- Email: [email protected]
Echocardiographic Findings in Continuous-Flow Left Ventricular Assist Device Pump Thrombosis
The 20th Annual Scientific Meeting
•
HFSA
S117
Surgery/Transplantation/Devices II 338 The Sub-Hemolytic Effect of Left Ventricular Assist Devices: Erythrocyte Survival as a Novel Biomarker of Device Thrombosis Ziad Taimeh 1 , Ryan Koene 1 , Julie Furne 2 , Ashish Singal 1 , Peter Eckman 3 , Michael Levitt2, Marc Pritzker1; 1Lillehei Heart Institute, University of Minnesota School of Medicine, Minneapolis, MN; 2Minneapolis Veterans Affairs Hospital, Minneapolis, MN; 3Minneapolis Heart Institute, Abbott Northwestern Hospital, Minneapolis, MN Background: Continuous-flow left ventricular assist devices (CF-LVAD) have transformed treatment for end-stage heart failure. Device thrombosis is a feared complication with challenging diagnosis. We investigated the sub-hemolytic injurious effect of CF-LVAD on circulating erythrocytes, in-vivo, using erythrocyte lifespan. Erythrocyte lifespan is a well-validated marker that employs the endogenous breath carbon monoxide excretion as a quantitative indicator of erythrocyte turnover. Methods: Sixty non-smoking subjects were prospectively enrolled. Twenty-five subjects had a CF-LVAD without thrombosis (mean age 60.5 ± 15.3 years, 76% male), 10 subjects had a CF-LVAD with thrombosis (67.3 ± 5.1 years, 90% male), and 25 subjects were normal controls (57.2 ± 14.1 years, 92% male). Results: The mean end-tidal breath carbon monoxide level was significantly higher in CF-LVAD subjects with (5.69 ± 2.54 ppm) compared to those without (2.88 ± 0.70 ppm) device thrombosis (P < .0001). The levels in these 2 cohorts were significantly higher compared to normal controls (2.25 ± 0.59 ppm, P < .0001, and P = .001, respectively). The calculated mean erythrocyte lifespan was significantly shorter in CF-LVAD subjects with (29.7 ± 14.9 days) compared to those without (65.0 ± 17.3 days) device thrombosis (P < .0001). The lifespans in these 2 cohorts were significantly shorter compared to normal controls (96.0 ± 24.9 days, both P < .0001). Receiver operator curve analysis demonstrated high sensitivity and specificity for use of erythrocyte lifespan to detect device thrombosis (area under the curve = 0.94). Conclusion: The inherent mechanical function of CF-LVAD induced significant sub-hemolysis, which was more pronounced in the presence of device thrombosis. Further studies are needed to elucidate the utility of erythrocyte lifespan for the detection of subtle pre-clinical device thrombosis, as well as to assess its temporal relationship to the mechanism of thrombus formation.
339 Echocardiographic Findings in Continuous-Flow Left Ventricular Assist Device Pump Thrombosis Juan M. Ortega-Legaspi, Sophia Airhart, Hoang Nguyen, S. Carolina Masri, James Kirkpatrick; University of Washington, Seattle, WA Background: Left ventricular assist device (LVAD) pump thrombosis is not uncommon and confers a high mortality risk, with a probability of death within 3 months of 24%. Early diagnosis and treatment improves outcomes. Ramp studies may be of reduced utility to detect thrombosis, and other echocardiographic methods have been proposed but have not been widely adopted. We investigated changes in 2D and Doppler echo parameters in patients with and without pump thrombosis. Methods: Echo data was analyzed in 20 patients with known LVAD thrombosis and a control group of 20 without LVAD thrombosis in a cohort of end stage HF patients who underwent continuous flow LVAD implant from 2011 to 2014. Peak systolic to nadir diastolic flow velocity ratios (S/D) from Doppler interrogation of the LVAD inflow and outflow cannulas, mitral regurgitation (MR), aortic valve (AV) opening and right ventricular (RV) size and function were assessed and compared between echos from the control group, case group prior to suspected thrombosis (pre), and case group immediately prior to diagnosis of thrombosis. Results: In the LVAD thrombosis cohort, average age was 52, 79% males, 36% (12/33) of devices were implanted as destination therapy (DT) and 95% (19/20) were HeartMate II (HMII). In the control group, average age 56, 81% males, 31% (16/47) were implanted as DT and 80% (16/20) were HMII. Peak plasma free hgb and LDH were significantly higher in the thrombosis group at 81 mg/dl vs 22 mg/dl and 1872 mg/dl vs 548 mg/dl, respectively. The average S/D flow velocity ratios across the inflow cannulae (HMII device only) was 3.3 in the control group, 4.8 in echos performed pre-thrombosis and 3.8 at the time of thrombosis (P > .05). The outflow ratio was 2.9 in the control group, and 3.1 and 4.3 in pre-thrombosis and thrombosis echoes, respectively, with a significant difference between the control and the thrombosis echos (P = .01) and between pre-thrombosis and thrombosis echos (P = .02). AV opening on every cardiac cycle was not seen in the control group and was seen on 8 (40%) pre-thrombosis and 14 (70%) thrombosis echos. There were no significant differences between RV function and RV size between groups. Conclusion: This study confirms prior work indicating that an increase in the peak systolic to nadir diastolic velocity ratios (outflow cannula) serves as a marker for pump thrombosis. An increase in this ratio over time may be useful for early detection of pump thrombosis. Aortic valve opening in every cardiac cycle may also serve as a useful marker. Future research is necessary to compare the accuracy of this ratio to ramp studies examining changes in LV size.
340 Severity of Right Heart Failure Post LVAD Implantation Predicts 1 Year Mortality Mohamad Alghothani, Salil Kumar, Bryan Lee, Sakima Smith; The Ohio State University, Columbus, OH Background: The presence of right ventricular failure (RVF) following LVAD implantation has been identified as a predictor of mortality. Our goal was to investigate whether the severity of RVF using INTERMACS criteria was a better method to determine 1-year mortality post LVAD implantation. Methods: We performed a singlecenter retrospective analysis of 211 patients who received a Heartmate II LVAD between Jan 1st 2007 and Dec 31st 2014. RVF severity using INTERMACS criteria was deter-
•
HFSA
S117
Surgery/Transplantation/Devices II 338 The Sub-Hemolytic Effect of Left Ventricular Assist Devices: Erythrocyte Survival as a Novel Biomarker of Device Thrombosis Ziad Taimeh 1 , Ryan Koene 1 , Julie Furne 2 , Ashish Singal 1 , Peter Eckman 3 , Michael Levitt2, Marc Pritzker1; 1Lillehei Heart Institute, University of Minnesota School of Medicine, Minneapolis, MN; 2Minneapolis Veterans Affairs Hospital, Minneapolis, MN; 3Minneapolis Heart Institute, Abbott Northwestern Hospital, Minneapolis, MN Background: Continuous-flow left ventricular assist devices (CF-LVAD) have transformed treatment for end-stage heart failure. Device thrombosis is a feared complication with challenging diagnosis. We investigated the sub-hemolytic injurious effect of CF-LVAD on circulating erythrocytes, in-vivo, using erythrocyte lifespan. Erythrocyte lifespan is a well-validated marker that employs the endogenous breath carbon monoxide excretion as a quantitative indicator of erythrocyte turnover. Methods: Sixty non-smoking subjects were prospectively enrolled. Twenty-five subjects had a CF-LVAD without thrombosis (mean age 60.5 ± 15.3 years, 76% male), 10 subjects had a CF-LVAD with thrombosis (67.3 ± 5.1 years, 90% male), and 25 subjects were normal controls (57.2 ± 14.1 years, 92% male). Results: The mean end-tidal breath carbon monoxide level was significantly higher in CF-LVAD subjects with (5.69 ± 2.54 ppm) compared to those without (2.88 ± 0.70 ppm) device thrombosis (P < .0001). The levels in these 2 cohorts were significantly higher compared to normal controls (2.25 ± 0.59 ppm, P < .0001, and P = .001, respectively). The calculated mean erythrocyte lifespan was significantly shorter in CF-LVAD subjects with (29.7 ± 14.9 days) compared to those without (65.0 ± 17.3 days) device thrombosis (P < .0001). The lifespans in these 2 cohorts were significantly shorter compared to normal controls (96.0 ± 24.9 days, both P < .0001). Receiver operator curve analysis demonstrated high sensitivity and specificity for use of erythrocyte lifespan to detect device thrombosis (area under the curve = 0.94). Conclusion: The inherent mechanical function of CF-LVAD induced significant sub-hemolysis, which was more pronounced in the presence of device thrombosis. Further studies are needed to elucidate the utility of erythrocyte lifespan for the detection of subtle pre-clinical device thrombosis, as well as to assess its temporal relationship to the mechanism of thrombus formation.
339 Echocardiographic Findings in Continuous-Flow Left Ventricular Assist Device Pump Thrombosis Juan M. Ortega-Legaspi, Sophia Airhart, Hoang Nguyen, S. Carolina Masri, James Kirkpatrick; University of Washington, Seattle, WA Background: Left ventricular assist device (LVAD) pump thrombosis is not uncommon and confers a high mortality risk, with a probability of death within 3 months of 24%. Early diagnosis and treatment improves outcomes. Ramp studies may be of reduced utility to detect thrombosis, and other echocardiographic methods have been proposed but have not been widely adopted. We investigated changes in 2D and Doppler echo parameters in patients with and without pump thrombosis. Methods: Echo data was analyzed in 20 patients with known LVAD thrombosis and a control group of 20 without LVAD thrombosis in a cohort of end stage HF patients who underwent continuous flow LVAD implant from 2011 to 2014. Peak systolic to nadir diastolic flow velocity ratios (S/D) from Doppler interrogation of the LVAD inflow and outflow cannulas, mitral regurgitation (MR), aortic valve (AV) opening and right ventricular (RV) size and function were assessed and compared between echos from the control group, case group prior to suspected thrombosis (pre), and case group immediately prior to diagnosis of thrombosis. Results: In the LVAD thrombosis cohort, average age was 52, 79% males, 36% (12/33) of devices were implanted as destination therapy (DT) and 95% (19/20) were HeartMate II (HMII). In the control group, average age 56, 81% males, 31% (16/47) were implanted as DT and 80% (16/20) were HMII. Peak plasma free hgb and LDH were significantly higher in the thrombosis group at 81 mg/dl vs 22 mg/dl and 1872 mg/dl vs 548 mg/dl, respectively. The average S/D flow velocity ratios across the inflow cannulae (HMII device only) was 3.3 in the control group, 4.8 in echos performed pre-thrombosis and 3.8 at the time of thrombosis (P > .05). The outflow ratio was 2.9 in the control group, and 3.1 and 4.3 in pre-thrombosis and thrombosis echoes, respectively, with a significant difference between the control and the thrombosis echos (P = .01) and between pre-thrombosis and thrombosis echos (P = .02). AV opening on every cardiac cycle was not seen in the control group and was seen on 8 (40%) pre-thrombosis and 14 (70%) thrombosis echos. There were no significant differences between RV function and RV size between groups. Conclusion: This study confirms prior work indicating that an increase in the peak systolic to nadir diastolic velocity ratios (outflow cannula) serves as a marker for pump thrombosis. An increase in this ratio over time may be useful for early detection of pump thrombosis. Aortic valve opening in every cardiac cycle may also serve as a useful marker. Future research is necessary to compare the accuracy of this ratio to ramp studies examining changes in LV size.
340 Severity of Right Heart Failure Post LVAD Implantation Predicts 1 Year Mortality Mohamad Alghothani, Salil Kumar, Bryan Lee, Sakima Smith; The Ohio State University, Columbus, OH Background: The presence of right ventricular failure (RVF) following LVAD implantation has been identified as a predictor of mortality. Our goal was to investigate whether the severity of RVF using INTERMACS criteria was a better method to determine 1-year mortality post LVAD implantation. Methods: We performed a singlecenter retrospective analysis of 211 patients who received a Heartmate II LVAD between Jan 1st 2007 and Dec 31st 2014. RVF severity using INTERMACS criteria was deter-