Pressure–strain loops unveil haemodynamics behind mechanical circulatory support systems

Abstract

Aims

Mechanical circulatory support (MCS) systems are increasingly employed in cardiogenic shock and advanced heart failure. A thorough understanding of the complex interactions occurring among heart, vasculature, and device is essential to optimize patient's management. The aim of this study is to explore non‐invasive haemodynamic profiling of patients undergoing MCS based on pressure–strain (PS) analysis.

Methods

Clinical and echocardiographic data from consecutive patients undergoing different MCS systems positioning/implantation admitted to the third level cardiological intensive care unit of Siena Hospital from August 2021 to November 2021 were retrospectively reviewed. Patients without a useful echocardiographic exam or without arterial blood pressure recording at the time of echocardiography were excluded. Myocardial work analysis was performed in the included patients.

Results

We reviewed 18 patients, of which nine were excluded. Included patients were three patients with intra‐aortic balloon pump (IABP), two patients with durable left ventricular assist device (dLVAD), two patients with Impella®, one patient with extracorporeal membrane oxygenation (ECMO), and one patient with ECMO and IABP. Myocardial work analysis was feasible in each included patient. The use of IABP shifted the PS curve rightward and downward. Global work index (GWI) and global wasted work (GWW) decreased after IABP positioning, whereas global work efficiency (GWE) increased. The use of continuous‐flow pumps, whether temporaneous (Impella®) or long term (dLVAD), induced a change in the PS loop morphology, with a shift towards a triangular shape. ECMO positioning alone resulted in a narrowing of the PS loop, with a decrease in GWI and GWE and an increase in GWW and mean arterial pressure. The combined used of IABP with ECMO widened the PS loop and improved GWI and GWE.

Conclusions

PS loops analysis in patients undergoing MCS seems to be feasible and may unveil MCS‐induced haemodynamic variations. Myocardial work could be used to monitor ventricular–arterial–device coupling and guide tailored MCS management.

Introduction

Indications for MCS are progressively expanding, ¹ , ² , ³ , ⁴ , ⁵ , ⁶ , ⁷ , ⁸ , ⁹ because they have become rapid, minimally invasive, and effective. ¹⁰ However, MCS use is still burdened by a significant rate of serious complications, such as infections, bleedings, and vascular complications. ¹ , ¹¹ , ¹² , ¹³ The successful management of patients with MCS is dependent on the expertise of the Heart Team, ¹⁴ , ¹⁵ requiring a comprehensive understanding of the complex interactions occurring among the heart, the vasculature, and the specific devices adopted. ¹⁶ , ¹⁷

Pressure–volume (PV) loops synthetize and expand clinically adopted cardiovascular parameters such as cardiac volumes and pressures providing a deeper understanding on ventricular–arterial coupling and, when on MCS, ventricular–arterial–device(s) one. Not only is PV loop helpful in understanding cardiovascular mechanics but also a mirror of the energetic burden. Thus, qualitative and quantitative PV loops evaluation could potentially provide a powerful tool to visualize treatment goals, to assess MCS response, and to guide MCS management across escalation, de‐escalation, and device weaning. ¹⁷ However, a PV loop‐guided approach to MCS management cannot routinely be employed because it is cumbersome and costly, requires specific advanced training, and is not currently widely available. ¹⁸

A novel non‐invasive echocardiographic method to estimate PV loop could potentially overcome these limits. ¹⁹ Myocardial work (MW) analysis uses speckle‐tracking echocardiography (STE) as a surrogate of volume to compute pressure–strain (PS) loops, which have demonstrated to closely correlate with invasively derived PV loops. ¹⁹ Despite a sound rationale for their use in MCS management, concerns on their feasibility in this setting may have currently limited their implementation. Therefore, the aim of this study is to explore the feasibility of MW analysis in patients with different MCS systems and explore their haemodynamic profile.

Material and methods

Patient selection

Consecutive patients admitted to the third‐level cardiological intensive care unit (CICU) of our Institution (Policlinico Santa Maria alle Scotte, Siena Hospital, Siena, Italy) between August 2021 and November 2021 undergoing different MCS systems positioning/implantation were retrospectively enrolled. Patients were included if having a transthoracic or transesophageal echocardiographic exam with electrocardiographic (ECG) tracking during MCS along with concomitant arterial blood pressure measurement. Patients were excluded if the four chambers, two chambers, and long‐axis/three‐chamber clips on any echocardiographic examination were not available, or if segmental endo‐myocardial definition at apical windows was inadequate for strain analysis. The study was performed in accordance with the Declaration of Helsinki.

Data collection and standard echocardiography

Patients' arterial blood pressure at the time of echocardiography and echocardiographic exams performed during MCS at least at a distance of 24 h from MCS implantation/positioning were retrospectively collected. Pre‐MCS implantation/positioning echocardiographic exams were collected as well, if available.

All echocardiographic examinations were performed by experienced operators using GE Vivid iQ ultrasound system equipped with an adult transthoracic 1.5–4.0 MHz or an adult transesophageal 3.0–8.0 MHz phased array transducers, and with an ECG continuously traced, according to the American Society of Echocardiography/European Association of Cardiovascular Imaging recommendations. ²⁰ , ²¹ , ²²

Arterial blood pressure data were obtained from arterial accesses, as for CICU practice, at the time of echocardiography. In patients with IABP, arterial blood pressure data were obtained from device console instead. In patients with durable left ventricular assist device (dLVAD), Doppler measured arterial blood pressure was considered.

Myocardial work analysis

For STE analysis, endocardial borders and/or myocardium of all segments (four chambers, two chambers, and apical long axis) had to be visualized throughout the whole cardiac circle. Left ventricular (LV) speckle‐tracking strain is semi‐automatically performed by the software in the three apical views and adjusted by the operator, in terms of region of interest (ROI) width and positioning, to optimize endomyocardial tracking. The software warns the operator whether a specific wall segment is not automatically recognized and they may manually adapt ROI and force analysis.

For subsequent MW analysis, markers for aortic and mitral valves opening and closure are required to set the beginning and the end of each main phase of cardiac cycle (isovolumetric contraction, ejection, isovolumetric relaxation, filling), and they were visually set from the apical long‐axis view. In those situations, in which aortic valve opening and closure timing was difficult or even impossible to be determined, such as with Impella® use, aortic markers were set using ventricular contraction and ECG track as a reference. Particularly, aortic valve opening was set at the beginning of the ejection phase, whereas aortic valve closure was set at the end of the ejection phase. Finally, arterial blood pressure is needed to warp in time and amplitude the reference curve for LV pressures estimation.

The software's output is a series of indices that depict the PS loop from various perspectives. ²³ , ²⁴ In addition, a graphic representation of the PS loop is provided. The main MW indices are Global Work Index (GWI), which is the total work performed by the heart from mitral valve closure to mitral valve opening; Global Constructive Work (GCW), which is the work performed during shortening in systole adding work during lengthening in isovolumetric relaxation; Global Wasted Work (GWW), which is the work performed during lengthening in systole adding work performed during shortening in isovolumetric relaxation; and Global Work Efficiency (GWE), which is constructive work divided by the sum of constructive and wasted work. MW analysis was performed using EchoPAC software v204 (GE HealthCare). The analysis was performed by an experienced operator in MW analysis, blinded to serial exams.

The study was reviewed and approved by the local ethics committee, which waived the need for written informed consent for collection and publication of the study data.

Results

Patient population

Of 18 consecutive patients admitted to the CICU undergoing MCS during the study period, six patients were excluded because of incomplete echocardiographic exam storage (absence of four chambers, two chambers, and/or long‐axis/three‐chamber clips or absence of ECG tracking), and three patients were excluded because inadequate acoustic window for strain analysis. Seven patients with temporary MCS and two patients with long‐term MCS were included. Overall, three patients underwent positioning of an intra‐aortic balloon pump (IABP) device with 1:1 setting, four patients with a continuous‐flow pump, of whom two were treated with a percutaneous temporary microaxial pump (Impella® CP, the first one with 2.7 L/min support, the second one with 2.5 L/min support) and two with a dLVAD (HeartMate® III devices, the first one with pump speed 5600 rpm, pump flow 4.7 lpm, pump power 4.3 W, pulse index 3.5; the second one with pump speed 4900 rpm, pump flow 4.2 lpm, pump power 3.3 W, pulse index 3.0), one patient with extracorporeal membrane oxygenation (ECMO, 4.4 L/min support) and one patient with ECMO (3.5 L/min support) and IABP. Patients' characteristics are listed in Table 1 . Echocardiographic parameters of LV unloading along with time occurring between MCS positioning/implantation and echocardiographic examination during MCS and time between before (if available) and during MCS echocardiographic examinations are reported in Table 2 .

Table 1.

Baseline patients' characteristics

	IABP			Impella®		ECMO		dLVAD
	First patient	Second patient	Third patient	First patient (Impella® CP, 2.7 L/min)	Second patient (Impella® CP, 2.5 L/min)	First patient (ECMO, 3.5 L/min + IABP 1:1)	Second patient (ECMO, 4.4 L/min)	First patient (HeartMate® III)	Second Patient (HeartMate® III)
Age	70	52	62	58	76	42	62	69	73
Sex	Male	Male	Male	Male	Male	Male	Male	Male	Male
Smoke	No	Ex	Yes	No	Ex	Ex	Yes	Ex	Ex
Hypertension	Yes	No	No	Yes	Yes	No	No	No	Yes
Dyslipidaemia	No	No	Yes	Yes	Yes	No	Yes	Yes	Yes
Diabetes	No	Yes	Yes	Yes	No	No	Yes	Yes	No
Obesity	No	No	No	No	No	No	No	No	No
CKD (eGFR <60 mL/min/1.73 mq)	No	No	Yes	No	No	No	Yes	Yes	Yes
COPD	No	No	No	No	No	No	No	No	No
Known CAD	Yes	No	No	Yes	Yes	No	No	Yes	No
Prior MI	No	No	No	Yes	Yes	No	No	Yes	No
Prior PCI	No	No	No	Yes	Yes	No	No	Yes	No
Prior CABG	Yes	No	No	No	No	No	No	Yes	No
Ischaemic aetiology of HF	Yes	No	No	Yes	Yes	No	No	Yes	No
EF (%)	20	13	15	15	25	27	10	20	20
SCAI class	B	C	C	C	C	C	E	A	A
Inotropes/vasopressors	Dobutamine 5 mcg/kg/minDopamine 3 mcg/kg/min	Dobutamine 4 mcg/kg/min	Noradrenaline 0.23 mcg/kg/minDobutamine 3.2 mcg/kig/min	Noradrenaline 0.4 mcg/kg/min Adrenaline 0.05 mcg/kg/min	Noradrenaline 0.04 mcg/kg/min Adrenaline 0.01 mcg/kg/min	Dobutamine 4 mcg/kg/min	Noradrenaline 0.2 mcg/kg/minDobutamine 3 mcg/kig/min	Noradrenaline 0.05 mcg/kg/min	n.a.
Diuretics	Furosemide 1000 mg/24 h Metolazone 5 mg	Furosemide 500 mg/24 h	Furosemide 800 mg/24 h	Furosemide 1000 mg/24 h	Furosemide 1000 mg/24 h	Furosemide 800 mg/24 h	Furosemide 600 mg/24 h	Furosemide 25 mg/die	Furosemide 125 mg/die
MAP (mmHg)	70	88	69	65	67	62	68	70	83
HR (bpm)	105	90	130	110	80	60	80	90	60
Rhythm	Sinus	AF	Sinus	Sinus	AF	Sinus	Sinus	Sinus	Sinus
PM/CRT	No	CRT	CRT	No	No	No	CRT	CRT	CRT
LBBB	No	n.a.	n.a.	No	Yes	No	n.a.	n.a.	n.a.
Hb (g/dL)	11.6	12.2	10.3	11.3	10.0	10.5	9.9	10.3	12.2
PLT (103/mm2)	187	296	221	156	186	156	213	181	157
WBC (103/mm2)	12.4	18.3	7.9	19.1	20.7	30.7	9.2	5.8	7.3
CRP (mg/dL)	9.93	1.60	1.29	17.22	12.00	31.16	5.01	0.40	0.59
Creatinine (mg/dL)	1.65	1.40	3.57	4.30	1.91	3.68	3.47	1.50	1.72
Sodium (mEq/L)	130	135	129	147	137	143	133	137	141
Bilirubin (mg/dL)	2.9	0.7	0.7	3.7	0.7	0.6	1.6	0.3	0.5
Troponin (ng/L)	384	69	204	21 325	17 183	403	255	29	44
Lactate (mmol/L)	1.79	2.10	2.90	3.74	2.52	2.87	5.18	1.96	1.65
PA sat (%)	65	45	27	44	46	42	32	47	54
CVP (cmH2O)	+13	+8	+18	+12	+12	+13	+20	+6	+3

Vital parameters, laboratory exams, and medical therapy are the latest available before beginning of mechanical circulatory support. Medical therapy has remained the same at the time of echocardiographic exam under MCS.

AF, atrial fibrillation; CABG, coronary artery bypass graft; CAD, coronary artery disease; CKD, chronic kidney disease; COPD, chronic obstructive pulmonary disease; CRP, C‐reactive protein; CRT, cardiac resynchronization therapy; CVP, central venous pressure; dLVAD, durable left ventricular assist device; ECMO, extracorporeal membrane oxygenation; EF, ejection fraction; eGFR, estimated glomerular filtration rate; Hb, haemoglobin; HF, heart failure; HR, heart rate; IABP, intra‐aortic balloon pump; LBBB, left bundle branch block; MAP, mean arterial pressure; MCS, mechanical circulatory support; MI, myocardial infarction; PA, pulmonary artery; PCI, percutaneous coronary intervention; PLT, platelets; PM, pacemaker; SCAI, Society for Cardiovascular Angiography and Interventions; WBC, white blood cells.

Table 2.

Echocardiographic parameters of LV unloading

Device	Parameter	Pre‐MCS	During MCS	Time between MCS positioning/implantation and echo during MCS	Time between pre‐ and during MCS echo exams
IABP
First patient	LV EDD (mm)	49	48	36 h	48 h
	LV ESD (mm)	42	40
	LV EDV (mL)	80	76
	LV ESV (mL)	66	61
	LV EF (%)	17	20
	LV mass (g)	268	274
	LA size (mL)	71	68
	E‐wave (m/s)	0.94	0.90
	E/A ratio	2.33	2.29
	Mitral DT (ms)	137	146
	e′ (m/s)	0.04	0.04
	E/e′	24	22
Second patient	LV EDD (mm)	72	70	48 h	8 days
	LV ESD (mm)	68	65
	LV EDV (mL)	273	260
	LV ESV (mL)	238	221
	LV EF (%)	13	15
	LV mass (g)	297	285
	LA size (mL)	172	130
	E‐wave (m/s)	1.20	1.05
	E/A ratio	2.4	2.3
	Mitral DT (ms)	141	150
	e′ (m/s)	0.04	0.04
	E/e′	30	26
Third patient	LV EDD (mm)	82	82	24 h	36 h
	LV ESD (mm)	73	72
	LV EDV (mL)	319	315
	LV ESV (mL)	267	255
	LV EF (%)	16	19
	LV mass (g)	373	374
	LA size (mL)	123	115
	E‐wave (m/s)	1.07	1.00
	E/A ratio	2.2	2.1
	Mitral DT (ms)	144	152
	e′ (m/s)	0.04	0.04
	E/e′	27	25
Impella®
First patient	LV EDD (mm)	58	55	24 h	28 h
	LV ESD (mm)	50	45
	LV EDV (mL)	151	145
	LV ESV (mL)	133	125
	LV EF (%)	12	14
	LV mass (g)	182	181
	LA size (mL)	69	59
	E‐wave (m/s)	0.88	0.74
	E/A ratio	2.0	1.8
	Mitral DT (ms)	147	180
	e′ (m/s)	0.03	0.04
	E/e′	29	19
Second patient	n.a.			24 h	n.a.
dLVAD
First patient	LV EDD (mm)	70	61	67 days	74 days
	LV ESD (mm)	65	49
	LV EDV (mL)	299	144
	LV ESV (mL)	239	108
	LV EF (%)	20	25
	LV mass (g)	290	254
	LA size (mL)	65	55
	E‐wave (m/s)	0.96	0.43
	E/A ratio	1.6	0.60
	Mitral DT (ms)	185	210
	e′ (m/s)	0.04	0.05
	E/e′	24	9
Second patient	n.a.			24 h	48 h
ECMO	LV EDD (mm)	79	81	36 h	72 h
	LV ESD (mm)	70	73
	LV EDV (mL)	304	354
	LV ESV (mL)	258	319
	LV EF (%)	15	10
	LV mass (g)	360	367
	LA size (mL)	115	123
	E‐wave (m/s)	0.9	1.2
	E/A ratio	1.9	2.1
	Mitral DT (ms)	165	147
	e′ (m/s)	0.04	0.04
	E/e′	22	30
ECMO and IABP	LV EDD (mm)	51	42	72 h	5 days
	LV ESD (mm)	45	28
	LV EDV (mL)	130	111
	LV ESV (mL)	114	78
	LV EF (%)	12	30
	LV mass (g)	155	147
	LA size (mL)	54	45
	E‐wave (m/s)	0.85	0.83
	E/A ratio	2.2	1.1
	Mitral DT (ms)	127	188
	e′ (m/s)	0.03	0.08
	E/e′	28	10

In those cases with an available pre‐MCS positioning/implantation echocardiographic exam, echocardiographic parameters of LV unloading are reported in order to show variations induced by the MCS in question. Time occurring between MCS positioning/implantation and echocardiographic examinations during MCS and time between pre‐MCS and during MCS echocardiographic examinations are reported as well.

dLVAD, durable left ventricular assist device; DT, deceleration time; ECMO, extracorporeal membrane oxygenation; EDD, end‐diastolic diameter; EDS, end‐systolic diameter; EDV, end‐diastolic volume; EF, ejection fraction; ESV, end‐systolic volume; IABP, intra‐aortic balloon pump; LA, left atrium; LV, left ventricle; MCS, mechanical circulatory support.

Intra‐aortic balloon pump

Pre‐IABP positioning and during IABP support results of MW analysis along with relative arterial blood pressures are reported in Table 3 . Figure 1 shows pre‐positioning and during IABP support PS loops and MW bull's eyes of the first patient.

Table 3.

Global longitudinal strain, myocardial work indices and arterial blood pressure of study population undergoing mechanical circulatory support systems

Device	Parameter	Pre‐MCS	During MCS	Pre‐during delta
IABP
First patient	GLS (%)	−8	−7	+1
	GWI (mmHg%)	443	389	−54
	GCW (mmHg%)	500	424	−76
	GWW (mmHg%)	56	30	−26
	GWE (%)	85	87	+2
	PA (mmHg)	93/58	70/58
Second patient	GLS (%)	−2	−2	0
	GWI (mmHg%)	145	134	−11
	GCW (mmHg%)	339	306	−33
	GWW (mmHg%)	223	135	−88
	GWE (%)	59	66	+7
	PA (mmHg)	110/77	91/68
Third patient	GLS (%)	−3	−4	−1
	GWI (mmHg%)	232	228	−4
	GCW (mmHg%)	388	366	−22
	GWW (mmHg%)	132	85	−47
	GWE (%)	70	80	+10
	PA (mmHg)	95/57	74/44
Impella®
First patient	GLS (%)	−7	−7	0
	GWI (mmHg%)	441	292	−149
	GCW (mmHg%)	608	324	−284
	GWW (mmHg%)	175	201	+26
	GWE (%)	76	60	−16
	PA (mmHg)	104/75	85/50
Second patient	GLS (%)		−2
	GWI (mmHg%)		237
	GCW (mmHg%)		328
	GWW (mmHg%)		533
	GWE (%)		39
	PA (mmHg)		90/65
dLVAD
First patient	GLS (%)	−7	−8	−1
	GWI (mmHg%)	579	229	−350
	GCW (mmHg%)	867	402	−465
	GWW (mmHg%)	340	320	−20
	GWE (%)	70	56	−14
	PA (mmHg)	110/55	75
Second patient	GLS (%)	−6	−4	+2
	GWI (mmHg%)	486	72	−414
	GCW (mmHg%)	629	190	−439
	GWW (mmHg%)	129	99	−30
	GWE (%)	78	61	−17
	PA (mmHg)	110/70	59
ECMO	GLS (%)	−4	−1	+3
	GWI (mmHg%)	249	100	−149
	GCW (mmHg%)	484	297	−187
	GWW (mmHg%)	146	210	+64
	GWE (%)	76	55	−21
	PA (mmHg)	95/55	93/73
ECMO and IABP	GLS (%)	−4	−7	−3
	GWI (mmHg%)	123	425	+302
	GCW (mmHg%)	362	754	+392
	GWW (mmHg%)	127	291	+164
	GWE (%)	67	72	+5
	PA (mmHg)	75/55	88/65

Cases with an available pre‐MCS positioning/implantation echocardiographic exam are reported in order to show variations induced by the MCS in question. Absolute variation between before and during support in GLS and MW indices are shown, if available. Arterial blood pressure is reported as systolic and diastolic pressure, but only mean value in the HeartMate® III cases.

dLVAD, durable left ventricular assist device; ECMO, extracorporeal membrane oxygenation; GCW, global constructive work; GLS, global longitudinal strain; GWE, global work efficiency; GWI, global work index; GWW, global wasted work; IABP, intra‐aortic balloon pump; MCS, mechanical circulatory support; MW, myocardial work; PA, arterial pressure.

Figure 1.

Pre‐positioning and during IABP support pressure–strain loops and global work index bull's eyes of the first patient with IABP. (A) The red curve represents the pressure–strain loop before IABP positioning, the yellow one the pressure–strain loop during IABP support. (B) Global work index bull's eye before IABP positioning. (C) Global work index bull's eye during IABP support. LVP, left ventricular pressure.

Impella®

Pre‐Impella® positioning and during Impella® support results of MW analysis along with relative arterial blood pressures are reported in Table 3 . Figure 2 shows pre‐positioning and during Impella® support PS loops and MW bull's eyes of the first patient.

Figure 2.

Pre‐positioning and during Impella® support pressure–strain loops and global work index bull's eyes of the first patient with Impella®. (A) The red curve represents the pressure‐strain loop before Impella® positioning, the yellow one the pressure–strain loop during Impella® support. (B) Global work index bull's eye before Impella® positioning. (C) Global work index bull's eye during Impella® support. LVP, left ventricular pressure.

Durable left ventricular assist devices

Pre‐HeartMate® III implantation and during HeartMate® III support results of MW analysis along with relative arterial blood pressures are reported in Table 3 . Figure 3 shows pre‐implantation and during HeartMate® support PS loops and MW bull's eyes of the first patient.

Figure 3.

Pre‐implantation and during HeartMate® III support pressure–strain loops and global work index bull's eyes of the first patient with durable left ventricular assist device. (A) The red curve represents the pressure–strain loop before HeartMate® III positioning, the yellow one the pressure–strain loop during HeartMate® III support. (B) Global work index bull's eye before HeartMate® III implantation. (C) Global work index bull's eye during HeartMate® III support. LVP, left ventricular pressure.

Extracorporeal membrane oxygenation

Pre‐ECMO (and IABP) positioning and during ECMO (and IABP) support results of MW analysis along with relative arterial blood pressures are reported in Table 3 . Figure 4 shows before and during ECMO and IABP support PS loops and MW bull's eyes of the first patient. Figure 5 shows before and during ECMO support PS loops and MW bull's eyes of the second patient.

Figure 4.

Pre‐positioning and during ECMO and IABP support pressure–strain loops and myocardial work bull's eyes of the first patient with ECMO. (A) The red curve represents the pressure–strain loop before ECMO and IABP positioning, the yellow one the pressure–strain loop during ECMO and IABP support. (B) Global work index and global work efficiency bull's eye before ECMO and IABP positioning. (C) Global work index and global work efficiency bull's eye during ECMO and IABP support. LVP, left ventricular pressure.

Figure 5.

Pre‐positioning and during ECMO support pressure–strain loops and global work index bull's eyes of the second patient with ECMO. (A) The red curve represents the pressure–strain loop before ECMO positioning, the yellow one the pressure–strain loop during ECMO support. (B) Global work index bull's eye before ECMO positioning. (C) Global work index bull's eye during ECMO support. LVP, left ventricular pressure.

Discussion

To the best of our knowledge, this is the first study to report non‐invasive PS analysis in patients with different MCS systems. We demonstrated the feasibility of MW analysis in such patients, which often present suboptimal acoustic windows and evaluated variations in the haemodynamic profile induced by MCS.

We will separately discuss the haemodynamic impact of each considered MCS.

Intra‐aortic balloon pump

Due to inflation during diastole and deflation during systole, IABP has multiple effects on haemodynamics. Firstly, diastolic pressure is increased with consensual increase in peripheral and coronary perfusion. Secondly, end‐diastolic deflation reduces impedance after which left ventricle must eject. For these reasons, ejection fraction is increased, end‐diastolic pressure decreased, and wall tension reduced. Systolic and end‐diastolic blood pressure are reduced as well. The higher myocardial oxygen supply/demand ratio reverses the vicious cycle of the failing heart, eventually translating into a slight increase in cardiac output (between 0.5 and 1.0 L/min or up to 30%). ²⁵ Consequently, PV loop is shifted leftward and downward, because the heart works at lower pressures and volumes, increasing stroke volume even though stroke work and total energy expenditure are reduced.

Our results from the MW analysis are in agreement with these pathophysiological premises. In fact, GWI, representing stroke work, GCS, and GWW were reduced after IABP positioning, whereas GWE was increased, proof of better heart energetics when IABP is used. Regarding PS loops, the one during IABP support is shifted downward and rightward instead of leftward. This is due to the strain as a surrogate of volume being a negative measure.

Impella® and durable left ventricular assist devices

These two different MCS systems are both LV‐to‐arterial MCS. The most typical impact of LV‐to‐arterial MCS on the haemodynamic profile is the change in the shape of the PV loop. Suction of blood from ventricular chamber irrespective of cardiac cycle results in the disappearance of isovolumic contraction and relaxation phases, with PV loop transforming from a ‘trapezoidal’ to a ‘triangular’ shape. Besides this, these MCS represent parallel pumps to the heart, vicaring its function in varying amount in dependence of their speed. Therefore, heart is at least partially relieved from high‐filling pressures. Peripheral and coronary perfusion is once again increased, whereas LV intracavitary pressures and wall tension decreased. As a consequence, better myocardial oxygen supply/demand ratio is established. Secondary effects may occur, such as an increase in LV contractility, or end‐systolic elastance (Ees), and a decrease in peripheral resistances, or effective arterial elastance (Ea), with subsequent enhanced ventricular‐arterial coupling.

Regarding the first patient with Impella®, GLS value did not change before and after MCS positioning and traditional echocardiographic parameters of LV unloading are not as much affected. This could possibly be explained by the short time gap between MCS positioning and echocardiographic examination during MCS and changes in medical therapy, such as reduction in inotropic support because of the MCS positioning. Regarding the first patient with dLVAD, GLS value have only slightly reduced, possibly because of suboptimal device speed, right heart failure development, or progression of the underlying myocardial disease.

Our results from MW analysis in the patients with Impella® and dLVAD consistently showed the shift towards the characteristic ‘triangular’ shape of the PS loop. Moreover, most of MW indices have reduced after LV‐to‐arterial MCS, but GWW not so much. In the first patient with Impella® and the first patient with dLVAD, as well as GWW, GLS did not vary much, but substantial variations were observed in GWI and GCW, resulting in reduced GWE. Therefore, haemodynamic changes are most likely explained by the substantial variation in GWI and GCW induced by the presence of a parallel pump to the heart rather than the subtle variations in GWW, which are more likely explained by inter‐examination variability.

MW analysis seems to be feasible also with suboptimal transthoracic acoustic windows (see Video S1 ). However, in the second patient with HeartMate® III, due to the complete absence of transthoracic acoustic window in supine position during intensive care unit (ICU) stay, a transesophageal echocardiographic (TEE) exam to evaluate right ventricular (RV) function was performed in the suspicion of RV failure. Therefore in this case we used a TEE exam to perform MW analysis, the first time to be reported in human (see Video S2 ). ²⁶ , ²⁷ , ²⁸ Thanks to its ease of performance in sedated patients and its high endomyocardial borders definition, we suggest that TEE could be a valuable option in ICU to monitor RV function in patients who underwent dLVAD implantation, with MW being the most reliable tool to load‐independently measure RV contractility. MW analysis could help in optimizing patient's management, either suggesting RV assist device necessity in case of severe RV failure or guiding weaning when RV function improves.

Extracorporeal membrane oxygenation and intra‐aortic balloon pump

While providing total cardiopulmonary support in emergency situations, ECMO itself may bring detrimental effects on left ventricular mechanics and energetics. In fact, the failing heart may difficultly overcome the increase in afterload at expense of increased filling pressures. This translates in a rightward and upward shift and a narrowing of the PV loop. The increase in left ventricular diastolic pressures pours on the pulmonary circulation with risk of pulmonary oedema. Thus, response to ECMO is variable, and myocardial oxygen supply/demand ratio may either improve, because of the increase in mean arterial pressure and more efficient gas exchange, or worsen, when left ventricle is unable to eject and pulmonary oedema leads to hypoxia and hypercapnia. In the second case, some treatment options may favour LV unloading, such as IABP, Impella®, septostomy, or apical canulation. ²⁹ , ³⁰ , ³¹

In our case, MW analysis and traditional echocardiographic indices of LV unloading performed before and during ECMO and IABP positioning showed improvement in left ventricular function. GWW increased proportionally to GLS and all other MW indices; therefore, it was the substantial increase in GLS that justified the haemodynamic improvement. On the contrary, ECMO support without additional systems for LV unloading showed narrowing of the PS strain loop with unfavourable effects on LV function, demonstrated by the decrease in GLS, GWI, GCS, and GWE, discordant increase in GWW, and decrease in echocardiographic parameters of LV unloading (see Table 2 ). As such, both LV systolic function and LV diastolic function parameters worsened with ECMO only, differently from ECMO with IABP. Mean arterial blood pressure with ECMO increased as expected.

Clinical perspectives

CICU patients usually present suboptimal acoustic windows due to supine decubitus and reduced possibilities of mobilization and/or invasive ventilatory support. However, echocardiography is still an essential tool to guide clinicians in their decisions. STE strain is a sensitive and reproducible parameter to evaluate myocardial function and could be useful in patients with MCS as well. ³² Results from our analysis suggest that MW analysis is feasible even in the context of MCS. Despite requiring post‐processing analysis, it is a rapid and easy method to non‐invasively assess mechanics and energetics not only of the left ventricle but also of the right one. In addition, the possibility of using TEE clips whereby TTE is not informative makes MW analysis flexible and more precise.

MW seems to provide precious information about haemodynamic profile variations induced by MCS systems and ventricular–arterial–device coupling (Figure S1 ). The management of MCS is particularly challenging and mostly rely on Heart Team experience with the different devices. Estimation of PV loops through MW analysis could unmask myocardial contribution to circulation, define severity of myocardial impairment, and assess energetics. More importantly, serial assessments could show dynamic variations in response to therapies and guide further clinical decisions, either increasing drug/mechanical support or decreasing them during the weaning process. Regarding dLVAD, besides invasive haemodynamic monitoring with pulmonary artery catheter during ramp studies, MW could have a complementary role in this context to better define the optimal device settings thanks to its unique feature of providing insights not only into myocardial mechanics but also energetics.

Limitations

This was a single‐centre and retrospective analysis. Hence, it presents some important limitations:

1. Few patients were included, and at most three patients for each MCS system were available, reducing the validity of our results. However, the purpose of the study was only to evaluate the feasibility of MW analysis in patients with MCS, which usually present suboptimal acoustic windows, and to explore haemodynamic variation induced by MCS.

2. A pre‐MCS positioning/implantation echocardiographic exam was not available for all patients, therefore limiting the evaluation of the haemodynamic changes induced by the MCS in question. However, for each category of MCS included (IABP, LV‐to‐arterial MCS and ECMO) at least one pre‐positioning/implantation echocardiographic exam was available.

3. In patients with dLVAD, the presence of the inflow cannula may alter strain measurement in some LV segments. However, EchoPAC software for MW analysis allows at most one segment to be left out from the analysis. Nonetheless, the possible impact of the inflow cannula on results should be limited considering that at most two to three segments out of 17 segments may be altered by the presence of the inflow cannula.

4. Intra‐observer and inter‐observer variability has not been assessed due to the shortage of patients included. However, other studies have previously assessed the reproducibility of MW analysis, even though not in the context of MCS systems.

5. Due to the absence of complete echocardiographic exams for each consecutive patient admitted to CICU in the study period, it was not possible to quantitatively assess MW feasibility.

Conclusions

MW analysis in the context of MCS seems to be feasible and to provide precious information about haemodynamic profile variations through PS loops. Therefore, this study provides a rationale to test whether MW could be useful to monitor ventricular–arterial–device coupling in such patients and guide tailored MCS management.

Conflict of interest

None declared.

Funding

This research did not receive any specific grant from funding agencies in the public, commercial, or not‐for‐profit sectors.

Supporting information

Video S1. Myocardial work analysis performed in a patient with Impella^® support. This demonstrate feasibility of myocardial work analysis also with suboptimal transthoracic acoustic windows.

Video S2. Myocardial work analysis performed with a transesophageal echocardiographic exam in a patient with recent implant of a HeartMate^® III. This was the first time that myocardial work analysis was performed with a transesophageal echocardiographic exam in a human.

Figure S1. Supporting illustration Hemodynamic variations induced by different MCS systems. A) Red curve represents PS loop previously to IABP support, yellow curve represents PS loop during IABP support. B) Red curve represents PS loop previously to ECMO support, yellow curve represents PS loop during ECMO support. C) Red curve represents PS loop previously to ECMO support, yellow curve represents PS loop during ECMO and IABP support D) Red curve represents PS loop previously to Impella^® support, yellow curve represents PS loop during Impella^® support E) Red curve represents PS loop previously to HeartMate^® III support, yellow curve represents PS loop during HeartMate^® III support. PS = pressure‐strain; IABP = intra‐aortic balloon pump; ECMO = extracorporeal membrane oxygenation; dLVAD = durable left ventricular assist device.

Landra, F. , Mandoli, G. E. , Sciaccaluga, C. , Gallone, G. , Bruno, F. , Fusi, C. , Barilli, M. , Focardi, M. , Cavigli, L. , D'Ascenzi, F. , Bernazzali, S. , Maccherini, M. , Cameli, M. , and Valente, S. (2023) Pressure–strain loops unveil haemodynamics behind mechanical circulatory support systems. ESC Heart Failure, 10: 2607–2620. 10.1002/ehf2.14339.