GVL mismatch combinations
| Mismatch combination, donor-recipient . | n . | HR (95% CI) . | P . |
|---|---|---|---|
| Cw*0102-Cw*1402*† | 13 | ND | ND |
| Cw*0801-Cw*0102*† | 10 | ND | ND |
| Cw*1402-Cw*0304† | 20 | ND | ND |
| Cw*1502-Cw*1402 | 43 | 0.28 (0.09-0.88) | .030 |
| DPB1*0402-DPB1*0201* | 54 | 0.32 (0.12-0.87) | .026 |
| DPB1*0501-DPB1*0201* | 301 | 0.67 (0.50-0.91) | .009 |
| DPB1*0501-DPB1*0401* | 48 | 0.36 (0.13-0.98) | .046 |
| DPB1*0501-DPB1*0402* | 112 | 0.55 (0.33-0.93) | .026 |
| DPB1*0901-DPB1*0201* | 43 | 0.37 (0.14-0.96) | .042 |
| DPB1*1301-DPB1*0201*† | 20 | ND | ND |
| Mismatch combination, donor-recipient . | n . | HR (95% CI) . | P . |
|---|---|---|---|
| Cw*0102-Cw*1402*† | 13 | ND | ND |
| Cw*0801-Cw*0102*† | 10 | ND | ND |
| Cw*1402-Cw*0304† | 20 | ND | ND |
| Cw*1502-Cw*1402 | 43 | 0.28 (0.09-0.88) | .030 |
| DPB1*0402-DPB1*0201* | 54 | 0.32 (0.12-0.87) | .026 |
| DPB1*0501-DPB1*0201* | 301 | 0.67 (0.50-0.91) | .009 |
| DPB1*0501-DPB1*0401* | 48 | 0.36 (0.13-0.98) | .046 |
| DPB1*0501-DPB1*0402* | 112 | 0.55 (0.33-0.93) | .026 |
| DPB1*0901-DPB1*0201* | 43 | 0.37 (0.14-0.96) | .042 |
| DPB1*1301-DPB1*0201*† | 20 | ND | ND |
As an example of the mismatch combination analysis, the Cw*0102-Cw*1402 mismatch combination meant that the donor has HLA-Cw*0102, the recipient has HLA-Cw*1402 and another HLA-Cw allele of each donor and recipient was identical. Each mismatch pair in HLA-Cw was compared with the HLA-Cw allele match, and each mismatch pair in HLA-DPB1 was compared with the HLA-DPB1 allele match. All indicated results were concurrently significant in both the base analysis and validation analysis using bootstrap resampling.
ND indicates not determined.
Mismatch combinations that were not significantly associated with a higher occurrence of severe acute GVHD in our previous study.8 However, the Cw*0102-Cw*1402 mismatch combination has a marginal effect on the occurrence of severe acute GVHD; that is, Cw*0102-Cw*1402 was significantly associated with a higher occurrence of severe acute GVHD in base analysis, but not in validation analysis.
HR was not estimated due to the lack of an event in this group.