulsatile flow would offer benefits for reducing infarct size. We have already reported':' that the ICAC system as a pulsatile LV bypass is preferable in ameliorating the myocardial blood supply/demand relationship and maintaining peripheral circulation compared with IABP support alone. We designed the present experiment to investigate whether the synchronized diastolic pulsatility of this ICAC system would provide benefits concerning LV unloading and myocardial blood flow, as well as providing hemodynamic stability, as compared with nonpulsatile LV bypass.

Materials and methods Ten mongrel dogs weighing 18 to 28 kg were anesthetized with intramuscular ketamine hydrochloride (20 mg/kg) and intravenous pentobarbiturate anesthesia (25 mg/kg). After intubation, respiration was controlled with positive-pressure mechanical ventilation with a mixture of room air and low-flow oxygen. After making a transverse thoracotomy in the fourth intercostal space, a catheter was placed in the LA and in the LV via the LV apex and left carotid artery and connected to a pressure transducer for monitoring pressures. The diastolic pressure time index and the tension time index were calculated from simultaneous measurements of LV and aortic pressure. The subendocardial oxygen supply/demand ratios were estimated by diastolic pressure time index/tension-time index calculations. Electromagnetic flow probes were secured around the main pulmonary artery and the bypass tube, and the total cardiac output (output of LV + bypass flow) and bypass flow were measured. A laser Doppler flowprobe was placed on the LV free wall to measure myocardial blood flow with the value expressed in milliliters per minute per 100 gm myocardial weight." The prototype ICAC used in the canine experiments measured the following: outer diameter 18F, inner diameter 9F, balloon size 20 or 30 ml, and length 60 em. The catheter tube was made of polyvinyl chloride. After the lower end of the abdominal aorta was exposed by median laparotomy and the right carotid artery was exposed through an incision in the neck, heparin (2.5 mg/kg) was administered intravenously. The ICAC was inserted via an abdominal cutdown to the descending aorta. The guidewire with its soft tip and a pigtail catheter were introduced into the LV through the ICAC cannula with the aid of fluoroscopy. The tip of the ICAC was advanced into the LV cavity, guided by the pigtail catheter. An arterial perfusion cannula (14F) was then introduced into the right carotid artery. Both cannulas were then connected to a tube (318 inch) of the bypass circuit that contained a centrifugal pump (Biopump; Medtronic Bio-Medicus, Eden Prairie, Minn.) primed with heparinized saline solution. After obtaining baseline measurements, we administered intravenously a (J-blocker (propranolol hydrochloride, 4 to 12

mg) and low-molecular weight dextran to achieve basic biventricular failure and also to control heart rate. In addition, we ligated the branches of left anterior descending coronary artery to induce LV dysfunction of various levels of severity, as determined by hemodynamic monitoring mentioned previously. To control the heart rate, we performed right atrial overdrive pacing after administration of (J-blocker or right ventricular pacing with atrioventricular block formation 15 to avoid the influence of the heart rate. In the various states of heart failure under control, nonpulsatile LV bypass, and pulsatile LV bypass measurements were obtained after hemodynamic stability was confirmed. Pulsatile LV bypass was performed by exerting an intraaortic balloon console with the aortic pressure waveforms synchronized to the heart beat on the electrocardiogram, which was adjusted to provide maximal diastolic phase pulsatility, and compared with nonpulsatile LV bypass under the same bypass flow.The bypass flow ranged from 500 to 1700 ml/rnin. Each group was randomly and sequentially prepared and compared at the same heart rate to avoid the influence of hemodynamic changes over time. All animals were treated humanely in strict accordance with the "Guide for the Care and the Use of Laboratory Animals" published by the National Institute of Health (NIH publication No. 85-23, revised 1985). Values were expressed as mean ± standard deviation. Paired or nonpaired Student's t tests (double-tailed) were used to compare differences between groups, with the use of the commercially available Stat View software package (Brain Power Inc., Calabasas, Calif.); p values less than 0.05 were considered to be statistically significant.

Results With the method of pigtail catheter-guided insertion of the ICAC with the aid of fluoroscopy, it took about 3 minutes to place the ICAC into the LV cavity properly. We encountered no complications concerning the LV wall, the aortic valve, or the aortic wall during experiments; this was confirmed at postmortem examination. The baseline hemodynamic values after full instrumentation are as follows: mean aortic pressure 116.8 ± 8.7 mm Hg, total cardiac output 2.7 ± 0.4 L/min, left atrial pressure 4.0 ± 1.6 mm Hg, mean heart rate 114.6 ± 11.2 beats/min. The mean values of the random and sequential measurements of the hemodynamic parameters in the three groups are summarized in Table I. The.mean bypass flow set up in each parameter is shown in Table I, and its mean LV bypass flow was approximately 40% of the normal cardiac output before ,a-blocker administration. The mean aortic pressure showed significant changes from baseline in the pulsatile group (22.2% increase; p < 0.001 versus control), whereas mean aortic pressure did not fluctuate significantly in the nonpulsatile group (3.5% increase; p > 0.05 versus control), leading to a significant improvement of mean aortic pressure in the pulsatile group as compared with the nonpulsatile group (p < 0.00 I )(Fig. 2). Concerning the total cardiac output,

The Journal of Thoracic and Cardiovascular Surgery Volume 107, Number 2

the use of both nonpulsatile and pulsatile LV bypass at the same bypass flow resulted in a significant increase in total cardiac output (13% increase; p = 0.001 versus control, and 21.7% increase; p < 0.00 I versus control, respectively), with significant differences observed between the nonpulsatile and pulsatile groups (p = 0.004). Furthermore, the LV bypass flow in both groups (mean bypass flow 1090 ± 420 ml/min) exceeded the actual increase of total cardiac output in any measurements in this series, indicating a reduction of real LV output and leading to volume unloading of the LV in both groups. As for the preload of LV estimated by left atrial pressure, the use of both the pulsatile and non pulsatile LV bypass systems led to a significant reduction in left atrial pressure. No significant differences were observed between these two groups (p > 0.05), indicating the predominant effect of the LV flow support itself in comparison with its pulsatility. Myocardial blood flow measured by the epicardial laser Doppler method showed that, in both groups, the LV bypass contributed to the improvement of myocardial blood flow (nonpulsatile group 9.5% increase versus controlp < 0.001, and pulsatile group 21.8% increase versus control; p < 0.001). This increase in myocardial blood .flow in the pulsatile group significantly exceeded that of the nonpulsatile group (p < 0.001), showing a preferable effect of pulsatility on the myocardial oxygen supply. Moreover, calculation of diastolic pressure time index/ tension time index ratio as an index of the myocardial oxygen supply/demand ratios, as well as of LV pressure loading status, elucidated a marked increase in the diastolic pressure time index/tension time index ratio only in the pulsatile group (p < 0.001 versus control and the nonpulsatile group; nonpaired t test) as shown in Fig. 3. In contrast, no significant fluctuation of diastolic pressure time index/tension time index ratio was observed in the nonpulsatile group (p> 0.05 versus control). These observations show the benefit of the added pulsatility on LV bypass in that it ameliorates the myocardial oxygen supply /demand ratios. Discussion Previous reports- 3. 7-9 have demonstrated that LV bypass, mainly LA-FA bypass, primarily maintains peripheral circulation, irrespective of pulsatile or nonpulsatile flow support and is undoubtedly effective in managing profound LV impairment. To meet the demands of recent advances in interventional cardiology or cardiac surgery, the following issues have precluded widespread clinical application of the various kinds of conventional LV bypass methods. The ideal LV bypass methods should be implemented rapidly, and preferably, percutaneously, for the treatment

Ide et al.

57 I

intraaortic balloon

double lumen catheter for counterpulsation and LV bypass

I ABP driVing system centrifugal pump

Fig. 1. Clinical applicationof the ICAC. The perfusion cannula can be also be placedfrom the subclavianartery. AV, Aortic valve; L V left ventricle. of abrupt cardiogenic shock or an evolving myocardial infarction because these conditions can occur not only in the surgical suite after cardiotomy but also in the catheter laboratory and intensive care unit. However, the conventional LV bypass methods, 1-5,7 with their inherent mechanisms that necessitate surgical or complicated procedures for placement, limit their prevalence, unlike IABP. As Axelrod and associates I I, 12 pointed out, percutaneous cardiopulmonary bypass (VA bypass) may surpass the LV bypass system in this respect. The new ICAC system for LV bypass offers ease of percutaneous insertion. The present experiments in dogs demonstrated that the draining tube could be placed in the LV through the aortic valve in about 3 minutes with the aid of fluoroscopy and an internal guidewire catheter, as previously mentioned, without complications. In theory, in the clinical setting, this device could be installed promptly at the bedside with the aid of echocardiography and initiated quickly. Thus, this new system, unlike conventional LV bypass, overcomes the difficulty of insertion that is associated with the transatrial septal puncture required with LA-FA bypass,': 2,10 which precludes its widespread application. The ICAC system can be implemented with a centrifugal pump and an IABP console, in contrast to

572

The Journal of Thoracic and Cardiovascular Surgery February 1994

Ideetal.

~

groupNP group P

30 - A - - - - - - - - - r - - - - - - , - - - - - -.....- - - - - . , 20

glO c 8

-20 -30 . . L.!;

-20.2

====

: : : !:

====

mAoP

: : : !:

-20.2

====

LAP

total CO

: : !: :

====

~

myocardial Bloodflow

Fig. 2. Comparison of percentage of changes of hemodynamic parametersbetween pulsatile (P) and nonpulsatile (NP) groups. NS, Not significant; AoP, aortic pressure; CO, cardiac output; LAP, left atrial pressure. Table I. Comparison of hemodynamic variables No. of samples

mAoP (mm Hg)* Total CO (L/min)t LAP (mm Hg):j: Myocardial blood flow (ml/min/100 gm)§

27 24 26 22

Prebypass 76.5 1.15 8.9 70.3

± 26.9 ± 0.5 ± 3.8 ± 29.2

Nonpulsatile group 79 1.3 7.1 77

± 27.7 ± 0.45 ± 2.6 ± 29.6

Pulsatile group 93.5 1.4 7.1 85.6

± 27.3 ± 0.43 ± 2.7 ± 32.7

Mean bypassf/ow (mljmin) 990 1090 1080 1270

± 390 ± 420 ± 400 ± 900

p Value

<0.001 0.004

NS

<0.001

Paired t tests were used for comparison; p values on the table show the comparison between non pulsatile and pulsatile groups. Prebypass, Before bypass; AoP, aortic pressure; CO, cardiac output; LAP, left atrial pressure; NS, not significant. "Baseline value was 116.8 ± 8.7 mm Hg: mean heart rate was 114.6 ± 11.2; p = 0.084 (nonpulsatile group versus prebypass) and p versus prebypass).

< 0.001 (pulsatile group

tBaseline value was 2.7 ± 0.4 Lyrnin: p = 0.001 (nonpulsatile group versus prebypass) and p < 0.001 (pulsatile group versus prebypass). :j:Baseline value was 4.0 ± 1.6 mm Hg: p = 0.008 (nonpulsatile group versus prebypass) and p = 0.015 (pulsatile group versus prebypass). §p < 0.001 (nonpulsatile group versus pre bypass and pulsatile group versus prebypass).

other methods. 2, 5, 8,9, II, 12, 16 Another clinical advantage of the ICAC for LV-FA bypass is that it does not induce stagnation of the blood in the LV, as seen with traditional LA-FA bypass, which presents the risk of LV thrombus formation and subsequent systemic embolism. However, the recently developed intraarterial axial-flow blood pump, the Hemopump (Nimbus Medical, Inc., Rancho Cordova, Calif.) may have some superiority over the ICAC8 because it necessitates minimal surgical procedure without thoracotomy like the ICAC system and also supplies LV bypass flow support of up to 4 Lzrnin; once it is introduced, its drive shaft is only 9F, preventing lower limb ischemia of the inserted site, unlike the ICAC, although it requires its specificpower console and provides nonpulsatile flow.

Thus, theoretically the ICAC system has some potential superiority over the conventional LV assist device l -3, 5, 7-10, 16 for practical use. The other pulsatile bypass systems- 5, 7, 9-12,16 have adopted an intrinsic way of using the pulsatile pump to achieve pulsatility for perfusion as a possible alternative to the conventional intraaortic balloon counterpulsation method we used. However, with regard to simplicity and effectiveness, the following disadvantages exist for an intrinsic method in comparison with conventional intraaortic balloon pulsatilit y2, 7, 9-12,16: 1) intrinsic method necessitates a pulsatile pump system unlike the widely used IABP console, (2) effective pulsatility would be obtained only with a large perfusion catheter, (3) synchronized pulsation would not be possible in the presence of severe tachycardia, and (4)

The Journal of Thoracic and Cardiovascular Surgery Volume 107, Number 2

perfusion from the FA would not produce intense pulsation at the aortic coronary ostium, where maximal pulsatility is required. Therefore, we consider our extrinsic approach to be preferable. In addition, the ICAC bypass allows easy weaning to IABP counterpulsation. The integration of the balloon portion and the draining cannula is indispensable, in that the independent insertion of the cannula and IABP creates two problems. First, the bilateral FAs are occupied by perfusion and draining cannulation, and the IABP would not be inserted percutaneously. Second, the balloon would not be implemented effectivelyin the descending aorta, where a drainage tube is placed as an obstacle in the event of independent use. As for myocardial salvage after the revascularization of the ischemic myocardium, the efficacy of cardiac assist devices has been evaluated mainly according to whether they have the ability of limiting the size of myocardial infarct.'" 12, 17-20 Instead, this experimental study was designed to evaluate the pulsatility of ICAC system as a LV-aorta (FA) bypass in view of myocardial oxygen supply and unloading status especially for salvaging ischemic myocardium and hemodynamic stability. In this canine model, two points should be taken into consideration. First, we adopted the insertion site of the left carotid artery instead of the site distal to the intraaortic balloon, simulating the insertion site of the perfusion catheter via the subclavian artery, not the FA. The main reason was to avoid the technical difficulty of returning the blood from the FA without causing lower limb ischemia, The preliminary experiments comparing these two insertion sites showed no significant difference in the aortic root waveforms and actual pressure, indicating that the one-aortic chamber model exerts minimal effect on its perfusion site, although there might be some difference in clinical situations. Second, we administered ,6-blocker and ligated the coronary arteries in the LV regions to achieve various degrees of biventricular failure to simulate the clinical condition of LV-dominated heart failure. Its extent of cardiac failure compared with that of normal canine hearts after full instrumentation for this experiment has been previously described and is shown in Table 1. The hemodynamic findings estimated from mean aortic pressure and total cardiac output demonstrated the significant benefit of adding electrocardiogram-gated pulsatility on the maintenance of the peripheral circulation, provided that the ICAC system is used under certain LV dysfunction protocols, such as ours; the results in other settings might not agree with other experimental results of normal cardiac function 10, 21 or even the LAD ligation model. 19 Those studies have shown no significant hemodynamic improvement. In their models, LV bypass flowand LV output itself are considered to have the rela-

Ide et al.

573

0,8

o

~

0,6

0.2

o

Control (N=10)

I

NS

group NP (N=20) group P (N=30)

II

pco.ool

I

p-co.oo l

Fig. 3. Comparison of diastolic pressure time index (DPTI) /tension time index (TTl) values among the three groups. Mean bypass flow of the nonpulsatile (NP) and pulsatile (P) groups were 1040 ml/rnin and 1060 ml/rnin, respectively (statistically not significant). Vertical bars show the standard deviation. Nonpaired t test was used for comparison. NS, Not significant.

tionship of mutual compensation, and pulsatile LV bypass itself brings about no significant hemodynamic improvements. In our model of cardiac failure, pulsatile LV bypass increased the aortic pressure by 22.2% and the cardiac output by 21.7%, with a flow support of 990 ± 390 mljmin and 1090 ± 420 ml/rnin, respectively. In the clinical setting, these values would theoretically be obtained under a flowsupport of approximately 2 to 2.5 Ljmin (about 40% of the normal cardiac output) and preclude the ICAC size and length for clinical application. Although those maximal flows might be sufficient for a clinical report," the potential disadvantage of the restricted maximal flow determined by the long draining tube of a limited size is shown to be compensated by its pulsatility from our experiment. In the clinical setting, as is expected, this catheter should be long, inserted from the FA to the LV, and preferably smaller in its outer diameter to prevent lower limb ischemia of the inserted site, thus limiting its maximum bypass flow. Our experimental data show also that a maximal flow of more than 2 Ljmin is necessary to achieve the same hemodynamic effect as was discussed previously. With these requirements in mind, we created a clinical model of the ICAC. In brief, its length is 90 em, outer diameter 20F, and inner diameter 15F. It is made of heparin-coated polyurethane. In vivo and in vitro flow characteristics in dogs showed that its maximal flow is of more than 2 Ljmin without cavitation in a centrifugal pump head and hemolysis (unpublished data). Weare now ready to use it clinically.

574

The Journal of Thoracic and Cardiovascular Surgery February 1994

Ide et at.

No significant difference in left atrial pressure was observed between the pulsatile and nonpulsatile groups, indicating that the reduction in the LV preload could be attributed to the flow support itself, which is in agreement with previous study. 10 Pulsatility is considered to have little contribution. This phenomenon and its explanation are supported by the results of other investigators- 21 and by the results of our previous study.l ' which examined and compared the IABP and ICAC and showed the superiority of the!CAC as a pulsatile LV-aorta bypass over the IABP alone and in which pulsatility itself did not reduce left atrial pressure values. With regard to the salvage of the ischemic myocardium, we did not evaluate the actual infarct size or infarct size/area at risk ratios during the reperfusion of ischemic myocardium, as is used by other investigators to validate various assist devices.?: 12, 17-20 Instead, we measured the myocardial blood flow and calculated the LV afterload (pressure load), which is the theoretic basis of myocardial salvage. Calculation of tension-time index and myocardial oxygen consumption has been used exclusively* to assess and quantitate the unloading status of the heart. Some articles have shown that a nonpulsatile VA bypass is ineffective for decreasing the tension-time index values 11, 12 and myocardial oxygen consumption, 11, 22 indicating the potential inability of VA bypass.to unload the LV, whereas even a nonpulsatile LA-aorta bypass can reduce tension-time index values'" 17, 18 and myocardial oxygen consumption 17,22-24 and LV stroke work." Thus, Spencer and colleagues'' have applied the LA- FA bypass in the treatment of postcardiotomy LV impairment with good clinical results. Their more recent reports have shown, however, that the pulsatility of the VA bypass is a highly useful adjunct for salvaging the ischemic myocardium.Ui '? Subsequent studies that compared the LV-aorta bypass and LA-aorta bypass revealed an advantage of the former in reducing tension-time index values,'? myocardial oxygen consumption.P and myocardial infarct size. I9, 26 This advantage may be a result of a difference in the efficacy of inducing LV decompression, which supports the preferable effect of our bypass method. It should be noted that nonpulsatile flow delivers its unloading condition secondary to a complete LV decompression.U: 19,21,22,23 In contrast from our findings, adding pulsatility to the LV bypass appears to be an important means of improving the myocardial oxygen supply/demand ratios and the actual myocardial blood flow under incomplete decompression. This outcome is generally consistent with the results of other investigations of pulsatile LV bypass.v 5 Grossi and associates 10, 20 *References 9-12, 17-20, 22-24.

showed a preferable effect ofthe pulsatile LA-FA bypass in reducing tension time index values, epicardial/endocardial flow ratio, and myocardial infarction size, whereas Rose and associates" suggested the inability of pulsatility of a partial, as compared with a total, LV bypass. Others? showed no significant difference of LV afterload estimated by tension-time index measurement between pulsatile and nonpulsatile flow in LA-aorta bypass. The present study showed a significant improvement of myocardial blood supply/demand ratio, as reflected by diastolic pressure time index/tension-time index measurements with its pulsatility even under incomplete LV decompression. Also, our previous study with a similar protocol':' showed a significant increase of diastolic pressure time index/tension-time index ratios as compared with the use of counterpulsation alone, indicating that constant LV flow support may augment the effect of counterpulsation. Furthermore, as mentioned previously, LV output estimated by total cardiac output and bypass flow was shown to decease under LV bypass, indicating that LV bypass, whether pulsatile or nonpulsatile, reduces LV volume loads. In conclusion, the ICAC systems for conducting partial LV-FA bypass ameliorated the myocardial oxygen supply/demand relationship, in addition to maintaining the peripheral circulation in LV-dominated profound heart failure in dogs. We acknowledge the assistance of Muneyasu Saitoh, MD, Professor of the Department of Cardiology, Omiya Medical Center, Jichi Medical School, in statistical analysis. REFERENCES I. Dennis C, Hall DP, Moreno JR, Senning A. Left atrial cannulation without thoracotomy for total left heart bypass. Acta Chir Scand 1962;123:267-79. 2. Glassman E, Engelman RM, Boyd AD, Lipson D, Ackerman B, Spencer Fe. A method of closed-chest cannulation of the left atrium for left atrial-femoral artery bypass. J THORAC CARDIOVASC SURG 1975;69:283-90. 3. Rose DM, Colvin SB, Culliford AT, et al. Long-term survival with partial left heart bypass following perioperative myocardial infarction and shock. J THORAC CARDIOVASC SURG 1982;83:483-92. 4. Rose DM, Colvin SB, Culliford AT, et al. Late functional and hemodynamic status of surviving patients following insertion of the left heart assist device. J THoRAc CARDlOVASC SURG 1983;86:639-45. 5. Igo SR, Migliore JJ, Fuqua JM Jr, Morman J'C, Effects of abdominal left ventricular assist device (AL VAD) on myocardial contractility during acute coronary occlusion. J Surg Res 1974;17;177. 6. Spencer FC, Eiseman B, Trinkle JK, Rossi NP. Assisted circulation for cardiac failure following intracardiac sur-

The Journal of Thoracic and Cardiovascular Surgery Volume 107, Number 2

7.

8.

9.

10.

II.

12.

13.

14.

IS.

gery with cardiopulmonary bypass. J THoRAc CARDIOVASC SURG 1965;12:56-73. Litwak R, Koffsky RM, Jurado RA, et al. Support of severelyimpaired cardiac performance with left-heart assist device following intracardiac operation. Heart Lung 1978; 7:622-6. Frazier OH, Nakatani T, Duncan J, Parnis SM, Fuqua J. Clinical experience with the Hemopump. Trans Am Soc Artif Intern Organs 1989;35:604-6. Laschinger JC, Cunningham IN Jr, Catinella FP, Knopp EA, Glassman E, Spencer Fe. "Pulsatile" left atrial-femoral artery bypass. Arch Surg 1983;118:965-9. Grossi EA, Hunter CE, Culliford AT, Colvin SB, Baumann FG, Spencer Fe. Percutaneous assist device provides simple technique for total left ventricular support. Surg Forum 1985;36:208-10. Axelrod HI, Galloway AC, Murphy MS, et al. Percutaneous cardiopulmonary bypass with a synchronous pulsatile pump combines effective unloading with ease of application. J THoRAc CARDIOVASC SURG 1987;93:358-65. Axelrod HI, Murphy MS, Galloway AC, et al. Percutaneous cardiopulmonary bypass limits myocardial injury from ischemic fibrillation and reperfusion. Circulation 1988; 78(SuppI):148-52. Ide H, Yamaguchi A, Ino T, Matsumoto H, Fujimasa I. Hemodynamic evaluation of a new left ventricular assist device. An Integrated Cardioassist Catheter as a pulsatile left ventricle-femoral artery bypass. Artif Organs 1992; 16:286-90. Mizutani T, Katayama Y, Onoda K, et al. Evaluation of aortocoronary bypass operation using a He-Ne Laser flowmeter; comparison between the internal mammary artery grafts and the sapheneous vein grafts. J Jpn Assoc Thorac Surg 1990;38:389-95. Steiner C, Kovalik TW. A simple technique for the production of chronic complete heart block in dogs. J Appl Physiol 1968;25:631-6.

Ide et al.

575

16. Zwart HH, Kralios A, Kwan-Gett CS, et al. First clinical application of transarterial closed-chest left ventricular (TaCL V) bypass. Trans Am Soc Artif Intern Organs 1970;56:386-91. 17. Pennock JP, Pierce WS, Waldhausen JA. Quantitive evaluation of left ventricular bypass in reducing myocardial ischemia. Surgery 1976;79:523-33. 18. Laks H, Ott RA, Standeven JW, Hahn JW, Blair OM, William VL. The effect ofleft atrial-to aortic assistance on infarct size. Circulation I977;56(Suppl):1138-43. 19. Pennock JL, Pae WE Jr, Pierce WS, Waldhausen JA. Reduction of myocardial infarct size: comparison between left atrial and left ventricular bypass. Circulation 1979; 59:275-9. 20. Grossi EA, Krieger KH, Cunningham IN Jr, et aI. Time course of effective interventionalleft heart assist for limitation of evolving myocardial infarction. J THORAC CARDIOVASC SURG 1985;624-9. 21. Rose EA, Marrin CAS, Bregman 0, Spotnitz HM. Left ventricular mechanics of counterpulsation and left heart bypass, individually and in combination. .J THORAC CARDIOVASC SURG 1979;77:127-37. 22. Salisbury PF, Cross CE, Rieben PA, Lewin RJ. Comparison of two types of mechanical assistance in experimental heart failure. Circ Res 1960;8:431-9. 23. Dennis C, Hall DP, Moreno JR, Senning A. Reduction of the oxygen utilization of the heart by left heart bypass. Circ Res 1962;10:298-305. 24. Pierce WS, Aaronson AE, Prophet GA, Williams DR, Waldhausen JA. Hemodynamic and metabolic studies during two types of left ventricular bypass. Surg Forum 1972;23:176-8. 25. Miller DR. Comparative effects of left atrial or left ventricular bypass on coronary sinus flow and oxygen usage in dogs. Ann Surg 1974;179:830-5.