- Email: [email protected]
269 Aortic Valve Repair at the Time of Continuous-Flow Left Ventricular Assist Device Implantation
S94
The Journal of Heart and Lung Transplantation, Vol 30, No 4S, April 2011
Stage 1 (mild) WRF, 40 (2%) had Stage 2 (moderate) WRF, and 84 (4%) had Stage 3 (severe) WRF. All WRF stages were independently associated with post- heart transplantation, in-hospital mortality regardless of whether traditionally defined renal dysfunction was present or not; mild WRF (adjusted odds ratio [AOR] 2.1, 95% confidence intervals [CI] 1.2-3.5), moderate WRF (AOR 2.7, 95% CI 1.1-6.7), and severe WRF (AOR 3.6, 95% CI 2.0-6.5). Among hospital survivors, the association of WRF with subsequent survival did not reach statistical significance (hazard ratio 1.2, 95% confidence interval 0.9-1.7). Conclusions: Worsening renal function occurs frequently in children awaiting heart transplantation and is associated with early post-transplant mortality even when traditional definitions of renal dysfunction are not met. Worsening renal function does not appear to affect post-discharge survival. 267 The Use of Post Auricular Pedestal Is a Winning Strategy in Reducing Driveline Infections during Long-Term Mechanical Support with LVADs G. Gerosa,1 G. Di Giammarco,2 G. Sani,3 M. Maccherini,4 M. Rinaldi,5 M. De Bonis,6 F. Gazzoli,7 A. Renzulli,8 G. Mazzesi,9 G. Di Credico,10 M. Zogno,11 U. Livi.12 1Department of Cardiac, Thoracic and Vascular Sciences, Division of Cardiac Surgery, University of Padova, Padova, Italy; 2Dipartimento di Scienze Cliniche e Bioimmagini, Sezione di Cardiochirurgia, Cattedra di Cardiochirurgia Università G.D’Annunzio, Chieti, Italy; 3Medical and Surgical Critical Care Department, Cardiac Surgery, University of Florence, Florence, Italy; 4Department of Cardiovascular Diseases, Cardiac Surgery, University of Siena, Siena, Italy; 5Department of Cardiac Surgery, Cardiac Surgery, University of Turin, Turin, Italy; 6Department of Cardiac Surgery, Cardiac Surgery, San Raffaele University Hospital, Milan, Italy; 7Department of Cardiac Surgery, Cardiac Surgery, Fondazione IRCCS Policlinico San Matteo, University of Pavia, Pavia, Italy; 8Department of Cardiac Surgery, Cardiac Surgery Unit, Magna Graecia, University of Catanzaro, Catanzaro, Italy; 9Department of Thoracic Surgery, Cardiac Surgery, University of Rome La Sapienza, Policlinico Umberto I, Rome, Italy; 10 Department of Cardiac Surgery, Cardiac Surgery, Civil Hospital, Legnano, Italy; 11Department of Cardiac Surgery, Cardiac Surgery Unit, Azienda Ospedaliera C. Poma, Mantova, Italy; 12Cardiothoracic Department, Cardiac Surgery, University Hospital S Maria della Misericordia, Udine, Italy. Purpose: Driveline infections are a major source of morbidity during longterm mechanical support.Unlike other currently available LVADs Jarvik 2000 exhibits the unique feature of a power cable exit site at the post auricular area by means of a titanium pedestal firmly secured to the patient’s skull.In order to investigate whether this technical solution might reduce the incidence of driveline infections we reviewed the data of patients included in the Jarvik 2000 prospective, controlled Italian Registry. Methods and Materials: From June 2006 to September 2010 45 consecutive patients (40 males, mean age 62 ⫾ 8 yrs, mean BSA 1.9 ⫾ 0.2 m2) were enrolled in the Jarvik 2000 Italian Registry. 93% of the patients were uneligible to heart transplant at the time of VAD implant. The indication for VAD implantation was end-stage cardiomiopathy either ischemic (56%) or idiopathic (44%). Patients were prospectively followed up for all complications. Results: After implant 76% of the patients were discharged at home. Mean time on VAD was 307 days [range 57-1071], 60% of patients being on device ⬎6 months. As far as pedestal complications are concerned: 1 patient required surgical revision due to bleeding on 1st postoperative day. 2 patients presented with episodes of serum oozing around the pedestal with no signs of inflammation or infection which resolved by accurate wound dressing. 1 patient required repositioning of the pedestal 145 days post-implant due to osteoporosis:patient is still on device 408 days after pedestal repositioning without further complications. During the entire follow-up only 2 patients developed localized cable exit site infection successfully treated by surgical re-exploration and antibiotic therapy. Conclusions: The post auricular exit site of Jarvik 2000 LVAD power cable showed minimum incidence of complications and no VAD driveline infection. The post auricular pedestal appears to be an attractive solution granting significant reduction in driveline morbidity being cost-effective during long term mechanical support.
268 Predictors of Gastrointestinal Bleeding in Patients with Continuous Flow Left Ventricular Assist Devices S.R. Patel, R. Bello, S.J. Jooyoung, G. Stevens, P. Edwards, D. D’Alessandro, M. Simon, D. Goldstein. Center for Advanced Cardiac Therapy, Division of Cardiology, Albert Einstein College of Medicine, Montefiore Medical Center, Bronx, NY. Purpose: The high rate of gastrointestinal bleeding (GIB) with continuous flow (CF) LVADs has been discouraging. All CF LVAD recipients develop an acquired von Willebrand syndrome; however, only a fraction experience GIB. Therefore, we reviewed our experience to identify pre-operative predictors of GIB. Methods and Materials: We reviewed all 64 recipients of a CF LVAD between June 2006 and August 2010. GIB was defined as any episode with heme positive stool requiring transfusion of packed red blood cells (PRBC). As per institutional protocol, all pts were maintained on Asa 81mg and warfarin to achieve an INR of 1.5-2.5 unless contraindicated. Results: Median length of follow-up was 190 days. Follow-up was 100%. A total of 28 GIB episodes occurred in 20 CF LVAD recipients (31%). Median time from implantation to GIB was 23d (6-112 d). Freedom from GI bleeding is depicted in Figure 1.
At the time of initial bleed, the mean INR was 2.23 and the mean PRBC transfusion requirement was 5.15 units. Endoscopy records demonstrated the cause of GIB to be angiodysplasia in 32% (5 gastric, 4 small bowel), an unidentified source (presumed small bowel angiodysplasia) in 54%, and due to other causes in 14% of subjects. Older age (67 y vs 54 y, p ⫽ 0.003), pre-op bilirubin (1.9 vs 1.2, p ⫽ 0.020), and pre-op prealbumin (13.1 vs 17.9, p ⫽ 0.011) were identified as independent predictors of GIB. On multivariate analysis, age and pre-op total bilirubin remained as significant predictors of GIB. Conclusions: GIB occurs early and frequently after implantation in recipients of CF LVADs. The most common identified etiology of GIB is angiodysplasia. Older age and increased pre-op total bilirubin are significant predictors of GIB after CF LVAD implantation. 269 Aortic Valve Repair at the Time of Continuous-Flow Left Ventricular Assist Device Implantation S. Melnitchouk,1 H. Takayama,1 U. Jorde,2 N. Uriel,2 P. Colombo,2 J. Yang,1 D. Mancini,2 Y. Naka.1 1Department of Cardiothoracic Surgery, Columbia University Medical Center, New York, NY; 2 Department of Cardiology, Columbia University Medical Center, New York, NY. Purpose: Left Ventricular Assist Devices (LVADs) direct flow from LV apex to the aorta. Aortic insufficiency (AI) “short circuits” flow back to LV and must be corrected at the time of LVAD implant. Our indication for concomitant aortic valve repair (AVr) has evolved over time as recent studies have shown a high incidence of AI after implantation of continuous flow LVADs. Methods and Materials: AVr was performed for moderate or severe AI until end of 2009 and for any AI afterwards. Aortic valve was repaired by approximating the nodules of Arantius with a pledgetted suture. Retrospective chart review was performed for the period from 9/2007 to 8/2010.
Abstracts Results: During the study period, 19 patients (mean age 64 years, 17 males) underwent AVr at LVAD implantation (15.2% of 125 patients) - 8 for mild, 8 for moderate, and 3 for severe AI. Immediate postoperative echocardiography showed none to trace AI. 6 of 19 patients developed at least mild AI (3 mild, 2 moderate, and 1 severe) with a median time-to-failure of 85 days. Four of 6 patients had initially trace AI but eventually progressed to mild AI (3 patients) or moderate AI (1 patient). The other 2 patients had moderate or severe AI on the first postoperative TTE. One patient developed severe eccentric AI in the early postoperative period and had to undergo earlier heart transplantation after being upgraded to UNOS 1A.
Conclusions: This simple repair technique significantly reduced baseline AI in patients undergoing LVAD implantation. However, a third of patients developed at least mild AI after a short period of observation. Larger series with longer follow up are needed to assess the long-term durability of the aortic valve repair in these patients. 270 Simultaneous Aortic Valve Replacement in Left Ventricular Assist Device Recipients: Single-Center Experience N. Dranishnikov, A. Stepanenko, E.V. Potapov, M. Dandel, H. Siniawski, Y.-G. Weng, T. Krabatsch, R. Hetzer. Department of Cardiothoracic and Vascular Surgery, Deutsches Herzzentrum Berlin, Berlin, Germany. Purpose: Aortic valve regurgitation or presence of mechanical aortic valve is a contraindication for isolated left ventricular assist device (LVAD) support. However, LVAD implantation with concomitant aortic valve replacement (AVR) by a biological prosthesis may be performed in this situation. We report our recent experience. Methods and Materials: Between 1.01.2008 and 1.06.2010 191 patients were supported with a long-term LVAD in our institution. In 9, simultaneous aortic valve procedure (3 re-replacement with biovalve, 6 primary replacement) was performed. Patients were divided into 2 groups according to INTERMACS level: Group 1 (n ⫽ 4) consisted of patients with level 1-2, Group 2 (n ⫽ 5) with level 3⫹ (no prior aortic valve procedures). Preoperative and intraoperative data were analyzed. Results: All patients were men and median age was 58 years (range 32-67 years). In group 1 three out of four patients had status post AVR. Three patients from Group 1 died on 22nd, 45th and 193rd day of support due to multiorgan failure because of sepsis. In 3 of 4 patients from Group 1 postoperative RVAD support was necessary, with successful RVAD weaning in 2 patients after 10 and 30 days of support. In Group 2 four out of five patients survived to discharge. Conclusions: Simultaneous aortic valve replacement is not associated with an increased surgical risk when semielective operation is performed before onset of cardiogenic shock. In patients with cardiogenic shock, however, simultaneous aortic valve procedure may be related to negative clinical outcome. In this situation other procedure (e.g. patch closure or remaining of mechanical valve in situ) may be an valuable option.
S95 271 Development of Aortic Insufficiency (AI) in Patients Supported with Continuous Flow Left Ventricular Assist Devices (LVAD) B. Soleimani, A. Manoskey, A. Haouzi, E. Stephenson, A. Nunez, C. Brehm, A. El-Banayosy, W. Pae. Heart & Vascular Institute, Penn State Hershey Medical Center, Hershey, PA. Purpose: LVADs are increasingly used in long-term therapy of patients with advanced heart failure. Development of AI can adversely affect pump performance. In this study, we examined the prevalence of AI following LVAD implant at our institution. Methods and Materials: Echocardiograms (n ⫽ 195) of 66 patients who received HeartMate II (HMII, n ⫽ 58) or Heartware (HW, n ⫽ 8) LVAD at our institution since June 2008 were reviewed. Studies were obtained pre-operatively, and post-operatively at 1 month and then at 3 monthly intervals. AI was regarded significant if mild-to-moderate or greater. Patients with concomitant aortic valve replacement (2) or significant baseline AI (1) were excluded from analysis. Results: Median LVAD support duration was 221 days (265 days for HMII and 134 days for HW). New AI developed in 6 of 63 patients (9.5%). There were no cases of severe or symptomatic AI. The median time to AI development was 374 days. There was no significant difference between the AI prevalance in HMII and HW recipients (6/65 (10.9%) vs. 0/8 (0%), z value 0.298). For patients who remained on LVAD support at 6 and 12 months, freedom from AI was 100% and 68.4% respectively (numbers at risk of 38 and 19 respectively). There was no difference between aortic root diameters at baseline between those who developed AI and those who did not (3.5 ⫾ 0.41 vs. 3.38 ⫾ 0.41, p ⫽ 0.53). AI occured with similar frequency in those LVAD supported patients in whom the aortic valve opened versus those that did not (2/26 vs. 4/30, z value 0.247). Conclusions: AI develops frequently during long-term support with continuous flow LVADs, particularly in those supported for longer than 6 months. However, we found no cases of severe or symptomatic AI in our cohort. As we move to the era of long-term LVAD support and destination therapy, further studies with longer follow ups are required to determine the progression and clinical significance of AI in these patients. 272 The Need for Device Exchange in the Era of Continuous Flow LVADs S. Aggarwal,1 M. Borkon,1 Y. Li,2 S.L. Myers,3 K.B. Allen,1 J. Chichetti,5 R. John,4 F.D. Pagani.6 1Cardithoracic Surgery, Mid America Heart and Vascular Institute, Saint Luke’s Hospital, Kansas City, MO; 2Biostatistics, Mid America Heart and Vascular Institute, Saint Luke’s Hospital, Kansas City, MO; 3Surgery, University of Alabama at Birmingham, Birmingham, AL; 4Surgery, University of Minnesota, Minneapolis, MN; 5Surgery, Robert Wood Johnson University Hospital, New Brunswick, NJ; 6Surgery, University of Michigan Medical Center, Ann Arbor, MI. Purpose: The use of continuous flow(CF) ventricular assist devices has led to improved clinical outcomes when compared to pulsatile flow (PF) pumps. We hypothesize that the use of CF devices has reduced the need for device exchange and has altered the indications. Here we describe the incidence, indications, and outcomes following device exchange in a large cohort of LVAD patients from the Interagency Registry for Mechanically Assisted Circulatory Support (INTERMACS). Methods and Materials: The study cohort included prospectively entered patients from the INTERMACS registry undergoing primary LVAD implantation from June 2006 to March 2010. Patients receiving BiVAD support or RVADs were excluded. Kaplan Meier analysis was performed to examine freedom from device exchange and for conditional survival following device exchange. A total of 1930 patients were analyzed (1451 CF and 479 PF). Results: 85 patients underwent device exchange (30 CF, 55 PF). Patients with CF devices demonstrated significantly increased freedom from device exchange when compared to PF patients (Figure 1A, p ⬍ 0.0001). Indications for device exchange in PF patients included device failure (91%) and infection (6%). In CF flow patients, indications for exchange included device failure (73%), infection (16.7%), hemolysis (3.3%), and pump thrombosis (3.3%). No significant difference in
The Journal of Heart and Lung Transplantation, Vol 30, No 4S, April 2011
Stage 1 (mild) WRF, 40 (2%) had Stage 2 (moderate) WRF, and 84 (4%) had Stage 3 (severe) WRF. All WRF stages were independently associated with post- heart transplantation, in-hospital mortality regardless of whether traditionally defined renal dysfunction was present or not; mild WRF (adjusted odds ratio [AOR] 2.1, 95% confidence intervals [CI] 1.2-3.5), moderate WRF (AOR 2.7, 95% CI 1.1-6.7), and severe WRF (AOR 3.6, 95% CI 2.0-6.5). Among hospital survivors, the association of WRF with subsequent survival did not reach statistical significance (hazard ratio 1.2, 95% confidence interval 0.9-1.7). Conclusions: Worsening renal function occurs frequently in children awaiting heart transplantation and is associated with early post-transplant mortality even when traditional definitions of renal dysfunction are not met. Worsening renal function does not appear to affect post-discharge survival. 267 The Use of Post Auricular Pedestal Is a Winning Strategy in Reducing Driveline Infections during Long-Term Mechanical Support with LVADs G. Gerosa,1 G. Di Giammarco,2 G. Sani,3 M. Maccherini,4 M. Rinaldi,5 M. De Bonis,6 F. Gazzoli,7 A. Renzulli,8 G. Mazzesi,9 G. Di Credico,10 M. Zogno,11 U. Livi.12 1Department of Cardiac, Thoracic and Vascular Sciences, Division of Cardiac Surgery, University of Padova, Padova, Italy; 2Dipartimento di Scienze Cliniche e Bioimmagini, Sezione di Cardiochirurgia, Cattedra di Cardiochirurgia Università G.D’Annunzio, Chieti, Italy; 3Medical and Surgical Critical Care Department, Cardiac Surgery, University of Florence, Florence, Italy; 4Department of Cardiovascular Diseases, Cardiac Surgery, University of Siena, Siena, Italy; 5Department of Cardiac Surgery, Cardiac Surgery, University of Turin, Turin, Italy; 6Department of Cardiac Surgery, Cardiac Surgery, San Raffaele University Hospital, Milan, Italy; 7Department of Cardiac Surgery, Cardiac Surgery, Fondazione IRCCS Policlinico San Matteo, University of Pavia, Pavia, Italy; 8Department of Cardiac Surgery, Cardiac Surgery Unit, Magna Graecia, University of Catanzaro, Catanzaro, Italy; 9Department of Thoracic Surgery, Cardiac Surgery, University of Rome La Sapienza, Policlinico Umberto I, Rome, Italy; 10 Department of Cardiac Surgery, Cardiac Surgery, Civil Hospital, Legnano, Italy; 11Department of Cardiac Surgery, Cardiac Surgery Unit, Azienda Ospedaliera C. Poma, Mantova, Italy; 12Cardiothoracic Department, Cardiac Surgery, University Hospital S Maria della Misericordia, Udine, Italy. Purpose: Driveline infections are a major source of morbidity during longterm mechanical support.Unlike other currently available LVADs Jarvik 2000 exhibits the unique feature of a power cable exit site at the post auricular area by means of a titanium pedestal firmly secured to the patient’s skull.In order to investigate whether this technical solution might reduce the incidence of driveline infections we reviewed the data of patients included in the Jarvik 2000 prospective, controlled Italian Registry. Methods and Materials: From June 2006 to September 2010 45 consecutive patients (40 males, mean age 62 ⫾ 8 yrs, mean BSA 1.9 ⫾ 0.2 m2) were enrolled in the Jarvik 2000 Italian Registry. 93% of the patients were uneligible to heart transplant at the time of VAD implant. The indication for VAD implantation was end-stage cardiomiopathy either ischemic (56%) or idiopathic (44%). Patients were prospectively followed up for all complications. Results: After implant 76% of the patients were discharged at home. Mean time on VAD was 307 days [range 57-1071], 60% of patients being on device ⬎6 months. As far as pedestal complications are concerned: 1 patient required surgical revision due to bleeding on 1st postoperative day. 2 patients presented with episodes of serum oozing around the pedestal with no signs of inflammation or infection which resolved by accurate wound dressing. 1 patient required repositioning of the pedestal 145 days post-implant due to osteoporosis:patient is still on device 408 days after pedestal repositioning without further complications. During the entire follow-up only 2 patients developed localized cable exit site infection successfully treated by surgical re-exploration and antibiotic therapy. Conclusions: The post auricular exit site of Jarvik 2000 LVAD power cable showed minimum incidence of complications and no VAD driveline infection. The post auricular pedestal appears to be an attractive solution granting significant reduction in driveline morbidity being cost-effective during long term mechanical support.
268 Predictors of Gastrointestinal Bleeding in Patients with Continuous Flow Left Ventricular Assist Devices S.R. Patel, R. Bello, S.J. Jooyoung, G. Stevens, P. Edwards, D. D’Alessandro, M. Simon, D. Goldstein. Center for Advanced Cardiac Therapy, Division of Cardiology, Albert Einstein College of Medicine, Montefiore Medical Center, Bronx, NY. Purpose: The high rate of gastrointestinal bleeding (GIB) with continuous flow (CF) LVADs has been discouraging. All CF LVAD recipients develop an acquired von Willebrand syndrome; however, only a fraction experience GIB. Therefore, we reviewed our experience to identify pre-operative predictors of GIB. Methods and Materials: We reviewed all 64 recipients of a CF LVAD between June 2006 and August 2010. GIB was defined as any episode with heme positive stool requiring transfusion of packed red blood cells (PRBC). As per institutional protocol, all pts were maintained on Asa 81mg and warfarin to achieve an INR of 1.5-2.5 unless contraindicated. Results: Median length of follow-up was 190 days. Follow-up was 100%. A total of 28 GIB episodes occurred in 20 CF LVAD recipients (31%). Median time from implantation to GIB was 23d (6-112 d). Freedom from GI bleeding is depicted in Figure 1.
At the time of initial bleed, the mean INR was 2.23 and the mean PRBC transfusion requirement was 5.15 units. Endoscopy records demonstrated the cause of GIB to be angiodysplasia in 32% (5 gastric, 4 small bowel), an unidentified source (presumed small bowel angiodysplasia) in 54%, and due to other causes in 14% of subjects. Older age (67 y vs 54 y, p ⫽ 0.003), pre-op bilirubin (1.9 vs 1.2, p ⫽ 0.020), and pre-op prealbumin (13.1 vs 17.9, p ⫽ 0.011) were identified as independent predictors of GIB. On multivariate analysis, age and pre-op total bilirubin remained as significant predictors of GIB. Conclusions: GIB occurs early and frequently after implantation in recipients of CF LVADs. The most common identified etiology of GIB is angiodysplasia. Older age and increased pre-op total bilirubin are significant predictors of GIB after CF LVAD implantation. 269 Aortic Valve Repair at the Time of Continuous-Flow Left Ventricular Assist Device Implantation S. Melnitchouk,1 H. Takayama,1 U. Jorde,2 N. Uriel,2 P. Colombo,2 J. Yang,1 D. Mancini,2 Y. Naka.1 1Department of Cardiothoracic Surgery, Columbia University Medical Center, New York, NY; 2 Department of Cardiology, Columbia University Medical Center, New York, NY. Purpose: Left Ventricular Assist Devices (LVADs) direct flow from LV apex to the aorta. Aortic insufficiency (AI) “short circuits” flow back to LV and must be corrected at the time of LVAD implant. Our indication for concomitant aortic valve repair (AVr) has evolved over time as recent studies have shown a high incidence of AI after implantation of continuous flow LVADs. Methods and Materials: AVr was performed for moderate or severe AI until end of 2009 and for any AI afterwards. Aortic valve was repaired by approximating the nodules of Arantius with a pledgetted suture. Retrospective chart review was performed for the period from 9/2007 to 8/2010.
Abstracts Results: During the study period, 19 patients (mean age 64 years, 17 males) underwent AVr at LVAD implantation (15.2% of 125 patients) - 8 for mild, 8 for moderate, and 3 for severe AI. Immediate postoperative echocardiography showed none to trace AI. 6 of 19 patients developed at least mild AI (3 mild, 2 moderate, and 1 severe) with a median time-to-failure of 85 days. Four of 6 patients had initially trace AI but eventually progressed to mild AI (3 patients) or moderate AI (1 patient). The other 2 patients had moderate or severe AI on the first postoperative TTE. One patient developed severe eccentric AI in the early postoperative period and had to undergo earlier heart transplantation after being upgraded to UNOS 1A.
Conclusions: This simple repair technique significantly reduced baseline AI in patients undergoing LVAD implantation. However, a third of patients developed at least mild AI after a short period of observation. Larger series with longer follow up are needed to assess the long-term durability of the aortic valve repair in these patients. 270 Simultaneous Aortic Valve Replacement in Left Ventricular Assist Device Recipients: Single-Center Experience N. Dranishnikov, A. Stepanenko, E.V. Potapov, M. Dandel, H. Siniawski, Y.-G. Weng, T. Krabatsch, R. Hetzer. Department of Cardiothoracic and Vascular Surgery, Deutsches Herzzentrum Berlin, Berlin, Germany. Purpose: Aortic valve regurgitation or presence of mechanical aortic valve is a contraindication for isolated left ventricular assist device (LVAD) support. However, LVAD implantation with concomitant aortic valve replacement (AVR) by a biological prosthesis may be performed in this situation. We report our recent experience. Methods and Materials: Between 1.01.2008 and 1.06.2010 191 patients were supported with a long-term LVAD in our institution. In 9, simultaneous aortic valve procedure (3 re-replacement with biovalve, 6 primary replacement) was performed. Patients were divided into 2 groups according to INTERMACS level: Group 1 (n ⫽ 4) consisted of patients with level 1-2, Group 2 (n ⫽ 5) with level 3⫹ (no prior aortic valve procedures). Preoperative and intraoperative data were analyzed. Results: All patients were men and median age was 58 years (range 32-67 years). In group 1 three out of four patients had status post AVR. Three patients from Group 1 died on 22nd, 45th and 193rd day of support due to multiorgan failure because of sepsis. In 3 of 4 patients from Group 1 postoperative RVAD support was necessary, with successful RVAD weaning in 2 patients after 10 and 30 days of support. In Group 2 four out of five patients survived to discharge. Conclusions: Simultaneous aortic valve replacement is not associated with an increased surgical risk when semielective operation is performed before onset of cardiogenic shock. In patients with cardiogenic shock, however, simultaneous aortic valve procedure may be related to negative clinical outcome. In this situation other procedure (e.g. patch closure or remaining of mechanical valve in situ) may be an valuable option.
S95 271 Development of Aortic Insufficiency (AI) in Patients Supported with Continuous Flow Left Ventricular Assist Devices (LVAD) B. Soleimani, A. Manoskey, A. Haouzi, E. Stephenson, A. Nunez, C. Brehm, A. El-Banayosy, W. Pae. Heart & Vascular Institute, Penn State Hershey Medical Center, Hershey, PA. Purpose: LVADs are increasingly used in long-term therapy of patients with advanced heart failure. Development of AI can adversely affect pump performance. In this study, we examined the prevalence of AI following LVAD implant at our institution. Methods and Materials: Echocardiograms (n ⫽ 195) of 66 patients who received HeartMate II (HMII, n ⫽ 58) or Heartware (HW, n ⫽ 8) LVAD at our institution since June 2008 were reviewed. Studies were obtained pre-operatively, and post-operatively at 1 month and then at 3 monthly intervals. AI was regarded significant if mild-to-moderate or greater. Patients with concomitant aortic valve replacement (2) or significant baseline AI (1) were excluded from analysis. Results: Median LVAD support duration was 221 days (265 days for HMII and 134 days for HW). New AI developed in 6 of 63 patients (9.5%). There were no cases of severe or symptomatic AI. The median time to AI development was 374 days. There was no significant difference between the AI prevalance in HMII and HW recipients (6/65 (10.9%) vs. 0/8 (0%), z value 0.298). For patients who remained on LVAD support at 6 and 12 months, freedom from AI was 100% and 68.4% respectively (numbers at risk of 38 and 19 respectively). There was no difference between aortic root diameters at baseline between those who developed AI and those who did not (3.5 ⫾ 0.41 vs. 3.38 ⫾ 0.41, p ⫽ 0.53). AI occured with similar frequency in those LVAD supported patients in whom the aortic valve opened versus those that did not (2/26 vs. 4/30, z value 0.247). Conclusions: AI develops frequently during long-term support with continuous flow LVADs, particularly in those supported for longer than 6 months. However, we found no cases of severe or symptomatic AI in our cohort. As we move to the era of long-term LVAD support and destination therapy, further studies with longer follow ups are required to determine the progression and clinical significance of AI in these patients. 272 The Need for Device Exchange in the Era of Continuous Flow LVADs S. Aggarwal,1 M. Borkon,1 Y. Li,2 S.L. Myers,3 K.B. Allen,1 J. Chichetti,5 R. John,4 F.D. Pagani.6 1Cardithoracic Surgery, Mid America Heart and Vascular Institute, Saint Luke’s Hospital, Kansas City, MO; 2Biostatistics, Mid America Heart and Vascular Institute, Saint Luke’s Hospital, Kansas City, MO; 3Surgery, University of Alabama at Birmingham, Birmingham, AL; 4Surgery, University of Minnesota, Minneapolis, MN; 5Surgery, Robert Wood Johnson University Hospital, New Brunswick, NJ; 6Surgery, University of Michigan Medical Center, Ann Arbor, MI. Purpose: The use of continuous flow(CF) ventricular assist devices has led to improved clinical outcomes when compared to pulsatile flow (PF) pumps. We hypothesize that the use of CF devices has reduced the need for device exchange and has altered the indications. Here we describe the incidence, indications, and outcomes following device exchange in a large cohort of LVAD patients from the Interagency Registry for Mechanically Assisted Circulatory Support (INTERMACS). Methods and Materials: The study cohort included prospectively entered patients from the INTERMACS registry undergoing primary LVAD implantation from June 2006 to March 2010. Patients receiving BiVAD support or RVADs were excluded. Kaplan Meier analysis was performed to examine freedom from device exchange and for conditional survival following device exchange. A total of 1930 patients were analyzed (1451 CF and 479 PF). Results: 85 patients underwent device exchange (30 CF, 55 PF). Patients with CF devices demonstrated significantly increased freedom from device exchange when compared to PF patients (Figure 1A, p ⬍ 0.0001). Indications for device exchange in PF patients included device failure (91%) and infection (6%). In CF flow patients, indications for exchange included device failure (73%), infection (16.7%), hemolysis (3.3%), and pump thrombosis (3.3%). No significant difference in