Predicting Major Outcomes after MCSD Implant Risk Factors for Death, Transplant, and Recovery James Kirklin, MD David Naftel, PhD 1
I have no financial disclosures (I am the Principle Investigator for the NIH contract for INTERMACS) James K Kirklin
Predicting Major Outcomes after MCSD Implant Risk Factors for Death, Transplant, and Recovery What do we have so far? 3
Continuous Flow LVAD/BiVAD Implants: 2008 2013, n=9359 Event: Death (censored at transplantation or recovery) n=9359, Deaths=2280 % Survival Months % Survival 1 95% 12 80% 24 69% 36 59% 48 47% Hazard (early + late) Deaths/months n at risk: 9359 6195 4024 2736 1820 1194 711 408 174 Months post implant 4
Continuous Flow LVAD/BiVAD Implants: 2008 2013, n=9359 Event: Death (censored at transplantation or recovery) n=9359, Deaths=2280 % Survival Months % Survival 1 95% 12 80% 24 69% 36 59% 48 47% Hazard (early + late) Deaths/months n at risk: 9359 6195 4024 2736 1820 1194 711 408 174 Months post implant 5
Figure 10 Implants: June 2006 December 2013, n=10542 LVAD: Continuous n=9112, Deaths=2154 TAH: Pulsatile n=239, Deaths=59 % Survival at Pump 1 yr post implant CF LVAD 81% CF Bi-VAD 57% PF LVAD 65% PF Bi-VAD 45% TAH 59% % Survival BiVAD: Pulsatile n=319, Deaths=123 BiVAD: Continuous n=260, Deaths=118 LVAD: Pulsatile n=612, Deaths=230 overall p <.0001 Event: Death (censored at transplant and recovery) Months post implant 6
Implants: June 2006 December 2013, n=10542 LVAD: Continuous n=9112, Deaths=2154 TAH: Pulsatile n=239, Deaths=59 % Survival at Pump 1 yr post implant CF LVAD 81% CF Bi-VAD 57% PF LVAD 65% PF Bi-VAD 45% TAH 59% % Survival BiVAD: Pulsatile n=319, Deaths=123 BiVAD: Continuous n=260, Deaths=118 LVAD: Pulsatile n=612, Deaths=230 overall p <.0001 Event: Death (censored at transplant and recovery) Months post implant 7
Adult Primary Continuous Flow LVADs and BiVADs Implants: June 2006 December 2013, n=9372 Risk Factors for Death Demographics Early hazard Late hazard Hazard Ratio p-value Hazard Ratio p-value Age (older) 1.36 <.0001 Female 1.20.007 BMI (higher) 1.13 <.0001 Clinical Status History of Stroke 1.30.03 INTERMACS Level 1 1.69 <.0001 INTERMACS Level 2 1.44 <.0001 Destination Therapy 1.24.0005 Non-Cardiac Systems Albumin (lower) 0.90.02 Creatinine (higher) 1.05.0003 Dialysis 2.37 <.0001 BUN (higher) 1.06 <.0001 1.06.01 Right Heart Dysfunction Right Atrial Pressure (higher) 1.11.02 RVAD in same operation 2.45 <.0001 Bilirubin (higher) 1.21 <.0001 Ascites 1.27.01 Surgical Complexities History of cardiac surgery 1.43 <.0001 Concommitant Cardiac Surgery 1.21.0008 8
Adult Primary Continuous Flow LVADs and BiVADs Implants: June 2006 December 2013, n=9372 Early hazard Late hazard Risk Factors for Death Hazard Ratio p-value Hazard Ratio p-value Demographics Age (older) 1.36 <.0001 Female 1.20.007 BMI (higher) 1.13 <.0001 Clinical Status History of Stroke 1.30.03 INTERMACS Level 1 1.69 <.0001 INTERMACS Level 2 1.44 <.0001 Destination Therapy 1.24.0005 Non-Cardiac Systems Albumin (lower) 0.90.02 Creatinine (higher) 1.05.0003 Dialysis 2.37 <.0001 BUN (higher) 1.06 <.0001 1.06.01 Right Heart Dysfunction Right Atrial Pressure (higher) 1.11.02 RVAD in same operation 2.45 <.0001 Bilirubin (higher) 1.21 <.0001 Ascites 1.27.01 Surgical Complexities History of cardiac surgery 1.43 <.0001 Concommitant Cardiac Surgery 1.21.0008 9
Continuous Flow LVAD/BiVAD Implants: 2008 2013, n=9372 Age < 50 yrs n=2403, deaths=446 % Survival Age > 70 yrs, n=1301 deaths=451 Overall P <.0001 Age 50-64 yrs n=4184 Deaths= 942 Age 65-70 yrs n=1484, deaths=433 Event: Death (censored at transplant and recovery) Months post implant 10
June 2006 December 2012 AGE GROUP (yr) 0-18 19-39 40-59 60-79 80+ Unspecified TOTAL IMPLANT DATE PERIOD < 2010 2010-2011 2012 n % n % n % 40 1.7 % 30 0.8 % 8 0.3 % 354 15.8 % 422 11.7 % 223 10.6 % 1125 50.5 % 1469 40.8 % 806 38.5 % Age Category 708 31.7 % 1651 45.9 % 1037 49.5 %.. 22 0.6 % 18 0.8 %.... 1 0.0 % 2227 100.0 % 3594 100.0 % 2093 100.0 % 11
Adult Primary Continuous Flow LVADs and BiVADs Implants: June 2006 December 2013, n=9372 Early hazard Late hazard Risk Factors for Death Hazard Ratio p-value Hazard Ratio p-value Demographics Age (older) 1.36 <.0001 Female 1.20.007 BMI (higher) 1.13 <.0001 Clinical Status History of Stroke 1.30.03 INTERMACS Level 1 1.69 <.0001 INTERMACS Level 2 1.44 <.0001 Destination Therapy 1.24.0005 Non-Cardiac Systems Albumin (lower) 0.90.02 Creatinine (higher) 1.05.0003 Dialysis 2.37 <.0001 BUN (higher) 1.06 <.0001 1.06.01 Right Heart Dysfunction Right Atrial Pressure (higher) 1.11.02 RVAD in same operation 2.45 <.0001 Bilirubin (higher) 1.21 <.0001 Ascites 1.27.01 Surgical Complexities History of cardiac surgery 1.43 <.0001 Concommitant Cardiac Surgery 1.21.0008 12
Continuous Flow LVAD/BiVAD Implants: 2008 2013, n=9372 Levels 4-7, n=1789 Deaths=405 Level 1: n=1391 Deaths=381 % Survival Level 3: n=2591 Deaths=544 Level 2: n=3601 Deaths=942 P <.0001 Event: Death (censored at transplant and recovery) Months post implant 13
Patient Profile at time of implant Continuous Flow Devices CF-LVAD/BiVAD Implants: January 2008 December 2013, n=9372 Implant Date Era 2008-2010 2011-2013 TOTAL N % N % N % 1 Critical Cardiogenic Shock 464 16.0% 927 14.3% 1391 14.8% 2 Progressive Decline 1250 43.0% 2351 36.4% 3601 38.4% 3 Stable but Inotrope Dependent 659 22.7% 1932 29.9% 2591 27.7% 4 Resting Symptoms 370 12.7% 941 14.6% 1311 14.0% 5 Exertion Intolerant 84 2.9% 192 3.0% 276 2.9% 6 Exertion Limited 49 1.7% 79 1.4% 129 1.4% 7 Advanced NYHA Class 3 30 1.0% 43 1.0% 73 1.0% TOTAL 2906 100.0% 6465 100.0% 9371 100.0% 14
Patient Profile at time of implant Continuous Flow Devices CF-LVAD/BiVAD Implants: January 2008 December 2013, n=9372 Implant Date Era 2008-2010 2011-2013 TOTAL N % N % N % 1 Critical Cardiogenic Shock 464 16.0% 927 14.3% 1391 14.8% 2 Progressive Decline 1250 43.0% 2351 36.4% 3601 38.4% 3 Stable but Inotrope Dependent 659 22.7% 1932 29.9% 2591 27.7% 4 Resting Symptoms 370 12.7% 941 14.6% 1311 14.0% 5 Exertion Intolerant 84 2.9% 192 3.0% 276 2.9% 6 Exertion Limited 49 1.7% 79 1.4% 129 1.4% 7 Advanced NYHA Class 3 30 1.0% 43 1.0% 73 1.0% TOTAL 2906 100.0% 6465 100.0% 9371 100.0% 15
Table 6 Adult Primary Continuous Flow LVADs and BiVADs Implants: June 2006 December 2013, n=9372 Early hazard Late hazard Risk Factors for Death Hazard Ratio p-value Hazard Ratio p-value Demographics Age (older) 1.36 <.0001 Female 1.20.007 BMI (higher) 1.13 <.0001 Clinical Status History of Stroke 1.30.03 INTERMACS Level 1 1.69 <.0001 INTERMACS Level 2 1.44 <.0001 Destination Therapy 1.24.0005 Non-Cardiac Systems Albumin (lower) 0.90.02 Creatinine (higher) 1.05.0003 Dialysis 2.37 <.0001 BUN (higher) 1.06 <.0001 1.06.01 Right Heart Dysfunction Right Atrial Pressure (higher) 1.11.02 RVAD in same operation 2.45 <.0001 Bilirubin (higher) 1.21 <.0001 Ascites 1.27.01 Surgical Complexities History of cardiac surgery 1.43 <.0001 Concommitant Cardiac Surgery 1.21.0008 LVAD, left ventricular assist device; BiVAD, biventricular assist device; DT, destination therapy; BTT, bridge to transplant; BMI, body mass index; BUN, blood urea nitrogen; RVAD, right ventricular assist device. 16
Continuous Flow LVAD/BiVAD Implants: 2008 2013, n=9372 BTR, n= 46 Deaths= 4 BTT (including BTC), n= 5886 Deaths= 1218 % Survival Rescue Therapy, n= 27 Deaths= 11 DT, n= 3373 Deaths= 1027 p <.0001 Other, n= 40 Deaths= 12 Event: Death (censored at transplant and recovery) Months post implant 17
Continuous Flow Devices CF-LVAD/BiVAD Implants: June 2006 December 2013, n=9372 Device Strategy at time of implant Implant Date Era 2008-2010 2011-2013 TOTAL N % N % N % BTT Listed 1133 39.0% 1342 26.4% 2475 26.4% BTT Likely 765 26.3% 1387 21.5% 2152 23.0% BTT Moderate 296 10.2% 663 10.3% 959 10.2% BTT Unlikely 82 2.8% 218 ¾% 300 3.2% Destination Therapy 591 20.3% 2781 43.0% 3373 36.0% BTR 15 1.0% 31 1.0% 46 1.0% Rescue Therapy 10 0.3% 17 0.3% 27 0.3% Other 14 0.5% 26 0.4% 40 0.4% TOTAL 2906 100.0% 6465 100.0% 9371 100.0% 18
Table 6 Adult Primary Continuous Flow LVADs and BiVADs Implants: June 2006 December 2013, n=9372 Early hazard Late hazard Risk Factors for Death Hazard Ratio p-value Hazard Ratio p-value Demographics Age (older) 1.36 <.0001 Female 1.20.007 BMI (higher) 1.13 <.0001 Clinical Status History of Stroke 1.30.03 INTERMACS Level 1 1.69 <.0001 INTERMACS Level 2 1.44 <.0001 Destination Therapy 1.24.0005 Non-Cardiac Systems Albumin (lower) 0.90.02 Creatinine (higher) 1.05.0003 Dialysis 2.37 <.0001 BUN (higher) 1.06 <.0001 1.06.01 Right Heart Dysfunction Right Atrial Pressure (higher) 1.11.02 RVAD in same operation 2.45 <.0001 Bilirubin (higher) 1.21 <.0001 Ascites 1.27.01 Surgical Complexities History of cardiac surgery 1.43 <.0001 Concommitant Cardiac Surgery 1.21.0008 LVAD, left ventricular assist device; BiVAD, biventricular assist device; DT, destination therapy; BTT, bridge to transplant; BMI, body 19 mass index; BUN, blood urea nitrogen; RVAD, right ventricular assist device.
Figure 14 Continuous Flow LVAD/BiVAD Implants: 2008 2013, n=9372 Survival by Renal Risk Factors Low n=8237, deaths=1862 % Survival Severe (dialysis) n=142, deaths=71 Moderate (creat > 2 or BUN > 60) n=993, deaths=339 Overall P <.0001 Event: Death (censored at transplant and recovery) Months post implant 20
Figure 14 Continuous Flow LVAD/BiVAD Implants: 2008 2013, n=9372 Survival by Renal Risk Factors Low n=8237, deaths=1862 % Survival Severe (dialysis) n=142, deaths=71 Moderate (creat > 2 or BUN > 60) n=993, deaths=339 Overall P <.0001 Event: Death (censored at transplant and recovery) Months post implant 21
Adult Primary Continuous Flow LVADs & BIVADs, DT and BTT, n=4917 Implants: June 2006 March 2012: Creatinine Time course of Creatinine according to pre-implant Renal Dysfunction Creatinine (mg/dl) Severe (n=282) Moderate (n=1475) Mild/None (n=3160) ** ** ** ** ** ** ** ** ** ** ** ** ** ** * ** ** ** ** * p <.05 ** p <.001 Paired comparisons to pre-implant Follow-up Time Period 22
Adult Primary Continuous Flow LVADs and BiVADs Implants: June 2006 December 2013, n=9372 Early hazard Late hazard Risk Factors for Death Hazard Ratio p-value Hazard Ratio p-value Demographics Age (older) 1.36 <.0001 Female 1.20.007 BMI (higher) 1.13 <.0001 Clinical Status History of Stroke 1.30.03 INTERMACS Level 1 1.69 <.0001 INTERMACS Level 2 1.44 <.0001 Destination Therapy 1.24.0005 Non-Cardiac Systems Albumin (lower) 0.90.02 Creatinine (higher) 1.05.0003 Dialysis 2.37 <.0001 BUN (higher) 1.06 <.0001 1.06.01 Right Heart Dysfunction Right Atrial Pressure (higher) 1.11.02 RVAD in same operation 2.45 <.0001 Bilirubin (higher) 1.21 <.0001 Ascites 1.27.01 Surgical Complexities History of cardiac surgery 1.43 <.0001 Concommitant Cardiac Surgery 1.21.0008 23
Continuous Flow LVAD/BiVAD Implants: 2008 2013, n=9372 Survival by Right Heart Failure Risk Factors Low n=6596, deaths=1481 % Survival Severe (BiVAD) n=260, deaths=120 Moderate (ascites or RAP >= 18 or Bilirubin >= 2) n=2516, deaths=690 Overall P <.0001 Event: Death (censored at transplant and recovery) Months post implant 24
Adult Primary Continuous Flow LVADs and BiVADs Implants: June 2006 December 2013, n=9372 Early hazard Late hazard Risk Factors for Death Hazard Ratio p-value Hazard Ratio p-value Demographics Age (older) 1.36 <.0001 Female 1.20.007 BMI (higher) 1.13 <.0001 Clinical Status History of Stroke 1.30.03 INTERMACS Level 1 1.69 <.0001 INTERMACS Level 2 1.44 <.0001 Destination Therapy 1.24.0005 Non-Cardiac Systems Albumin (lower) 0.90.02 Creatinine (higher) 1.05.0003 Dialysis 2.37 <.0001 BUN (higher) 1.06 <.0001 1.06.01 Right Heart Dysfunction Right Atrial Pressure (higher) 1.11.02 RVAD in same operation 2.45 <.0001 Bilirubin (higher) 1.21 <.0001 Ascites 1.27.01 Surgical Complexities History of cardiac surgery 1.43 <.0001 Concommitant Cardiac Surgery 1.21.0008 25
Tricuspid Repair Following VAD Implantation in INTERMACS 04/04/2014 Howard Song, MD, PhD Co-authors James Mudd, MD; Jill Gelow, MD, MPH; Christopher Chien, MD; Fred Tibayan, MD; Kathryn Hollifield, BSN, RN; David Naftel, PhD; and James Kirklin, MD "This project has been funded in whole or in part with Federal funds from the National Heart, Lung, and Blood Institute, National Institutes of Health, Department of Health and Human Services, under Contract No. HHSN268201100025C" 26
INTERMACS: June 2006 March 2013: TVR Adult Destination Therapy CF primary LVAD/BiVAD, n=2527 Phase of Hazard Early Constant Risk Factors for Death Hazard ratio p-value Hazard ratio p-value Age (years) 1.58 <.0001 BMI (higher) 1.49.0003 Not Married 1.93.0009 Hx of CABG 1.56.02 BUN (higher) 1.10.006 1.06.01 Total Bilirubin (higher) 1.78.02 Creatinine (higher) 1.10.001 Ventilator 2.18.005 INTERMACS Level 1 2.74.0001 INTERMACS Level 2 1.85.004 Bi-VAD 5.05 <.0001 Tricuspid Regurgitation 1.35*.009 * Represents the increased risk for 1 level increase in TV Regurgitation 27
So, is the real risk factor the procedure or the condition? 28
INTERMACS: June 2006 March 2013: TVR Adult Destination Therapy CF primary LVAD/BiVAD, n=2527 Tricuspid Regurg Echo at pre-implant: Moderate/Severe, n=989 Concomitant Tricuspid Surgery n deaths None 757 225 Repair 215 60 % Survival Overall p =.83 Event: Death with a device in place (censored at transplant and recovery Months post implant 29
% Moderate/Severe Tricuspid Regurgitation 100% INTERMACS: June 2006 March 2013: TVR Moderate/Severe Tricuspid Regurgitation over time by TV Repair vs. No TVR Procedure, n=989 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% 100% 100% n=215 n=757 42% n=136 21% n=21 30% n=50 25% 26% n=8 39% 39% n=88 n=17 Pre-implant 1 week 1 month 3 month 6 month 12 month 18 month 24 month TV Repair No TV Procedure Note: N s represent the number of Moderate/Severe Tricuspid Regurg n=75 29% n=17 No TV Procedure, n=757 TV Repair, n= 215 37% n=46 23% n=9 40% n=31 19% n=3 n=12 33% 30% n=3
Continuous Flow LVAD Implants: 2008 2013, n=9111 Era 1: 2008-2010 n=2800, Events=273 Pump Thrombosis % Freedom from Exchange Era 2: 2011-2013 N=6311, events=522 p <.0001 Event: Time to 1 st Exchange Months post implant 31
Figure 20 Continuous Flow LVAD/BiVAD Implants: 2008 2013, n=9372 Operation 1: N=8565, deaths=2039 p <.0001 % Survival Operation 2: n=734, Deaths=242 Operation 3: n=73, Deaths=29 Event: Death (censored at transplant and recovery) Months post implant 32
Education War Era Cardiac Surgery These stratified actuarial depictions are very useful, but They have limited value for predicting outcomes for an individual patient. The stratified actuarial method forces conditions like.. assuming he/she does not receive a heart transplant, or.. assuming there is not explant for recovery Ideally, INTERMACS could provide software or website applications that allow the physician to provide the patient with the most useful predictions of the chances of various outcomes occurring to him/her. 33
Education War Era Cardiac Surgery Predictive modeling for outcome events based on initial implant strategy The goal is useful predictions of the likelihood (probability) of a specific patient (with his/her individual risk profile) experiencing a given outcome depending on the initial implant strategy. The basic methodology utilizes multivariable hazard function analyses on the competing outcomes platform. The basic strategies are BTT, BTC, DT, and BTR. The outcome event of interest are Transplant, Alive on device, Explant for recovery, and death. David Naftel will review our initial steps in this process. 34
Continuous Flow LVAD/BiVAD Implants: 2008 2013, n=9372 BTT: Listed CFLVAD implants 2011-2013, n=1309 Proportion of Patients Outcome % at 1 year Alive (device in place) 50% Transplanted 37% Dead 12% Recovery 1% Figure 6 Months after Implant 35
Continuous Flow LVAD/BiVAD Implants: 2008 2013, n=9372 BTC CFLVAD implants 2011-2013, n=2205 Proportion of Patients Outcome % at 1 year Alive (device in place) 64% Transplanted 20% Dead 15% Recovery 1% Figure 7 Months after Implant 36