Outflow Cannula Research Articles

PEDS6: Virtual Anatomic Fit Study of PediaFlow Implantable VAD in Infants (4 Kg To 8 Kg)

Study: Because of the limited anatomic space available for an implantable VAD in infants, anatomic compatibility in infants is a critical design challenge. Unfortunately, there is limited anatomic data available for infants in the first six months of life. Yet the weight of the child nearly doubles during this period (4 kg to 8 kg for the 50 percentile). Therefore the goals of this research was to create a broad 3D anatomic database for the infant, derived from a limited set of serial CT scans of infants (4 kg – 8 kg) and to conduct a virtual fit study of the PediaFlow PF5 pediatric VAD, including inflow and outflow cannula. Methods: The series of axial images and two orthogonal images of 4 kg and 8 kg infants were imported into commercially available reconstruction software (Mimics) to create a 3-D surface model of the chest cavity. The clipping tool in Mimics was used to perform qualitative validation on two patients to assess the accuracy of the reference models for the database. The tool allowed the superimposition of axial, coronal, and sagittal CT imaging views onto 3D models, as shown in Figure 1. Interpolation of anatomic data from two reference patients resulted in an animation of the growth of the thoracic cavity and internal organs from 4 kg to 8 kg. Results: Figure 2 displays the virtual fitting of the PediaFlow PF5 in 5 intermediate anatomic models interpolated between 4 kg and 8 kg. Based on this analysis, we can appreciate the lower limit of patients for whom the PediaFlow and cannula can be fully implanted. Conclusion: Three-dimensional anatomic data for virtual fit studies in infants is limited. This study produced a scalable 3D model, derived from two reference CT scan data of two patients, that can be used for virtual fit studies, and can also be exported for 3D printing and 4D reconstruction. Figure 1. Qualitative validation was performed on the reconstructed anatomy of a 4-day-old patient infant. Figure 2. Results of virtual fit studies in 5 different anatomic models

Association of Outflow Cannula Angulation with Adverse Neurological Events in Patients with CF-LVAD

<h3>Purpose</h3> Adverse neurological outcomes are the Achilles heel of the current day LVAD technology and are associated with worse survival. Our previous data suggests that outflow cannula (OFC) angles measured using cardiac CT may correlate with stroke in continuous-flow HeartMate II patients. Therefore, the aim of our study was to assess the relationship of inflow cannula (IFC) and outflow cannula (OFC) position with neurological outcomes centrifugal flow pumps (HM3 and HVAD). <h3>Methods</h3> 108 patients underwent LVAD implantation between 2015 and 2019 at our institution. Of these, 68 patients underwent cardiac CT, and 6 were excluded due to poor image quality. Of these patients, a total of 32 neurological events (NE) occurred in 20 patients. These cases were compared to patients with CT data who did not have NEs. 2D angle measurements were performed on positions of OFC relative to the ascending aorta and IFC relative to the mitral valve. Further we used a novel point-based 3D measurement system to determine the angles in an x-y-z plane. Baseline demographic data and outcome data were collected for all patients. Optimal cut-point of the OFC angle was determined using ROC curve analysis. <h3>Results</h3> Of the 62 patients recruited for this study, 65% had HM3 and 35% had HVAD. The median OFC angle on 2D in patients with NE was 59.8 degrees relative to the aorta, compared with a median angle of 51.7 degrees in patients without NE (p=0.02). Similarly on 3D angle measurements, a median OFC angle of 86.7 degree was observed in patients with NE, compared to a median angle of 77.4 degrees in patients without NE (p=0.01). IFC angle was not associated with an increased risk of NE. The ROC curve analysis showed optimal OFC angle was < 53 degrees in 2D measurements, with sensitivity of 79% and specificity of 61% at predicting NEs. In 3D measurements the optimal OFC angle was less than < 83 degrees with a sensitivity of 79% and a specificity of 61% at predicting NEs. <h3>Conclusion</h3> Optimal angulation of the outflow cannula in LVAD patients can lead to decreased risk for HRAE. Using cardiac CT chest imaging for 2D and 3D measurements, an optimal outflow cannula angulation of < 53 degrees and < 83 degrees, respectively, should be preferred to minimize the risk for HRAE in patients with centrifugal flow LVAD support.

Three-Dimensional CT Reconstruction for Diagnosis of LVAD Outflow Graft Twist

<h3>Introduction</h3> Third generation LVAD, HeartMate3 (HM3) are associated with favorable prognosis. A rare life-threatening adverse event is a sudden outflow graft obstruction due to a rotational twist. There is limited data regarding the natural history of this complication. <h3>Case Report</h3> A 70-year-old male with dilated cardiomyopathy was implanted with an HM3 LVAD on September 2019. Echocardiography and cardiac CT immediately following the implantation showed appropriate left ventricular (LV) unloading and normal outflow cannula flows and patency. On April 2021 the patient complained of weakness. Clinical examination was normal, but LVAD interrogation identified decline of LVAD flow. Laboratory tests revealed INR in therapeutic range and normal LDH. An echocardiography revealed impaired unloading of the LV and opening of the aortic valve. A CTA showed a large irregular filling defect sized 7cm at the proximal outflow cannula. Differential diagnosis included outflow cannula twist or thrombosis. Three-dimensional CT volume rendering demonstrated a full rotational twist of the outflow cannula. An untwist operation using a lateral thoracotomy approach identified a >270º twist of the outflow graft. The cannula was untwisted and using a fixation device, the cannula was fixed, resulting in immediate improvement of the flow rate. Repeated transient low flow events at the end of the procedure mandated a repeat CTA to confirm the adequacy of cannula untwisting. Four months following the untwist procedure, the patient is well with stable LVAD flows. <h3>Summary</h3> This is the first case of an outflow graft twist obstruction with comprehensive CTA imaging at 3 critical time points: immediately after LVAD implantation, at twist occurrence and after surgical repair. We believe that it is very important to maintain a high degree of active awareness that a sudden drop in pump flow without signs of pump thrombosis may indicate outflow graft obstruction due to a twist, which is best recognized by three-dimensional CT reconstruction of the outflow graft.

Nutritional Status and Cannula Infections in Pediatric VAD Patients

<h3>Purpose</h3> Ventricular assist devices (VADs) have emerged as an effective treatment option to bridge pediatric patients to heart transplantation. Paracorporeal VADs require the placement of an inflow and outflow cannula, which can create an environment for infections. We aimed to characterize cannula infections in pediatric VAD patients and examine the impact of nutritional status. <h3>Methods</h3> This was a single-center retrospective study (2005-2020) of all patients <20 years old implanted with a VAD using Berlin Heart cannulas. A cannula infection was defined by symptoms in the presence of a positive culture and the need for antibiotic therapy. Malnutrition was defined as per the ASPEN guidelines. <h3>Results</h3> There were 54 patients eligible for the study. The median age at implant was 0.9 years (IQR 0.3, 4.1), 59% were male, with the most common etiology being non-congenital heart disease (67%). The median time on VAD support was 89.5 days (IQR 34.3, 174.8). During the study period, 29.6% of patients developed a cannula infection with a median time to infection of 52 days (IQR 23.3, 109.0). There was no difference in terms of pre-VAD weight, height, malnutrition scores, BMI or albumin levels between patients with and without a cannula infection. However, lower pre-VAD pre-albumin levels was associated with cannula infections (p=.022), as was older age (p=.048). Seventy-five percent (n= ¾) of the patients with a g-tube developed a cannula infection and almost 75% of patients who developed a cannula infection were malnourished based on their pre-VAD weight z-scores. <h3>Conclusion</h3> Less than one-third of pediatric VAD patients with Berlin cannulas developed a cannula infection. Pre-VAD pre-albumin levels and older age were associated with the development of a cannula infection. In addition, the majority of the children with an infection were malnourished, suggesting that pre-implant nutritional status may be a modifiable risk factor. These results suggest a relationship between nutritional status and infection that warrants further study in the pediatric VAD population.

Optimization of Flow Dynamics During the HeartWare HVAD to HeartMate 3 Exchange: A Computational Study Assessing Differential Surgical Techniques

<h3>Purpose</h3> With the discontinuation of the HeartWare HVAD, patients who experience a device complication will require exchange to another pump. The HeartMate 3 (HM3) is now the only commercially available LVAD in the United States. The optimal surgical configuration for connecting the 10 mm HVAD outflow graft to the 14 mm HM3 outflow graft has yet to be defined. <h3>Methods</h3> A lumped-parameter computational fluid dynamic (CFD) model of the circulatory system, including HQ curves of the HM3 was created. The model was coupled with 3D CFD model, in which outflow cannula flow served as the boundary conditions for blood flow simulations in the aorta. In total, 4 variations of the outflow cannula were tested: 1) Anastomosis near the pump (long 10 mm); 2) Anastomosis near the aorta (long 14 mm); 3) Anastomosis 5 cm from the aorta (mid transition); 4) The entire cannula 14 mm (complete exchange). The flow patterns were characterized based on the results of 3D CFD simulations. <h3>Results</h3> Maximum velocity was highest in the configurations with longer sections of 10 mm graft (long 10 mm, 4.20 m/s; long 14 mm, 4.12 m/s, mid transition, 4.18 m/s; complete exchange, 2.27 m/s). This resulted in a 176% larger blood volume exposed to very high shear rate (SR) >5000 s⁻¹ for the long 10 mm vs the long 14 mm configuration (0.59 ml vs 0.21 ml) and a 805% increase compared to the complete exchange (0.065 ml). There was a 21% increase in right coronary flow (0.29 ml/s vs 0.24 ml/s) and a 22% increase in left coronary flow (1.41 ml/s vs 1.16 ml/s) with the long 10 mm compared to the 14 mm configuration. When the long 10 mm cannula was compared against complete exchange, the flow increase was 26% (0.29 ml/s vs 0.23 ml/s) for right coronary and 62% (1.41 ml/s vs 0.87 ml/s) for left coronary. <h3>Conclusion</h3> When planning for a HVAD to HM3 exchange, the location of the anastomosis and relative length of the 10 and 14 mm grafts has important implications related to shear rate and coronary blood flow which should be accounted for during surgery.

Successful Percutaneous Mechanical Suction Thrombectomy of Extracorporeal Filtration System Following Bilateral Lung Transplantation Secondary to COVID-Pneumonia

IntroductionCOVID infections show increased risk of thromboembolic events. We report a case of a 43 year old male with acute Covid-19 pneumonia necessitating veno-venous ECMO and RVAD support as bridge to pulmonary transplantation. At transplant, he had thrombus along his extra-corporeal pulmonary artery cannula necessitating percutaneous mechanical thrombectomy.Case ReportThe patient presented as a transfer to our institution with COVID-19 related ARDS in refractory respiratory failure with multiple bronchopleural fistulas. Shortly after admission, veno-venous ECMO was initiated and over time was fully ECMO dependent due to extensive tissue destruction with essentially no functional lung tissue. He was converted to right internal jugular-left subclavian vein ECMO-RVAD configuration while assessing for transplantation. After 135 days of support, a suitable donor was identified and was taken for bilateral lung transplantation with ECMO/RVAD support. This was complicated by a frozen chest, massive transfusion, and primary graft dysfunction necessitating postoperative maintenance of circulatory support. Intraoperatively, a large thrombus burden was found along the pulmonary artery outflow cannula. His chest was left open at that time while his graft recovered. Three days later, a percutaneous suction thrombectomy device was inserted through his right femoral vein and under TEE guidance, he underwent suction thrombectomy of the pulmonary artery cannula clot burden (Figure 1). He was decannulated and underwent chest closure thereafter. He was anticoagulated post-operatively and has not had any further thromboembolic events.SummaryAcute COVID-19 infection leads to a known increased risk of thromboembolic phenomena. We present an interesting approach to removal of ECMO-cannula associated thrombus in severe SARS-CoV-2 infection necessitating bilateral lung transplantation.

Long-Term Outcomes in Ventricular Assist Device Outflow Cannula Anastomosis to the Descending Aorta

BackgroundLeft ventricular assist device (LVAD) implantation via thoracotomy with outflow cannula anastomosis to the descending aorta is an alternative implantation technique that uses a single incision and avoids anterior mediastinal planes. We evaluated long-term survival and hospital readmissions after LVAD implantation via thoracotomy with outflow cannula anastomosis to the descending aorta. MethodsAdult patients implanted with a continuous flow centrifugal LVAD at an academic center were retrospectively analyzed. Patients were assigned to 1 of 2 cohorts based on the anastomosis site of the LVAD outflow cannula: ascending aorta cohort (Asc-Ao) and descending aorta cohort (Desc-Ao). Primary and secondary outcomes were survival and hospital readmissions during device support. Readmission analysis included patients with ≥30-day survival after discharge. Multivariable analysis and propensity score matching were performed. ResultsSurvival analysis included 330 patients (Asc-Ao: 272, Desc-Ao: 58). Readmission analysis included 277 patients (Asc-Ao: 231, Desc-Ao: 46) and a total of 1028 readmissions during 654 patient-years of follow-up were analyzed. There was no significant difference in in-hospital, 6-month, 1-year, 3-year, and 5-year mortality between the two cohorts. Readmission-free survival, 30-day readmission, number of admissions per year and hospital length of stay per year were not significantly different between the 2 cohorts after adjustment for patient characteristics. ConclusionsThis study found no difference in long-term survival or hospital readmissions between LVAD implantation via thoracotomy with outflow cannula anastomosis to the descending aorta and standard implantation.

Outflow cannula position for left ventricular assist device: A propensity score-matched study.

The clinical consequences of alternative outflow cannula positioning in patients undergoing left ventricular assist device implantation are unknown. We evaluated clinical outcomes in patients who underwent implantation with the outflow cannula implanted from the right side into the ascending aorta versus the left side of the pericardium. Fifty consecutive patients with terminal left heart failure underwent implantation using the Medtronic Heartware Ventricular Assist Device at Toronto General Hospital. Patients were divided between left and right outflow cannula positioning during implantation where anastomosis occurred in the ascending aorta. Propensity score matching using exact matching on the following pre-specified covariates: Interagency Registry for Mechanically Assisted Circulatory Support score, previous cardiac surgery, and preoperative inotrope use. Fifty consecutive patients (25 left implantation and 25 right implantation) were included in the unmatched cohort and 45 patients (25 left implantation and 20 right implantation) were included in the matched cohort. No significant differences in baseline demographics. Pump thrombosis occurred in 10% (n = 2) receiving right-sided implantation and 8% (n = 2) with left implantation (p = 1.00). Postoperative stroke occurred in 10% (2/20) with right implantation and 16% (4/25) with left implantation (p = .88). No difference in 1-year mortality between right 20% (5/25) and left 25% (5/20) implantation (p = .97). No observed difference in mortality when adjusting for competing risk of heart transplantation. There was also no difference in stroke, pump thrombosis, driveline infection. Larger studies are required to confirm these findings. These preliminary data support the use of left-sided anastomoses to facilitate subsequent re-entry during heart transplantation.

Percutaneous venopulmonary artery extracorporeal membrane oxygenation for right heart failure after left ventricular assist device insertion.

Right ventricular failure after left ventricular assist device (LVAD) insertion is associated with significant mortality and morbidity. Mechanical support options include right ventricular assist devices, venoarterial extracorporeal membrane oxygenation (ECMO) and venopulmonary artery ECMO, the latter often involving central cannulation. We sought to evaluate the feasibility and early outcomes of a truly percutaneous venopulmonary artery (pVPA) ECMO strategy, with the potential advantage of bedside removal once weaned. Data from a single tertiary centre were reviewed retrospectively from January 2014 to January 2019. During this time, 54 patients underwent LVAD insertion, with 19 requiring mechanical support for right ventricular failure. Among them, 10 patients received pVPA ECMO. Implantation of the pVPA ECMO was performed under transoesophageal echocardiography and fluoroscopy guidance, with an inflow cannula placed in the right atrium via the right femoral vein and an outflow cannula placed in the left pulmonary artery (PA) via the right internal jugular vein. Cannula insertion was 100% successful with no need for repositioning. Eight patients (80%) were able to be successfully weaned (at the bedside); 6 were discharged from the hospital and there were no cases of early sepsis, mediastinitis or thromboembolism. At follow-up, 5 patients had received transplants (50%), with 1 on LVAD support as destination therapy (10%). Survival was 60 ± 15% and 50 ± 16% at 6 and 12 months, respectively. pVPA ECMO is 100% technically feasible and is an efficient method for temporary right ventricular support after LVAD insertion with the advantage of simple bedside removal and avoidance of a PA graft remnant in the chest cavity.

Exercise capacity following ventricular assist device implantation via thoracotomy with outflow cannula anastomosis to the descending aorta.

Left ventricular assist device (LVAD) implantation via left lateral thoracotomy with outflow cannula anastomosis to the descending aorta is an alternative technique that avoids anterior mediastinal planes and requires a single incision. This study compares changes in exercise capacity following LVAD implantation with outflow cannula anastomosis to the descending aorta versus ascending aorta. Adult patients who received a continuous flow centrifugal LVAD implantation and completed both pre- and postimplantation cardiopulmonary exercise tests (CPETs) and or 6-minute walk tests (6MWT) were included. Change in CPET parameters (maximum oxygen intake: vO2 max, oxygen uptake efficiency ratio: OUES, ventilatory efficiency ratio: vE/vCO2 Slope) and 6MWT distance were compared between ascending and descending aorta anastomosis groups. Ascending and descending aorta anastomosis cohorts included 59 and 14 patients, respectively. Pre- and postimplantation CPETs were performed 63±12days before and 216±17days following implantation. The improvement in CPET parameters (vO2 max, OUES, vE/vCO2 Slope) or 6MWT distance was not significantly different between the ascending and descending aorta anastomosis groups. This study found no significant difference in the improvement of CPET parameters or 6MWT distance between LVAD implantation via thoracotomy with outflow cannula anastomosis to descending aorta and standard implantation via sternotomy with outflow cannula anastomosis to ascending aorta.

Heart Mate II Deactivation Using a Left Atrial Appendage Occluder in the Outflow Cannula

Introduction Reverse cardiac remodeling may occur in some LVAD recipients. Although considered the standard therapy, surgical device explantation with repeat sternotomy might be undesirable or very high risk. On the other hand, there are few data reporting minimally invasive percutaneous LVAD deactivation. We describe a case of a man with LVAD malfunction due to driveline fracture and LV function recovery who had a Heart Mate II deactivated with a percutaneous technique using a left atrial appendage occluder (LAAO) positioned inside the outflow cannula. Case Report A 48 yo male patient with previous diagnosis of dilated cardiomyopathy and HMII implanted in 2017 as bridge to transplant was admitted due to device malfunction and alarm triggering after coughing crisis. On admission, the pump was off probably due to driveline fracture, although the patient was asymptomatic and hemodynamically stable. An echocardiogram showed cardiac remodeling and ejection fraction improved from 29% to 69%. The heart team proposed to keep the pump off and to perform a minimally invasive outflow cannula occlusion. A cardiac CT was performed and according to the measured land zone, the LAAO device Lambre (Lifetech Scientific Corp., Shenzhen, China)1824 mm was selected. Through a 5F sheath in the left common femoral artery, a pig-tail catheter was used to obtain an aortography. Through an 8F sheath in the right femoral artery, a 5F pig-tail catheter was placed inside the outflow cannula, then an Amplatz Super-Stiff (Boston Scientific, Marbourough, MA, US) 0.035’’inch wire was advanced to the distal aspect of the graft. Using an over-the-wire technique, a 9F Lambre delivery sheath was positioned, and the Lambre 1824 mm device was released, occluding the outflow cannula from the anastomosis. A new aortography showed decreased flow to the graft. After the procedure, the patient was maintained on anticoagulation and persisted asymptomatic. Echocardiogram one week later showed no flow in the outflow graft. LV ejection fraction remained 60%. He is now in hospital discharge planning. Summary To the best of our knowledge this the first report of LVAD deactivation with the fully recapturable Lambre LAAO device. We propose that the use of a LAA occlude to obstruct HM II outflow cannula is feasible and safe.

Left Ventricular Assist Device Graft Obstruction Relieved by Transcutaneous Stenting in a Teenage Patient

<h3>Introduction</h3> We describe a case of LVAD outflow graft stenosis from external outflow graft compression in a teenage patient successfully treated with percutaneous stenting. <h3>Case Report</h3> 17-year-old female with a continuous flow left ventricular assist device implanted 3 years prior for heart failure secondary to cardiomyopathy due to a TTN gene mutation. She presented with 1 week of worsening shortness of breath and 3 kilogram weight gain. LVAD flows were 3 liters-per-minute (L/min). The patient was referred for transcatheter stent therapy due to concern for outflow cannula obstruction. Hemodynamics during initial catheterization showed a gradient of 30 mmHg across the conduit. A stent was deployed in the distal aspect of the outflow cannula. A second stent was placed to reinforce the area of persistent stenosis. Post-intervention hemodynamics showed an improvement in the gradient to 13 mmHg. She re-presented 3 months later due to LVAD low flow alarm, decreased flows to 0.3 L/min, and decreased wattage of 2.9 watts (Figure 1). CT angiography (CTA) demonstrated 80% narrowing of the outflow conduit, proximal to previously placed outflow graft stents (Figure 2a-b). Low density material between the outflow graft and circumferential wrap was identified by radiology. The previously placed stents were first re-dilated. Three additional stents were placed within the LVAD outflow cannula. We placed the stents from aortic end toward LVAD inflow. In total, the outflow graft was stented for a length of approximately 7.5cm (Figure 3). Angiography showed mild residual narrowing at the anastomotic site. LVAD flows improved to 3.5L/min. Repeat CTA showed no residual evidence of stenosis. <h3>Summary</h3> To our knowledge this is the first reported case of external compression of an LVAD outflow graft from an overwrap treated with successful transcatheter stent treatment in the pediatric population. By stenting the outflow graft we believe we pushed the built-up material into the mediastinum, relieving the obstruction.

Left Ventricular Assist Device Flow Pattern Analysis Using Computational Fluid Dynamics at the Time of Invasive Hemodynamic Ramp Study: Using Patient-Specific Data to Optimize the Ramp Study

<h3>Purpose</h3> Invasive hemodynamic ramp studies have been shown to decrease the rates of left ventricular assist device (LVAD) complications however hemocompatibility related adverse events (HRAE) still remain unacceptably high. Computational fluid dynamics (CFD) allows for a more thorough analysis of shear rate (SR) and shear stress (SS) which influences endothelial function and thrombotic potential. <h3>Methods</h3> An invasive hemodynamic ramp study was performed on a patient with a HeartMate 3 (HM3). An anatomical model of the patient was reconstructed from computed tomography (CT) images, and the LVAD outflow cannula was used as the inflow boundary. The LVAD flow was calculated separately using a lumped-parameter-model (LPM) of the circulation. Ventricular volumes were reconstructed from CT and the patient's hemodynamic tracings. The contribution of flow from the HM3 was implemented in the LPM via published H-Q curves. CFD simulations were then carried out over the entire clinical operating condition of the HM3 from 4600 to 6400 RPM. SS and SR were quantified in the outflow cannula, ascending aorta, aortic arch and cerebral circulation. <h3>Results</h3> As speed was increased from 4600 to 5500, and then to 6400 RPM, blood volume exposed to a SR greater than 500 s⁻¹ increased, respectively, by 1.46 fold and 2.18 fold (2.00 to 2.93 to 4.37 cm³). Increase of SS in the cannula was 1.37 fold and 1.74 fold (1.88 to 2.56 to 3.26 Pa), in the ascending aorta was 1.44 fold and 1.92 fold (0.60 to 0.86 to 1.14 Pa), and in the aortic arch and cerebral circulation was 1.40 fold and 1.88 fold (0.46 to 0.65 to 0.87 Pa). <h3>Conclusion</h3> Hemodynamic optimization during ramp studies needs to be balanced with increased rates of SS and SR that occur with incremental speed increases. CFD when applied to a ramp study can provide additive patient-specific information that may further mitigate HRAE.

HeartMate III™ Implantation: A Novel Technique to Reduce the Incidence of Perioperative Bleeding and the Use of Intraoperative Blood Products

<h3>Purpose</h3> We describe a method for implanting the Abbot HeartMate 3™ (HM3) Left Ventricular Assist Device (LVAD) that may reduce perioperative bleeding and the use of blood products. <h3>Methods</h3> First a median sternotomy is performed and the HM3 prepared. The driveline is tunneled through the subcutaneous tissue and the patient is placed on cardiopulmonary bypass (CPB). The HM3 is implanted via the "cut then sew" method. A ring is made from Teflon™ felt and placed over the cuff of the HM3 sewing ring (Figure 1A-C). After coring the LV apex, a second donut shaped felt ring (1D) is placed (1E). Sutures are placed through the felt and LV wall and passed through the sewing cuff. The cuff is lowered, sutures tightened (1F-G), and more sutures are placed through the cuff and passed through the myocardium and felt strip (1H). The inflow and outflow cannula are attached, and patient removed from CPB. To determine the effect of this method on perioperative bleeding, we reviewed the charts of 114 patients that underwent HM3 placement in our institution between October 2015 and June 2019 by 4 surgeons, 1 of whom used this method. <h3>Results</h3> Table 1 shows our results and suggests that the novel method results in less perioperative bleeding, as indicated by both comparison of pre and postoperative hemoglobin and hematocrit and the amount of blood products given intraoperatively. The novel method also resulted in a shorter procedure time, possibly due to improved hemostasis. <h3>Conclusion</h3> Our addition of Teflon™ felt strips around the sewing cuff of the HM3 may reduce perioperative bleeding and the use of perioperative blood products.

Outcomes in LVAD Implantation via Lateral Thoracotomy with Outflow Cannula Anastomosis to the Descending Aorta

<h3>Purpose</h3> LVAD implantation via left lateral thoracotomy with outflow cannula anastomosis to the descending aorta is an alternative technique that uses a single incision and avoids anterior mediastinal planes in cases with hostile mediastinal anatomy. However, the long-term outcomes of outflow cannula to descending aorta anastomosis using contemporary LVADs are unknown and are evaluated in this study. <h3>Methods</h3> Adult patients implanted with a continuous flow centrifugal LVAD (HeartWare HVAD) in an academic center were included to the study. Hospital stay following implantation, short & long-term survival and readmission patterns were compared between ascending and descending aorta anastomosis groups. Readmission analysis only included patients that survived >30 days following the index hospitalization. <h3>Results</h3> 338 patients (Asc Ao: 280, Desc Ao:58) were included to the study. Readmission analysis included 277 patients (Asc Ao: 231, Desc Ao: 46) and a total of 1043 readmissions during 667 years of follow-up were analyzed (Table 1). There was no significant difference in hospital-stay following implantation or 90-day, 1-year, 3-year and 5-year mortality between the two groups. Readmission-free survival, 30-day readmission, number of admissions per year and hospital length of stay per year were similar. Number of admissions for specific causes were not significantly different (Table 2). <h3>Conclusion</h3> LVAD implantation via lateral thoracotomy with outflow cannula anastomosis to the descending aorta has a comparable survival and readmission profile with standard ascending aorta anastomosis in long-term follow-up.

Low Flow Alarms! A Case of Percutaneous LVAD Decommissioning

<h3>Introduction</h3> Low flow alarms (LFA's) in LVAD can occur due to a multitude of conditions. The differential diagnosis includes: hypovolemia, tamponade, right ventricular failure, hypertension, arrythmias, inflow cannula malposition, outflow cannula obstruction, and pre-pump inflow cannula thrombosis. We present a case of a patient with an HVAD with persistent LFA's and apparent myocardial recovery. <h3>Case Report</h3> A 44 y/o female who underwent destination therapy HVAD placement three years prior presented with one week of asymptomatic LFA's with low pulsatility. Device speed prior to presentation was 2400 RPM (flow 3 L/min, 6.4 L pulsatility, power 2.7 W); on presentation mean flows were 0.5 L/min, peak and trough 2.4 and 0 L/min, MAP was 59 mm Hg. TTE showed normal cannula position, closed aortic valve, mild right ventricular dysfunction. No arrythmias were present. LDH and plasma free Hgb were within normal limits. A controller exchange did not resolve LFA's. CT angiography showed no outflow graft obstruction. Invasive hemodynamics at 3100 RPM showed (mm Hg): RA pressure 8, RV pressure 42/8, PA pressure 40/14/28, PCWP 14/14/11, and cardiac index 4.3 L/min/m². Given no obvious cause of LFA and apparent myocardial recovery, we opted for percutaneous device decommissioning, given poor candidacy for surgical explant. Under general anesthesia and full heparinization, device speed was lowered to 1800 RPM for 5 minutes, during which invasive hemodynamics remained stable. A 0.035 inch Rosen wire was placed into the outflow graft with subsequent exchange to a long 6-French sheath into the outflow graft. A 12-mm Amplatzer vascular plug was then deployed into the outflow graft with subsequent ventriculogram demonstrating retrograde flow (Figure). The device was powered off, the driveline buried, and the patient discharged with lifelong coumadin (INR 2-3) & Aspirin 81 mg. <h3>Summary</h3> This case demonstrates feasibility of percutaneous LVAD decommissioning in patients with myocardial recovery or device-related complications not amenable to surgical explant.

Combining Invasive Cardiopulmonary Exercise Testing with Computational Fluid Dynamics to Better Understand LVAD Fluid Mechanics during Exercise

<h3>Purpose</h3> Invasive cardiopulmonary exercise testing (iCPET) is helpful in evaluating dyspnea but is infrequently utilized in patient supported by left ventricular assist devices (LVAD). Computational fluid dynamics (CFD) allows for a more thorough analysis of fluid mechanics including shear stress (SS) and shear rate (SR). The additive information from iCPET and CFD may allow for optimal understanding of the pump-patient interaction during exercise. <h3>Methods</h3> Hemodynamic and ventilatory gas measurements were recorded as part of an iCPET study on a patient with a HeartMate 3 (HM3). Exercise using a bicycle ergometer by increments of 20 watts every 2 minutes was performed. An anatomical model of the vasculature was reconstructed from computed tomography (CT) images, and the LVAD outflow cannula was used as the inflow boundary. The LVAD outflow was calculated separately using a lumped-parameter-model (LPM) of the circulation. Ventricular volumes were reconstructed from CT and the patient's hemodynamic tracings. The contribution of flow from the HM3 was implemented in the LPM via published H-Q curves. CFD simulations were then carried out over the stages of exercise. Blood velocities, SS and SR were quantified in the vascular beds. <h3>Results</h3> Compared to rest, blood volume exposed to SR >1000 s⁻¹ increased 2.8 fold (0.47 to 1.31 cm³), while average SS increased 1.45 fold (2.61 to 3.79 Pa) in the cannula, 1.48 fold (0.89 to 1.31 Pa) in the ascending aorta, and 1.61 fold (0.67 to 1.08 Pa) in the aortic arch and cerebral circulation. Compared to rest, peak velocities increased nearly 5 fold in the aortic arch. <h3>Conclusion</h3> CFD applied to iCPET provides valuable information about variations in flow to the vasculature with exercise as well as shear rate and shear stress augmentation. This patient-specific information may help with optimization of LVAD speeds to meet the physiological needs of exercise.

Using Aortic Arch Short Axis Views during Transesophageal Echocardiographic Examination Facilitates Right Ventricular Assist Device Imaging

<h3>Introduction</h3> Recent increased utilization of percutaneous mechanical circulatory support for the management of cardiogenic shock and end-stage heart failure has resulted in the development and more widespread use of percutaneous right ventricular assist device (RVAD) technologies in right ventricular failure. RVADs require appropriate imaging to confirm correct positioning and ensure function. The proper imaging for this new technology has not been well described. We present two cases in which the upper esophageal (UE) aortic arch (AA) short-axis (SAX) view with transesophageal echocardiography (TEE) played an important role in RVAD management. <h3>Case Report</h3> <b>Case 1.</b> Device malposition: Four days after initial placement of a ProtekDuo cannula (CardiacAssist, Pittsburg, PA, USA) with extracorporeal RVAD under fluoroscopy, cannula malposition was noted intraoperatively. Under TEE UE AA SAX visualization, the cannula was advanced such that the tip remained in the main pulmonary artery throughout the cardiac cycle, ensuring proper flow (Figure 1A-C). <b>Case 2.</b> Device malfunction: One month after extracorporeal RVAD implantation, acute RVAD flow reduction and significant hemolysis were observed. The UE AA SAX view demonstrated thrombus formation on the tip of the RVAD outflow cannula (Figure 1D). The RVAD outflow cannula was exchanged surgically, and upon examination of the old cannula, the thrombus was noted to cause significant cannula exit obstruction. <h3>Summary</h3> These cases support the importance of the UE AA SAX TEE view in RVAD outflow imaging. When obtainable, this view should be included in routine RVAD assessment.

Dynamic LVAD Inflow Cannula Obstruction Due to Mobile Thrombi

<h3>Introduction</h3> While LVAD outflow cannula obstructions are commonly described, we present a rare case of fulminant heart failure secondary to dynamic inflow cannula obstruction due to mobile thrombi within the left ventricular cavity. <h3>Case Report</h3> A 65 year old man with a history of ischemic cardiomyopathy and chronic kidney disease (CKD) underwent implantation of a continuous flow left ventricular assist device (LVAD; HeartWare™ HVAD™, Medtronic plc) as a bridge to transplantation. Three years following implantation, he presented with progressive fulminant right heart failure (RHF) followed by a sudden decrease in LVAD flow from 3.5 to 1.5 liters/min (Figure 1A). INR was 2.6, lactate dehydrogenase (LDH) level was normal and transthoracic echocardiogram showed a midline septum and the aortic valve opening with every beat. On admission, flows did not improve despite inotropic support. Outflow graft obstruction was suspected, however imaging was not pursued due to the risk of further renal injury with intravenous contrast, and that ultimately definitive treatment strategy remained heart transplantation. His UNOS listing status was upgraded from 4 to 2E. Anticoagulation was not escalated beyond aspirin and warfarin due to recurrent gastrointestinal bleeding. He continued to experience recurrent abrupt changes in LVAD flows (Figure 1B) without additional evidence to support pump thrombosis. Five weeks later, at the time of heart-kidney transplant, a partially obstructive thrombus within the inflow cannula as well as a mobile thrombus attached to the endocardium, prolapsing intermittently into the inflow cannula were visualized (Figure 1C-1D). <h3>Summary</h3> Dynamic inflow cannula obstruction should be suspected in the setting of heart failure and sudden changes in LVAD pump power consumption even when a lack corroborating imaging and laboratory evidence exists.

Vertical Ventricular Assist Device Placement to Systemic Right Ventricle for a Patient with Corrected Transposition of the Great Arteries and Mesocardia

Introduction Patients with physiologically repaired corrected transposition of the great arteries (cTGA) have morphological right ventricle (RV) as systemic ventricle. They may develop systemic RV failure in long-term and subsequently require heart transplantation with or without ventricular assist device (VAD) placement. However, VAD placement for systemic RV carries technical challenges, including trabeculated and diverse shapes of RV, close proximity to tricuspid valve and various apex positions. Case Report A case was 38-year-old female with cTGA and pulmonary stenosis. She had surgical history of physiological repair and redo pulmonary artery plasty at 10 and 19 years of age, respectively. She had history of heart failure symptoms with systemic RV dysfunction since 27 years of age. She was re-admitted for recurrent heart failure and listed for heart transplantation at 37 year of age. Her cardiac position was mesocardia and systemic RV was longitudinally elongated with severely dilated left atrium. The aorta was left anterior to the pulmonary trunk. Therefore, the standard VAD placement from cardiac apex may cause interference of sternal closure by pump body, inflow obstruction due to close proximity to the tricuspid valve, and difficult anastomosis to the short anterior ascending aorta. In contrast, the vertical VAD placement from diaphragmatic surface towards the aortic valve seemed to resolve these challenges. At the operation, a cardiopulmonary bypass was established through femoral cannulation. Under transesophageal echocardiogram guidance, the optimal inflow cannulation site was decided. After aortic cross-clamp, the systemic RV trabeculation was resected and the 3rd generation VAD pump was placed vertically from RV diaphragmatic surface. The outflow graft was laid along left side of the heart and anastomosed to left lateral ascending aorta. Postoperatively, the VAD flow had been stable with no evidence of sucking or thrombus formation. She was discharged and currently waiting for heart transplantation. Summary It is important to determine VAD placement technique with thorough understanding of unique characteristics of cTGA. Vertical VAD placement to systemic RV and anastomosis of outflow cannula to the left side of the aorta can be a good option for patients with systemic RV failure with mesocardia.