Allogeneic hematopoietic stem cell transplant (AlloSCT) from related or unrelated histocompatible donors has been well established as potentially curative therapy for children and adults with selected hematologic malignancies 1. The concept of AlloSCT as a treatment option for hematologic malignancies has long been based on the assumption that myeloablative doses of cytotoxic therapy were required for both disease eradication and host immunosuppression. However, over the last decade there has been a paradigm shift towards curative potential of graft versus leukemia or tumor (GVL/GVT) effect 2. The impact of GVL effect is strengthened by observation that recipients of syngeneic SCT or T-cell-depleted SCT are at increased risk of relapse compared to unmanipulated AlloSCT, patients who develop acute or chronic graft-versus-host (aGVHD/cGVHD) disease have less risk of relapse 3–7; and also donor lymphocyte infusion (DLI) can induce remission in patients whose disease has relapsed after AlloSCT 8–10. The concept behind reduced intensity AlloSCT (RI-AlloSCT) is that instead of eradicating tumors through intensive/toxic chemoradiation, the SCT donor's immune cells might be used for tumor eradication, relying on allogeneic GVT effects. However, in non-malignant diseases with RI-AlloSCT, the aim is to create an immunologic platform of host and donor tolerance using pre- and post-transplantation immunosuppression. Hence, elimination of high-dose cytotoxic therapy would allow medically infirm patients to be treated with AlloSCT 2.
Non-myeloablative (NMA) regimens are different from reduced intensity conditioning (RIC) regimens. A truly NMA regimen should not eradicate host hematopoiesis and should allow relatively prompt hematopoietic recovery (<28 days) without a transplant 2. Upon engraftment, mixed chimerism should be present. If the graft is rejected, prompt autologous recovery should occur. However, RIC regimens require a transplant for hematologic recovery, and if the graft is rejected, prolonged aplasia typically occurs. This distinction in intensity of regimens is crucial to differentiate the GVL effects of donor engraftment from the anti-tumor effect of the conditioning regimen. Most commonly used immunosuppressive and RIC regimens are depicted in Figure 1.
Socie et al. 11 reported long-term survival and late effects after myeloablative allogeneic bone marrow transplantation (BMT) in 6,691 patients who were free of their original disease for at least two years after AlloSCT. A large number of patients died of other secondary complications including GVHD (31%), infection (6%), secondary malignancies (6%), and organ failure (6%). Children with non-malignant disease who require myeloablative therapy and AlloSCT for curative intent, fear the above-mentioned long-term late complications associated with this therapy, including growth failure, gonadal failure, secondary malignancies and secondary myelodysplastic syndrome (MDS). These potential long-term complications sometimes weigh heavily on their decision to proceed with curative-intent therapy especially in children with non-malignant diseases. It remains to be determined whether a select group of children receiving RI-AlloSCT will benefit from a reduced risk of disease reoccurrence and at the same time a reduced risk of long-term complications. Here, we review the recent experience of NMA/RI-AlloSCT in children and adolescents with both malignant and non-malignant diseases and discuss the challenges for the future.
RI-AlloSCT IN CHILDREN FOR NON-MALIGNANT DISEASES
Hereditary anemias caused by β-thalassemia and sickle cell disease (SCD) are the most common worldwide genetic diseases. Supportive therapies have significantly ameliorated clinical manifestations of these diseases but cannot eliminate disease and treatment-related complications may result in end-organ damage. Allogeneic hematopoietic SCT is the only cure for patients with hemoglobinopathies. Results of transplants have steadily improved over the last few decades due to effective control of transplant-related complications.
The historical experience of myeloablative AlloSCT in patients with SCD or β-thalassemia major indicates that stable mixed hematopoietic chimerism may be sufficient to cure patients with these hemoglobinopathies 12, 13. Thus, myeloablation per se may not be mandatory for the treatment of SCD, raising the possibility that NMA/RIC regimens could be investigated as an alternative method to reduce long-term complications in this subpopulation. Studies in a murine model of SCD and β-thalassemia suggest this approach could even be curative in the human leukocyte antigen (HLA)-disparate setting 14, 15. Given the preliminary results of reduced acute toxicity following NMA/RI-AlloSCT, this approach seems attractive for patients with SCD, particularly older patients and those with acquired organ toxicity and it may offer children the potential to retain gonadal function and normal fertility. There have been few published reports of NMA/RI-AlloSCT in patients with SCD and β-thalassemia. Krishnamurti et al. 16 has reported the success of matched sibling BMT in a child with homozygous SCD with a conditioning regimen of fludarabine, busulfan, anti-thymocyte globulin (ATG) and total lymphoid irradiation. The patient was reported to be disease free, without GVHD and with 100% donor engraftment at 14 months post-BMT. Iannone et al. 17 in a prospective multi-center study, reported an NMA SCT approach in seven pediatric patients with SCD and thalassemia who received pre-transplantation conditioning with fludarabine and 200 cGy total body irradiation (TBI) and ATG. Regimen-related toxicity was minimal but there was only transient donor engraftment in six of seven patients. These preliminary results suggest that more intensive conditioning than 2 Gy TBI and fludarabine is required for previously transfused patients with hemoglobinopathies (Table I). However, RIC regimens with intense immune suppression and myelosuppression are able to prevent GVHD and promote engraftment 18. Moreover, in some cases RI regimens were associated with an increased risk of regimen-related toxicity and severe GVHD 19. Incidence of infertility following RI-AlloSCT in patients with hemoglobinopathies is yet to be determined.
Table I. Reduced Intensity/Non-Myeloablative Allogeneic Stem Cell Transplant in Children and Adolescents With Non-Malignant Disease
FLU (150 mg/m2)/MEL (125–140 mg/m2) or BU (8 mg/kg)/±ATG or Campath
AlloSCT is curative for congenital immunodeficiencies and few studies with small numbers of patients with primary immunodeficiencies have demonstrated that the RI-AlloSCT approach permits rapid engraftment from sibling and unrelated donors with minimal toxicity, even in the presence of severe organ dysfunction, thus establishing host tolerance to the donor immune cells responsible for immune reconstitution (Table I) 20. However, some of the patients who are fully engrafted initially may develop mixed chimerism. Despite the mixed chimerism, children with high- and low-level chimerism can remain well and free of disease and patients who suffer graft failure can be successfully re-transplanted with myeloablative or RIC regimens 21. In comparison to myeloablative regimens patients receiving RIC regimens have significantly better overall survival (OS; Fig. 2; Table I) 21. Overall, the RIC regimens in patients with primary immunodeficiencies are well tolerated even in high-risk patients with pre-existing organ dysfunction.
AlloSCT is the treatment of choice for Fanconi anemia (FA) patients with severe hematologic manifestations. FA patients are very sensitive to conventional conditioning protocols due to high chromosome fragility. Chemotherapy, especially alkylating agents and ionizing radiation, has been associated with a high incidence of transplantation-related toxicity and mortality 22. Low dose cytoxan and limited field irradiation is currently the standard AlloSCT preparative regimen for FA patients with related donors 23–25. However, the impact of AlloSCT on late malignancies in FA patients has yet to be defined. Two small studies in patients with FA utilizing fludarabine based conditioning regimens were associated with minimal toxicity and very excellent overall (100%) and disease free survival (82–100%) 26, 27. In both of the above-mentioned studies, although performed in only a small group of patients, this approach seems promising and safe.
Children with multisystem Langerhans cell histiocytosis (LCH) and organ involvement who fail to respond to conventional chemotherapy have an extremely poor prognosis. Twenty-nine pediatric patients who underwent myeloablative AlloSCT for LCH with organ involvement had an OS of 48% and transplant-related mortality (TRM) of 45% 28. Steiner et al. 28 used a fludarabine based RIC regimen for nine pediatric patients with chemotherapy refractory, high-risk LCH. Seven out of the nine patients survived and demonstrated no signs of disease activity after median follow-up of 390 days post-SCT. Hemophagocytic lymphohistiocytosis (HLH) poses major diagnostic and therapeutic challenges. The primary autosomal recessive form in children is fatal without adequate treatment 29, 30. In 86 children that received HLH-94 therapy followed by myeloablative AlloSCT, the OS at 3 years was 64% and TRM was 30% 31. Cooper et al. 32 reported the results in 12 pediatric patients with primary HLH with significant coexisting morbidity that received RI-AlloSCT. Seventy-five percent of patients achieved complete remission (CR) after RI-AlloSCT and 25% of patients died due to TRM. In patients with HLH and pre-existing severe organ toxicity before transplantation, RI-AlloSCT might be the best approach.
RI-AlloSCT IN CHILDREN WITH MALIGNANT DISORDERS
There are multiple studies published on the use of RI-AlloSCT in adults with malignant diseases. However, there are only a few studies in pediatric patients with malignant diseases. Pediatric patients who undergo intensive chemotherapy are at risk of developing co-morbidities which preclude them from receiving standard myeloablative AlloSCT.
As we have previously reported, the preliminary results in 14 children who received fludarabine-based regimens followed by AlloSCT for malignant diseases showed a 1-year OS of 78% (Table II) 33. This study demonstrated that a high degree of sustained donor chimerism and myeloid engraftment can be achieved in children with malignant diseases following RI conditioning therapy with either a matched family donor or unrelated cord blood donor. In an another study, Gomez-Almaguer et al. 34 reported their experience with RI-AlloSCT in 16 children and adolescents with malignant diseases (Table II). Two-year survival of patients with malignant diseases was 44%. Overall, this regimen was very well tolerated and cost-effective as half of the patients received transplant on an outpatient basis. However, these studies included a heterogeneous small group of patients with different donor sources and different RI conditioning regimens. Since a variety of RIC regimens were used, attribution of regimen to transplant outcome is difficult to ascertain. The results of these studies should be interpreted with caution until larger groups of children with a homogenous diagnosis and disease status, uniform cell source, and RIC regimens are utilized.
Table II. Reduced Intensity Allogeneic Stem Cell Transplant in Children and Adolescents With Malignant Disease
There are only two malignant disease specific studies published. In the first study, 11 children with high-risk acute lymphoblastic leukemia received fludarabine and busulfan based RI-AlloSCT (Table II) 35. As noted in a previous study, a high degree of donor chimerism was achieved in this study also 33. In this very high-risk group of patients who were ineligible for standard intensity conditioning, 36% of patients achieved remission following RI-AlloSCT, however, TRM was very high (50%) 35. In the second study, eight pediatric patients with CD33+ AML in CR1 or CR2 received targeted immunotherapy with gemtuzumab ozogamicin after AlloSCT during the time of potential minimal residual disease following RIC 36. RIC regimens were well tolerated without any significant morbidity or mortality and OS (63%) was comparable with the standard intensity regimen.
ACUTE AND CHRONIC GVHD FOLLOWING RI-AlloSCT
The fundamental differences between RI and myeloablative allogeneic conditioning regimens may have impact on the incidence or severity of GVHD. These differences are first, RI-AlloSCT conditioning appears to cause only limited tissue damage; second, development of transient mixed donor-host chimerism may facilitate establishment of mutual tolerance, which, in turn, may down-regulate GVH activity; third, the type and duration of immunosuppressive agents administered before and after RI-AlloSCT conditioning may sufficiently differ from those used after myeloablative AlloSCT; and fourth, the number and function of recipient antigen presenting cells may be higher after RI-AlloSCT as compared to myeloablative AlloSCT 37. These cells may play a major role in the initiation of GVH responses early after SCT 38. In order to facilitate engraftment and decrease the incidence of GVHD, intense immunosuppression is required for a longer length of time which results in delayed immune reconstitution. This might increase the incidence of viral infection and also aGVHD may occur later than the usual first 100 days. Due to lack of large and/or randomized studies, incidence of aGVHD after RI-AlloSCT in pediatric patients is yet to be defined. However, in various small studies the incidence of aGVHD seems to be lower than myeloablative AlloSCT (Tables I and II). Because the length of follow-up is relatively short, it is difficult to accurately assess the incidence and severity of cGVHD after RI-AlloSCT. However, adult studies from many centers have reported that the incidence of extensive cGVHD is at least comparable or higher for patients who undergo RIC regimens as compared to conventional myeloablative allografting 39.
HEMATOLOGIC RECONSTITUTION AND ENGRAFTMENT FOLLOWING RI-AlloSCT
The median time for myeloid and platelet recovery after RI-AlloSCT might be shorter than myeloablative regimens. However, many patients do not drop to an absolute neutrophil count <0.5 × 109 L and platelet count below 20,000 × 109/L. Assessment of engraftment might be difficult based on peripheral blood counts. Donor chimerism should be performed as early as 2 weeks following RI-AlloSCT and serial monitoring of chimerism (every 2–4 weeks) can be helpful to determine the need for immune manipulation to promote engraftment. Achievement of high donor chimerism (>95%) may take up to 3 months. Since there is no established definition of graft failure following RI-AlloSCT, various centers perform DLI based on their clinical experience. Various graft- and host-related factors contribute to the engraftment. The graft-related factors include histocompatibility, stem cell source, and stem cell dose, T-cell dose and host-related factors includes pre- and post-transplant immune-suppression, conditioning regimen, early post-transplant infections, disease (malignant or non-malignant), prior history of chemotherapy, and prior history of allosensitization secondary to blood transfusions.
The incidence of primary graft failure following RI-AlloSCT can be as high as 20–25% 40. Patients receiving RI-AlloSCT using HLA-matched unrelated adult donors, unrelated cord blood transplantation or heavily pre-transfused children with non-malignant diseases are at higher risk of graft rejection. It remains to be determined what degree of intensity is required for different pediatric subpopulations. Patients with primary refractory hematologic disease, including hemoglobinopathies, MDS, HLH, and/or severe aplastic anemia, may require more intense conditioning and immunosupression, particularly for heavily pre-transfused patients. It is well established that cell dose is a critical determinant of hematopoietic recovery and survival after single unrelated donor cord blood transplant. A recent study from the University of Minnesota demonstrated that co-infusion of two umbilical cord blood units appears safe and may improve upon the rate of engraftment anticipated after transplantation with an available single cord blood unit 41. Infusion of two cord blood units after RIC may increase the chances of engraftment.
INFECTIOUS COMPLICATIONS FOLLOWING RI-AlloSCT
The potential mechanism(s) of reduced infectious morbidity following RI-AlloSCT includes decreased duration of severe neutropenia, reduced grade of mucositis, enhanced immune reconstitution, and/or decreased rates of severe aGVHD 42, 43. However, due to intense immunosuppression, patients receiving RIC regimens are at higher risk for viral and fungal infections. Data on infections in RI-AlloSCT recipients are preliminary and differences in the degrees of myeloablation may lead to considerable variation in outcomes. Junghanss et al. 44 reported that risk of bacterial infection during the first 100 days was significantly reduced in RI-AlloSCT recipients. Identified risk factors for developing viral diseases post-RIC include lymphopenia, CMV serological status of the donor and recipients and use of alemtuzumab for immunosuppression 45, 46. Pediatric patients with primary immunodeficiency and HLH who received RI-AlloSCT had a significantly higher incidence of not only CMV but also EBV and adenovirus 21, 32. Nevertheless, with rigorous monitoring by PCR based assays and availability of new anti-viral drugs most of these patients can be successfully treated 21, 32. The result of infectious disease reported from small studies should be verified in multiple larger studies, especially considering the substantial variation between RIC regimens.
POTENTIAL OF ADOPTIVE CELLULAR IMMUNOTHERAPY FOLLOWING RI-AlloSCT
The chimeric state after RI-AlloSCT provides an ideal platform for adoptive cellular immunotherapy. A variety of cells can be used for adoptive immunotherapy, including plain lymphocytes, selected T cells, T-cell clones, and natural killer (NK) cells. DLI can be given after RI-AlloSCT for conversion from mixed to full donor chimerism, as pre-emptive therapy to prevent relapse, or for the treatment of relapse of hematologic malignancies and post-transplant lymphoproliferative disorder using donor-derived cells. DLI is limited by the development of aGVHD and cGVHD in up to 60% of the patients, which can be associated with significant morbidity and mortality. Several groups have investigated the preparation and infusion of purified, T-cell-depleted, donor NK lymphocytes with the aim to consolidate engraftment and to induce GVL effects in patients after hematopoietic SCT from haploidentical or other donors. However, clinical data on efficacy of NK-DLI is very limited 47–50.
CRITIQUE AND FUTURE CONSIDERATIONS
Multiple studies have been published demonstrating benefits of RI-AlloSCT in the adult population. However, most of the studies are non-randomized and meaningful comparisons cannot be performed due to multiple factors. The RI-AlloSCT experience in children and adolescents is scant. Most of the studies are single institution experience with small heterogeneous group of patients with variable conditioning regimens and immunosuppressive therapy and short follow-up. The major impetus for performing RI-AlloSCT in children is due to the fact that following myeloablative SCT children are predisposed to growth failure, gonadal failure, secondary malignancies, and secondary MDS. There is a need for well-planned and well-executed follow-up studies to evaluate the effect of RIC on growth, fertility, cGVHD, and secondary malignancies.
Regardless of technical approaches to RI-AlloSCT, there is a need to perform controlled clinical trials in children and adolescents with well-defined hematologic and non-hematologic malignancies to evaluate the role of RI-AlloSCT more fully. Most appropriate would be in patients with hematologic or non-hematologic malignancies who have failed initial therapies, who have a poor prognosis, and for whom salvage therapies exist that might be used for comparisons against RI-AlloSCT. Curative treatment modalities work best when used early in the disease course and this strategy likely will also apply to RI-AlloSCT. This consideration must be balanced by the fact that treatment-related mortality with RI-AlloSCT is significant and exceeds that of other frequently used therapies. One might anticipate that technical refinements and use in earlier stage patients will enhance the safety of RI-AlloSCT and thus increase acceptance of this modality for future clinical trials.
This study was supported in part by the Pediatric Cancer Research Foundation, Swim Across America Foundation, National Institute of Arthritis and Musculoskeletal and Skin diseases (R21AR49330; MSC), Marisa Fund, Sonia Scaramella Fund, Brittany Barron Fund, and Bevanmar Foundation.