Haplotype mismatches have been associated with higher risk of GVHD in patients receiving 8/8 HLA matched HCT. In addition a CIBMTR genome wide association study identified SNP mismatches within the MHC region in association with higher risk of GVHD. However, current standard of care relies only on matching at class I (A, B, C) and class II (DRB1) loci. The gamma block (GB) is located in the central MHC region between the genomic blocks that include HLA-C/B and DRB1/DQB1 and is the only major genomic block within the MHC not currently considered in HCT donor selection. We tested the hypothesis that mismatches in the GB increase the risk of GVHD.
Methods
We investigated 25 GB SNPs in 236 HCT recipient/unrelated donor pairs transplanted at our center (2000-2010). High resolution typing of HLA A, B, C, DRB1, DQB1, DPB1 loci and for MHC class I related chain A (MICA) was performed by sequence specific oligonucleotide probes and/or sequencing. GB SNP typing was performed by sequence specific primers. SNPs associated with severe acute GHVD (grades III-IV) at P<0.20 were combined into a composite variable. Multivariate analysis using Fine and Gray regression was used to identify risk factors for severe acute GVHD with adjustment for other significant risk factors.
Results
The study included AML (n=89), ALL (41), MDS (34), NHL (28) and other diagnoses (44). Stem cell source was BM (vs. PBSC) in 146 (62%) recipients. Disease status was early, intermediate and advanced in 98, 69 and 69 recipients, respectively. Conditioning intensity was myeloablative (vs. reduced intensity) in 176 (75%) recipients. GVHD prophylaxis was MTX based, MMF based and FK in 152, 74 and 10 recipients, respectively. HLA match was 8/8 and 10/10 in 184 (78%) and 179 (76%) pairs, respectively. HLA-DPB1 and MICA were matched in 44 (19%) and 197 (85%) recipients, respectively. Recipients included 137 (58%) males, 222 (94%) white. Mean age was 44±14 years.
SNPs c.2918+98G, c.3316C, and c.4385C (reference sequence C4A NG_011638.1) were associated with severe GVHD and analyzed as a composite variable. Mismatch at ≥1 of the 3 SNPs occurred in 49 pairs and 187 were matched across all 3 SNPs. SNP mismatch was associated with increased risk of severe GVHD in univariable analysis (HR 2.43, 95% CI 1.32-4.47, P= 0.004). There were no significant differences among SNP match and mismatch pairs regarding HLA match (10/10 vs. <10/10, 8/8 vs. <8/8, DPB1), MICA match, and patient, disease or transplant characteristics. SNP mismatch remained significantly associated with severe GVHD in a multivariable analysis (HR 2.54, P= 0.002) after adjusting for graft source, HLA and MICA mismatch.
Conclusion
GB SNP mismatch is associated with a higher risk of severe acute GVHD independent of the HLA and MICA match and other clinical risk factors. Avoiding GB SNP mismatch can potentially reduce the risk of severe acute GVHD and further investigation in larger cohorts is warranted.