Susan Martin Mark S. Davies Amanda J. Robson

S. Martin (IXI) . M. S. Davies A. J. Robson NW Regional Tissue Typing Laboratory, St. Mary’s Hospital, Hathersage Road, Manchester M I 3 OJH, UK

A retrospective flow cytometric crossmatch study in transplant recipients with autoreactive lymphocytotoxic anti bodies

Abstract Our previous data shows renal transplant recipients with autoreactive lymphocytotoxic antibodies to have a reduced transplant survival when compared to patients without autoantibodies. This could have been due to the presence of weak IgG antibodies inhibited by the dithiothreitol used to remove IgM antibodies in the pretransplant cytotoxicity cross- match. That possibility was in- vestigated in a retrospective study of 52 recipients of 57 renal trans- plants who were recrossmatched using a more sensitive flow cyto- metry crossmatch (FCXM) to detect recipient IgG antibodies to donor T and/or B cell splenic lympho- cytes. Fourteen of the 57 (24%)

transplants failed. Six losses were within the 1st month posttransplant and four of these were immunolog- ical failures. None of the transplant failures had a positive pretrans- plant FCXM. These results showed that the recipients with autoanti- bodies did not have pretransplant IgG anti-donor antibodies. The transplant failures did not, there- fore, relate to the presence of antibodies undetected by the dithiothreitol-treated cytotoxicity crossmatch.

Key words Autoreactive lymphocytotoxic antibodies Autoantibodies . Renal transplant survival . Flow cytometry crossmatch

Introduction

Autoreactive lymphocytotoxic antibodies are often identified in the sera of patients awaiting renal transplan- tation. They can exhibit different reactivity profiles, reacting with normal B lymphocytes, normal T and B lymphocytes or T and B lymphocytes and lymphocytes from patients with B cell chronic lymphocytic leukaemia (CLL) [I]. These different reactivity profiles may reflect a difference in sensitivity of the target cells with sensitivity increasing in the order, CLL cells, T cells, B cells.

It has become accepted that these autoreactive anti- bodies are not associated with transplant failure even

when they cause a positive crossmatch and are present at the time of transplant [2-51. We have previously evaluated transplant survival for patients receiving renal transplants at this centre and found recipients with IgM autoreactive lymphocytotoxic antibodies to have a reduc- tion in their initial transplant survival [6] . This was despite qur rigorous protocol for the identification of autoanti- bodies before using dithiothreitol (DTT) to remove IgM antibodies in the pretransplant cytotoxicity crossmatch.

One possible explanation was that recipient sera also contained weak IgG anti-donor antibodies that were inhibited by the DTT in the pretransplant crossmatch, giving a false-negative result. The flow cytometry cross-

S 528

match (FCXM) is a more sensitive technique for the detection of complement fixing as well as non- complement fixing antibodies [7-lo]. We therefore, per- formed a retrospective FCXM study on 52 recipients of 57 renal transplants carried out at this centre between 1985 and 1992. Pretransplant sera were retrospectively recrossmatched using two-colour flow cytometry in order to determine whether weak or non-cytotoxic IgG anti- donor antibodies had been present in any of the recipients prior to transplantation.

~~~~~~

Patients and methods

Patients

Between 1985 and 1992, 961 patients received renal transplants a t this centre (824 first, 137 retransplants; 889 cadaver, 72 live related). Immunosuppressive therapy was as described previously [l 11 with all patients receiving cyclosporin. Recipients were selected for trans- plantation on the basis of ABO compatibility, a negative cytotoxi- city crossmatch using all historic positive and current sera and with a policy of minimising HLA antigen mismatches and no more than three mismatches being acceptable. Transplant failure was defined as transplant nephrectomy, return to dialysis or death of the patient for any reason.

Lymphocytotoxic antibody screening

All patients were regularly screened pretransplant for HLA-specific antibodies. They were screened in a standard microlymphocytotox- icity assay using a selected 40-cell panel of tray-frozen peripheral blood lymphocytes (PBLs) [I 21 from HLA-typed volunteers and patients with CLL, as a source of B cells. Whenever an HLA specificity was identified, that antigen was avoided in a donor for the patient concerned.

Autoreactive lymphocytotoxic antibodies

Autoreactive antibodies were initially identified during pretrans- plant lymphocytotoxic antibody screening by their reactivity with PBLs but not CLL cells, and then shown to be IgM by digestion with 0.005 M DTT. The presence of autoantibodies was confirmed by a cytotoxicity crossmatch with autologous lymphocytes and absorp- tion with autologous lymphoblastoid cell line cells [13]. Patients with confirmed IgM autoantibodies were crossmatched with and without DTT, a negative crossmatch in the presence of DTT being taken as acceptable for transplantation.

Complement dependent cytotoxicity crossmatch

Lymphocytes were prepared from monocyte-depleted donor spleen cell suspension and cross-matched against recipient sera as described previously [14]. Cytotoxicity was assessed following staining with a cocktail of ethidium bromide and acridine orange and cytotoxicity of greater than 10% above background was considered to be positive.

Flow cytometry crossmatch

Two-colour flow cytometry T and B cell crossmatches were per- formed retrospectively using cryopreserved donor splenic lympho- cytes. A R-phycoerythrin-conjugated mouse monoclonal anti- human CD3 (PE-CD3; R810, DAKO) and an R-phycoerythrin- conjugated mouse monoclonal anti-human CD 19 (PE-CD 19; R808; DAKO) were used for T cell and B cell identification respectively. A FITC-conjugated F (ab'), rabbit anti-human IgG (FITC-IgG, F315 DAKO) was used to identify IgG alloantibodies bound to T and/or B cells. Controls included a pool of sera from patients with high titre broadly reactive alloantibodies as a positive control and pooled human AB serum from non-transfused males as a negative control.

The method has been described previously [lo]. In brief, 10' lymphocytes in 30 11 PBS + 1 YO sodium azide were incubated with 20 p1 serum for 1 h at 22 "C. After washing, the cells were then incubated for 20 min on ice with FITC-IgG, washed and incubated for a further 20 min on ice with either PE-CD3 or PE-CD 19. After further washing, the samples were run on a FACScan flow cytometer (Becton Dickinson). Consort 30 software was used to acquire data that was analysed using LYSIS I software.

For T and B cell crossmatches, FITC staining was assessed for CD3- or CD 19-positive cells, respectively, using histograms. The sample mean channel of fluorescence was compared to the negative and positive controls. A positive FCXM was defined as a shift in the mean of more than 20 channels for the T or B cell peak when a patient serum was compared with the negative control.

Serum samples

For each recipient, three pretransplant sera were selected for FCXM against lymphocytes from the donor of the recipient's kidney. The samples included the one with the highest level of autoantibodies, the one with peak panel reactive alloantibodies (if applicable) and the latest serum prior to transplantation. All of the samples tested by FCXM had been cytotoxicity crossmatch negative prior to trans- plantation. In addition, where available a sample from the time of transplant failure was tested, or for patients with a functioning transplant, a sample taken at 3 months posttransplant.

Effect of DTT on IgG

In order to assess the effect of 0.005 M DTT on IgQ, the positive control pooled alloantiserum was tested in dilution from neat to 1 :512 with and without DTT by flow &tometry.

Results

Transplant outcome

For 58 of the 961 (6%) transplants, the recipients had clearly defined autoantibodies pretransplant and a posi- tive donor crossmatch that became negative in the presence of DTT. Overall, actuarial graft survival (GS, log rank) at 1 year was 82 % for recipients with autoanti- bodies and 87% for those without.

S 529

Table 1 Transplant outcome for recipients with autoantibodies (ns not significant)

Transplant function No. autoantibody-positive transplants

plants and 22 % of the retransplants with autoantibodies failed (Table 1).

~

First transplant Retransplant

Failed < 1 month 4 2 Failed 1-i2months 3 1 Failed > 1 year 3 2 Function 25 18

P = ns

Table 2 Cause of transplant failure for recipients with autoanti- bodies

Cause of failure

Following transplant biopsy, 12 of the failures were histologically defined as immunological losses (n = 7), non-immunological losses (n = 4) or failure due to a combination of factors (n = 1). The cause of failure in relation to transplant number and to time of failure is shown in Tables 2 and 3, respectively.

Cause of failure No. autoantibody-positive transplants

First transplant Retransplant

Immunoiogical 6 1 Non-immunological 2 2 Combination of factors 1 0

Table 3 Time of transplant failure in relation to cause for recipients with autoantibodies

Cause of failure Number of failed transplants

< 1 year > 1 year

Immunological 5 Non-immunological 3 Combination of factors 1

2 1 0

Table 4 Incidence of alloantibodies in recipients with autoanti- bodies (Txp transplant)

Outcome Number of recipients

Autoantibodies Auto- and alloantibodies

is t tpx Retpx 1st tpx Retpx

Failed 10 1 0 4 Function 21 6 4 12

For recipients of primary transplants, GS at 1 year was 77 % for patients with autoantibodies and 87 YO for those without. This is compared with our previously reported 85% l-year GS for highly sensitised primary transplant recipients with alloantibodies who received HLA- matched, cytotoxicity crossmatch negative transplants [15]. At 5 years, the difference in GS between patients with and without autoantibodies was reduced to 70% and 73%, respectively. Thirty-five of the 824 recipients (4%) of first transplants and 23 of the 137 recipients (1 7 %) of retransplants had autoantibodies; 29 % of the first trans-

HLA matching

There was no difference in the degree of HLA matching between the patients with and without autoantibodies. The percentage of recipients with 0, 1 + 2 and 3 HLA- B + DR mismatches was 10.3, 84.5 and 5.2 respectively, for the recipients with autoantibodies and 8.3, 83.4 and 8.3, respectively, for those without.

Alloreactive lymphocytotoxic antibodies

The relationship between recipient alloantibody status, transplant number and transplant outcome for the reci- pients with autoantibodies is shown in Table 4.

Flow cytometry mismatch

All of the 57 autoantibody-positive recipients tested by FCXM were T cell negative pretransplant. Two of the 57 were B cell positive pretransplant: one was a primary transplant who suffered two biopsy-proven, rejection episodes but whose transplant is still functioning: the other was a second transplant wfio had one rejection episode and who returned to dialysis after 3.5 years.

Eleven of the 14 failures were also tested by FCXM posttransplant. All 11 were T cell FCXM negative but one was B cell positive 8 months posttransplant, 3 months prior to returning to dialysis. Of the 33 successful transplants tested posttransplant, all were T cell negative and one was B cell positive, as he had been pretransplant.

Effect of DTT on IgG

The positive control alloantiserum was titred from neat to 1 : 51 2. When tested by flow cytometry, the serum became

S530

negative at 1 :256. When tested with and without DTT, there was no significant reduction in the level of IgG binding to T or B cells by treatment with DTT.

Discussion

The data presented here confirmed our previous finding that for primary transplant recipients with autoanti- bodies, GS is 10 % lower at 1 st year than for those with- out autoantibodies, although this did not reach sig- nificance. We found that, with two exceptions, the recipients with autoreactive lymphocytotoxic antibodies did not have pretransplant anti-donor IgG antibodies. None of the transplant failures, not even the five immun- ological losses within the 1 st year posttransplant, could be attributed to the presence pretransplant of IgG unde- tected in the DTT cytotoxicity crossmatch. Low levels of anti-donor IgG would have been detected by the more sensitive FCXM. This was T cell negative in all cases, suggesting that none of the recipients had IgG antibodies specific for HLA class I.

The significance of the two pretransplant B cell FCXM positive results remains unclear. It is known that the key factor in determining the relevance of donor-reactive B cell antibodies is the specificity of the antibodies [16]. Only HLA-specific antibodies are detrimental to trans- plant outcome. Our study did not investigate the specific- ity of the antibodies and blocking studies would be necessary to draw any further conclusions.

One patient developed anti-donor B cell IgG anti- bodies detected by FCXM posttransplant, prior to trans- plant failure. Again, blocking studies are required to determine the specificity of the antibodies, although one can speculate that anti-donor IgM antibodies were pre-

sent pretransplant that were removed in the DTT cytotox- icity crossmatch and not detected by FCXM but that then switched to IgG posttransplant and contributed to trans- plant failure. However, there was no evidence for this in the other cases.

IgM HLA specific antibodies can persist in the sera of renal patients [17] and it remains possible that a patient could have IgM allo- and autoreactive antibodies, neither of which would be detected by the DTT cytotoxicity crossmatch or FCXM. Transplantation in the presence of IgM HLA specific antibodies results in poorer outcome [2, 51, so every effort should be made pretransplant to remove autoantibodies by absorption with autologous lymphocytes or lymphoblastoid cell line cells. The specif- icity of any HLA-specific antibodies can then be defined and those antigens avoided in a donor for that patient.

IgM autoreactive antibodies have broad-binding spe- cificity and may result from the polyclonal stimulation of CD5-positive B cells by a variety of stimuli including bacterial viral and parasitic antigens [l]. The reason for a reduced initial GS in patients with these antibodies remains unresolved, but our initial concern that low levels of IgG were destroyed in the DTT cytotoxicity cross- match were not borne out either by the FCXM studies with patient sera or by the studies on the effect of DTT on dilutions of alloantiserum in the FCXM. We, therefore, remain confident that for patients with proven ZgM autoantibodies, the DTT cytotoxicity crossmatch is suffi- ciently sensitive to detect IgG anti-donor antibodies.

Patients with autoreactive lymphocytotoxic anti- bodies can clearly be successfully transplanted. However, for reasons as yet unresolved but unrelated to the presence of anti-donor IgG antibodies pretransplant, primary transplant recipients with autoantibodies initi- ally do less well than those without.

References

1. Taylor CJ, Ting A, Morris PJ (1991) Production and characterization of human monoclonal lymphocytotoxic antibodies from a renal dialysis patient. Tissue Antigens 37: 3 12-120

2. Chapman JR, Taylor CJ, Ting A, Morris PJ (1986) Immunoglobulin class and specificity of antibodies causing positive T cell crossmatches: relationship to renal transplant outcome. Transplantation 42:608-613

3. Ettenger R, Robertson L (1985) The evaluation and relevance of auto- lymphocytotoxic antibody in the highly presensitized patient. Trans- plantation 43: 302-304

4. Ting A, Morris PJ (1983) Successful transplantation with a positive T and B cell crossmatch due to autoreactive antibodies. Tissue Antigens 21 :219- 226

Morris PJ (1989) Characterization of lymphocytotoxic antibodies causing a positive crossmatch in renal transplan- tation. Transplantation 48 :953-958

5. Taylor CJ, Chapman JR, Ting A,

6. Martin S, Dyer PA, Scott PD (1992) Autoreactive IgM non-HLA anti- bodies are not entirely clinically irrelevant. Transplant Proc 24: 2504

7. Garovoy MR, Colombe BW, Melzer J, Feduska N, Shields C, Cross D, Amend W, Vincenti F, Hopper S, Duca R, Salvatierra 0 (1985) Flow cytometry crossmatching for donor- specific transfusion recipients and cadaveric transplantation. Transplant Proc 17:693-695

S531

8. Cook DJ, Terasaki PI, Iwaki Y, Terashita G, Takeda A, Fiyikawa J (1987) The flow cytometry crossmatch in kidney transplantation. In: Terasaki PI (ed) Clinical Transplants 1987. UCLA Tissue Typing Laboratory, Los Angeles, pp 409-414

9. Mahoney RJ, Auk KA, Given SR, Adams RJ, Breggia AC, Paris PA, Dalonaki GE, Hitchcox SA, White BW, Himmelgarb J, Lecher D (1990) The flow cytometric crossmatch and early transplant loss. Transplantation 49 527-535

10. Martin S, Liggett H, Robson AJ, Connolly J, Johnson RWG (1993) The association between a positive T and B cell flow cytometry crossmatch and renal transplant failure. Transplant Immunol (in press)

11. Johnson RWG, Mallick NP, Bakran A, Pearson RC, Scott PD, Dyer P, Donaghue D, Morris D, Gokal R (1989) Cadaver renal transplantation without maintenance steroids. Trans- plant Proc 21:1581-1582

12. Sinnott PJ, Kippax RL, Sheldon S, Dyer PA (1985) A rapid and simple method for the detection of lympho- cytotoxic antibodies using cell panels frozen on Terasaki trays. Tissue Antigens 26: 31 8 -322

13. Martin S, Claas F (1993) In: Dyer PA, Middleton D (eds) Methods in clinical histocompatibility testing. IRL Press, Oxford, pp 81 - 105

14. Dyer PA, Johnson RWG, Martin S, Bakran A, Gokal R, Harris R, Mallick NP, Manos J, Orr WM, Pearson RC, Scott PD, Sheldon S, Short C (1989) Evidence that matching for HLA antigens significantly in- creases transplant survival in 1001 renal transplants performed in the North West Region of England. Transplantation 48:131-I35

15. Martin S, Connolly J, Jos V (1993) Importance of HLA matching in highly sensitised renal transplant recipients with high transplant survival rates. Transplant Proc 25:265-266

16. Karuppan SS, Lindholm A, Moller E (1 990) Characterization and sig- nificance of donor reactive B cell antibodies in current sera of transplant patients. Transplantation 49: 510-515

17. Tyler JD, Bray JE, Looney SD, Foster JH, Goldman MH (1991) Analysis of specificity of DTT sensitive antibodies in renal serum samples. Transplant Proc 23 : 1742 - 1743