Using pulsatility responses to breath-hold maneuvers to predict readmission rates in continuous-flow left ventricular assist device patients.

BACKGROUND: Dynamic respiratory maneuvers induce heterogenous changes to flow-pulsatility in continuous-flow left ventricular assist device patients. We evaluated the association of these pulsatility responses with patient hemodynamics and outcomes. METHODS: Responses obtained from HVAD (Medtronic) outpatients during successive weekly clinics were categorized into three ordinal groups according to the percentage reduction in flow-waveform pulsatility (peak-trough flow) upon inspiratory-breath-hold, (%∆P): (1) minimal change (%∆P ≤ 50), (2) reduced pulsatility (%∆P > 50 but <100), (3) flatline (%∆P = 100). Same-day echocardiography and right-heart-catheterization were performed. Readmissions were compared between patients with ≥1 flatline response (F-group) and those without (NF-group). RESULTS: Overall, 712 responses were obtained from 55 patients (82% male, age 56.4 ± 11.5). When compared to minimal change, reduced pulsatility and flatline responses were associated with lower central venous pressure (14.2 vs. 11.4 vs. 9.0 mm Hg, p = 0.08) and pulmonary capillary wedge pressure (19.8 vs. 14.3 vs. 13.0 mm Hg, p = 0.03), lower rates of ≥moderate mitral regurgitation (48% vs. 13% vs. 10%, p = 0.01), lower rates of ≥moderate right ventricular impairment (62% vs. 25% vs. 27%, p = 0.03), and increased rates of aortic valve opening (32% vs. 50% vs. 75%, p = 0.03). The F-group (n = 28) experienced numerically lower all-cause readmissions (1.51 vs. 2.79 events-per-patient-year [EPPY], hazard-ratio [HR] = 0.67, p = 0.12), reduced heart failure readmissions (0.07 vs. 0.57 EPPY, HR = 0.15, p = 0.008), and superior readmission-free survival (HR = 0.47, log-rank p = 0.04). Syncopal readmissions occurred exclusively in the F-group (0.20 vs. 0 EPPY, p = 0.01). CONCLUSION: Responses to inspiratory-breath-hold predicted hemodynamics and readmission risk. The impact of inspiratory-breath-hold on pulsatility can non-invasively guide hemodynamic management decisions, patient optimization, and readmission risk stratification.

Importance of electromagnetic interactions between ICD and VAD devices-Mechanistic assessment.

BACKGROUND: Implanted cardioverter defibrillators (ICDs) and left ventricular assist devices (LVADs) are established interventions that prolong life in advanced heart failure, but their combination has not been demonstrated as beneficial. Electromagnetic interference (EMI) produced by a LVAD can preclude ICD interrogation with external programmers. We undertook a systematic evaluation of the LVAD-ICD interaction "in-vitro" to clarify the extent of this interaction. METHODS: Using explanted ICDs and VADs in a mock physiological rig, we assessed interrogation and reprogramming of ICD devices in the presence of a running LVAD. When connectivity between the ICD programmer and the ICD failed, we attempted three different techniques to re-establish connectivity: (1) Electromagnetic shielding of the ICD with a pseudo-faraday cage; (2) altering the LVAD speed; and (3) increasing the distance between the VAD and the ICD. RESULTS: We tested a total of 24 ICDs from different manufacturers in the presence of the Heartware (HW) and HeartMate 3 (HM3) LVADs. With HW, we only observed interaction with Biotronik ICD devices at very close range (0-6 cm). With HM3, only Medtronic ICD devices showed no interaction. Interactions could be mitigated by increasing the VAD-ICD distance. CONCLUSIONS: LVADs, notably the HM3, produce EMI that interferes with the communication between an ICD and its respective programmer. This may need to be considered when choosing the type of VAD to implant in patients with a previously implanted left-sided ICD. The only safe way to regain connectivity is to increase the distance between the VAD and the ICD, with patients raising their arm above their head.

Phenotyping of Stable Left Ventricular Assist Device Patients Using Noninvasive Pump Flow Responses to Acute Loading Transients.

BACKGROUND: Although it has been established that continuous flow left ventricular assist devices are sensitive to loading conditions, the effect of acute load and postural changes on pump flow have not been explored systematically. METHODS AND RESULTS: Fifteen stable outpatients were studied. Patients sequentially transitioned from the seated position to supine, passive leg raise, and standing with transition effects documented. A modified Valsalva maneuver, consisting of a forced expiration with an open glottis, was performed in each position. A sustained, 2-handed handgrip was performed in the supine position. The pump flow waveform was recorded continuously and left ventricular end-diastolic diameter measured during each stage using transthoracic echocardiography. Transitioning from seated to supine posture produced a significant increase in the flow and the ventricular end-diastolic diameter, consistent with an increased preload. The transition from supine to standing produced a transient increase in the mean flow and decreased the flow pulsatility index. At steady state, these changes were reversed with a decrease in the mean and trough flow and increased pulsatility index, consistent with venous redistribution and possible baroreflex compensation. Four distinct patterns of standing-induced flow waveform effects were identified, reflecting varying preload, afterload, and individual compensatory effects. A sustained handgrip produced a significant decrease in flow and increase in flow pulsatility across all patients, reflecting an increased afterload pressure. A modified Valsalva maneuver produced a decrease in the flow pulsatility while seated, supine, and standing, but not during leg raise. Five patterns of pulsatility effect during Valsalva were observed: (1) minimal change, (2) pulsatility recovery, (3) rapid flatline, (4) slow flatline with delayed flow recovery, and (5) primary suction. CONCLUSIONS: Acute disturbances in loading conditions produce heterogeneous pump flow responses reflecting their complex interactions with pump and ventricular function as well as reflex compensatory mechanisms. Differences in responses and individual variabilities have significant implications for automated pump control algorithms.

Is it safe to irradiate the newest generation of ventricular assist devices? A case report and systematic literature review.

An increasing number of mechanical assist devices, especially left ventricular assist devices (VADs), are being implanted for prolonged periods and as destination therapy. Some VAD patients require radiotherapy due to concomitant oncologic morbidities, including thoracic malignancies. This raises the potential of VAD malfunction via radiation-induced damage. So far, only case reports and small case series on radiotherapy have been published, most of them on HeartMate II (HMII, Abbott, North Chicago, IL, USA). Significantly, the effects of irradiation on the HeartMate 3 (HM3, Abbott) remain undefined, despite the presence of controller components engineered within the pump itself. We report the first case of a patient with a HM3 who successfully underwent stereotactic hypofractionated radiotherapy due to an early-stage non-small-cell lung cancer. The patient did not suffer from any complications, including toxicity or VAD malfunction. Based on this case report and on published literature, we think that performing radiotherapy after VAD implantation with the aid of a multidisciplinary team could be performed, but more in vitro studies and cases series are needed to reinforce this statement.

Beat-to-beat detection of aortic valve opening in HeartWare left ventricular assist device patients.

Continuous flow left ventricular assist devices (cfLVADS) result in a significant reduction in aortic valve (AV) opening, which has been associated with several complications. Reliable monitoring of AV opening is needed to determine whether pump speed adjustment may be able to minimize adverse outcomes. We assessed AV status continuously by echocardiography for 4 minutes in 3 states in 18 HeartWare HVAD patients: 2 minutes at rest, and 1 minute each following Valsalva maneuver and supine leg-raising. Using a previously described algorithm, beat-to-beat AV status was compared with the area under the curve of the normalized power spectral density analysis (PSD-AUC) for the corresponding beats of the pump speed waveform. Five thousand five hundred twenty-seven beats were analyzed. AV opening varied between 0% and 100% for the cohort with the median AV opening frequency 21.5%, and median duration of opening of 124 msec (range 0-279). The receiver operating characteristic (ROC) curve area for AV opening by the PSD-AUC algorithm was 0.95 (P < 0.0001). A PSD-AUC cut-off of 0.82 distinguished between an open and closed AV with 86% sensitivity and 93% specificity. Accuracy was similar in regular cardiac rhythm, atrial fibrillation or with frequent ventricular ectopic beats. Valsalva maneuver and leg-raising had no impact on accuracy. The PSD-AUC was strongly predictive of AV opening duration (P < 0.0001). We found that AV status and opening duration can be determined with high accuracy on a beat-to-beat basis irrespective of cardiac rhythm and with low level exercise and changes in filling.

The 2018 ISHLT/APM/AST/ICCAC/STSW Recommendations for the Psychosocial Evaluation of Adult Cardiothoracic Transplant Candidates and Candidates for Long-term Mechanical Circulatory Support.

The psychosocial evaluation is well-recognized as an important component of the multifaceted assessment process to determine candidacy for heart transplantation, lung transplantation, and long-term mechanical circulatory support (MCS). However, there is no consensus-based set of recommendations for either the full range of psychosocial domains to be assessed during the evaluation, or the set of processes and procedures to be used to conduct the evaluation, report its findings, and monitor patients' receipt of and response to interventions for any problems identified. This document provides recommendations on both evaluation content and process. It represents a collaborative effort of the International Society for Heart and Lung Transplantation (ISHLT) and the Academy of Psychosomatic Medicine, American Society of Transplantation, International Consortium of Circulatory Assist Clinicians, and Society for Transplant Social Workers. The Nursing, Health Science and Allied Health Council of the ISHLT organized a Writing Committee composed of international experts representing the ISHLT and the collaborating societies. This Committee synthesized expert opinion and conducted a comprehensive literature review to support the psychosocial evaluation content and process recommendations that were developed. The recommendations are intended to dovetail with current ISHLT guidelines and consensus statements for the selection of candidates for cardiothoracic transplantation and MCS implantation. Moreover, the recommendations are designed to promote consistency across programs in the performance of the psychosocial evaluation by proposing a core set of content domains and processes that can be expanded as needed to meet programs' unique needs and goals.

Derivation of indices of left ventricular contractility in the setting of continuous-flow left ventricular assist device support.

It is important to accurately monitor residual cardiac function in patients under long-term continuous-flow left ventricular assist device (cfLVAD) support. Two new measures of left ventricular (LV) chamber contractility in the cfLVAD-unloaded ventricle include IQ, a regression coefficient between maximum flow acceleration and flow pulsatility at different pump speeds; and K, a logarithmic relationship between volumes moved in systole and diastole. We sought to optimize these indices. We also propose RIQ, a ratio between maximum flow acceleration and flow pulsatility at baseline pump speed, as an alternative to IQ. Eleven patients (mean age 49 ± 11 years) were studied. The K index was derived at baseline pump speed by defining systolic and diastolic onset as time points at which maximum and minimum volumes move through the pump. IQ across the full range of pump speeds was markedly different between patients. It was unreliable in three patients with underlying atrial fibrillation (coefficient of determination R(2) range: 0.38-0.74) and also when calculated without pump speed manipulation (R(2) range: 0.01-0.74). The K index was within physiological ranges, but poorly correlated to both IQ (P = 0.42) and RIQ (P = 0.92). In four patients there was excellent correspondence between RIQ and IQ, while four other patients showed a poor relationship between these indices. As RIQ does not require pump speed changes, it may be a more clinically appropriate measure. Further studies are required to determine the validity of these indices.

Normalisation of haemodynamics in patients with end-stage heart failure with continuous-flow left ventricular assist device therapy.

BACKGROUND: New generation continuous-flow left ventricular assist devices (LVADs) utilise centrifugal pumps. Data concerning their effect on patient haemodynamics, ventricular function and tissue perfusion is limited. We aimed to document these parameters following HeartWare centrifugal continuous-flow LVAD (HVAD) implantation and to assess the impact of post-operative right heart failure (RHF). METHODS: We reviewed 53 consecutive patients (mean age 49.5 ± 14.1 yrs) with HVAD implanted in the left ventricle, at St. Vincent's Hospital, Sydney, between January 2007 and August 2012. Available paired right heart catheterisation (n=35) and echocardiography (n=39) data was reviewed to assess response of invasive haemodynamics and ventricular function to LVAD support. RESULTS: A total of 28 patients (53%) were implanted from interim mechanical circulatory support. Seventeen patients (32%) required short-term post-implant veno-pulmonary artery extracorporeal membrane oxygenation. At 100 ± 61 days post-implant, mean pulmonary artery pressure and mean pulmonary capillary wedge pressure decreased from 38.8 ± 7.7 to 22.9 ± 7.7 mmHg and 28.3 ± 6.4 to 13.4 ± 5.4 mmHg respectively (p<0.001). LV end diastolic diameter decreased from 71.3 ± 12.7 to 61.1 ± 13.7 mm and LV end-systolic diameter from 62.7 ± 12.3 to 53.9 ± 14.4mm (p<0.001). Aortic regurgitation remained trivial. Serum sodium increased from 133.3 ± 5.7 to 139.3 ± 2.8 mmol/L and creatinine decreased from 109.1 ± 42.5 to 74.3 ± 26.2 µmol/L (p<0.001). Across the entire cohort, the six-month survival/transplant rate was significantly lower for RHF patients (72.2%, n=18) compared to those without (96.9%, n=35, p=0.01). CONCLUSIONS: HVAD support improves haemodynamics, LV dimensions and renal function. Following implantation with a centrifugal continuous-flow LVAD, RHF remains a significant risk with a tendency to worse outcomes in the short to medium term.

Longitudinal analysis left ventricular chamber responses under durable LVAD support.

BACKGROUND: Left ventricular assist device (LVAD) support offers remodeling potential in some patients. Our goal was to use noninvasively derived pressure-volume (PV) loops to understand the effect of demographic and device variables on serial changes in cardiac function under pump support. METHODS: Thirty-two consecutive Medtronic HeartWare Ventricular Assist Device (HVAD) patients (mean 55.9 ± 12.3 years, 81.3% male) were prospectively recruited. Single-cycle ventricular pressure and volume were estimated using a validated algorithm. PV loops (n = 77) and corresponding cardiac chamber dynamics were derived at predefined postimplant timepoints (1, 3, 6 months). Changes in PV loop parameters sustained across the 6-month period were characterized using mixed-effects modeling. The influence of demographic and device variables on the observed changes was assessed. RESULTS: Across a 6-month period, the mean ventricular function parameters remained stable. Significant predictors of monthly improvement of stroke work include: lower pump speeds (2400 rpm vs 2500-2800 rpm) [0.0.051 mm Hg/liter/month (p = 0.001)], high pulsatility index (>1.0 vs <1.0) [0.052 mm Hg/liter/month (p = 0.012)], and ischemic cardiomyopathy indication for LVAD implantation (vs nonischemic) [0.0387 mm Hg/liter/month (p = 0.007)]. Various other cardiac chamber function parameters including cardiac power, peak systolic pressure, and LV elastance also showed improvements in these cohorts. CONCLUSIONS: Factors associated with improvement in ventricular energetics and hemodynamics under LVAD support can be determined with noninvasive PV loops. Understanding the basis of increasing ventricular load to optimize myocardial remodeling may prove valuable in selecting eligible recovery candidates.

Interaction between physical activity and continuous-flow left ventricular assist device function in outpatients.

OBJECTIVE: Studies have shown that pump output by continuous-flow left ventricular assist devices (LVADs) increases with graded exercise testing. However, data on pump behavior during activities of daily living and sleep, where cardiac output requirements vary markedly, are lacking. We sought to determine pump parameters and activity levels in stable patients receiving outpatient LVAD therapy. METHODS AND RESULTS: Eleven outpatients (mean age 51 ± 14 years, 9 male) with centrifugal continuous-flow LVADs underwent monitoring of LVAD flow, heart rate (HR), energy expenditure, and physical activity over 1 week in an outpatient setting. Physical activity was recorded with the use of a combined pedometer, accelerometer, and calorimeter Sensewear armband. Pump, HR, and physical activity parameters were time matched for correlation analysis. Outpatients had an average pump flow of 5.67 ± 1.27 L/min and engaged predominately in low levels of physical activity (mean daily step count 3,249/day). Across the entire cohort, pump flow exhibited strong univariate relationships with patients' energy expenditure (r = 0.73), step count (r = 0.69), HR (r = 0.73), sleep (r = -0.89), and skin temperature (r = -0.85; P < .0001 for all). Multivariate analysis suggested that pump output was predominantly affected by recumbent position, energy expenditure and skin temperature (r(2) = 0.84; P < .0001). Pump flow and power consumption were significantly lower during sleep than during wake periods (5.48 ± 1.31 L/min vs 5.80 ± 1.26 L/min; P < .001). CONCLUSIONS: Pump output from continuous-flow LVADs is adaptive to changes in activities of daily living. Circadian variation in pump flow is mostly explained by recumbency and activity levels. Despite adequate pump flow, many LVAD patients continue to live sedentary lifestyles.

Thrombolysis for suspected intrapump thrombosis in patients with continuous flow centrifugal left ventricular assist device.

The current recommended anticoagulation regimen during continuous flow centrifugal left ventricular device support is a combination of antiplatelet therapy as well as oral anticoagulation. Despite this, pump thrombosis occurs in rare situations. We report the risk factors and nonsurgical management and outcomes of five patients implanted with continuous flow centrifugal left ventricular assist devices who displayed clinical, hemodynamic, and laboratory features of intrapump thrombosis. This information may support the use of intravenous thrombolytics for suspected pump thrombus in these newer generation devices.

Death on an LVAD-Two sides of a coin.

LVADs are increasingly being used to support patients with end stage heart failure. As such, the traditional definition of death, the absence of a pulse, requires re-examination as it is no longer clinically relevant. We present two contrasting cases of "death" on LVAD support and present some ethical issues surrounding the end of life on LVAD support.

Safety and tolerability of sodium-glucose cotransporter-2 inhibitors in bridge-to-transplant patients supported with centrifugal-flow left ventricular assist devices.

BACKGROUND: The safety and tolerability of sodium-glucose cotransporter-2 inhibitors (SGLT2i) in patients with end-stage heart failure supported with left-ventricular-assist-devices (LVADs), irrespective of diabetes mellitus, is not known. METHODS: A retrospective analysis of 31 outpatients implanted with LVADs as bridge-to-transplant (BTT) was conducted. Patients with biventricular support, aged under 18 years, who were discharged from the index hospitalisation, or were prescribed SGLT2i prior to their first outpatient clinic were excluded. Patient demographics, laboratory studies, pump haemodynamic and adverse event data was collected. RESULTS: Sixteen (51.6%) of 31 patients were prescribed SGLT2i over median 101.5 days (37.5-190.8). No patients discontinued SGLT2i use or reported attributable adverse symptoms. No significant differences between patients prescribed SGLT2i compared to those SGLT2i-naïve were seen in: [1] renal function; [2] weight; [3] mean arterial pressure. There were numerically lower infection-related (n = 4 vs 7, HR 0.32 (0.08-1.28), p = 0.11) and haemocompatibility-related (n = 3 vs 4, HR 0.52 (0.09-2.83), p = 0.45) adverse events in the SGLT2i group, albeit non-significant. CONCLUSIONS: We found SGLT2i to be safe and well-tolerated in the BTT LVAD cohort with no significant difference in rates of infection or haemocompatibility-related adverse events with SGLT2i use. Larger studies will inform further beneficial effects of SGLT2i prescription in this cohort.

Major adverse kidney events predict reduced survival in ventricular assist device supported patients.

AIMS: There is limited data describing major adverse kidney events (MAKE) in patients supported with ventricular assist devices (VAD). We aim to describe the association between MAKE and survival, risk factors for MAKE, and renal trajectory in VAD supported patients. METHODS AND RESULTS: We conducted a single-centre retrospective analysis of consecutive VAD implants between 2010 and 2019. Baseline demographics, biochemistry, and adverse events were collected for the duration of VAD support. MAKE was defined as the first event to occur of sustained drop (>50%) in estimated glomerular filtration rate (eGFR), progression to stage V chronic kidney disease, initiation or continuation of renal replacement therapy beyond implant admission or death on renal replacement therapy at any time. One-hundred and seventy-three patients were included, median age 56.8 years, 18.5% female, INTERMACS profile 1 or 2 in 75.1%. Thirty-seven patients experienced MAKE. On multivariate analysis, post-implant clinical right ventricular failure and the presence of chronic haemolysis, defined by the presence of schistocytes on blood film analysis, were significantly associated with increased risk of MAKE (adjusted odds ratio 9.88, P < 0.001 and adjusted odds ratio 3.33, P = 0.006, respectively). MAKE was associated with reduced survival (hazard ratio 4.80, P < 0.001). Patients who died or experienced MAKE did not demonstrate the expected transient 3-month improvement in eGFR, seen in other cohorts. CONCLUSIONS: MAKE significantly impacts survival. In our cohort, MAKE was predicted by post-implant right ventricular failure and chronic haemolysis. The lack of early eGFR improvement on VAD support may indicate higher risk for MAKE.

Arterial Compliance and Continuous-Flow Left Ventricular Assist Device Pump Function.

Durable continuous-flow left ventricular assist devices (cfLVADs) demonstrate superior survival, cardiac functional status, and overall quality of life compared to medical therapy alone in advanced heart failure. Previous studies have not considered the impact arterial compliance may have on pump performance or developed arterial pressure. This study assessed the impact of alterations in arterial compliance, preload, and afterload on continuous-flow pump function and measured hemodynamics using an in-vitro pulsatile mock circulatory loop. Decreased arterial compliance was associated with a significant increase in arterial pressure pulsatility which was not evident in the flow pulsatility, as displayed in pump flow waveforms. There were marked changes in the pump flow waveforms due to the significant alteration in the aortoventricular gradient during diastole according to the changes in compliance. This study demonstrates that changes in systemic blood pressure, afterload, and left ventricular contractility each significantly affects the flow waveform. The association of hypertension with lower aortic compliance results in markedly decreased diastolic flow rates which may be important in contributing to a greater risk of adverse events under cfLVAD support.

Global best practices consensus: Long-term management of patients with hybrid centrifugal flow left ventricular assist device support.

OBJECTIVES: Six months after withdrawal of the HeartWare HVAD System (HVAD; Medtronic) from sale, approximately 4000 patients continue ongoing support with this device. In light of the diminishing experience, this global consensus document summarizes key management recommendations. METHODS: International experts with experience in the management of patients with ongoing HVAD support were invited to summarize key aspects of patient and pump management and highlight differences in the current HeartMate 3 (Abbott Laboratories) ventricular assist device. Clinicians from high-implanting HVAD sites reviewed current literature and reported experience to generate a consensus statement. RESULTS: Specific guidelines to assist in the management of ongoing HVAD patients are developed. Key management protocols and helpful techniques developed from experienced clinicians are combined into a short guideline document. As experience with HeartMate 3 increases, key differences in approach to management are highlighted, where appropriate. CONCLUSIONS: With decreasing worldwide experience in the ongoing management of HVAD-supported patients, this consensus guideline provides a summary of best practice techniques from international centers. Differences in HeartMate 3 management are highlighted.

Impact of frailty on mortality and morbidity in bridge to transplant recipients of contemporary durable mechanical circulatory support.

BACKGROUND: Frailty is associated with adverse outcomes in advanced heart failure. We studied the impact of frailty on postoperative outcomes in bridge to transplant (BTT) durable mechanical circulatory support (MCS) recipients. METHODS: Patients undergoing left ventricular assist device (LVAD, n = 96) or biventricular support (BiV, n = 11) as BTT underwent frailty assessment. Frailty was defined as ≥ 3 physical domains of the Fried's Frailty Phenotype (FFP) or ≥ 2 physical domains of the FFP plus cognitive impairment on the Montreal Cognitive Assessment (MoCA). RESULTS: No difference in mortality at 360 days was observed in frail (n = 6/38, 15.8%) vs non-frail (n = 4/58, 6.9%) LVAD supported patients, p = 0.19. However, there was a significant excess mortality in frail BiV (n = 4/5) vs non-frail BiV (n = 0/6) supported patients, p = 0.013. In all patients, frail patients compared to non-frail patients experienced longer intensive care unit stay, 12 vs 6 days (p < 0.0001) and hospital length of stay, 48 vs 27 days (p < 0.0001). There was no difference in hemocompatibility and infection related adverse events. The majority (n = 22/29, 75.9%) of frail patients became non-frail following MCS; contrastingly, a minority (n = 3/42, 7.1%) became frail from being non-frail (p = 0.0003). CONCLUSIONS: Abnormal markers of frailty are common in patients undergoing BTT-MCS support and those used herein predict mortality in BiV-supported patients, but not in LVAD patients. These findings may help us better identify patients who will benefit most from BiV-BTT therapy.

Alarms and Their Outcomes in Left Ventricular Assist Device Patients.

Low flow and suction alarms are provided to alert caregivers of changes in left ventricular assist device pump function but may be reset in clinical practice. We investigated the incidence and underlying causes of these alarms as well as their prognostic significance. HeartWare ventricular assist device patients (n = 113) were divided into quartiles based on their frequency of low flow and suction alarms. Survival and adverse events (thrombus, stroke, bleeding, and right heart failure) were compared between quartiles. Low flow alarms peaked in the first few months of pump support before dropping down to near negligible levels. Suction alarm frequency remained relatively constant throughout pump support. Although pump speeds (p < 0.001) and flow (p = 0.01) decreased over time, there was an increase in suction alarm frequency (p = 0.018), with no changes in low flow alarms. Patients with smaller body size (p = 0.016) and lower pump flows (p = 0.008) had higher frequencies of low flow alarms on multiple regression (p < 0.001). Patients with the highest low flow alarm frequency demonstrated poorer survival (p = 0.026). There was no relationship between suction alarm frequency and survival. There was also no relationship between either low flow or suction alarm frequency with strokes, gastrointestinal bleeds, pump thrombus, or right ventricular failure. Duration of alarm and intervention in response to the alarm was not assessed in this study. Further studies examining alarm duration and responses may inform future pump alarm algorithms.

Intermittent Papillary Muscle Suction: A Rare Cause of Left Ventricular Assist Device Low Flow.

Low flow alarms represent a management challenge in patients with left ventricular assist devices because they are often a consequence of complex patient-device interactions. We present a case of intermittent suction of the postero-medial papillary muscle into the left ventricular assist device inflow cannula during diastole, causing low flows. This case highlights the importance of a systematic approach and use of multiple investigation modalities in making an accurate diagnosis. (Level of Difficulty: Advanced.).

Ex Vivo Assessment of Different Oral Anticoagulant Regimens on Pump Thrombosis in a HeartWare Ventricular Assist Device.

BACKGROUND: In light of decreased intracranial hemorrhage with direct oral anticoagulants and concerns about their safety in continuous flow left ventricular assist devices, we conducted an ex vivo study of thrombus formation using multiple anticoagulation agents. METHODS: A continuous flow left ventricular assist device (HeartWare ventricular assist device) hemocompatibility loop was run using human blood under 7 conditions: control (no anticoagulation or antiplatelet); in vitro addition of aspirin; in vitro addition of apixaban at low dose (equivalent 2.5 mg twice daily); addition of apixaban at high dose (equivalent 5 mg twice daily); patients on warfarin; patients on apixaban (5 mg twice daily); and patients on dabigatran (150 mg twice daily). The primary outcome was time to formation of intrapump thrombosis. Secondary outcomes were reduction in clotting times over 1 hour, hemolysis, reduced platelet aggregation, and von Willebrand activity. RESULTS: Twenty-one runs were completed. Times to thrombosis in median (interquartile range) were control, 131 (127-134.5); in vitro aspirin, 124 (114.5-137); and patients on dabigatran, 131 (130.5-135.5) minutes, respectively. Times in patients on warfarin were, 137 (136.5-143.5); in vitro low-dose apixaban, 141 (138.5-142); and patients on apixaban, 140 (138-142.5) minutes, respectively. No thrombus formed in the in vitro high-dose apixaban group. There were no significant differences between the individual groups. When all apixaban groups were compared with nonapixaban groups, the time to thrombosis formation was significantly longer, 143 (137-150) versus 133.5 (128.5-140) minutes, P=0.02. There were similar changes in lactate dehydrogenase levels and other secondary end points. CONCLUSIONS: In an in vitro study of anticoagulation using human blood in a mock loop with a HeartWare HVAD, we demonstrated similar thrombosis times for apixaban and warfarin. Time to clotting was longer in the combined apixaban groups compared with combined other groups, but thrombosis times between individual groups were not significantly different.