Authors
References
n
HLA loci
Evidence level
Remarks
Hoffmann et al.
[19]
20
A, B/DR
II
Beneficial
Boisjoly et al.
[11]
438
A, B
II
Beneficial
CCTS
[38]
419
A, B/DR/AB0
I
No effect, typing inaccurate
Hoffmann et al.
[18]
248
A, B/DR
II
Beneficial
Vail et al.
[39]
602
A, B/DR
II
A/B: beneficial; DR: adverse
Munkhbat et al.
[28]
81
DR, DQ, DP
II
Beneficial
Baggensen et al.
[5]
74
DR/DQ
II
Beneficial
Völker-Dieben et al.
[40]
1681
A, B/DR
II
Beneficial (i.e., DR)
Khaireddin et al.
[24]
459
A, B/DR
II
Beneficial
Bartels et al.
[6]
303
A, B
II
Beneficial
Reinhard et al.
[32]
48
A, B/DR
II
Beneficial
Reinhard et al.
[31]
418
A, B/DR
II
Beneficial
Böhringer et al.
[8]
1561
A, B/DR
II
A/B: beneficial; DR: depends on A/B matching
Prediction of the Waiting Time in HLA Matching
HLA matching inevitably prolongs the time on the waiting list. This is because all grafts but the first with very few HLA mismatches are rejected for the patient. Quality of life usually limits acceptable waiting periods to one year at maximum. The additional waiting period strongly depends on the HLA type. Patients with more common HLA phenotypes usually receive a match after few months. This is because their HLA alleles are also common among the donors. However, patients with a rare HLA alleles (i.e., when additionally homocygotic) may remain on the waiting list for years. It is nowadays possible to identify these patients in advance with a computer program and a database of the haplotype frequencies in the donor population [9]. This method is essential for discussing HLA matching with the patients as early as when discussing the indication of keratoplasty with them.
Evidence on Anti-HLA Antibodies in Keratoplasty
Anti-HLA antibodies had been originally detected in macro-agglutination assays. In this method, the patient serum is incubated with HLA-coated test erythrocytes. After adding patient serum, the test erythrocytes agglutinated in the presence of specific antibodies against that HLA allele. Nowadays, flow-based bead assays are used to detect anti-HLA antibodies. Donor-specific anti-HLA antibodies are presumed to deteriorate the prognosis of penetrating keratoplasty [15]. On the other hand, several other investigations failed to observe an antibody effect. A new method for reliable detection of these antibodies has recently strengthened the hypothesis that donor-specific anti-HLA antibodies play an important role in graft rejections after penetrating keratoplasty [35]. This is in line with the success of HLAMatchmaker in keratoplasty. This method (detailed in a previous section) is based on antibody epitopes. However, HLA crossmatching is still not performed as part of the clinical routine in keratoplasty. This is most likely due to lack of clear level I evidence at the time of writing.
Minor Histocompatibility Antigens (H Antigens)
Graft reactions may occur even when all HLA loci are perfectly matched. In some transplantation models, these graft reactions take a milder course in comparison to rejections caused by HLA mismatches. The underlying antigens have therefore been coined minor antigens [17]. Later, these have been identified as the aforementioned targets of indirect allorecognition embedded in MHC class II molecules on APCs. Another source of minor antigens are the intracellular fragments that are embedded into the HLA class I molecules of all nucleated cells. These convey a proteomic cellular fingerprint to the outer membrane. The antigens originate from somatic proteins that are constantly degraded by proteasomes. Proteasomes are organelles that recycle the amino acids of freshly synthesized and sorted out proteins by means of enzymatic fragmentation. Sometimes the proteasomes fuse with the endoplasmatic reticulum. Here, the peptides are placed in the binding groove of freshly synthesized HLA class I molecules with the help of tapasin. The endoplasmatic reticulum eventually fuses with the outer cell membrane and exposes the loaded HLA molecules to the aforementioned methods of allorecognition. It is important to note that each HLA allele has a specific repertoire of minor antigens that it can hold. This specificity is a consequence of the physical properties of its binding groove.
Discussion on Selected H Antigens
H-Y
The Y chromosome encodes several cytosolic proteins. These give rise to the H-Y group. Male grafts can thus be rejected by female recipients. H-Y antigens are supposedly expressed in the human cornea. H-Y antigens can be embedded into HLA A1 or HLA A2. A 20 % reduction of graft rejections was observed in 252 keratoplasties when avoiding the HLA A1/H-Y mismatch. In the same trial, the HLA A2/H-Y epitope was not relevant [10]. The prevalence of HLA A1 male donors is only 13 % in, e.g., Germany. For this reason, generally avoiding transplantation of male donors to female recipients does not make sense [22]. However, allocating male HLA-A1 donors to female recipients may be a discrete risk factor for immune reactions after penetrating keratoplasty.
HA-3
The HA-3 epitope is also HLA-A1 restricted. This epitope is derived from the lymphoid blast crisis (Lbc) oncoprotein and H antigen that has been expressed in the cornea. The HA-3 epitope comes in two alleles: VTEPGTAQY (HA-3 T) and VMEPGTAQY (HA-3 M). However, only grafting into the direction of HA-3 T is considered immunogenic. This does not seem to be highly relevant in penetrating keratoplasty, though [10].
Blood Group (ABO)
Blood group antigens are sometimes also referred as minor antigens. The allelic nature of the synthesizing enzyme gives rise to the ABO system. These are immunogenic glycoproteins attached to erythrocyte membranes but also present on a wide variety of human tissues. The ABO antigens are not physiologically expressed in the corneal stroma and corneal endothelium. However, they have been detected in failed corneal grafts [3].
The evidence on blood group matching in keratoplasty is controversial. A beneficial effect has been observed in high-risk penetrating keratoplasty, but not in normal-risk keratoplasties [12,23,38]. Other blood group antigens may play a role in normal-risk keratoplasties [33]. More research is warranted to work out the exact mechanisms in blood group histocompatibility. These retrospective nonrandomized results may well be confounded by peptidic H antigens originating, e.g., from the ABO-specific glycosyltransferases or other factors.
Conclusions and Recommended Clinical Practice
The recent rise of lamellar grafting certainly reduces the need for tissue typing in the clinical routine. It is nowadays possible, e.g., to replace a failed graft endothelium with Descemet membrane transplantation. This almost completely avoids subsequent rejection episodes. However, tissue typing still makes sense for specialty centers that deal particularly with high-risk transplantations in vascularized grafts or with limbal allografts. Here, all corneal donors should be HLA typed at least at the loci A, B, and DR. DNA typing is the method of choice. An alternative source of typed grafts is, e.g., Bio Implant Services, Leiden, the Netherlands. Blood groups may be additionally typed. The potential benefit from this is at a lower level of evidence, though. Lobbying is still required because costs from HLA typing of the donor are poorly reimbursed in most health systems. However, from the payer’s perspective, the additional cumulative costs from HLA typing have been recently calculated as low as 4.62 EUR per additional day of graft survival after penetrating keratoplasty [7].
Patients awaiting penetrating keratoplasty should be provided a histocompatible graft (HLA and AB0) whenever possible. This is especially true for high-risk keratoplasties. The expected time on the waiting list should be calculated and discussed with the patient in advance. The HLAMatchmaker algorithm can help in discriminating between “harmless” and “dangerous” mismatches to increase the donor pool and shorten the waiting time.