Pulsatile mode of operation of left ventricular assist devicesand pulmonary haemodynamics

Michael Poullis*

Department of Cardiothoracic Surgery, Liverpool Heart and Chest Hospital, Liverpool, UK

* Corresponding author. Department of Cardiothoracic Surgery, Liverpool Heart and Chest Hospital, Thomas Drive, Liverpool L14 3PE, UK. Tel: +44-151-2281616;fax: +44-151-2932254; e-mail: mpoullis@hotmail.com (M. Poullis).

Received 2 December 2013; received in revised form 18 January 2014; accepted 22 January 2014

Abstract

OBJECTIVES: To determine the effect of differing modes of left ventricular assist device (LVAD) operation: synchronous, independent(asynchronous or pseudosynchronous) or counter pulsation (antisynchronous), on left atrial pressure, pulmonary artery pressure, pulmon-ary blood ow and right ventricular work load, utilizing a previously published electrical analogy of the systemic and pulmonary circulationand the heart.

METHODS: A previously published electrical analogy of the systemic and pulmonary circulation was utilized. The Simulation Package withIntegrated Circuit Emphasis (LTSPICE IV) was utilized. Three LVAD operation mode scenarios were analysed: synchronous, counter pulsa-tion and independent pulsatile. The root mean square of the pulmonary artery pressure (PAP), left atrial pressure (LAP) and pulmonaryblood ow (PBF) were calculated, as was the right ventricular work load.

RESULTS: Counter pulsation LVAD operation resulted in the lowest LAP, PAP, right ventricular work load and the highest pulmonary blood ow.Independent pulsation resulted in the highest LAP, PAP and the lowest pulmonary blood ow. This technique actually increased RV work load.

CONCLUSIONS: If an LVAD is to be operated in a pulsatile mode, the counter pulsation mode reduces pulmonary artery and left atrial pressureand increases pulmonary blood ow and thus cardiac output. This is in addition to the reduced right ventricular energetic requirement, anding previously described. Clinical validation of our ndings is necessary.

Keywords: Left ventricular assist device Electrical analogy Simulation

INTRODUCTION

Left ventricular assist devices (LVADs) are utilized for failing left ven-tricle function [1, 2]. LVADs can operate in a number of differentmodes: continuous or pulsatile mode; the latter can be independent,synchronous or counter pulsation with respect to the native heart.

LVADs work in a manner analogous to intra-aortic balloonpumps increasing forward ow and decreasing the work of the leftventricle [3].

The left and right ventricles contract essentially simultaneously.This results in the pulmonary and aortic valves being open simultan-eously and the mitral and tricuspid valves being closed simultan-eously. As the mitral valve is closed when the right ventriclecontracts, right ventricular stroke volume has to be accommodatedby the pulmonary vasculature [4]. The rise in pulmonary artery pres-sure depends on the native pulmonary vascular compliance. If themitral valve was open when the right ventricle contracts or the LVADis lling during right ventricle systole, then the left atrial pressure willbe lower and the forward ow higher, i.e. counter pulsation.

Previous work has identied that operating an LVAD in acounter pulsation manner results in reduced left ventricular work[5]. We extend this analysis by analysing the differing modes ofLVAD operation: synchronous, independent or counter pulsation

(antisynchronous), on pulmonary artery pressure and ow utilizinga previously published electrical analogy of the systemic and pul-monary circulation and the heart [6].

METHODS

The previously published electrical analogy of the systemic andpulmonary circulation was utilized [6], Fig. 1. The Simulation Packagewith Integrated. Circuit Emphasis (SPICE) circuit listing is detailed inthe Supplementary material.

Three LVAD operation mode scenarios were analysed: synchron-ous, counter pulsation (antisynchronous) and independent pulsatile(pseudosynchronous or asynchronous).

Baseline variables for model

Systemic blood pressure of 140 mmHg, a cardiac output of 5.5 l/min.It was assumed that 1 V ;1 mmHg of pressure. The intrinsic heartrate was assumed to be 60 beats/min for all analyses, and 1 mA ;1 l/min. The root mean square of the pulmonary artery pressure(PAP), left atrial pressure (LAP) and pulmonary blood ow (PBF)

The Author 2014. Published by Oxford University Press on behalf of the European Association for Cardio-Thoracic Surgery. All rights reserved.

WORKIN

PROGRESS

REPORT

Interactive CardioVascular and Thoracic Surgery 19 (2014) 1115 WORK IN PROGRESS REPORT ADULT CARDIACdoi:10.1093/icvts/ivu073 Advance Access publication 9 April 2014

were calculated utilizing LTSPICE (CTRL and left click on the vari-able to be measured). Simulation assumed no aortic regurgitationand the absence of a patent foramen ovale.

Incorporation of left ventricular assist deviceinto circulatory model

The different modes of operation of the LVAD were simulated byadjusting the mode of operation of the voltage source, labeled asleft ventricle in Fig. 1. For the settings of the voltage source shownin Fig. 1, for synchronous operation was DC offset (V) 0, amplitude(V) 120, frequency (Hz) 1.00 and delay (s) 0.0, counter pulsation oper-ation was DC offset (V) 0, amplitude (V) 120, frequency (Hz) 1.00and delay (s) 0.5, and asynchronous operation was achieved viatwo different frequency voltage sources, DC offset (V) 0, amplitude(V) 120, frequency (Hz) 1.00 and delay (s) 0.0, and DC offset (V) 0,amplitude (V) 120, frequency (Hz) 0.75 and delay (s) 0.0equivalentto a native ventricular rate of 45/s and an LVAD rate of 60.

Right ventricular work load

The right ventricular work load was calculated form the cardiacoutput multiplied by the root mean square PAP. The results were

normalized (referenced) to the baseline function of the RV in anormally functioning heart.

Effect of heart rate

The effect of varying the native heart rate from 30 to 120 beats/min was investigated. For asynchronous operation, the nativeheart rate was kept at 60/min and the LVAD rate was varied.

Software

LTSPICE IV (www.linear.com) was utilized for circuit simulation.The oscilloscope function was utilized to read blood pressureaortic, LAP and PAP (voltage) and pulmonary blood ow (current)values from the simulated circuit.

RESULTS

Baseline

The baseline output of the model for comparison purposes isshown in Fig. 2.

Figure 1: Electrical analogy of systemic and pulmonary circulation.

Figure 2: Baseline aortic, pulmonary artery pressure, left atrium pressure traces and pulmonary blood ow. I[Lungs1] represents pulmonary blood ow which is essen-tially cardiac output assuming no intracardiac shunts.

M. Poullis / Interactive CardioVascular and Thoracic Surgery12

Counter pulsation

The effect of counter pulsation LVAD operation on LAP and PBF isshown in Fig. 3. It should be noted that the aortic trace is 180 outof phase compared with Fig. 2, which is due to the techniquecounter pulsation.

Asynchronous

The effect of asynchronous LVAD function on pulmonary bloodow and left atrial pressure is shown in Fig. 4. It can be seen thatasynchrony causes highly variable differences in LAP and pulmon-ary blood ow.

Synchronous

The effect of synchronous LVAD operation on LAP and PBF is shownin Fig. 5.

Comparative analysis

The effect of differing LVAD pulsatile support strategies on LAP,PAP and PBF is shown in Fig. 6. It can readily be seen that

asynchronous operation is associated with a higher pulmonaryartery pressure, increased LAP and reduced PBF. Counter pulsa-tion results in the lowest PAP, LAP and the greatest PBF. The in-crease in cardiac output possible with counter pulsation assumesthe right ventricle to be unlimited in its function and the pulmon-ary compliance to be normal, a situation that is uncommon inclinical practice.

Right ventricular work load

The normalized RV work load for the differing LVAD pulsatilesupport strategies is shown in Fig. 7. It can be seen that thecounter pulsation technique has the lowest RV work load (84%),and the asynchronous mode has the highest (118%, which indi-cated that the technique actually increases RV work load), com-pared with a normal RV (100%).

Effect of heart rate

The effect of heart rate on left atrial pressure is shown in Fig. 8. Itcan be seen that regardless of heart rate, the counter pulsationmode is associated with a lower left atrial pressure. A lower leftatrial pressure results in a reduced pulmonary artery pressureand/or increases pulmonary blood ow (data not shown).

Figure 3: The effect of counter pulsation left ventricular assist device function on aortic, pulmonary artery and left atrium pressure traces and pulmonary blood ow.

Figure 4: The effect of asynchronous left ventricular assist device operation on aortic, pulmonary artery, left atrium pressure traces and pulmonary blood ow.

Figure 5: The effect of synchronous left ventricular assist device function on aortic, pulmonary artery, left atrium pressure traces and pulmonary blood ow.

WORKIN

PROGRESS

REPORT

M. Poullis / Interactive CardioVascular and Thoracic Surgery 13

DISCUSSION

LVAD operation in a counter pulsation mode results in lower pul-monary artery pressure, lower left atrial pressure and increased pul-monary artery ow, and hence cardiac output. The results of ouranalysis compliment previous modelling and clinical work [5, 7, 8].

Increasing pulmonary blood ow increases cardiac output, asno right-to-left shunt exists. Reducing left atrial pressure, so in-creasing pulmonary blood ow, also reduces the stroke workindex of the right ventricle. This may reduce the incidence of rightventricle failure post LVAD implantation requiring replacement fora biventricular assist device [9]. Reducing left atrial pressure mayalso reduce capillary uid exudation secondary to Starlings law;however, our model cannot conrm this.

Counter pulsation (antisynchronous) is analogous to intra-aorticballoon pump mechanics [10]. Intra-aortic balloon pump hasbeen shown to reduce left ventricular and right ventricular workload [1113].

In the setting of left ventricular dysfunction, especially if long-standing, pulmonary vascular compliance is frequently raised, therearises the need for assessment for biventricular assist devicesupport. Counter pulsation may be particularly important in thissubset of patients. The benet in previously t patients with normalpulmonary compliance (e.g. acute massive myocardial infarction) ofcounter pulsation will be reduced but still present.

LVADs providing continuous ow is the most common mode ofoperation [14]. Continuous ow from an electrical analogy wouldprovide the lowest left atrial and pulmonary artery pressures. Anelectrical model is unable to assess the merits and detrimentaleffects of continuous (non-pulsatile ow). We did not model con-tinuous ow as the electrical analogy model has not been veriedunder direct current (DC or continuous ow) conditions.

If an LVAD functions in a pulsatile manner, but independently ofthe heart, it can be described as functioning in a pseudosynchro-nous or isochronous mode. Our analysis predicts that this is po-tentially highly detrimental to pulmonary ow and pulmonarypressure. This nding explains and conrms previous clinicalndings [7], and we would strongly suggest against this mode ofoperation.

Our model assumes normal valvular function and the absenceof cardiac arrhythmias. Atrial brillation is quite common clinicallyand, depending on the mode of LVAD operation, could be quitedetrimental, as this situation is analogous to the asynchronousmode of operation. We hypothesize, but we offer no direct evi-dence, that rate control and pacing a failing heart to create aregular rhythm and operating an LVAD in the counter pulsationmode may offer the best haemodynamic solution.

As the LV ejection fraction reduces to zero, the haemodynamics(LAP, PAP and PBF) gradually switch to the LVAD mode modelledabove. It should be noted that the counter pulsation mode is the onlymode capable of increasing the cardiac output above the restingstate, due to the lowering of left atrial pressure below pre-LVAD levels.

The use of an electrical analogy to analyse cardiovascularhaemodynamics is well described [1517] and complements com-putational uid dynamics analysis with regard to cardiovascular pres-sures and ows [18, 19]. Randomized trials are logistically difcult,time consuming and would have to be large to avoid being under-powered due to biological variation, unlike electrical analogies.

LIMITATION

No model can accurately reect complex biological systems, andour data should be interpreted as such. A strength and a potentialweakness of using an electrical analogy is the fact that the effect ofbiological variation is essentially removed. Simulation assumed novalvular heart disease or intra cardiac shunt such as a patentforamen ovale. Dynamic regulation and interaction between theheart/lung block and the peripheral circulation was not modelledfor. At present, the counter pulsation mode of LVAD operation isnot routinely available, although working models have been previ-ously described [8].

The model assumed normal right ventricular function, i.e. pureleft ventricular dysfunction. This is clinically an unusual situation;however in the clinical setting, any cause of a raised pulmonaryartery pressure will be detrimental in the setting of reduced rightventricular function.

LVADs are frequently operational in patients with sepsis andvasodilatation. The electrical analogy model we utilized has notbeen validated in this scenario and its use may lead to errors.

CONCLUSION

If an LVAD is to be operated in a pulsatile mode, the counter pul-sation mode reduces pulmonary artery and left atrial pressure andincreases pulmonary blood ow and thus cardiac output. This is in

Figure 6: Comparison between left atrial pressure, pulmonary atrial pressureand pulmonary blood ow for different left ventricular assist device operationmodes.

Figure 7: Normalized right ventricular work load for the differing left ventricu-lar assist device pulsatile support strategies.

Figure 8: The effect of left ventricular assist device (LVAD) rate on left atrial pres-sure for differing LVAD pulsatile support strategies.

M. Poullis / Interactive CardioVascular and Thoracic Surgery14

addition to the reduced right ventricular energetic requirement, anding previously described. Clinical validation of our ndings isnecessary.

SUPPLEMENTARY MATERIAL

Supplementary material is available at ICVTS online.

Conict of interest: none declared.

REFERENCES

[1] Garbade J, Bittner HB, Barten MJ, Mohr FW. Current trends in implantableleft ventricular assist devices. Cardiol Res Pract 2011;2011:290561.

[2] Stevenson LW, Rose EA. Left ventricular assist devices: bridges to trans-plantation, recovery, and destination for whom? Circulation 2003;108:305963.

[3] Chang Y, Gao B. Modeling and identication of an intra-aorta pump.ASAIO J 2010;56:5049.

[4] Piene H. Pulmonary arterial impedance and right ventricular function.Physiol Rev 1986;66:60652.

[5] Drzewiecki GM, Pilla JJ, Welkowitz W. Design and control of theatrio-aortic left ventricular assist device based on O2 consumption. IEEETrans Biomed Eng 1990;37:12837.

[6] Abdolrazaghi M, Navidbakhsh M, Hassani K. Mathematical modelling andelectrical analog equivalent of the human cardiovascular system.Cardiovasc Eng 2010;10:4551.

[7] Maybaum S, Williams M, Barbone A, Levin H, Oz M, Mancini D.Assessment of synchrony relationships between the native left ventricleand the HeartMate left ventricular assist device. J Heart Lung Transplant2002;21:50915.

[8] Ando M, Takewa Y, Nishimura T, Yamazaki K, Kyo S, Ono M et al. A novelcounterpulsation mode of rotary left ventricular assist devices canenhance myocardial perfusion. J Artif Organs 2011;14:18591.

[9] Patlolla B, Beygui R, Haddad F. Right-ventricular failure following left ven-tricle assist device implantation. Curr Opin Cardiol 2013;28:22333.

[10] Kolyva C, Pantalos GM, Pepper JR, Khir AW. How much of the intraaorticballoon volume is displaced toward the coronary circulation? J ThoracCardiovasc Surg 2010;140:1106.

[11] Nordhaug D, Steensrud T, Muller S, Husnes KV, Myrmel T. Intraaorticballoon pumping improves hemodynamics and right ventricular efciencyin acute ischemic right ventricular failure. Ann Thorac Surg 2004;78:142632.

[12] Powell WJ Jr, Daggett WM, Magro AE, Bianco JA, Buckley MJ, Sanders CAet al. Effects of intra-aortic balloon counterpulsation on cardiac perform-ance, oxygen consumption, and coronary blood ow in dogs. Circ Res1970;26:75364.

[13] Liakopoulos OJ, Ho JK, Yezbick AB, Sanchez E, Singh V, Mahajan A. Rightventricular failure resulting from pressure overload: role of intra-aorticballoon counterpulsation and vasopressor therapy. J Surg Res 2010;164:5866.

[14] Carrel T, Englberger L, Martinelli MV, Takala J, Boesch C, Sigurdadottir Vet al. Continuous ow left ventricular assist devices: a valid option forheart failure patients. Swiss Med Wkly 2012;142:w13701.

[15] Ridgway T, Al Rawi O, Palmer K, Pullan M, Poullis M. Theoretical treatise:arterial pressure during aortic surgery. J Extra Corpor Technol 2012;44:1514.

[16] Palmer K, Ridgway T, Al Rawi O, Johnson I, Poullis M. Lissajous gures: anengineering tool for root cause analysis of individual cases--a preliminaryconcept. J Extra Corpor Technol 2011;43:1536.

[17] Warwick R, Poole R, Palmer K, Johnson I, Poullis M. Patient directed perfu-sion pressure on bypass, an analogy from electrical engineering--a newconcept. J Extra Corpor Technol 2010;42:5760.

[18] Heim L, Poole RJ, Warwick R, Poullis M. The concept of aortic replacementbased on computational uid dynamic analysis: patient-directed aorticreplacement. Interact CardioVasc Thorac Surg 2013;16:5838.

[19] Poullis M. Computational uid dynamic analysis to prevent aortic root andvalve clots during left ventricular assist device support. J Extra CorporTechnol 2012;44:2105.

WORKIN

PROGRESS

REPORT

M. Poullis / Interactive CardioVascular and Thoracic Surgery 15

/ColorImageDict > /JPEG2000ColorACSImageDict > /JPEG2000ColorImageDict > /AntiAliasGrayImages false /CropGrayImages true /GrayImageMinResolution 150 /GrayImageMinResolutionPolicy /OK /DownsampleGrayImages true /GrayImageDownsampleType /Bicubic /GrayImageResolution 175 /GrayImageDepth -1 /GrayImageMinDownsampleDepth 2 /GrayImageDownsampleThreshold 1.50286 /EncodeGrayImages true /GrayImageFilter /DCTEncode /AutoFilterGrayImages false /GrayImageAutoFilterStrategy /JPEG2000 /GrayACSImageDict > /GrayImageDict > /JPEG2000GrayACSImageDict > /JPEG2000GrayImageDict > /AntiAliasMonoImages true /CropMonoImages true /MonoImageMinResolution 1200 /MonoImageMinResolutionPolicy /OK /DownsampleMonoImages true /MonoImageDownsampleType /Bicubic /MonoImageResolution 175 /MonoImageDepth 4 /MonoImageDownsampleThreshold 1.50286 /EncodeMonoImages true /MonoImageFilter /CCITTFaxEncode /MonoImageDict > /AllowPSXObjects true /CheckCompliance [ /None ] /PDFX1aCheck false /PDFX3Check false /PDFXCompliantPDFOnly false /PDFXNoTrimBoxError true /PDFXTrimBoxToMediaBoxOffset [ 0.00000 0.00000 0.00000 0.00000 ] /PDFXSetBleedBoxToMediaBox true /PDFXBleedBoxToTrimBoxOffset [ 0.00000 0.00000 0.00000 0.00000 ] /PDFXOutputIntentProfile (None) /PDFXOutputConditionIdentifier () /PDFXOutputCondition () /PDFXRegistryName () /PDFXTrapped /False

/Description >>> setdistillerparams> setpagedevice