ctToooahreb

Risk Factor Analysis of Late Survival After Heart TransplantationAccording to Donor Profile: A Multi-Institutional Retrospective Studyof 512 Transplants

A.I. Fiorelli, J.N. Branco, J.J. Dinkhuysen, J.L. Oliveira Junior, T.V. Pereira, L.F.L. Dinardi, M.M. Santos,R.R. Dias, L.A. Pereira, and N.A.G. Stolf

ABSTRACT

Introduction. Patients with terminal heart failure have increased more than the availableorgans leading to a high mortality rate on the waiting list. Use of Marginal and expandedcriteria donors has increased due to the heart shortage.Objective. We analyzed all heart transplantations (HTx) in Sao Paulo state over 8 yearsfor donor profile and recipient risk factors.Method. This multi-institutional review collected HTx data from all institutions in thestate of Sao Paulo, Brazil. From 2002 to 2008 (6 years), only 512 (28.8%) of 1777 availableheart donors were accepted for transplantation. All medical records were analyzedretrospectively; none of the used donors was excluded, even those considered to benonstandard.Results. The hospital mortality rate was 27.9% (n � 143) and the average follow-up timewas 29.4 � 28.4 months. The survival rate was 55.5% (n � 285) at 6 years after HTx. Univariateanalysis showed the following factors to impact survival: age (P � .0004), arterial hypertension(P � .4620), norepinephrine (P � .0450), cardiac arrest (P � .8500), diabetes mellitus (P �.5120), infection (P � .1470), CKMB (creatine kinase MB) (P � .8694), creatinine (P � .7225),and Na� (P � .3273). On multivariate analysis, only age showed significance; logisticregression showed a significant cut-off at 40 years: organs from donors older than 40 yearsshowed a lower late survival rates (P � .0032).Conclusions. Donor age older than 40 years represents an important risk factor forsurvival after HTx. Neither donor gender nor norepinephrine use negatively affected early

survival.

pl

THE SHORTAGE of heart donors for transplantationhas been one of the main reasons to use marginal or

expanded criteria donors.1–3 In contrast, the number ofandidates with terminal heart failure has increased morehan available organs, with increased waiting list mortality.he clinical experience has shown good results with the usef marginal donors. However, it is well known that inferiorrgan quality interferes with immediate and late transplantutcomes. Efforts to amplify donor numbers are commend-ble when they do not damage the recipient. The literatureas polemically discussed the use of 2 waiting lists for adultecipients with marginal donors for critical recipients; how-ver, this proposal is not accepted in our country and has

een strongly criticized for ethical reasons.4

© 2012 by Elsevier Inc. All rights reserved.360 Park Avenue South, New York, NY 10010-1710

Transplantation Proceedings, 44, 2469–2472 (2012)

Donor risk factors that play important roles in hearttransplantation (HTx) evolution are controversial becauseof differences in the local factors that can interfere with ananalysis. Unfortunately, most of the information comesfrom multi-institutional databases that may be nonhomoge-neous.5,6 Therefore, they may not apply to a specific

opulation. Each community may have a distinct waitingist, patient profile, quality of captured organs, as well as

From the Heart Institute of Sao Paulo University MedicalSchool, Sao Paulo, Brazil.

Address reprint requests to Alfredo I. Fiorelli, Rua Morgado deMateus 126/81, Sao Paulo/SP, Brazil, CEP: 04015-050. E-mail:

fiorelliai@uol.com.br

0041-1345/–see front matterhttp://dx.doi.org/10.1016/j.transproceed.2012.07.025

2469

cf

ic

mcwo

davndmaiSss

2470 FIORELLI, BRANCO, DINKHUYSEN ET AL

potential donor family susceptibility to donation.7–9 Thehoice of heart and lung donors is more selective than thator other organs.8,10,11 This study has the main objective to

analyze all HTx performed in the state of Sao Paulo, Brazilover 6 years, emphasizing donor profile and risk factors forpatient immediate and late evolution.

MATERIALS AND METHODSStudy Population

The collected data in this multi institutional review were obtainedfrom all institutions that perform HTx in the state of Sao Paulo,Brazil whose data are registered at the Transplantation Depart-ment of the Health State Secretary. multi-institutional review.From 2002 to 2008 (6 years) 512/1777 available donor hearts(28.8%) were accepted; their medical records were analyzed ret-rospectively. No donor was excluded from study, even those whowere nonstandard (marginal donors). The characteristics of thedonor population obtained from medical records were analyzed asrisk variables: gender, age, racial/ethnic group, body mass index(BMI), blood group, brain death etiology, arterial hypertension(AH), diabetes mellitus (DM), smoking, chronic obstructive pul-monary disease, infection, alcoholism, cardiorespiratory arrest,intensive care unit stay, norepinephrine administration, as well asmeasurements of urea, creatinine, sodium, hemoglobin, hemato-crit, lactate, blood glucose, amylase, serum glutamic-oxalocetictransaminase (SGOT), serum glutamic-pyruvic transaminase(SGPT), alkaline phosphatase (ALKP), gama-glutamyl transpepti-dase (GGT), lactic acid dehydrogenase (LDH), bilirubin, andcreatine phosphokinase-MB (CKMB). The demographic, clinical,and laboratory characteristics of the population are described inTable 1.

Statistical Analysis

Descriptive analysis was performed after the confirmation of theconsistency of the collected data. Categorical variables were pre-sented as absolute values and percentages; continuous variableswere presented as minimum and maximum values with calculatedaverages and standard deviations. For univariate analysis we ap-plied the chi-square test to compare distributions, and Student ttest for 2 independent samples to compare continuous variables.We also used the nonparametric Mann-Whitney test and log-ranktest. The possible association of donor characteristics was alsoanalyzed with regard to the primary endpoint of hospital death.The univariate analysis utilized odds ratios (OR) to measure thestrength of associations between each exposure and the endpointalone. We calculated 95% confidence levels for each OR. After-ward, a multivariate analysis of a logistic regression model was usedto examine binary categorical variables endpoint (variables with yesor no only, as death). To construct comparisons we initially selectedunivariate analysis data that showed values of P � .05 for the testedhypothesis, the cutoff value for data selection in the multipleregression model construction. Datum entry order was made withincreasing P values. The ROC (receiver operating characteristic)curve representing the probability of patient survival regardless ofthe follow-up was obtained from the logistic regression model.Subsequently, 2 survival curves (Kaplan-Meier) were built fordonors younger than versus older than 40 years. The mortalityendpoint was considered in the period of hospitalization. Statisticalanalysis was performed using Stata 9.0 (StataCorp LP, USA); the

level of significance was set at P � .05. w

RESULTS

The hospital mortality rate was 27.9% (n � 143). Theaverage follow-up was 29.4 � 28.4 months (range, 0–84.6).At 6 years after HTx the survival rate was 55.5% (n � 285;Fig 1). Statistical analysis of studied variables is described inTable 1 The multivariate analysis showed only donor age tobe significant; the logistic regression evaluation indicated asignificant cut off at 40 years. Donors older than 40 years ofage were associated with lower HTx recipient late survivals(P � .0032; Fig 1 and Table 1). Because of incompletenformation we could not evaluate the impact of lactatehronic obstructive pulmonary disease.

DISCUSSION

Currently, HTx has been totally incorporated into clinicalpractice as the best method to treat terminal heart failureand normalize hemodynamic status. However, the numberof heart donors will not probably increase in the comingyears, requiring tightening selection criteria for candidateswho will really benefit from HTx.1 Organ availability hasincreased with the use of expanded criteria donors. Thedonor clinical profile has direct implications for early andlate HTx survival due to allograft quality.5,12 Certain char-acteristics present risk factors the objective of this retro-spective cohort evaluation. The donor shortage has stimu-lated multiorgan retrieval, which have not negativelyinterfered with HTx survival.3–5,13 The classical donationcriteria have been enlarged over the years, such as ischemiatime, older donors, or brain death due to stroke. However,the results of these organs have been controversial due tononuniformity of the study protocols.

Previous studies have identified risk factors linked to therecipient or to the donor that may influence HTx survival:the need for pretransplantation mechanical support, ele-vated pulmonary vascular resistance, previous heart opera-tion, retransplantation, older donor age, and prolongedischemic time. Female donors show an adverse effect onearly HTx outcomes due to weight mismatches and later,immunologic phenomena.12,14 In the present analysis, fe-

ale donor hearts and ischemia time did not show signifi-ant impacts on immediate mortality perhaps because theyere always used with stricter criteria. Most of our HTxccurred with ischemia times less than 4 hours.The use of vasoactive agents for a prolonged time or at high

oses is harmful to the myocardium and may be considered to ben exclusion factor because it predisposes to immediate rightentricular failure after HTx. In this series, a qualitative study oforadrenaline pharmacological treatment showed a significantifference only in the univariate analysis, more likely becauseost donors received this agent. Possibly, a donor-stratified

nalysis of noradrenaline doses and time could offer additionalnformation about the strength of this factor on HTx survival.imilar considerations may be applied to systemic AH that onlyhowed a significant difference upon univariate analysis. Exclu-ion of donors with isolated severe or moderate AH associated

ith other morbidities perhaps explains these findings.

stiuo

cosa

nhdmr

G

WBR

NAABBCSSGCHN

SURVIVAL AFTER HTX 2471

Previous studies have shown difficulty to predict HTxpatient outcomes when donor or recipient selection onlyconsidered parameters outlined in the guidelines. Tsai et alanalyzed a single-institution large retrospective cohort ofhomogenous patients wherein they discarded high-risk re-cipients or donors.15 Despite this selection, the studyhowed hospital-mortality risk factors of higher preopera-ive pulmonary vascular resistance, pretransplantation renalnsufficiency, pretransplantation intra-aortic balloon pumpse, and female donors. Recipient age older than 60 years

Table 1. The Demographic, Clinical, and Laboratory Profile ofthe Variab

Variable/Description No. (%) or Range* M

Age Years 1 ↔ 62 29.8�40 y 273 (53.3%) 23.2

1 ↔ 3940–50 y 127 (24.8%) 45.2

40 ↔ 50�50 y 112 (21.9%) 57.5

51 ↔ 57ender Male 367 (71.7%)

Female 145 (28.3%)eight kg 11 ↔ 120 71.3MI kg/m2 11 ↔ 30 24.5ace‡ — —

ICU days d 1 ↔ 30 5.0Blood group† O 320 (62.5%)

A 156 (30.5%)B 34 (6.6%)AB 2 (0.4%)

Cause of death TBI 306 (59.7%)HS 155 (30.3%)IS 19 (3.7%)Others 32 (6.3%)

AH — 74 (14.5%)Alcoholism — 64 (12.5%)Cardiac arrest — 53 (10.3%)DM — 40 (7.8%)Infection§ — 134 (26.2%)

orepinephrine — 301 (58.7%)LKP IU/L 35 ↔ 323 138.8mylase IU/L 10 ↔ 2398 160.6ilirubin mg/dL 0.1 ↔ 7.5 0.7lood glucose mg/dL 11 ↔ 516 163.9KMB IU/L 1.0 ↔ 1209 67.3GPT IU/L 5.0 ↔ 4000 82.8GOT IU/L 5.0 ↔ 3743 118.3GT IU/L 3.0 ↔ 600.0 78.1reatinine mg/dL 0.5 ↔ 8.1 2.5ematocrit % 19.1 ↔ 41.1 34.3a� mEq/L 108 ↔ 200 153.2

Abbreviations: HS, hemorrhagic stroke; CI, confidence interval; ICU, intensivetraumatic brain injury.

*(↔) Range with minimum value and maximum.†The criterion for inclusion of the multivariate analysis was univariate analys‡The races were tested but do not show significant differences (P � .05).§Nonsystemic infection.

f age that was present in 35% of patients was not e

onsidered to be a risk factor.11,12 This practice was unlikeur investigation that evaluated all used donors during thetudy period seeking to understand the real world situationnd the impact of donor variables on HTx mortality.

Controlled studies that have excluded nonstandard do-ors seeking to simulate ideal conditions have reported aospital mortality rate of 30% with right ventricular failureue to graft primary dysfunction without rejection as theajor cause of death. The immediate mortality was not

elated exclusively to older donors, or to prolonged isch-

12 Donors Studied and Univariate and Multivariate Analysis ofxamined

P

CI ORUnivariate Multivariate†

.4 — —

.1 .0004 .0560 1.02–1.30 1.08

.0040 .0032 0.78–4.27 2.79

.1100 — — —

.1280 — 1.02–1.28 1.02—

.0 .0868 — 1.02–1.16 1.16.2140 — 1.2–1.34 0.98.0960 — — —.2735 — 1.46–2.34 0.88.0830 — 0.86–0.91 0.96

—— —— —

.3478 — 0.80–0.98 1.02—

— —— —

.0462 .5470 1.02–1.06 1.02

.3970 — 1.12–1.56 1.22

.5450 — 1.23–1.50 0.88

.6330 — 1.34–1.52 0.87

.8800 — 1.66–2.34 0.98

.0450 .2030 1.02–1.88 2.892.6 .7225 — 2.18–3.67 1.286.9 .9452 — 2.89–5.16 1.08

.9709 — 3.09–4.76 1.52.4 .9112 — 2.99–4.16 1.98.9 .7631 — 1.89–3.34 1.323.2 .0336 .5600 1.90–2.86 1.261.4 .6825 — 2.02–4.10 1.024.9 .4049 — 1.89–2.78 1.23

.7225 — 2.90–6.56 1.88.9 .1090 — 3.12–4.56 1.28.4 .2550 — 2.98–3.66 1.48

unit; IS, ischemic stroke; M, mean; OR, odds ratio; SD, standard deviation; TBI,

P � .5.

the 5les E

� SD

� 12� 10

� 3.1

� 6.0

—

� 16� 4.0—� 5.1——————————————� 10� 24� 0.8� 70� 96� 21� 24� 10� 1.9� 10� 15

care

is with

mic time suggesting that other factors play important roles

aa3esiHmyv1ain

2472 FIORELLI, BRANCO, DINKHUYSEN ET AL

such as elevated pulmonary vascular resistance. In thepresent open study, we included all used donors showingdonor age to singly be a death predictive factor, principallyin the first year after the HTx. However, after this periodthe survival curves became parallel in relation to thegeneral group, indicating that the age factor lost signifi-cance over time. Our findings are consistent with thosereported by the International Society for Heart and LungTransplantation.12 The International Registry has shownlso every decade of the donor are older. In 1980 theverage age was about 25 years and in 2007 it increased to3 years, demonstrating the worldwide trend toward donorxpansion.12 However, a more refined analysis of these datahowed that the continuous variables, such as age andschemia time, were risk factors for 1-year mortality after

Tx. Donor age is a relatively significant risk for 1-yearortality, which increases exponentially above 40 or 45

ears old. Ischemia time and donor age are continuousariables that remain as risk factors for 1-, 5-, 10-, and5-year mortality rates, independent of the recipientge.14,16 The association of an older donor, a prolongedschemia time, and myocardial hypertrophy is a bad combi-ation for survival after HTx.14,16

In conclusion, donor age older than 40 years representedan important risk factor for survival after HTx. Neitherdonor gender nor norepinephrine use affected early survivalnegatively.

ACKNOWLEDGMENT

We acknowledge the Transplantation Department of the HealthState Secretary of Sao Paulo, Brazil.

REFERENCES

1. Antunes PE, Prieto D, Eugénio L, et al: Donor mitral valverepair in cardiac transplantation. J Thorac Cardiovasc Surg 129:227, 2005

2. Pêgo-Fernandes PM, Samano MN, Fiorelli AI, et al: Recom-

Fig 1. Actuarial survival curveafter HTx according to donor ageand general group (512 cases).The estimated probability of sur-vival at 7 years is 57% in thegroup with �40 years and 41% inthe group with �40 years.

mendations for the use of extended criteria donors in lung trans-plantation. Transplant Proc 43:216, 2011

3. Fiorelli AI, Stolf NA, Pego-Fernandes PM, et al: Recommen-dations for use of marginal donors in heart transplantation: Bra-zilian Association of Organs Transplantation guideline. TransplantProc 43:211, 2011

4. Laks H, Marelli D, Fonarow GC, et al: Use of two recipientlists for adults requiring heart transplantation. J Thorac CardiovascSurg 125:49, 2003

5. Darracott-Cankovic S, Stovin PGI, Wheeldon D, et al: Effectof donor heart damage on survival after transplantation. EurJ Cardiothorac Surg 3:525, 1989

6. Baroldi G, Pasquale GD, Silver MD, et al: Type and extent ofmyocardial injury related to brain damage and its significance inheart transplantation: a morphometric study. J Heart Lung Trans-plant 16:994, 1997

7. Salim A, Velmahos GC, Brown C, et al: Aggressive organdonor management significantly increases the number of organsavailable for transplantation. J Trauma 58:991, 2005

8. Hornby K, Ross H, Keshavjee S, et al: Non-utilization ofhearts and lungs after consent for donation: Canadian multicentrestudy. Can J Anesth 53:831, 2006

9. Moraes BN, Bacal F, Teixeira MC, et al: Behavior profile offamily members of donors and nondonors of organs. TransplantProc 41:799, 2009

10. Boin IF, Kajikawa P, Palmiero HO, et al: Profile of cadavericliver donors of the OPO-UNICAMP from 2002 to 2006. TransplantProc 40:657, 2008

11. Lietz K, John R, Mancini DM, et al: Outcomes in cardiactransplant recipients using allografts from older donors versusmortality on the transplant waiting list. Implications for donorselection criteria. JACC 43:1553, 2004

12. Aurora P, Edwards LB, Kucheryavaya AY, et al: TheRegistry of the International Society for Heart and Lung Trans-plantation: thirteenth official pediatric lung and heart-lung trans-plantation report-2010. J Heart Lung Transplant 29:1129, 2010

13. Smits JM, Vanhaecke J, Haverich A, et al: Three-yearsurvival rates for all consecutive heart-only and lung-only trans-plants performed in Eurotransplant, 1997–1999. Clin Transpl 89,2003

14. Loebe M, Potapov EV, Hummel M, et al: Medium-termresults of heart transplantation using older donor organs. J HeartLung Transplant 19:957, 2000

15. Tsai FC, Marelli D, Bresson J, et al: Use of hearts trans-planted from donors with atraumatic intracranial bleeds. J HeartLung Transplant 21:623, 2002

16. Pedotti P, Mattucci DA, Gabbrielli F, et al: Analysis of the

complex effect of donor’s age on survival of subjects who under-went heart transplantation. Transplantation 80:1026, 2005