Clinical and laboatory study MHC class II antigens in alopecia areata 
K E Y 
WORDS 
Alopecia 
areata, HLA 
Class II 
antigens, 
DR,DQ 
JMHC cUlSs II antigens in awpecia areata 
M. Poljački, Z. Gajinov, B. Belic, S. Vojvodic and M. Matic 
ABSTRACT 
lntroduction. In accordance with assumption ot a possible autoimmune nature ot alopecia areata, the 
purpose ot this study was to investigate associations between alopecia areata and MHC class II anti-
gens (DR and DQ loci). 
Materials and methods. The study was performed on 23 patients with alopecia areata, classified in two 
groups, on the basis ot severity ot clinical picture: 13 alopecia totalis/universalis and 1 O unilocular/mul-
tilocular alopecia areata patients. 
MHC class II antigens were determined by double staining immunofluorescence. Seventeen antigens 
were determined tor DR locus, and 9 antigens tor DQ locus. For all antigens phenotypic frequency and 
relative risk were calculated. The Control group consisted ot 114 healthy tissue and organ donors. Chi 
square test tor small samples was used in statistical analysis. 
Results. Significantly higher frequencies ot DQ2, and smaller ot DR1 antigens were detected in the 
whole group. In alopecia totalis/universalis subgroup a significantly higher frequency ot the DQ2 and a 
smaller one ot DQ3(9) antigens were observed. In the unilocular/multilocular alopecia areata subgroup 
only a significant decrease ot the DQ3(9) frequency was tound. 
Conclusion. It can be assumed that an increased frequency ot DQ2 increases the risk ot severe forms ot 
alopecia areata while the absence ot DQ3(9) may have some protective role. Further studies on a larger 
number ot patients are however necessary. 
Introduction 
Alopecia areata (AA) is a non-cicatricial alopecia of 
unknown etiology. Findings of CD4+, CDS+ lympho-
cytes and CDl + cells in perivascular, peribulbar and 
intrafollicular infiltrate, an up-regulated expression of 
MHC class I molecules in hair follicles, hair follicle spe-
cific auto-antibodies, possible associations with immune 
disorders as well asa beneficial effect of lmmune-modu-
lating agents support the assumption that AA is an auto-
immune disorder. In accordance with such a supposi-
tion we decided to investigate the associations between 
AA and MHC class II antigens, the DR and DQ loci in 
patients from Vojvodina. 
Acta Dermatoven APA Vol 11, 2002, No 2 - --------- -------_)_) 
MHC class II antigen., in alopecia areata Clinical an d laboa to r y stucl y 
Table 1 . DO locus antigen frequency in total group of alopecia areata patients 
DQ antigens AA FFAA 
DQl 15 
DQ l (5) o 
DQl (6) o 
DQ2 10 
DQ3 6 
DQ3 (7) 3 
DQ3 (8) o 
DQ3 (9) o 
DQ4 o 
DQX 12 
''"'' - statistical significance p< 0,0001 
AA - alopecia areata patients group 
0,652 
o 
o 
0,434 
0,260 
0,130 
o 
o 
o 
0,520 
K 
91 
9 
4 
18 
51 
14 
2 
1 
8 
88 
FFK RR X2 
0,631 1,09 0,0021 
0,0625 o 2,99 
0,027 o 0,12 
0,125 5,45 11,51*" 
0,354 0,64 1,24 
0,097 1,39 0,01 
0,013 o 2,56 
0,006 o 3,45 
0,055 o 2,84 
0,611 0,69 1,08 
FFAA - phenotypic frequency in patients with alopecia areata 
K- control group of healthy organ ancl bloocl donors 
FFK - phenotypic frequency in control group 
RR - relative risk 
Materials and methods 
Twenty-three patients with AA were separated into 
two groups accorcling to the severity of clinical symp-
toms: 13 alopecia totalis/ universalis (AT/ U) ancl 10 
unilocular/multilocular (AU/M) AA patients. In 12 of the 
patients alopecia lasted more than one year, ancl in 11 
less than that. Familial occurrence of AA was not notecl 
in any of our patients , but in 3 patients other presum-
ably autoirnrnune clisorclers were identifiecl: vitiligo, 
prima1y hypothyreoiclism ancl psoriasis vulgaris, ancl 
insulin clependent diabetes mellitus. 
MHC class II antigens were determined by two-col-
o red immunofluorescent methocl, using immuno-
magnetic technique . 17 antigens were cletermined for 
DR locus, and 9 antigens far DQ locus. Phenotypic fre-
quency (PF) ancl relative risk (RR) were calculated for 
all antigens. A control group (K) consisted of 114 heal-
thy, tissue and organ donors. Chi square test (X2) far 
small samples was used for the statistical analysis 
(p<0.05 statistical significance) . 
Results 
In the whole group of AA patients a significantly 
higher frequency of DQ2 (RR=S.45; X2= 11.51) , and 
smaller frequency of DRl antigens (RR= 0.206; X2= 3.85) 
were detected. Table l. In AT/ U subgroup, significant 
higher frequency of DQ2 antigen (RR= 16.33; X2= 18.70) 
was confirmed, together w ith a lower frequency of 
DQ3(9) antigen (RR= O; X2= 5.29). Table 2. In AU/M 
subgroup only a significantly clecreasecl frequency of 
DQ3(9) antigen (RR= O; X2= 4.50) was founcl. Table 3. 
Discussion 
The HLA system has an important role in immune 
response and in autoimmunity regulation (1). By pre-
senting self and non-self antigens to the immune sys-
tem, it controls the production of T lyrnphocytes by 
determining their selection processes in the thymus. 
Genes of MHC class I, code for polymorphous mem-
brane glyco-proteins , presenting self and 11011-self pro-
teins to CDS+ lymphocytes, while class II MHC anti-
gens as membrane molecules, present self and non-self 
antigens to CD4+ T cells. Taking into consideration the 
facts that AA is genetically determined disease of auto-
immune nature, HLA typification can contribute to the 
hereditary aspect of this disorder (2 , 3). 
In the past investigations of a possible association 
between HLA and AA were directed towards HLA class 
I antigens (A and B loci). The results obtained were 
controversial. Kiant et al. fauncl a significantly higher 
frequency of HLA B12 antigen , while Hachman-Zadeh 
observed an increased frequency of the HLA B18 anti-
gen ( 4,5) . Kuntz et al. clid not confirm a statistically sig-
nificant correlation between HLA class I antigens and 
AA (6) . Similar results were obtained by Zlotogorski, 
who proved that even familial cases of AA were not 
associated with any of HLA antigens (7). Considering 
these and similar reports, it can be concluded that no 
consensus bas been reachecl, concerning a possible 
J6 ------- ------------------------ - ---Acta Dermatoven APA Vol 11, 2002, No 2 
Clinical and l ahoator y study MHC class II antigens in alopecia areata 
Table 2. DQ locus antigen frequency in alopecia totalis/ universalis group 
DQ antigens AT/UFFAT/U 
DQl 5 0,5 
DQl (5) o o 
DQl (6) o o 
DQ2 7 0,7 
DQ3 3 0,3 
DQ3 (7) 1 0,1 
DQ3 (8) o o 
DQ3 (9) o o 
DQ4 o o 
DQX 4 0,4 
* - statistical significance p< 0,05 
*''' - statistical significance p< 0,0001 
AT/U - alopecia totalis/ universalis 
K 
91 
9 
4 
18 
51 
14 
2 
1 
8 
88 
FFK RR x2 
0,631 0,5824 1,37 
0,0625 o 2,06 
0,027 o 2,44 
0,125 16,33 18,70** 
0,354 0,78 0,48 
0,097 1,031 0,27 
0,013 o 3,31 
0,006 o 5,29* 
0,055 o 2,26 
0,611 0,424 2,72 
FFAT/U - phenotypic frequency in patients with alopecia totalis/ universalis 
K- control group of healthy organ and blood donors 
FFK - phenotypic frequency in control group 
RR - relative risk 
correlation between AA and MHC class I system. 
Studies on HLA class II antigens in AA are more con-
sistent, especially for HLA DR4, DR5 and DQ3 antigens. 
Colombe et al. have confirmed a higher frequency of 
DQ3 (DQ7, DQ8), DR5 (DRll) , and DR4 antigens (8) 
According to their research, AT/ U group shows a sig-
nificant association with the DQ3- subtype DQ7, and 
DR5 (DRll), w hich is not the case in patients w ith a 
less severe form of the disease, or in alopecia lasting 
less than 6 months. Such results favor the connection 
of alopecia totalis to DQ3 (DQ7) and DRl 1 antigens. 
This is in accordance with Welsh's statement about DQ3 
antigen being a major marker of AA (9). According to 
Colombe et al. , specific aminoacid sequence in the epi-
thope common to ali DQ3 antigens, (through molecu-
lar mimicry) may be responsible for AA. Further epi-
thopes, such as DQ7 and DRl 1 exert influence on pro-
gression ofthe disease (8). Zhang's investigation on 55 
Caucasian patients confirmed an association between 
AA and DR4, but not with DR5 or DQ antigens (10). 
Table 3. DQ locus antigen frequency in alopecia unilocularis/ multilocularis group 
DQ antigens AU/M FFU/M K FFK RR 
DQl 10 0,769 91 0,631 1,941 
DQl (5) o o 9 0,0625 o 
DQl (6) o o 4 0,027 o 
DQ2 3 0,23 18 0,125 2,1 
DQ3 3 0,23 51 0,354 0,547 
DQ3 (7) 2 0,153 14 0,097 1,688 
DQ3 (8) o o 2 0,013 o 
DQ3 (9) o o 1 0,006 o 
DQ4 o o 8 0,055 o 
DQX 8 0,615 88 0,611 1,2727 
* - statistical significance p< O.OS 
AU/M - alopecia unilocularis/ multilocularis 
FFU/M - phenotypic frequency in patients with alopecia unilocularis/ multilocularis 
K- control group of healthy organ and blood donors 
FFK - phenotypic frequency in control group 
RR - relative risk 
X2 
0,17 
2,41 
2,33 
0,42 
1,44 
0,03 
2,95 
4,50* 
2,34 
0,07 
Acta Dermatoven APA Vol 11, 2002, No 2 ------------------J7 
MHC class II antigens in alopecia areata Clinical a nd laboatory study 
Our results for the whole group of AA patients support 
the role of the DQ2 antigen, w hich was also confirmed 
in AT/ U patients. An increased frequency of this anti-
gen was not confirmed in patients with less severe forms 
of the disease. A significantly decreased frequency of 
DQl was observed in the whole group of AA patients, 
and a decreased frequency of DQ3(9) in both AT / U and 
AU/ M groups. 
Conclusion 
It can be assumed that an increased frequency of 
DQ2 increases the risk of severe forms of AA, while the 
absence of DQ3 (9) may have a protective role. 
Further studies on a larger number of patients are 
however necessary. 
11.EFERENCES 
AUTHORS' 
ADDRESSES 
l. Oliveira DBG, Peters DK. The immunogenetic basis of autoimmunity. Autoimmunity 1990; 5: 293-306. 
2. Michael AJ. The genetic epidemiolO!,ry and autoimmune pathogenesis of alopecia areata. J Eur Acad 
Dermatol Venereol 1997; 9: 36-43. 
3. Galbraith GMP, Thiers BH, Vasily DB, Fudenberg HH. Immunological profiles in alopecia areata. Br J 
Dermatol 1984; 110: 163-70. 
4. Kianto U, Reunala T, KarvonenJ, Lassus A, Tiilikainen A. HLA-B12 in alopecia areata. Arch Dermatol 
1977; 113: 1716. 
5. Hachem-Zadeh S, Brautbar C, Cohen T. HLA and alopecia areata inJerusalem. Tissue Antigens 1981; 
18: 71-4. 
6. Kuntz B, Selze D, Braun-Falco O, Scholz S, Albert ED. HLA antigens in alopecia areata. Arch Dermatol 
1977; 113:1717. 
7. Zlotogorski A, Weinrauch L, Brautbar C. Familial alopecia areata: No linkage with HLA. Tissue Antigens 
1990; 35: 40-1. 
8. Colombe B W, Price VH, Khomy EL, Garovoy MR and Lou CD. HLA class II antigen associations help to 
define two types of alopecia areata. J Am Acad Dermatol 1995; 33 (5): 757-65. 
9. Welsh EA, Clark HH, Epstein SZ, Reveille JD, Duvic M. Human leukocyte antigen-DQB1*03 alleles are 
associated with alopecia areata. J Invest Dermatol 1994; 103: 758-63. 
10. Zhang L, Weetman P, Friedman PS, Oliveira DBG. HLA associations with alopecia areata. Tissue Anti-
gens 1991; 38: 89-91. 
Poljački Mirjana MD, PhD, Associate Projessor, Clinical Centre, Glinic 
far Dermatovenerologic Diseases, Hajduk Veljkova 1-3, 21000 Novi Sad, 
Yugoslavia 
Gajinov Zorica MD, MSc, same address. 
Belic Branislava MD, PhD, Assistant Projessor, Transjusiology 
Department, Hajduk Veljkova 1-3, 21000 Novi Sad, Yugoslavia. 
Vojvodič Svetlana MD, MSc, Transjusiology Department, same address. 
Matic Milan MD, Jv!Sc, Clinical centre, Clinicjor Dermatovenerologic 
Diseases, Hajduk Veljkova 1-3, 21000 Novi Sad, Yugoslavia 
J8 - - ----- - ---- - --------- - -----------Acta Dermatoven APA Vol 11, 2002, No 2