Anastas Pashov1,2,3,4, Thierry Calvez5, Laurent Gilardin1,2,3, Bernard Maillère6, Yohann Repessé7, Johannes Oldenburg8, A. Pavlova8, Srini V. Kaveri1,2,3,9, Sébastien Lacroix‐Desmazes1,2,3,9
1Centre de Recherche des Cordeliers Université Paris Descartes UMR S 872 Paris France
2Centre de Recherche des Cordeliers Université Pierre et Marie Curie‐Paris6 UMR S 872 Paris France
3Centre de Recherche des Cordeliers, INSERM, UMR S 872, Paris, France
4Department of Immunology, Institute of Microbiology, BAS, Sofia, Bulgaria
5Inserm et UPMC, UMR S 943, Paris, France
6CEA iBiTecS Service d'Ingénierie Moléculaire des Protéines Labex LERMIT et VRI Gif Sur Yvette France
7Laboratoire d'Hématologie, CHU de Caen, Caen, France
8Institute of Experimental Haematology and Transfusion Medicine, University Clinic Bonn, Bonn, Germany;
9International Associated Laboratory IMPACT (INSERM, France–Indian Council of Medical Research, India) National Institute of Immunohaematology Mumbai India
Tóm tắt
SummaryForty per cent of haemophilia A (HA) patients have missense mutations in the F8 gene. Yet, all patients with identical mutations are not at the same risk of developing factor VIII (FVIII) inhibitors. In severe HA patients, human leucocyte antigen (HLA) haplotype was identified as a risk factor for onset of FVIII inhibitors. We hypothesized that missense mutations in endogenous FVIII alter the affinity of the mutated peptides for HLA class II, thus skewing FVIII‐specific T‐cell tolerance and increasing the risk that the corresponding wild‐type FVIII‐derived peptides induce an anti‐FVIII immune response during replacement therapy. Here, we investigated whether affinity for HLA class II of wild‐type FVIII‐derived peptides that correspond to missense mutations described in the Haemophilia A Mutation, Structure, Test and Resource database is associated with inhibitor development. We predicted the mean affinity for 10 major HLA class II alleles of wild‐type FVIII‐derived peptides that corresponded to 1456 reported cases of missense mutations. Linear regression analysis confirmed a significant association between the predicted mean peptide affinity and the mutation inhibitory status (P = 0.006). Significance was lost after adjustment on mutation position on FVIII domains. Although analysis of the A1‐A2‐A3‐C1 domains yielded a positive correlation between predicted HLA‐binding affinity and inhibitory status (OR = 0.29 [95% CI: 0.14–0.60] for the high affinity tertile, P = 0.002), the C2 domain‐restricted analysis indicated an inverse correlation (OR = 3.56 [1.10–11.52], P = 0.03). Our data validate the importance of the affinity of FVIII peptides for HLA alleles to the immunogenicity of therapeutic FVIII in patients with missense mutations.
