http://www.jhltonline.org

1053-2498/$ - see fronhttp://dx.doi.org/10.10

E-mail address: jul

Reprint requests:Cardiovascular SurgerGuillou, 35033 Renne299282496.

Comparison of cardiac allograft vasculopathy inheart and heart–lung transplantations: A 15-yearretrospective study

Julien Guihaire, MD,a Olaf Mercier, MD, PhD,b Erwan Flécher, MD, PhD,a

Marie Aymami, MD,a Soly Fattal, MD,b Céline Chabanne, MD,a Francois Leroy Ladurie, MD,b

Bernard Lelong, MD,a Jacques Cerrina, MD,b Thierry Langanay, MD,a Sacha Mussot, MD,b

Dominique Fabre, MD,b Bertrand De Latour, MD,a Hervé Corbineau, MD,a

Jean-Philippe Verhoye, MD, PhD,a Philippe Dartevelle, MD,b Alain Leguerrier, MD,a andElie Fadel, MD, PhDb

From the aDepartment of Thoracic and Cardiovascular Surgery, Pontchaillou University Hospital of Rennes, University of Rennes 1,

Rennes; and the bDepartment of Thoracic and Vascular Surgery and Heart–Lung Transplantation, Marie LannelongueHospital, University of Paris Sud, Le Plessis Robinson, Paris, France.

KEYWORDS:cardiac allograftvasculopathy;heart–lungtransplantation;immune tolerance;long-term outcomes

t matter r 2014 I16/j.healun.2014.0

ien.guihaire@u-ps

Julien Guihaire, My, Pontchaillou Uns, France. Telepho

BACKGROUND: Cardiac allograft vasculopathy (CAV) is a major factor limiting long-term survivalafter heart transplantation (HTx). Specific determinants of CAV and long-term outcome after CAVoccurrence have been poorly investigated after heart–lung transplantation (HLTx).METHODS: Between January 1996 and December 2006, 79 patients underwent HLTx (36.3 � 12.2years old; 47% men) and 141 patients underwent HTx (49.2 � 12.3 years old; 77% men) at twodifferent institutions. CAV grading was reviewed in both groups according to the 2010 standardizednomenclature of the International Society for Heart and Lung Transplantation. The mean post-transplantfollow-up was 94 (1 to 181) months.RESULTS: Overall 10-year survival rate was 58% after HTx and 43% after HLTx (p ¼ 0.11). The Grade1 (or higher) CAV-free survival rate was 95% at 4 years and 69% at 10 years after HLTx, and 77% and39%, respectively, after HTx (p o 0.01). Mean cyclosporine blood levels were similar between thegroups at 3, 6, 12, 24 and 36 months. The main causes of mortality beyond 5 years after HTx and HLTxwere malignancies and bronchiolitis obliterans, respectively. By multivariate analysis, recipients whodeveloped 43 acute myocardial rejections during the first year post-transplant were exposed to a higherrisk of CAV (95% CI 1.065 to 2.33, p ¼ 0.02). Episodes of acute pulmonary rejection and bronchiolitisobliterans were not associated with an increased risk of CAV (p ¼ 0.52 and p ¼ 0.30).CONCLUSION: HLTx recipients appeared protected from CAV compared with HTx patients in thisretrospective study. Repeated acute cardiac rejections were independent predictors of CAV. Unlikebronchiolitis obliterans, CAV had a very low impact on long-term survival after HLTx.J Heart Lung Transplant 2014;33:636–643r 2014 International Society for Heart and Lung Transplantation. All rights reserved.

nternational Society for Heart and Lung2.007

ud.fr

D, Department of Thoracic andiversity Hospital, 2 rue Henri Lene: 0033-299282497. Fax: 0033-

Heart–lung transplantation (HLTx) remains the onlytreatment for end-stage cardiopulmonary disease. Transplantsfor refractory congenital heart disease or pulmonary arterialhypertension currently account for the majority of HLTxs.1

Transplantation. All rights reserved.

Guihaire et al. CAV After HTx 637

The 1-year survival rate after HLTx has progressivelyimproved to 70% in recent years, compared with 64% inthe previous era, whereas the 10-year survival rate remainedstable around 40% over the last 3 decades.2–4 Malignancies,non-cytomegalovirus infections and chronic allograft dys-function are the main causes of morbidity and mortalitybeyond 5 years in HLTx recipients.4 Among these, lateallograft failure may not be the consequence of long-termexposure to immunosuppressive therapies, but instead may beconsidered a manifestation of chronic allograft rejection.

Bronchiolitis obliterans syndrome (BOS) is the mostfrequent form of chronic rejection, affecting 59% of HLTxrecipients by 5 years, with nearly 90% free of coronaryallograft vasculopathy (CAV), the other manifestation ofchronic rejection after HLTx.4 After experimental findingsshowing a progressive fibrous remodeling of the trans-planted vessels in animal models, CAV was first describedin human heart transplants in 1970 by investigators atStanford University.5 This accelerated form of neointimalhyperplasia results in significant narrowing of the coronaryvessels beyond 1 year after heart transplantation (HTx) andis responsible for chronic allograft dysfunction due tomyocardial ischemia.6 Potential causes of CAV after HTxhave included both immunologic determinants, such asacute rejection and human leukocyte antigen mismatches,and non-immunologic factors, such as metabolic disorders,ischemia–reperfusion injury and cytomegalovirus (CMV)infection.6

When compared with HTx recipients, HLTx recipientsappear to be relatively spared from CAV, although BOSdevelops at a rate similar to that seen in double-lungtransplant recipients.2,4,7 Specific determinants of CAV afterHLTx and the mechanisms of its lower frequency comparedwith BOS remain unclear in this population. Therefore, wesought to compare our experience in HLTx to a series ofHTxs over the same period to determine whether lungallografts could provide protection from CAV development.We also aimed to investigate risk factors and the impact ofCAV on long-term outcomes after combined HLTx.

Methods

Data collection

Patients who underwent HLTx at Marie Lannelongue Hospital orisolated HTx at Rennes University Hospital between January 1996and December 2006 were analyzed retrospectively. Donorcharacteristics were reviewed from the national transplant databaseof the "Agence de la Biomédecine" (La Plaine Saint-Denis, France)with regard to age, gender and cardiovascular comorbidity. Patientso18 years old at time of transplantation and those who receivedrepeat transplantations were excluded. A total of 220 patients wereanalyzed according to type of transplant (heart–lung vs isolatedheart) with respect to survival, acute rejection, CMV infection,BOS and CAV rates. The mean post-transplant follow-up was 94(range 1 to 181) months. A third group of double-lung transplant(LTx) recipients from Marie Lannelongue Hospital (transplantedfrom 1996 to 2006, n ¼ 92) was compared with HTx and HLTxgroups in analyses of overall survival, causes of death, and eventsrelated to level of prolonged immunosuppression.

The institutional review boards of the University Hospital ofRennes and Marie Lannelongue Hospital approved the study andwaived the need for informed consent.

Heart and heart–lung procurement andimplantation

Donor assessment included arterial blood gas measurement,transthoracic echocardiography, chest radiography and visualinspection of heart and lungs. Myocardial preservation consistedof a 1-liter infusion of conventional ice-cold hyperkalemiccrystalloid cardioplegia solution into the ascending aorta to achievecomplete diastolic arrest. The heart was then placed in sterile coldRinger’s solution for transport. For heart–lung procurement,cardiac preservation was associated with lung preservation,including an initial injection of 500 mg of prostaglandin E1 intothe pulmonary artery, followed by infusion of 4 liters of modifiedPapworth solution using leukocyte-depleted blood until December2005, after which low-potassium dextran (50 ml/kg; Perfadex;Vitrolife, Goteborg, Sweden) was used for lung preservation.A patent foramen ovale was systematically closed when present.The heart–lung block was then placed in cold saline for transport.

The surgical approach was mid-line sternotomy for all HTxsand in most of HLTxs. All allografts were implanted undercardiopulmonary bypass with mild hypothermia (321C). The HTxand HLTx techniques have been described previously and weremodified toward bicaval anastomosis since 2005.8,9

Immunosuppressive regimen

HTx and HLTx recipients received the same immunosuppressiveregimen following the Stanford protocol.3 Details are reported inthe Supplementary Material (available online at www.jhltonline.org). Briefly, induction therapy was based on rabbit anti-thymocyteglobulin at a dose of 1.5 mg/kg intravenously until Day 6post-transplant.2 Methylprednisolone (1 g) was started between1 and 2 hours before transplantation and the daily dose wasprogressively decreased thereafter to reach the level of 0.5 mg/kg.Maintenance immunosuppression consisted of cyclosporine(3 mg/kg) and azathioprine (2 mg/kg). Starting in 2002,mycophenolate mofetil replaced azathioprine, with a target bloodlevel of between 3 and 8 ng/ml. Acute rejections were treated byinfusion of methylprednisolone, which was combined withthymoglobulin in cases of severe allograft dysfunction. Thereafter,the maintenance immunosuppression regimen was transientlyincreased. In patients with repeated acute rejections, tacrolimusand mycophenolate mofetil replaced cyclosporine and azathio-prine.

Long-term follow-up and CAV definition

Patients were closely followed by the post-transplant teams. Acuterejection of the pulmonary allografts was defined as any episode ofGrade Z2 rejection on transbronchial biopsy. Endomyocardialbiopsies were performed weekly at the beginning (first 6 weeks)and then regularly spaced. Additional biopsies were performedwhen clinically indicated. BOS was defined as a 420% decline inforced expiratory volume in 1 second or in forced expiratory flow,mid-expiratory phase, in the absence of acute rejection or infectionon bronchoscopy. Because most patients in this study didnot benefit from recent technologies to detect post-transplant

Table 1 Pre-operative Characteristics and Risk Factors forCAV Among the 220 Patients Who Underwent Either CombinedHeartLung Transplantation or Isolated Heart Transplantation

Variables

HLTxrecipients(n ¼ 79)

HTxrecipients(n ¼ 141) p-valuea

RecipientsAge (years) 36.3 � 12.2 49.2 � 12.3 o0.01Male gender 37 (47%) 108 (77%) o0.01Creatinineclearance (ml/min)

76 � 23 69 � 23 0.13

BMI (kg/m2) 20.9 � 3.6 24.2 � 3.3 o0.01Smoking history 7 (9%) 59 (42%) o0.01Hypertension 3 (4%) 59 (42%) o0.01Diabetes mellitus 0 (0%) 8 (6%) 0.07Hyperlipidemia 0 (0%) 44 (31%) o0.01

DonorsAge (years) 31.2 � 12.5 34.7 � 13 0.11Male gender 48 (61%) 115 (82%) o0.01Smoking history 25 (32%) 45 (31%) 0.95Hypertension 7 (9%) 8 (6%) 0.53Diabetes mellitus 2 (2%) 0 (0%) 0.13

MismatchesCMVRþ/D� 27 (34%) 29 (20%) 0.06Dþ/R� 25 (32%) 27 (19%) 0.08

GenderR_M/D_F 6 (7%) 12 (8%) 0.80R_F/D_M 16 (20%) 20 (14%) 0.33

Data expressed as mean � standard deviation or as number (%).Creatinine clearance estimated from the Cockroft Gault formulaaccording to age, body mass and gender. BMI, body mass index;CMV, cytomegalovirus; D, donor; F, female; HLTx, heart lungtransplantation; HTx, heart transplantation; M, male; R, recipient.

ap o 0.05, significant difference between groups.

The Journal of Heart and Lung Transplantation, Vol 33, No 6, June 2014638

donor-specific antibodies (DSA), this variable was not included inour analysis of the predictors of CAV.

Heart transplant recipients had cardiac echography and chestradiography at each follow-up visit during the first post-operativeyear and at least every 4 months thereafter. Endomyocardialbiopsies were considered as early as possible in case of clinicalsuspicion of allograft rejection. Acute cardiac rejection (ACR) wasdefined as any episode of GradeZ2 rejection. Asymptomatic acuterejections were treated when severe (Grade 3R), even in theabsence of cardiac allograft dysfunction.10

Conventional coronary angiography (CCA) was first performedin both groups 2 years after transplant and thereafter routinelyrepeated every 2 years or when clinically warranted. Since 2003,thallium scintigraphy, dobutamine stress echocardiography andcomputed tomography coronary angiography have been usedalternatively rather than CCA for monitoring CAV. Each time anon-invasive investigation was positive, a CCA was performedthereafter to confirm (or not) a diagnosis of CAV. Evidence ofcoronary lesions was defined as the presence of any lesions withstenosis 450% on coronary angiography. The severity of CAVwas scored every 2 years according to the ISHLT grading score forCAV.11 Patients not screened by CCA during follow-up wereconsidered CAV 0 when non-invasive tests were negative.

Statistics

Continuous variables are reported as mean � standard deviation(SD) and categorical variables expressed as proportion (%).Unpaired 2-tailed t-test and 1-way analysis of variance were usedto compare continuous variables between groups. Categoricalvariables were compared using the chi-square test. The actuarialsurvival and the freedoms from CAV and BOS were studied usingthe Kaplan–Meier method and all graphs are shown with 95%confidence intervals. The log-rank test was used to compare groupsregarding survival estimates, freedom from BOS and freedom fromCAV. All statistical analyses were performed using STATVIEW V(Abacus Concepts, Inc., Berkeley, CA) and GRAPHPAD PRISM(GraphPad, Inc., La Jolla, CA) software. p o 0.05 was consideredsignificant. To identify independent predictors for the developmentof CAV, Cox multivariate proportional hazards regression methodswere used. After exploratory analysis of all donor and recipientvariables, all factors with p o 0.05 by univariate analysis wereincluded in the multivariate model.

Results

Patients’ demographics

There were 108 male and 33 female HTx recipients with amean age of 49.2 � 12.2 years, and 37 male and 42 femaleHLTx recipients with a mean age of 36.3 � 12.2 years(Table 1).The etiology of cardiopulmonary disease leadingto HTLx was pulmonary arterial hypertension in 54 patients(68%), pulmonary hypertension (PH) associated with leftheart disease in 9 (11%), chronic thromboembolic PH in 6(8%), PH associated with lung disease in 6 (8%) andmiscellaneous in 4 (5%). Indication for isolated HTx wasdilated cardiomyopathy in 66 patients (47%), ischemiccardiomyopathy in 40 (28%) and other causes in 35 (25%).Mean donor age, as well as male gender prevalence, werelower in the HLTx group compared with the HTx group.Mean cold ischemia time of the transplant was not different

between groups (204 minutes for HLTx vs 207 minutes forHTx, p ¼ 0.77). HTx recipients were more exposed tocardiovascular risk factors before transplant, except fordiabetes mellitus. Additional pre-operative characteristics ofthese two populations are presented in Table 1. Indicationsfor LTx were chronic obliterative pulmonary disorder(30%), pulmonary hypertension (27%), primary fibrosis(22%), dilation of the bronchi (9%), leiomyomatosis (4%),bronchiolitis obliterans (4%) and cystic fibrosis (3%), with amean recipient age of 39 � 14 (range 13 to 67) years.

Post-transplant events related to level ofimmunosuppression

HTx and HLTx were exposed to the same level of immuno-suppression regarding absence of a significant differencebetween groups for mean cyclosporine and tacrolimus bloodlevels at 3, 6, 12, 24 and 36 months after transplantation(Table 2). The rate of patients requiring steroid bolus for ACRduring the first year was higher in the HTx group (70% vs 14%,po 0.01), but was similar to the rate in patients treated for acutepulmonary rejection (APR) in the HLTx group (66%). Duringpost-transplant follow-up, the mean number of ACRs per patient

Table 2 Post-transplant Events Related to Level of Immunosuppression.

HLTx (n ¼ 79) HTx (n ¼ 141) LTx (n ¼ 94)

Mean cyclosporine blood levels (mg/liter)3 months 192 � 70 194 � 53 262 � 92b,c

6 months 182 � 76 185 � 57 234 � 92b,c

12 months 157 � 84 155 � 62 196 � 46b,c

24 months 130 � 67 123 � 56 168 � 47b,c

36 months 137 � 72 112 � 40 140 � 49b

Mean tacrolimus blood levels (ng/ml)3 months 8.9 � 3.3 9.4 � 3.5 9.1 � 2.06 months 6.9 � 2.4 9.7 � 3.9 8.9 � 3.012 months 7.9 � 3.7 7.6 � 2.7 7.8 � 2.924 months 8.5 � 2.6 7.0 � 2.9 8.7 � 3.736 months 12.5 � 5.9 5.7 � 1.6 5.2 � 1.5a,c

Acute graft rejection requiring steroid bolusPatients with 41 ACR during the first year 11 (14%) 99 (70%) —

ACR/patient during the follow-up (mean � SD) 0.2 � 0.1 3.0 � 0.4 —

Patients with 41 APR during the first year 52 (66%) — 48 (51%)APR/patient during the follow-up (mean � SD) 1.7 � 0.9 — 1.5 � 1.1

Complications related to immunosuppressionOpportunistic infections 38 (48%) 73 (52%) 52 (55%)CMV disease 31 (39%) 32 (23%) 28 (30%)PTLD 5 (6%) 3 (2%) 3 (3%)

Data expressed as mean � standard deviation or as number (%). ACR, acute cardiac rejection; APR, acute pulmonary rejection; HLTx, heart lungtransplantation; HTx, heart transplantation; PTLD, post-transplant lymphoproliferative disorder. Opportunistic infections refers to any infection due toopportunistic pathogens, such as Aspergillus sp, Candida albicans, Toxoplasma gondii, Cryptococcus neoformans, Histoplasma, Mycobacterium tuberculosis,Microsporidium, Streptococcus pyogenes, Pseudomonas aeruginosa, Pneumocystis jirovecii and Legionella pneumophila. CMV disease: symptomatic infectiondue to cytomegalovirus, including hepatitis, encephalitis, pneumonitis and ocular of gastrointestinal manifestations.

ap o 0.05 for HTx vs HLTx.bp o 0.05 for HTx vs LTx.cp o 0.05 for HLTx vs LTx.

Guihaire et al. CAV After HTx 639

was higher in the HTx group (3.0 � 0.4 vs 0.2 � 0.1, p o0.01), and HLTx patients had a mean APR of 1.7 � 0.9. Wealso observed similar rates of opportunistic infections and post-transplant lymphoproliferative disorders between groups (p ¼0.37 and p ¼ 0.06), and a higher risk of CMV infection amongHTx recipients (p ¼ 0.02). When compared with HLTxrecipients, LTx patients shared similar rates of APR andcomplications related to chronic immunosuppression.

Figure 1 Comparison of survival curves between heart–lungtransplant, heart transplant and lung transplant recipients. Overallsurvival rates were not significantly different between the groups.See text for abbreviations.

Survival rates in heart and heart–lung transplantrecipients

The overall survival rates were 69% and 58% at 5 and 10years, respectively, in the HTx group compared with 51%and 43%, respectively, in the HLTx group (p ¼ 0.10;Figure 1). Despite having the lowest median survival, LTxpatients had similar survival rates according to Kaplan–Meier curve analysis (52% at 5 years and 38% at 10 years,p ¼ 0.10; Figure 1). The conditional transplant survivalrates at 5 and 10 years among recipients who survived atleast 1 year post-operatively were 94% and 79% after HTxand 74% and 61% after HLTx, respectively (p ¼ 0.04). Theprimary cause of death among 1-year survivors after HLTxand LTx was BOS, whereas CAV and malignancies werethe main causes of long-term mortality among HTx patients.Two patients died from severe CAV after HLTx (11%) and5 after HTx (21%) (Table 3). Respective 5- and 10-year

survival rates were 80.5% and 76.3% among HLTx patientswithout BOS, and 45.5% and 18.2% in HLTx patients withchronic rejection of the lung (p o 0.01).

Cardiac allograft vasculopathy in heart–lungtransplant and heart transplant recipients

Mean time between transplantation and diagnosis of CAVwas 5.6 � 3.2 years in HTx recipients and 7 � 2.8 years in

Table 3 Causes of Death in HeartLung, Lung and HeartTransplant Patients Who Survived Beyond Post-operative Year 1

Cause of death HLTx (n ¼ 19) HTx (n ¼ 24) LTx (n ¼ 34)

BOS 6 (32%) — 12 (35%)Neoplasm 1 (5%) 7 (29%) 3 (9%)CAV 2 (11%) 5 (21%) —

Sepsis 4 (21%) 4 (17%) 7 (21%)Stroke — 3 (12%) 1 (3%)Acute rejection 1 (5%) 2 (8%) 1 (3%)Other 5 (26%) 3 (12%) 10 (29%)

Data expressed as number (%). BOS, bronchiolitis obliteranssyndrome; CAV, cardiac allograft vasculopathy; HLTx, heart lungtransplantation; HTx, heart transplantation; LTx, lung transplantation.

Figure 2 CAV-free survival from and changes of CAV gradeover the time among heart–lung transplant and heart transplantrecipients. Survival of CAV Grade Z1 was better among HLTxrecipients (A). Similarly, CAV grade after HTx was higher at each2-year follow-up compared with HLTx. CAV severity increasedprogressively over the time after HLTx, suggesting that CAV wasdelayed rather than completely prevented in this group. #p o 0.01(HTx vs HLTx).

The Journal of Heart and Lung Transplantation, Vol 33, No 6, June 2014640

HLTx recipients. Results of the non-invasive modalitiesused as an alternative for CCA to routinely screen for CAVare listed in the Table 4. Two HTx and no HTLx patientswith CAV were missed by routine non-invasive screening.In these 2 cases, CAV was diagnosed by CCA performed at5 and 6 years post-transplant, respectively, because ofcardiac allograft dysfunction without evidence of acuterejection. Grade 1 CAV-free survival rates were 98%, 95%,84%, 74% and 69% respectively, at 2, 4, 6, 8 and 10 yearspost-HLTx, and 84%, 77%, 71%, 52% and 39%, respec-tively post-HTx (p o 0.01; Figure 2A). Similarly, meanCAV grade was lower after HLTx at 2, 4, 6, 8 and 10 years(p o 0.01). Dynamic changes of CAV severity over timeare displayed in Figure 2B. Among HLTx recipients, 5(6.3%) underwent a coronary angioplasty associated withstenting for the treatment of CAV lesions. The averagenumber of treated lesions per patient was 1.4 � 0.2 in thisgroup. Seven HTx recipients (4.9%) underwent repeatcoronary angioplasty for CAV with an average of 2 � 0.3

Table 4 Results of Non-invasive Modalities for CAV Screeningin Heart and HeartLung Transplant Recipients

HTx HLTx

DSEn 113 58Positives 41 (36%) 9 (15%)True positives 32 (78%) 5 (55%)

CCTAn 12 41Positives 5 (42%) 9 (22%)True positives 4 (80%) 6 (67%)

D-thallium SPECTn 19 8Positives 3 (16%) 1 (12%)True positives 3 (100%) 1 (100%)

CCTA, cardiac computed tomography angiography; DSE, dobutaminestress echocardiography; D-thallium SPECT, dobutamine thalliumsingle-photon emission computed tomography; n refers to the overallnumber of tests performed; positives are defined by a regionalmyocardial dysfunction (DSE), or by a luminal narrowing of 470%on coronary branches 41.5 mm in diameter (CCTA), or by an abnormalmyocardial perfusion (SPECT); positive cases confirmed by conventionalinvasive coronary angiography (luminal narrowing 450%) wereconsidered true positives.

treated lesions per patient. Two of them died from CAV-related allograft dysfunction. No HLTx patient was referredfor iterative transplant because of severe CAV.

Risk factors for development of CAV

Univariate analysis identified 43 episodes of ACR over thefirst year post-transplant as a risk factor for the developmentof CAV in both HTx and HLTx recipients (p o 0.01),whereas donor age did not reach statistical significance(p ¼ 0.05). More than 3 ACR episodes in the first year post-transplant was the only significant determinant of CAV bymultivariate analysis (95% confidence interval 1.065 to2.33, p ¼ 0.02). Several other risk factors listed in Table 5were not found to correlate with CAV, such as number ofAPR episodes, diagnosis of BOS and CMV-relatedinfectious events.

Discussion

In the modern era, long-living recipients after solid-organtransplant represents a growing population with a higherproportion developing chronic allograft rejection.12

Previous reports demonstrated that HLTx recipients exhibitBOS at a rate similar to that of double-lung transplant

Table 5 Independent Risk Factors for Development of CardiacAllograft Vasculopathy After HeartLung and Heart Transplanta-tion

Variable p-valuea

ACR during the first year 0.01Donor age 0.05APR during the first year 0.52BOS 0.30

ACR, treated episodes of acute cardiac rejection; APR, treatedepisodes of acute pulmonary rejection; BOS, bronchiolitis obliteranssyndrome.

ap o 0.05 considered significant as risk factor for developmentof CAV.

Guihaire et al. CAV After HTx 641

recipients.7 The present 15-year comparative study is thefirst to investigate determinants and progressive develop-ment of CAV after HTx and combined HLTx. Bycomparing long-term outcomes, we have demonstratedthat HLTx patients have a lower risk of CAV than HTxpatients.

Our study has confirmed previous clinical reports fromStanford showing a low incidence of CAV after HLTx.3,13

Unlike our findings, those studies did not analyzeprogression of coronary lesions over time or the clinicalrole of CAV in long-term survival. We showed that repeatedACR episodes during the first year post-transplant werepredictive of CAV, reflecting that host immunity against thecardiac allograft is an important determinant of endothelialcell injury.14 Although HTx patients shared a higher risk ofACR during the first year in our study, this wascompensated by a similar rate of APRs in the HLTx group.Moreover, the rates of complications related to prolongedimmunosuppression were similar between HTx and HLTxpatients, except for CMV disease. The higher incidence ofCMV disease may be related to the higher rate of CMVmismatch between donors and recipients in the HLTx group(Table 1). These findings underscore that HLTx patientswere not exposed to greater immunosuppression than HTxrecipients, as the treatment for CMV infection requires lessimmunosuppression. As a consequence, the lower rate ofCAV after HLTx may not result from a stronger immuno-suppressive regimen in this group compared with HTxpatients.

Considering the higher CAV-free survival rate and thelower severity of coronary lesions over the time comparedwith HTx recipients, our findings suggest that lung allo-grafts may provide an immunoprotective effect to the heart.The bronchial-associated lymphoid tissue may concentratethe immune system into the lungs. Lung allografts may thusreduce the host response against cardiac allograft aftercombined HLTx.15 This immune tolerance is known as the“combi-effect.” Animal experiments have demonstrated thata high level of donor lymphocytes in one allograft mayimprove the survival of other organs transplanted at thesame time.15 The numbers of complement activationfragments are also lower in the myocardium after combinedtransplantation.16 This lower rate of CAV compared withHTx is not restricted to combined HLTx, as showed by

Raichlin et al, who also reported a higher CAV-free survivalrate after combined heart and kidney transplantation.17

The overall survival rate after transplant was not differentbetween HLTx and HTx in this study. Earlier and moresevere CAV after HTx was not associated with a lowersurvival rate compared with HLTx patients. We hypothesizethat CAV after HLTx may not impact long-term mortality toa similar extent compared with HTx recipients. BOS-relatedmortality may thus compensate for the low impact of CAVon long-term survival after HLTx. As previously reported,double-lung transplantation and HLTx recipients share thesame rate of BOS in our experience.2 In the present study,BOS was the main cause of mortality in both HLTx and LTxrecipients. Because the lungs may act as a positiveimmunologic filter for antibodies and allogeneic responsecells, HLTx recipients are at high risk of both acute andchronic rejection of the lungs. On the other hand, thesequestration of the specific antigenic response against thelung allograft may explain why repeated acute rejections ofthe lung as well as development of BOS were not predictiveof CAV in our study.

As expected, HTx recipients were more exposed tocardiovascular comorbidities before transplantation becauseischemic cardiomyopathy was a main etiology leading toHTx in the present study. Unlike post-transplant cardiovas-cular risk factors, the critical role of pre-transplant como-rbidities in the promotion of CAV has not been clearlydemonstrated.6,18 Moreover, donor comorbidities werecomparable in the 2 groups, suggesting the same rate ofpre-existing coronary lesions transmitted to the recipient,despite a lower rate of male donors in the HLTx group. Thispopulation included mainly young women because pulmo-nary arterial hypertension was the primary cause of HLTx.Regarding post-transplant characteristics, we considered thetwo populations comparable as they did not show asignificant difference in mean cyclosporine blood levelsduring the first 3 years post-Tx.

Despite major advances in immunosuppressive therapyover the last 3 decades, CAV remains a treatment-refractorycomplication after HTx and is the primary cause of chronicallograft failure. We showed a relatively low rate ofcoronary angioplasty among HTx with CAV, comparablewith that reported in the HLTx group. The higher severityand complexity of CAV lesions in HTx patients, charac-terized by multivessel disease with long and circumferentialstenosis, may not always be amenable to treatment withangioplasty. Re-transplantation is currently the only defin-itive treatment in selected patients. Our study supports theneed for further investigations focusing on protective factorsagainst CAV development. The lower incidence andseverity of CAV among HLTx may not only relate to thecombi-effect. The lymphatic drainage of the pericardium tothe mediastinum is preserved in HLTx, whereas thelymphatics that connect the mediastinal lymph nodes tothe thoracic duct are disrupted in both HTx and HLTx.19

The regeneration of lymphatic vessels after LTx or HLTxmay occur earlier than after isolated HTx due to thepresence of endothelial progenitor cells in the transplantedlungs.20 We hypothesize that preserved lymphatic clearance

The Journal of Heart and Lung Transplantation, Vol 33, No 6, June 2014642

of the cardiac allograft may play a role in the prevention ofCAV after HLTx. However, the precise role of thelymphatic circulation in thoracic transplantation has notbeen well described.21–24 Other protective approaches toimprove graft survival may arise from complete examinationof the molecular profile of mild to moderate CAVphenotypes among long-term survivors after HTx.25–27

Limitations of the study

In addition to being a retrospective analysis, our study hasother limitations, including absence of case-matching for age,gender and all other cardiovascular risk factors that couldinterfere with CAV pathogenesis. However, as previouslymentioned, both groups were comparable with regard to donorcharacteristics, and the level of immunosuppression wasequivalent. Moreover, for practical reasons, intravascularultrasound screening of coronary arteries was deliberately notused in our patients. Even if considered as less sensitive, non-invasive investigations were preferred because of a lower riskof complications. This is a major limitation of the studybecause non-invasive modalities may be positive only at anadvanced stage of the disease. Significant coronary lesionsmay thus be missing in patients with negative non-invasivetesting. Although associated with chronic graft failure, post-transplant DSA were excluded from our final analysis becauseof the major heterogeneity in frequency and mode of detectionover time in this study. Nevertheless, the clinical relevance oflow levels of detected DSA remains unknown.28–30

Clinical implications

With improvements in long-term survival after solid-organtransplantation, chronic rejection has become a major causeof morbidity and mortality. Combined HLTx may delayCAV progression when compared with isolated HTx,although BOS remains the main cause of mortality, asshown in the LTx recipients in our study. Our findingshighlight the fact that HLTx patients are more like LTxpatients than HTx patients over long-term follow-up. CAVmay not be considered a major complication after HLTx, butshould always be investigated, especially in long-termsurvivors without BOS. Moreover, major advances haverecently improved the sensitivity and specificity of non-invasive modalities to screen CAV. We suggest that CCAnot be used as a first approach to detect CAV after HLTx.If further investigated by experimental studies and lymphaticimaging technologies in HTx patients, improvement ofcardiac lymph drainage may represent a promising approachin the prevention of CAV.

In this study we have reported long-term follow-up afterHLTx and HTx and showed that CAV is less frequent andless severe in HLTx recipients. Our results also show thatrepeated acute cardiac rejection during the first post-operative year is predictive of CAV development. Immunetolerance of the heart as provided by the lung allografts mayresult in lessening the impact of CAV on graft survivalafter HLTx.

Disclosure statement

The authors have no conflicts of interest to disclose.

Supplementary data

Supplementary data associated with this article can be foundin the online version at www.jhltonline.org/.

References

1. Stehlik J, Edwards LB, Kucheryavaya AY, et al. The registry of theInternational Society for Heart and Lung Transplantation: 29th officialadult heart transplant report—2012. J Heart Lung Transplant2012;31:1052-64.

2. Fadel E, Mercier O, Mussot S, et al. Long-term outcome of double-lung and heart-lung transplantation for pulmonary hypertension:a comparative retrospective study of 219 patients. Eur J CardiothoracSurg 2010;38:277-84.

3. Deuse T, Sista R, Weill D, et al. Review of heart–lung transplantationat Stanford. Ann Thorac Surg 2010;90:329-37.

4. Christie JD, Edwards LB, Kucheryavaya AY, et al. The registry of theInternational Society for Heart and Lung Transplantation: twenty-eighth adult lung and heart–lung transplant report—2011. J Heart LungTransplant 2011;30:1104-22.

5. Bieber CP, Stinson EB, Shumway NE, et al. Cardiac transplantation inman. VII. Cardiac allograft pathology. Circulation 1970;41:753-72.

6. Schmauss D, Weis M. Cardiac allograft vasculopathy: recent develop-ments. Circulation 2008;117:2131-41.

7. Moffatt-Bruce SD, Karamichalis J, Robbins RC, et al. Are heart–lungtransplant recipients protected from developing bronchiolitis obliteranssyndrome? Ann Thorac Surg 2006;81:286-91;discussion 91.

8. Vouhe PR, Dartevelle PG. Heart–lung transplantation. Technicalmodifications that may improve the early outcome. J ThoracCardiovasc Surg 1989;97:906-10.

9. Logeais Y, Lelong B, Langanay T, et al. Results of heart trans-plantation: experience of 233 grafts [in French]. Bull Acad Natl Med2003;187:325-40;discussion 40-3.

10. Costanzo MR, Dipchand A, Starling R, et al. The International Societyfor Heart and Lung Transplantation guidelines for the care of hearttransplant recipients. J Heart Lung Transplant 2010;29:914-56.

11. Mehra MR, Crespo-Leiro MG, Dipchand A, et al. International Societyfor Heart and Lung Transplantation working formulation of astandardized nomenclature for cardiac allograft vasculopathy—2010.J Heart Lung Transplant 2010;29:717-27.

12. Prins K, Masri C, Colvin-Adams M, et al. Characterization of 25þ yearsurvivors of cardiac transplantation. J Heart Lung Transplant 2013;32:560-2.

13. Lim TT, Botas J, Ross H, et al. Are heart–lung transplant recipientsprotected from developing transplant coronary artery disease? A case-matched intracoronary ultrasound study. Circulation 1996;94:1573-7.

14. Zimmer RJ, Lee MS. Transplant coronary artery disease. J Am CollCardiol 2010;3:367-77.

15. Kutlu HM, Sadeghi AM, Norton JE, et al. Effect of simultaneous lungtransplantation on heart transplant survival in rats. J Heart Transplant1987;6:29-33.

16. Kelm C, Adams J, Buhr J, et al. Specificity of "combi-effect" afterxenogenic discordant heart–lung transplantation [in German]. Langen-becks Arch Chir Suppl 1998;115:73-7.

17. Raichlin E, Kushwaha SS, Daly RC, et al. Combined heart and kidneytransplantation provides an excellent survival and decreases risk ofcardiac cellular rejection and coronary allograft vasculopathy. Trans-plant Proc 2011;43:1871-6.

18. Colvin-Adams M, Agnihotri A. Cardiac allograft vasculopathy: currentknowledge and future direction. Clin Transplant 2011;25:175-84.

19. Riquet M, Souilamas R, Hubsch JP, et al. Lymphatic drainage of heart andlungs: comparison between pig and man. Surg Radiol 2000;22:47-50.

Guihaire et al. CAV After HTx 643

20. Schniedermann J, Rennecke M, Buttler K, et al. Mouse lung containsendothelial progenitors with high capacity to form blood and lymphaticvessels. BMC Cell Biol 2010;11:50.

21. Kong XQ, Wang LX, Kong DG. Cardiac lymphatic interruption is amajor cause for allograft failure after cardiac transplantation. LymphatRes Biol 2007;5:45-7.

22. Cui Y. The role of lymphatic vessels in the heart. Pathophysiology2010;17:307-14.

23. Loukas M, Abel N, Tubbs RS, et al. The cardiac lymphatic system.Clin Anat 2011;24:684-91.

24. Riquet M, Le Pimpec-Barthes F, Hidden G. Lymphatic drainage of thepericardium to the mediastinal lymph nodes. Surg Radiol Anat2001;23:317-9.

25. Le Texier L, Thebault P, Lavault A, et al. Long-term allograft toleranceis characterized by the accumulation of B cells exhibiting an inhibitedprofile. Am J Transplant 2011;11:429-38.

26. Wei L, Wang M, Qu X, et al. Differential expression of microRNAsduring allograft rejection. Am J Transplant 2012;12:1113-23.

27. Mehra MR, Uber PA, Benitez RM. Gene-based bio-signature patternsand cardiac allograft rejection. Heart Failure Clin 2010;6:87-92.

28. Hodges AM, Lyster H, McDermott A, et al. Late antibody-mediatedrejection after heart transplantation following the development of denovo donor-specific human leukocyte antigen antibody. Transplanta-tion 2012;93:650-6.

29. Topilsky Y, Gandhi MJ, Hasin T, et al. Donor-specific antibodies toclass II antigens are associated with accelerated cardiac allograftvasculopathy: a three-dimensional volumetric intravascular ultrasoundstudy. Transplantation 2013;95:389-96.

30. Zeevi A, Lunz J, Feingold B, et al. Persistent strong anti-HLA antibodyat high titer is complement binding and associated with increased riskof antibody-mediated rejection in heart transplant recipients. J HeartLung Transplant 2013;32:98-105.