- Email: [email protected]
Speed Reduction Does Not Restore High Molecular Weight Von Willebrand Multimers during Heart Mate II Support: An In-Vivo Analysis
S244
The Journal of Heart and Lung Transplantation, Vol 35, No 4S, April 2016
6( 62) Effect of an Institutional Protocol for Gastrointestinal Bleeding in Left Ventricular Assist Device Patients A.H. Healy ,1 J. Colarusso,1 S.H. McKellar,1 A. Koliopoulou,1 S.G. Drakos,2 C.H. Selzman.1 1Surgery, University of Utah, Salt Lake City, UT; 2Medicine, University of Utah, Salt Lake City, UT. Purpose: Gastrointestinal (GI) bleeding in patients with left ventricular assist devices (LVADs) is a common problem. We implemented an institutional GI bleeding protocol for the management of these patients to determine if a systematic approach to this problem would result in improved outcomes. Methods: We conducted a retrospective cohort study by reviewing our institutional experience with GI bleeds in LVAD patients between October 2010 and December 2014. In January 2013, an institutional GI bleeding protocol was initiated to systematically approach the diagnosis and treatment of these events. We compared the frequency of GI bleeds, the rate of interventions such as endoscopy and colonoscopy, and the use of resources such as transfusion of packed red blood cells for patients before (BP) and after (AP) the initiation of the protocol. Standard statistical methods were utilized. Results: During the study period, 43 patients were readmitted with 90 GI bleeds. The BP group consisted of 16 patients with 38 bleeds, while the AP group consisted of 27 patients with 52 bleeds. There were no differences between the groups with regard to age, gender, device type, or history of GI bleeding prior to LVAD implantation. The BP group had 2.37 ± 1.41 bleeds/patient compared with 1.93 ± 1.49 bleeds/patient in the AP group (p = 0.336). After protocol initiation, the number of esophagogastroduodenoscopies (EGDs)/bleed increased (0.68 ± 0.66 vs. 1.10 ± 0.82, p = 0.013), while the number of colonoscopies, capsule studies, push EGDs, and units of blood transfused was not different. Tagged red blood studies/bleed decreased after the initiation of the protocol (0.32 ± 0.53 vs. 0.02 ± 0.14, p = 0.002), as did the number of angiograms/bleed (0.18 ± 0.46 vs. 0 ± 0, p = 0.017). Conclusion: The initiation of a GI bleeding protocol led to a shift towards less resource intensive treatment modalities. Further patient accrual may lead to additional reduction in resource utilization, such as transfusion of red blood cells and patient admissions. 6( 63) Severity of Hemolysis Is Associated with Death and Ischemic Stroke during Veno-Arterial Extracorporeal Membrane Support O. Saeed , W. Jakobleff Jr., M. Chau, S. Rangasamy, M. Algodi, M. Makkiya, M. Cruz, S. Patel, S. Murthy, D. Sims, J. Shin, D. Goldstein, U. Jorde. Medicine, Albert Einstein College of Medicine Montefiore Medical Center, Bronx, NY. Purpose: Hemolysis has emerged as a marker of adverse outcomes after CF LVAD placement, however, its impact during Veno-Arterial Extracorporeal Membrane (VA ECMO) support is uncertain. Methods: A single center retrospective analysis of all adult patients placed on VA ECMO from May 2011 to September 2015 was conducted. Major demographics, ECMO characteristics and clinical outcomes including cerebrovascular accidents (CVAs) were retrieved. As a marker of hemolysis severity, the peak lactate dehydrogenase (LDH) level was collected during VA ECMO support. To analyze the impact of hemolysis on in-hospital mortality patients were categorized as survivors and non survivors. Results: 111 patients underwent VA ECMO placement. The mean age was 57±14 years, 41 were female and 60 patients expired during VA ECMO. Both survivors and non survivors had similar distributions of major demographics and laboratory parameters except for baseline creatinine (survivors: 1.6±1.2 vs. non survivors: 2.4±1.7, p= 0.003). The median peak LDH for survivors was lower at 771 (IQR: 456-1616) U/L in comparison to 1444 (IQR: 8473080) U/L for non survivors (p= 0.005, figure). Patients who eventually had an ischemic CVA (n= 11) had a greater median peak LDH 1863 (IQR: 8004014) U/L in comparison to those with no CVA 961 (IQR: 509-1710) U/L, p= 0.039, figure. Conclusion: Severity of hemolysis measured by LDH is a marker of higher mortality during VA ECMO and may predict onset of thrombotic complications such as ischemic CVA. Further studies are warranted on usage of hemolysis severity to guide VA ECMO management, including circuit exchange.
6( 64) Speed Reduction Does Not Restore High Molecular Weight Von Willebrand Multimers during Heart Mate II Support: An In-Vivo Analysis S. Patel , J. Patel, J. Rand, D. Goldstein, U. Jorde, O. Saeed. Medicine, Albert Einstein College of Medicine Montefiore Medical Center, Bronx, NY. Purpose: Acquired Von Willebrand Factor (vWF) deficiency due to the loss of high molecular weight multimers (HMWM) has been well documented during CF LVAD support. It has been proposed that lowering pump speed in response to clinical gastrointestinal bleeding (GIB) may decrease shear stress allowing for the return of HMWMs. In-vivo data supporting this practice is lacking. Methods: Subjects at least 30 days post implantation of a Heart Mate (HM) II were prospectively recruited from the LVAD clinic. After confirming INR was > 2.0, pump speed was decreased to 8000 rpm and maintained for 6 hours. Blood samples obtained at baseline and 6 hours were compared for 2 measures of acquired vWF deficiency: 1) the ristocetin cofactor activity to vWF antigen ratio (Rco:Ag) and 2) gel electrophoresis for vWF multimer distribution. Results: Four patients agreed to participation. They were 57±15 years old, all were male and had been on HM II support for 401±199 days. All patients tolerated speed reduction without any adverse events. At baseline speed, HMWMs were reduced in all 4 patients. After 6 hours at 8000 rpm, there was no change in the HMWM profile (Figure 1). Similarly, the Rco:Ag ratio was reduced (nl > 0.65) in 3 of 4 patients at baseline and did not significantly change after speed reduction (0.56 → 0.56, 0.74 → 0.67, 0.58 → 0.65, 0.47 → 0.45; p = 0.437). Conclusion: Decreasing pump speed during HM II support does not lead to restoration of HMW Von Willebrand multimers. These findings suggest there may be no benefit to speed reduction in response to GIB.
Abstracts S245 6( 65) A Multicenter Evaluation of Octreotide for Ventricular Assist Device Related Gastrointestinal Bleeding G.B. Smallfield ,1 S. Gunda,1 S. Emani,2 M.K. Kanwar,3 N. Uriel,4 P.C. Colombo,5 P.A. Uber,1 M.L. Sears,1 K.B. Shah.1 1Virginia Commonwealth University, Richmond, VA; 2Ohio State University, Columbus, OH; 3Allegheny General Hospital, Pittsburgh, PA; 4University of Chicago, Chicago, IL; 5Columbia, New York, NY. Purpose: Gastrointestinal bleeding (GIB) is a frequent complication after implantation of a left ventricular assist device (LVAD). Octreotide is used to treat GIB as well as bleeding prophylaxis for vascular ectasias. The data to support octreotide use in LVAD associated bleeding is sparse. We therefore evaluated the efficacy of outpatient administration of octreotide to prevent reoccurrence of an LVAD related GIB. Methods: We present an interim analysis from a multicentered retrospective cohort including data from five centers and 34 patients who were discharged from the hospital with subcutaneous and depot (LAR) injectable octreotide. Patients were included if they received outpatient octreotide rebleeding prophylaxis on discharge after bleeding with LVAD. We measured forward from the bleeding event until their next bleeding event, transplant, explant or death. Results: The patients in the study had a mean (SD) age of 62.6 (9.6) years of age and were predominantly white (62%) and male (55%). At the time of receiving octreotide, patients were on device support 30.1(SD= 15) months and 63% of patients had experienced a prior GIB. Of the 34 patients analyzed, 5 were excluded for incomplete rebleeding data. Ten (34.5%) of patient re-bled at median 2.0 [IQR: 1.3, 5.8] months. There was no difference in rebleeding rate between those receiving LVAD for destination v. bridge to transplant (30% v. 37%). Though not statistically significant, there were several factors that appear associated with rebleeding after treatment with octreotide. Patients who had GIB before LVAD had a higher risk rebleeding after octreotide (75% v. 29%). Exposure to sildenafil showed a 50% rebleeding rate v. 32%. Compared to daily injections, the use of depot monthly LAR formulation showed a lower rate of bleeding 29% v. 42% for short acting formulations. Conclusion: This preliminary analysis from an ongoing multicenter cohort demonstrates a promising role for the long acting formulation octreotide for LVAD related GIB to reduce the risk of rebleeding. Prespecified multivariate regression and time-dependent outcomes modeling on the complete cohort of 75 patients will further elucidate the efficacy treatment at the time of data presentation. 6( 66) Eptifibatide in the Treatment of Pump Thrombosis: What Is the Prescription? L.M. Peters , D.T. Majure, M.E. Rodrigo, F.H. Sheikh, M. Hofmeyer, E.J. Molina, S.W. Boyce, S.S. Najjar. MedStar Washington Hospital Center, Washington, DC. Purpose: The optimal medical therapy for LVAD pump thrombosis (PT) is unknown. Past reports have discouraged the use of platelet GPIIb/IIIa receptor inhibitors due to concern for lack of efficacy and increased bleeding. Given the limited published data on the use of eptifibatide, we sought to describe our single center experience. Methods: We reviewed all cases of PT at our center from 1/2010 through 9/2015 who received eptifibatide. Our center utilizes eptifibatide in treatment of PT if unfractionated heparin (UFH) therapy is unsuccessful. Eptifibatide is typically started at 0.5 mcg/kg/min and titrated up to a maximum of 2 mcg/ kg/min depending on clinical response. Results: A total of 45 distinct PT events occurred for a cumulative incidence of 11.5% out of 278 implants. Of the 44 events treated with medical therapy, 20 (45%) were treated with eptifibatide. Events were treated with UFH alone for an average of 5±4 days prior to addition of eptifibatide. The mean dose of eptifibatide was 1.6±0.5 mcg/kg/min, and it was administered for a mean of 6±3 days (range 3-11). Eptifibatide led to clinical resolution (CR) in 11 (55%) events. Nine of 13 patients who received the maximum dose of 2 mcg/kg/min had CR. In contrast, 2 of 7 patients who received ≤ 1 mcg/kg/min had CR. Of the patients who failed eptifibatide, 5 patients required pump exchange, 3 required a heart transplant, and 1 died five months after failed treatment for
pump thrombosis as he was deemed not a candidate for pump exchange or heart transplant. Bleeding during eptifibatide occurred in 8 events: 3 patients developed gastrointestinal hemorrhage, 2 had epistaxis, 1 had intracranial hemorrhage with modified Rankin score of 1 at last follow-up, and 2 developed other sources of bleeding. Only one patient required blood transfusion (1 unit) due to menorrhagia. Five (62.5%) of the bleeding events occurred in patients who received the maximum dose of eptifibatide. One patient developed an embolic cerebrovascular accident. No patient developed significant thrombocytopenia during therapy. Conclusion: In our single center experience, addition of eptifibatide to UFH in medical management of PT had a moderate efficacy (clinical resolution > 50%) with low risk of bleeding requiring transfusion. Further studies are needed to define the optimal dosing strategy and timing, as well as patient characteristics that predict a successful outcome with this therapy. 6( 67) Abnormal Pre-Operative Mixing Studies Identify VAD Patients at High Risk for Spontaneous Bleeding J.D. Remick ,1 L. Holton,2 M. Mangum,1 K. Evenson,1 R. Lewis,1 M. Puhlman,1 C. Moore,3 G. Ott,1 J. Abraham.1 1Center for Advanced Heart Disease, Providence Heart & Vascular Institute, Portland, OR; 2Department of Medicine, Providence St. Vincent Medical Center, Portland, OR; 3Hematology and Oncology, Providence Cancer Center, Portland, OR. Purpose: Bleeding and thromboembolic (TE) complications are important causes of morbidity and mortality in VAD patients. Pre-operative screening for coagulopathies may influence patient selection, management, and outcomes. Methods: We performed a retrospective study of 33 patients with no prior history of coagulopathy who underwent HeartMate II VAD implant. Mixing studies were performed on patients with abnormal baseline PT or PTT. A bleeding event was defined as an episode resulting in one or more of the following: death, reoperation, hospitalization or transfusion (> 2 units) and further characterized as spontaneous or procedure-related. A TE was defined as either a neurologic event (stroke or TIA), pump malfunction requiring exchange, or LDH elevation requiring hospitalization. Bleeding and TE were compared between patients with normal vs abnormal mixing studies (AMS) using Poisson regression. Results: Of 33 patients, 20 had AMS. 15/20 had a total of 24 bleeding events, of which 80% were spontaneous. 5/13 patients with normal mixing studies had a total of 6 bleeding events, of which 66% were spontaneous. The AMS cohort had 0.99 bleeding events per patient year vs. 0.39 events per patient year in the normal group (risk ratio 2.54; p = 0.04). 10 thrombotic events occurred in 6/20 patients with AMS (0.41 events per patient-year) compared to 6 thrombotic events in 4/13 patients with normal mixing studies (0.39 events per patient year; p= 0.91). Conclusion: AMS resulting from factor deficiencies or circulating inhibitors are frequently detected in HF patients without prior history of coagulopathy. Their presence confers a 2.54-fold increase in risk of bleeding following VAD implantation. Routine pre-operative screening for coagulation disorders is warranted.
The Journal of Heart and Lung Transplantation, Vol 35, No 4S, April 2016
6( 62) Effect of an Institutional Protocol for Gastrointestinal Bleeding in Left Ventricular Assist Device Patients A.H. Healy ,1 J. Colarusso,1 S.H. McKellar,1 A. Koliopoulou,1 S.G. Drakos,2 C.H. Selzman.1 1Surgery, University of Utah, Salt Lake City, UT; 2Medicine, University of Utah, Salt Lake City, UT. Purpose: Gastrointestinal (GI) bleeding in patients with left ventricular assist devices (LVADs) is a common problem. We implemented an institutional GI bleeding protocol for the management of these patients to determine if a systematic approach to this problem would result in improved outcomes. Methods: We conducted a retrospective cohort study by reviewing our institutional experience with GI bleeds in LVAD patients between October 2010 and December 2014. In January 2013, an institutional GI bleeding protocol was initiated to systematically approach the diagnosis and treatment of these events. We compared the frequency of GI bleeds, the rate of interventions such as endoscopy and colonoscopy, and the use of resources such as transfusion of packed red blood cells for patients before (BP) and after (AP) the initiation of the protocol. Standard statistical methods were utilized. Results: During the study period, 43 patients were readmitted with 90 GI bleeds. The BP group consisted of 16 patients with 38 bleeds, while the AP group consisted of 27 patients with 52 bleeds. There were no differences between the groups with regard to age, gender, device type, or history of GI bleeding prior to LVAD implantation. The BP group had 2.37 ± 1.41 bleeds/patient compared with 1.93 ± 1.49 bleeds/patient in the AP group (p = 0.336). After protocol initiation, the number of esophagogastroduodenoscopies (EGDs)/bleed increased (0.68 ± 0.66 vs. 1.10 ± 0.82, p = 0.013), while the number of colonoscopies, capsule studies, push EGDs, and units of blood transfused was not different. Tagged red blood studies/bleed decreased after the initiation of the protocol (0.32 ± 0.53 vs. 0.02 ± 0.14, p = 0.002), as did the number of angiograms/bleed (0.18 ± 0.46 vs. 0 ± 0, p = 0.017). Conclusion: The initiation of a GI bleeding protocol led to a shift towards less resource intensive treatment modalities. Further patient accrual may lead to additional reduction in resource utilization, such as transfusion of red blood cells and patient admissions. 6( 63) Severity of Hemolysis Is Associated with Death and Ischemic Stroke during Veno-Arterial Extracorporeal Membrane Support O. Saeed , W. Jakobleff Jr., M. Chau, S. Rangasamy, M. Algodi, M. Makkiya, M. Cruz, S. Patel, S. Murthy, D. Sims, J. Shin, D. Goldstein, U. Jorde. Medicine, Albert Einstein College of Medicine Montefiore Medical Center, Bronx, NY. Purpose: Hemolysis has emerged as a marker of adverse outcomes after CF LVAD placement, however, its impact during Veno-Arterial Extracorporeal Membrane (VA ECMO) support is uncertain. Methods: A single center retrospective analysis of all adult patients placed on VA ECMO from May 2011 to September 2015 was conducted. Major demographics, ECMO characteristics and clinical outcomes including cerebrovascular accidents (CVAs) were retrieved. As a marker of hemolysis severity, the peak lactate dehydrogenase (LDH) level was collected during VA ECMO support. To analyze the impact of hemolysis on in-hospital mortality patients were categorized as survivors and non survivors. Results: 111 patients underwent VA ECMO placement. The mean age was 57±14 years, 41 were female and 60 patients expired during VA ECMO. Both survivors and non survivors had similar distributions of major demographics and laboratory parameters except for baseline creatinine (survivors: 1.6±1.2 vs. non survivors: 2.4±1.7, p= 0.003). The median peak LDH for survivors was lower at 771 (IQR: 456-1616) U/L in comparison to 1444 (IQR: 8473080) U/L for non survivors (p= 0.005, figure). Patients who eventually had an ischemic CVA (n= 11) had a greater median peak LDH 1863 (IQR: 8004014) U/L in comparison to those with no CVA 961 (IQR: 509-1710) U/L, p= 0.039, figure. Conclusion: Severity of hemolysis measured by LDH is a marker of higher mortality during VA ECMO and may predict onset of thrombotic complications such as ischemic CVA. Further studies are warranted on usage of hemolysis severity to guide VA ECMO management, including circuit exchange.
6( 64) Speed Reduction Does Not Restore High Molecular Weight Von Willebrand Multimers during Heart Mate II Support: An In-Vivo Analysis S. Patel , J. Patel, J. Rand, D. Goldstein, U. Jorde, O. Saeed. Medicine, Albert Einstein College of Medicine Montefiore Medical Center, Bronx, NY. Purpose: Acquired Von Willebrand Factor (vWF) deficiency due to the loss of high molecular weight multimers (HMWM) has been well documented during CF LVAD support. It has been proposed that lowering pump speed in response to clinical gastrointestinal bleeding (GIB) may decrease shear stress allowing for the return of HMWMs. In-vivo data supporting this practice is lacking. Methods: Subjects at least 30 days post implantation of a Heart Mate (HM) II were prospectively recruited from the LVAD clinic. After confirming INR was > 2.0, pump speed was decreased to 8000 rpm and maintained for 6 hours. Blood samples obtained at baseline and 6 hours were compared for 2 measures of acquired vWF deficiency: 1) the ristocetin cofactor activity to vWF antigen ratio (Rco:Ag) and 2) gel electrophoresis for vWF multimer distribution. Results: Four patients agreed to participation. They were 57±15 years old, all were male and had been on HM II support for 401±199 days. All patients tolerated speed reduction without any adverse events. At baseline speed, HMWMs were reduced in all 4 patients. After 6 hours at 8000 rpm, there was no change in the HMWM profile (Figure 1). Similarly, the Rco:Ag ratio was reduced (nl > 0.65) in 3 of 4 patients at baseline and did not significantly change after speed reduction (0.56 → 0.56, 0.74 → 0.67, 0.58 → 0.65, 0.47 → 0.45; p = 0.437). Conclusion: Decreasing pump speed during HM II support does not lead to restoration of HMW Von Willebrand multimers. These findings suggest there may be no benefit to speed reduction in response to GIB.
Abstracts S245 6( 65) A Multicenter Evaluation of Octreotide for Ventricular Assist Device Related Gastrointestinal Bleeding G.B. Smallfield ,1 S. Gunda,1 S. Emani,2 M.K. Kanwar,3 N. Uriel,4 P.C. Colombo,5 P.A. Uber,1 M.L. Sears,1 K.B. Shah.1 1Virginia Commonwealth University, Richmond, VA; 2Ohio State University, Columbus, OH; 3Allegheny General Hospital, Pittsburgh, PA; 4University of Chicago, Chicago, IL; 5Columbia, New York, NY. Purpose: Gastrointestinal bleeding (GIB) is a frequent complication after implantation of a left ventricular assist device (LVAD). Octreotide is used to treat GIB as well as bleeding prophylaxis for vascular ectasias. The data to support octreotide use in LVAD associated bleeding is sparse. We therefore evaluated the efficacy of outpatient administration of octreotide to prevent reoccurrence of an LVAD related GIB. Methods: We present an interim analysis from a multicentered retrospective cohort including data from five centers and 34 patients who were discharged from the hospital with subcutaneous and depot (LAR) injectable octreotide. Patients were included if they received outpatient octreotide rebleeding prophylaxis on discharge after bleeding with LVAD. We measured forward from the bleeding event until their next bleeding event, transplant, explant or death. Results: The patients in the study had a mean (SD) age of 62.6 (9.6) years of age and were predominantly white (62%) and male (55%). At the time of receiving octreotide, patients were on device support 30.1(SD= 15) months and 63% of patients had experienced a prior GIB. Of the 34 patients analyzed, 5 were excluded for incomplete rebleeding data. Ten (34.5%) of patient re-bled at median 2.0 [IQR: 1.3, 5.8] months. There was no difference in rebleeding rate between those receiving LVAD for destination v. bridge to transplant (30% v. 37%). Though not statistically significant, there were several factors that appear associated with rebleeding after treatment with octreotide. Patients who had GIB before LVAD had a higher risk rebleeding after octreotide (75% v. 29%). Exposure to sildenafil showed a 50% rebleeding rate v. 32%. Compared to daily injections, the use of depot monthly LAR formulation showed a lower rate of bleeding 29% v. 42% for short acting formulations. Conclusion: This preliminary analysis from an ongoing multicenter cohort demonstrates a promising role for the long acting formulation octreotide for LVAD related GIB to reduce the risk of rebleeding. Prespecified multivariate regression and time-dependent outcomes modeling on the complete cohort of 75 patients will further elucidate the efficacy treatment at the time of data presentation. 6( 66) Eptifibatide in the Treatment of Pump Thrombosis: What Is the Prescription? L.M. Peters , D.T. Majure, M.E. Rodrigo, F.H. Sheikh, M. Hofmeyer, E.J. Molina, S.W. Boyce, S.S. Najjar. MedStar Washington Hospital Center, Washington, DC. Purpose: The optimal medical therapy for LVAD pump thrombosis (PT) is unknown. Past reports have discouraged the use of platelet GPIIb/IIIa receptor inhibitors due to concern for lack of efficacy and increased bleeding. Given the limited published data on the use of eptifibatide, we sought to describe our single center experience. Methods: We reviewed all cases of PT at our center from 1/2010 through 9/2015 who received eptifibatide. Our center utilizes eptifibatide in treatment of PT if unfractionated heparin (UFH) therapy is unsuccessful. Eptifibatide is typically started at 0.5 mcg/kg/min and titrated up to a maximum of 2 mcg/ kg/min depending on clinical response. Results: A total of 45 distinct PT events occurred for a cumulative incidence of 11.5% out of 278 implants. Of the 44 events treated with medical therapy, 20 (45%) were treated with eptifibatide. Events were treated with UFH alone for an average of 5±4 days prior to addition of eptifibatide. The mean dose of eptifibatide was 1.6±0.5 mcg/kg/min, and it was administered for a mean of 6±3 days (range 3-11). Eptifibatide led to clinical resolution (CR) in 11 (55%) events. Nine of 13 patients who received the maximum dose of 2 mcg/kg/min had CR. In contrast, 2 of 7 patients who received ≤ 1 mcg/kg/min had CR. Of the patients who failed eptifibatide, 5 patients required pump exchange, 3 required a heart transplant, and 1 died five months after failed treatment for
pump thrombosis as he was deemed not a candidate for pump exchange or heart transplant. Bleeding during eptifibatide occurred in 8 events: 3 patients developed gastrointestinal hemorrhage, 2 had epistaxis, 1 had intracranial hemorrhage with modified Rankin score of 1 at last follow-up, and 2 developed other sources of bleeding. Only one patient required blood transfusion (1 unit) due to menorrhagia. Five (62.5%) of the bleeding events occurred in patients who received the maximum dose of eptifibatide. One patient developed an embolic cerebrovascular accident. No patient developed significant thrombocytopenia during therapy. Conclusion: In our single center experience, addition of eptifibatide to UFH in medical management of PT had a moderate efficacy (clinical resolution > 50%) with low risk of bleeding requiring transfusion. Further studies are needed to define the optimal dosing strategy and timing, as well as patient characteristics that predict a successful outcome with this therapy. 6( 67) Abnormal Pre-Operative Mixing Studies Identify VAD Patients at High Risk for Spontaneous Bleeding J.D. Remick ,1 L. Holton,2 M. Mangum,1 K. Evenson,1 R. Lewis,1 M. Puhlman,1 C. Moore,3 G. Ott,1 J. Abraham.1 1Center for Advanced Heart Disease, Providence Heart & Vascular Institute, Portland, OR; 2Department of Medicine, Providence St. Vincent Medical Center, Portland, OR; 3Hematology and Oncology, Providence Cancer Center, Portland, OR. Purpose: Bleeding and thromboembolic (TE) complications are important causes of morbidity and mortality in VAD patients. Pre-operative screening for coagulopathies may influence patient selection, management, and outcomes. Methods: We performed a retrospective study of 33 patients with no prior history of coagulopathy who underwent HeartMate II VAD implant. Mixing studies were performed on patients with abnormal baseline PT or PTT. A bleeding event was defined as an episode resulting in one or more of the following: death, reoperation, hospitalization or transfusion (> 2 units) and further characterized as spontaneous or procedure-related. A TE was defined as either a neurologic event (stroke or TIA), pump malfunction requiring exchange, or LDH elevation requiring hospitalization. Bleeding and TE were compared between patients with normal vs abnormal mixing studies (AMS) using Poisson regression. Results: Of 33 patients, 20 had AMS. 15/20 had a total of 24 bleeding events, of which 80% were spontaneous. 5/13 patients with normal mixing studies had a total of 6 bleeding events, of which 66% were spontaneous. The AMS cohort had 0.99 bleeding events per patient year vs. 0.39 events per patient year in the normal group (risk ratio 2.54; p = 0.04). 10 thrombotic events occurred in 6/20 patients with AMS (0.41 events per patient-year) compared to 6 thrombotic events in 4/13 patients with normal mixing studies (0.39 events per patient year; p= 0.91). Conclusion: AMS resulting from factor deficiencies or circulating inhibitors are frequently detected in HF patients without prior history of coagulopathy. Their presence confers a 2.54-fold increase in risk of bleeding following VAD implantation. Routine pre-operative screening for coagulation disorders is warranted.