To the Editor:

Acute graft-versus-host disease (aGVHD) is a major cause of morbidity and mortality after allogenic bone marrow transplantation (BMT) even in related BMT. Incompatibility at the major histocompatibility complex is the major factor of aGVHD. However, because aGVHD occurs in approximatively 30% of HLA-identical sibling transplants, molecules other than major histocompatibility complex (MHC), generally referred to as minor histocompatibility antigens, could be responsible for the developpement of aGVHD. Recently, Behar et al1 reported that the polymorphic adhesion CD31 molecule (leucine or valine at codon 125) acts as a minor histocompatibility antigen. Contradictory results were obtained by Nichols et al2 in a larger patient population. To test this hypothesis, we genotyped the CD31 molecule by sequence specific primers amplification according to a previously reported protocol1 in 96 HLA-identical sibling pairs and in 4 HLA-identical unrelated BMT. Our data showed similar proportions to those reported by Behar et al(1), of heterozygotes (Leu/Val), homozygotes (Leu/Leu), and (Val/Val) (respectively, 48.5%, 30.5%, and 21%). We did not find that CD31 mismatches between recipients and their HLA-identical donors were significantly associated with an increased risk of aGVHD (aGVHD grade I, II, III, and IV v grade 0: P = .83; aGVHD grade III and IV v 0: P = .16; or aGVHD grade II, III, and IV v 0 and I: P = .62). To comply with the patients' characteristics of the two previous studies,1,2 we excluded the four unrelated pairs, the cases with chronic or grade I and II aGVHD, and the patients with a follow-up less than 100 days (the numbers of sibling pairs without GVHD [n = 29] and aGVHD grade III/IV [n = 12] were similar to those in the study of Behar et al). Again, CD31 mismatches or incompatibilities for aGVHD (ie, the compatible recipient had no CD31 alleles that were foreign to the donor) (Table 1) were not significantly associated with an increased risk of severe aGVHD (P = .15 and P corrected = .65, respectively). Because children have GVHD less often than adults, we analyzed the data after excluding the recipients under the age of 18 years. Again, no significant difference was observed (P corrected = .89).

Table 1.

Relation of CD31 Polymorphism to the Occurrence of aGVHD

No aGVHD (n = 29)aGVHD Grade III, IV
(n = 12)
CD31 identity 19 (66%) 5 (42%) 
CD31 mismatch 10 
CD31 compatibility 23 (79%) 8 (66%) 
CD31 incompatibility 
No aGVHD (n = 29)aGVHD Grade III, IV
(n = 12)
CD31 identity 19 (66%) 5 (42%) 
CD31 mismatch 10 
CD31 compatibility 23 (79%) 8 (66%) 
CD31 incompatibility 

In conclusion, as with the data from Nichols et al, our results did not support the hypothesis that CD31 molecule can act as a minor histocompatibility antigen; however, as it has already been shown that peptides derived from male antigen H-Y and from the non-filament-forming class I myosin family were minor histocompatibility antigens,3 it clearly appears that the identification of clinically relevant minor histocompatibility antigens is very important to improve BMT.

REFERENCES

REFERENCES
1
Behar
E
Chao
NJ
Hiraki
D
Krishnaswami
S
Brown
BW
Zehnder
JL
Grumet
F
Polymorphism of adhesion molecule CD31 and its role in acute graft-versus-host disease.
N Engl J Med
334
1996
286
2
Nichols
WC
Antin
JH
Lunetta
KL
Terry
VH
Hertel
CE
Wheatley
MA
Arnold
ND
Siemieniak
DR
Boehnke
M
Ginsburg
D
Polymorphism of adhesion molecule CD31 is not a significant risk factor for graft-versus-host disease.
Blood
88
1996
4429
3
Den Haan
JMM
Sherman
NE
Blokland
E
Huczko
E
Koning
F
Drijfhout
JW
Skipper
J
Shabanowitz
J
Hunt
DF
Engelhard
VH
Goulmy
E
Identification of a graft versus host disease-associated human minor histocompatibility antigen.
Science
238
1995
1476