Venous excess ultrasound score detects systemic venous congestion in a left ventricular assist device patient: A case report

Abstract

BACKGROUND

Left ventricular assist devices (LVADs) are a well-known therapeutic option for patients with advanced heart failure. Pump speed directly influences ventricular dimensions, septal position, right ventricular (RV) loading conditions, and overall haemodynamics. Echocardiography remains crucial for the assessment of such patients, whereas the venous excess ultrasound score (VExUS) is an emerging bedside tool for evaluating systemic venous congestion.

CASE SUMMARY

A 45-year-old woman with advanced ischaemic cardiomyopathy was admitted to the intensive care unit after HeartMate 3 LVAD implantation. On the fourth postoperative day, she developed severe hypotension with low estimated pump flow and low pulsatility index (PI) while the LVAD was set at 5100 rpm. After an empiric 250 mL saline bolus, urgent transthoracic echocardiography showed a small left ventricular cavity, RV dilation, leftward septal shift, and features of ventricular interdependence. VExUS assessment demonstrated severe systemic venous congestion, with hepatic vein systolic flow reversal, marked portal vein pulsatility, and severely abnormal intrarenal venous Doppler flow. Based on combined echocardiographic, venous Doppler, and device findings, hypotension was attributed to suction-related ventricular interdependence and right-sided congestion rather than isolated hypovolaemia. LVAD speed was decreased from 5100 rpm to 4900 rpm, followed by rapid haemodynamic stabilisation, improved pump flow and PI, restoration of a midline septal position, and partial improvement in portal venous flow.

CONCLUSION

This case highlights the potential value of integrating VExUS with conventional echocardiography and LVAD parameter interpretation during postoperative haemodynamic instability. Serial venous Doppler assessment may help distinguish congestion-related ventricular interdependence from true hypovolaemia and support individualized adjustment of pump speed and fluid strategy. Further validation in LVAD populations is required.

Key Words: Left ventricular assist device; Venous excess ultrasound score; Echocardiography; Venous congestion; Haemodynamic assessment; Case report

Core Tip: In left ventricular assist device (LVAD)-supported patients, hypotension and low-flow alarms do not necessarily indicate hypovolaemia. This case highlights how integrating transthoracic echocardiography with venous excess ultrasound score (VExUS) can help identify suction-associated ventricular interdependence with systemic venous congestion and guide timely LVAD speed optimisation rather than unnecessary fluid loading. VExUS should be interpreted as an adjunctive bedside tool within the broader clinical, echocardiographic, and device-parameter context. Serial venous Doppler reassessment may further support bedside monitoring of congestion and response to interventions during postoperative haemodynamic instability. Further validation is required before routine implementation in LVAD patients.

INTRODUCTION

Left ventricular assist devices (LVADs) are an established treatment option for people with advanced heart failure. They may serve as a bridge to transplantation, a bridge to recovery, or as a destination therapy[1]. LVADs improve cardiac output (CO), systemic perfusion, and end-organ function by continuously mechanically unloading the left ventricle (LV)[2]. Nevertheless, optimal device performance requires careful titration of pump speed to maintain an adequate balance between LV unloading and right ventricular (RV) adaptation, as excessive unloading may adversely affect biventricular interaction[3].

Haemodynamic disturbances in LVAD-supported patients are often multifactorial and difficult to interpret. Hypotension could be a sign of true hypovolaemia, excessive pump speed causing interventricular septal shift, or RV dysfunction. In this context, depending on single haemodynamic parameters like arterial pressure, central venous pressure, or estimated pump flow may be insufficient to identify the mechanism of instability. In particular, suction-associated ventricular interdependence and RV preload limitation may be difficult to distinguish from true hypovolaemia, potentially leading to inappropriate fluid administration or delayed optimisation of LVAD settings[4].

Point-of-care ultrasound is an essential component of the bedside evaluation of patients with mechanical circulatory support, permitting a dynamic assessment of ventricular size, septal position, and ventricular interdependence[5]. The venous excess ultrasound score (VExUS) is a comprehensive assessment of systemic venous congestion through an evaluation of Doppler waveforms of the hepatic, portal, and renal veins. VExUS has shown clinical utility in critically ill populations for directing fluid management and evaluating organ congestion[6]; however, its use in patients with LVADs support has not been thoroughly examined. This case report emphasizes the utility of VExUS as an adjunctive diagnostic instrument for evaluating ventricular interaction and facilitating LVADs speed optimization.

CASE PRESENTATION

Chief complaints

A 45-year-old woman with advanced ischaemic cardiomyopathy was admitted to the intensive care unit (ICU) after HeartMate 3 LVAD implantation. On 4^th day postoperatively, the patient developed severe hypotension, with systolic arterial pressure dropped to 50 mmHg.

History of present illness

The patient underwent coronary artery bypass grafting with left internal mammary artery to the left anterior descending artery (LAD) and mitral valve repair with annuloplasty. Following cardiopulmonary bypass, she developed refractory haemodynamic instability, and a durable LVAD was inserted (HeartMate 3). Postoperatively, LVAD speed was gradually increased over the first days of intensive care unit stay, reaching 5100 rpm with an estimated flow of approximately 4.4 L/minute.

History of past illness

A 45-year-old woman (body weight 105 kg, height 168 cm) with advanced ischaemic cardiomyopathy was admitted to our center for management of end-stage heart failure.

Personal and family history

Her medical history included anterior ST-elevation myocardial infarction treated with two percutaneous coronary interventions to the LAD, chronic heart failure with reduced ejection fraction, implantable cardioverter-defibrillator upgraded to cardiac resynchronization therapy with defibrillator, paroxysmal atrial fibrillation, and type 2 diabetes mellitus. Preoperative echocardiography demonstrated severe biventricular dysfunction, with a markedly dilated LV (end-diastolic diameter 70 mm, ejection fraction 20%), RV dilation with slightly impaired systolic function (TDI S′ 9 cm/s), severe secondary mitral regurgitation, moderate tricuspid regurgitation, and severe pulmonary hypertension (estimated pulmonary artery systolic pressure 75-80 mmHg).

Physical examination

She was mechanically ventilated and sedated, receiving inotropic and vasopressor support. No sustained arrhythmia was documented on continuous monitoring during the episode.

Laboratory examinations

There were no laboratory findings suggestive of active bleeding or sepsis as the primary cause of hypotension. The patient did not have a pulmonary artery catheter in place at the time of the event; therefore, additional invasive haemodynamic parameters, including right atrial pressure, pulmonary artery pressures, and CO measurements, were not available. The LVAD was running at 5100 rpm with a low estimated pump flow of 1.2 L/minute and a low pulsatility index (PI) of 1.1. Device malfunction was considered less likely based on the absence of pump power abnormalities or device alarms suggestive of mechanical failure or pump thrombosis.

Imaging examinations

An empiric measure was to give a limited bolus of 250 mL normal saline 0.9% before definitive ultrasound assessment. A transthoracic echocardiogram was then performed. Transthoracic echocardiography revealed a small LV cavity, marked RV dilation, and leftward shift of the interventricular septum (Video 1). There were no echocardiographic findings suggestive of pericardial tamponade.

A VExUS assessment was subsequently performed by an operator experienced in critical care echocardiography and venous Doppler ultrasound. In this case, acoustic windows were sufficient to obtain interpretable hepatic, portal, and intrarenal venous Doppler waveforms. Doppler interrogation demonstrated severely abnormal venous waveforms, including systolic flow reversal in the hepatic veins, marked pulsatility of the portal vein, and discontinuous intrarenal venous flow, consistent with severe systemic venous congestion and a VExUS of 3 (Figure 1).

Figure 1 Severely abnormal venous Doppler waveforms during the episode of haemodynamic instability. A: Demonstrating hepatic vein systolic flow reversal; B: Marked portal vein pulsatility > 50% (middle); C: Monophasic intrarenal venous flow (right), consistent with severe systemic venous congestion (VExUS score 3).

FINAL DIAGNOSIS

Based on the integrated echocardiographic and VExUS findings, hypotension was attributed to RV failure and venous congestion related to excessive LVAD unloading rather than intravascular volume depletion.

TREATMENT

The sequence of management was as follows: Recognition of severe hypotension with low LVAD flow at 5100 rpm, administration of a limited 250 mL saline bolus, urgent transthoracic echocardiography with VExUS assessment, interpretation of the findings as congestion-related ventricular interdependence, and immediate reduction of LVAD speed to 4900 rpm.

OUTCOME AND FOLLOW-UP

This intervention resulted in a rapid haemodynamic stabilisation and prompt improvement in device performance (pump flow increased to 3.7 L/minute and PI rised to 3.5). Repeated echocardiography demonstrated improvement in LV dimensions, restoration of a midline interventricular septal position (Video 2), and partial normalization of portal venous flow (Figure 2), despite fluid administration. The patient remained in the intensive care unit for an additional 15 days without recurrence of hypotensive episodes. During this period, she underwent daily echocardiographic and VExUS monitoring to guide haemodynamic assessment and congestion surveillance. Renal function remained stable during the ICU course, without need for renal replacement therapy. She was subsequently transferred uneventfully to the general ward and was later discharged from hospital without further complications.

Figure 2 Following left ventricular assist device speed reduction and haemodynamic stabilization. A-C: Portal vein flow demonstrated improvement with pulsatility index about 40 (B) despite the limited fluid bolus, whereas hepatic (A) and intrarenal (C) venous Doppler waveforms remained severely abnormal, indicating persistent systemic venous congestion.

DISCUSSION

In this paper, we present a case report of a patient supported with a durable LVAD who acutely developed hypotension related to suspected suction-associated ventricular interdependence with concomitant systemic venous congestion. Although a small volume bolus was empirically administered, subsequent integrated echocardiographic assessment combined with VExUS evaluation demonstrated that hypotension was unlikely to be driven by isolated hypovolaemia. Instead, the combination of a small LV cavity, leftward septal shift, RV dilation, low pump flow, low PI, and severe venous Doppler abnormalities supported LVAD speed reduction rather than further volume administration. Following speed reduction, haemodynamics and LVAD parameters rapidly improved, and the patient experienced no further hypotensive episodes during the subsequent ICU course.

Managing haemodynamics in patients with LVADs is inherently complicated and necessitates meticulous interpretation of bedside findings alongside device parameters[3]. Echocardiography is the imaging modality of choice for initial evaluation of patients with LVADs presenting with haemodynamic instability, because it allows rapid assessment of ventricular size and function, valves, septal position, and potential device-related complications. In particular, echocardiography plays a central role in identifying RV failure, and suction-related phenomena[7]. However, assessment of LV dimensions and volume status may be challenging in LVAD patients because of poor acoustic windows and altered cardiac anatomy, potentially limiting the ability of echocardiography alone to fully characterize the haemodynamic substrate driving hypotension[8].

RV function represents a critical determinant of successful LVAD support, as the RV effectively serves as the flow-limiting component of the system by providing preload to the LV. Although LVAD implantation unloads the LV and improves systemic output, it simultaneously increases the workload of the RV, making an appropriate balance between ventricular loading conditions essential. Excessive unloading of the LV can make the LV cavity smaller and change the position of the septum. This can increase the RV afterload, worsen the RV systolic performance, and promote ventricular interdependence. RV systolic function is facilitated to a significant extent by septal contraction. This loss of septal mechanics post-LVAD implantation may further impair effective RV output. Changes in shape and function may also lead to tricuspid annular dilation and worsen functional tricuspid regurgitation, which would further slowdown forward RV flow and contribute to systemic venous congestion[9].

VExUS is a point-of-care ultrasound-based method for the assessment of systemic venous congestion by combining the diameter of the inferior vena cava (IVC) with Doppler evaluation of the hepatic, portal, and intrarenal veins. It provides physiologic information about the haemodynamic consequences of elevated right-sided filling pressures that extends beyond cardiac morphology and static pressure estimates. Prior studies have established a significant correlation between severe VExUS grades, increased right atrial pressure, and negative outcomes, including acute kidney injury and heart failure-related morbidity, underscoring its capacity to detect clinically significant venous congestion[10]. However, in the present case, no pulmonary artery catheter was in place at the time of haemodynamic deterioration. Therefore, right atrial pressure, pulmonary artery pressures, and invasive CO measurements were not available to corroborate the noninvasive diagnosis of venous congestion. This should be acknowledged as an important limitation, and VExUS findings should be considered as complementary physiologic evidence rather than a direct substitute for invasive haemodynamic assessment when such data are clinically required.

Recent studies have shown that the portal vein Doppler is a particularly sensitive and physiologically relevant marker of systemic venous congestion. The portal vein pulsatility fraction was significantly reduced following ultrafiltration and decongestion in patients with end-stage kidney disease, and was superior to changes in the diameter of the IVC[11]. These results suggest that portal vein Doppler can better describe dynamic variations of venous pressure and congestion than traditional static parameters[12]. In our patient, the partial improvement in portal venous flow with LVAD speed reduction was consistent with a dynamic change in venous congestion. Nonetheless, caution is warranted in interpretation as an empiric 250 mL fluid bolus was given before the first VExUS assessment and may have affected venous Doppler patterns.

This is particularly important in patients with LVADs, where it is critical to distinguish true hypovolaemia from congestion-related ventricular interdependence. VExUS supplies valuable real-time feedback on haemodynamic changes that may not be immediately obvious from echocardiographic morphology alone. In our case, the partial normalisation of portal vein flow pattern after LVAD speed reduction provides further support to the utility of this marker as a dynamic marker of right-sided congestion relief. If VExUS had been evaluated prior to empiric volume administration, unnecessary fluid loading may have been avoided and LVAD speed optimisation achieved in a timely fashion. However, in the absence of large multicentre studies, it should be interpreted as part of an integrated haemodynamic phenotype combining baseline right-sided vulnerability, postoperative physiology, ventricular interdependence and device settings.

Importantly, in this patient, venous Doppler abnormalities should not be ascribed to LVAD speed or suction-associated physiology alone. Pre-existing pulmonary hypertension, moderate tricuspid regurgitation, RV dysfunction and the immediate postoperative state may independently affect hepatic, portal and intrarenal venous Doppler waveforms. Therefore, VExUS should be interpreted in the full clinical and echocardiographic context, including ventricular geometry, septal position, RV function, valvular lesions, LVAD parameters, and the overall postoperative trajectory.

Therefore, the use of splanchnic vein Doppler (VExUS) into the routine haemodynamic evaluation might improve diagnostic accuracy and help with interventions, such as LVAD speed titration, especially in the management of hypotension of uncertain aetiology. Invasive measurement of right atrial pressure remains the reference standard for estimating venous congestion, but is not routinely feasible. Commonly used noninvasive surrogates such as IVC diameter have important limitations. VExUS is a non-invasive bedside technique that can record the downstream effects of increased venous pressures in various venous territories, thus providing a dynamic assessment of congestion and its response to therapeutic interventions[13].

Although VExUS is increasingly utilized in critically ill populations to evaluate venous congestion and inform fluid management[14], evidence regarding its application in patients receiving durable mechanical circulatory support is still lacking. This case demonstrates that the integration of VExUS assessment into conventional echocardiographic evaluation may enhance diagnostic precision in the management of LVAD-associated haemodynamic instability. In practical terms, VExUS may be incorporated as an adjunctive step in the bedside evaluation of LVAD-supported patients with unexplained hypotension, low-flow alarms, suspected suction physiology, RV dysfunction, or during postoperative haemodynamic optimisation and weaning from inotropic or mechanical support. Initial assessment of LVAD parameters should be followed by focused echocardiography to evaluate LV size, septal position, RV size and function, valvular lesions, and pericardial effusion. Venous Doppler assessment may then help determine whether systemic venous congestion is present. Because validated VExUS thresholds for LVAD speed adjustment are currently lacking, VExUS should not be used as a stand-alone trigger for pump speed changes, fluid removal, or fluid administration, but rather as part of a multimodal haemodynamic assessment[15]. Further studies are necessary to assess and validate the clinical utility of VExUS in patients with LVAD implantation.

Several limitations should be noted. First, this is a single case report and should therefore be considered hypothesis-generating rather than evidence of the efficacy or safety of a VExUS-guided strategy in LVAD patients. Second, invasive haemodynamic confirmation was not available because the patient did not have a pulmonary artery catheter at the time of the event. Third, the empiric 250 mL saline bolus administered before the ultrasound assessment represents a potential temporal confounder when interpreting early venous Doppler changes. Fourth, venous Doppler abnormalities may have been influenced by pre-existing tricuspid regurgitation, pulmonary hypertension, and RV dysfunction. Baseline venous Doppler assessment was not available; therefore, we could not determine whether the severe VExUS abnormalities were entirely new or partly reflected pre-existing RV dysfunction. Finally, obtaining high-quality Doppler signals requires appropriate training, Doppler alignment, and adequate acoustic windows. These requirements may be particularly challenging in LVAD patients after cardiac surgery because of altered thoracic anatomy, mechanical ventilation, body habitus, and limited patient positioning. As a result, feasibility, reproducibility, interobserver and intraobserver agreement, and the learning curve of VExUS in LVAD populations require further investigation.

CONCLUSION

VExUS assessment may be a useful adjunctive diagnostic tool in selected LVAD-supported patients presenting with low-flow episodes or unexplained hypotension. When integrated with transthoracic echocardiography, device parameters, and clinical assessment, VExUS may provide additional information on systemic venous congestion and help distinguish congestion-related ventricular interdependence from isolated hypovolaemia. However, this single case remains hypothesis-generating, and further studies are required to validate the feasibility, reproducibility, and clinical impact of VExUS-informed assessment in LVAD populations.