Immunoglobulin M-to-Immunoglobulin G Anti-Human Leukocyte Antigen Class II Antibody Switching in Cardiac Transplant Recipients Is Associated With an Increased Risk of Cellular Rejection

Immunoglobulin M-to-Immunoglobulin G Anti-HumanLeukocyte Antigen Class II Antibody Switching in CardiacTransplant Recipients Is Associated With an Increased Risk

of Cellular Rejection and Coronary Artery DiseaseKatherine Lietz, MD, PhD; Ranjit John, MD; Elizabeth Burke, ANP; Michael Schuster, MS;

Tyson B. Rogers, MS; Nicole Suciu-Foca, PhD; Donna Mancini, MD; Silviu Itescu, MD

Background—Activation of T cells induces immunoglobulin (Ig)M-to-IgG B-cell isotype switching via costimulatoryregulatory pathways. Because rejection of transplanted organs is preceded by alloantigen-dependent T-cell activation,we investigated whether B-cell isotype switching could predict acute cellular rejection and the subsequent developmentof transplantation-related coronary artery disease (TCAD) in cardiac transplant recipients.

Methods and Results—Among 267 nonsensitized heart transplant recipients, switching from IgM to IgG anti-humanleukocyte antigens (HLA) antibodies directed against class II but not against class I antigens was associated with ashorter duration to high-grade rejection, defined as International Society for Heart and Lung Transplantation grade 3Aor higher (P�0.001), a higher cumulative rejection frequency (P�0.002), accelerated development of TCAD (P�0.04),and decreased late survival (P�0.03). Conversely, the persistence of IgM anti-HLA antibodies against class II but notagainst class I antigens for �30 days and the lack of IgG isotype switching were associated with protection against bothacute rejection (P�0.02) and TCAD (P�0.05). Alloisotype switching coincided with T-cell activation, as evidenced byincreased serum levels of soluble CD40 ligand costimulatory molecules. Finally, a case-control study showed thatreduction of cardiac allograft rejection by mycophenolic acid was accompanied by reduced CD40 ligand serum levelsand the prevention of IgM-to-IgG anti-HLA class II antibody switching.

Conclusions—T-cell–dependent B-cell isotype switching and the consequent production of IgG anti-HLA class IIantibodies are strongly correlated with acute cellular rejection, a high incidence of recurrent rejections, TCAD, and poorlong-term survival. Detecting this isotype switch is a clinically useful surrogate marker for in vivo T-cell activation andmay provide a noninvasive approach for monitoring the efficacy of T-cell targeted immunosuppressive therapy in hearttransplant recipients. (Circulation. 2005;112:2468-2476.)

Key Words: antibodies � coronary disease � transplantation

The long-term success of cardiac transplantation is cur-rently limited by the high incidence of transplantation-

related coronary artery disease (TCAD),1 a complicationprincipally related to antecedent episodes of high-gradeallograft rejection1,2 and/or ongoing immune responsesagainst donor human leukocyte antigens (HLAs).3–8 Between42% and 88% of cardiac transplant recipients produce anti-HLA antibodies at some point after transplantation,6,9–11 andwhen these are first detected within the first year aftertransplantation, 5-year heart allograft survival is reducedfrom 91% to 78%.9 This is the result of both an increase inepisodes of acute rejection and the development of acceler-ated TCAD. Adverse outcomes have been predominantlyassociated with the immunoglobulin (Ig) G isotype of anti-

HLA antibodies,12 and switching of isotypes from IgM to IgGalloantibodies has been reported to increase the risk of acuteand chronic renal and liver allograft rejection.13,14

The risk of developing high-grade acute cellular rejectionis greatest during the first 3 months after cardiac transplan-tation, when most alloreactive recipient T-cell clones recog-nize foreign HLA molecules directly.15–18 With adequateimmunosuppression, over time there is a reduction in thefrequency of recipient T-cell clones directly recognizingimmunodominant alloantigenic determinants derived fromdonor HLA class II molecules and a gradual shift to indirectrecognition of multiple, new epitopes from these molecules,ie, presented by antigen-presenting cells such as dendriticcells, macrophages, and B cells.19–21 These processes, termed

Received June 15, 2004; revision received June 14, 2005; accepted July 29, 2005.From the Division of Cardiology (D.M.), Division of Cardiothoracic Surgery (K.L., R.J., E.B., M.S., S.I.) and Department of Pathology (N.S.-F.),

Columbia-Presbyterian Medical Center, New York, NY, and the Division of Cardiology (K.L.), Division of Cardiothoracic Surgery (R.J.), andDepartment of Internal Medicine (T.R.), University of Minnesota Medical Center, Minneapolis, Minn.

Correspondence to Silviu Itescu, MD, Director of Transplantation Immunology, Departments of Medicine and Surgery, Columbia-Presbyterian MedicalCenter, PH14 East, 630W 168th St, New York, NY 10032. E-mail [email protected]

© 2005 American Heart Association, Inc.

Circulation is available at http://www.circulationaha.org DOI: 10.1161/CIRCULATIONAHA.104.485003

2468

Transplantation

by guest on September 16, 2015http://circ.ahajournals.org/Downloaded from

intermolecular and intramolecular spreading, respectively,are postulated to involve close T-cell and B-cell cooperationand to be associated with an increased risk for long-term graftloss.22–24 Consequently, a major objective in cardiac trans-plantation is to develop immunomodulatory therapies thatinduce an early and long-lasting state of T-cell and B-cellimmune nonreactivity, or tolerance, to donor HLAs.

Because IgM-to-IgG isotype switching by B cells is aprocess that is tightly regulated by both T-cell costimulatorysignals and recognition of cognate antigen, in the presentstudy we examined (1) whether there was a correlationbetween B-cell isotype switching from IgM to IgG antibodiesdirected against HLA molecules and the complications ofcellular rejection and accelerated TCAD, (2) whether thedevelopment of isotype switching was correlated with sys-temic markers of T-cell activation, and (3) whether isotypeswitching could be prevented by novel, more potent immu-nosuppressive regimens that reduce T-cell–mediated rejec-tion.25 Our results indicate that isotype switching and theproduction of IgG anti-HLA class II antibodies are associatedwith an increased risk of recurrent rejections, progression toTCAD, and poor long-term survival and that this presumablyreflects insufficient immunosuppression of T-cell/B-cell co-operation. Detecting this isotype switch is a clinically usefulsurrogate marker for in vivo T-cell activation and mayprovide a noninvasive approach for monitoring the efficacyof T-cell targeted immunosuppressive therapy in heart trans-plant recipients.

MethodsPatientsBetween January 1992 and January 2000, 475 adult patients receivedheart transplants at the Columbia-Presbyterian Medical Center, NewYork, NY. The study included 267 previously unsensitized primaryheart transplant recipients. Patients who were sensitized beforetransplantation with anti-HLA antibodies (n�22), who received asecond transplant (n�32), or who were supported with left ventric-ular–assist devices before transplantation (n�90) were excludedfrom the study. In addition, patients without follow-up data forposttransplantation anti-HLA antibody production (n�31) and thosewho died within the first month after transplantation (n�33) wereexcluded. All perioperative deaths were due to nonimmune causes,with the exception of 2 patients who died of acute rejection withinthe first week after transplantation. The demographic characteristicsof patients who were excluded from analysis were similar to those ofthe studied cohort.

Immunosuppressive RegimenAll patients who received transplants before 1996 were given a tripleimmunosuppression regimen consisting of cyclosporine, steroids,and azathioprine (AZA). Cyclosporine (Neoral, Novartis Pharmaceu-ticals) was administered preoperatively and then 1 to 2 mg/kg dailyas a continuous infusion until the patient could receive oral medica-tions. Cyclosporine doses were titrated to a whole-blood trough levelof 300 to 350 ng/mL. All patients received 4 mg/kg AZA preoper-atively and then 2 mg/kg intravenously until they were able toreceive oral medications. All patients who received transplants after1996 received mycophenolate mofetil (MMF) given at a startingdose of 1000 mg twice daily, instead of AZA, in combination withcyclosporine and steroids. Intravenous methylprednisolone was ad-ministered at a dose of 1 g intraoperatively and then at 125 mg q8hfor 3 doses. Prednisone was given in tapering doses of 1 mg · kg�1

· d�1 postoperatively to 0.1 mg · kg�1 · d�1 at the fourth month.Rejection episodes were treated with oral or intravenous steroid

pulses of 100 mg/d for 3 days followed by a tapered dose for 1 weekto the baseline dose. Nonresponders were treated with cytolytictherapy (OKT3 or ATGAM).

HLA TypingSerological typing of HLA-A and HLA-B loci was done by standardmicrocytotoxicity techniques. HLA-DR typing was performed byanalysis of serology samples and DNA techniques with sequence-specific oligonucleotide primers and polymerase chain reaction.

Anti-HLA AntibodiesAt the time of each endomyocardial biopsy, serum was screened forcomplement-mediated lytic activity, in the presence or absence ofdithiothreitol, against T and B lymphocytes from a panel of 70 of themost frequently encountered HLA class I and II antigens. In addition,donor-specific alloantibodies were assessed. Persistent serum reac-tivity after dithiothreitol treatment identified IgG alloantibodies,whereas the loss of reactivity identified IgM alloantibodies. Anti-HLA class I antibodies were identified when serum reacted with bothT-cell and B-cell panels. Anti-HLA class II antibodies were identi-fied when serum reacted with B-cell but not with T-cell panels.Antibodies against both HLA class I and II antigens were identifiedwhen serum reacted against both panels and when B-cell reactivityexceeded T-cell reactivity by �2-fold. The presence of autoantibod-ies was excluded by autologous serum cross-match with recipient Tcells and B cells. Overall, the use of these combined criteriaidentified anti-HLA class I and II antibodies with a high degree ofsensitivity and specificity.2

Detection of IgM-to-IgG Anti-HLA Antibody SwitchingDevelopment of isotype switching was considered to have occurredat the time when IgG anti-HLA antibodies were first detected in theserum of a transplant recipient whose previous samples demonstratedeither no anti-HLA reactivity or only the presence of IgM anti-HLAantibodies.

Detection of Soluble CD40 Ligand ProductionELISA was used to detect levels of soluble CD40 ligand (Chemicon)from patient sera. Absorbance was measured spectrophotometricallywithin 10 minutes after adding the phosphoric acid stop solution (450nm as the primary wavelength and 620 nm as the reference wavelength).

Acute RejectionRejection was diagnosed by routine endomyocardial biopsy: weeklyfor the first 4 weeks, then every 2 weeks for the next month, monthlyfor 4 months, every 2 months for the next 6 months, every 3 monthsfor the next 6 months, and then every 6 to 12 months. The biopsyfragments were graded according to the International Society forHeart and Lung Transplantation criteria.26 High-grade cellular allograftrejection was defined pathologically as grade 3A or 3B.

Transplantation-Related CADThe diagnosis of TCAD was based on the results of annuallyperformed coronary angiograms and defined as (1) discrete lesionsresulting in �50% obstruction of the proximal or middle portions ofmajor graft vessels or (2) diffuse, concentric narrowing of the wholevessels, including their branches. Patients were not given routinevasodilators before coronary injections. In patients with possiblediffuse CAD, intimal thickening was documented by vascularultrasound. All coronary angiograms were compared with the previ-ous year’s films to detect the presence of luminal irregularities,discrete stenoses, loss of third-order branches, or “pruning” ofvessels. Because of concerns related to donor-transmitted CAD, ourcenter generally did not accept hearts from female donors �45 yearsand male donors �40 years with angiographic evidence of CAD in�1 major vessel. In hearts from younger donors, baseline angio-grams were not routinely performed. Hearts explanted before re-transplantation and postmortem heart autopsy specimens were ex-

Lietz et al IgM-to-IgG Antibody Switching in Heart Transplant 2469

by guest on September 16, 2015http://circ.ahajournals.org/Downloaded from

amined for evidence of vessel occlusion and irregularities, ischemicdamage, and the presence of acute cardiac rejection.

The results of annual coronary angiography were available for 222patients. Three patients with a follow-up period �3 months and 2patients diagnosed with donor-transmitted coronary artery diseaseshortly after transplantation were excluded from this analysis.

Statistical AnalysesSurvival curves were estimated with the Kaplan-Meier method andcompared by log-rank tests.27 The Kaplan-Meier estimates describ-ing the time to the first episode of high-grade acute rejection used thedevelopment of anti-HLA antibodies as a time-varying covariate,wherein each patient was coded as a nonproducer until the time whenrelevant antibodies were detected. Patients who developed anti-HLAantibodies before the first rejection episode had the time to rejectioncensored at the time when anti-HLA antibodies were first detected.At that time, these patients were recategorized to the antibodyproducers group, and the time origin for the time-to-rejection curvewas set to the day when anti-HLA antibodies were detected. Allpatients were followed up until the first episode of high-grade acuterejection or the day of the last negative endomyocardial biopsy.Patients who developed anti-HLA antibodies after the first acuterejection episode were treated as nonproducers of alloantibodies forthis analysis. Because most cases of anti-HLA antibody productionwere detected before the first annual coronary angiography, Kaplan-Meier estimates describing the time to development of TCAD usedthe development of IgM and IgG anti-HLA antibody switching as astatic variable.

The Cox proportional-hazards model was used for multivariableanalyses. Separate Cox models were developed to identify riskfactors for high-grade acute rejection, TCAD, and IgG anti-HLAclass II isotype switching. The multivariable Cox models evaluatingthe effects of covariates on the risk of high-grade acute rejection usedanti-HLA antibody isotype switching as a time-varying covariate,whereas analysis of the risk for progression of TCAD used antibodydetection as a static variable. In addition, to accommodate theuncertainty of the time when TCAD developed between annualangiograms, a discrete-time version of the proportional-hazardsmodel was used according to the EXACT option in the SAS-sPHREG procedure. Covariates that achieved a significance ofP�0.15 in individual analyses were entered into the multivariablemodel by stepwise selection to produce a final set of risk factors. Therisk factors analyzed included donor and recipient ages; gender;ethnicity; pretransplantation diagnosis of heart disease; donor-recipient complete mismatch at HLA-A, HLA-B, and HLA-DR loci;ischemic time; and triple immunosuppressive regimen with MMFversus AZA.

Because nonfatal morbid events such as cellular rejection canoccur repeatedly in the same patient, cumulative frequencies forthese events were modeled by the method of Wei et al,28 which takesinto account the fact that repeated episodes may be correlated within

each patient. We did not control for multiple comparisons; therefore,type I errors are inflated above their nominal level of 0.05. Resultswere considered significant for P�0.05. Data were analyzed withSAS software version 7.1 (SAS Institute, Inc).

ResultsHigh Frequency of B-Cell Isotype Switching andProduction of IgG Anti-HLA Antibodies inCardiac Transplant RecipientsOf the 267 transplant cases studied, 80% were male, 80%were white, and the mean�SD age was 51.1�12.4 years. Themajor causes of heart disease were CAD (41%) and idio-pathic cardiomyopathy (41%). The 1- and 5-year graft sur-vival was 95% and 79%, respectively, with a mean follow-upof 3.1 years (range, 0.1 to 7 years).

IgG anti-HLA alloantibody class switching was detectedin 54% (143/267) of studied patients, 20% (53/267)produced IgM anti-HLA antibodies only, and 26% (71/267) did not demonstrate anti-HLA antibody production.Among patients who switched isotype, 20% (29/143)developed IgG anti-HLA class I antibodies, 40% (57/143)produced IgG anti-HLA class II antibodies, and 40%(55/143) of recipients demonstrated IgG antibodies againstboth classes of HLA molecules. As illustrated in Figure 1,by 1 year after transplantation, 28% of patients haddeveloped anti-HLA class I antibodies and 44% haddeveloped anti-HLA class II antibodies. In the majority ofpatients who switched alloantibody isotype during follow-up, most switching occurred during the first year aftertransplantation: 86% (74/86) and 88% (100/114) of anti-HLA class I and II producers, respectively. Isotype switch-ing against MHC class I molecules occurred earlier (me-dian, 15 days; range, 2 to 1539 days) than against class IImolecules (median, 71 days; range; 5 to 1294 days).

B-Cell Isotype Switching and Production of IgGAnti-HLA Class II Antibodies Are CorrelatedWith the Risk for High-Grade Acute RejectionAfter Cardiac TransplantationAs shown in Figure 2, B-cell isotype switching and theproduction of IgG antibodies directed against HLA class IIbut not against class I molecules were correlated with anincreased risk of high-grade acute rejection. Acute rejection

Figure 1. Kaplan-Meier survival curve illus-trates the time to B-cell antibody switchingand production of IgG anti-HLA class I (A)and II (B) molecules. Abbreviations are asdefined in text.

2470 Circulation October 18, 2005

by guest on September 16, 2015http://circ.ahajournals.org/Downloaded from

developed in 53.4% (47/88) of patients who switched isotypeto IgG anti-HLA class II compared with only 30.4% (80/263)of nonswitchers (log-rank P�0.03). By multivariable analy-sis, 2 factors were correlated with the increased risk ofhigh-grade acute rejection: switching to IgG anti-HLA classII molecules (relative risk [RR]�3.30, 95% confidence inter-val [CI]�2.18 to 4.99, P�0.001) and an AZA-based immu-nosuppressive regimen relative to one based on MMF(RR�1.71, 95% CI�1.20 to 2.44, P�0.003). Moreover,producers of IgG anti-HLA class II antibodies experienced ahigher incidence of recurrent rejections (23%, versus 12% ofnonswitchers, P�0.06), and their overall cumulative fre-quency of high-grade rejection within the first year aftertransplantation was higher (0.86, versus 0.51 in nonswitchers,P�0.002).

In contrast to these findings, isotype switching againstHLA class I molecules was not correlated with rejection bymultivariable analysis (RR�0.95, 95% CI�0.60 to 1.51,P�0.95). Producers and nonproducers of anti-HLA class Iantibodies did not differ with respect to the overall

cumulative frequency of rejections (0.81 versus 0.63,P�0.16) or the development of recurrent rejections (18%versus 17%, P�0.84).

B-Cell Isotype Switching and Production of IgGAnti-HLA Class II Antibodies Are CorrelatedWith the Risk for TCAD and Decreased SurvivalBecause recurrent episodes of acute rejection are known to beassociated with and predispose to accelerated TCAD, we nextinvestigated whether there was a correlation between aposttransplantation alloantibody isotype switch and progres-sion to TCAD. As shown in Figure 3, the presence of IgGantibodies directed against MHC class II but not against classI molecules was correlated with the risk for TCAD. Amongrecipients who switched isotype to IgG anti-HLA class IIantibodies, TCAD developed within a median of 2.0 years,compared with 3.1 years in nonswitchers (log-rank P�0.02).Multivariable analysis confirmed that the isotype switch andthe presence of IgG antibodies directed against MHC class IImolecules were risk factors associated with accelerated

Figure 2. A and B, B-cell isotype switchingand production of IgG anti-HLA class II anti-bodies increases the risk of the first high-grade acute rejection after cardiac transplan-tation. Development of IgG anti-HLAantibodies in this analysis was treated as atime-varying covariate, wherein each patientwas coded as a nonproducer of IgG anti-HLAantibodies until the day when the antibodieswere detected. Patients who developed IgGanti-HLA antibodies were recategorized to theantibody producers group, and their time torejection as antibody nonproducers was cen-sored, as described in Methods. Abbrevia-tions are as defined in text.

Figure 3. A and B, Increased risk for TCADaccompanying the B-cell isotype switch iscorrelated with the presence of IgG anti-HLAclass II but not class I antibodies. The analy-sis included only those patients with availableresults from the annual angiograms who hadno evidence of donor-transmitted CAD.Abbreviations are as defined in text.

Lietz et al IgM-to-IgG Antibody Switching in Heart Transplant 2471

by guest on September 16, 2015http://circ.ahajournals.org/Downloaded from

TCAD (RR�1.47, P�0.04). All patients in our studyswitched the anti-HLA antibody isotype against class IImolecules before the development of TCAD. However, laterepisodes of isotype switching against HLA class II moleculesafter the first 10 weeks after transplantation (median time toanti-HLA class II antibody switching) portended a higher riskof TCAD (RR�1.99, P�0.003) than did earlier episodes(RR�1.23, P�0.40) when compared with nonswitchers. Nosignificant associations were seen between earlier and laterepisodes of isotype switching against HLA class I moleculesin comparison with nonproducers of these antibodies.

Because recurrent rejections and accelerated progressionof TCAD predict poor long-term allograft survival, wenext sought to investigate the impact of antibody switchingon patient survival beyond the first year after transplanta-tion. Among 206 one-year survivors of heart transplanta-tion, those who switched antibody isotype to IgG anti-HLAclass II had a lower 5-year survival (91% [96/105]) thancontrols (81% [81/100], log-rank P�0.03). There was nosignificant difference in survival of patients who did anddid not switch antibody isotype against IgG MHC class Imolecules.

Persistence of IgM Anti-HLA Antibodies AgainstClass II but Not Against Class I Molecules for>30 Days Protects Against High-Grade AcuteRejection and TCADWe next investigated whether individuals who demonstrateda persistence of IgM antibodies directed against HLA mole-cules for at least 30 days and who did not isotype-switch toIgG antibodies were protected from acute rejection andTCAD. Patients who switched isotype and produced IgGanti-HLA antibodies were excluded from this analysis. Theresults showed that the persistent production of IgM anti-HLA antibodies directed against class II (log-rank P�0.02)but not against class I molecules (log-rank P�0.23) wasassociated with protection against acute rejection (Figure 4).Only 17% (4/23) of patients with persistent production ofIgM anti-HLA class II antibodies had a high-grade rejectionduring follow-up compared with 39% (56/144) of those who

did not produce these antibodies. Moreover, as illustrated inFigure 5, persistent production of IgM anti-HLA antibodiesagainst class II (log-rank P�0.05) but not against class I(log-rank P�0.47) molecules was associated with protectionagainst TCAD. During the follow-up period, 51% (14/28) ofpatients with persistent IgM anti-HLA class II antibodiesdeveloped TCAD compared with 74% (71/96) of patientswho did not produce these antibodies.

Increase in Circulating Soluble CD40 LigandCoincided With B-Cell Isotype Switching FromIgM to IgG Anti-HLA Class II AntibodiesBecause B-cell immunoglobulin isotype switching and pro-duction of IgG antibodies are regulated by interactionsbetween CD40 ligand (CD40L) on activated T cells andCD40 on the B-cell surface, we measured serum levels ofsoluble CD40L in 4 groups of patients as an indirect estimatorof T-cell dependent B-cell activation. The 4 groups consistedof (1) 8 heart transplant recipients on the day when IgGanti-HLA class II antibodies first appeared in their sera; (2) 6heart transplant recipients who never developed anti-HLAantibodies of the IgG isotype at a time after transplantationthat was matched for duration to the patients in group 1; (3)4 end-stage heart failure patients listed as status I for hearttransplantation whose sera were obtained on the day oflisting; and (4) 4 healthy volunteers who served as controls.Soluble CD40L serum concentrations were significantly in-creased in all patients who underwent heart transplantationcompared with both status I heart transplant candidates (0.139ng/mL) and healthy individuals (0 ng/mL). Moreover, solubleCD40L levels were 2-fold higher in the sera of patients on theday when their B cells switched isotype to produce IgGanti-HLA class II antibodies than in the sera of nonswitchersat similar times after transplantation (0.67 versus 0.31 ng/mL,P�0.012).

Immunosuppressive Regimen Influenced theIgM-to-IgG Anti-HLA Class II Antibody SwitchWe next sought to identify clinical risk factors for anti-HLAclass II isotype switching. By individual analysis, the only

Figure 4. A and B, Persistence of IgM anti-HLA class II but not class I molecules for �30days protects against the first episode ofhigh-grade acute rejection. The developmentof IgM anti-HLA antibodies in this analysiswas treated as a time-varying covariate,wherein each patient was coded as a nonpro-ducer of IgM anti-HLA antibodies until the daywhen the antibodies were detected. Patientswho developed IgM anti-HLA antibodies wererecategorized to the antibody producersgroup, and their time to rejection as antibodynonproducers was censored, as described inMethods. Patients who switched isotype toproduce IgG anti-HLA antibodies wereexcluded from this analysis. Abbreviations areas defined in text.

2472 Circulation October 18, 2005

by guest on September 16, 2015http://circ.ahajournals.org/Downloaded from

risk factor identified was modulation of the triple immuno-suppressive regimen used. Multivariable analysis confirmedthat substitution of MMF for AZA in patients who were alsoreceiving steroids and cyclosporine significantly reduced therisk of developing IgG anti-HLA class II antibody produc-tion, which was 1.9-fold higher for AZA- than for MMF-based regimens (P�0.004, the Table). Moreover, treatmentwith a regimen containing AZA rather than MMF was astronger risk factor for production of IgG anti-HLA class IIantibodies than more conventional risk factors for alloreac-tivity, such as donor-recipient mismatches at HLA-A,HLA-B, or HLA-DR loci.

Case-Control Study Demonstrated That MMFPrevented T-Cell Activation and B-Cell Anti-HLAAntibody Isotype SwitchingTo directly evaluate the effect of MMF on anti-HLA class IIisotype switching, we performed a case-control study com-paring alloantibody production in 30 cardiac transplant recip-ients matched according to age and sex. These patientsparticipated in a prospective pilot study conducted in 1994that compared MMF (cases) versus AZA (controls) as part ofa triple immunosuppressive regimen. Treatment with MMFdelayed the onset of B-cell isotype switching and anti-HLAclass II antibody production, with only 29% of MMF-treatedrecipients developing anti-HLA class II antibodies by the endof the first year compared with 55% of AZA-treated patients(P�0.25). Notably, whereas both groups had undetectablelevels of soluble CD40L at the time of transplantation, MMFtreatment prevented the significant rise in soluble CD40Llevels seen at 6 months in patients treated with AZA (0.01versus 0.36 ng/mL, P�0.001). Because MMF has previouslybeen reported to significantly reduce the risk of acute,high-grade cellular rejection compared with AZA,25 ourresults indicate that the mechanism of this protective effectinvolves prevention of T-cell activation and of IgG anti-HLAclass II isotype switching.

DiscussionIn this study, we have shown that de novo isotype switchingof IgM to IgG anti-HLA antibodies directed against class IIantigens was associated with earlier and more frequenthigh-grade cellular rejections, an increased risk of acceleratedTCAD, and decreased long-term survival after heart trans-plantation. In contrast, a lack of isotype switching withpersistent production of IgM anti-HLA class II antibodieswas associated with a decreased incidence of acute cellularrejection and protection against the development of TCAD.Moreover, alloantibody isotype switching was associatedwith higher levels of soluble CD40 ligand. A case-controlstudy showed that reduction of cardiac allograft rejection byMMF was accompanied by both reduced CD40L serum

Figure 5. A and B, Persistence of IgM anti-HLA class II but not class I molecules for �30days protects against TCAD. The analysisincluded only those patients with availableresults from the annual angiograms who hadno evidence of donor-transmitted CAD.Patients who switched isotype to produce IgGanti-HLA antibodies were excluded from thisanalysis. Abbreviations are as defined in text.

Multivariable Analysis of Risk Factors for Antibody IsotypeSwitching to IgG Anti-HLA Class II Antibodies Within the FirstYear After Transplantation

Predictors of Antibody Isotype Switchingto IgG Anti-HLA Class II P RR 95% CI

Immunosuppressive regimen

Double therapy with AZA vs MMF 0.004 1.90 1.22–2.94

Recipient demographics

Age �50 y 0.92 1.02 0.65–1.61

Female sex 0.28 1.32 0.80–2.20

Ethnicity (white) 0.11 1.75 0.88–3.50

Cause of heart disease:

Ischemic cardiomyopathy 1

Idiopathic cardiomyopathy 0.86 0.94 0.48–1.83

Donor demographics

Age �30 y 0.94 0.98 0.63–1.53

Female sex 0.62 0.89 0.55–1.43

Ethnicity (white) 0.46 0.80 0.43–1.46

Donor-recipient HLA mismatch (complete)

A locus 0.99 0.99 0.64–1.56

B locus 0.13 0.68 0.41–1.11

DR locus 0.61 0.89 0.56–1.40

Ischemic time �4 h 0.25 0.58 0.23–1.47

Abbreviations are as defined in text.

Lietz et al IgM-to-IgG Antibody Switching in Heart Transplant 2473

by guest on September 16, 2015http://circ.ahajournals.org/Downloaded from

levels and prevention of IgM-to-IgG anti-HLA class IIantibody switching. These results indicate that IgM-to-IgGanti-HLA class II antibody switching is the result of concom-itant T-cell activation and B-cell stimulation via CD40L-CD40 interactions.

The relation identified in this study between cellularrejection and production of IgG antibodies against MHCclass II antigens may reflect a heightened activity ofrecipient T cells, which recognize processed forms ofsoluble HLA alloantigens released from the graft incontext of self-MHC.29 –31 Such T cells, engaged in theindirect recognition pathway, produce lymphokines re-quired for the growth and maturation of alloantibody-producing B cells and cytotoxic T lymphocytes. Activationof B cells by soluble MHC class II products, particularlyHLA-DR molecules, and the subsequent presentation ofmultiple HLA-DR allopeptides by self–B cells to CD4 Tcells has been postulated to initiate the cascade of intramo-lecular and intermolecular epitope spreading and diversi-fication of the immune response,22–24 resulting in theamplification of T-cell responses, recurrent high-gradecellular rejections, and the early development of TCAD. Inthis view, an amplification loop of increased alloreactivityand activation of B cells set up by the primary T-cellrecognition of donor HLA class II peptide would result inB-cell isotype switching and surface expression of IgGantibodies against donor MHC class II molecules, whichwould then bind cognate antigen and allow the B cell topresent diverse peptides in the context of self–MHC classII molecules to the CD4 T cells. Because injury of theallograft during acute rejection is an important source ofsoluble alloantigens, this may explain the high incidence ofIgG anti-HLA class II antibody switching directly after thefirst rejection occurrence and the subsequent high fre-quency of recurrent rejections.

B-cell isotype switching to IgG antibodies against MHCclass I antigens does not appear to have the same adverseeffects on allograft outcome. One possible explanationmay be that MHC class I antigens expressed by the graftare recognized predominantly by the recipient’sCD8�CD28� T-suppressor cells. Activation of the sup-pressor cell pool has been shown to inhibit activation ofHLA class II–reactive T-helper cells32–35 and therebymaintain quiescence after organ transplantation.

The observed patterns of progressive intramolecular andintermolecular T-cell HLA-DR epitope spreading in car-diac transplant recipients with recurrent rejections andaccelerated TCAD,19 –21 as well as the increased frequencyof cellular rejection reported when IgG antibodies againstMHC class II antigens are present in sensitized patientsbefore transplantation,36 provide support for the concept ofa B-cell/T-cell amplification loop.

Although it is generally agreed that HLA disparity iscorrelated with the strength of alloreactivity,4 our studyhas shown no association between B-cell anti-HLA class IIantibody switching and the degree of HLA-DR mismatch.The ability of B cells to switch isotype in response toT-cell–presented HLA-DR molecules relies on the intactfunction of all elements of the T- and B-cell interaction,

such as CD4� T-cell ability to recognize HLA-DR anti-gens, express appropriate surface receptors,37 and producesufficient numbers of cytokines to stimulate B-cell differ-entiation.38 In transplant recipients, the function of manyof these components is inhibited by the action of immu-nosuppressive drugs, the efficacy of which may vary inindividual patients, depending on some genetic determi-nants (eg, variation in cytokine genes39) and other un-known factors. The heterogeneity of alloresponses toallografts with the same degree of HLA-DR disparity hasbeen previously shown.4 Therefore, escape of B-cell acti-vation in immunosuppressed patients and generation ofanti-HLA class II antibody switching would identify pa-tients with heightened responses to donor MHC molecules,irrespective of the degree of HLA-DR mismatch. This testwould therefore allow assessment of individual allorespon-siveness independent of routinely monitored immunosup-pressive drug levels and donor MHC disparity.

After alloantigen recognition, activated T cells in trans-plant recipients upregulate CD40L and amplify the allo-reactive response by ligating CD40 molecules, not only onantigen-presenting cells but also on allograft endothelialcells.37 Increased T-cell expression of CD40L has beenassociated with acute and chronic renal allograft rejec-tion,40 and interruption of the CD40L-CD40 costimulatorypathway results in long-term survival of heart, kidney,liver, and skin allografts in animal models of transplanta-tion.41– 44 Because IgM-to-IgG isotype switching by B cellsis tightly regulated by both T-cell costimulatory signalsand recognition of cognate antigen, our results implicateCD40L-CD40 interactions in the induction of B-cell iso-type switching after organ transplantation.

Several lines of evidence support the idea that monitor-ing recipient sera for switching of IgM to IgG anti-HLAclass II antibodies can provide a clinically useful param-eter for evaluating the efficacy of T-cell directed immu-nosuppressive therapy. First, we have previously shownthat induction of IgG anti-HLA class II antibodies isaccompanied by intragraft T-helper cell activation andinterleukin-2 receptor expression.2,45 Second, we recentlyshowed that use of a humanized monoclonal antibodycapable of selectively blocking the interleukin-2 receptoron T cells without cellular activation reduced both cellularallograft rejection and the formation of IgG anti-HLAantibodies.46 Third, we have demonstrated that a tripleimmunosuppressive regimen containing MMF, which issuperior to that containing AZA for prevention of T-cellactivation and cellular rejection,25 is also superior forprevention of posttransplantation B-cell isotype switchingof IgM to IgG antibodies directed against MHC class IImolecules.

Collectively, these observations indicate that monitoringcardiac allograft recipients for de novo production of IgGanti-HLA class II antibodies may provide a simple indirectassessment of whether inhibition of CD4 T-cell alloacti-vation in a given patient is adequate and may be used as asurrogate marker in clinical trials for determining theefficacy of novel T-cell immunomodulatory therapies.Moreover, detection of de novo production of IgG anti-

2474 Circulation October 18, 2005

by guest on September 16, 2015http://circ.ahajournals.org/Downloaded from

HLA class II may allow relatively early posttransplantionidentification of patients at increased risk for poor long-term outcomes who would benefit from either modificationof their current immunosuppressive regimen or institutionof more effective immunomodulatory therapies aimed atslowing or preventing the progression of TCAD.

AcknowledgmentsThe research was supported by the National Institutes of Health-Grant No. AI055234-03 and the Interuniversitary Organ Transplan-tation Consortium, Rome, Italy.

References1. Zerbe T, Uretsky B, Kormos R, Armitage J, Wolyn T, Griffith B,

Hardesty R, Duquesnoy R. Graft atherosclerosis: effects of cellularrejection and human lymphocyte antigen. J Heart Lung Transplant.1992;11:S104–S110.

2. Itescu S, Tung TC, Burke EM, Weinberg AD, Mancini D, Michler RE,Suciu-Foca NM, Rose EA. An immunological algorithm to predictrisk of high-grade rejection in cardiac transplant recipients. Lancet.1998;352:263–270.

3. Kerman RH, Kimball P, Scheinen S, Radovancevic B, Van Buren CT,Kahan BD, Frazier OH. The relationship among donor-recipient HLAmismatches, rejection, and death from coronary artery disease incardiac transplant recipients. Transplantation. 1994;57:884 – 888.

4. Smith JD, Rose ML, Pomerance A, Burke M, Yacoub MH. Reductionof cellular rejection and increase in longer-term survival after hearttransplantation after HLA-DR matching. Lancet. 1995;346:1318 –1322.

5. Constanzo-Nordin MR. Cardiac allograft vasculopathy: relationshipwith acute cellular rejection and histocompatibility. J Heart LungTransplant. 1992;11:S90 –S103.

6. Suciu-Foca N, Reed E, Marboe C, Harri P, Yu PX, Sun YK, Ho E,Rose E, Reemtsma K, Kind DW. The role of anti-HLA antibodies inheart transplantation. Transplantation. 1991;51:716 –724.

7. Hess M, Hastillo A, Mohanakumar T, Cowley MJ, Vetrovac G,Szentpetery S, Wolfgang TC, Lower RR. Accelerated atherosclerosisin cardiac transplantation: role of cytotoxic B cell antibodies andhyperlipidemia. Circulation. 1983;68(suppl II):II-94 –II-101.

8. Rose EA, Pepino P, Barr ML, Smith CR, Ratner AJ, Ho E, Berger C.Relation of HLA antibodies and graft atherosclerosis in human cardiacallograft recipients. J Heart Lung Transplant. 1992;11:S120 –S123.

9. Barr ML, Cohen DJ, Benevisty AI, Hardy M, Reemtsma K, Rose EA,Marboe CC, D’Agati V, Suciu-Foca N, Reed E. Effect of anti-HLAantibodies on the long-term survival of heart and kidney allografts.Transplant Proc. 1993;25:262–264.

10. Smith JD, Danskine AJ, Rose ML, Yacoub MH. Specificity of lym-phocytotoxic antibodies formed after cardiac transplantation and cor-relation with rejection episodes. Transplantation. 1991;53:1358 –1362.

11. George JF, Kirklin JK, Shroyer TW, Naftel DC, Bourge RC, McGiffinDC, White-Williams C, Noreuil T. Utility of posttransplantationpanel-reactive antibody measurements for the prediction of rejectionfrequency and survival of heart transplant recipients. J Heart LungTransplant. 1995;14:856 – 864.

12. Chapman JR, Taylor CJ, Ting A, Morris PJ. Immunoglobulin classand specificity of antibodies causing positive T cell crossmatches(relationship to renal transplant outcome). Transplantation. 1986;42:608 – 613.

13. Farges O, Buffello D, Shi YM, Berth A, Bismuth H. Anti-donorantibody class switching after liver transplantation. Transplantation.1995;60:296 –300.

14. Sumethkul V, Mongkolsuk T, Sujirachato K, Chiewsilp P. Immuno-globulin class switch of posttransplant panel reactive antibody and theimpact on kidney allograft outcome. Transplant Proc. 1998;30:1167.

15. Rizeq M, Masek MA, Billingham ME. Acute rejection: significance ofelapsed time after transplantation. J Heart Lung Transplant. 1994;13:862– 868.

16. Winters GL, Loh E, Schoen FJ. Natural history of focal moderatecardiac allograft rejection: is treatment warranted. Circulation. 1995;91:1975–1980.

17. White JA, Guirandon C, Pfluglelder PW, Kostuk WJ. Routine sur-veillance myocardial biopsies are unnecessary beyond one year afterheart transplantation. J Heart Lung Transplant. 1995;14:1052–1056.

18. Kubo SH, Naftel DC, Mills RM Jr, O’Donnell J, Rodeheffer RJ,Cintron GB, Kenzora JL, Bourge RC, Kirklin JK. Risk factors for laterecurrent rejection after heart transplantation: a multiinstitutional,multivariable analysis; Cardiac Transplant Research Database Group.J Heart Lung Transplant. 1995;14:409 – 418.

19. Liu Z, Colovai AI, Tugulea S, Reed EF, Fisher PE, Mancini D, RoseEA, Cortesini R, Michler RE, Suciu-Foca N. Indirect recognition ofdonor HLA-DR peptides in organ allograft rejection. J Clin Invest.1996;98:1150 –1157.

20. Tugulea S, Ciubotariu R, Colovai AI, Liu Z, Itescu S, Schulman LL,Fisher PE, Hardy MA, Rose EA, Michler RE, Cortesini R, Suciu-FocaN. New strategies for early diagnosis of heart allograft rejection.Transplantation. 1997;64:842– 847.

21. Ciubotariu R, Liu Z, Colovai AI, Ho E, Itescu S, Ravalli S, HardyMA, Cortesini R, Rose EA, Suciu-Foca N. Persistent allopeptidereactivity and epitope spreading in chronic rejection of organ allo-grafts. J Clin Invest. 1998;101:398 – 405.

22. Vanderlugt CJ, Miller SD. Epitope spreading. Curr Opin Immunol.1996;8:831– 836.

23. Mamula MJ, Janeway CA Jr. Do B cells drive the diversification ofimmune responses? Immunol Today. 1993;14:151–154.

24. Reed EF, Hong B, Ho E, Harris PE, Weinberger J, Suciu-Foca N.Monitoring of soluble HLA alloantigens and anti-HLA antibodiesidentifies heart allograft recipients at risk of transplant associatedcoronary artery disease. Transplantation. 1996;61:556 –572.

25. Kobashigawa J, Miller L, Renlund D, Mentzer R, Alderman E, BourgeR, Constanzo M, Eisen H, Dueau G, Ratkovec R, Hummel M, Ipe D,Johnson J, Keogh A, Mamelok R, Mancini D, Smart F, Valantine H.A randomized active-controlled trial of mycophenolate mofetil inheart transplant recipients; Mycophenolate Mofetil Investigators.Transplantation. 1998;66:507–515.

26. Billingham ME, Cary NR, Hammond ME, Kemnitz J, Marboe C,McCallister HA, Snovar DC, Winters GL, Zerbe A. A working for-mulation for the standardization of nomenclature in the diagnosis ofheart and lung rejection: Heart Rejection Study Group, the Interna-tional Society for Heart Transplantation. J Heart Transplant. 1990;9:587–593.

27. Kaplan EL, Meier P. Nonparametric estimation from incompleteobservations. J Am Stat Assoc. 1958;53:457– 481.

28. Wei LJ, Lin DY, Weissfeld L. Regression analysis of multivariateincomplete failure time data by modeling marginal distributions. J AmStat Assoc. 1989;84:1065–1073.

29. Krieger NR, Yin DP, Fathman CG. CD4� but not CD8� cells areessential for allorejection. J Exp Med. 1996;184:2013–2018.

30. Zeevi A, Pavlick M, Lombardozzi S, Banas R, Pappo O, Rao AS,Fontes P, Demetris J, Shapiro R, Dodson F. Immune status ofrecipients following bone marrow-augmented solid organ transplan-tation. Transplantation. 1995;59:616 – 620.

31. Creemers P, Du Toit E, Cassidy MJ, Kahn D. Sequential mixedlymphocyte culture after kidney transplantation: induction of tol-erance or sensitization. Nephron. 1997;75:166 –170.

32. Liu Z, Tugulea S, Cortesini R, Suciu-Foca N. Specific suppression ofT helper alloreactivity by allo-MHC class I restricted CD8�CD28� Tcells. Int Immunol. 1998;10:775–783.

33. Ciubotariu R, Colovai AI, Pennesi G, Liu Z, Smith D, Berlocco P,Cortesini R, Suciu-Foca N. Specific suppression of human CD4� Thcell responses to pig MHC antigens by CD8�CD28� regulatory Tcells. J Immunol. 1998;161:5193–5202.

34. Li J, Liu Z, Jiang S, Cortesini R, Lederman S, Suciu-Foca N. Tsuppressor lymphocytes inhibit NF-�B-mediated transcription ofCD86 gene in APC. J Immunol. 1999;163:6386 – 6392.

35. Chang CC, Ciubotariu R, Manavalan JS, Yuan J, Colovai AI, PiazzaF, Lederman S, Colonna M, Cortesini R, Dalla-Favera R, Suciu-FocaN. Tolerization of dendritic cells by T(S) cells: the crucial role ofinhibitory receptors ILT3 and ILT4. Nat Immunol. 2002;3:237–243.

36. Itescu S, Tung T, Burke E, Weinberg A, Moazami N, Artrip JH,Suciu-Foca N, Rose EA, Oz MC, Michler RE. Preformed IgG anti-bodies against major histocompatibility class II antigens are a majorrisk factor for high-grade cellular rejection in recipients of hearttransplantation. Circulation. 1998;98:786 –791.

37. Hancock WW, Sayegh MH, Zheng XG, Peach R, Linsley PS, TurkaLA. Costimulatory function and expression of CD40 ligand, CD80,

Lietz et al IgM-to-IgG Antibody Switching in Heart Transplant 2475

by guest on September 16, 2015http://circ.ahajournals.org/Downloaded from

and CD86 in vascularized murine cardiac allograft rejection. ProcNatl Acad Sci U S A. 1996;93:13967–13972.

38. Theze J, Alzari PM, Bertoglio J. Interleukin-2 and its receptors: recentadvances and new immunological functions. Immunol Today. 1996;17:481– 486.

39. Holweg CT, Peeters AMA, Balk AH, Uitterlinden AG, Niesters HG,Maat AP, Weimer W, Baan CC. Effect of HLA-DR matching on acuterejection after clinical heart transplantation might be influenced by anIL-2 gene polymorphism. Transplantation. 2002;73:1353–1356.

40. Shoker A, George D, Yang H, Baltzan M. Heightened CD40 ligandgene expression in peripheral CD4� T cells from patients with kidneyallograft rejection. Transplantation. 2000;70:497–505.

41. Fu F, Li W, Lu L, Thomson AW, Fung JJ, Qian S. Systemic admin-istration of CTLA4-Ig or anti-CD40 ligand antibody inhibitssecond-set rejection of mouse liver allografts. Transplant Proc. 1999;31:629 – 632.

42. Lu L, Li W, Fu F, Chambers FG, Qian S, Fung JJ, Thomson AW.Blockade of the CD40-CD40 ligand pathway potentiates the capacity

of donor-derived dendritic cell progenitors to induce long-termcardiac allograft survival. Transplantation. 1997;64:1808 –1815.

43. Larsen CP, Elwood ET, Alexander DZ, Ritchie SC, Hendrix R,Tucker-Burden C, Cho HR, Aruffo A, Hollenbaugh D, Linsley PS,Winn KJ, Pearson TC. Long-term acceptance of skin and cardiacallografts after blocking CD40 and CD28 pathways. Nature. 1996;381:434 – 438.

44. Kirk AD, Harlan DM, Armstrong NN, David TA, Dong Y, Gray GS,Hong X, Thomas D, Fechner JH Jr, Knechtle SJ. CTLA4-Ig andanti-CD40 ligand prevent renal allograft rejection in primates. ProcNatl Acad Sci U S A. 1997;94:8789 – 8794.

45. Fisher PE, Suciu-Foca N, Ho E, Michler RE, Rose EA, Mancini D.Additive value of immunologic monitoring to histologic grading ofheart allograft biopsy specimens: implications for therapy. J HeartLung Transplant. 1995;14:1156 –1161.

46. Beniaminovitz A, Itescu S, Lietz K, Donovan M, Burke EM, GroffBD, Edwards N, Mancini DM. Prevention of rejection in cardiactransplantation by blockade of the interleukin-2 receptor with a mono-clonal antibody. N Engl J Med. 2000;342:613– 619.

2476 Circulation October 18, 2005

by guest on September 16, 2015http://circ.ahajournals.org/Downloaded from

Suciu-Foca, Donna Mancini and Silviu ItescuKatherine Lietz, Ranjit John, Elizabeth Burke, Michael Schuster, Tyson B. Rogers, Nicole

Risk of Cellular Rejection and Coronary Artery DiseaseAntibody Switching in Cardiac Transplant Recipients Is Associated With an Increased

Immunoglobulin M-to-Immunoglobulin G Anti-Human Leukocyte Antigen Class II

Print ISSN: 0009-7322. Online ISSN: 1524-4539 Copyright © 2005 American Heart Association, Inc. All rights reserved.

is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231Circulation doi: 10.1161/CIRCULATIONAHA.104.485003

2005;112:2468-2476Circulation. 

http://circ.ahajournals.org/content/112/16/2468World Wide Web at:

The online version of this article, along with updated information and services, is located on the

  http://circ.ahajournals.org//subscriptions/

is online at: Circulation Information about subscribing to Subscriptions: 

http://www.lww.com/reprints Information about reprints can be found online at: Reprints:

  document. Permissions and Rights Question and Answer this process is available in the

click Request Permissions in the middle column of the Web page under Services. Further information aboutOffice. Once the online version of the published article for which permission is being requested is located,

can be obtained via RightsLink, a service of the Copyright Clearance Center, not the EditorialCirculationin Requests for permissions to reproduce figures, tables, or portions of articles originally publishedPermissions:

by guest on September 16, 2015http://circ.ahajournals.org/Downloaded from