Modulation of Idiotypic and Antiidiotypic Immunoglobulin G Responses in an Immune Thrombocytopenic Purpura Patient as a Consequence of Extracorporeal Protein A Immunoadsorption

Abstract: Immune thrombocytopenic purpura (ITP) is an autoimmune disorder characterized by the presence of antiplatelet antibody which sensitizes platelets resulting in their clearance by the reticuloendothelial system. Extracorporeal protein A immunoadsorption has been demonstrated to be of benefit in the treatment of this autoimmune disorder. In the present study, a patient with underlying systemic lupus erythematosus (SLE) presented with ITP. The patient received 14 immunoadsorption treatments and responded to therapy. During the course of immunoadsorption treatments, there was a decline in circulating immune complex (CIC) levels, antinu-clear antibody (ANA) levels, and antiplatelet IgG antibody levels. In addition, elevated levels of antiidiotypic IgG antibody detected before initiation of therapy were significantly reduced during the course of immunoadsorption treatments. This study suggests that specific autoimmune idiotypic IgG antibody and corresponding antiidiotypic IgG antibody responses may be modulated in association with extracorporeal immunoadsorption employing protein A/silica columns.     

Acute glomerulonephritis occurring during immunoadsorption with staphylococcal protein A column (Prosorba).

BACKGROUND: Apheresis of patient plasma by immunoadsorption with a staphylococcal protein A (SPA) column is used in a variety of autoimmune disorders. Leukocytoclastic vasculitis is an uncommon severe complication that can occur during immunoadsorption with SPA (Prosorba. METHODS: We report a case of immune complex glomerulonephritis occurring during Prosorba immunoabsorption in a patient with rheumatoid arthritis (RA). Using a Medline literature search and information provided by Cypress Bioscience/Fresenius Hemocare, we review renal complications associated with Prosorba immunoadsorption. RESULTS: We identified seven additional potential cases of glomerulonephritis (GN) in association with Prosorba immunoadsorption. Five of these patients were being treated for RA, and two for idiopathic thrombocytopenia purpura (ITP). Renal biopsies were performed on four patients, all of whom had evidence of immune complex GN. Among RA patients treated with Prosorba, the incidence of GN closely paralleled that of leukocytoclastic vasculitis at 1.75%. The presence of leukocytoclastic vasculitis was a significant risk factor for the development of GN (relative risk = 75.95, CI 7-1869, P = 0.00021). In contrast, among more than 10 000 ITP patients treated with Prosorba, there were only two potential cases of GN. The risk of developing GN in association with Prosorba immunoadsorption was significantly greater for patients with RA than for those with ITP (relative risk = 62.95, CI 10-453, P = 0.00002). CONCLUSION: This case series highlights the risk of GN among patients undergoing SPA immunoadsorption. The development of GN is associated with the presence of leukocytoclastic vasculitis. Patients with RA seem to be at particular risk.     

Immune Modulation Associated with Extracorporeal Immunoadsorption Treatments Utilizing Protein A/Silica Columns

Abstract: The Prosorba column is designed for the removal of IgG and IgG containing immune complexes from plasma. Clinical studies employing patients presenting with idiopathic thrombocytopenic purpura (ITP) indicate that this new form of therapy is effective in approximately 40% of treated patients. Responding patients exhibit a significant increase in platelet numbers associated with decreases in antiplatelet antibody and immune complexes suggesting the induction of immune modulation. Preliminary studies indicate that ITP patients presenting with antiplatelet IgG antibody are those most likely to respond. In addition, this subgroup of ITP patients also exhibit elevated levels of antiidiotypic IgG antibody, which may contribute to an exacerbation of the autoimmune process due to antigen mimicry of the platelet autoantigen. Interestingly, antiidiotypic IgG antibody levels appear to decrease in association with antiplatelet IgG autoantibody levels suggesting that removal of immune complexes composed of IgG autoantibody and platelet autoantigen and/or antiidiotypic IgG antibody may be related to the observed clinical responses. Additional studies with alloimmune patients refractory to platelet transfusion suggest that transfused platelet retention time may be increased as a consequence of immunoadsorption therapy. This clinical response appears to be related to decreases in IgG alloantibody, again suggesting the induction of immune modulation. Alloimmune thrombocytopenic patients also appear to present with elevated levels of antiidiotypic IgG antibody which may contribute to an exacerbation of the alloimmune process due to antigen mimicry of platelet alloantigen(s). Preliminary studies indicate that both IgG alloantibody and corresponding antiidiotypic IgG antibody levels appear to decrease during immunoadsorption therapy, which suggests that removal of these antibodies, possibly in the form of immune complexes, may be related to clinical responses. Finally, studies in rheumatoid arthritis patients suggest that immunoadsorption therapy may be of clinical benefit in this autoimmune disorder. Consistent with the results observed above, preliminary studies in patients responding to immunoadsorption treatments again suggest that there is a concomitant decrease in idiotypic IgG (rheumatoid factor) and antiidiotypic IgG antibodies levels during therapy.     

Changes of Circulating Antibody Levels Induced by ABO Antibody Adsorption for ABO-Incompatible Kidney Transplantation

ABO-incompatible kidney transplantation using immunoadsorption to remove anti-A/B antibodies has become a successful clinical practice. Since the data on the specificity of the ABO columns are controversial, the present study assessed the efficiency and specificity of the ABO immunoadsorption, the effect on total immunoglobulins and antibodies previously induced by vaccination. Anti-A/B antibodies were measured by agglutination and ABO flow cytometry, total IgG/IgM, carbohydrate- and protein-specific antibodies by nephelometry and ELISA. The first immunoadsorption not only efficiently reduced donor-specific anti-A/B IgM (81%) and IgG (56%) but also reduced compatible anti-A/B IgM (59%) and IgG (34%). The measurements of antidonor A/B antibodies by direct agglutination (IgM) or flow cytometry better represented the effective antibody levels than the indirect agglutination test (IgG). The median reduction of total IgM and total IgG levels after a single immunoadsorption was 34% and 18%, respectively. Antibodies against pneumococcus and haemophilus polysaccharide antigens were significantly reduced, whereas antitetanus and antidiphtheria protein antibodies were not affected. Intravenous immunoglobulin administration restored the protective anticarbohydrate antibody levels. In summary, immunoadsorption efficiently removed antidonor A/B antibodies, but was not specific for A/B antigens. Anti-A/B antibody levels as determined by ABO flow cytometry are useful to establish the minimal number of immunoadsorptions needed for successful ABO-incompatible transplantation.     

Protein A Immunoadsorption Therapy in HIV-Related Immune Thrombocytopenia: A Preliminary Report

Nine homosexual patients with immune thrombocytopenia were treated with autologous plasma that had been perfused over silica-immobilized Staphylococcus aureus protein A (SpA). Pretreatment platelet counts ranged from 10,000 to 98,000 cells/mm3 (mean: 54,000 cells/mm3). Six patients responded to therapy. Platelets increased by a mean of 95,000 cells/mm3 (p less than 0.007) and reached normal levels (greater than 150,000 cells/mm3) in four patients. Increased platelet counts are presently sustained in these four individuals after 5 months of follow-up. Increases in platelet counts significantly correlated with decreases in platelet-associated IgG (PAIgG), platelet-directed IgG (PDIgG), and immune complexes (CIC). PAIgG and PDIgG declined by a mean of 67% (p less than 0.003) and 58% (p less than 0.007), respectively. CIC decreased by a mean of 37% (p = 0.02). Complement was concomitantly activated in all four examined patients. C3a and C5a increased 23-fold and 2.6-fold, respectively, while total hemolytic complement decreased by 50%. Activated complement components and removal of CIC and IgG thus may contribute to the platelet-enhancing activity of SpA immunoadsorption therapy.     

In vitro removal of anti-HLA IgG antibodies from highly sensitized transplant recipients by immunoadsorption with protein A and protein G sepharose columns: a comparison

In the present study we compared the capabilities of sepharose-bound protein A versus protein G columns to remove in vitro lymphocytotoxic anti-HLA antibodies from sera of four highly sensitized renal transplant recipients (PRA70%). In none of the patients were protein A sepharose columns capable of completely removing anti-HLA antibodies, as demonstrated by the presence of residual alloreactive lymphocytotoxic activity in IgG 3 antibodies containing fractions eluted at pH 7. In contrast, no residual anti-HLA lymphocytotoxic antibody activity was found in fractions eluted at pH 7 from protein G columns. These data demonstrate that: (1) IgG 3 antibodies can be partly responsible for lymphocytotoxic anti-HLA reactivity in some sensitized renal transplant recipients and (2) at least in this patient category, in vitro immunoadsorption with protein G is more efficient than protein A sepharose columns in completely removing anti-HLA IgG antibodies from sera.     

Effect of protein A immunoadsorption on panel lymphocyte reactivity in hyperimmunized patients awaiting a kidney graft

Six hyperimmunized patients awaiting a kidney graft underwent immunoadsorption of separated plasma through a protein-A column to remove HLA antibodies. This procedure was partially limited by constant and rapid antibody resynthesis in spite of strong immunosuppression with cyclophosphamide and prednisolone. Only two patients could be grafted with previously positive--currently negative crossmatches. The first died of infection on day 40, having developed early chronic vascular rejection and the other returned to hemodialysis on day 285 after the development of transplant glomerulopathy. However, immunoadsorption seems to be effective in removing HLA antibodies having a titer below 1/2. Such extremely hyperimmunized patients should probably be excluded from the immunoadsorption program.     

Characterization of the anti-A antibody binding in an ABO-incompatible living donor renal transplantation

A 57-year-old female, blood group B, with polycystic kidneydisease, received an ABO-identical, HLA-A,B,DR 5-mismatchedrenal allograft in 1986. Due to graft artery thrombosis andvascular rejection, she lost the kidney 6 months after transplantationand developed HLA antibodies with a panel reactivity of 99%.Despite 5 years on a European waiting list for highly immunizedpatients, she was not offered a second kidney. An attempt toremove her HLA antibodies by plasmapheresis combined with cyclophosphamidetherapy did not succeed. Her 53-year-old HLA-identical, butABO-incompatible sister (blood group A₁) was then accepted asa donor. After immunoadsorption on Biosynsorb-A columns, transplantationwas performed. The post-transplant course was uneventful withoutany signs of rejection. Studies on the anti-A antibody bindingcharacteristics before and after immunoadsorption and aftertransplantation, showed that IgM and IgG antibodies recognizingthe A trisaccharide epitope based on the type 1,2, and 4 coresaccharide chains, were effectively removed by Biosynsorb-Aadsorption, but the column failed to remove anti-A antibodiesrecognizing the A type 3 antigen. These antibodies probablyrequires part of the core saccharide chain for binding. Thepresence of these antibodies did not seem to influence the outcomeof the ABO-incompatible transplantation.     

Mercury exposure in protein A immunoadsorption.

BACKGROUND: Immunoadsorption is increasingly used to treat antibody-mediated autoimmune diseases. To prevent microbial growth during storage, reusable protein A-Sepharose gel columns are primed with ethyl mercury thiosalicylate (thiomersal, 0.1% solution) and rinsed with phosphate buffer before use. In this study, we tested the hypothesis of systemic mercury exposure in protein A immunoadsorption. METHODS: Whole blood mercury levels were measured by atomic absorption spectroscopy before and after protein A immunoadsorption (11 patients, 26 treatments), anti-IgG immunoadsorption (eight patients, 13 treatments) and LDL apheresis (DALI and Therasorb systems; nine patients, 14 treatments). RESULTS: Patients treated with protein A immunoadsorption had significantly elevated baseline mercury levels compared with the other groups, which were not different from healthy controls. Following protein A immunoadsorption, mercury levels increased from 5.9+/-1.4 microg/l (mean+/-SEM, normal, <5 microg/l) to 32.3+/-5.7 microg/l, P<0.001). In one intensively treated patient, acute neurological toxicity developed at a mercury level of 107 microg/l. Symptoms abated slowly and did not recur after switching to a thiomersal-free system and chelation therapy. No mercury release to patients occurred in anti-IgG immunoadsorption or LDL apheresis treatments. CONCLUSION: This preliminary report suggests that protein A immunoadsorption columns primed with thiomersal during storage may cause a sustained increase of systemic mercury concentrations, which exceed current safety recommendations in a proportion of patients. Considering the potential for mercury-induced toxicity, every effort should be undertaken to reduce systemic mercury exposure, either by adding chelators to the rinsing solution or ideally by replacement of thiomersal.     

Anticipation of highly sensitised renal patients' immunoadsorption requirements by prescreening using protein A minicolumns

Abstract  We are able to subdivide highly sensitised renal patients who wish to enter our immunoadsorption programme into two groups; those who will require acute pretransplant immunoadsorption only and those requiring regular immunoadsorption prior to transplantation. This division of patients is based on the results obtained from laboratory assessment using protein A minicolumns. Patient's plasma is passed down a minicolumn for 6 times 10 min cycles, a sample of plasma is kept after each cycle for analysis by cell flow cytometric cross-match (FCXM). The samples are screened against cells from two normal volunteers, one expressing a previously mismatched Class I HLA antigen (MMA) to which the patient has raised persistent IgG antibodies, the other, whilst not expressing any MMAs, should express a cross-reactive HLA Class I antigen (XRA) to which the patient has formed persistent IgG antibodies. Patients are allocated into the acute pretransplant immunoadsorption group if, after 6 minicolumn cycles, the T cell FCXM vs XRA and MMA is reduced to less than 1 Log median fluorescence intensity shift above the negative control and that both these values have been reduced by at least 15 % from the preimmunoadsorption figure. If these criteria are not met, regular immunoadsorption is required under cover of cyclophosphamide. Eleven patients who have been allocated by these criteria have subsequently been transplanted without any incidence of hyperacute rejection.     

In Vivo and In Vitro Optimization of Depletion of IgM and IgG Xenoantibodies by Immunoadsorption Using Cell Membrane Proteins

Abstract: A system of immunoadsorption was developed for in vitro depletion of xenoreactive natural antibodies of classes IgG and IgM from monkey and human plasma. Porcine endothelial cell membrane proteins, platelet membrane proteins, and endothelial cells were used as affinity ligands, and cyanogen bromide-activated Sepharose 6 Fast Flow and Sepharose CL4B gels were used for chromatography. Adsorption capacity was evaluated by means of ELISA, immunonephelometry, and cytotoxicity testing. Several consecutive adsorption-desorption cycles were performed. Different parameters influencing immunoadsorption were examined: ligand density on the column gel, adsorbent-plasma contact time, ratio of plasma volume to immunoadsorbent volume, desorption conditions, and temperature. After 2 adsorption-desorption cycles, 99% and 82 to 85% of IgG and IgM antipig antibodies were adsorbed, respectively. Furthermore, there was a 74 to 77% decrease in cytotoxicity. In vivo, we observed that after one adsorption-desorption cycle, 97% of antipig IgG antibodies and 96% of antipig IgM antibodies were adsorbed, and there was an 85% decrease in cytotoxicity. The immunoadsorption method studied and optimized in vitro and in vivo therefore efficiently depleted xenoantibodies and reduced the cytotoxicity. Thus, it can be used in xenotransplantation experiments without eliminating nonspecific antibodies.     

Modulation of Idiotypic and Antiidiotypic Immunoglobulin G Responses in an Alloimmune Thrombocytopenic Patient Associated with Extracorporeal Protein A Immunoadsorption

Abstract: In the present case study, a patient with Non-Hodgkins Lymphoma underwent combination chemotherapy resulting in severe pancytopenia requiring transfusion support with blood products. The patient became refractory to random donor platelet transfusions and subsequently received five immunoadsorption treatments. The patient's clinical response to immunoadsorption therapy was assessed by monitoring platelet transfusion recovery and survival. In addition, changes in antibody responses were assessed. Early during the course of immunoadsorption therapy, antiplatelet immunoglobulin G (IgG) alloantibody was detected. There was a decline in antiplatelet IgG alloantibody levels by the last immunoadsorption treatment associated with increases in platelet corrected count increments after completion of immunoadsorption therapy. In addition, elevated levels of antiidiotypic IgG antibody detected early during the course of therapy were significantly reduced by the last immunoadsorption treatment. This case study suggests that specific alloimmune idiotypic IgG antibody and corresponding antiidiotypic IgG antibody responses may be modulated in association with extracorporeal immunoadsorption employing protein A/silica columns.     

Food Antigens, IgA-immune Complexes and IgA Mesangial Nephropathy

To investigate whether patients with IgA nephropathy have anexaggerated serum IgA response to ubiquitous food antigens wemeasured serum IgA antibodies to gliadin, ovalbumin, bovineserum albumin (BSA), ß-lactoglobulin and casein in120 patients and 53 normal controls, using ELISA. No significantdifferences were observed between patients and controls in serumIgA antibodies against each of the antigens tested. Moreover,no correlation was found between serum IgA antibodies and IgA-immunecomplexes (IgA CIC). However, nine patients but no controlshad an association of two or more IgA antibodies to dietaryantigens. Sixty-six per cent of these patients (vs 24% in the remainingpopulation) had IgA CIC, suggesting a possible involvement ofthese antibodies in the constitution of IgA CIC. Analysis ofsera by HPLC revealed that both monomeric and higher molecularforms of IgA antibodies were present, the latter being coincidentwith the peak of IgA CIC. Preincubation of sera with serial concentrations of the specificantigen decreased significantly IgA CIC, suggesting that inthis subgroup of patients IgA antibodies to food antigens (mainlyBSA) are involved in the formation of IgA CIC. BSA-containingIgA CIC were in fact demonstrated by ELISA using rabbit IgGanti-BSA coated plates and peroxidase-conjugated anti-humanIgA. The role of these CIC in the pathogenesis of IgA nephropathyneeds to be further elucidated.     

Studies on the removal of anti-pig xenoantibodies in the human by plasmapheresis/immunoadsorption

Abstract: Removal of human preformed natural anti-pig antibodies from the blood is a prerequisite before xenografting between pig and man can be performed. This work explores the effect of plasmapheresis and immunoadsorption (protein-A sepharose) on the reduction and recurrence of anti-pig antibodies in 14 patients. The anti-pig antibody changes were evaluated by lymphocy to toxic, hemagglutinating, and endothelial cell ELISA techniques. The changes induced showed a similar pattern with all three techniques used. In addition, plasma from plasmapheresis treatments were perfused through pig kidneys and the reduction of anti-pig antibodies was estimated by the mentioned in vitro techniques. The anti-pig antibody titers could be reduced to low levels, but not completely eliminated, by 3–4 plasmapheresis sessions. The titers gradually returned to pretreatment levels or higher in a period of 1–2 weeks. A few patients showed signs of a more rapid resynthesis reaching pretransplant levels in 3–4 days. Protein A immunoadsorption satisfactory removed IgG but not IgM antibodies. In vitro perfusion of pig kidneys at 37°C showed a rapid reduction of anti-pig antibody titers of 3–4 titer steps. The combination of 3–4 plasma exchanges followed by in vitro pig kidney perfusion completely removed all anti-pig antibodies. Reduction of the anti-pig lymphocyte and erythrocyte antibody titers by soluble oligosaccharides carrying terminal Galoc-epitopes was only partly successful. A 40–60% inhibition was achieved by 5–10 mg saccharide/ml serum and no clear inhibition difference between di- and trisaccharides was found. Inhibition of plasma obtained after 3–4 plasmapheresis treatments with soluble Galα1-di- and trisaccharides resulted in very low anti-pig titers. Therefore one feasible pretreatment procedure, before pig to human xenotransplantation could be plasmapheresis for major reduction of anti-pig antibody titer followed by neutralisation of the remaining antibodies by addition of soluble oligosaccharides or immunoadsorption with Galα-1-columns.     

Circulating immune complexes in regularly dialyzed patients with chronic renal failure

The prevalence of circulating immune complexes (CIC) was investigated using the C1q binding assay (C1q BA) and the conglutinin binding assay (Kg BA) in 200 patients undergoing maintenance hemodialysis. Increased C1q binding was found in 45% (87 of 194) of the patients, and the modified Kg BA gave elevated values in 31% (20 of 65). The prevalence of CIC was similar in American and Swiss patients, and in patients undergoing hemodialysis, self-dialysis or peritoneal dialysis. In patients with 'nonimmunological' renal diseases, CIC were detected with similar frequency. No change in CIC was noted during hemodialysis in 6 additional patients tested. The abnormality was not related to age, sex, duration of dialysis, hepatitis B antigenemia, bacterial infections, or transfusions. Anti-DNA antibodies were absent in all subjects tested and the results of the C1q BA were not changed by DNase digestion of eight sera with high C1q binding. Rheumatoid factor activity (RF) was detected in approximately one-fifth of the patients, and there was a direct correlation between positive C1q binding and RF. There was no correlation between CIC and lymphocytotoxic antibodies. This study demonstrated a high prevalence of CIC in dialyzed uremic patients and established its relationship to other immunological abnormalities.     

Bone marrow purging for multiple myeloma by avidin-biotin immunoadsorption

Avidin-biotin immunoadsorption, a technique based on the high affinity between the protein avidin and the vitamin biotin, has been used to remove neoplastic plasma cells from the bone marrow of patients with multiple myeloma. Buffy coat cells obtained from 25 patients were first incubated with monoclonal antibodies (MoAb) capable of recognizing plasma cell-associated antigens (i.e., 8A, 8F6, 62B1, and cocktails of 8A plus 8F6 or 62B1), then with biotinylated goat antimouse immunoglobulin, and passed over a column containing avidin conjugated to Sepharose GMB. Both non-linked and linked cells were analyzed by immunofluorescence and morphological staining. The results showed that over 98% of plasma cells were removed by using 8A or 8F6 alone, while 99.5% +/- 0.4 SD of plasma cell purging was achieved with 2 associated MoAb. In addition, the overall recovery of committed granulocyte-macrophage (CFU-GM),* erythroid (BFU-e), and multilineage (CFU-GEMM) progenitors after column treatment ranged from 39% +/- 15 SD to 50% +/- 6 SD, from 15% +/- 2 SD to 39% +/- 7 SD, and from 16 +/- 4 SD to 64% +/- 10 SD, according to the MoAb employed. On this basis avidin-biotin immunoadsorption appears to be a suitable technique for ex-vivo manipulation of bone marrow infiltrated by neoplastic plasma cells.     

Heart transplantation across the antibodies against HLA and ABO

We have intentionally performed heart transplantation in a 5-year-old child, despite the most unfavourable risk factors for patient survival; the presence of high level of antibodies against donor's human leucocyte antigen (HLA) class I/II and blood group antigens. Pretransplant treatment by mycophenolate mofetil, prednisolone, tacrolimus, intravenous immunoglobulin, rituximab, protein-A immunoadsorption (IA) and plasma exchange reduced antibody titres against the donor's lymphocytes from 128 to 16 and against the donor's blood group antigen from 256 to 0. The patient was urgently transplanted with a heart from an ABO incompatible donor (A(1) to O). A standard triple-drug immunosuppressive protocol was used. No hyperacute rejection was seen. Antibodies against the donor's HLA antigens remained at a low level despite three acute rejections. Rising anti-A(1) blood group antibodies preceded the second rejection and were reduced by two blood group-specific IAs and remained at a low level. The patient is doing well despite the persistence of donor-reactive antibodies.     

Circulating immune complexes in patients following clinically curative resection of colorectal cancer.

Sixty-nine patients have been followed prospectively after curative resection of Dukes-Kirklin B-2 or C colorectal cancer. Serial plasma samples were studied in selected patients to determine changes in circulating immune complex concentrations (CIC) following primary tumor resection, and to compare serial plasma CIC and carcinoembryonic antigen (CEA) levels. CIC was determined in an average of seven serial samples per patient by inhibition of antibody-dependent cell-mediated cytotoxicity (ADCC). CEA assays were performed by the Hanson Z-gel method. Two distinct patterns of serial CIC have emerged. In seven patients with no known tumor recurrences, serial CEA levels and CIC oscillated regularly and were inversely related. In seven of eight patients whose tumors recurred, both CEA and CIC rose together. In three patients with elevated plasma CEA levels due to inflammatory bowel disease, serial Ag-Ab complex concentrations did not vary, nor did separated Ag or Ab fractions inhibit ADCC. These data suggest that, in patients following curative resection of colorectal cancer, serial changes in circulating immune complexes may discriminate between transient CEA elevations which occur despite no known tumor recurrence and tumor recurrence which is beyond the capacity of adequate host antitumor defense.     

Immunoadsorption of anti-HLA antibodies for highly sensitized patients awaiting renal transplantation.

Fifteen end-stage renal disease patients with high titres of panel reactive (PRA) antibodies were treated with immunoadsorption (IA) on sepharose-bound protein A columns in order to remove anti-HLA antibodies and facilitate transplantation. Infectious complications were not observed after IA and transplantation, and the procedure was well tolerated. In spite of the use of adjunctive immunosuppressive treatment with cyclophosphamide and prednisolone, this method produced only variable effects in lowering panel reaction antibodies, and was hampered by high de novo resynthesis of anti-HLA antibodies. Patients whose pre-IA antibody titre was greater than or equal to 1:64 clearly did not benefit from the procedure, but other immunological criteria were not predictive of efficacy. Twelve patients were transplanted on the basis of a negative cross-match with current serum, historical sera being retrospectively tested. Surprisingly, seven patients received a well-matched graft with both pre- and post-IA negative cross-matching. Graft survival was 86% in this group. Conversely, in the group of five transplants which were performed in recipients having a positive historical cross-match with the donor, graft survival was only 40%. One patient died with a functional graft, and two grafts failed due to hyperacute humoral rejection. Humoral rejection in a third patient was successfully treated by a second IA course and administration of polyclonal IgG. We conclude that IA is a safe procedure for managing hyperimmunized transplant candidates. However, its efficacy remains variable, and a better definition of patients who should benefit from IA needs to be found.     

Extracorporeal removal of anti-HLA antibodies in transplant candidates

We report on the results of a clinical trial in which 14 transplant candidates were treated with an extracorporeal immunoadsorption system using Protein A that selectively removes immunoglobulin from plasma; we also assessed the dynamics of anti-HLA antibody as a model of IgG removal and re-equilibration, as well as the clinical safety of the procedure. At the end of a treatment course, plasma IgG levels were reduced by 90% +/- 8% of control values (P less than 0.01). In contrast, albumin levels were reduced by only 15% (P less than 0.05). Specific cytotoxic anti-HLA antibody titers were reduced by approximately 18-fold. Panel reactivity was measured as the proportion of a 40-member cell donor panel killed by patients' serum in the presence of complement; in nine of the 14 patients, there was a significant reduction in this parameter (range, 23% to 87%). During the 4-week follow-up period, anti-HLA antibody titers returned to baseline levels. There were no remarkable changes observed in blood chemistries, nor were there any unanticipated adverse reactions seen in the patients treated. We conclude that selective extracorporeal immunoadsorption is a safe and effective way of removing IgG-type antibodies, with potential application to reduction of HLA antibodies in transplant candidates.