|Home | About | Journals | Submit | Contact Us | Français|
For decades the application of allogeneic blood and marrow transplantation has been a curative treatment for inherited non-malignant disease such as thalassemia, sickle cell disease, immunodeficiency diseases, and storage disorders and for acquired hematopoietic disorders such as aplastic anemia. The premise that lymphoablative conditioning and allogeneic or autologous transplant could prevent progression or reverse organ damage from inherited or acquired autoimmune disease derives from the pioneering animal experiments of Good, Ikehara and van Bekkum (1, 2). Still, a decade elapsed between the preclinical studies and human trials. Appreciation of the potential for cure of autoimmune disease was bolstered by illustrative experiments in nature wherein patients with coincident autoimmune disease and hematologic malignancy or aplastic anemia remained in long-term remission of both diseases after allogeneic transplantation (3). As predicted by Thomas, a variety of problems and opportunities encompassed these first clinical translations (4). Now with initial trials in several autoimmune diseases published with sufficient follow-up for determinations of safety and efficacy, it is opportune to review the biology and results to date and glimpse from both sides of the Atlantic into the future of transplantation for autoimmune diseases.
Current concepts on the pathogenesis of autoimmune disorders attribute a crucial role to T and B cells inappropriately recognizing self antigens and initiating a cell-mediated or humoral reaction, or both, resulting in inflammatory tissue and vascular damage (5). Treating autoimmune disease with antigen-specific tolerization has been an ambitious but largely elusive goal and both pharmaceutical and academic-driven drug development efforts have targeted shared effector or regulatory pathways with immuno-suppressive/modulatory compounds.
Autologous hematopoietic cell transplantation (HCT) is being evaluated as treatment for severe forms of immune-mediated disorders including multiple sclerosis (MS), systemic lupus erythematosus (SLE), systemic sclerosis (SSc) and rheumatoid (RA) or juvenile idiopathic arthritis (JIA). The goal of this therapy is to induce medication-free remission from disease activity by correcting the immune aberrations that promote the attack against self tissue (“immune repair”).
Collagen-induced arthritis (CIA) and experimental autoimmune encephalomyelitis (EAE) are examples of antigen-induced autoimmune diseases, and serve as models for human RA and MS, respectively. Data in both disease models suggest that tolerance induced by autologous HCT can prevent autoimmunity even after antigenic re-encounter. In van Bekkum’s studies of bone marrow transplantation (BMT) in experimental models, amelioration of autoimmune disease was observed not only after syngeneic but also after autologous or pseudo-autologous BMT [reviewed in (6)]. Remarkably, arthritic rats treated with syngeneic BMT did not relapse with CIA even after being re-immunized with the antigen (2). Syngeneic transplantation also conferred protection from disease relapse in EAE (7). To explain these observations we must postulate the induction of protective changes of the immune system not linked to a correction of an underlying stem cell defect. We have recently observed that BMT applied to mice in the late phase of EAE development resulted in different clinical outcomes. Numbers of activated macrophage/microglial cells were significantly greater in mice that progressed, and tracking of green fluorescent protein-transduced BM cells showed the endogenous origin of the activated microglia (8). Therefore, tissue-specific factors such as the persistence of local inflammatory cell types may influence the clinical outcome independent of the effects of BMT on the peripheral adaptive immune system (T and B lymphocytes).
Early studies on immune reconstitution following autologous transplant for both autoimmune diseases and cancer showed a profound lymphopenia in the first year after transplantation. The cytopenia was observed to affect the lymphocyte subsets differently, the kinetics of reconstitution likely depending on different timing of recovery for each cell type. Whereas B cells, natural killer (NK) cells, and CD8+ T cells display a rapid and complete reconstitution to pre-transplantation levels, the recovery of CD4+ T cells has consistently been observed to be delayed, and often incomplete. By extending longitudinal follow-up of patients, recent studies have shown a recovery of the number of CD4+ T cells after a 2 year follow-up in young adults treated for MS (9) and RA (10), and after 12 months in children with JIA (11). The observation that quantitative recovery of lymphocytes was not correlated to inflammatory activity or disease relapse revealed that numeric immune deficit is an insufficient explanation for a prolonged absence of autoimmune disease activity after autologous HCT.
The rationale for using autologous HCT in autoimmune disease is to purge the existing immune system and regenerate a new and healthy repertoire of immune cells. However, the notion of an immune “resetting” remained conjectural until Muraro et al. demonstrated the regeneration of a new, naïve T cell repertoire emerging from the thymus of patients with MS who had been treated with ablative conditioning and autologous HCT (9). In this study, a detailed analysis of T cell receptor repertoire showed the regeneration of a different and more diverse TCR repertoire post-transplant. Thymic reactivation, expansion of naïve T cells following autografting and improved repertoire diversity were subsequently also demonstrated in individuals with SLE (12).
Important lessons also come from studies on antibody responses to foreign antigens such as after vaccination or re-vaccination preceding or following autologous transplant. In a recent study, immunoablative conditioning and autologous HCT eliminated immunological memory for a neoantigen given after the graft harvest, and diminished, although did not completely eliminate, the immunological memory for a recall antigen boosted before harvest following non-rigorous T-cell depletion of the autograft (13). Serological evidence of attenuation of immunological memory suggests that the B cell compartment, until now less extensively investigated, may also undergo a renewal through autologous HCT.
Potentially pathogenic autoimmune responses are generated not infrequently and failure of tolerance towards self may also require the failure of protective immunoregulatory mechanisms. Autoimmune disease can therefore be regarded as the final outcome of a series of events that likely include not only a genetic susceptibility, but also the failure of the checkpoints available to prevent autoimmunity following exposure to environmental challenges, such as infections (Figure 1). It is reasonable to postulate the normalization of immune regulatory mechanisms could play a role in the suppression of autoimmunity following autologous HCT. The CD4+ CD25+ expressing the forkhead transcription factor 3 (FoxP3) cells are potent suppressors of immune responses that are generated in the thymus both in rodents (14) and in humans (15). CD25high FoxP3+ CD4+ T cells were reported to be more resistant to irradiation than effector cells and mediated the amelioration of experimental graft-versus-host disease (16). In EAE rats, there was an increase of CD4+CD25+ T cells after syngeneic BMT and this was seen in connection with attenuation of active disease and protection from induction of relapses (17). Longitudinal enumeration of CD4+CD25 high T cells in children with JIA studied following autologous HCT showed recovery of the pre-treatment frequency at 6 months post-transplant and a continued increase for the remaining 12-month follow-up. Their frequency correlated directly with clinical remission. Therefore, reinstallation of immune regulation could be involved in long-term tolerance post-transplant.
As detailed above, autoimmune diseases result from failure of an organism to recognize its own parts as “self”, thereby producing an auto-aggressive response. The components of this response are as pleiomorphic as the immune system itself, ranging from cells and molecules of the innate immune system such as dendritic cells and NK cells to the tightly regulated members of the adaptive immune system. Targeting these individual components with chemical or biological agents has been extremely effective in controlling symptoms in many patients, but still some patients do not respond sufficiently and may lose life or vital organ function or suffer severe toxicity from treatment. Also, no therapy to date has induced long-term drug-free remission in any autoimmune disorder.
Based on animal models and anecdotal experience of HCT patients with coincidental autoimmune diseases mentioned above, it was proposed that by eradicating the whole immune system, tolerance could be re-established during the immune reconstitution which follows lymphoablation. As in other HCT protocols, it was hoped that the early increased transplant-related toxicity, compared with standard of care, would be offset by a later increased disease-free survival.
Colleagues from hematology, rheumatology, immunology, neurology and gastroenterology sat together 14 years ago in Basel (18) and Seattle (19) to work out a structured research agenda which had as its mission statement two main objectives:
Autologous HCT was initially chosen due to its lesser toxicity compared with allogeneic HCT. A limited number of protocols were proposed to allow comparison of more aggressive myeloablative and reduced intensity regimens. At the time it was not known whether complete eradication of all autoreactive immune competent cells was required, or rather simply an autoimmune “debulking” to allow natural immune regulation to be re-established. A gratifying international scientific collaboration became established which remains today.
From the first case report, a 45 year female with untreatable pulmonary hypertension and SSc (20), through the small case series and then phase I/II studies, there was a strong impression that autologous HCT influenced the natural history of several autoimmune disorders, including SSc, MS, RA, JIA, SLE, and Crohn’s Disease (CD). In addition, impressive positive results were also seen for less common disorders such as chronic inflammatory demyelinating polyneuropathy (CIDP) and various vasculitides.
A first mega analysis on 473 patients from the EBMT in 2005 showed that 11% has died, either from treatment-related toxicity (TRM) (7%) or disease progression (5%), adjusted to an average three year follow-up (21). Of the patients evaluable for response (n= 299), 81% responded and this was sustained in 71%. Not reported then, but subsequently demonstrated, were many of these responders with apparent long-term, drug-free remission demonstrating tissue remodelling. Conditioning regimens were classified as low, intermediate or high intensity, and although responses were better in the high intensity regimes, the associated increased toxicity with these initial trials was considered unacceptably high. Subsequent protocol designs for the prospective RCTs in Europe were based on the intermediate intensity regimens, mostly consisting of mobilization with cyclophosphamide (CY) and G-CSF followed by CY (or BEAM) and ATG plus or minus CD34+ cell selection. Randomized trials have been designed for SSc, MS, CD, RA and CIDP.
Events Along the Way, In December 1993, a landmark paper was published showing a dramatic and hitherto unseen response in 20 RA patients over 8 weeks to the TNF-a blocking agent, the chimeric monoclonal antibody, infliximab (22). Toxicity was modest and this experience opened an era of biopharmaceuticals, first revolutionising the treatment of RA, then rapidly spreading to CD, MS and now becoming established or investigational therapy in most autoimmune diseases.
Although none of these offer “cure”, the toxicity is rather low and the trials are supported by the pharmaceutical industry. This impacted on the recruitment of patients with autoimmune disease onto transplant trials, since while best results with HCT are seen in early, reversible autoimmune disease, such patients are also suitable for less toxic biopharmaceuticals. Of note, so far an effective disease modifying agent for SSc is not available for this highly morbid and mortal disorder.
In 2001, the European Parliament enacted the European Union’s Clinical Trials Directive 2001/20/EC which has had a profound effect on clinical trials in Europe. This is a lengthy document specifying all aspects of the clinical trials process and has proven complicated for investigator-initiated trials such as studies of HCT for autoimmune disease which are often university based and extend across national borders. This regulatory burden has hindered the clinical trialist, especially as it relates to professional indemnity.
Table 1 presents the current status of RCTs in Europe as of August 2009. The Autologous Stem cell Transplantation International Scleroderma Trial (ASTIS) has almost completed enrollment and although deaths have occurred in both arms, the independent safety committee has adjudicated that no unexpected toxicity occurred. The Autologous Stem cell Transplantation International Rheumatoid Arthritis trial (ASTIRA) never started due to a plethora of biopharmaceutical agents for RA including anti TNF-a, IL-1, IL-6 and co-stimulation blockade.
SSC: some patients have achieved complete remission including unexpected and dramatic clinical and biopsy regression of dermal fibrosis as well as normalization of the microvasculature (23). SLE: in a small series of patients, complete clinical remissions as well as loss of autoantibody (antinuclear antibody) have been described, suggesting a true resetting of autoimmunity. This has been attributed to eradication of long lived plasma cells. MS: both clinical improvement and loss of active MRI lesions have been described (24). These improvements persist despite return of a normal immune repertoire.
Since the 2005 analysis, additional patients have been registered in the EBMT/ EULAR data base (status update March 2009, courtesy of D Farge): total n=1,031 consisting of MS n=379, SSc n=207, SLE n=92, RA n=88, JIA n=70, ITP n=23, CD n=23. A recent analysis of the EBMT / EULAR data base (D Farge, personal communication) suggests a reduction in TRM attributable to more precise patient selection. However, some long-term follow-up data are missing, and one must assume that HCT will always be associated with some degree of TRM. The EBMT, EULAR and the other learned societies are committed to completing these prospective RCTs and, if positive, developing more focused future protocols including mechanistic side studies to exploit the effectiveness and reduce the toxicity of HCT in autoimmune disorders. In the meantime, members of the Autoimmune Disease Working Party of the EBMT “consider it non-contributory to transplant patients with autoimmune disease outside the context of an approved, prospective RCT”.
Initial published experience focused on autologous HCT for three autoimmune disorders: severe diffuse scleroderma with internal organ involvement (SSc), SLE, and MS. Among 34 patients with life-threatening SSc followed up to 8 years after autologous transplant, stabilization/improvement of pulmonary disease and significant improvement in dermal sclerosis and functioning were observed (25). Figure 2 depicts these transplant outcomes. This experience formed the basis of a phase III randomized comparison of intensive immunosuppression with 12 monthly pulses of iv CY (750mg/m2) vs. immunoablation followed by CD34+ selected autologous HCT in the SCOT (Scleroderma: Cyclophosphamide Or Transplantation) protocol (www.sclerodermatrial.org). Encouraging results have also been reported in 50 patients with SLE followed to 7.5 years after autografting (26). A subsequent phase III LIST (Lupus Immunosuppressive/immunomodulatory therapy or Stem cell Transplant) trial will be discussed below. In addition, two phase II US trials in MS with 4 year follow-up have been reported with apparent reduction in MS episodes (27, 28).
Table 2 lists currently open HCT trials for adults with autoimmune disorders as registered with ClinicalTrials.gov. Six major diseases each with 2–6 accruing studies are presented. Of these 26 trials, 15 are autologous and 11 allogeneic transplant studies. Only five of the 26 are randomized trials and only four are NIH supported.
For patients with severe SSc, co-morbid organ impairment is common; for those with MS, timing of HCT is an issue; and for individuals with SLE, a host of conventional treatments are competing options. With experience gained in autologous HCT for autoimmune disease, and with considerable allograft experience with reduced intensity regimens for hematologic malignancies, allogeneic transplant regimens are being investigated for autoimmune disorders (29). Collaboration across disciplines promotes recruitment of patients onto HCT trials despite restricted funding by insurers which remains a significant issue, as predicted at the beginning of this clinical research (19). What was less clear a decade ago was the relentless growth in regulatory steps in protocol development. For example, for 16 trials of the Eastern Cooperative Oncology Group, some 481 distinct processes were required to activate a protocol consuming a median of 808 days effort from study concept to enrollment (30). This delay is not unique to oncology. The LIST study was an NIH supported randomized transplant trial in SLE (31). It was never activated due to delays in protocol development.
Despite formation of cooperative and community research groups, annual enrollment onto oncology trials in the US remains stuck at 3% of all newly diagnosed adults with cancer (32). Factors cited by physicians as road blocks to patient recruitment include extra uncompensated time required to enroll subjects, regulatory paperwork and insurance overload. It has recently been estimated that $31 billion represents the collective cost to US physician practices for time spent on interactions with insurers for approval of non-protocol tests and treatments (33). As the authors of the study note, this cost to interact with health plans is equal to 6.9% of all US expenditures for physician and clinical services. Health insurers may further restrict protocol recruitment by denial or delay of coverage for treatment on a study. Because of uncertainty about insurer reimbursement for clinical trials, 26 states have enacted legislation requiring all third party payers to cover participation in cancer clinical trials (21 states) or in trials for life-threatening diseases (5 states). Laws require that trials be approved by the NIH, NIH Cooperative Group or Center, FDA, or the Departments of Defense or Veterans Affairs (www.cancer.gov/clinialtrials/ctlaws-home). Unfortunately, only five of the 50 states mandate insurers to support clinical trials for life-threatening, non-malignant diseases.
Although adults in America are either very willing (32% of those surveyed) or inclined (38%) to participate in a clinical trial, most do not (34). Fewer than 1% of the US population enrolls each year in approximately 80,000 open clinical trials. When surveyed, major barriers to enrollment were not patient attitude; rather, the unavailability of an appropriate trial, disqualification due to co-morbidities and concerns about transportation or insurance coverage were the major impediments (34).
Unique among the wealthy nations, the United States has tens of millions of citizens without any health care coverage from private insurer or third party agency. The national debate on healthcare for the uninsured continues, but it remains true that HCT treatment trials are not available for the uninsured. Less obviously, clinical trials in HCT may also be unavailable for the insured. Health insurers make determinations for coverage which vary widely across and within plans. In a study by Peters and Rogers, 533 women with stages II-IV breast cancer were tracked for decisions on treatment coverage for high-dose chemotherapy and autologous HCT (35). Requests for coverage were denied in 121 (23%) of the women. In an accompanying editorial in 1994, a call was made to bring fairness, rationality and public accountability to this reimbursement process (36).
Fifteen years on, problems not only persist but, arguably, are worse. Data from the first 95 patients with severe SSc submitted for health insurance coverage for the SCOT trial found that 51 (54%) of subjects were initially denied coverage even though both treatments had been published in top-tier, peer-reviewed journals and found to have promising results (25, 37). Nevertheless, insurers judged the trial as “experimental or investigation” and denied coverage. Coverage decisions varied widely both across and within plans and funding decisions within the same company were inconsistent or persistent in denial despite repeated prior reversals of denial on outside independent review. As stated in a recent Technology Assessment report prepared for the Agency for Healthcare Research and Quality (AHRQ), although published data are near non-existent to quantify the magnitude of the effect of third party insurance denials on recruitment into clinical trials, insurance policies do restrict recruitment onto NIH supported clinical studies and thus impede clinical research and the evidence needed to advance healthcare of the nation (38).
The irony is that as protocol development and recruitment become more difficult and more laden with processes, the opportunity for mechanistic and genomic studies generated within clinical trials flourishes. For allogeneic HCT, studies of non-HLA encoded genes and their influence on outcomes is just beginning. Genes controlling drug metabolism and immune response are being discovered and will predict greater individualization of treatment as biobanks of relevant materials are being established world-wide (39). Among subjects funded by insurance and randomized on the SCOT trial, over 4,000 samples to date have been stored at baseline and 10 time points after randomization. Stored materials include serum, plasma, cells, DNA and RNA. Requests for proposal for scientific studies using repository specimens were recently sent to over 5,100 rheumatologists and 180 medical school directors of HCT programs and basic science departments in North America. In addition to the biobank, the NIH has funded 9 additional mechanistic research studies from materials from consenting subjects enrolled in SCOT to determine the molecular mechanisms of SSc, immune regulation and responses to treatment.
Considerable advances have been gained in our understanding of the immunobiology of autoimmune disease and HCT. Transient numeric depletion of immune cells does not explain the prolonged remissions observed after autologous HCT in patients with autoimmune disorders. The sustained clinical effects are better explained by qualitative change in the reconstituted immune repertoire. Autologous HCT induces substantial post-transplant modifications in the adaptive immune system. It remains to be determined whether and to what extent the suppression of inflammation observed post-therapy depends on the eradication of disease-associated T and/or B cell populations or on their regulation of function.
Pilot clinical trials of HCT for autoimmune diseases have paved the way for controlled RCTs which are underway in Europe and the United States. These studies illustrate the challenges of conducting clinical treatment trials in an intense regulatory environment found on both continents. While impediments from the US health insurance system can restrict clinical trial entry, there is vibrant opportunity and NIH support of basic science investigations within pivotal clinical trials.
Financial disclosure: Supported in part by NIH award AI-05419. The authors otherwise have nothing to disclose.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.