Granulocytotoxic antibodies in a patient with propylthiouracil-induced agranulocytosis.

Agranulocytosis developed in a patient who was receiving propylthiouracil. Using a microgranulocytotoxicity assay, serum taken from the patient was shown to be strongly granulocytotoxic when tested against the patients granulocytes and those obtained from two of eight normal subjects. Tests for granulocyte agglutinins and for lymphocytotoxicity were negative. Granulocytotoxic activity decreased as the patient's peripheral granulocyte count recovered. Cytotoxicity was shown to be mediated by a complement-dependent IgM antibody.

Acquired agranulocytosis with granulocyte specific cytotoxic autoantibody.

Multiple infections and severe neutropenia were found in a previously healthy 29 year old man with no history of similar syndromes in the family, drug ingestion or exposure to environmental toxins. There was no evidence at the time of presentation of diseases previously associated with agranulocytosis (e.g., neoplasia, thyrotoxicosis, chronic infection, collagen-vascular disease or leukoagglutinating antibody). His serum contained a nonagglutinating, complement-dependent, cytotoxic antibody, however, reactive with peripheral blood granulocytes from 35 per cent of normal donors. The neutropenia was not affected by steroids but resolved promptly after splenectomy. Microscopic examination of the spleen revealed ingestion of polymorphonuclear leukocytes by splenic macrophages. Family studies indicated that the target antigen was non-HLA and that the antibody was not absorbed by lymphocytes or platelets. We conclude that the agranulocytosis was autoimmune in origin and suggest that similar myeloid-specific immune responses could influence granulocyte tranfusion and bone marrow transplantation by alloimmune "rejection" that would not be avoided by matching only for HLA specificities.

Nosocomial Legionnaires' disease. Occurrence in recipients of bone marrow transplants.

Nosocomial pneumonia caused by Legionella pneumophila serogroup 1 occurred in five patients after bone marrow transplantation for hematologic malignancies. Two patients died as a result of the infection despite treatment with erythromycin. Serologic screening revealed no other cases of Legionnaires' disease in 40 consecutive recipients of bone marrow transplants, giving a frequency of infection of 13 percent. These five cases represent 23 percent of the pneumonia occurring in this group of patients. Patients undergoing bone marrow transplantation are highly susceptible to infectious complications. Legionnaires' disease must now be added to the list of pathogens infecting this group of patients. Erythromycin is not generally a part of standard empiric antibiotic regimens in febrile neutropenic patients, but appears to be a reasonable addition when pneumonia does not respond to conventional, empiric treatment. Even with appropriate therapy, Legionnaires' disease remains a highly lethal infection in immunocompromised hosts.

A 2-year trial of prospective HLA-DR matching: effects on renal allograft survival and rate of transplantation.

In a 2-year trial at a single center, prospective HLA-DR matching for cadaver renal transplantation exerted a stronger effect than either HLA-A and B matching or blood transfusion. One-year cumulative graft survivals for two-DR-matched organs was 92%. Grafts matched for one DR antigen had a cumulative 1-year graft survival of 65% whereas grafts matched for zero DR antigens had a cumulative 1-year survival of 41%. When all cadavers with less than two identifiable DR antigens were excluded from this analysis, however, the graft survivals of the groups known to be mismatched for one or two DR antigens were similar (61% versus 59%). Grafts matched for three or four HLA-A and B antigens did somewhat better than those matched for only zero, one, or two HLA A and B antigens (74% versus 59%, 1-year survival). This effect was only demonstrable in the zero- and one-DR-matched group. Similarly, prior blood transfusion exerted a modest effect (transfused versus nontransfused, 71% versus 56% 1-year graft survival) that was also most evident in the zero- and one-DR-matched groups. The institution of this trial was also associated with a 35% annual increase in the rate of transplantation and a 50% reduction in median patient waiting time. DR typing of cadaveric donors is feasible and highly desirable. Multicenter pooling of DR-typed donors is thus predicted to lead to optimal matching for a high proportion of renal transplant candidates.

Successful DR-incompatible cadaver kidney transplantation. Combined effect of HLA A and B matching and blood transfusion.

The purpose of this retrospective analysis of DR-incompatible cadaver renal transplantation was to evaluate the effect of HLA A and B matching and blood transfusion status on actual one-year graft survival. There were 31 2-DR, 111 1-DR, and 27 0-DR grafts at risk during the study period. First, a comparison was made between preoperative (PRE) and peroperative (PER) transfusions alone. Graft survivals were 70% vs. 92% (2 DR), 67% vs. 52% (1 DR) and 71% vs. 39% (0 DR) for the PRE and PER groups, respectively. Statistical significance was not found between the two values in each DR subgroup, although the difference approached significance in the O DR group (0.1 greater than P greater than 0.05). Matching for greater than or equal to 2 A and B antigens significantly improved graft survival in the 1 DR-matched group when compared with those matched for less than 2 antigens (76% vs. 44%, P less than 0.005). While marked differences between the greater than or equal to 2 and less than 2 A and B matched groups were observed for both the 2 DR (92% vs. 68%, P greater than 0.1) and O DR groups (59% vs. 40%, P greater than 0.3) these differences were not significant. Stratifying the data for transfusion status revealed that the positive influence of HLA A and B matching in the 1 DR group was dependent upon the presence of preoperative blood administration. Graft survival of 87% for the PRE transfused recipients of grafts matched for greater than or equal to 2 A and B antigens was significantly better (P less than 0.001) than the 42% survival observed in similarly transfused recipients of poorer matched organs. Conversely, A and B matching was not significantly beneficial in the 1 DR recipients transfused only at the time of transplant with graft survivals of 57% vs. 43% for those matched for greater than or equal to 2 or less than 2 A and B antigens, respectively (P greater than 0.3). This analysis suggests that a combined effect of both HLA A and B matching and preoperative blood transfusions may allow for highly successful first cadaver renal transplantation in the face of DR incompatibility.

Differential requirements for class II MHC antigen in human T cell activation.

We have employed three different models of major histocompatibility complex (MHC) antigen to investigate the role of these antigens in some aspects of human T cell activation in in vitro; specifically the induction of lymphokine synthesis and receptors, and in antigen presentation. Those models were (1) allogeneic platelets (class I alone); (2) D/DR compatible lymphocytes (class I plus nonimmunogenic class II); and (3) allogeneic heat-treated lymphocytes. After heating at 45 C for 1 hr, the latter are completely viable and express both class I (HLA A and B) and class II (HLA, DR, MT/MB) MHC antigens but do not stimulate in the mixed lymphocyte reaction (MLR). These models were compared with conventional irradiated MLR stimulating or antigen-presenting cells. Only the conventional cells were able to stimulate the synthesis of interleukin-1 and interleukin-2. Also, irradiated but not heat-treated non-T cells could present soluble antigen to autologous T cells. This implies that intact and unmodified Ia molecules are required for those responses. On the other hand, heat treated allogeneic lymphocytes and D/DR-compatible lymphocytes but not allogeneic platelets could induce responsiveness to interleukin-2. This function appears to require a heat stable but nonallogeneic component of the class II molecules. Our previous observations employing these models have also shown that there is yet a third category of T cell responses, including memory cell priming and suppressor cell induction, that can occur in the presence of predominantly class I antigen. Taken together, these data illustrate the functional heterogeneity of T cell responses to Ia-like antigens. The possible relevance of these findings to clinical transplantation is discussed.

Failure of cyclosporine to prevent in vivo T cell priming in man. Studies in allogeneic spleen transplantation.

Cyclosporine is a potent new immunosuppressive agent utilized in clinical organ transplantation. Available evidence suggest that it interferes with the secretion of interleukin-2. However, the long term efficacy of cyclosporine in preventing allograft rejection may depend on a relative sparing of suppressor cells early in the allogeneic response, allowing them to mature and effect a state of operational tolerance. If this is the case, cyclosporine must not affect antigen priming or recognition. Two patients in our center underwent allogeneic spleen transplant in conjunction with renal and pancreatic transplant. Both patients were treated with therapeutic levels of cyclosporine during the course of transplant. Neither developed any clinical signs of renal or pancreatic transplant rejection. Both patients developed graft-versus-host disease and eventually required allogeneic (donor) splenectomy. Studies performed on the splenocytes recovered from these specimens demonstrate alloantigen-specific cytotoxic T cell precursors. These studies demonstrate that although cyclosporine can prevent allograft rejection it does not necessarily prevent or ameliorate graft-versus-host disease. Furthermore, cyclosporine does not prevent in vivo T cell priming of alloantigen recognition. The primed cytotoxic precursors can be expanded in the presence of exogenous interleukin-2 to become fully active cytoxic cells.

The use of partially matched, unrelated donors in clinical bone marrow transplantation.

It is estimated that 60-70% of patients who might benefit from a bone marrow transplant will not have a suitably matched, related donor. We have, therefore, designed a clinical experiment to test the safety and feasibility of using marrow from partially matched, unrelated donors. This paper details our transplant experience in the first eight patients with leukemia. The first four patients had advanced leukemia at the time of transplantation. Each showed hematopoietic recovery, but all died from septic complications largely related to extended neutropenia encompassing both the pre-marrow-grafting and the post-marrow-grafting period. The next four patients were in remission at the time of transplantation. Each showed prompt and sustained hematopoiesis with variable graft-versus-host disease (GVHD). No acute or chronic GVHD was seen in two patients, grade II (skin only) was seen in one patient, and grade IV (skin, liver, and gut) was seen in one patient. One patient has died from sepsis five-and-one-half months following transplantation, and three are alive and well six-and-one-half to nine-and-one-half months postengraftment. This preliminary experience, together with several case reports in the literature, leads us to conclude that bone marrow transplantation with partially matched, unrelated marrow is a safe and feasible approach. If these results are confirmed by longer follow-up in a larger group of patients, the development of marrow donor pools would appear to be justified.

Outcome of renal transplantation following a positive cross-match with historical sera: the ASHI survey.

Analysis of retrospective data, obtained on 216 patients from 27 centers transplanted across some form of positive lymphocyte cross-match in a noncurrent serum, revealed that actuarial 1-yr graft survival was 69% in first transplants and 53% in recipients of second or subsequent transplants. Graft outcome did not correlate with peak antibody levels, change in antibody from peak to current, remoteness in time of the most recent positive serum, the number or timing of sera cross-matched, the technique or target cell cross-matched, or the degree of positivity of the most recent positive serum. Although a concurrent control population was not available, these results support the concept that acceptable graft survival can be achieved despite a positive cross-match with noncurrent sera.

Differential stimulatory requirements and regulation of naive and primed human lymphocytes.

In these investigations, human lymphocytes primed in vitro in MLR have been employed as a model for human memory cells and have been compared to naive lymphocytes from the same donor. Both the stimulatory requirements and the regulation of these cells were found to differ significantly. The dose of stimulators giving a maximal primary (I) response was less than 10% the dose of restimulating cells giving a maximal secondary (II) response. II responses were further found to be inversely related to the original I response. This was associated with at least two separate regulatory phenomena. Suppressor cell induction was enhanced at high priming doses while memory cell precursors were preferentially stimulated at very low priming doses. Priming of memory cells could also be demonstrated to occur in the absence of any detectable I proliferation by utilizing platelets or heat treated stimulators. Memory cells were also a much more resistant than naive cells to both alloantigen induced suppressor cells and to culture activated monocyte suppressor cells. This in vitro model suggests that the human I and II responses to alloantigen have both distinct triggering requirements and differential sensitivity to regulatory cells. It is suggested that preferential formation of memory cells under conditions that require no proliferation and which are suboptimal for suppressor cell generation and the acquired resistance of memory cells to down regulation by suppressor cells may contribute to the poor graft prognosis of sensitized renal transplant patients.

Is secreted IL-1 necessary for normal human T-cell activation?

The role of secreted and membrane-bound IL-1 in the activation of human T cells by monocyte-associated, processed antigen was examined. The IL-1 is secreted from antigen-pulsed plastic-adherent monocytes for only 24 h after isolation. After extensive washing, however, these monocytes are fully capable of stimulating T cells to proliferate. The T cells require less than 24 h of exposure to the monocytes to become activated and can then be cultured alone in fresh media. Addition of exogenous IL-1 does not enhance T-cell responsiveness in this model. Anti-IL-1 beta antibody does not inhibit the response, although we observed that pulsed monocytes have significant membrane-bound IL-1 assayed as biologic activity in the murine thymocyte costimulator assay. These studies suggest that secreted IL-1 is not required for the activation of human T cells and that membrane-bound IL-1 serves this function. The data further suggest that these two forms of IL-1 may be functionally distinct and that IL-1 beta is not the major component of membrane-bound IL-1. The possible relationship of these findings to the clinical efficacy of IL-1 inhibitors, such as corticosteroids, is discussed.

Immunoregulatory activity of T-cell subsets activated in human mixed lymphocyte reaction.

Unseparated human T cells or isolated OKT4+ and OKT8+ subsets were stimulated in allogeneic MLR and subsequently tested for their capacity to suppress fresh autologous responding cells in a second MLR. Contrary to the bulk of evidence regarding the immunoregulation of B-cell function, both OKT4+ and OKT8+ cells could be activated to form suppressors of the MLR and were equally radioresistant. An early peak of blastogenesis was observed when OKT4+ suppressors were used, and this event was completely radiosensitive. When the fresh responding cells were also fractionated into OKT4+ and OKT8+ subsets, no preference for suppressibility of either subset could be demonstrated, indicating that OKT4+ suppressors were acting directly and not through an inductive mechanism requiring OKT8+ effectors. It is suggested that when alloantigenic responses are involved, the human T-cell subsets defined by these monoclonal antibodies may be more accurately defined by the class of MHC antigen to which they respond than by their effector function.

Resistance to suppression and acquisition of memory by human T cells.

Human memory cells acquire resistance to several types of suppressor cells, including MLR generated suppressor cells. These data suggest one possible mechanism of that resistance, namely retention of IL-2 receptors in the resting state. Cells from a 7-day MLR were separated on a single step percoll gradient. All proliferating cells were found in the interface. Pellet cells were nondividing. Interface and pellet cells had equivalent memory function in a secondary MLR. Thus, there appear to be at least two subpopulations of memory cells, including one that is primed without undergoing division. These subpopulations are functionally distinct. Interface memory cells were 30-50% more resistant than pellet memory cells to MLR generated suppressor cells. On culture day 10, neither pellet nor interface cells displayed significant spontaneous proliferation but exogenous Interleukin 2 (IL-2) produced up to five times as much proliferation in interface cells as in pellet cells. Further, FACS analysis with an anti-TAC equivalent antibody also showed that significantly more interface cells have surface receptors for IL-2. Thus, cells that had previously divided continue to have more and/or higher affinity receptors for IL-2 even after return to the resting state. If a mechanism of suppression in the mixed lymphocyte reaction is to reduce the synthesis/release of Il-2, memory cells may acquire their relative resistance to this suppression by virtue of the increased IL-2 sensitivity of this discrete subpopulation.

Suppressor cells induced in the human autologous mixed lymphocyte reaction.

Theophylline, a phosphodiesterase inhibitor, also reversibly inhibits the SRBC receptor of a subpopulation of human T lymphocytes. We have previously reported that the precursor for the Concanavalin A (Con A) inducible suppressor cell (SC) is found predominantly in the theophylline-sensitive (Ts) population; whereas, the precursor of the mitomycin resistant suppressor cell induced in the allogeneic mixed lymphocyte reaction (MLR) is predominantly in the theophylline-resistant (TR) subset. This study examines the theophylline sensitivity and functional characteristics of suppressor cells induced in autologous MLR (AMLR). Both TS and TR proliferate moderately in AMLR but differ in their ability to form suppressors. AMLR activated Ts and TR cells demonstrate early blastogenesis and weak suppression in the Con A assay system, but suppression in an MLR assay system appeared to be confined largely to the TR subset. AMLR suppressors were mitomycin sensitive. These data illustrate that suppressors generated in allogeneic and autologous MLR have significant functional and phenotypic similarities, and strengthen the association between these two immunologic phenomena.

Human suppressor cells: evidence of functionally distinct subsets.

Human suppressor cells (SC) differing in both kinetic characteristics and degree of specificity were induced in vitro either by Concanavalin A (CON A) (10 microgram/ml) or by repeated stimulation with allogeneic cells. CON A SC caused a characteristic early peak in blastogenesis (Day 2) when cocultured with fresh autologous cells in the presence of either CON A or augmented response fell rapidly to values 50--90% below positive controls by the usual optimal day for the particular stimulus. Treatment with 5-bromodeoxyuridine and light during rapid proliferation ablated the early responses but increased the later responses (i.e., reduced the suppression). The cells mediating both the early rise and late suppression were found to belong to a subset of T cells that lost their ability to form SRBC rosettes (theophylline sensitive) in the presence of the phosphodiesterase inhibitor, theophylline, and were inactivated by mitomycin C. In contrast, SC induced by repeated allogenic stimulation were suppressive in MLC but not CON A cultures, did not initiate early blastogenesis in the presence of naive cells and were partially mitomycin resistant. The mitomycin resistant SC induced by allogeneic stimulation are found in the theophylline resistant T-cell subset. These data are consistent with a model of human SC differentiation in which at least two subsets may be present: (1) a mitomycin sensitive, nonspecific cell that participates in an obligatory early autologous recognition event (induction?) and (2) a later occurring, more mitomycin resistant effector cell that appears to acquire specificity for the response that it affects.

Alteration of human accessory cell function by heat treatment: role of IL-1 and class II MHC antigens.

Heat-treated monocytes (1 hr, 45 degrees C) cannot present soluble antigen or mitogen to purified autologous T cells. This is despite normal viability and normal expression of class II MHC antigens. They do not secrete IL-1 nor stimulate secretion of IL-2 by T cells. Addition of exogenous IL-1 or IL-2 does not, however, reconstitute the response to soluble antigen. Furthermore, even after overnight pulsing with antigen prior to heat treatment under circumstances in which the antigen is known to be appropriately processed, stimulation of T-cell proliferation still does not occur. Thus there appear to be at least two discrete lesions produced by heating: failure of IL-1 production, per se, and intrinsic failure to present previously processed antigen. It is also hypothesized that heat treatment may produce alterations in Ia molecules which specifically disallow transduction of the proliferation signal to T cells.

Critical role of HLA-DR beta 1 residue 58 in multiple polymorphic epitopes recognized by xenogeneic and allogeneic antibodies.

In a previous study, we identified glutamic acid at position 58 in DR (beta 1*1101) as critical for the epitopes recognized by the DRw11-specific mAb GS88.2, as well as the I-LR1 mAb that recognizes a polymorphic epitope on DR(alpha,beta 1*1101) and some DP molecules. The purpose of this study was to determine whether other polymorphic residues contribute to these epitopes and whether DR beta glutamic acid or alanine 58 and DP beta glutamic acid 56, the analogous position in DP beta, contribute to epitopes recognized by other anti-class-II mAb and allosera. Site-directed mutagenesis and transfection were used to produce cells bearing wild-type or mutant class II molecules that were analyzed with mAbs by flow cytometry and with human allosera by absorption and subsequent microcytotoxicity assays. These studies demonstrate that the residue at DR beta position 58 plays a central role in at least three different mAb epitopes and an epitope recognized by anti-DRw11 allosera. Substitution of glutamic acid for alanine at position 58 of eight DR beta chains caused gain of binding of four mAbs to all of the mutant molecules, except DR(alpha,beta 4*0101). These data suggest that the side chains of DR beta 58 and DP beta 56 point outward from the alpha-helix and directly contact antibody.

The relationship between HLA-A, B, DQ, and DR antigens and asbestos-induced lung disease.

To evaluate the relationship between human leukocyte antigen (HLA) and both asbestos-induced pulmonary fibrosis and pleural fibrosis, we obtained HLA-A, B, C, DQ, and DR phenotypes in 42 long-term asbestos-exposed workers. Among these exposed workers, 15 had asbestosis (ILO > or = 1/0) on the chest radiograph and 18 had asbestos-induced pleural fibrosis. We found that there was an increased percentage of HLA-A29, HLA-B44, and HLA-Bw4 in the subjects with asbestosis. In addition, we observed a marginally positive relationship between HLA-A29 and the severity of pulmonary fibrosis. Similarly, there was a higher prevalence of HLA-DRw53 and DQ2 in the subjects with asbestos-induced pleural fibrosis. The presence of HLA-DQ2 was significantly related to the extent and type of asbestos-induced pleural fibrosis while HLA-DRw53 was not consistently related to the type or extent of pleural disease. Importantly, we observed that HLA-A29, HLA-B44, HLA-Bw4, HLA-DRw53, and HLA-DQ2 do not have a significantly shorter duration or latency of asbestos exposure. Moreover, none of the HLA haplotypes (A29, B44, Bw4, DRw53, and DQ2) that we found to be associated with radiographic manifestations of asbestos-induced lung disease were associated with the physiologic abnormalities that have been traditionally associated with asbestos-induced lung disease. The only exception was an isolated decrease in the residual volume associated with the presence of HLA-A29. These results suggest that the HLA-A29 phenotype may be associated with the development of asbestosis and the HLA-DQ2 phenotype may be associated with the development of asbestos-induced pleural fibrosis. However, these associations are not particularly strong, physiologic correlation is lacking, and previous studies do not support our findings.

HLA-B7 and HLA-DR2 antigens and acute posterior multifocal placoid pigment epitheliopathy.

Acute posterior multifocal placoid pigment epitheliopathy is a chorioretinal inflammatory disease occurring in young, healthy adults. Its cause is unknown, although it frequently follows a flulike illness. We reexamined 30 patients with documented acute posterior multifocal placoid pigment epitheliopathy to determine their HLA class I antigen (A and B) and class II antigen (DR and DQ) distribution. The HLA class I antigen B7 was found in 12 patients (40.0%) compared with 63 controls (16.6%) (relative risk, 3.38). The class II antigen DR2 was present in 17 patients (56.7%) compared with 107 controls (28.2%) (relative risk, 3.34). The specific role of HLA antigens in uveitis is unknown, but the finding of an increased prevalence of HLA-B7 and HLA-DR2 antigens in patients with acute posterior multifocal placoid pigment epitheliopathy suggests an immunogenetic predisposition to acquiring this disease.