- Email: [email protected]
Cardiac transplant outcome of patients supported on left ventricular assist device vs intravenous inotropic therapy
MECHANICAL CIRCULATORY SUPPORT
Cardiac Transplant Outcome of Patients Supported on Left Ventricular Assist Device vs Intravenous Inotropic Therapy Brian E. Jaski, MD, Joseph C. Kim, MPH, David C. Naftel, PhD, John Jarcho, MD, Maria Rosa Costanzo, MD, Howard J. Eisen, MD, James K. Kirklin, MD, and Robert C. Bourge, MD for the Cardiac Transplant Research Database Research Group Background: Although the left ventricular assist device (LVAD) has been increasingly used as a bridge to transplant, its effect on post-transplant outcome is uncertain. We, therefore, designed this study using the Cardiac Transplant Research Database to compare patients supported on an LVAD before transplant with those treated with intravenous inotropic medical therapy. Methods and Results: Of the 5,880 patients transplanted between 1990 and 1997, a total of 502 received support from LVADs and 2,514 received intravenous inotropic medical therapy at the time of transplant. Kaplan-Meier analysis showed no significant difference in post-transplant survival between the LVAD and medical-therapy groups (p ⫽ 0.09). Results of a multivariate Cox regression analysis were consistent with that of the Kaplan-Meier analysis and did not identify LVAD as a significant risk factor for mortality. The percentage of patients who received LVADs as a function of total transplants increased from 2% in 1990 to 16% in 1997. Furthermore, although the number of extracorporeal LVADs remained relatively constant, the number of intracorporeal LVADs increased over time. Multivariate parametric analysis found that the risk factors for post-transplant death in the LVAD group were extracorporeal LVAD use (p ⫽ 0.0004), elevated serum creatinine (p ⫽ 0.05), older donor age (p ⫽ 0.03), increased donor ischemic time (p ⬍ 0.0001), and earlier year of transplant (p ⫽ 0.03). Conclusions: Given a limited donor supply, the intracorporeal LVAD helps the sickest patients survive to transplant and provides post-transplant outcome similar to that of patients supported on inotropic medical therapy. Therefore, patients supported on LVADs before transplant may receive the greatest marginal benefit when compared with other transplant candidates. J Heart Lung Transplant 2001;20:449–456.
From the Cardiac Transplant Research Database Research Group, University of Alabama, Birmingham, Alabama. Submitted July 7, 2000; accepted October 20, 2000. Reprint requests: Brian E. Jaski, MD, San Diego Cardiac Center, Sharp Memorial Hospital, 3131 Berger Ave., San Diego, CA 92123. Telephone: 858-244-6804. Fax: 858-244-6809.
This study was presented in part at the 19th Scientific Session of the International Society for Heart and Lung Transplantation, San Francisco, California, April 1999. Copyright © 2001 by the International Society for Heart and Lung Transplantation. 1053-2498/01/$–see front matter PII S1053-2498(00)00246-1
449
450
Jaski et al.
B
ecause of a limited donor supply, patients at imminent risk of death may require mechanical hemodynamic support before cardiac transplant. Although other forms of mechanical support exist, the left ventricular assist device (LVAD) has been increasingly used to provide circulatory support for patients awaiting transplant who are refractory to intravenous inotropic medical therapy. Nevertheless, the effect of LVAD support on long-term post-transplant outcome is uncertain, and the risk factors for post-transplant outcome associated with LVAD use are unknown.1–3 Furthermore, LVAD patients may have distinct risk factors compared with patients supported on medical therapy and may require a different approach to clinical management. Previous studies that examined post-transplant outcome in LVAD patients were limited by small sample size and inadequate duration of patient follow-up.1,3 We, therefore, designed this study to compare patients supported on LVAD before transplant with those treated with intravenous inotropic medical therapy. We also assessed the risk factors associated with post-transplant mortality in the LVAD population.
METHODS Study Population The Cardiac Transplant Research Database (CTRD) is a multi-institutional, prospectively collected compilation of the total heart transplant experience at participating institutions. Data collection began in January 1990, with data entry, checking, and maintenance housed at the CTRD coordinating center at the University of Alabama at Birmingham. Data for this study included all adult patients (ⱖ18 years of age) transplanted at 38 participating institutions from January 1990 through December 1997. All patients were followed through December 31, 1998. Data compiled included extensive donor and recipient demographic data before transplantation and data regarding major posttransplantation events including death, rejection, infection, retransplantation, malignancy, and coronary angiography.
Study Endpoints and Definitions The primary endpoint of this study was death from all causes. Secondary endpoints were post-transplant infection and rejection. Each study center independently verified cause of death, and the Data Coordination and Analysis Center at the University of Alabama, Birmingham, subsequently verified the cause. We defined an infection episode as an infection that required intravenous antibiotics, or one treated with
The Journal of Heart and Lung Transplantation April 2000
oral antibiotics and thought to be life threatening in the investigator’s opinion. We defined a cardiac rejection episode as a clinical event, usually but not always accompanied by an abnormal endomyocardial biopsy, that resulted in augmentation of immunosuppression. We compared primary and secondary endpoints between patients who received LVAD placement before transplantation and those who received intravenous inotropic medical therapy. We compared risk factors associated with each endpoint between treatment groups.
Definition of Treatment Groups We included in the LVAD treatment group all patients supported on LVADs at the time of transplant. We considered LVADs of various manufacturers and types in this study. Intracorporeal LVADs (n ⫽ 417) included the Novacor and the TCI HeartMate pneumatic and electric models. Extracorporeal LVADs (n ⫽ 85) included the Thoratec, Abiomed, Biomedicus, and Sarns models. We excluded from the analysis patients who required other forms of circulatory mechanical support, including the intra-aortic balloon pump, biventricular mechanical support, extracorporeal life support, or total artificial heart. Patients classified as United Network for Organ Sharing (UNOS) Status 1 and treated with intravenous inotropic therapy (e.g., dobutamine, dopamine, milrinone, etc.) at the time of transplant were included in the medical-therapy group. Status 2 patients who received outpatient inotropic therapy were excluded from the analysis. In this study, we chose the comparison group of UNOS Status 1 patients because their baseline clinical characteristics more closely matched those of the group that received ventricular assist support.
Power of the Study We performed a power calculation using a standard formula for comparing 2 binomial proportions, with a 2-sided test at a significance level of 0.05, and sample sizes of 2,514 and 502 in the medical-therapy and LVAD groups, respectively.4 With this formula, this study had 99% power to detect an 8% difference in 1-year post-transplant survival between LVAD and medical-therapy groups.
Statistical Analysis We compared baseline characteristics between LVAD and medical-therapy groups using the Wilcoxon rank sum test. Kaplan-Meier analysis was performed to
The Journal of Heart and Lung Transplantation Volume 20, Number 4
FIGURE 1 Percentage of patients with LVADs at the
time of transplant. Since 1990, the percentage of patients with LVADs at the time of transplant has increased steadily each year. CTRD, Cardiac Transplant Research Database; LVAD, left ventricular assist device.
determine the probability estimates of time-related freedom from outcome events, and comparisons between groups were made using the log-rank test. We performed multivariate analysis using both Cox regression and parametric analysis in the hazard function domain with stepwise addition of covariates.5,6 All statistical analyses were performed using SAS (SAS Institute, version 6.12, 1996; Cary, NC). We considered a p value of ⱕ0.05 to be statistically significant. All data are represented as mean ⫾ standard deviation unless otherwise noted.
RESULTS Of the 5,880 patients transplanted between January 1990 and December 1997, a total of 502 patients received support from LVADs at the time of transplant. Of these, 409 (81%) received intracorporeal LVADs, 85 (17%) received extracorporeal LVADs, and 8 (2%) received unspecified LVADs. We included 2,514 patients in the Status 1 medical-therapy group. The percentage of patients who received LVADs as a function of total transplants monotonically increased from 2% in 1990 to 16% in 1997 (Figure 1). Furthermore, although the numbers of extracorporeal LVADs remained relatively constant, the numbers of intracorporeal LVADs increased over time (Figure 2). Although equal numbers of intracorporeal and extracorporeal devices were placed into patients before January 1994, intracorporeal devices were placed more often compared with extracorporeal devices after January 1994 (89% vs 11%, respectively).
Jaski et al.
451
FIGURE 2 Number of extracorporeal LVADs and
intracorporeal LVADs used. Since 1990, the number of extracorporeal LVADs used for a bridge to transplant has remained small and constant, whereas the number of intracorporeal LVADs used has increased each year. CTRD, Cardiac Transplant Research Database; LVAD, left ventricular assist device.
Following transplant, a total of 107 (21%) deaths occurred in the LVAD group and 560 (22%) deaths occurred in the medical-therapy group. The LVAD group had 408 episodes of infection, whereas the medical-therapy group had 1,841 episodes. The LVAD and medical-therapy groups experienced 548 and 3,132 episodes of rejection, respectively. The most frequent causes of death in both treatment groups were infection (20%), acute rejection (15%), and early graft failure (12%). Among LVAD patients, infection was the most frequent cause of death (27%), followed by early graft failure (18%) and acute rejection (10%).
Baseline Characteristics We followed LVAD and medical-therapy groups for an average of 20 months and 31 months, respectively (Table I, p ⬍ 0.0001). Treatment groups were comparable at baseline with respect to donor waiting time, serum creatinine, recipient race, donor age, and donor race. However, patients in the LVAD group were slightly younger (50 years vs 52 years, p ⫽ 0.005), more often male (86% vs 79%, p ⬍ 0.001), and slightly taller and heavier (176 cm vs 174 cm, p ⬍ 0.0001, and 78 kg vs 75 kg, p ⫽ 0.0002, respectively) than those in the medical-therapy group. Furthermore, a greater proportion of LVAD patients received blood transfusions before transplant (37% vs 14%, p ⬍ 0.001) and had lower total serum cholesterol levels (149 mg/dl vs 169 mg/dl,
452
Jaski et al.
The Journal of Heart and Lung Transplantation April 2000
TABLE I Baseline characteristics Medical therapy (n ⴝ 2,514)
Follow-up (months) Waiting time (months) Recipient characteristics Gender (male) Race (white) Age (years) Height (cm) Weight (kg) Total cholesterol (mg/dl)a Creatine (mg/dl) Prior transfusion Donor characteristics Gender (male) Race (white) Age (years) Height (cm) Weight (kg) Ischemic time (minutes) a
LVAD (n ⴝ 502)
n
Mean ⴞ SE
n
Mean ⴞ SE
P-value
2,514 2,511
30.6 ⫾ .507 5.1 ⫾ 1.5
502 501
20.4 ⫾ .875 4.6 ⫾ .27
⬍.0001 .15
1,980 2,066 2,513 2,509 2,509 645 2,454 345
79% 82% 51.8 ⫾ .221 174 ⫾ .194 75.2 ⫾ .303 169 ⫾ 2.13 1.35 ⫾ .011 14%
430 409 502 496 499 174 491 188
86% 81% 50.2 ⫾ .518 176 ⫾ .391 77.9 ⫾ .685 149 ⫾ 3.94 1.30 ⫾ .032 37%
⬍.001 .7 .005 ⬍.0001 .0002 ⬍.0001 .07 ⬍.001
1,746 2,006 2,510 2,425 2,495 2,499
70% 80% 30.7 ⫾ .259 173 ⫾ .240 74.2 ⫾ .346 171 ⫾ 1.16
377 395 502 487 502 500
75% 79% 29.7 ⫾ .536 175 ⫾ .485 76.6 ⫾ .719 187 ⫾ 2.64
.01 .5 .10 .003 .005 ⬍.0001
Data on total cholesterol were not collected until January 1, 1996. LVAD, left ventricular assist device.
p ⬍ 0.0001). The LVAD patients were more likely to receive hearts from male donors (75% vs 70%, p ⫽ 0.01), taller and heavier individuals (175 vs 173 cm, p ⫽ 0.003, and 77 vs 74 kg, p ⫽ 0.005), and those with greater ischemic times at the time closest to transplant (187 minutes vs 171 minutes, p ⬍ 0.0001) compared to medical-therapy patients. The average time from LVAD implant to cardiac transplant was greater in those who received intracorporeal devices compared with those who received extracorporeal devices (2.3 months vs 0.6 months, respectively).
than those who received LVADs between 1990 and 1993 (Figure 4, p ⫽ 0.05). Kaplan-Meier analysis showed no significant difference in post-transplant survival between LVAD and medical-therapy groups in either the cohort of patients transplanted between 1990 and 1993 (p ⫽ 0.13) or those transplanted between 1994 and 1997 (p ⫽ 0.09). Results of the multivariate Cox regression analysis were consistent with those of the Kaplan-Meier analysis for each
LVAD vs Medical Therapy Kaplan-Meier analysis showed no significant difference in long-term survival following transplant between the LVAD and medical-therapy groups (Figure 3, p ⫽ 0.09). Multivariate Cox regression analysis did not identify LVAD as a significant risk factor for mortality (p ⫽ 0.054). Furthermore, the p value for this covariate increased after excluding those with extracorporeal devices in the LVAD group (p ⫽ 0.232). Survival at 1-year post-transplant was 82% in the LVAD group and 85% in the medical-therapy group. The 5-year survival rates for LVAD and medical-therapy groups were 72% and 71%, respectively. Patients who received LVADs more recently, between 1994 and 1997, had significantly better survival
FIGURE 3 Kaplan-Meier survival curves for LVAD
and medical-therapy groups (p ⫽ 0.09). CTRD, Cardiac Transplant Research Database; IABP, intra-aortic balloon pump; LVAD, left ventricular assist device.
The Journal of Heart and Lung Transplantation Volume 20, Number 4
FIGURE 4 Kaplan-Meier survival curve comparing
Jaski et al.
453
FIGURE 6 1990 to 1997 freedom from rejection for the
1990 to 1993 vs 1994 to 1997 transplanted cohort (p ⫽ 0.05). CTRD, Cardiac Transplant Research Database; LVAD, left ventricular assist device.
LVAD and medical-therapy groups (p ⫽ 0.8). CTRD, Cardiac Transplant Research Database; IABP, intra-aortic balloon pump; LVAD, left ventricular assist device.
cohort (i.e., 1990 to 1993 cohort, p ⫽ 0.286 vs 1994 to 1997 cohort, p ⫽ 0.059, respectively). The LVAD patients had a significantly lower overall freedom from first infection compared with medicaltherapy patients (Figure 5, p ⫽ 0.0002). However, we found no significant difference in overall freedom from first rejection between LVAD and medicaltherapy groups (Figure 6, p ⫽ 0.8).
post-transplant survival (Figure 7A). Patients who received either intracorporeal devices or inotropic therapy before transplant had significantly greater post-transplant survival than those who received extracorporeal LVADs (Figure 7A, p ⫽ 0.04). Although we noted a trend for better survival in the Thoratec-treated patients compared with those treated with other extracorporeal devices, the small number of patients does not permit definite conclusions (Figure 7B). Multivariate Cox regression analysis performed for the 1994 to 1997 cohort did not identify LVAD as a significant risk factor for mortality. The predicted 1-year probability of death was fit for a “typical” subject (i.e., a subject with an average ischemic time of 180 minutes, serum creatinine level of 1.2 mg/dl, and donor age of 25 years) (Figure 8). The figure shows that this typical subject implanted with an extracorporeal device had a consistently greater probability of death following transplant compared with another typical subject implanted with an intracorporeal device, independent of transplant year.
Intracorporeal vs Extracorporeal Device Placement Patients who received intracorporeal LVADs or intravenous inotropic medical therapy before transplant between 1994 and 1997 had nearly identical
Risk Factors for Death Following LVAD Placement
FIGURE 5 1990 to 1997 freedom from infection for the
LVAD and medical-therapy groups (p ⫽ 0.0002). CTRD, Cardiac Transplant Research Database; IABP, intra-aortic balloon pump; LVAD, left ventricular assist device.
Parametric analysis revealed a single, rapidly declining hazard, with the highest risk during the first month after transplantation (i.e., early phase), whereas after 6 months, the hazard fell to a relatively constant level (i.e., constant phase) (Figure 9). Table II. shows risk factors for post-transplant death in the LVAD group based on multivariate parametric analysis. In the early phase, patients who received extracorporeal LVADs before transplant had higher risk of death compared with those who
454
Jaski et al.
The Journal of Heart and Lung Transplantation April 2000
FIGURE 8 Parametric analysis for a “typical” patient.
The figure shows that this typical patient with an extracorporeal device before transplant has a consistently greater probability of death compared with a subject implanted with an intracorporeal device independent of the year of transplant. CTRD, Cardiac Transplant Research Database; LVAD, left ventricular assist device.
In the early series of 1990 to 1993 post-transplant Kaplan-Meier survival was lower if the duration of LVAD support before transplant was ⬍1 month (p ⫽ 0.03). We did not find this in the larger and more recent 1994 to 1997 cohort (p ⫽ 0.6) or in the sub-group of intracorporeal LVADs from the entire time period (p ⫽ 0.4).
FIGURE 7 1994 to 1997 (A) Kaplan-Meier survival
curves for extracorporeal LVAD, intracorporeal LVAD, and medical-therapy groups. From 1994 to 1997, there was a significant difference in survival between the intracorporeal LVAD or medical-therapy groups and the extracorporeal LVAD group (p ⫽ 0.04). (B) 1994 to 1997 Kaplan-Meier survival curves for Thoratec, other extracorporeal LVAD, intracorporeal LVAD, and medical-therapy groups. From 1994 to 1997, there was a trend for a difference in survival among the intracorporeal LVAD, medical-therapy, and the Thoratec or other extracorporeal LVAD groups (p ⫽ 0.11). CTRD, Cardiac Transplant Research Database; IABP, intra-aortic balloon pump; LVAD, left ventricular assist device.
received intracorporeal devices (p ⫽ 0.0004). The only significant risk factor for death in the constant phase was earlier year of transplant (p ⫽ 0.03).
FIGURE 9 Parametric hazard and survival plots for
LVAD group. Parametric analysis revealed a single, rapidly declining hazard, with the highest risk during the first month after transplantation, reflecting the “early” hazard phase. CTRD, Cardiac Transplant Research Database; LVAD, left ventricular assist device.
The Journal of Heart and Lung Transplantation Volume 20, Number 4
Jaski et al.
TABLE II Risk factors for post-transplant death in LVAD group (n ⫽ 502) p values Risk factor Extracorporeal LVAD Creatinine at transplant (higher) Donor age (older) Ischemic time (longer) Year of transplant (earlier)
Early hazard 0.0004 0.05 0.03 ⬍0.0001
Constant hazard
0.03
* LVAD, left ventricular assist device.
DISCUSSION The number of LVADs used in bridge to transplant has steadily increased since 1990. Whereas the annual use of extracorporeal LVADs remained consistently small, the use of intracorporeal LVADs increased steadily. As a result, the percentage of patients supported on LVADs at the time of heart transplant has increased substantially from 2% in 1990 to 16% in 1997. This study was powered to detect an 8% difference in mortality between treatment groups, from 502 LVAD patients and 2,514 medical-therapy patients in the CTRD. We found that LVAD patients had post-transplant survival similar to UNOS Status 1 patients supported on inotropic medical therapy. This finding is consistent with some, but not all, previous studies comparing post-transplant survival between LVAD and non-LVAD groups.7–12 The results of this study contrast and extend the findings from multivariate logistic regression analysis of 1-, 3-, and 5-year mortality for adult heart transplants performed since 1982 and reported by the Registry of the International Society for Heart and Lung Transplantation (ISHLT) between 1993 and 1999.9,14 –19 These 7 analyses found a statistically significant odds ratio for increased 1-year mortality for pre-transplant use of LVAD or “ICU and life support,” ranging from 1.21 to 1.88. This apparent discordant finding may relate to our analysis comparing LVAD patients with only inotrope-dependent patients rather than all transplant patients. Of note, the ISHLT database, which extends back to 1982, includes a population of LVAD recipients that represents implantation with a variety of LVADs very early in the learning curve. Our analysis of the CTRD began in a more recent era with higher use of intracorporeal LVADs. Finally, our comparison of
455
Kaplan-Meier survival curves may yield a result different from statistical analysis at a single time point. The time-related difference in survival between the 1990 to 1993 and the 1994 to 1997 cohorts may be attributed to improvements in LVAD technology and better post-operative management. Furthermore, intracorporeal devices appear to provide better posttransplant survival than do extracorporeal devices. We found other risk factors for survival in the LVAD group similar to those previously reported in the entire CTRD database, including elevated creatinine, older donor age, and increased donor ischemic time.13,20,21 Currently, LVAD placement is considered for only those patients in imminent risk of death. Previous studies have shown that chronic LVAD support for 1 to 3 months prevents or reverses end-organ dysfunction through improvements in hemodynamic performance and enhancements in end-organ perfusion.22–24 It has been reported that patients supported on LVADs for ⬎30 days have better post-transplant survival than those supported for ⬍30 days, because of overall improvements in physiologic function.25 Multivariate analysis in our series did not identify duration of LVAD support as a significant determinant of early or late survival. We found similar survival in patients supported with intracorporeal LVADs for more or less than 30 days. In this study, patients in the LVAD group received support for an average of 72 ⫾ 67 days and had significantly lower serum creatinine levels than did patients in the medical-therapy group at the time of transplant. This suggests that LVAD patients in this study received sufficient support to achieve improvements in end-organ function before transplant.
Limitations Because LVADs are currently offered to only those patients who fail inotropic medical therapy before transplant, a randomized study could not be performed. Thus, differences in freedom from outcome events using Kaplan-Meier analysis in this study may have been confounded by small but statistically significant differences in baseline characteristics between LVAD and medical-therapy groups. Baseline characteristics, which may have contributed to better survival in the LVAD group, include recipient age and gender, and serum creatinine, whereas those characteristics that may have contributed to worse survival include history of blood transfusion and donor ischemic time. Although we observed statistical differences in baseline characteristics between treatment groups, these differences, in general, were clinically small. Furthermore, we adjusted for all of these covariates in the
456
Jaski et al.
multivariate analysis. Variability in the diagnosis and treatment of infection and rejection episodes between study centers limited this study.9,26 In addition, we excluded from this study patients who received LVADs but did not survive to transplant because the CTRD is limited only to transplanted patients.
CONCLUSIONS Bridge to transplant with an LVAD is a relatively common procedure, and its use will likely continue to increase in the future. In current clinical practice, LVAD implantation is offered only to patients who deteriorate following chronic intravenous inotropic therapy. Earlier in this decade, a greater proportionate use of extracorporeal LVADs was associated with lower post-transplant survival compared with more recent experience. Between 1994 and 1997, intracorporeal LVAD and intravenous medicaltherapy survival post-transplant were similar. Given a limited donor supply, intracorporeal LVADs may help the sickest patients survive to transplant, while providing post-transplant outcome comparable to inotropic medical therapy. Therefore, patients supported on LVADs before transplant may receive the greatest marginal benefit when compared with other transplant candidates. REFERENCES 1. Argenziano M, Catanese KA, Moazami N, et al. The influence of infection on survival and successful transplantation in patients with left ventricular assist devices. J Heart Lung Transplant 1997;16:822–31. 2. Oz MC, Rose EA, Levin HR. Selection criteria for placement of left ventricular assist devices. Am Heart J 1995;129:173–7. 3. Oz MC, Goldstein DJ, Pepino P, et al. Screening scale predicts patients successfully receiving long-term implantable left ventricular assist devices. Circulation 1995;92:II169 –73. 4. Rosner B. Fundamentals of biostatistics, 4 ed. New York: Duxbury Press, 1995. 5. Blackstone EH. Outcome analysis using hazard function methodology. Ann Thorac Surg. 1996;61:S2–7. 6. Cox DR, Oakes D. Analysis of survival data. New York: Chapman and Hall, 1984. 7. Birovljev S, Radovancevic B, Burnett CM, et al. Heart transplantation after mechanical circulatory support: four years’ experience. J Heart Lung Transplant 1992;11:240 –5. 8. Smart FW, Naftel DC, Costanzo MR, et al. Risk factors for early, cumulative, and fatal infections after heart transplantation: a multiinstitutional study. J Heart Lung Transplant. 1996;15:329 – 41. 9. Hosenpud JD, Novick RJ, Breen TJ, Keck B, Daily P. The Registry of the International Society for Heart and Lung Transplantation: twelfth official report—1995. J Heart Lung Transplant 1995;14:805–15. 10. McCarthy PM, Hoercher K. Clinically available intracorporeal left ventricular assist devices. Prog Cardiovascular Dis 2000;43(1):37– 46.
The Journal of Heart and Lung Transplantation April 2000 11. Massad MG, McCarthy PM, Smedira NG, et al. Does successful bridging with the implantable left ventricular assist device affect cardiac transplantation outcome? J Thorac Cardiovasc Surg. 1996;112:1275– 81. 12. Bank AJ, Mir SH, Nguyen DQ, et al. Effects of left ventricular assist devices on outcomes in patients undergoing heart transplantation. Ann Thorac Surg 2000;69(5):1369 –74. 13. Bourge RC, Naftel DC, Costanzo-Nordin MR, et al. Pretransplantation risk factors for death after heart transplantation: a multiinstitutional study. The Transplant Cardiologists Research Database Group. J Heart Lung Transplant 1993;12:549 – 62. 14. Kaye, MP. The Registry of the International Society for Heart and Lung Transplanation: tenth official report—1993. J Heart Lung Transplant 1993;12:541– 8. 15. Hosenpud JD, Novick RJ, Breen TJ, Daily OP. The Registry of the International Society for Heart and Lung Transplantation: eleventh official report—1994. J Heart Lung Transplant 1994;13:561–70. 16. Hosenpud JD, Novick RJ, Bennett LE, Keck B, Fiol B, Daily OP. The Registry of the International Society for Heart and Lung Transplantation: thirteenth official report—1996. J Heart Lung Transplant 1996;15:655–74. 17. Hosenpud JD, Bennett LE, Keck B, Fiol B, Novick RJ. The Registry of the International Society for Heart and Lung Transplantation: fourteenth official report—1997. J Heart Lung Transplant 1997;16:691–712. 18. Hosenpud JD, Bennett LE, Keck B, Fiol B, Boucek MM, Novick RJ. The Registry of the International Society for Heart and Lung Transplantation: fifteenth official report— 1998. J Heart Lung Transplant 1998;17:656 – 68. 19. Hosenpud JD, Bennett LE, Keck B, Fiol B, Boucek MM, Novick RJ. The Registry of the International Society for Heart and Lung Transplantation: sixteenth official report— 1999. J Heart Lung Transplant 1999;18:611–26. 20. Young JB, Naftel DC, Bourge RC, et al. Matching the heart donor and heart transplant recipient. Clues for successful expansion of the donor pool: a multivariable, multiinstitutional report. The Cardiac Transplant Research Database Group. J Heart Lung Transplant 1994;13:353– 64. 21. Rodeheffer RJ, Naftel DC, Stevenson LW, et al. Secular trends in cardiac transplant recipient and donor management in the United States, 1990 to 1994. A multi-institutional study. Cardiac Transplant Research Database Group. Circulation 1996;94:2883–9. 22. Burnett CM, Duncan JM, Frazier OH, Sweeney MS, Vega JD, Radovancevic B. Improved multiorgan function after prolonged univentricular support. Ann Thorac Surg 1993;55:65–71. 23. Frazier OH, Macris MP, Myers TJ, et al. Improved survival after extended bridge to cardiac transplantation. Ann Thorac Surg 1994;57:1416 –22. 24. McCarthy PM. HeartMate implantable left ventricular assist device: bridge to transplantation and future applications. Ann Thorac Surg 1995;59:S46 –51. 25. Ashton RCJ, Goldstein DJ, Rose EA, Weinberg AD, Levin HR, Oz MC. Duration of left ventricular assist device support affects transplant survival. J Heart Lung Transplant 1996;15:1151–7. 26. Kobashigawa JA, Kirklin JK, Naftel DC, et al. Pretransplantation risk factors for acute rejection after heart transplantation: a multiinstitutional study. The Transplant Cardiologists Research Database Group. J Heart Lung Transplant 1993; 12:355– 66.
Cardiac Transplant Outcome of Patients Supported on Left Ventricular Assist Device vs Intravenous Inotropic Therapy Brian E. Jaski, MD, Joseph C. Kim, MPH, David C. Naftel, PhD, John Jarcho, MD, Maria Rosa Costanzo, MD, Howard J. Eisen, MD, James K. Kirklin, MD, and Robert C. Bourge, MD for the Cardiac Transplant Research Database Research Group Background: Although the left ventricular assist device (LVAD) has been increasingly used as a bridge to transplant, its effect on post-transplant outcome is uncertain. We, therefore, designed this study using the Cardiac Transplant Research Database to compare patients supported on an LVAD before transplant with those treated with intravenous inotropic medical therapy. Methods and Results: Of the 5,880 patients transplanted between 1990 and 1997, a total of 502 received support from LVADs and 2,514 received intravenous inotropic medical therapy at the time of transplant. Kaplan-Meier analysis showed no significant difference in post-transplant survival between the LVAD and medical-therapy groups (p ⫽ 0.09). Results of a multivariate Cox regression analysis were consistent with that of the Kaplan-Meier analysis and did not identify LVAD as a significant risk factor for mortality. The percentage of patients who received LVADs as a function of total transplants increased from 2% in 1990 to 16% in 1997. Furthermore, although the number of extracorporeal LVADs remained relatively constant, the number of intracorporeal LVADs increased over time. Multivariate parametric analysis found that the risk factors for post-transplant death in the LVAD group were extracorporeal LVAD use (p ⫽ 0.0004), elevated serum creatinine (p ⫽ 0.05), older donor age (p ⫽ 0.03), increased donor ischemic time (p ⬍ 0.0001), and earlier year of transplant (p ⫽ 0.03). Conclusions: Given a limited donor supply, the intracorporeal LVAD helps the sickest patients survive to transplant and provides post-transplant outcome similar to that of patients supported on inotropic medical therapy. Therefore, patients supported on LVADs before transplant may receive the greatest marginal benefit when compared with other transplant candidates. J Heart Lung Transplant 2001;20:449–456.
From the Cardiac Transplant Research Database Research Group, University of Alabama, Birmingham, Alabama. Submitted July 7, 2000; accepted October 20, 2000. Reprint requests: Brian E. Jaski, MD, San Diego Cardiac Center, Sharp Memorial Hospital, 3131 Berger Ave., San Diego, CA 92123. Telephone: 858-244-6804. Fax: 858-244-6809.
This study was presented in part at the 19th Scientific Session of the International Society for Heart and Lung Transplantation, San Francisco, California, April 1999. Copyright © 2001 by the International Society for Heart and Lung Transplantation. 1053-2498/01/$–see front matter PII S1053-2498(00)00246-1
449
450
Jaski et al.
B
ecause of a limited donor supply, patients at imminent risk of death may require mechanical hemodynamic support before cardiac transplant. Although other forms of mechanical support exist, the left ventricular assist device (LVAD) has been increasingly used to provide circulatory support for patients awaiting transplant who are refractory to intravenous inotropic medical therapy. Nevertheless, the effect of LVAD support on long-term post-transplant outcome is uncertain, and the risk factors for post-transplant outcome associated with LVAD use are unknown.1–3 Furthermore, LVAD patients may have distinct risk factors compared with patients supported on medical therapy and may require a different approach to clinical management. Previous studies that examined post-transplant outcome in LVAD patients were limited by small sample size and inadequate duration of patient follow-up.1,3 We, therefore, designed this study to compare patients supported on LVAD before transplant with those treated with intravenous inotropic medical therapy. We also assessed the risk factors associated with post-transplant mortality in the LVAD population.
METHODS Study Population The Cardiac Transplant Research Database (CTRD) is a multi-institutional, prospectively collected compilation of the total heart transplant experience at participating institutions. Data collection began in January 1990, with data entry, checking, and maintenance housed at the CTRD coordinating center at the University of Alabama at Birmingham. Data for this study included all adult patients (ⱖ18 years of age) transplanted at 38 participating institutions from January 1990 through December 1997. All patients were followed through December 31, 1998. Data compiled included extensive donor and recipient demographic data before transplantation and data regarding major posttransplantation events including death, rejection, infection, retransplantation, malignancy, and coronary angiography.
Study Endpoints and Definitions The primary endpoint of this study was death from all causes. Secondary endpoints were post-transplant infection and rejection. Each study center independently verified cause of death, and the Data Coordination and Analysis Center at the University of Alabama, Birmingham, subsequently verified the cause. We defined an infection episode as an infection that required intravenous antibiotics, or one treated with
The Journal of Heart and Lung Transplantation April 2000
oral antibiotics and thought to be life threatening in the investigator’s opinion. We defined a cardiac rejection episode as a clinical event, usually but not always accompanied by an abnormal endomyocardial biopsy, that resulted in augmentation of immunosuppression. We compared primary and secondary endpoints between patients who received LVAD placement before transplantation and those who received intravenous inotropic medical therapy. We compared risk factors associated with each endpoint between treatment groups.
Definition of Treatment Groups We included in the LVAD treatment group all patients supported on LVADs at the time of transplant. We considered LVADs of various manufacturers and types in this study. Intracorporeal LVADs (n ⫽ 417) included the Novacor and the TCI HeartMate pneumatic and electric models. Extracorporeal LVADs (n ⫽ 85) included the Thoratec, Abiomed, Biomedicus, and Sarns models. We excluded from the analysis patients who required other forms of circulatory mechanical support, including the intra-aortic balloon pump, biventricular mechanical support, extracorporeal life support, or total artificial heart. Patients classified as United Network for Organ Sharing (UNOS) Status 1 and treated with intravenous inotropic therapy (e.g., dobutamine, dopamine, milrinone, etc.) at the time of transplant were included in the medical-therapy group. Status 2 patients who received outpatient inotropic therapy were excluded from the analysis. In this study, we chose the comparison group of UNOS Status 1 patients because their baseline clinical characteristics more closely matched those of the group that received ventricular assist support.
Power of the Study We performed a power calculation using a standard formula for comparing 2 binomial proportions, with a 2-sided test at a significance level of 0.05, and sample sizes of 2,514 and 502 in the medical-therapy and LVAD groups, respectively.4 With this formula, this study had 99% power to detect an 8% difference in 1-year post-transplant survival between LVAD and medical-therapy groups.
Statistical Analysis We compared baseline characteristics between LVAD and medical-therapy groups using the Wilcoxon rank sum test. Kaplan-Meier analysis was performed to
The Journal of Heart and Lung Transplantation Volume 20, Number 4
FIGURE 1 Percentage of patients with LVADs at the
time of transplant. Since 1990, the percentage of patients with LVADs at the time of transplant has increased steadily each year. CTRD, Cardiac Transplant Research Database; LVAD, left ventricular assist device.
determine the probability estimates of time-related freedom from outcome events, and comparisons between groups were made using the log-rank test. We performed multivariate analysis using both Cox regression and parametric analysis in the hazard function domain with stepwise addition of covariates.5,6 All statistical analyses were performed using SAS (SAS Institute, version 6.12, 1996; Cary, NC). We considered a p value of ⱕ0.05 to be statistically significant. All data are represented as mean ⫾ standard deviation unless otherwise noted.
RESULTS Of the 5,880 patients transplanted between January 1990 and December 1997, a total of 502 patients received support from LVADs at the time of transplant. Of these, 409 (81%) received intracorporeal LVADs, 85 (17%) received extracorporeal LVADs, and 8 (2%) received unspecified LVADs. We included 2,514 patients in the Status 1 medical-therapy group. The percentage of patients who received LVADs as a function of total transplants monotonically increased from 2% in 1990 to 16% in 1997 (Figure 1). Furthermore, although the numbers of extracorporeal LVADs remained relatively constant, the numbers of intracorporeal LVADs increased over time (Figure 2). Although equal numbers of intracorporeal and extracorporeal devices were placed into patients before January 1994, intracorporeal devices were placed more often compared with extracorporeal devices after January 1994 (89% vs 11%, respectively).
Jaski et al.
451
FIGURE 2 Number of extracorporeal LVADs and
intracorporeal LVADs used. Since 1990, the number of extracorporeal LVADs used for a bridge to transplant has remained small and constant, whereas the number of intracorporeal LVADs used has increased each year. CTRD, Cardiac Transplant Research Database; LVAD, left ventricular assist device.
Following transplant, a total of 107 (21%) deaths occurred in the LVAD group and 560 (22%) deaths occurred in the medical-therapy group. The LVAD group had 408 episodes of infection, whereas the medical-therapy group had 1,841 episodes. The LVAD and medical-therapy groups experienced 548 and 3,132 episodes of rejection, respectively. The most frequent causes of death in both treatment groups were infection (20%), acute rejection (15%), and early graft failure (12%). Among LVAD patients, infection was the most frequent cause of death (27%), followed by early graft failure (18%) and acute rejection (10%).
Baseline Characteristics We followed LVAD and medical-therapy groups for an average of 20 months and 31 months, respectively (Table I, p ⬍ 0.0001). Treatment groups were comparable at baseline with respect to donor waiting time, serum creatinine, recipient race, donor age, and donor race. However, patients in the LVAD group were slightly younger (50 years vs 52 years, p ⫽ 0.005), more often male (86% vs 79%, p ⬍ 0.001), and slightly taller and heavier (176 cm vs 174 cm, p ⬍ 0.0001, and 78 kg vs 75 kg, p ⫽ 0.0002, respectively) than those in the medical-therapy group. Furthermore, a greater proportion of LVAD patients received blood transfusions before transplant (37% vs 14%, p ⬍ 0.001) and had lower total serum cholesterol levels (149 mg/dl vs 169 mg/dl,
452
Jaski et al.
The Journal of Heart and Lung Transplantation April 2000
TABLE I Baseline characteristics Medical therapy (n ⴝ 2,514)
Follow-up (months) Waiting time (months) Recipient characteristics Gender (male) Race (white) Age (years) Height (cm) Weight (kg) Total cholesterol (mg/dl)a Creatine (mg/dl) Prior transfusion Donor characteristics Gender (male) Race (white) Age (years) Height (cm) Weight (kg) Ischemic time (minutes) a
LVAD (n ⴝ 502)
n
Mean ⴞ SE
n
Mean ⴞ SE
P-value
2,514 2,511
30.6 ⫾ .507 5.1 ⫾ 1.5
502 501
20.4 ⫾ .875 4.6 ⫾ .27
⬍.0001 .15
1,980 2,066 2,513 2,509 2,509 645 2,454 345
79% 82% 51.8 ⫾ .221 174 ⫾ .194 75.2 ⫾ .303 169 ⫾ 2.13 1.35 ⫾ .011 14%
430 409 502 496 499 174 491 188
86% 81% 50.2 ⫾ .518 176 ⫾ .391 77.9 ⫾ .685 149 ⫾ 3.94 1.30 ⫾ .032 37%
⬍.001 .7 .005 ⬍.0001 .0002 ⬍.0001 .07 ⬍.001
1,746 2,006 2,510 2,425 2,495 2,499
70% 80% 30.7 ⫾ .259 173 ⫾ .240 74.2 ⫾ .346 171 ⫾ 1.16
377 395 502 487 502 500
75% 79% 29.7 ⫾ .536 175 ⫾ .485 76.6 ⫾ .719 187 ⫾ 2.64
.01 .5 .10 .003 .005 ⬍.0001
Data on total cholesterol were not collected until January 1, 1996. LVAD, left ventricular assist device.
p ⬍ 0.0001). The LVAD patients were more likely to receive hearts from male donors (75% vs 70%, p ⫽ 0.01), taller and heavier individuals (175 vs 173 cm, p ⫽ 0.003, and 77 vs 74 kg, p ⫽ 0.005), and those with greater ischemic times at the time closest to transplant (187 minutes vs 171 minutes, p ⬍ 0.0001) compared to medical-therapy patients. The average time from LVAD implant to cardiac transplant was greater in those who received intracorporeal devices compared with those who received extracorporeal devices (2.3 months vs 0.6 months, respectively).
than those who received LVADs between 1990 and 1993 (Figure 4, p ⫽ 0.05). Kaplan-Meier analysis showed no significant difference in post-transplant survival between LVAD and medical-therapy groups in either the cohort of patients transplanted between 1990 and 1993 (p ⫽ 0.13) or those transplanted between 1994 and 1997 (p ⫽ 0.09). Results of the multivariate Cox regression analysis were consistent with those of the Kaplan-Meier analysis for each
LVAD vs Medical Therapy Kaplan-Meier analysis showed no significant difference in long-term survival following transplant between the LVAD and medical-therapy groups (Figure 3, p ⫽ 0.09). Multivariate Cox regression analysis did not identify LVAD as a significant risk factor for mortality (p ⫽ 0.054). Furthermore, the p value for this covariate increased after excluding those with extracorporeal devices in the LVAD group (p ⫽ 0.232). Survival at 1-year post-transplant was 82% in the LVAD group and 85% in the medical-therapy group. The 5-year survival rates for LVAD and medical-therapy groups were 72% and 71%, respectively. Patients who received LVADs more recently, between 1994 and 1997, had significantly better survival
FIGURE 3 Kaplan-Meier survival curves for LVAD
and medical-therapy groups (p ⫽ 0.09). CTRD, Cardiac Transplant Research Database; IABP, intra-aortic balloon pump; LVAD, left ventricular assist device.
The Journal of Heart and Lung Transplantation Volume 20, Number 4
FIGURE 4 Kaplan-Meier survival curve comparing
Jaski et al.
453
FIGURE 6 1990 to 1997 freedom from rejection for the
1990 to 1993 vs 1994 to 1997 transplanted cohort (p ⫽ 0.05). CTRD, Cardiac Transplant Research Database; LVAD, left ventricular assist device.
LVAD and medical-therapy groups (p ⫽ 0.8). CTRD, Cardiac Transplant Research Database; IABP, intra-aortic balloon pump; LVAD, left ventricular assist device.
cohort (i.e., 1990 to 1993 cohort, p ⫽ 0.286 vs 1994 to 1997 cohort, p ⫽ 0.059, respectively). The LVAD patients had a significantly lower overall freedom from first infection compared with medicaltherapy patients (Figure 5, p ⫽ 0.0002). However, we found no significant difference in overall freedom from first rejection between LVAD and medicaltherapy groups (Figure 6, p ⫽ 0.8).
post-transplant survival (Figure 7A). Patients who received either intracorporeal devices or inotropic therapy before transplant had significantly greater post-transplant survival than those who received extracorporeal LVADs (Figure 7A, p ⫽ 0.04). Although we noted a trend for better survival in the Thoratec-treated patients compared with those treated with other extracorporeal devices, the small number of patients does not permit definite conclusions (Figure 7B). Multivariate Cox regression analysis performed for the 1994 to 1997 cohort did not identify LVAD as a significant risk factor for mortality. The predicted 1-year probability of death was fit for a “typical” subject (i.e., a subject with an average ischemic time of 180 minutes, serum creatinine level of 1.2 mg/dl, and donor age of 25 years) (Figure 8). The figure shows that this typical subject implanted with an extracorporeal device had a consistently greater probability of death following transplant compared with another typical subject implanted with an intracorporeal device, independent of transplant year.
Intracorporeal vs Extracorporeal Device Placement Patients who received intracorporeal LVADs or intravenous inotropic medical therapy before transplant between 1994 and 1997 had nearly identical
Risk Factors for Death Following LVAD Placement
FIGURE 5 1990 to 1997 freedom from infection for the
LVAD and medical-therapy groups (p ⫽ 0.0002). CTRD, Cardiac Transplant Research Database; IABP, intra-aortic balloon pump; LVAD, left ventricular assist device.
Parametric analysis revealed a single, rapidly declining hazard, with the highest risk during the first month after transplantation (i.e., early phase), whereas after 6 months, the hazard fell to a relatively constant level (i.e., constant phase) (Figure 9). Table II. shows risk factors for post-transplant death in the LVAD group based on multivariate parametric analysis. In the early phase, patients who received extracorporeal LVADs before transplant had higher risk of death compared with those who
454
Jaski et al.
The Journal of Heart and Lung Transplantation April 2000
FIGURE 8 Parametric analysis for a “typical” patient.
The figure shows that this typical patient with an extracorporeal device before transplant has a consistently greater probability of death compared with a subject implanted with an intracorporeal device independent of the year of transplant. CTRD, Cardiac Transplant Research Database; LVAD, left ventricular assist device.
In the early series of 1990 to 1993 post-transplant Kaplan-Meier survival was lower if the duration of LVAD support before transplant was ⬍1 month (p ⫽ 0.03). We did not find this in the larger and more recent 1994 to 1997 cohort (p ⫽ 0.6) or in the sub-group of intracorporeal LVADs from the entire time period (p ⫽ 0.4).
FIGURE 7 1994 to 1997 (A) Kaplan-Meier survival
curves for extracorporeal LVAD, intracorporeal LVAD, and medical-therapy groups. From 1994 to 1997, there was a significant difference in survival between the intracorporeal LVAD or medical-therapy groups and the extracorporeal LVAD group (p ⫽ 0.04). (B) 1994 to 1997 Kaplan-Meier survival curves for Thoratec, other extracorporeal LVAD, intracorporeal LVAD, and medical-therapy groups. From 1994 to 1997, there was a trend for a difference in survival among the intracorporeal LVAD, medical-therapy, and the Thoratec or other extracorporeal LVAD groups (p ⫽ 0.11). CTRD, Cardiac Transplant Research Database; IABP, intra-aortic balloon pump; LVAD, left ventricular assist device.
received intracorporeal devices (p ⫽ 0.0004). The only significant risk factor for death in the constant phase was earlier year of transplant (p ⫽ 0.03).
FIGURE 9 Parametric hazard and survival plots for
LVAD group. Parametric analysis revealed a single, rapidly declining hazard, with the highest risk during the first month after transplantation, reflecting the “early” hazard phase. CTRD, Cardiac Transplant Research Database; LVAD, left ventricular assist device.
The Journal of Heart and Lung Transplantation Volume 20, Number 4
Jaski et al.
TABLE II Risk factors for post-transplant death in LVAD group (n ⫽ 502) p values Risk factor Extracorporeal LVAD Creatinine at transplant (higher) Donor age (older) Ischemic time (longer) Year of transplant (earlier)
Early hazard 0.0004 0.05 0.03 ⬍0.0001
Constant hazard
0.03
* LVAD, left ventricular assist device.
DISCUSSION The number of LVADs used in bridge to transplant has steadily increased since 1990. Whereas the annual use of extracorporeal LVADs remained consistently small, the use of intracorporeal LVADs increased steadily. As a result, the percentage of patients supported on LVADs at the time of heart transplant has increased substantially from 2% in 1990 to 16% in 1997. This study was powered to detect an 8% difference in mortality between treatment groups, from 502 LVAD patients and 2,514 medical-therapy patients in the CTRD. We found that LVAD patients had post-transplant survival similar to UNOS Status 1 patients supported on inotropic medical therapy. This finding is consistent with some, but not all, previous studies comparing post-transplant survival between LVAD and non-LVAD groups.7–12 The results of this study contrast and extend the findings from multivariate logistic regression analysis of 1-, 3-, and 5-year mortality for adult heart transplants performed since 1982 and reported by the Registry of the International Society for Heart and Lung Transplantation (ISHLT) between 1993 and 1999.9,14 –19 These 7 analyses found a statistically significant odds ratio for increased 1-year mortality for pre-transplant use of LVAD or “ICU and life support,” ranging from 1.21 to 1.88. This apparent discordant finding may relate to our analysis comparing LVAD patients with only inotrope-dependent patients rather than all transplant patients. Of note, the ISHLT database, which extends back to 1982, includes a population of LVAD recipients that represents implantation with a variety of LVADs very early in the learning curve. Our analysis of the CTRD began in a more recent era with higher use of intracorporeal LVADs. Finally, our comparison of
455
Kaplan-Meier survival curves may yield a result different from statistical analysis at a single time point. The time-related difference in survival between the 1990 to 1993 and the 1994 to 1997 cohorts may be attributed to improvements in LVAD technology and better post-operative management. Furthermore, intracorporeal devices appear to provide better posttransplant survival than do extracorporeal devices. We found other risk factors for survival in the LVAD group similar to those previously reported in the entire CTRD database, including elevated creatinine, older donor age, and increased donor ischemic time.13,20,21 Currently, LVAD placement is considered for only those patients in imminent risk of death. Previous studies have shown that chronic LVAD support for 1 to 3 months prevents or reverses end-organ dysfunction through improvements in hemodynamic performance and enhancements in end-organ perfusion.22–24 It has been reported that patients supported on LVADs for ⬎30 days have better post-transplant survival than those supported for ⬍30 days, because of overall improvements in physiologic function.25 Multivariate analysis in our series did not identify duration of LVAD support as a significant determinant of early or late survival. We found similar survival in patients supported with intracorporeal LVADs for more or less than 30 days. In this study, patients in the LVAD group received support for an average of 72 ⫾ 67 days and had significantly lower serum creatinine levels than did patients in the medical-therapy group at the time of transplant. This suggests that LVAD patients in this study received sufficient support to achieve improvements in end-organ function before transplant.
Limitations Because LVADs are currently offered to only those patients who fail inotropic medical therapy before transplant, a randomized study could not be performed. Thus, differences in freedom from outcome events using Kaplan-Meier analysis in this study may have been confounded by small but statistically significant differences in baseline characteristics between LVAD and medical-therapy groups. Baseline characteristics, which may have contributed to better survival in the LVAD group, include recipient age and gender, and serum creatinine, whereas those characteristics that may have contributed to worse survival include history of blood transfusion and donor ischemic time. Although we observed statistical differences in baseline characteristics between treatment groups, these differences, in general, were clinically small. Furthermore, we adjusted for all of these covariates in the
456
Jaski et al.
multivariate analysis. Variability in the diagnosis and treatment of infection and rejection episodes between study centers limited this study.9,26 In addition, we excluded from this study patients who received LVADs but did not survive to transplant because the CTRD is limited only to transplanted patients.
CONCLUSIONS Bridge to transplant with an LVAD is a relatively common procedure, and its use will likely continue to increase in the future. In current clinical practice, LVAD implantation is offered only to patients who deteriorate following chronic intravenous inotropic therapy. Earlier in this decade, a greater proportionate use of extracorporeal LVADs was associated with lower post-transplant survival compared with more recent experience. Between 1994 and 1997, intracorporeal LVAD and intravenous medicaltherapy survival post-transplant were similar. Given a limited donor supply, intracorporeal LVADs may help the sickest patients survive to transplant, while providing post-transplant outcome comparable to inotropic medical therapy. Therefore, patients supported on LVADs before transplant may receive the greatest marginal benefit when compared with other transplant candidates. REFERENCES 1. Argenziano M, Catanese KA, Moazami N, et al. The influence of infection on survival and successful transplantation in patients with left ventricular assist devices. J Heart Lung Transplant 1997;16:822–31. 2. Oz MC, Rose EA, Levin HR. Selection criteria for placement of left ventricular assist devices. Am Heart J 1995;129:173–7. 3. Oz MC, Goldstein DJ, Pepino P, et al. Screening scale predicts patients successfully receiving long-term implantable left ventricular assist devices. Circulation 1995;92:II169 –73. 4. Rosner B. Fundamentals of biostatistics, 4 ed. New York: Duxbury Press, 1995. 5. Blackstone EH. Outcome analysis using hazard function methodology. Ann Thorac Surg. 1996;61:S2–7. 6. Cox DR, Oakes D. Analysis of survival data. New York: Chapman and Hall, 1984. 7. Birovljev S, Radovancevic B, Burnett CM, et al. Heart transplantation after mechanical circulatory support: four years’ experience. J Heart Lung Transplant 1992;11:240 –5. 8. Smart FW, Naftel DC, Costanzo MR, et al. Risk factors for early, cumulative, and fatal infections after heart transplantation: a multiinstitutional study. J Heart Lung Transplant. 1996;15:329 – 41. 9. Hosenpud JD, Novick RJ, Breen TJ, Keck B, Daily P. The Registry of the International Society for Heart and Lung Transplantation: twelfth official report—1995. J Heart Lung Transplant 1995;14:805–15. 10. McCarthy PM, Hoercher K. Clinically available intracorporeal left ventricular assist devices. Prog Cardiovascular Dis 2000;43(1):37– 46.
The Journal of Heart and Lung Transplantation April 2000 11. Massad MG, McCarthy PM, Smedira NG, et al. Does successful bridging with the implantable left ventricular assist device affect cardiac transplantation outcome? J Thorac Cardiovasc Surg. 1996;112:1275– 81. 12. Bank AJ, Mir SH, Nguyen DQ, et al. Effects of left ventricular assist devices on outcomes in patients undergoing heart transplantation. Ann Thorac Surg 2000;69(5):1369 –74. 13. Bourge RC, Naftel DC, Costanzo-Nordin MR, et al. Pretransplantation risk factors for death after heart transplantation: a multiinstitutional study. The Transplant Cardiologists Research Database Group. J Heart Lung Transplant 1993;12:549 – 62. 14. Kaye, MP. The Registry of the International Society for Heart and Lung Transplanation: tenth official report—1993. J Heart Lung Transplant 1993;12:541– 8. 15. Hosenpud JD, Novick RJ, Breen TJ, Daily OP. The Registry of the International Society for Heart and Lung Transplantation: eleventh official report—1994. J Heart Lung Transplant 1994;13:561–70. 16. Hosenpud JD, Novick RJ, Bennett LE, Keck B, Fiol B, Daily OP. The Registry of the International Society for Heart and Lung Transplantation: thirteenth official report—1996. J Heart Lung Transplant 1996;15:655–74. 17. Hosenpud JD, Bennett LE, Keck B, Fiol B, Novick RJ. The Registry of the International Society for Heart and Lung Transplantation: fourteenth official report—1997. J Heart Lung Transplant 1997;16:691–712. 18. Hosenpud JD, Bennett LE, Keck B, Fiol B, Boucek MM, Novick RJ. The Registry of the International Society for Heart and Lung Transplantation: fifteenth official report— 1998. J Heart Lung Transplant 1998;17:656 – 68. 19. Hosenpud JD, Bennett LE, Keck B, Fiol B, Boucek MM, Novick RJ. The Registry of the International Society for Heart and Lung Transplantation: sixteenth official report— 1999. J Heart Lung Transplant 1999;18:611–26. 20. Young JB, Naftel DC, Bourge RC, et al. Matching the heart donor and heart transplant recipient. Clues for successful expansion of the donor pool: a multivariable, multiinstitutional report. The Cardiac Transplant Research Database Group. J Heart Lung Transplant 1994;13:353– 64. 21. Rodeheffer RJ, Naftel DC, Stevenson LW, et al. Secular trends in cardiac transplant recipient and donor management in the United States, 1990 to 1994. A multi-institutional study. Cardiac Transplant Research Database Group. Circulation 1996;94:2883–9. 22. Burnett CM, Duncan JM, Frazier OH, Sweeney MS, Vega JD, Radovancevic B. Improved multiorgan function after prolonged univentricular support. Ann Thorac Surg 1993;55:65–71. 23. Frazier OH, Macris MP, Myers TJ, et al. Improved survival after extended bridge to cardiac transplantation. Ann Thorac Surg 1994;57:1416 –22. 24. McCarthy PM. HeartMate implantable left ventricular assist device: bridge to transplantation and future applications. Ann Thorac Surg 1995;59:S46 –51. 25. Ashton RCJ, Goldstein DJ, Rose EA, Weinberg AD, Levin HR, Oz MC. Duration of left ventricular assist device support affects transplant survival. J Heart Lung Transplant 1996;15:1151–7. 26. Kobashigawa JA, Kirklin JK, Naftel DC, et al. Pretransplantation risk factors for acute rejection after heart transplantation: a multiinstitutional study. The Transplant Cardiologists Research Database Group. J Heart Lung Transplant 1993; 12:355– 66.