Secondary Logo

Journal Logo

Cardiovascular Anesthesiology: Echo Rounds

Intraoperative Use of Vascular Ultrasound to Localize Thrombus in Left Ventricular Assist Device Exchange

Smith, Eva S. MD; Nelson, Mark T. MD; Tang, Daniel G. MD

Author Information
doi: 10.1213/ANE.0000000000000980

A 61-year-old woman with a nonischemic cardiomyopathy underwent HeartMate II Left Ventricular Assist Device (LVAD) implantation (Thoratec, Pleasanton, CA; Fig. 1). On postoperative day 42, the patient became dyspneic, with the LVAD showing low flow and increasing power surges. Her blood chemistry was remarkable for increased lactate dehydrogenase, plasma-free hemoglobin, and low haptoglobin. A transthoracic echocardiogram was performed, which failed to identify thrombus but demonstrated increased left ventricular (LV) end-diastolic diameter and consistent aortic valve opening with each LV contraction. With these indications of LVAD thrombosis, LVAD exchange was scheduled. The LVAD was powered off because of continuous low-flow alarms.

Figure 1
Figure 1:
HeartMate II Left Ventricular Assist Device. This diagram illustrates the longer length of the outflow graft compared with the inflow and also shows that access to the inflow graft could be limited because of its position at the apex of the heart. Reprinted with the permission from Thoratec Corporation (Pleasanton, CA).

A preoperative transesophageal echocardiogram (TEE) was performed with the LVAD disabled. No thrombus was visualized, but a modest bidirectional Doppler flow within the outflow graft and proximal portion of the inflow cannula was noted, with anterograde systolic flow and retrograde diastolic flow (Fig. 2; Supplemental Digital Content 1, Video 1, This Doppler flow pattern is observed in disabled but patent axial flow devices and can be also observed in LVAD obstruction, although diastolic flow reversal is less in the latter.1 Because the LVAD was disabled in this case, the flow patterns seen in the figures/videos cannot be conclusively attributed to thrombosis only.

Figure 2
Figure 2:
Preoperative transesophageal echocardiogram with left ventricular assist device disabled. Left, Midesophageal 2-chamber view with color Doppler showing no flow convergence or flow inside inflow cannula during mid-systole. Right, Midesophageal ascending aorta long-axis view showing intermittent retrograde flow within the outflow graft. No distinct thrombus was visualized in these views. LA = left artery; LV = left ventricle.

With the chest open, the surgeon directly evaluated the outflow graft using a L15-7io Broadband Compact Linear Array transducer (Philips Healthcare, Bothell, WA). A sterile sheath with sterile ultrasound gel was used, and gel was also applied to the surface of the graft. A stand-off was not used, which resulted in a loss of 2 to 3 mm of near-field resolution. The increased frequency and compact size of the vascular probe made it ideal for this application.

Long-axis 2-dimensional views of the outflow graft showed a clear delineation of thrombus extending within several centimeters of the aortic anastomosis (Fig. 3). Short-axis images were not obtained, because they were not necessary for the purpose of determining the site of graft division. Color Doppler revealed minimal flow into and out of the outflow graft (Supplemental Digital Content 2, Video 2, During cardiopulmonary bypass (CPB), the outflow graft was divided distal to the thrombus, which completely obliterated the graft lumen, and a graft-to-graft anastomosis was sewn. The thrombosed LVAD impeller was also replaced. After CPB, TEE evaluation showed proper position and function of the LVAD. Color Doppler examination of the cannulas now demonstrated significantly increased antegrade flow and flow convergence at the inflow cannula (Fig. 4; Supplemental Digital Content 3, Video 3,

Figure 3
Figure 3:
Intraoperative long-axis view of left ventricular assist device (LVAD) outflow graft with thrombus identified using the vascular probe.
Figure 4
Figure 4:
After left ventricular assist device replacement. Left, Midesophageal 2-chamber view with color Doppler showing flow convergence at and flow into the inflow cannula. Right, Midesophageal ascending aorta long-axis view showing consistent antegrade flow through outflow graft.

After surgery, the patient was shifted to the intensive care unit where she had a lengthy recovery but was discharged to home after several months. Consent for publication of this case was obtained from the patient’s family.


Moazami et al.2 reviewed HeartMate II LVAD implantations from 2005 to 2010 and found a 6.4% incidence of operative LVAD replacement. Device alarms, worsening dyspnea, and abnormal hemolysis laboratories are all consistent with LVAD obstruction, commonly due to thrombosis. Two-dimensional echo findings consistent with LVAD occlusion include increased frequency of aortic valve opening and an increased LV diastolic internal diameter. Doppler echo findings include decreased diastolic device flow and increased systolic to diastolic velocity ratio3 measured at either inflow or outflow cannula.

Preoperative TEE showed modest amounts of bidirectional flow in the inflow and outflow grafts as described earlier but failed to identify thrombus in either. Comparing the inflow Doppler velocities with the pump thrombosed (Fig. 2; Supplemental Digital Content 1, Video 1, to the postoperative inflow examination (Fig. 4; Supplemental Digital Content 3, Video 3,, the presence of flow convergence at the inflow cannula and increased antegrade flow within the cannula are noted. Again, because the LVAD in this case was completely thrombosed and powered off, the difference in observed Doppler findings may not necessarily have been due to the presence of thrombus but because the pump was off.

The inability to visualize thrombus within the LVAD or cannula is not uncommon. In a study by Raman et al.4 comparing cardiovascular computed tomography (CCT) with transthoracic echocardiogram and TEE, echocardiography was found to be insensitive compared with CCT for identifying thrombus within LVAD cannulas. Although TEE was found to be less affected by acoustic window, it routinely did not provide adequate visualization of the entire inflow and outflow cannula because of limited depth of imaging and volume of coverage. In this patient, CCT was not performed before LVAD exchange because the echo findings were overwhelmingly suggestive of LVAD obstruction. As described by Uriel et al.,5 a ramp test, which considers mitral regurgitation, LV diastolic diameter, and aortic valve opening frequency and duration at different LVAD device speeds, can also aid in the diagnosis of LVAD malfunction.

The surgeon’s use of the vascular probe on the outflow graft to determine the presence and precise location of thrombus was critical. If thrombus was not visualized, the outflow cannula could simply be disconnected from the device and reattached to the replacement impeller. This would significantly reduce surgical and CPB times. If thrombus was identified within the outflow graft, its localization would determine where to divide the outflow graft. Dividing the outflow cannula blindly and physically removing thrombus distal to the division could result in graft endothelial disruption and an increased risk of thromboembolism. For this reason, the graft division needed to be accurate and beyond any visualized thrombus. Adequate visualization of the inflow cannula with the vascular probe was not feasible because of its short length, close proximity to the diaphragm, and limited physical access.

In summary, TEE evaluation of dysfunctional LVADs can present a unique challenge to the echocardiographer. However, examination of the LVAD outflow graft with a high-frequency compact vascular transducer can identify the presence, location, and extent of thrombus formation and provide valuable information to the surgeon on where to divide the graft during LVAD exchange.

Clinician’s Key Teaching Points

By Kent H. Rehfeldt, MD, Roman M. Sniecinski, MD, and Martin J. London, MD

  • Obstruction due to thrombosis is a well-described complication in patients with axial-flow left ventricular assist devices (LVADs). Clinical findings include dyspnea and fatigue, as well as laboratory markers of hemolysis. In addition, the device itself may go through “power spikes” and trigger low-flow alarms when significant thrombus forms around the pump bearings.
  • Although thrombus within the LVAD is best visualized with computed tomography, echocardiography can often provide the indirect signs of thrombotic obstruction. Normal flow within the inflow and outflow cannulas is low velocity (<2 m/s), laminar, and unidirectional. With obstruction, these velocities will increase and show evidence of turbulent flow on color flow Doppler. Because there are no valves within an axial flow device, diastolic flow toward the device in the inflow cannula should raise the suspicion of device malfunction. To exclude thrombus, a “ramp test” is often performed by incrementally increasing the LVAD speed during transthoracic echocardiogram or transesophageal echocardiogram (TEE) imaging. A normally functioning LVAD will show decreased frequency of aortic valve opening and decreased end-diastolic diameter; an obstructed LVAD will fail to show these expected responses.
  • In this case, intraoperative TEE was unable to directly visualize a thrombus within the device, although clinical suspicion remained high. This prompted the novel use of a linear ultrasound probe by the authors, directly scanning on top of the LVAD outflow graft, localizing the thrombus, and indicating the site where the graft could be safely clamped and divided to remove it.
  • Acoustic shadowing, imaging artifact, and distance from the transducer limit the ability of TEE and transthoracic echocardiogram to directly visualize thrombus within an LVAD. However, in a manner similar to epiaortic scanning, intraoperative use of linear vascular probes in the surgical field may be useful to precisely localize thrombus and indicate which components of the device need to be replaced.


Name: Eva S. Smith, MD.

Contribution: This author helped conduct the study and write the manuscript.

Attestation: Eva S. Smith approved the final manuscript.

Name: Mark T. Nelson, MD.

Contribution: This author helped design the study, conduct the study, and write the manuscript.

Attestation: Mark T. Nelson approved the final manuscript.

Name: Daniel G. Tang, MD.

Contribution: This author helped design the study and conduct the study.

Attestation: Daniel G. Tang approved the final manuscript.

This manuscript was handled by: Martin London, MD.


1. Myers TJ, Frazier OH, Mesina HS, Radovancevic B, Gregoric ID. Hemodynamics and patient safety during pump-off studies of an axial-flow left ventricular assist device. J Heart Lung Transplant. 2006;25:379–83
2. Moazami N, Milano CA, John R, Sun B, Adamson RM, Pagani FD, Smedira N, Slaughter MS, Farrar DJ, Frazier OHHeartMate II Investigators. . Pump replacement for left ventricular assist device failure can be done safely and is associated with low mortality. Ann Thorac Surg. 2013;95:500–5
3. Fine NM, Topilsky Y, Oh JK, Hasin T, Kushwaha SS, Daly RC, Joyce LD, Stulak JM, Pereira NL, Boilson BA, Clavell AL, Edwards BS, Park SJ. Role of echocardiography in patients with intravascular hemolysis due to suspected continuous-flow LVAD thrombosis. JACC Cardiovasc Imaging. 2013;6:1129–40
4. Raman SV, Sahu A, Merchant AZ, Louis LB IV, Firstenberg MS, Sun B. Noninvasive assessment of left ventricular assist devices with cardiovascular computed tomography and impact on management. J Heart Lung Transplant. 2010;29:79–85
5. Uriel N, Morrison KA, Garan AR, Kato TS, Yuzefpolskaya M, Latif F, Restaino SW, Mancini DM, Flannery M, Takayama H, John R, Colombo PC, Naka Y, Jorde UP. Development of a novel echocardiography ramp test for speed optimization and diagnosis of device thrombosis in continuous-flow left ventricular assist devices: the Columbia ramp study. J Am Coll Cardiol. 2012;60:1764–75
© 2016 International Anesthesia Research Society