Summary of the article

Christoph Frohn et al study 2001
Retrospective Study
The aim of the study was to determine whether HLA-C is an independent factor for acute rejection. HLA match status and occurrence of acute rejection episodes were retrospectively examined by analysing 106 consecutive cadaveric kidney donor/recipient pairs.

Study Subjects and Methods: 
The Study Population  were 106 consecutive cadaveric kidney donor/recipient pairs from their renal transplantation unit from 1995 to 1997. The mean follow-up time was 42 months. 

• One pair was excluded because of severe perioperative complications(cardiac infarction), which led to exclusion from the standard immunosuppressive treatment, and one patient died intra-operatively. 
• Eighty-five out of 104 patients (81%) were transplanted for the first time.
• Patients receiving their first transplant were treated initially with 7.5 mg/kg cyclosporin; 2 mg/kg azathioprine, and 500, 250, 125 and 20 mg methylprednisolone. 
• Patients receiving a second transplant were also treated with (ATG) Doses were adjusted according to T-cell counts and graft function.
• Rejection episodes were treated initially with methylprednisolone, and additional antibody-based therapy was per- formed if necessary (17/35 cases). 
• The rejection therapy avoided acute graft loss in 30/48 cases.

The Study Methods: 

• In all individuals, complete HLA class-I typing was performed using a complement-mediated cytotoxicity assay with 174 test sera according to National Institutes of Health(NIH) standards on EDTA-anticoagulated blood.
• To ascertain HLA-C type, frozen material was analysed using a commercially available PCR kit, which discriminates 18 currently known broad specificities (HLA-C low resolution SSP, Dynal, Sweden). 
• A self-designed, group- specific PCR protocol was also used that discriminated two alternative amino acids at positions 77 and 80 of the HLA-C locus to ascertain the results obtained by allele-specific PCR. 
• HLA-DR type was also ascertained by genotyping in all cases at the time of transplantation (Dynal DR high resolution PCR kit). 
• Due to the retrospective design of the study, no frozen material from 34 donor/recipient pairs was available to perform HLA-C genotyping. Sixteen had two different HLA-C alleles; it was therefore not possible that a serologically undetectable allele was overlooked in these cases.
• Considering the above fact and the very low discrepancy between serology typing and genotyping for HLA-C in centre laboratory (19 out of 318 genotyped alleles were mistyped by serology, whereby 18 mistypings have been the result of the non-existence of a serological equivalent for these alleles w12x), these patients were included in the study. The 18 resting pairs, for which DNA was not available and in which a second undetectable allele was possible, were analysed separately. Mismatches were always defined on the serological split level, thus HLA-C/-DR subsplit discrepancies were not regarded a mismatch, but HLA-A or -B split discrepancies were. 
• If a rejection episode was clinically suspected a biopsy was performed immediately. Histological evaluation was performed by the local pathologist according to the Banff scheme. 
• To exclude CMV infection as the cause of graft function loss, pp65 early antigen was tested on lymphocyte smears counting 400 000 cells. 
• The following parameters were also initially considered in the analysis: mismatches on the HLA-A, -B or -DR loci; time of ischaemia; age and gender of donor/recipient; pre- existence of cytotoxic antibodies; and underlying diagnosis leading to renal failure.
• All patients stayed in hospital until blood creatinine was stabilized to a constant level. The creatinine level at the day of discharge was also documented as an indicator of graft function.

Study Statistics 

• First, a univariate analysis was performed for all considered variables. 
• Evaluation of all resulting cross tables was performed by Fisher’s exact test using the SPSS software program. 
• The influence of the time of cold and warm ischaemia was calculated by the Wilcoxon ranked sums test. 
• A graft survival analysis was also performed by calculating a Kaplan±Meier blot and log rank statistics (2 degrees of freedom) using SPSS software. 

Study Results 
HLA typing:

• Genotyping for HLA-C revealed unambiguous results in all cases. 
• Differences between serotyping and genotyping were observed in 18/140 (13%) genotyped individuals and were always due to alleles for which no serological equivalent exists (most frequently HLA-C *12xx). 
• Serotyping for HLA-C never indicated alleles that did not actually exist but sometimes overlooked alleles, especially those with no serological equivalent. 
• As expected, HLA-A/-B mismatch was strongly correlated with HLA-C mismatch (P< 0.001) as a consequence of linkage disequilibrium. 

Rejection-univariate analysis:

• Time of warm and cold ischaemia, number of previous transplantations and gender match had no demonstrable influence on acute rejection episodes. 
• HLA-B mismatch was weakly correlated with the rejection probability (P=0.1). 
• HLA-A and -DR had no influence on rejection, and the same was observed for all other considered variables. 
• The strongest influence on acute rejection as determined by univariate analysis was exerted by HLA-C (P=0.002), whereby the cut-off point was between 0 or 1 and 2 mismatches. 

Analysis of the independent influence of HLA-C:

• Within the subgroup of patients showing one HLA-B mismatch , HLA-C was found to exert a significant influence (P=0.004, or P=0.022) if only pairs with HLA-C determination on a genomic level were considered on the probability of acute rejection episodes. 
• No influence of C mismatch was observed in the subgroup of patients with no additional B mismatches . 
• In summary, the HLA-C mismatch status is an independent factor in acute rejection, at least when paired with additional B mismatches. 

Graft survival:

• There was a tendency towards better graft survival in the HLA-C-matched pairs. This effect was again more pronounced in pairs with one additional B mismatch, nearly reaching significance when pairs with no and two HLA-C mismatches were compared in this subgroup (P=0.055). 

Creatinine: 

• The creatinine level at the time of discharge was not influenced by either HLA-A/-B or HLA-C mismatching. 

Study Discussion:
Despite the limited number of cases, univariate analysis revealed a statistically significant correlation with rejection probability for HLA-C and a trend for HLA-B.

By subgroup analysis the independent contribution of the C locus could be established. The C effect could not be demonstrated in those pairs without additional B mismatches. Apart from problems due to the relatively small observation cohort, immunological reasons may also exist, presumably a form of `hyper-additive’ effect of the stimulation of B- and C-specific T-cell receptors which may explain the latter observation. All other variables, including HLA-A and -DR mismatch, revealed no correlations. 

The most remarkable finding was the significant correlation between HLA-C mismatching and rejection. This factor has never been considered in clinical practice and has never before been analysed in the context of solid organ transplantation. 

• This may in part be due to the fact that serological typing has been considered to be less reliable for the HLA-C locus than for other HLA loci. 
• Moreover, the discrepancy between serotyping and genotyping does not exceed 7.5% if sufficient HLA-C sera are used in the standard NIH cytotoxicity test.
• Another reason for the lack of interest in HLA-C may be that its expression on cell surfaces is quite low.

Subsplit mismatches may appear if serotyping is replaced by genotyping (e.g. *4402/*4403; *3901/*3906) but have never been reported to exert an in ̄uence on rejection. 

The B and C loci are in close proximity to one another on chromosome 6; therefore the linkage disequilibrium is very strong. This phenomenon is also reflected by our data, which show a highly significant correlation between HLA-B and -C match status. As we were not able to perform genotyping for HLA-B for logistical reasons, one might argue that HLA-C may only be an indicator for additional B mismatches which have been over- looked by serological typing. However, the problems of serotyping are much less important for HLA-B alleles than reported for HLA-DR alleles.

Comparing this Study with other studies 

• Many studies in large patient cohorts have shown a positive effect of HLA matching on renal allograft survival time.
• The undesirable nature of the longer allocation procedure was more than compensated for by the benefits of matching.
• The occurrence of acute rejection episodes as confirmed by histopathological diagnosis was not analysed in any of these studies. 
• The impact of HLA matching on acute rejection episodes of renal allografts has been examined specifically in a few smaller studies.
• This study monitored matching effects for two reasons: (i) acute rejection is an a priori immunological phenomenon and thus well suited for monitoring the effects of HLA matching; (ii) around 50% of all cases of graft function loss are actually caused by immunological rejection, despite the fact that immunosuppressive regimes are given in such cases.
• The data presented in other reports are consistent with this study data to the extent that a correlation was observed between HLA mismatch and the incidence of rejection episodes.
• With respect to the individual HLA loci involved, results are contradictory. In the studies cited, the HLA-DR locus has the greatest influence, followed by HLA-B. In contrast, this study’s data do not demonstrate any influence of HLA-DR match-not even a tendency-on rejection probability. One explanation for this discrepancy might be the study’s local matching policy, which places high priority on matching for DR(60% no mismatch; 34% one mismatch; 6% two mismatches), thus the number of DR-mismatched donor/recipient pairs is probably too small to result in a statistically significant clinical DR-effect.