Humoral rejection in cardiac transplantation: risk factors, hemodynamic consequences and relationship to transplant coronary artery disease

CARDIAC REJECTION

Humoral Rejection in CardiacTransplantation: Risk Factors,Hemodynamic Consequences andRelationship to Transplant CoronaryArtery DiseasePaul J. Michaels, MD,a Maria L. Espejo, BA,b Jon Kobashigawa, MD,b

Juan C. Alejos,c Caron Burch, RN,c Steve Takemoto, PhD,d

Elaine F. Reed, PhD,d and Michael C. Fishbein, MDa

Background: Acute cellular rejection is the mechanism of most immune-related injuryin cardiac transplant recipients. However, antibody-mediated humoral rejection (HR)has also been implicated as an important clinical entity following orthotopic hearttransplantation. Humoral rejection has been reported to play a role in graft dysfunctionin the early post-transplant period, and to be a risk factor for the development oftransplant coronary artery disease. Some involved in transplantation pathology doubtthe existence of clinically significant humoral rejection in cardiac allografts. Those whorecognize its existence disagree on its possible role in graft dysfunction or graftcoronary artery disease. In this study, we report clinical features of patients with thepathologic diagnosis of HR at our institution since July 1997, when we began systematicsurveillance for humoral rejection.

Methods: We reviewed medical records of patients with the pathologic diagnosis of HRwithout concurrent cellular rejection between July 1997 and January 2001. Diagnosiswas based on routine histology (“swollen cells” distending capillaries, interstitial edemaand hemorrhage) and immunofluorescence (capillary deposition of immunoglobulin andcomplement with HLA-DR positivity), or immunoperoxidase staining of paraffin-embedded tissue (numerous CD68-positive macrophages and fewer swollen endothelialcells distending capillaries).

Results: A total of 44 patients (4 to 74 years old) showed evidence of HR withoutconcurrent cellular rejection at autopsy or on one or more biopsies. Although femalescomprised only 26% of our transplant population, 23 patients (52%) with HR werefemale. A positive peri-operative flow cytometry T-cell crossmatch was observed in 32%of HR patients compared with 12% of controls (p � 0.02). Hemodynamic compromiseconsisting of shock, hypotension, decreased cardiac output/index and/or a rise in

From the Divisions of aAnatomic Pathology, bCardiology andcPediatric Cardiology and the dImmunogenetics Center, Uni-versity of California at Los Angeles, Los Angeles, California.

Submitted January 8, 2002; revised February 1, 2002; acceptedMay 22, 2002.

Reprint requests: Michael C. Fishbein, MD, Department ofPathology and Laboratory Medicine, A7-149 CHS, University

of California at Los Angeles, 10833 Le Conte Avenue, LosAngeles, California 90095-1732. Telephone: 310-825-9731.Fax: 310-794-4161. E-mail: [email protected]

Copyright © 2003 by the International Society for Heart andLung Transplantation.

1053-2498/03/$–see front matter PII S1053-2498(02)00472-2

58

capillary wedge or pulmonary artery pressure was observed in 47% of patients at thetime of diagnosis of HR. Six patients (5 females) died (14% mortality) with evidence ofHR at or just before autopsy, 6 days to 16 months after transplantation. The incidenceof transplant coronary artery disease was 10% greater at 1 year, and 36% greater at 5years, in patients with HR when compared with non-HR patients.

Conclusion: Humoral rejection was associated with acute hemodynamic compromise in47% of patients, and was the direct cause of death in 6 patients (13%). Humoralrejection is a clinicopathologic entity with a high incidence in women and is associatedwith acute hemodynamic compromise, accelerated transplant coronary artery diseaseand death. J Heart Lung Transplant 2003;22:58–69.

The role of the humoral arm of the immunesystem in acute cardiac allograft rejection has beenreported to be a major cause of morbidity andmortality.1–9 Humoral rejection (HR) has long beenacknowledged as a significant cause of allograftdysfunction in kidney transplant patients,10 and hy-peracute cardiac rejection.11 Antibody-mediated re-jection is the major obstacle to success in xenotrans-plantation.

Humoral rejection is often reported to be de-tected early after transplantation and has beenlinked to hemodynamic compromise, graft arteri-opathy and subsequent poor survival.9 In fact, someinvestigators have suggested that patients diagnosedwith HR experience increased hemodynamic com-promise compared with patients who have histologicevidence of acute cellular rejection.5 Patients be-lieved to be at an increased risk of developingHR include females,5 those with an elevated panel-reactive antibodies (PRA),12,13 cytomegalovirus(CMV) seropositivitiy,14–16 a positive cross-match12,13,17 and those with prior sensitization toOKT3.18 It is likely that the manifestations of HRrepresent the consequences of antibody-inducedand complement-mediated activation of endothelialcells, secretion of cytokines, increased endothelialcell adherence of leukocytes and subsequent isch-emic damage to the graft.3,19 Immunofluorescentstaining in HR reveals deposition of immunoglobu-lin and complement within capillaries and upregu-lation of MHC Class II antigens (HLA-DR) withevidence of significant interstitial fibrin depositionsaid to be related to serum extravasation secondaryto capillary leakage.4 In addition to the immuno-globulin staining seen in myocardial tissue, thepresence of antibodies in the serum also appears tobe related to histologic evidence of HR.20 In renaltransplantation, the appearance of anti-endothelialcell antibodies has been shown to be a major riskfactor for the development of humoral rejection and

progression to graft failure shortly after transplan-tation.10 Along similar lines, a significant correlationbetween serum anti-endothelial cell antibody posi-tivity and HR detected on endomyocardial biopsyhas also been demonstrated.21,22 Mechanisms otherthan anti-endothelial cell antibody generation hy-pothesized to be involved in HR of heart transplantrecipients include immune-complex deposition, an-ti-HLA cytotoxic antibodies21 and antibodies di-rected against mouse monoclonal OKT3.18,20,22 Inaddition to their role in humoral rejection, thesealloantibodies are also thought to play a major rolein transplant coronary artery disease (TCAD). In-deed, recent data have shown that ligation of HLAClass I molecules with anti-HLA antibodies resultsin increased tyrosine phosphorylation of intracellu-lar proteins and induction of fibroblast growth factorreceptor expression on endothelium and smoothmuscle cells.23 These findings support a role foranti-HLA antibodies in the transduction of prolifer-ative signals, which may stimulate development ofmyointimal hyperplasia associated with chronic re-jection of allografts.24

Despite the increasing recognition of humoralrejection by many groups,1–9,12 there is still debateregarding the existence, etiology, incidence and clin-ical significance of this entity. Even those whorecognize the entity, often do not monitor for it, orattempt to establish or confirm the diagnosis bypathologic studies. In this report we retrospectivelyevaluated patients with a pathologic diagnosis ofHR with respect to their demographics, crossmatchprofile, hemodynamic status and subsequent devel-opment of TCAD.

MATERIALS AND METHODSPatients and Tissues Studied

The pathologic diagnosis of humoral rejection (HR)was made in a total of 116 endomyocardial biopsy(EMB) specimens from 56 patients between July

The Journal of Heart and Lung Transplantation Michaels et al. 59Volume 22, Number 1

1997 and January 2001. During this period, approx-imately 600 cardiac transplant patients were beingfollowed with surveillance biopsies and coronaryangiography at our institution. Of these patientswith HR, 44 (77 biopsies) were diagnosed with HRwithout concurrent acute cellular rejection (ISHLTGrade 0). Individual EMB specimens were obtainedfrom as early as 2 days post-transplant up to as longas 9 years following cardiac transplantation.

A group of 73 patients from the adult hearttransplantation program were selected as concur-rent controls. All control patients were hemodynam-ically stable, had few to no episodes of prior acutecellular rejection, and never had a diagnosis of HR.The study and control patients were closely matchedfor age, gender and elapsed time from transplanta-tion. The two groups were compared on the basis ofpercentage of individuals with a positive flow cytom-etry T-cell crossmatch and presence of documentedtransplant coronary artery disease either on angio-gram or at autopsy.

In our laboratory, following right ventricularEMB, a portion of the cardiac tissue is immediatelyput into Bayley fixative, processed routinely, andembedded in paraffin. Paraffin blocks are seriallycut into 3-�m-thick sections. Three slides with mul-tiple sections each are subsequently stained withhematoxylin and eosin (H&E). A small portion ofthe EMB is wrapped in saline-soaked gauze andfresh frozen in chilled isopentane at �70°C in OCTcompound (Tissue-Tek OCT compound, Miles, Inc,Elkhart, IN) for later immunofluorescence studies.Those antibodies routinely used for our immunoflu-orescence studies include immunoglobulins M, Gand A (IgM, IgG, IgA) complements 3 and 1q (C3,C1q), fibrinogen and HLA-DR. Immunofluores-cence studies were performed routinely on speci-mens from the first two biopsies, usually 1 and 2weeks after transplantation. If HR was present,immunofluorescence was performed on tissuestaken from succeeding biopsies until the tissue wasnegative for HR. If specimens from a later biopsyshowed histologic evidence of HR and if frozentissues were available, then immunofluorescencewas performed. If not, we performed immunohisto-chemical studies on fixed, paraffin-embedded sec-tions to further evaluate the swollen cells observedwithin the myocardial microvasculature. We usedantibodies to endothelial cells (anti-factor VIII–related antigen and/or anti-CD34) and macrophages(CD68) using a standard labeled strepavidin–biotinhorseradish immunoperoxidase technique. The im-munoperoxidase studies were also done if patients

had unexplained hemodynamic compromise at thetime of biopsy or autopsy.

Criteria for Humoral Rejection

To diagnose humoral rejection, we evaluated EMBsfor histologic evidence of so-called “capillary endo-thelial cell swelling,” interstitial edema, interstitialhemorrhage and neutrophil infiltration8,9 (Figure 1).In addition, tissue was examined using immunoflu-orescence for the detection of IgG, IgM and IgAand C1q and C3 deposition in capillaries in a linearpattern or for immunoperoxidase evidence ofCD68-positive cells distending capillaries (Figure 1).If frozen tissues were not available the finding ofnumerous CD68-positive macrophages within capil-laries in the absence of T-cell infiltrates in the tissuewas considered diagnostic of HR (Figures 2 and 3).As we and Caple et al have reported, the majority ofswollen cells within the microvasculature in HR areactually macrophages and not endothelial cells.7,8

We graded HR based on the degree of “cell swell-ing,” interstitial edema and interstitial hemorrhage(0 to �3).

Pre-Transplant HLA Antibody Screening

Pre-transplant HLA antibody screens were per-formed using the NIH cytotoxicity assay when pa-tients were listed for a transplant. Antibody reactiv-ity was measured as a percentage of the panelreactive to the patient’s serum (PRA). Samples withPRA �10% were re-tested with dithiothreotol toremove IgM antibody.

Serum samples collected at the time of transplan-tation were crossmatched directly with the donor’s Tlymphocytes using the two-color flow cytometrycrossmatch technique. Donor lymphocytes (0.2 to0.4 � 106/tube) were incubated with 25 �l of recip-ient serum or negative control serum for 30 minutesat room temperature. Following the incubation, thecells were washed three times in phosphate-bufferedsaline (PBS) containing 5% fetal calf serum (FCS)and 0.1% sodium azide, and then resuspended in 20�l of pre-titered flourescein isothiocyanate (FITC)-conjugated F(ab')2 Fc-fragment-specific, goat anti-human IgG (Jackson Immunoresearch Laborato-ries, West Grove, PA), and 10 �l of phycoerythrin-conjugated anti-CD3 monoclonal antibody (BectonDickinson). After 30 minutes at 4°C, the cells werewashed three times and suspended in 0.1 ml of PBSand analyzed by FACScan (Becton Dickinson).Electronic gates were set to include only viablelymphocytes. During the period prior to 1994, me-dian fluorescence intensity, displayed on a 256-

60 Michaels et al. The Journal of Heart and Lung TransplantationJanuary 2003

FIGURE 1 (A) Cross-section of a heart from a patient who died of humoral rejection(HR). There is diffuse myocardial hemorrhage within both the left and right ventricularwalls. (B, C) Microscopic sections show intravascular macrophages and neutrophils. Thereis congestion of capillaries with focal interstitial hemorrhage (original magnification �33).(D) Immunofluorescence studies show capillary positivity (green fluorescence) for IgM (d),C1q (e) and HLA-DR (f). Yellow fluorescence represents lipofucsin granules in myocytes(original magnification �120).


FIGURE 2 (A) Hematoxylin–eosin stain of myocardium in a patient with HR showing“swollen cells” (macrophages) within the capillaries (arrows). (B) Immunohistochemicalstaining for CD68 highlights the macrophages within the capillaries. (C)Immunohistochemical staining for CD34 highlights the endothelium of the capillaries,which are distended by macrophages (original magnification �66).

FIGURE 3 (A) Hematoxylin–eosin staining in a patient with HR showing very littleobvious inflammation at low power. (B) An immunohistochemical stain for CD3-positive Tlymphocytes shows very little staining. (C, D) Low and high power of this biopsy followingimmunohistochemical staining for CD68 shows significant macrophage infiltrate within thecapillaries.


channel scale, was measured. Positive T-cell flowcytometry crossmatch results were defined as me-dian channel shift values of �20. During the period1994 to 2001, median fluorescence intensity, dis-played on a 1024-channel, four-decade log scale wasmeasured. Positive T-cell flow cytometry cross-matches results were defined as median channelshift values of �50.

Definition of Allograft Dysfunction

HR-associated allograft dysfunction was diagnosedin heart transplant recipients who had a �30%increase in pulmonary arterial wedge pressure and a�30% decrease in cardiac index when comparedwith previous right heart catheterization measure-ments during timepoints at which no evidence ofrejection (cellular or humoral) was seen on EMB.25

Coronary Angiography

All patients underwent coronary angiography annu-ally following heart transplantation. TCAD wasdefined as any angiographic coronary lesion with�20% luminal stenosis or occlusive TCAD con-firmed at autopsy.

Immunosuppressive Therapy

Between July 1997 and January 2001, routine immu-nosuppression consisted of cyclosporine, corticoste-

roids and azathioprine or mycophenolate mofetil(routine after April 1999). No cytolytic inductiontherapy was used. After 6 months, for most patients(those with little or no rejection), corticosteroidswere weaned off. Asymptomatic HR was nottreated. Those HR patients with a �10% fall inechocardiographic left ventricular ejection fractionreceived high-dose oral corticosteroids. In addition,those patients on azathioprine were switched tomycophenolate mofetil. Depending on availability,some HR patients were treated with gammaglobulinin association with high-dose corticosteroids. Forthose with hemodynamic compromise, part or all ofthe following combination therapy was adminis-tered: intravenous (IV) solumedrol (500 mg fourtimes daily for 3 days); cyclophosphamide (1 to 3mg/kg IV four times daily); plasmapheresis (fourtimes daily for 5 days), OKT3 (5 mg four times dailyfor 14 days given immediately after each plasma-pheresis); and IV heparinization for 7 days.

Statistical Methods

The patients diagnosed with HR were comparedwith control patients using the chi-square test (Ta-bles I and II). Transplant coronary artery–free sur-vival analysis (Figure 4) was performed by Kaplan–Meier product-limit estimates with the end-pointbeing either a documented diagnosis of transplant

TABLE I Comparison of patients with pathologic diagnosis of HR and matched controls

HR patientsMatched control patients

without HR p value

Total patients 44 73Female 23/44 (52%) 35/68 (49%) p � 0.70PRA � 10% 4/33 (12%) 2/65 (3%) p � 0.08Positive crossmatch 12/37 (32%) 8/64 (12%) p � 0.02TCAD at 1 year from OHT 4/27 (15%) 3/61 (5%) p � 0.09TCAD at 3 years from OHT 5/15 (33%) 1/37 (3%) p � 0.001TCAD �5 years from OHT 6/7 (86%) 5/23 (22%) p � 0.001

HR, humoral rejection; OHT, orthotopic heart transplantation; PRA, panel-reactive antibodies; TCAD, transplant coronary arterydisease.

TABLE II Comparison of patients with pathologic diagnosis of HR and all UCLA patients transplanted atthis institution from July 1997 until January 2001

HR patientsAll UCLA patients without HR

transplanted since 7/97 p value

Total patients 44 298Female 23/44 (52%) 78/298 (26%) p � 0.001History of prior OHT failure 2/44 (4.5%) 1/298 (0.3%) p � 0.005

See Table I for abbreviations.


coronary artery disease or patient death. Statisticalsignificance was determined by log-rank compari-sons of survival curves using two-sided p values.Statistical software (STATA; College Station, TX)was used for all statistical analyses.

RESULTSPatient Demographics

A total of 44 patients (4 to 74 years old) showedevidence of HR without concurrent acute cellularrejection. In addition to the findings on H&E-stained histologic sections, the diagnosis was madeby immunofluorescence (n � 7) and/or immunohis-tochemical criteria (n � 37), as indicated. Up toJanuary 2001, a total of 1,090 patients have receivedorthotopic heart transplantation (OHT) at our in-stitution. Although only 78 (26%) of the 298 pa-tients without HR transplanted since July 1997 werefemale, in this study, 23 patients (52%) with HRwere female (p � 0.001) (Table II). Of the patientswith HR transplanted since July 1997 (n � 35), 2(4.5%) had a previous history of re-transplantationsecondary to transplant coronary artery disease.However, only 1 patient (0.3%) of a total of 298non-HR patients transplanted after July 1997 had aprior history of re-transplantation (p � 0.001). Sixpatients died (14% mortality), 5 of whom werefemale. In 3 patients, the first diagnosis of HR wasmade at autopsy. In the other 3, the diagnosis of HRwas made 6 days to 16 months after transplantation.

Baseline immunosuppression was similar in adultswith HR and controls. Primary therapy consisted oftacrolimus rather than cyclosporine for 15% of theHR patients compared with 10% of controls. Aza-thioprine was given as an adjunctive agent in 79% ofpatients in both groups.

Hemodynamic Dysfunction in Humoral RejectionPatients

Hemodynamic dysfunction consisting of shock, hy-potension, decreased cardiac output/index and/or arise in capillary wedge or pulmonary artery pressurewas observed in 47% of patients (20 of 43), at thetime of the pathologic diagnosis of HR. Of thesepatients with hemodynamic dysfunction, 65% werefemale. Of the 21 male patients with HR in thisstudy, only 7 (33%) had hemodynamic compromise,compared with 59% for females, suggesting thatfemales with HR may be more likely to exhibithemodynamic dysfunction than males (p � 0.09).

Hemodynamic dysfunction in HR patients wasanalyzed in relationship to duration since transplan-tation to first pathologic diagnosis of HR (Figure 5).Patients first diagnosed with HR within 1 monthfollowing transplantation had hemodynamic evi-dence of acute allograft dysfunction 68% of thetime, compared with only a 13% frequency ofhemodynamic dysfunction in patients diagnosedwith HR between 1 month and 2 years followingtransplantation (p � 0.001). Also, a significantly

FIGURE 4 Correlation of diagnosis of humoral rejection with freedom of diagnosis oftransplant coronary artery disease (TCAD) following cardiac transplantation.


greater number of patients first diagnosed with HRat least 2 years after transplantation had hemody-namic dysfunction (56%) as compared with individ-uals diagnosed with HR between 1 month and 2years (p � 0.03). There was no correlation of thegrade of HR with hemodynamic dysfunction.

Pre-Transplant HLA Antibody

HLA antibody was detected in 12% (4 of 33) ofpatients diagnosed with HR compared with 3% (2 of65) of those in the control group (Table I, p � 0.08).All 4 of the HR patients with detectable HLAantibody were female, whereas 1 of the 2 non-HRpatient controls was female (p � 0.06).

Peri-operative Crossmatch

Of the 44 patients diagnosed with HR withoutconcurrent acute cellular rejection in this study, 37had available data regarding peri-operative cross-match results (Table I). Twelve (32%) of the 37patients who developed HR had a positive T-cellflow crossmatch compared with only 8 (12%) of the64 control individuals (32% vs 12%; p � 0.02).When the criteria for a positive T-cell flow cross-match was set at a higher threshold, at medianchannel of �100, 10 of the 37 patients with HR hada positive T-cell flow cytometry crossmatch, whereasonly 1 of the 64 controls developed HR (p �0.0001).

Transplant Coronary Artery Disease

Of the patients with HR in this study, 27 had beenfollowed for at least 1 year and have had at least onecoronary angiogram. At our institution pediatricheart transplant patients do not undergo annualscreening angiograms during subsequent follow up,which is why the individuals selected for the controlgroup were taken only from our adult population.Of the 27 patients with HR at 1 year after transplan-tation (Table I), 4 (15%) had angiographic evidenceof TCAD, compared with 3 of the 61 controlpatients at 1 year (15% vs 5%; p � 0.09). Of the 15HR patients followed for at least 3 years, 5 (33%)displayed evidence of TCAD on angiography com-pared with only 1 (3%) of the 37 control patients(33% vs 3%; p � 0.001).

Kaplan–Meier survival curves depicting the per-centage of patients free of TCAD are shown inFigure 4. The incidence of TCAD or death at 1 yearpost-transplant was higher in patients diagnosedwith humoral rejection than in controls (16% vs5%). This difference increased progressively at 3years (36% vs 17%) and 5 years (64% vs 30%; p �0.01) post-transplant.

All patients diagnosed with HR were comparedwith one another on the basis of presence or ab-sence of hemodynamic dysfunction at the time ofHR diagnosis and eventual diagnosis of TCAD.

FIGURE 5 Time after transplant when diagnosis of HR was made and percentage ofpatients with hemodynamic compromise at time of diagnosis.


When evaluated at 1, 2, 3, 4 and 5 years, there wasno difference in the development of TCAD betweenthose HR patients with hemodynamic dysfunctionand those without hemodynamic dysfunction (all pvalues �0.13).

DISCUSSION

The results of our study confirm the original obser-vations of Hammond and associates9 and furtherdelineate the features and sequelae of HR in cardiactransplant recipients. We have document an in-creased incidence of HR in women, patients with ahistory of re-transplantation and those with a posi-tive peri-operative T-cell flow cytometry crossmatch.In addition, patients who develop HR are at anincreased risk for early development of TCAD andhave a relatively high rejection-associated mortality.

Despite our results and the research of otherssupporting the diagnosis of humoral rejection incardiac transplantation, uniform agreement regard-ing the existence and significance of this form ofrejection has not been reached from a histopatho-logic or immunofluorescence standpoint.26 Thoseinvestigators who question the existence of cardiacHR believe many of the histologic and immunoflu-orescent findings are due to events relating toperi-operative graft ischemia and/or OKT3 post-transplant therapy.

Ischemic injury, like humoral rejection, usuallyoccurs in the early post-transplant period, is associ-ated with vascular injury, and can cause allograftdysfunction.1,8,27 Ischemia can also be associatedwith an inflammatory infiltrate, interstitial edemaand hemorrhage. However, in our study, 11 (25%)of the 44 patients with HR had their first histopatho-logic manifestation of rejection diagnosed �1 yearfollowing transplant, making a primary ischemicetiology unlikely in these patients. Also, anotherstudy showed no difference in graft ischemic timewhen comparing patients diagnosed with humoralrejection to those with the histopathologic diagnosisof acute cellular rejection.12 In addition, improve-ment in hemodynamically compromised patientswith evidence of HR on EMB in our study andothers17,27,28 following the use of high-dose intrave-nous corticosteroid therapy, plasma exchange, cyclo-phosphamide and/or intravenous immunoglobulinwould not be expected if these episodes were relatedprimarily to graft ischemia.

Those who question the entity of HR in cardiactransplants also cite literature that has determinedthat many patients diagnosed with humoral rejec-tion have received OKT3 immunoprophylaxis for 10

to 14 days immediately after transplantation. Thesestudies implied that the underlying mechanism ofdamage was caused by immune complexes formed inresponse to OKT3, and not an antigen-specificantibody response.5 Virtually all patients with sen-sitization to OKT3 immunoprophylaxis showed pro-duction of human anti-mouse antibody in the serumand exhibited immunohistologic changes in theirEMBs, which were indistinguishable from thoseseen in humoral rejection.3,18 Although the highincidence of HR previously described by other re-searchers may, in part, be due to prior use ofOKT3,8 the current study has identified a largegroup of patients who carry the histologic diagnosisof HR without a previous history of OKT3 therapy.

Although other smaller studies on HR have notfound a higher incidence of females in such pa-tients,5 we noted a significantly increased percent-age (p � 0.001) of HR patients who were female(52%) compared with the proportion of womenpresent in the total cardiac transplant population atour institution (26%). In addition, interestingly, amajority of the female HR patients (59%) hadconcurrent allograft dysfunction compared with only33% of males with HR (p � 0.09). These findingssuggest that not only are women at an increased riskfor developing HR following transplantation, butthey may be more likely to develop a more clinicallysignificant and severe allograft dysfunction, occa-sionally leading to death. Strikingly, 83% of thoseHR patients who died in this study were women. Alikely explanation for the increased risk of severeHR in women is due to the high antibody titerstypically found in this group secondary to a commonpresence of multiparity, with subsequent predispo-sition to prior high foreign antigen exposure. Recentdata suggest that the development of HR episodes isstrongly correlated with the development of anti-HLA antibodies.20,28,29 Cherry et al found thatanti-HLA antibodies were detected in 90% of serumspecimens drawn at an average of 1.8 days fromEMB specimens that revealed HR.20

In our study, 12 (32%) of 37 patients with avail-able T-cell flow crossmatch data showed peri-oper-ative positivity compared with only 12% positivity incontrol patients (p � 0.02). Other studies have alsofound a high association of positive crossmatch withhistologic evidence of HR,12,13,29,30 again suggestinga strong etiologic role of pre-formed alloantibodiesin HR. In addition to the increased risk for devel-oping acute HR associated with positive IgG lym-phocytotoxic crossmatches, there also exists agreater risk for developing refractory and fatal


rejection.17 Aggressive early intervention withheightened immunosuppression and plasmapheresisgenerally leads to both a lower incidence of HR,earlier resolution of HR and increased survival inthese patients.17,31

The presence of clinically significant allograftdysfunction related to HR has been questioned. Inour study, 47% of the patients with histologic con-firmation of HR had evidence of allograft dysfunc-tion compared with a normal hemodynamic statusnoted at the time of their previous unremarkablebiopsy. Those patients first diagnosed with HRwithin 1 month following transplantation werenoted to have acute allograft dysfunction 68% of thetime compared with only a 13% incidence of hemo-dynamic compromise in patients first diagnosed withHR between 1 month and 2 years following trans-plantation (p � 0.001). This high incidence ofallograft dysfunction in individuals diagnosed withHR soon after transplantation might represent acombination of immune system hyperactivity sec-ondary to donor antigen overload, hemodynamicand fluid instability following transplantation, accli-mation to new immunosuppressive agents, and ahigh antigen-specific antibody response to donorendothelium often noted in the early transplantperiod in patients with HR. HR was often observed5 to 7 days post-operatively. This was associatedwith a gradual rise in donor-specific crossmatch flowcytometry antibodies, supporting an amnestic anti-body response to the donor antigens. Conversely,the relatively low occurrence of allograft dysfunctionbetween 1 month and 2 years after transplant mayrepresent partial resolution of the aforementionedfindings, and/or the development of accommoda-tion.

Accommodation, initially described in the xeno-graft model,32–37 is the process by which a graftsurvives and functions despite the presence of anti-donor antibodies in the recipient’s circulation. Adownmodulation of antigen expression is believed tounderlie the process of accommodation.33 Com-pared with those patients who developed HR be-tween 1 month and 2 years following transplantation(13%), a significantly higher proportion of patientsfirst diagnosed with HR at least 2 years after trans-plantation (56%) had evidence of acute allograftdysfunction (p � 0.03). This increased dysfunctionseen late after transplant may represent the cumu-lative effects of acute vascular injury and chronicvascular impairment from either overt or sub-clini-cal TCAD. Hammond et al3 suggested a gradingsystem that could be applied to the morphologic

evaluation of biopsies from patients with HR. In ourseries, and in our previous studies,8 we did notdetect a pattern of pathologic findings that distin-guished patients with or without hemodynamic com-promise. This may have been due to the complexinteractions between the host and graft, which in-cludes immunologic and non-immunologic factors,immunosuppressive therapy and poorly understoodphenomena such as accommodation.

TCAD is the major cause of late graft failure inpatients following heart transplantation. Anti–en-dothelial-cell antibody levels in the plasma are ele-vated in 80% of allograft recipients who are infectedwith CMV,38 indicating that the elevated anti-endo-thelial cell antibodies may be the pathogenic linkbetween CMV infection and the development ofTCAD with subsequent poor allograft survival.15

Reed et al showed that recipients displaying circu-lating donor alloantigens for �26 weeks followingtransplantation are at an increased risk for TCAD.24

This study has also demonstrated that the frequencyof HR episodes and the production of anti-donorHLA antibodies were associated with an increasedrisk for developing TCAD. Similarly, Dunn et al39

retrospectively investigated anti-endothelial cell an-tibodies produced by cardiac transplant recipientsand showed that peptide-specific anti-endothelialcell antibodies were found in 15 of 21 patients whodeveloped TCAD within 2 years of transplantation,in contrast to only 1 of 20 who did not developTCAD. Recently, using a mouse model renderedtolerant by CD4 antibody injections, Russell et alshowed that mice with an associated immunoglobu-lin deficiency did not develop TCAD.40

Our results indicate that patients who developHR following cardiac transplantation progress toTCAD earlier and at an increased frequency whencompared with control patients. In addition, the vastmajority of these patients (86%) suffer from TCADby 5 years after transplantation. Furthermore, wefound no statistically significant difference in devel-opment of TCAD between HR patients with andwithout hemodynamic dysfunction. This suggeststhat even those patients with HR who have normalgraft function are at increased risk for TCAD. Thefinding of an increased risk for TCAD in patientswith HR is not surprising given that the combina-tion of alloantibody presence, fibrin depositionand endothelial cell activation represents an idealenvironment for the release of growth factorsfrom platelets, monocytes and endothelial cells,which likely then stimulate migration, prolifera-tion and functional modification of medial smooth


muscle cells, leading to progressive vascular inju-ry.1,2 Early detection of HR by means of histo-logic, immunofluorescence and immunoperoxi-dase methods, followed by augmentation ofimmunosuppression, may lead to improved graftsurvival. The best diagnostic criteria and immu-notherapy for HR remain to be determined. Suchprospective studies are currently underway at ourinstitution.

REFERENCES

1. Forester A. Vascular rejection in cardiac transplantation:a morphological study of 25 human cardiac allografts.APMIS 1992;100:367–76.

2. Miller LW. The diagnosis, incidence, and management ofvascular rejection in heart transplantation. J Heart LungTransplant 1993;12(suppl):S111–2.

3. Hammond EH, Hansen JK, Spencer LS, et al. Vascularrejection in cardiac transplantation: histologic, immuno-pathologic, and ultrastructural features. Cardiovasc Pathol1993;2:21–34.

4. Olsen SL, Wagoner LE, Hammond EH, et al. Vascularrejection in heart transplantation: clinical correlation,treatment options, and future considerations. J HeartLung Transplant 1993;12(suppl):S135–42.

5. Miller LW, Wesp A, Jennison SH, et al. Vascular rejectionin heart transplant recipients. J Heart Lung Transplant1993;12(suppl):S147–52.

6. Czerska B, Hobbs RE, James KB, et al. Clinical manifes-tation of acute vascular rejection in cardiac transplantrecipients. J Heart Lung Transplant 1995;14(suppl):S46.

7. Caple JF, McMahon JT, Myles JL, Hook S, Ratliff NB.Acute vascular (humoral) rejection in non-OKT2-treatedcardiac transplants. Cardiovasc Pathol 1995;4:11–8.

8. Lones MA, Czer LS, Trento A, et al. Clinico-pathologicfeatures of humoral rejection in cardiac allografts: a studyin 81 consecutive patients. J Heart Lung Transplant1995;14:151–62.

9. Hammond EH, Yowell RL, Nunoda S, et al. Vascular(humoral) rejection in heart transplantation: pathologicobservations and clinical implications. J Heart Transplant1989;8:430 –43.

10. Yard B, Spruyt-Gerritse M, Claas F, et al. The clinicalsignificance of allospecific antibodies against endothelialcells detected with an antibody-dependent cellular cyto-toxicity assay for vascular rejection and graft loss afterrenal transplantation. Transplantation 1993;55:1287–93.

11. Bonnaud EN, Lewis NP, Masek MA, Billingham ME.Reliability and usefulness of immunofluorescence in hearttransplantation. J Heart Lung Transplant 1995;14:163–71.

12. McCarthy JF, Cook DJ, Smedira NG, et al. Vascularrejection in cardiac transplantation. Transplant Proc 1999;28:160.

13. Loy TS, Bulatao IS, Darkow GVD, et al. Immunostaining ofcardiac biopsy specimens in the diagnosis of acute vascular(humoral) rejection: a control study. J Heart Lung Trans-plant 1993;12:736–40.

14. Hosenpud JD. Coronary artery disease after heart transplan-

tation and its relation to cytomegalovirus. Am Heart J1999;138(suppl):S469–72.

15. Toyoda M, Petrosian A, Jordan SC. Immunological charac-terization of anti-endothelial cell antibodies induced by cy-tomegalovirus infection. Transplantation 1999;68:1311–8.

16. Grattan MT, Moreno-Cabral CE, Starnes VA, et al. Cyto-megalovirus infection is associated with cardiac allograftrejection and atherosclerosis. JAMA 1989;261:3561–6.

17. Ratkovec RM, Hammond EH, O’Connell JB, et al. Outcomeof cardiac transplant recipients with a positive donor-specificcrossmatch-preliminary results with plasmapheresis. Trans-plantation 1992;54:651–5.

18. Hammond E, Wittwer C, Greenwood J, et al. Relationship ofOKT3 sensitization and vascular rejection in cardiac trans-plant patients receiving OKT3 rejection prophylaxis. Trans-plantation 1990;50:776–82.

19. Platt JL, Fischel RJ, Matas AJ, Reif SA, Bolman RM, BachFH. Immunopathology of hyperacute xenograft rejection in aswine-to-primate model. Transplantation 1991;52:214–20.

20. Cherry R, Nielsen H, Reed E, Reemtsma K, Suciu-Foca N,Marboe CC. Vascular (humoral) rejection in human cardiacallograft biopsies: relation to circulating anti-HLA antibod-ies. J Heart Lung Transplant 1992;11:24–30.

21. Leprince P, Fretz C, Dorent R, et al. Posttransplantationcytotoxic immunoglobulin G is associated with a high rate ofacute allograft dysfunctions in heart transplant recipients.Am Heart J 1999;138:586–92.

22. Fredrich R, Toyoda M, Czer LSC, et al. The clinical signifi-cance of antibodies to human vascular endothelial cells aftercardiac transplantation. Transplantation 1999;67:385–91.

23. Bian H, Reed EF. Alloantibody mediated class I signaltransduction in endothelial cells and smooth muscle cells:enhancement by INF-� and TNF-. J Immunol 1999;163:1010–8.

24. Reed EF, Hong B, Ho E, Harris PE, Weinberger J, Suciu-Foca N. Monitoring of soluble HLA alloantigens and anti-HLA antibodies identifies heart allograft recipients at risk oftransplant-associated coronary artery disease. Transplanta-tion 1996;61:566–72.

25. Constanzo-Nordin MR, Heroux AL, Radvany R, Koch D,Robinson JA. Role of humoral immunity in acute cardiacallograft dysfunction. J Heart Lung Transplant 1993;12(sup-pl):S143–6.

26. McManus BM, Winters GL. Pathology of heart allograftrejection. In: Kokbeck PC, Markin RS, McManus BM, eds.Transplant pathology. Chicago: American Society of ClinicalPathology Press, 1994, 197–217.

27. Billingham ME. The postsurgical heart: the pathology ofcardiac transplantation. Am J Cardiovasc Pathol 1988;3:319–34.

28. Jordan SC, Quartel AW, Czer LSC, et al. Posttransplanttherapy using high-dose human immunoglobulin (intrave-nous gammaglobulin) to control acute humoral rejection inrenal and cardiac allograft recipients and potential mecha-nism of action. Transplantation 1998;66:800–5.

29. Greger B, Brossmann T, Gartner H, Hopt U, Lauchart W.Positive post-operative donor-specific crossmatch correlateswith B-cell infiltration and poor graft prognosis. TransplantProc 1990;22:1900–2.

30. Ensley RD, Hammond EH, Renlund DG, et al. Clinicalmanifestations of vascular rejection in cardiac transplanta-tion. Transplant Proc 1991;23:1130–2.


31. Pisani BA, Mullen GM, Malinowska K, et al. Plasmapheresiswith intravenous immunoglobulin G is effective in patientswith elevated panel reactive antibody prior to cardiac trans-plantation. J Heart Lung Transplant 1999;18:701–6.

32. Parker W, Holzknecht ZE, Song A, et al. Fate of antigenxenotransplantation: implications for acute vascular rejectionand accommodation. Am J Pathol 1998;152:829–39.

33. Platt JL, Vercellotti GM, Dalmasso AP, et al. Transplanta-tion of discordant xenografts: a review of progress. ImmunolToday 1990;11:450–6.

34. Magee JC, Collins BH, Harland RC, et al. Immunoglobulinprevents complement mediated hyperacute rejection inswine-to-primate xenotransplantation. J Clin Invest 1995;96:2404–12.

35. Rose AG, Cooper DKC. A histopathologic grading systemfor hyperacute (humoral, antibody-mediated) cardiac xeno-graft and allograft rejection. J Heart Lung Transplant 1996;15:804–17.

36. Holzknecht ZE, Williams JM, Plummer TB, Platt JL. Apo-ptosis in acute vascular rejection of xenografts. TransplantProc 2000;32:973.

37. Holzknecht ZE, Coombes S, Blocher BA, et al. Immunecomplex formation after xenotransplantation: evidence oftype III as well as type II immune reactions provide clues topathophysiology. Am J Pathol 2001;158:627–37.

38. Toyoda M, Galfayan K, Galera OA, Petrosian A, Czer LSC,Jordan SC. Cytomegalovirus infection induces anti-endothe-lial cell antibodies in cardiac and renal allograft recipients.Transplant Immunol 1997;5:104–11.

39. Dunn MJ, Crisp SJ, Rose ML, Taylor PM, Yacoub MH.Anti-endothelial antibodies and coronary artery disease aftercardiac transplantation. Lancet 1992;339:1566–70.

40. Russell PS, Chase CM, Colvin RB. Alloantibody and Tcell-mediated immunity in the pathogenesis of transplantarteriosclerosis: lack of progression to sclerotic lesions in Bcell-deficient mice. Transplantation 1997;64:1531–6.


Humoral rejection in cardiac transplantation: risk factors, hemodynamic consequences and relationship to transplant coronary artery disease

Documents