Ct

S

PdImfHa1rsp

P

Fw

C3

T

orrelation Between HLA and Posttransplantation Diabetes Mellitus inhe Han Population in South China

.J. Yu, L. Peng , X.B. Xie, F.H. Peng, C.H. Fang, Y. Wang, and G.B. Lan

ABSTRACT

Background. Posttransplantation diabetes mellitus (PTDM) is a common metaboliccomplication in renal transplant recipients. Some studies have revealed predisposing orprotective HLA genes for PTDM.Objective. To describe the characteristics of PTDM in patients in the south of China.Methods. The present study included 195 living-donor kidney transplant recipients.Informed consent was obtained from all participants, and the study was approved by ourInstitutional Ethics Committee. Each donor-recipient pair was related. Twenty-twopatients had PTDM, whereas 173 did not. The frequency of each HLA phenotype wascompared between these 2 groups. The 195 patients were divided into 2 groups accordingto immunosuppression regimen, a cyclosporine group and a tacrolimus group, and theincidence of PTDM was compared between the groups. Patients were then subdivided into2 groups according to age, an elderly group (age 40 years or older) and a younger group(age younger than 40 years), and the incidence of PTDM was calculated and comparedbetween these 2 groups.Results. HLA-A30 and HLA-DR7 seem to be predisposing genes for PTDM in patientsin the south of China; Low dosages of calcineurin inhibitors were used in our center, Therewas no significant difference in the incidence of PTDM between the CsA and Tac groups;The incidence of PTDM in the elderly group was significantly higher than that in theyounger group.Conclusion. Patients receiving HLA-A30 and HLA-DR7 antigens, and elderly patientsare at higher risk of developing PTDM. Tacrolimus does not significantly increase the

incidence of PTDM.

rwap

eoda

C

Ot

OSTTRANSPLANTATION diabetes mellitus (PTDM)is one of the main causes of infection, proteinuria, graft

ysfunction, and cardiovascular and cerebrovascular events.ts incidence varies from 2.5% to 24%.1,2 Onset of diabetesellitus (DM) is influenced by genetic and environmental

actors. HLA is an important genetic determinant of DM.3

LA genes related to type 2 DM differ among nations andreas and nations,4–6 as do those related to PTDM (Table). The objective of the present study was to analyze theelationship between HLA and PTDM to identify bothusceptible and protective genes for PTDM in the Hanopulation in the south of China.

ATIENTS AND METHODS

rom 2003 to 2008, 211 living-donor transplantation procedures

ere performed at our center. Each donor-recipient pair was C

rown Copyright © 2010 Published by Elseiver Inc. All rights reserve60 Park Avenue South, New York, NY 10010-1710

ransplantation Proceedings, 42, 2509–2512 (2010)

elated. The local police station provided kinship certification,hich was closely scrutinized by our Institutional Ethics Committeend Second Xiang-Ya Hospital before kidney transplantation. Norisoners were used in the study, either as donors or recipients.Patient data (eg, sex, age, ethnic group, and telephone number),

xamination results, and HLA phenotypes of the recipients werebtained. All patients were followed up via the information input in aatabase. Each donor-recipient pair was ABO-compatible, with neg-tive complement-dependent cytotoxicity (death rate of lymphocytes

From Second Xiang-Ya Hospital, Central South University,hangsha, China.Address reprint requests to Longkai Peng, PhD, Urologicalrgan Transplantation Center, Second Xiang-Ya Hospital, Cen-

ral South University, No.139 in Middle RenMin Road, Changsha

ity, Hunan Province, China. E-mail: plk3000@163.com

d. 0041-1345/–see front matterdoi:10.1016/j.transproceed.2010.04.029

2509

�tAaH

I

T0pcGtpmhpmmm6mmn1wwpb0wta

D

Ttbcttgir

p

mbrpw

R

RnaPai

Tp1

(igs

D

Tpmc

srpppdw

mgbo

2510 YU, PENG, XIE ET AL

5%). The polymerase chain reaction with sequence-specific markersechnique was used for HLA-A, HLA-B, and HLA-DR phenotyping.n enzyme-linked immunosorbent assay was used for panel reactive

ntibody (PRA) testing, and when anti-DQ antibodies were detected,LA-DQ phenotyping was verified.

mmunosuppression

he immunosuppressive regimen included methylprednisolone,.5 g, during the operation, and 0.5 g/d in the first 3 daysosttransplantation, with an overall dose of 2.5 g or less. Cal-ineurin inhibitors and mycophenolate mofetil (MMF; CellCept,enentech, Inc, South San Francisco, California) were prescribed

he day after surgery. Methylprednisolone was replaced withrednisone on postoperative day 3. The cyclosporine (CsA) regi-en was as follows: CsA plus MMF (first dose, 0.75 g every 12

ours, with gradual tapering over 1 month to 0.5 g every 12 hours)lus prednisolone (first dose, 80 mg/d, reduced by 10 mg/d to 20g/d and maintained for 1 month, then reduced to 15 mg/d andaintained for 6 months, and finally reduced to 10 mg/d andaintained continuously). Cyclosporine therapy was begun atmg/kg/d, and was gradually reduced to 3 mg/kg/d for long-termaintenance. The CsA blood concentration was monitoredonthly, and the trough concentration was adjusted to 220 to 250

g/mL over 1 to 6 months, 180 to 220 ng/mL at 6 to 12 months, and50 to 180 ng/mL after 12 months. The tacrolimus (Tac) regimenas as follows: Tac plus MMF (first dose, 0.75 g every 12 hours,ith gradual tapering over 1 month to 0.5 g every 12 hours) plusrednisone (same usage as for CsA). Tacrolimus therapy wasegun at 0.1 to 0.2 mg/kg/d, and was gradually reduced to 0.05 to.1 mg/d for long-term maintenance. The Tac blood concentrationas monitored monthly, and the trough concentration was adjusted

o 6 to 8 ng/mL over 1 to 6 months, 5 to 6 ng/mL at 6 to 12 months,nd 4 to 5 ng/mL after 12 months.

iagnostic Criteria for PTDM

he recipients had not been diagnosed with DM before transplan-ation. Symptoms of DM developed after transplantation (randomlood glucose concentration �11.1 mmol/L or fasting blood glu-ose concentration (at least 8 hours after intake) �7.8 mmol/L). Ifhe symptoms of DM were not easily recognized, an oral glucoseolerance test (OGTT) was performed to determine the bloodlucose concentration at 2 hours after glucose intake. If the results more than 11.1 mmol/L, a diagnosis of PTDM can be made. Allesults were verified on another day.

Complete data for the 211 patients were obtained. Sixteen

Table 1. Reported PTDM-Related G

Source Country/District

David et al7 United StatesSumrani et al8 New Yorkvon Kiparski et al9 SwitzerlandHjelmesaeth et al10 NorwayNafar et al11 IranAddous et al12 Saudi ArabiaYildiz et al13 TurkeySaxena et al14 IndiaTang et al15 Beijing, ChinaWang et al16 Northeast China

PTDM, posttransplantation diabetes mellitus.

atients were excluded, including recipients with DM, those from t

inority populations, those who underwent allograft nephrectomyecause of uncontrollable rejection during the perioperative pe-iod, and those who died of pulmonary infection during theerioperative period. Thus, 195 patients were enrolled in the study,ith informed consent.

ESULTS

ecipients were divided into 2 groups: PTDM (n � 22) andon-PTDM (n � 173). The frequency of HLA-A30 (9.1%)nd HLA-DR7 (9.1%) was significantly higher in theTDM group compared with the non-PTDM group (1.4%nd 2.0%, respectively; (P � .05, �2 test). The relative riskn the 2 groups was 6.5 and 4.55, respectively.

Recipients were further divided into a CsA group and aac group. The incidence of PTDM was 15.28% (8 of 52atients) in the Tac group vs 9.8% in the CsA group (14 of43), which was not significant (P � .25, �2 test).

Recipients were further divided into an elderly groupage �40 years) and a younger group (age � 40 years). Thencidence of PTDM was 19.80% (20 of 101) in the elderlyroup vs 2.13% (2 of 94) in the younger group, which wasignificant (P � .05, �2 test).

ISCUSSION

he incidence of PTDM is 9-fold that of DM.17 Mostatients demonstrated no symptoms, and the diagnosis wasade on the basis of elevated blood or urine glucose

oncentration.Many centers have already performed correlation analy-

es between HLA and PTDM (Table 1). In contrast,esearchers from France18,19 and Puerto Rico20 have re-orted that PTDM is not associated with HLA. In theresent study, HLA-A30 and HLA-DR7 were the predis-osing genes to PTDM in the south of China, remarkablyifferent from the north of China and other parts of theorld.HLA class II molecules are important antigen presentingolecules, with polymorphism accounting for 50% of the

enetic risk of contracting type 1 DM.21 The peptide-inding specificities of HLA-II molecules depend not onlyn peptide-binding motifs but also on spatial conforma-

in Various Countries and Districts

Predisposing Genes Protective Genes

, Bw42

, DR8 DR6, DR52, DR4 B7, DR2, DR3, DR4, DR3

1*02, DRB1*03, DQB1*03 DRB3, DQB1*06

enes

A28A30B8B27A26DR3DR2B15B15DRB

ions. Thus, susceptible HLA molecules can bind autoanti-

gcMterc

i(dk(prdapsewsLsgihgbstasTltpTegtbwriaddoibHipptd

tpoSpi

iP(cpcvppnspapoC

rbjT

Ptidvpcc

titstcoltOs

R

dt

r

HLA AND POSTTRANSPLANTATION DIABETES 2511

enic peptides and present them to autoreactive T lympho-ytes, whereas protective alleles do not allow this reaction.

any studies have described both predisposing and protec-ive molecules. Another hypothesis is that protective mol-cules mediate intrathymic deletion of diabetogenic auto-eactive T-cell clones22 or positive selection of regulatoryells23 more efficiently than predisposing molecules do.

In 1986, Mosmann et al24 differentiated mouse T cellsnto type 1 T-helper cells (Th1) and type 2 T-helper cellsTh2). In subsequent studies, Th1 and Th2 subgroups wereifferentiated in human T cells. Th1 cells produce interleu-in (IL)-2, interferon (IFN)-�, and tumor necrosis factorTNF)-�, and have a role in cell immunization. Th2 cellsroduce IL-4, IL-5, IL-6, IL-9, IL-10, and IL-13, and have aole in humoral immunity. Stimulating signals produceifferentiation of Th0 cells, which produce IL-2, IL-4, IL-10,nd INF-� before being stimulated. Th1 and Th2 cellsroduce cytokines that interact with each other. The Th1ubtype produces INF-�, which inhibits Th2 cellular prolif-ration, and the Th2 subgroup produces IL-4 and IL-10,hich inhibits growth of Th1 cells. The Th1 and Th2

ubgroups are in dynamic equilibrium in the normal state.iblau25 called this phenomenon “clonal diversion.” Th1

ubgroup priority was observed in many recipients withraft rejection, whereas Th2 priority was observed in recip-ents with long-term graft survival. However, some workersave reported that in certain mouse models, although theraft survived for a long time, Th1 and Th2 subgroups wereoth downregulated.26 Some workers believe that Th2ubgroup priority is an important predictor of partial or fullolerance. One theory is that HLA-II molecules presentutoantigenic peptides to functionally different T-cell sub-ets, inducing a diversion of the phenotype of diabetogenic

cells,27 predisposing to alleles that drive a Th1 response,eading to �-cell autoimmune destruction, whereas protec-ive alleles stimulate a Th2 response. Another study re-orted that splenic cells of DR0401 transgenic mice expressh2 cytokines, whereas those of DQ8 transgenic micexpress Th1 cytokines. It was postulated that the diabeto-enic effects of DQ8 were downregulated by HLA-DR0401o strengthen Th2 responses. When HLA-DR4 moleculesound to specific autoantigenic peptides, the Th2 responseas stimulated, antagonizing an HLA-DQ8-stimulated Th1

esponse, which would lead to autoimmune destruction ofslet cells. This result implied that there may be somessociation between Th2 responses and decreased inci-ence of type 1 diabetes. However, some studies haveemonstrated that Th2 cells do not necessarily delay devel-pment of DM,28 and it was even potentially diabetogenic

n certain models. The etiology of type 1 DM cannot simplye attributed to diversion of Th1 and Th2 responses.owever, many studies have suggested that a Th1 response

s responsible for insulitis, and a Th2 response may have arotective role. Another hypothesis of determinant captureroposes that protective and predisposing alleles competeo bind overlapping autoantigenic peptides recognized by

iabetogenic T cells.29,30 Harfouch-Hammoud et al31 found

t1

hat protective alleles exert a dominant effect over predis-osition in individuals carrying both types of alleles. An-ther study showed that patients with type 2 DM fromardinian type 2 diabetes families had insulin deficiency, inart because of preexisting anti-GAD65 antibody-induced

slet autoimmune destruction.32

Cyclosporine and Tac are 2 widely used calcineurinnhibitors, and both can lead to an increased incidence ofTDM. The pathogenic mechanisms of this type of drug are1) the similarity of signaling pathways between immuneells and �-cells, leading to interference by immunosup-ression agents; (2) Tac leads to �-cell vacuolization,ausing insulin deficiency; and (3) CsA causes small bloodessel spasm, leading to effects on the blood supply to theancreas. Schnitzler and Hardinger33 reported that, com-ared with CsA, Tac increased the incidence of PTDM byearly 75%, even in those cases in which CsA therapy waswitched to Tac therapy. Many studies have demonstrated aositive relationship between the Tac blood concentrationnd �-cell toxicities. However, some researchers have pro-osed that low dosages of Tac demonstrate less influencen glucose metabolism, even to a degree similar to that withsA.34

China is a developing country. All patients in our centereceived low dosages of immunosuppression agents, andlood concentrations were routinely monitored, with ad-

ustment of the drug dosage. Our study demonstrated thatac did not obviously increase the incidence of PTDM.Many studies have demonstrated that the incidence of

TDM is increased in elderly patients,35,36 and we came tohe same conclusion. Elderly recipients exhibit a higherncidence of underlying cardiovascular or cerebrovascularisease; thus, they are more susceptible to cardiocerebro-ascular events if complicated by PTDM. In this group ofatients, routine monitoring of blood and urine glucoseoncentrations is necessary to detect and control hypergly-emia.

The pathogenesis of PTDM is more complicated thanhat of diabetes. It refers not only to insulin deficiency andnsulin resistance in the development of diabetes but also tohe influences of immunosuppression agents on insulinecretion and insulin sensitivity, and the immune status ofhe recipients and organ function. Posttransplantation glu-ose metabolism is involved in an enormous network systemf HLA genes, drug-metabolizing enzyme genes, and insu-

in secretion and insulin sensitivity-related genes, as well asheir genetic polymorphisms and linkage disequilibrium.ur ultimate goal is to blueprint the complete network

ystem.

EFERENCES

1. Cosio FG, Pesavento TE, Osei K, et al: Post-transplantiabetes mellitus: increasing incidence in renal allograft recipientsransplanted in recent years. Kidney Int 59:732, 2001

2. Numakura K, Satoh S, Tsuchiya N, et al: Clinical and geneticisk factors for posttransplant diabetes mellitus in adult renal

ransplant recipients treated with tacrolimus. Transplantation 80:419, 2005

gp2

a3

naD

tiA

s5

ar

dc

id

Hm

deT

hd7

me

f2

aJ

iA

at

eP

eT

n1

tm

cs1

ml1

cI

aC

pT

dJ

H

che

ocd

a2

lT

cd

dc

m

2512 YU, PENG, XIE ET AL

3. Ide A, Kawasaki E, Abiru N, et al: Association between IL-18ene promoter polymorphisms and CTLA-4 gene 49A/G polymor-hism in Japanese patients with type 1 diabetes. J Autoimmun2:73, 20044. Omar MA, Hammond MG, Motala AA, et al: HLA class I

nd II antigens in South African Indians with NIDDM. Diabetes7:796, 19885. Banerji MA, Chaiken RL, Huey H, et al: GAD antibody

egative NIDDM in adult black subjects with diabetic ketoacidosisnd increased frequency of human leukocyte antigen DR3 andR4: Flatbush diabetes. Diabetes 43:741, 19946. Acton RT, Bell DS, Collins J, et al: Genes within and flanking

he major histocompatibility region are risk factors for diabetes,nsulin resistance, hypertension, and microalbuminuria in African-merican women. Transplant Proc 29:3710, 19977. David DS, Cheigh JS, Braun DW Jr, et al: HLA-A28 and

teroid-induced diabetes in renal transplant patients. JAMA 243:32, 19808. Sumrani NB, Delaney V, Ding ZK, et al: Diabetes mellitus

fter renal transplantation in the cyclosporine era: an analysis ofisk factors. Transplantation 51:343, 1991

9. von Kiparski A, Frei D, Uhlschmid G, et al: Post-transplantiabetes mellitus in renal allograft recipients: a matched-pairontrol study. Nephrol Dial Transplant 5:220, 1990

10. Hjelmesaeth J, Hartmann A, Kofstad J, et al. Glucosentolerance after renal transplantation depends upon prednisoloneose and recipient age. Transplantation 64:979, 199711. Nafar M, Pour-Reza-Gholi F, Amouzegar A, et al: IsLA-DR6 a protective factor against posttransplantation diabetesellitus? Transplant Proc 37:3098, 200512. Addous A, Mohamed AS, Ismail G, et al. Post-transplant

iabetes mellitus in kidney transplant recipients with special refer-nce to association with HLA antigens. Saudi J Kidney Disransplant 11:559, 200013. Yildiz A, Tütüncü Y, Yazici H, et al: Association between

epatitis C virus infection and development of posttransplantationiabetes mellitus in renal transplant recipients. Transplantation4:1109, 200214. Saxena S, Dash SC, Guleria S, et al: Post transplant diabetesellitus in live related renal allograft recipients: a single centre

xperience. J Assoc Physicians India 44:477, 199615. Tang YW, Zhang YH, Jia BX: Clinical features and high risk

actors of post-transplantation diabetes mellitus. Chinese J Urol0:95, 199916. Wang L, Shi L, Li X: Relationship between HLA-DR, -DQ

nd diabetes mellitus following renal transplantation. ChineseUrol 28:43, 200717. Wyzgal J, Paczek L, Sanko-Resmer J, et al: Insulin resistance

n kidney allograft recipients treated with calcineurin inhibitors.nn Transplant 12:26, 200718. Vesco L, Busson M, Bedrossian J, et al: Diabetes mellitus

fter renal transplantation: characteristics, outcome, and risk fac-ors. Transplantation 61:1475, 1996

19. Tueche SG: Diabetes mellitus after liver transplant: newtiologic clues and cornerstones for understanding. Transplant

roc 35:1466, 2003 i

20. Torres-Romero LF, Santiago-Delpı´n EA, de Echegaray S,t al: HLA is not predictive of posttransplant diabetes mellitus.ransplant Proc 38:914, 200621. Todd JA, Farrall M: Panning for gold: genome-wide scan-

ing for linkage in type 1 diabetes. Hum Mol Genet 5spec No:1443,99622. Schmidt D, Verdaguer J, Averill N, et al: A mechanism for

he major histocompatibility complex-linked resistance to autoim-unity. J Exp Med 186:1059, 199723. Lühder F, Katz J, Benoist C, et al: Major histocompatibility

omplex class II molecules can protect from diabetes by positivelyelecting T cells with additional specificities. J Exp Med 187:379,99824. Mosmann TR, Cherwinski H, Bond MW, et al: Two types ofurine helper T cell clone, I: definition according to profiles of

ymphokine activities and secreted proteins. J Immunol 136:2348,98625. Liblau RS, Singer SM, McDevitt HO: Th1 and Th2 CD4� T

ells in the pathogenesis of organ-specific autoimmune diseases.mmunol Today 16:457, 1995.

26. Larsen CP, Elwood ET, Alexander DZ, et al: Long-termcceptance of skin and cardiac allografts after blocking CD40 andD28 pathways. Nature 381:434, 199627. Singer SM, Tisch R, Yang XD, et al: An Abd transgene

revents diabetes in nonobese diabetic mice by inducing regulatorycells. Proc Natl Acad Sci U S A 90:9566, 199328. Poulin M, Haskins K: Induction of diabetes in nonobese

iabetic mice by Th2 T cell clones from a TCR transgenic mouse.Immunol 164:3072, 200029. Nepom GT: A unified hypothesis for the complex genetics ofLA associations with IDDM. Diabetes 39:1153, 199030. Deng H, Apple R, Clare-Salzler M, et al: Determinant

apture as a possible mechanism of protection afforded by majoristocompatibility complex class II molecules in autoimmune dis-ase. J Exp Med 178:1675, 1993

31. Harfouch-Hammoud E, Walk T, Otto H, et al: Identificationf peptides from autoantigens GAD65 and IA-2 that bind to HLAlass II molecules predisposing to or protecting from type 1iabetes. Diabetes 48:1937, 199932. Maioli M, Tonolo G, Bekris L, et al: GAD65 and IA-2

utoantibodies are common in a subset of siblings of Sardinian typediabetes families. Diabetes Res Clin Pract 56:41, 200233. Schnitzler M, Hardinger KL: Development of diabetes mel-

itus after renal transplantation and tacrolimus dose adjustment.ransplantation 78:113, 200434. Yu LX, Yao B, Du CF, et al: Clinical study of treatment

onversion from CsA to tacrolimus after kidney transplantationue to side effects of CsA. Chinese J Urol 24:169, 200335. Marcén R, Morales JM, del Castillo D, et al: Posttransplant

iabetes mellitus in renal allograft recipients: a prospective multi-enter study at 2 years. Transplant Proc 38:3530, 2006

36. Sezer S, Bilgic A, Uyar M, et al: Risk factors for develop-ent of posttransplant diabetes mellitus in renal transplant recip-

ents. Transplant Proc 38:529, 2006