Clinical Tolerance: the end of the beginning

This is not the end; this is not even the beginning of the end, but it is the end of the beginning. Winston Churchill, 1943. 2003 was the fiftieth anniversary of the discovery of actively acquired immunological tolerance by Medawar and colleagues (1,2). This experiment is considered the beginning of our ability to manipulate the allograft rejection reaction. In Medawar’s own words, the discovery of tolerance had a moral significance-it encouraged surgeons to proceed with clinical transplantation. 2003 was also the year that transplant physicians and surgeons seriously began to report their clinical attempts at inducing some form of clinical tolerance. The Editors of Transplantation are pleased to present this Forum on current clinical attempts to induce tolerance. In a sense, it represents the end of the beginning stages of our quest for tolerance. Hopefully it will inspire transplantation specialists to pursue other clinical tolerance strategies. This Forum was generated by the paper of Starzl and colleagues, who describe lessons of “organ-induced” tolerance learned from their enormous experience and perceptive insights. Success rates in early transplantation were much lower than currently achieved, but the number of long-term, drug-free survivors apparently was much greater. Almost 20% of live-related renal allograft recipients transplanted in Denver in the early 1960s survived over almost four decades, seven of nine off immunosuppression for 3 to 34 years. Modern multi-drug immunosuppression has increased survival, reduced acute rejection, but has been associated with chronic rejection and the need for long-term immunosuppression. Calcineurin-inhibitor drugs lowered the exorbitant mortality and high rejection rates after liver transplantation, but the number of tolerant (drug-free) liver recipients did not increase. Starzl and colleagues propose that the typically dominant HVG (host-versus-graft) response to the transplant results from migration of passenger leukocytes to host lymphoid organs (3). This response proceeds to rejection in untreated recipients. In certain experimental models of spontaneous tolerance, the HVG reaction is too weak to eliminate migrating leukocytes (particularly with the liver) and the response is exhausted and deleted. Because the exhaustion-deletion is never complete, maintenance of deletion depends on persistence of peripheral donor leukocytes (microchimerism). Posttransplant immunosuppression (especially depleting agents) reduces the anti-graft response to a readily deletable range. The anti-graft response can also be deleted posttransplant by immunosuppression. Such immunosuppression must not be so powerful that antigen specific immune activation is prevented, since the derivative event of exhaustion-detection may be curtailed. Starzl and colleagues applied these concepts to 82 patients awaiting kidney and other organ transplants (4). Patients received preoperative thymoglobulin (5 mg/kg, single infusion) followed by only tacrolimus monotherapy twice daily. After four months, tacrolimus monotherapy was weaned to varying levels from once daily to once per week. Immunosuppression morbidity was virtually eliminated. Patient survival was 95% and graft survival was 88% at 13–18 months, with 43 of 72 patients surviving grafts on spaced monotherapy. There was a high incidence of biopsy proven rejection (called immune activation by the authors) in weaned kidney patients that responded to steroid boluses and/or adjusted tacrolimus dosage. The high incidence of steroid responsive rejections was reminiscent of the frequent low-grade, easily reversible rejections observed in successful recipients of donor specific blood transfusions. Patients responding to rejection therapy could resume weaning. This important and unique study showed that excellent short-term patient and graft survival could be obtained with relatively modest tacrolimus monotherapy after lymphocyte depletive induction therapy. However, the high incidence of rejection reactions (immune activation) with this strategy may presage late organ injury or loss. Only time will tell. It is incumbent on the authors to provide long-term follow-up of these patients. Sir Roy Calne’s pragmatic approach emphasizes that a strict mechanistic definition of tolerance is not helpful, since an operationally tolerant state that could be achieved by various mechanisms is all that is needed for the welfare of the patient. He argues that tolerance certainly can be complete or partial. It could be vulnerable to acute or chronic rejection and variations and exacerbations of innate immunity. He suggests that a minimal baseline, nontoxic dose of maintenance immunosuppression could guard against these possibilities-hence the concept of almost or “prope” tolerance. Indeed, some patients might not need even minimal maintenance immunosuppression, but at present this cannot be determined except by drug withdrawal. Although complete tolerance (zero requirement for immunosuppression) is the ideal goal, heterogenicity of donor-recipient combinations, immune status, underlying disease, etc. might make complete stable tolerance difficult to achieve. Furthermore, if tolerance is maintained by regulatory T cells, then this might be a relatively fragile balance easily upset by other events, such as a coincidental infection. In several experimental models, long-term graft acceptance is achieved with minimal immunosuppression after induction treatment with various anti-T cell antibodies. Calne used Campath-1H, alemtuzumab, a humanized anti-CD52 monoclonal antibody, which is a powerful lytic agent for both T and B cells that does not affect bone marrow cells. Thirty-one renal allograft recipients received two peritransplant doses of 20 mg Campath-1H. Neoral to achieve a concentration of 100±25 ng/ml was not started until day +3, leaving a 48 hour window in which a perceived engagement of certain donor and recipient donor marrow derived cells might beneficially occur (the so-called window of opportunity). Twenty-nine patients have sustained graft function five years after surgery, with 26 receiving only cyclosporine monotherapy; only three patients are on additional prednisone and Azathioprine. Only five patients had episodes of impaired renal function, four with proven rejection. Immunosuppressive side effects are relatively minimal. Overall this series is impressive, achieving successful monotherapy in a high percentage of patients for five years. It is easy to apply and surprisingly low cost. Late side effects of Campath 1H have not been detected. Other questions remain. What is the best low maintenance immunosuppression regimen (sirolimus, tacrolimus)? Could some patients be weaned further? What is the optimal dose and timing for Campath use? Would ATG or ALG be just as effective? Would donor specific bone marrow infusion enhance the effect? Strober’s group has used total lymphoid irradiation based protocols for inducing tolerance in humans. They treated 28 cadaver kidney recipients with a TLI for 7 to 9 weeks pretransplant, followed by ATG and no donor bone marrow infusion. Maintenance immunosuppression was only 10 mg prednisone per day. Eleven of 28 patients showed no rejection episodes through the first year and 9 of 11 showed specific unresponsiveness to donors in MLR. Five years later, three patients were completely off immunosuppression with excellent renal graft function, and exhibited donor specific unresponsiveness in vitro by MLR and cell mediated lymphocytotoxicity (CML) assays. One patient had normal function off all drugs 12 years later without detectable microchimerism. These authors developed a posttransplant TLI based regimen for tolerance induction in rodents using TLI, ATG, and donor specific bone marrow infusion. They describe four patients given combined donor hematopoietic cells and kidney transplants followed by posttransplant TLI and ATG administered over two weeks. Donor cells were CD34+ progenitor cells purified from peripheral blood, cytokine (G-CSF) mobilized bone marrow cells obtained by leukopheresis. Two were HLA mismatched unrelated and two were haploidentical relatives. Posttransplant immunotherapy was prednisone and cyclosporine; both drugs were to be withdrawn if patients met the following criteria: development of macrochimerisms in the first three posttransplant months, no clinical rejections or rejection on protocol biopsy, specific unresponsiveness to donors in MLR. Two patients fulfilled these criteria and were taken off immunosuppression at 12 months. These patients had mild rejection episodes five months later that resolved with steroids. All four patients have normal graft function without evidence of chronic rejection and are maintained on low doses of one or two immunosuppressive drugs. The authors suggest their inability to induce tolerance might have related to failure to achieve sustained chimerism and failure to withdraw cyclosporine in more rapidly, i.e. prolonged cyclosporine therapy prevented tolerance induction. Alternatively, their therapy withdrawal criteria in all patients may have been too severe: there is experimental evidence that the salutary effect of donor bone marrow can be obtained without macrochimerism and that positive MLR reactions exist in operationally tolerant recipients. Knechtle and Billingham describe tolerance as a metastable process that involves a dynamic regulated immune response to the allograft. They describe studies in Rhesus monkeys in which profound T lymphocyte depletion produced by anti-CD3 immunotoxin induced a metastable tolerance to renal allografts. Tolerance which could be broken during a time posttransplantation (metastable period) by a strong donor immunogen (donor skin graft) was mediated in part by donor antigen-specific regulatory T cells with a CD4+ TGFβ (latent)+ phenotype. Tolerant monkey kidneys showed the presence of such cells that were lost with tolerance breakdown and replaced with invasive, TGFβ-negative infiltrates in late rejecting kidneys. These donor specific regulatory T cells emerge concomitant with effector T cells from the initial state of non-specific immunosuppression induced by anti-CD3-IT. Over time, the equilibrium matures to a more stable donor specific tolerance evidenced by down-regulation of host anti-donor lymphocyte responses. In addition to T cell depletion by anti-CD3-IT, activation through CD3 may also be a necessary precursor to development of TGFβ-dependent T cell regulation. The immunohistologic feature of CD4+ TGFβ (latent)+ T-regulatory cell infiltrates may provide a diagnostic marker for metastatic tolerance. They describe an HLA identical patient off immunosuppression for over 30 years (creatinine 1.2). Using their unique trans-vivo DTH assay, they showed this tolerance is metastable in that donor minor antigen-specific CD8+ CTL effectors are still present in the peripheral blood held in check by CD8+ T-regulatory cells whose suppressive action is mediated by TGFβ and also IL-10. They noted that 100% of HLA identical recipients still taking immunosuppressive drugs showed donor-specific antigen regulatory T cells, versus only 10–50% of patients with lesser degrees of match. They propose that persistent donor leukocyte microchimerism can serve to stabilize metastable tolerance. Finally, they summarize attempts by several groups using Campath-IH for T cell depletion to induce tolerance to human renal allografts. Seven of 7 patients given only Campath-IH rejected early but were all rescued successfully. When combined with low dose rapamycin similar to the cyclosporine study by Calne, 30% of recipients sustained early rejection, half of which were aggressive humoral rejections. Again patients could be rescued, and approximately 60% remain on rapamycin monotherapy at 3 years. Addition of two drugs (calcineurin inhibitor plus mycophenolate) has reduced the rejection rate to 10–15% and when combined with three drugs (addition of prednisone) the rejection rate falls to 10%. Whether the intensive T cell depletion achieved by Campath IH will permit later immunosuppressive drug withdrawal remains to be seen. Such withdrawal will be aided by monitoring assays for tolerance markers; several such markers are proposed. Miller and colleagues present a critique to the concept that human tolerance is essentially achievable via only the clonal depletion and exhaustion mechanism(s). They argue that relatively little positive or negative functional data exists about immune mechanisms in patients off immunosuppression, and cell regulatory mechanisms have not been excluded. Also, reliance on only deletion/exhaustion to produce tolerance assumes that the adult vestigial human thymus is continually active in producing life-long clonal deletion as an additional participating mechanism for operational tolerance in some human transplants. Indeed the successful experimental induction of tolerance in adult-thymectomized rodent recipients mitigates against the necessity for this ongoing requirement. Miller suggests that rather than restricting our mechanistic thinking to just clonal exhaustion maintained by ongoing classical deletion we might broaden the concept to include elimination of the remaining undepleted alloreactive clones (after the initial ablation) by programmed cell death, anergy, and regulatory cell mechanisms-in short, peripheral tolerance! They offer convincing evidence for regulatory cell effects from their clinical studies of renal allograft recipients given donor-specific bone marrow-a biological reagent with built in immunoregulatory activity. In their studies, cadaveric kidney graft recipients given donor bone marrow exhibit higher graft survival and decreased chronic rejection after six years. Microchimerism levels increased in recipient iliac crest marrow over time (but not in the peripheral blood). Purified posttransplant iliac crest marrow chimeric cells of either donor or recipient phenotype inhibited residual anti-donor recipient immune responses in MLR indicating an ongoing regulatory role for chimeric cells of both donor and recipient origin with a degree of donor-directed specificity. They describe studies in non-chimeric bone marrow showing strong inhibitory property of several bone marrow cell subpopulations suggesting that they could promote clinical unresponsiveness. They also speculate that specificity of operational tolerance occurs late after transplantation, i.e. that afferent and efferent immune alterations eventually become established, allowing a protective state in the graft as more general host immune reconstruction recovers. The clinical tolerance studies of the Miami group are important in that they demonstrate clearly a late salutary effect of donor specific bone marrow infusions. Many of the patients are on minimal immunosuppressive therapy but none has been subjected to drug withdrawal. Cosimi and Sachs describe the use of a non-toxic, non-myeloablative tolerance protocol employing total body irradiation, thymic irradiation, polyclonal ALG (ATGAM), splenectomy, donor bone marrow, and a short course of cyclosporine which they applied to renal allograft non-human primate recipients. Approximately 70% of animals so treated survived long-term off any immunosuppression; the longest one is currently over 10 years with normal function and histology. No graft-versus-host disease occurred. Chimerism was transient and low level. Allografts applied as late as three months after the conditioning regimen enjoyed prolonged survival at a time when lymphocyte counts were near normal. Renal allografts placed after four months were rejected. They hypothesized that early tolerance develops primarily due to migration of appropriate bone marrow elements to host thymus where clonal deletion occurs. Later, the kidney itself appeared to be involved in inactivation of new and/or residual alloreactive T cells via peripheral mechanisms. The authors extended this regimen to patients with renal failure secondary to multiple myeloma. Five patients with HLA matched donors were prepared with cyclophosphamide (in place of TLI), thymic irradiation, and postoperative ATGAM for simultaneous kidney transplantation and donor bone marrow infusion. Cyclosporine was begun on day -1 and continued postoperatively for 2–3 months. All five patients sustained normal renal function in two days. Cyclosporine was discontinued in four patients from day 40 to 77 while one patient continued cyclosporine through 380 days because of mild GVHD. There was no rejection in any patient; renal function remained normal in all except one who sustained recurrent myeloma with myeloma nephropathy. Additional donor lymphocyte infusions (DLI) were given for anti-tumor effect. T cell chimerism gradually declined and was undetectable in three patients while the remaining two were converted to full donor hematoporesis after DLI with mild GVHD developing in both. In vitro studies revealed persisting anti-donor CTL responses in some patients but no responses in MLR. There has been no evidence of acute or chronic rejection in sequential biopsies of their renal allografts. The MGH group has extended their protocol to two patients with ESRD without hematological malignancy and without HLA-matched donors. The protocol used in the myeloma patients was modified so that medi-507 (a humanized anti-CD2 monoclonal antibody) replaced ATGAM and donor lymphoid infusions were not given. The first patient had suffered severe multiple complications of conventional immunosuppression following kidney transplantation eight years earlier, necessitating drug withdrawal and return to dialysis. At this time the patient is off all immunosuppression for four months and appears to be tolerant by clinical parameters and in vitro assays. A second subject is currently four months posttransplant; a one-month biopsy showed C4d deposition and conventional immunosuppression has been added to treat possible humoral rejection. Thus far, no alloantibody is detectable and current immunosuppression is being tapered. Nevertheless, the appearance of a B cell mediated anti-graft response after tolerogenic conditioning is a cause for some concern. The authors thus far have five of seven patients off all immunosuppression-a remarkable and commendable achievement. These studies underscore the importance of donor bone marrow infusion to achieve donor specific tolerance/hyporesponsiveness. In the final (and perhaps most important) for human as they to organ transplantation These are based on the single no transplantation must demonstrate over as currently transplantation is a to organ to the to which they from a relatively that clinical tolerance remains a that is an appropriate subject to but it is not an any cost. He from the of to underscore several to transplantation and that all clinical study this is of the of patient is or studies of patients to a specific without are and appropriate in so far as they provide a for that the concept of with appropriate in this is the that is are and patient monitoring is He also emphasizes that the organ transplanted in tolerance be such that failure does not with Also, have the most to from tolerance some they might not be ideal initial for tolerance No clinical proceed without protocol by an which rather than form or Transplantation for A is for long-term five year the of which is an of the Furthermore, in tolerance rejection can occur late after immunosuppression withdrawal so follow-up with protocol biopsies are essentially an requirement. that be based on studies evidence obtained in rodents to Nevertheless, to this requirement are with the of Campath-IH treatment from human treatment to transplantation therapy being a A is also for the concept that addition of any immunosuppressive drug to a regimen be by in In perhaps the most important of this that the of immunosuppressive the of in transplantation to be the not the tolerance in the to results for patients in the In studies, the group must of course still a to the In the protocol must be with the best when the of the the the results of therapy and the of normal be In of a tolerance regimen of that can be applied to only a number of patients if successful has than a therapy that can be applied to all similar where are we with clinical we have learned a All of the attempts drug or withdrawal have varying degrees of and they have not any over the short given some of induction over currently protocols of maintenance immunosuppression. of immunosuppression after lymphocyte in is not Also, it is that treated low-grade, rejection reactions in the of immune mediated does not of high dose immunosuppression. These permit of tolerance drug withdrawal studies without that be to rejecting recipients. Thus far, all clinical drug some of recipient lymphocyte it is that a number of such will be in appropriate the role of donor bone marrow infusion to be and Donor bone marrow infusions have induced specific to or graft in since the first of bone marrow induced tolerance in treated The salutary effect of donor bone marrow in the two studies in this Forum this clinical studies are needed to the general use of donor bone marrow as a biological reagent to tolerance. Finally, the for appropriate clinical tolerance have been and clearly the clinical transplant will be to pursue tolerance