Charles W. DeWitt

OUTCOME OF CARDIAC TRANSPLANT RECIPIENTS WITH A POSITIVE DONOR-SPECIFIC CROSSMATCH—PRELIMINARY RESULTS WITH PLASMAPHERESIS

To assess the influence of a positive T or B cell IgG crossmatch on the development of rejection and mortality following cardiac transplantation, we reviewed all cardiac transplants performed in Utah between March 1985 and October 1990. Of the 328 cardiac allograft recipients, 11 (3.4%) had an IgG positive crossmatch. Actuarial survival at 24 months in the positive crossmatch group was 57.3% +/- 0.02 while that of the controls was 86.1% +/- 2.1 (P < 0.05). Allograft rejection occurred earlier in recipients with a positive crossmatch (10.0 +/- 5.8 days versus 34.0 +/- 2.3 days, P < 0.001). The first allograft rejection episode in patients with a positive crossmatch was characterized by immunoglobulin and complement deposition in small blood vessels and interstitial edema and endothelial cell activation in the absence of a lymphocytic infiltrate. Furthermore, the allograft rejection in the positive crossmatch group was accompanied by hemodynamic compromise in a large proportion of the patients (73%). In addition to augmentation of immunosuppression, plasma exchange therapy was performed within the first week following transplantation in 8 of the 11 positive crossmatch patients. Survival in the patients treated with plasma exchange (75%) appears to be better than in those not receiving plasma exchange (33%) within one week of transplantation. While immunosuppressive therapy aimed at the humoral arm of the immune system and plasma exchange therapy may improve survival in recipients with a positive donor-specific crossmatch, survival is worse in patients with a positive crossmatch than in patients with a negative crossmatch. Thus, it would appear prudent to prospectively crossmatch cardiac transplant candidates with a greater risk of developing a positive crossmatch, such as those potential recipients with an elevated level of panel-reactive antibodies.

Possible association of the exetended MHC haplotype B44-SC30-DR4 with autism

We previously reported that the complement C4B null allele appears to be associated with infantile autism. Since the C4B null allele is known to be part of the extended or ancestral haplotype [B44-SC30-Dr4], we investigated the incidence of [B44-Sc30-DR4] in 21 autistic children and their parents. This extended haplotype was increased by almost six-fold in the autistic subjects as compared with healthy controls. Moreover, the total number of extended haplotypes expressed on chromosomes of autistic subjects was significantly increased as compared with those expressed on chromosomes of healthy subjects. We conclude that a gene related to, or included in, the extended major histocompatibility complex may be associated with autism.

HLA A, B and DR typing in idiopathic dilated cardiomyopathy: a search for immune response factors

Several autoimmune diseases have been associated with increased frequencies of various histocompatibility antigen (HLA) types that may be linked to immune response genes. Idiopathic dilated cardiomyopathy (IDC) has been proposed as a disease with autoimmune features, but HLA associations have not been evaluated. We performed HLA typing in 37 consecutive patients with IDC. Patients with habitual alcoholism were excluded. Results showed that no single HLA type could account for most cases; IDC is a genetically heterogeneous disease. However, uneven distributions were noted for certain types. Haplotype frequency of B27 was 0.145 in patients vs 0.033 in 5,726 local control subjects (p less than 0.001). Other A and B frequencies (except A2) were evenly distributed. HLA DR typing also revealed differences. The DR4 locus was present in 54% (19 of 35) of patients, vs 32% (26 of 82) of blood bank control subjects (p less than 0.02). The associated relative risk of DR4 was 2.2 and the etiologic fraction 0.29. Sex, disease chronicity, functional class, ejection fraction and biopsy evidence of myocarditis did not distinguish DR4-positive from DR4-negative patients, but they were older (54 +/- 12 vs 42 +/- 14 years, p less than 0.02). Of note, 68% were positive for DR4 or B27, or both. HLA DR6Y was underrepresented; it was present in 9% (3 of 35) of patients, vs 26% (21 of 82) of control subjects (p less than 0.04). The relative risk of DR6Y was 0.27 and the preventive fraction 0.19. These associations will require independent confirmation. However, they suggest that genetically determined immune response factors associated with HLA loci may play a role in pathogenesis in certain patients with IDC.

Histocompatibility in couples with recurrent spontaneous abortion and normal fertility

Histocompatibility between husband and wife at the HLA locus has been suggested as a determinant of recurrent spontaneous abortion. We measured the incidence of HLA antigen sharing within 12 couples with histories of unexplained recurrent abortion and in a fertile control population of 77 couples. In the recurrent abortion group, 6 of 12 (50%) of the couples shared no HLA antigens, whereas only 3 of 12 (25%) shared one antigen, 1 of 12 (8.3%) shared two antigens, and 2 of 12 (17%) shared three antigens. In the fertile group, 27 of 77 (35.1%) shared no antigens, 33 of 77 (42.8%) shared one antigen, 14 of 77 (18.2%) shared two antigens, 2 of 77 (2.6%) shared three antigens, and 1 of 77 (1.3%) shared four antigens. In 50 of these control couples who were available for complete reproductive histories, there were no significant correlations between the incidence of antigen sharing and the numbers of offspring, the incidence of spontaneous abortion, or infertility problems. Six of the women in the recurrent abortion group became pregnant during the study. Three of these (50%) delivered live infants independent of the degree of antigen sharing and without the benefit of immunologic treatment. Therefore, the degree of HLA antigen sharing did not define a population with increased pregnancy wastage or predict subsequent pregnancy outcome.

Immunogenetic control of experimental collagen-induced arthritis in rats. II. ECIA susceptibility and immune response to type II collagen (CALF) are linked to RT1.

The segregation of genes controlling ECIA susceptibility and the level of immune response to native calf type II collagen were determined in the F2 progeny of matings between WF (RTIu/u; ECIA‐susceptible; high responders) and LEW.B3 (RT1n/n; ECIA‐resistant; low to intermediate responders). RT1n/n F2 progeny showed resistance to ECIA, low skin test reactivity to type II collagen and intermediate levels of IgG anti‐collagen antibodies (‐log2 of 6.2 ± 2.6; mean ± SD, n= 10). RT1u/u and RT1u/u F2 progeny were susceptible to ECIA and were high responders to type II collagen by skin testing and IgG antibody titres (‐log2 of 12.1 ± 1.3, mean ± SD, n= 26). Although all rats that developed arthritis were also high responders to type II collagen one group of immature F2 progeny, RT1u/u and RT1u/n, showed high anti‐collagen immune responses in the absence of detectable arthritis. The data indicate that genes linked to RT1.A control susceptibility to ECIA and at least part of the immune response to native calf type II collagen in WF and LEW.B3 rats.

Ag-F: serological and genetic identification of a new locus in the rat governing lymphocyte membrane antigens.

A new genetic locus in the rat (Ag-F) is described. It controls expression of lymphocyte membrane antigens as determined by allospecific cytotoxie antibody. It is linked by approximately two recombination units to the gene for albinism and is therefore in linkage group I. It is suggested that Ag-F is the homologue in the rat of mouse 3–1. It appears to be highly polymorphic and yields antigens which are highly cross reactive serologically. Disparity at Ag-F is the probable cause of alloantibody responses attributable to reciprocal immunization between the Lewis and Fischer strains.

Efficacy of OKT3 retreatment for refractory cardiac allograft rejection.

Although OKT3 monoclonal antibody is a useful therapy for refractory cardiac allograft rejection, the use of OKT3 for prophylaxis may be limited by the potential of sensitization and subsequent loss of efficacy on retreatment. OKT3 was required for refractory rejection in 21 of 165 recipients transplanted between March 1985 and August 1988. Twelve of these patients had previously been exposed to OKT3, and the retreatment efficacy was evaluated. The study population averaged 42.1 +/- 15.3 years of age (mean +/- SEM) and had experienced 2 +/- 1 previous episodes of rejection. The prior episodes of rejection had been treated with pulse methylprednisolone and antithymocyte globulin, and in addition 3 patients (25%) also required a course of antilymphoblast globulin. Retreatment OKT3 for refractory rejection was required 120 +/- 94 days following transplantation. CD3+ lymphocytes were eliminated from the circulation within 24-48 hr in 11 of 12 patients, all of whom showed histologic improvement within the first week. Total resolution on the initial follow-up biopsy was noted in 9 (75%) during the course of therapy. Subsequent rejection episodes occurred in 9 (82%) of the survivors at 71 +/- 64 days. One-year survival was 83% in this vigorously rejecting patient population. Serious infections occurred within 3 months of therapy in 4 (36%). The side effects of OKT3 retreatment were similar to those seen with first exposure and did not require OKT3 discontinuation. Thus OKT3 may be administered with success in most patients who have previously been exposed to it.

Further studies on the use of mouse epidermal cells for the in vitro induction and detection of cell-mediated cytotoxicity.

Abstract Mouse epidermal cells (EC) and lymphoid cells (LC) were compared as targets of cellmediated cytotoxicity (CMC) in short-term chromium release assays where attacker cells were generated in primary mixed cultures using irradiated allogeneic EC or LC as stimulators. Three patterns of relative susceptibility to lysis of the two types of target cells were observed: (i) significantly greater lysis of LC than of EC targets; (ii) significantly greater lysis of EC than LC targets; and (iii) approximately equal susceptibility to lysis of the two targets. The first pattern was primarily associated with LC stimulators, whereas the second and third patterns were almost invariably associated with EC stimulators. Factors possibly contributing to the differences in in vitro immunogenicity and susceptibility to CMC of EC and LC were investigated, including the alteration of EC surface antigens during the trypsinization required to prepare EC suspensions, the differential expression of shared alloantigens, or the restricted expression of tissue-specific alloantigens on the two types of cells. Tests with intact and trypsinized LC on the one hand and fresh and short-term cultured EC on the other indicated that trypsinization is not responsible for the basic differences between EC and LC detected in the in vitro assays. Antibody absorption tests demonstrated that although EC and LC express approximately equal quantities of the cell surface antigens determined by the H-2D region of the H-2 complex, LC express significantly greater quantities of the antigens determined by the H-2K and I regions. In addition, the results of cold target inhibition tests suggest that tissue-specific antigens on both EC and LC also influence their relative immunogenicity and susceptibility to lysis.

The effect of neonatal thymectomy on antibody responses to histocompatibility antigens in the rat

Abstract Neonatally thymectomized rats produced alloantibody of only the IgM class in response to histocompatibility (H) antigens governed by the major H locus. They produced no antibody against minor H antigens. Unaltered rats produced only IgG antibody against minor H antigens and this response is therefore termed T cell obligatory. This is in distinction to the T cell-dependent response of unaltered rats to major H antigens, which is composed of an initial IgM component followed by conversion to production of IgG.

Strong and weak immune responses across the same major histocompatibility barrier in rats.

Inbred rat strains, Fischer 344 (F-344) and Lewis (LEW), share the serologicalAg-Bl allele and react very weakly in mixed lymphocyte culture (MLC). Despite this apparent identity atAg-B, these strains differ markedly in their immune responses to anAg-B disparate third strain Marshall 520 (M-520) (Ag-B6). F-344 recipients allowed M-520 heart grafts an extended survival, whereas LEW recipients rejected them rapidly. F-344 and M-520 showed a weak response in MLC in contrast to a strong response for LEW and M-520. F-344 produced antisera in response to injection of M-520 cells that had a relatively high antibody titer but low cytotoxic activity. F-344 responded to another strain, Buffalo (BUF) (alsoAg-B6), in a similar fashion. F-344 apparently can produce a strong allogeneic response, as it was able to rapidly reject heart grafts from (LEW x Brown-Norway) F1 donors (LBN) (Ag-B 1/3). The low response of F-344 to M-520 probably was not due to shared antigens between the two strains because M-520 heart grafts underwent rapid rejection in LEW hosts highly tolerant to F-344. To explain the contrasting response of F-344 and LEW to theAg-B6 disparity, we propose that it is controlled by an immune-response gene(s); that F-344 has a low-responding allele and LEW has a high-responding allele. The data do not reveal a location for this proposed gene. The high-responding allele appears to be dominant, as M-520 hearts were rejected rapidly by (F-344 x LEW) F1 recipients.