Table 3.
Distribution of HLA-C allotypes according to 3 models of HLA-C mismatching
| Patient’s nonshared allotype* | MFI | HLA-C mismatch model for the nonshared patient allotype N (%) | |||||
|---|---|---|---|---|---|---|---|
| Allele vs antigen† (N = 1971) | Residue 116‡ (N = 1955) | Residues 77/80§ (N = 1951) | |||||
| Allele (N = 389) | Antigen (N = 1582) | Matched (N = 847) | Mismatched(N = 1108) | Matched (N = 955) | Mismatched (N = 996) | ||
| C*07 | 111 | 104 (27)¶ | 288 (18) | 242 (29) | 147 (13) | 210 (22) | 179 (18) |
| C*03 | 114 | 238 (61)|| | 187 (12) | 268 (32) | 155 (14) | 321 (34) | 100 (10) |
| C*17 | 115 | 0 | 3 (<1) | 1 (<1) | 2 (<1) | 1 (<1) | 2 (<1) |
| C*05 | 154 | 5 (1) | 147 (9) | 50 (6) | 100 (9) | 43 (5) | 107 (11) |
| C*02 | 164 | 1 (<1) | 146 (9) | 31 (4) | 114 (10) | 43 (5) | 102 (10) |
| C*08 | 176 | 5 (1) | 28 (2) | 10 (1) | 22 (2) | 15 (2) | 17 (2) |
| C*16 | 180 | 10 (3) | 91 (6) | 43 (5) | 57 (5) | 29 (3) | 71 (7) |
| C*12 | 193 | 4 (1) | 118 (7) | 84 (10) | 36 (3) | 86 (9) | 34 (3) |
| C*04 | 200 | 5 (1) | 161 (10) | 10 (1) | 155 (14) | 39 (4) | 126 (13) |
| C*15 | 223 | 11 (3) | 111 (7) | 4 (<1) | 118 (11) | 51 (5) | 71 (7) |
| C*06 | 225 | 5 (<1) | 63 (4) | 36 (4) | 32 (3) | 25 (3) | 43 (4) |
| C*18 | 239 | 0 | 2 (<1) | 1 (<1) | 1 (<1) | 1 (<1) | 1 (<1) |
| C*01 | 254 | 0 | 152 (10) | 40 (5) | 112 (10) | 68 (7) | 84 (8) |
| C*14 | 294 | 1 (<1) | 85 (5) | 27 (3) | 56 (5) | 23 (2) | 59 (6) |
| Mean MFI of the patient’s mismatched HLA-C allotype | 123.2 | 176.7 | 148.6 | 179.8 | 154.2 | 177.4 | |
| P < .0001 | P < .0001 | P < .0001 | |||||
Patient allele mismatches as a group had significantly lower mean MFIs than patient antigen mismatches. The mean MFIs of the patients’ nonshared HLA-C allotypes were significantly different between allele and antigen mismatches, between residue 116 matches and residue 116 mismatches, and between residue 77/80 matches and residue 77/80 mismatches. Similar results were observed for donors’ nonshared allotypes (123.2 and 176.5, P < .0001 for allele and antigen mismatches; 147.0 and 179.9, P < .0001 for residue 116 matches and residue 116 mismatches; 150.5 and 180.2, P < .0001 for residue 77/80 matches and residue 77/80 mismatches). The mean MFIs of the shared matched allotypes, however, did not differ from one another (150.0 and 155.0, P = .06; 155.2 and 153.3, P = .44; 152.5 and 155.6, P = .15, respectively).
Patients’ mismatched (nonshared) allotypes are listed in order from lowest (C*07) to highest (C*14) MFI.22
HLA allele and antigens were defined according to WHO HLA Nomenclature.29 Four individuals each encoded novel HLA-C sequences that have not yet been characterized using serological reagents; these individuals were not included in the allele/antigen mismatch analysis.
Patients’ and donors’ nonshared HLA-C allotypes can be either matched or mismatched at residue 116. A total of 20 transplants were not included in the residue 116 analysis because they lacked sequence information at this position.
Patient’s and donors’ nonshared HLA-C allotypes can be either matched or mismatched at residues 77/80 that define the KIR C1 and C2 ligand groups. A total of 24 transplants were not included in the residue 77/80 analysis because they lacked sequencing information for residue 77 and/or 80 or did not have 77S-80N (C1) or 77N-80K (C2).
The most common patient-donor mismatch was C*07:01 vs C*07:02 or C*07:02 vs C*07:01; N = 79/104 (76%).
The most common patient-donor mismatch was C*03:03 vs C*03:04 or C*03:04 vs C*03:03; N = 216/238 (91%).