Stem cells in the treatment of inflammatory arthritis

Autologous haematopoietic stem cell transplantation in patients with rheumatoid arthritis (RA) resulted in a positive short-term outcome clinically with low treatment-related toxicity. However, early conditioning regimens were of low immunoablative intensity and most patients relapsed. Mechanistic studies suggest that residual lesional effector cells may have been responsible for the relapses. The introduction of biopharmaceuticals has, for the moment, reduced the need for further experimental studies.

Juvenile idiopathic arthritis patients, mostly of the systemic subgroup, have shown nearly 33% durable drug-free remission, but with significant toxicity, including fatal macrophage-activation syndrome early in the programme. Later modifications to the protocol have reduced this toxicity.

Mesenchymal stem cells (MSCs), derived from several sources including bone marrow and adipose tissue, are being tested as tissue-regenerative and immunomodulating agents in many autoimmune diseases and animal models of inflammatory arthritis have been positive. MSCs and other stromal cells derived from actively inflamed synovium and peripheral blood of RA patients do not always demonstrate a full range of differentiation potential compared with healthy MSCs, although their immunomodulalatory capacity is unimpaired.

Despite the implementation of new treatment strategies and introduction of biologicals, long-term drug-free remission remains an elusive goal in most patients with chronic inflammatory arthritis, notably those with rheumatoid arthritis (RA) and juvenile idiopathic arthritis (JIA). A recent systematic review showed that the efficacy of a second biological agent, irrespective of the mode of action, after a first biological in RA patients is limited with ACR70 (ACR70, American College of Rheumatology criteria 70) responses ranging from 5% to 15% and disease activity score (DAS) remission from 9% to 15.4% . These figures underscore the need for new treatment modalities. Although data on other chronic idiopathic inflammatory arthritides are less robust, it is generally accepted that JIA and psoriatic arthritis (PsA) also constitute an area of huge unmet need , which explains why there is scope for new (including cellular) therapies. The rationale of cellular therapies for chronic inflammatory arthritis (and other rheumatic autoimmune diseases) is based on the concept that immune dysregulation can be restored by ex vivo expansion and reinfusion of cells with immunoregulatory function or therapeutic ablation of autoaggressive lymphocytes allowing subsequent preferential in vivo homeostatic expansion of such cells. Technological advances in cell processing and improvements in the medical care of complex patients have facilitated studies in this field. This article focusses on haematopoietic stem cell transplantation (HSCT) and mesenchymal stromal cell (MSC) therapy as examples of different types of cellular therapies that are both aimed at tilting the balance towards improved immune regulation while being fundamentally different in approach.

Haematopoietic stem cell transplantation

HSCT is the short name of a complex therapeutic intervention, comprising mobilisation of haematopoietic progenitor cells using high-dose chemotherapy and granulocyte-colony stimulating factor (G-CSF) or harvest of bone marrow, followed by intensification with myelo- or lymphoablative doses of chemotherapy and/or lymphocyte-depleting antibodies and/or total body irradiation (TBI), and (re)infusion of the graft to reduce the duration of aplasia. RA and JIA were among the first diseases to be targeted with HSCT in the 1990s based on long-term remissions observed in RA patients who received an allogeneic bone marrow transplantation (BMT) for haematological conditions and in studies in experimental animal models of autoimmune disease. The latter showed a dose–effect relationship of TBI on arthritis, providing evidence that deletion of autoaggressive T lymphocytes was necessary for maximum effect . More recent studies, however, showed that less intensive conditioning in combination with major histocompatibility complex (MHC)-mismatched BMT was equally effective both in terms of disease suppression and elimination of autoreactivity . Nevertheless, the risks of treatment-related mortality and graft-versus-host disease (GvHD) related to myeloablative conditioning and allotransplanting were felt to be high to justify treatment of patients with a chronic disease; hence, autologous transplants were pursued after preclinical studies in experimental arthritis demonstrated their curative potential. It is thought that the conditioning regimen is the key component in the immunoablation, but the potential immunomodulatory role of the autologous graft has never been properly investigated.

Rheumatoid arthritis

Early dose-finding studies showed that higher doses of intravenous cyclophosphamide for mobilisation of HSCs led to superior disease control when compared with lower doses . Subsequent pilot studies in the UK and the Netherlands confirmed the remission-induction potential of high-dose cyclophosphamide, but relapses were universally seen in most patients within a follow-up period of 3 years . The temporary improvements of disease activity translated into significant improvement of quality of life and arrest of joint destruction . HSCT was well tolerated by most patients, and no unexpected adverse events were seen. More intensive conditioning with the use of anti-thymocyte globulin (ATG) and/or CD34-selection of autologous grafts did not seem to be more effective, although the total numbers of RA patients treated was low and controlled studies were not done .

The need for HSCT waned after the introduction of effective biopharmaceuticals and implementation of new treatment strategies aimed at remission induction in early disease. The transplant activity in RA has now almost come to a standstill according to a recently published analysis of the European Group for Bone Marrow Transplantation (EBMT) data registry, which includes 89 RA cases, but only a few cases since 2006 . All of these patients had active, destructive, disease-modifying anti-rheumatic drug (DMARD)-refractory disease with a median disease duration of 86 months. The registry analysis confirmed the data from the pilot studies and an earlier registry analysis on 70 RA patients, showing that HSCT was relatively well tolerated in RA patients with 100-day transplant-related mortality (TRM) of 1% and overall survival of 94% (95% confidence interval (CI): 87–100%), probably because vital organ function was preserved. The high incidence of relapses in the RA population was reflected in a relatively low 5-year progression-free survival of 18% (95% CI: 9–27%; Fig. 1 ). Of note, most RA patients had been treated with low-intensity conditioning, comprising high-dose cyclophosphamide (200 mg kg ⁻¹) and/or CD34 ⁺-selected peripheral blood grafts. It must be emphasised that the level of detail of clinical information – for example, on clinical responses and toxicity – in registry analyses is limited, and so robust conclusions cannot be drawn. In an Australian randomised, controlled trial, RA patients treated with high-dose cyclophosphamide (200 mg kg ⁻¹) were treated with either an unmanipulated graft or a CD34 ⁺-selected graft . This study failed to show a benefit of CD34-selection, although it was argued that this might be due to the absence of in vivo T-cell depletion. In a subsequent study, the authors showed that relapses could be treated effectively with rituximab . This was in keeping with similar observations in the European studies showing that responsiveness to methotrexate and cyclosporine were restored post transplant.