Proportional hazards regression
| Factor . | RR . | 95% CI . | P . | Favorable . |
|---|---|---|---|---|
| Grades III-IV acute GvHD* | ||||
| HLA-A, HLA-B allele match† | 1.00 | Reference | NA | NA |
| Minor mismatch | 1.52 | 1.21-1.91 | .001 | Match |
| Major mismatch | 1.45 | 1.21-1.91 | .005 | Match |
| HLA-C allele match | 1.00 | Reference | NA | NA |
| HLA-C single allele mismatch | 1.40 | 1.19-1.63 | < .001 | C allele match |
| Minor vs major | 1.04 | 0.76-1.43 | .784 | Neither |
| Overall survival‡ | ||||
| HLA-A, HLA-B allele match | 1.00 | Reference | NA | NA |
| Minor mismatch | 1.51 | 1.23-1.86 | < .001 | Match |
| Major mismatch | 1.40 | 1.19-1.64 | < .001 | Match |
| HLA-C allele match | 1.00 | Reference | NA | NA |
| HLA-C single allele mismatch | 1.25 | 1.12-1.40 | < .001 | C allele match |
| Minor vs major | 1.08 | 0.85-1.37 | .525 | Neither |
| Factor . | RR . | 95% CI . | P . | Favorable . |
|---|---|---|---|---|
| Grades III-IV acute GvHD* | ||||
| HLA-A, HLA-B allele match† | 1.00 | Reference | NA | NA |
| Minor mismatch | 1.52 | 1.21-1.91 | .001 | Match |
| Major mismatch | 1.45 | 1.21-1.91 | .005 | Match |
| HLA-C allele match | 1.00 | Reference | NA | NA |
| HLA-C single allele mismatch | 1.40 | 1.19-1.63 | < .001 | C allele match |
| Minor vs major | 1.04 | 0.76-1.43 | .784 | Neither |
| Overall survival‡ | ||||
| HLA-A, HLA-B allele match | 1.00 | Reference | NA | NA |
| Minor mismatch | 1.51 | 1.23-1.86 | < .001 | Match |
| Major mismatch | 1.40 | 1.19-1.64 | < .001 | Match |
| HLA-C allele match | 1.00 | Reference | NA | NA |
| HLA-C single allele mismatch | 1.25 | 1.12-1.40 | < .001 | C allele match |
| Minor vs major | 1.08 | 0.85-1.37 | .525 | Neither |
NA indicates not applicable.
Acute GvHD model was stratified on disease, T depletion, and year of transplantation. It was adjusted for radiation in the conditioning regimen (P = .311) and cell source (P ≤ .001).
No significant interaction was detected between HLA-C and CREG matching (grades III-IV GvHD, P = .53; overall survival, P = .25).
Overall survival model was stratified on patient age and sex match. It was adjusted for race (P = .002), donor-patient CMV status (P < .001), Karnofsky/Lansky score at transplantation (P < .001), disease (P < .001), and disease stage (P < .001).