Two-locus Involvement in the Association of Human

Leukocyte Antigen with the Extrahepatic Manifestations of

Autoimmune Chronic Active Hepatitis
YANINA MARCOS,’ HUGOA. FAINBOIM,3 MONICACAPUCCHIO,l JORGEFINDOR,’JORGE DARUICH,~
RE YES,^ MARCEL0 PANDO,’ GRACIELA
BEATRIZ NORA MENDEZ,3 M. LEONARDO
DEL c . THEILER,~ SATZ’ AND
LEONARDO FAINBOIM~
‘Laboratorio de Inmunogenktica and 2Division Gastroenterologia, Seccidn Hepatologia, Hospital de Clinicas, Universidad
de Buenos Aires, and “Hepatopatias Infecciosas, Hospital F.J. Muiiiz, Buenos Aires, Argentina

We investigated the association of human leukocyte human leukocyte antigen DR6 was recently observed
antigen antigens and type 1 chronic active “autoim- in patients from the same ethnic group and geographic
mune” hepatitis in a population of 66 white Argen- region. The clinical and genetic heterogeneity ob-
tinian patients, taking into account the different served in this study may explain the weak human
manifestations of the disease. Standard microlympho- leukocyte antigen associations reported previously for
cytotoxicitywas used for human leukocyte antigen A, autoimmune chronic active hepatitis and suggest that
B, C, DR and DQ typing. Human leukocyte antigen class the extrahepatic forms in patients with autoimmune
2 alleles were also typed on genomic DNA by means of chronic active hepatitis represents a separate clinical
polymerase chain reaction amplification and hybrid- entity. (HEPATOLOGY 1994;19:1371-1374.)
ization to sequence specific oligonucleotides. A pri-
mary association with human leukocyte antigen DR4 Autoimmune CAH (A-CAH) is a progressive liver
was present (human leukocyte antigen DR4: 44% in disease characterized by piecemeal necrosis accom-
patients vs. 29% in controls; 9, 6.6; p = 0.02, relative
risk, 2.1). However, a novel association was observed panied by hyperglobulinemia,circulating autoantibodies
with human leukocyte antigen A l l (31%in patients vs. and marked female prevalence. In up to 48% of these
6%in the contmk $,26.3; corrected p = 0.001: relative patients extrahepatic manifestations are also observed
risk, 6.8). Moreover, of the 20 human leukocyte antigen ( 1).Since the first report by Mackay and Morris, several
A l l patients, 18 had extrahepatic manifestations asso- studies have demonstrated the association of A-CAH
ciated with autoimmune chronic active hepatitis. This with human leukocyte antigens (HLAs) (2-4). These
represented 60% of the patients bearing this form of studies, performed in an Anglo-Saxon population,
the diae~ar3e(n = 30), conferring a relative risk of 22.2 showed a primary association with HLA-B8 and HLA-
($,46.3; corrected p = 0.00008). In this group, human DR3 antigens. Donaldson et al. ( 5 )described a secondary
leukocyte antigen DR3 and DR4 had a weak asso- association with HLA-DR4 and revised all the conflicting
ciation. When present together, human leukocyte an-
tigen DR4 and human leukocyte antigen A l l had a results obtained when the association of HLA with
synergistic a e c t , yielding an odds ratio of 367. Statis- A-CAH was analyzed in different ethnic groups. In this
tical anal* and family segregation studies suggest study, we investigated the association of HLA with
that the two loci products may represent independent A-CAH in Argentina, taking into account the different
risk factorsfor this form of autoimmune chronic active manifestations of the disease. Only patients with clas-
hepatitis. This synergistic effect was not evident with sical or “lupoid type 1” A-CAH, characterized by the
A l l @US DR3. In autoimmune chronic active hepatitis presence of antinuclear antibody (ANA),smooth muscle
patients without extrahepatic manifestations, a weak antibody (SMA) or both were selected for this work. A
association with human leukocyte antigen DR6 was new association of A-CAH with two HLA locus products
found. Interestingly, in autoimmune chronic active was found in patients with extrahepatic manifestations
hepatitis of childhood (in which extrahepatic manifes-
tations a m seldom observed) a strong association with of this disease.

PATIENTS AND METHODS

Received August 30, 1993;accepted January 18, 1994. Patients. Sixty-fivewhite Argentinian patients (52females)
L.Fainboim and M.L.Satz are members of CONICET. M.L.Satz is a recipient with the diagnosis of type 1 A-CAH were referred from two
of a Biotechnologg Career Fellowship from the Rockefeller Foundation.
Addreas reprint requests to: Dr. L. Fainboim, Hospital de Clinicas, Av.
different liver units for the HLA study. All were seronegative
C6rdoba 2351,pis0 3 (1120)Buenos Aires, Argentina. for HBV and hepatitis C virus. None had history of drug or
Copyright 0 1994 by the American Association for the Study of Liver alcohol abuse. All were positive for ANA, SMA or both, with
Diseases. titers of 1 :80 or greater, aU had the histopathological picture
0270-9139194$3.00 + 0 3111/54878 of CAH and all were negative for liver-kidney microsomal
1371
1372 MARCOS ET AL. HEPATOLOGY
June 1994

1. Frequency of selected HLA markers in A-CAH patients

TABLE
Marker Patients (%) Controls n (%) RR P p Value p, Value
~~~ ~~ ~~ ~~~ ~~ ~ ~

AU A-CAH patients
HLA-A11 20 (31) 14 (6) 6.8 25.3 0.00002 0.001
HLA-DR3 16 (25) 37 (19) 1.6 1.6 NS NS
HLA-DR4 28 (44) 57 (29) 2.1 5.6 0.02 NS
HLA-DR6 16 (25) 34 (17) 1.7 2.3 NS NS
A-CAH patients with extrahepatic autoimmunity
HLA-A11 18 (60) 14 (6) 22.2 46.3 0.000001 0.00008
HLA-DR3 10 (36) 37 (19) 2.4 4.3 0.036 NS
HLA-DR4 14 (50) 57 (29) 2.4 4.9 0.025 NS
HLA-DR6 3 (11) 34 (17) 0.65 0.5 NS NS
A-CAH patients without extrahepatic autoimmunity
HLA-A11 2 (6) 14 (6) 1.1 0.025 NS NS
HLA-DR3 6 (18) 37 (19) 1.0 0.003 NS NS
HLA-DR4 14 (41) 57 (29) 1.7 2.1 NS NS
HLA-DR6 13 (38) 34 (17) 3.0 7.64 0.006 NS

All A-CAH patients: n = 65 (HLA-A,B,C), and n = 62 (HLA-DR,DQ); A-CAH patients with extrahepatic autoimmune diseases: n = 30
(HLA-A,B,C),and n = 28 (HLA-DR,DQ); A-CAH patients without extrahepatic manifestations, n = 35 (HLA-A,B,C) and n = 34 (HLA-DR,DQ).
Controls: n = 231 (HLA-A,B,C); n = 197 (HLA-DR,DQ).

antibodies. Autoantibodies were tested by means of indirect complement-dependent microcytotoxicity assay using sera
immunofluorescenceusing appropriate rat tissues. At presen- from the X and XI International Histocompatibility Work-
tation of the disease, all but two patients (one with acute liver shops supplemented with local sera. HLA-DR typing was
failure and one who had previously received corticosteroids) performed with DR-positive cells isolated by use of immuno-
were carefully studied for the presence of associated au- magnetic beads (Dynabeads; Dynal Inc., Great Neck, NY).
toimmune diseases as follows. Patients were referred to the Antigen frequencies were compared against those of a panel
rheumatology department for evaluation of joint diseases. comprising 231 (HLA-ABC)and 197 (HLA-DR)normal indi-
Thyroid involvement was assessed on the basis of the levels of viduals of the same ethnic group.
TSH, triiodothyronine and thyroxine (by RIA) and the HLA C i b s ZZ DNA Icsping. HLA class I1 DNA typing was
presence of antibodies to thyroglobulin, thyroid epithelial performed in accordancewith protocols of the XI International
microsomal antigens or both (by means of indirect immuno- Histocompatibility Workshop (XI-IHW). Polymorphic do-
fluorescence).Coomb’s tests were performed in three patients mains of DRB, DQB and DQA loci were amplified from genomic
with anemia. The diagnosis of Sjogren’s syndrome in a patient DNA with the set of primers designed for the XI-IHW.
with subjective dryness of mouth and eyes was based on the Amplified DNA was dotted onto nylon membranes and
Schirmer test. Patients with abnormal urinary sediment were hybridized to 32P-labeledoligonucleotides probes distributed
further studied for immunological renal disease. Special care for the XI-IHW studies. Frequencies in patients were com-
was taken in the diagnosis of skin-associated diseases and type pared with those from a panel of 55 normal individuals of the
1diabetes. same ethnic group.
Extrahepatic manifestations were present in 30 patients (28 Statistics. HLA antigen frequencies in A-CAH and control
females; mean age, 43.0 yr; range, 16 to 68 yr). These included panels were compared for significant differences by means of
Coomb’s positive hemolytic anemia (3 patients), pericarditis x2 analysis. We estimated the strength of association by
and pleuritis (1 patient), Sjogren’s syndrome (1 patient), calculating odds ratios and expressed these values as relative
thyroid disease with thyroid autoantibodies (7 patients), risks (RRs). The method of Woolf-Haldanewas used for these
membranoproliferative glomerulonephritis (1patient), photo- calculations (6). For the previously reported associations
sensitivity (1 patient), psoriasis vulgaris (1 patient) and between A-CAH with DR3, DR4 (2-5), and DR6 (Fainboim L,
arthritis (18 patients). In this group, three patients had more et al., Human Immunol 1994: in press), the p values derived
than one extrahepatic disease. A second group of 35 A-CAH from x2 statistics were taken as significant without further
patients (24 females; mean age, 38.6 yr; range, 17 to 73 yr) correction. For the new associations found here, we corrected
showed only chronic liver disease without any detectable p values (P,) by multiplying for the total number of com-
extrahepatic manifestations. parisons involved (77 antigens). Analysis of the combined
At this writing, the patients have been followed in the clinic effect of two different loci on susceptibility to A-CAH was
for periods between 3 mo and 15yr (mean, 10 yr), during which performed by an analysis similar to that described by Svejgaard
time all but one has been treated with various combinations of et al. (7, 8).
prednisolone and azathioprine or prednisolone alone. All but
one have had sustained remission on this therapy throughout. RESULTS AND DISCUSSION
One patient first seen with cirrhosis died as a consequence of
portal hypertension during the follow-up period. A similar The analysis of the frequencies of HLA-A, B, C, DR
response to this standard therapy was observed in patients and DQ antigens in the 65 A-CAH patients studied
with or without extrahepatic manifestations. confirmed the previously described increased frequency
TiSam Icsping. HLA-ABC (65 patients) and DR,DQ (62 of HLA-DR4 in other populations (Table 1).However, no
patients) phenotyping was performed by means of standard association with HLA-DR3 and only a very weak
 Vol. 19, No. 6, 1994
HEPATOLOGY MARCOS ET AL. 1373

TABLE
2. Odds ratios for the effects of MA-A11 with MA-DR4 or DR3 in the extrahepatic manifestation of A-CAB
Combination No. of patients No. of controls Odds ratio Xp Value p. Value

HLA-A11AUA-DR4
+I+ 10 0 357.0 15.4 0.0003
+I- 7 13 9.4 14.5 0.0004
-/+ 4 57 1.3 0.2 NS
-I- 7 127 - - -
HLA-A1lkILA-DR3
+I+ 7 5 24.5 23.3 0.00003
+/- 10 8 22.2 28.5 0.00001
-I+ 3 32 1.9 1.0 NS
-I- 8 152 - - -

“Patients” refers to patients with A-CAH showing extrahepatic manifestations. HLA-DR typing was not available for two patients (one with
HLA-All); therefore this analysis was performed in 28 patients. A total of 13 HLA-A11 controls was included. (DR,DQ typing was not available
in one additional A l l control.)
Odds ratios and x2 and p, values were calculated by comparison of the combination in question with a - I - combination.

association with DR4 was present in our population. A HLA-A11 and DR4 were present in separate haplotypes,
new primary association with HLA-A11 was found (31% suggesting that the increased risk provided by these two
in patients vs. 6% in controls; RR, 6.8; P, = 0.001). antigens would not involve extended haplotypes.
Interestingly, 18 of the 30 patients (60%) with extrahe- To analyze whether a particular HLA-DR4 allelic
patic manifestations had HLA-A11 (x2, 46.3; p, = variant was involved in this association, DR4 subtypes
0.00008; RR, 22.2). In this clinically different group, a were studied by means of DNA amplification and
weak association with HLA-DR3 was also observed, and oligotyping in 14 DR4 patients with A-CAH, including
the frequency of HLA-DR4 remained significantly in- patients with and without extrahepatic manifestations.
creased (Table 1). In the group of A-CAH patients On comparing patient data with a panel of 19 DR4
without any detectable extrahepatic manifestations, normal individuals of the same ethnic group, no sig-
only a sigdicant association with HLA-DR6 was found. nificant differences were found between patients and
Interestingly, a recent study of our laboratory in a group controls (DRB1*0401, 14% patients vs. 5% controls;
of 52 pediatric A-CAH patients showed a strong asso- 0402,0% vs. 16%; 0403,7% vs. 16%; 0404,21% vs. 26%;
ciation with HLA-DR6 (DRB1*1301 haplotype), with an 0405,29% vs. 10%; 0406,0% vs. 5%;0407,14% vs. 21%;
RR of 16.2 (Fainboim L, et al., Human Immunol 1994: 041 1,14%vs. 0%).This finding is in agreement with the
in press). It is worth mentioning that in this group of DR4 association reported in A-CAH patients of Japanese
children, only two patients had extrahepatic manifesta- origin (9, 10). Also, the DQA and DQB alleles were
tions. Taken together, all the data support the idea that investigated in this group of DR4 patients so that we
A-CAH patients with extrahepatic manifestations may might analyze their extended haplotypes. The results
define a separate clinical entity with a different genetic showed that DQB1*0301 was present in 5 of 15 normal
predisposition. DR4 individuals and in 1 of 13 DR4 A-CAH patients.
As mentioned above, A-CAH patients with extrahe- HLA-DQB1*0302 was present in 10 of 13 patients (vs. 9
patic manifestations had increased frequencies of HLA- of 15 controls). These differences were not statistically
A l l , DR3 and DR4. To assess a possible synergistic different.
effect of the class I and class I1 HLA loci in determining In conclusion, a new and strong association of
susceptibility to this form of the disease, we followed the HLA-A11was present in A-CAH patients with extrahe-
approach described by Svejgaard et al. and Morling et al. patic manifestations. Although HLA-DR4 was asso-
(6,7) (Table 2). Whereas HLA-A11 alone (in the absence ciated with A-CAH, this class I1 product was a weak
of DR4) conferred high susceptibility (odds ratio, 9.41, primary risk factor for this form of the disease; however
HLA-DR4 alone (in the absence of HLA-All) was not a it showed a strong synergistic effect when present with
risk factor for the extrahepatic forms of A-CAH (odds A l l . This observation may reflect the fact that extra-
ratio, 1.3). The joint expression of HLA-A11 and hepatic forms of disease in patients with A-CAH rep-
HLA-DR4 showed a dramatic synergistic effect in resent a separate clinical entity. On the other hand, it is
conferring susceptibility to extrahepatic A-CAH. Indi- evident that in different ethnic groups A-CAH is
viduals positive for HLA-A11 and HLA-DR4 were highly associated with various HLA antigens: DR3 in Anglo-
predisposed to extrahepatic A-CAH, with an odds ratio Saxon patients, DR4 in Japanese patients and A l l in
of 357, a value si&cantly higher than that obtained by Latin American Caucasian individuals. It would be
multiplying the independent risk factors obtained with interesting to reevaluate the data obtained in other
HLA-A11 and DR4. This synergistic effect was not ethnic groups to see whether the HLA-DR3 and DR4
observed for HLA-AllLHLA-DR3 (Table 2). associations are restricted to either group of patients
It was possible for us to study segregation of HLA (with or without extrahepatic disease). This is relevant
antigens in one family, and our study showed that when one takes into account the fact that all the relative
1374 MARCOS ET AL. HEPATOLOGY
June 1994

risks obtained with these alleles reported in the lit- 3. Opelz G, Vogten AJ, Summerskill WH, Schalm SW, Terasaki PI.
erature are not as strong as those found in our study HLA determinants in chronic active liver disease: possible relation
with A l l and A-CAH patients with extrahepatic mani- 4. of HLA-Dw3 to prognosis. Tissue Antigens 1977;9:36-40.
Mackay IR, Tait BD. HLA association with autoimmune-type
festations. chronic active hepatitis: identification of B8-DRw3 haplotype by
In support of our observation, Czaja et al. have family studies. Gastroenterology 1980;79:95-98.
reported preliminary findings also showing an increase 5. Donaldson PT, Doherty DG, Hayllar KM, McFarlane IG, Johnson
of HLA-A11 in A-CAH patients with multiple extrahe- PJ, Williams R. Susceptibility to autoimmune chronic active
hepatitis: human leukocyte antigens DR4 and Al-Bg-DR3 are
patic immunological manifestations (11). One could independent risk factors. HEPATOLOGY 1991;13:701-706.
speculate that autoreactive CD4 T cells triggered in the 6. SvejgaardA, Ryder LP. Associationsbetween HLA and disease. In:
context of HLA-DR4 would interact with A11-specific Dausset J, Svejgaard A, eds. HLA and disease. Copenhagen:
CD8 lymphocytes recognizing a cross reactive autoan- Munksgaard, 1977:46-71.
Svejgaard A, Ryder LP. HLA genotype distribution and genetic
tigen present in liver and other target tissues, mainly 7. models of insulin-dependent diabetes mellitus. Ann Hum Genet
thyroid and joints. 1981;45:293-298.
The relevance of the involvement of two HLA loci in 8. Morling N, Friis J, Fugger L, Svejgaard A. DNA polymorphism of
the pathogenesis of certain types of autoimmune dis- HLA class I1 genes in pauciarticular juvenile rheumathoid
eases shown in this study is supported by two recent arthritis. Tissue Antigens 1991;38:16-23.
Seiki T, Kiyosawa K, Inoko H, Ota M. Association of autoimmune
observations made in different ethnic groups: Graves’ 9. hepatitis with HLA-Bw54 and DR4 in Japanese patients.
disease in Japanese (12) and pauciarticular juvenile HEPATOLOGY 1990;12:1300-1304.
arthritis in white patients (Fernindez-Vifia M, Personal 10. Ota M, Seki T, Kiyosawa K, Furuta S, Hino K, Kondo T,
communication, July 1993). Fukushima H, et al. A possible association between basic amino
acids of position 13 of DRBl chain and autoimmune hepatitis.
Acknowledgment: We thank A. Baker for his assis- Immunogenetics 1992;36:49-55.
tance with the statistical analyses. 11. Czaja AJ, Carpenter HA, Santrach PJ, Moore SB. Extrahepatic
immunologid features of chronic active heDatitis: genetic Dredis-
REFERENCES positions &d prognostic implications [Abstractf HEPAGOLOGY
1992;16:19OA.
1. Czaja AJ. Natural history, clinical features, and treatment of 12. Dong R-P, Kimura A, Okubo R, Shinagawa H, Tamai H,
autoimmune hepatitis. Semin Liver Dis 1984;4:1-12. Nishimura Y,Sasasuki T. HLA-A and DPBl loci confer suscepti-
2. MacKay IR, Moms PJ. Association of autoimmune chronic bility to Graves’ disease. Hum Immunol 1992;35:165-172.
hepatitis with HL-A1,8. Lancet 1972;2:793-795.

BOUND VOLUMES AVAILABLE TO SUBSCRIBERS

Bound volumes of HEPATOLOGY for 1994 are available to subscribers only. They may
be purchased from the publisher at a cost of $90 for domestic, $119.30 for Canadian and
$113 international for Vol. 19 (January to June) and Vol. 20 (July to December). Price
includes shipping charges. Each bound volume contains a subject and author index, and all
advertising is removed. Copies are shipped within 60 days after publication of the last issue
of the volume. The binding is durable buckram with the journal name, volume number, and
year stamped in gold on the spine. Payment must accompany all orders. Contact
Mosby-Year Book, Inc., Subscription Services, 11830 Westline Industrial Drive, St. Louis,
MO 63146-3318,USA, phone (800)453-4351, or (314)453-4351.
Subscriptions must be in force to qualify. Bound volumes are not available in place of
a regular HEPATOLOGY subscription.