Key points

Introduction

The first hematopoietic stem cell transplants (HSCT) for patients with primary immunodeficiencies (PIDS), specifically severe combined immunodeficiency (SCID) and Wiskott-Aldrich syndrome (WAS), were performed in 1968. Since then, HSCT has been regularly used to treat a growing spectrum of immunodeficiencies. A recent Primary Immune Deficiency Treatment Consortium (PIDTC) survey reported more than 1000 patients transplanted in North America alone for SCID, WAS, or chronic granulomatous disease (CGD), which are the three most common types of PID.

Further, more than 200 distinct immunodeficiencies have now been described and at least 10 new genetic forms of PIDS are described each year: many of these new diseases have successfully been treated by HSCT. Thus transplantation is a major treatment option for patients with genetic mutations in cells of the hematopoietic system that lead to an impaired immune response. However, the rarity of these disorders and the unique aspects intrinsic to each disease make it difficult to define a universal transplant regimen. This article discusses those PIDs most commonly treated by HSCT and disease-specific issues regarding transplantation for these patients.

Introduction

The first hematopoietic stem cell transplants (HSCT) for patients with primary immunodeficiencies (PIDS), specifically severe combined immunodeficiency (SCID) and Wiskott-Aldrich syndrome (WAS), were performed in 1968. Since then, HSCT has been regularly used to treat a growing spectrum of immunodeficiencies. A recent Primary Immune Deficiency Treatment Consortium (PIDTC) survey reported more than 1000 patients transplanted in North America alone for SCID, WAS, or chronic granulomatous disease (CGD), which are the three most common types of PID.

Further, more than 200 distinct immunodeficiencies have now been described and at least 10 new genetic forms of PIDS are described each year: many of these new diseases have successfully been treated by HSCT. Thus transplantation is a major treatment option for patients with genetic mutations in cells of the hematopoietic system that lead to an impaired immune response. However, the rarity of these disorders and the unique aspects intrinsic to each disease make it difficult to define a universal transplant regimen. This article discusses those PIDs most commonly treated by HSCT and disease-specific issues regarding transplantation for these patients.

Severe Combined Immunodeficiencies and Related Disorders

SCID comprises several disorders mainly characterized by a lack of functional T cells caused by (1) a failure to produce cytokines necessary for T-cell maturation (interleukin [IL]-2 receptor common gamma chain, Jak3, and IL7RA deficiencies) (2) accumulation of toxic purine metabolites (Adenosine Deaminase [ADA]-SCID), (3) defective differentiation of a neutrophil/T-cell precursor (reticular dysgenesis), or (4) defective VDJ rearrangements (Rag2, Rag2, Artemis). SCID disorders can also be characterized by the presence or absence of natural killer (NK) or B cells.

Patients with these disorders usually do not survive beyond the first year of life. The only chance of long-term survival is immunologic reconstitution by normal T cells. The SCID genotype has significant implications on the need for conditioning and transplant outcome. Although immunosuppression is typically not required for these patients because the absence of T lymphocytes prevents graft rejection, there are some SCID variants in which immunosuppressive agents may be required to facilitate long-term immunoreconstitution. Preexisting infections, particularly viral infections, such as cytomegalovirus (CMV) or adenovirus, may limit the ability to use myeloablative chemotherapy conditioning, although reduced intensity conditioning (RIC) may be tolerated. In addition, maternal engraftment of leukocytes during pregnancy is common in some SCID genotypes because there are no autologous T cells to reject transplacental maternal T cells. Maternal T lymphocytes do not provide effective immunity, but may cause maternofetal graft-versus-host disease (GvHD) manifesting as an eczematous rash, splenomegaly, and occasionally pneumonitis or other complications. There is also the potential complication of a graft-versus-graft (GvG) reaction depending on the donor used. Infusion of unfractionated human leukocyte antigen (HLA)–identical sibling marrow can lead to rejection of the maternal T lymphocytes, with or without a clinical GvG reaction. Using a maternal donor can result in rapid immune reconstitution with or without GvHD that is usually amenable to treatment with low-dose steroids. In contrast, rejection of a paternal graft can occur either with or without a clinical GvG reaction.

The best donor is an HLA-matched sibling donor, with a 90% overall survival rate in the most favorable circumstances. When a matched sibling or family donor is unavailable, an alternative donor is necessary. Given the lack of autologous T cells and the ready availability of a parental donor, haploidentical transplants are a common choice. Bone marrow grafts depleted of T cells were studied prospectively resulting in an 81% overall survival with 77 patients having received a haploidentical graft and 12 using an HLA-matched related donor. All grafts were infused without preconditioning (except for 2 receiving placental blood) or posttransplant GvHD prophylaxis and most deaths occurred within 12 months following transplantation, most from preexisting infection. Note that 68 of the 72 had normal T-cell function but 45 of the 72 survivors required intravenous immunoglobulin replacement. Although this method means that patients without an HLA-identical sibling donor can be offered transplantation and cure, there is a significant, albeit diminishing, mortality risk associated with T lymphocyte–depleted transplants.

The results from HSCT are superior when the procedure is performed early, before end-organ damage occurs from infection. Twenty-one infants diagnosed early because of a positive family history were transplanted within 28 days of birth, 16 of whom had IL-2 receptor gamma chain–deficient or JAK3-deficient SCID, and all received T lymphocyte–depleted stem cells, without any conditioning or GvHD prophylaxis. Twenty patients (95%) survived, with a maximum follow-up of 19 years. A 2-center study from the United Kingdom similarly compared outcomes of 48 probands and 60 older affected siblings with an overall survival in the early transplants of 90% versus 60% in the older siblings. A recent report from the PIDTC on 240 children with SCID transplanted between 2000 and 2009 showed that survival in older patients was as good as in those transplanted at less than 3 months, as long as they were infection free or previous infection had cleared. Initiatives to mandate testing for SCID at birth are therefore underway in many US states and Canadian provinces.

Attempts to perform in utero transplantation have been made in patients with a positive family history of SCID. However, confirming the diagnosis requires invasive procedures, which can lead to fetal loss, and T lymphocytopenia with diminished mitogen response has been reported after some procedures, leaving the patient at risk of opportunistic infection. Furthermore, occult maternofetal T-lymphocyte engraftment may lead to graft rejection, and obtaining maternal stem cells to treat the patient is not feasible during pregnancy. GvHD is also undetectable in utero, and treatment via the mother and monitoring of the fetal responses are either not possible or impractical. Thus, for those patients with a positive family history, the preferred option is to confirm the diagnosis, initiate prophylactic antimicrobial treatment at birth, and search for a donor to perform a transplant as soon as possible.

More recently with high sequencing analysis of HLA, the results of matched unrelated donor transplantation using adult donors or umbilical cord blood (UCB) is gaining support as a viable alternative. HSCT using HLA-matched unrelated donors provides similar results to those of HLA-matched siblings. UCB products that are already banked are more readily available, allowing early transplantation. Fernandes and colleagues compared the results of UCB transplant with mismatched related donor transplant in a retrospective study involving 249 transplants (74 UCB vs 174 mismatched related donor). Most UCB transplants were done with a myeloablative conditioning regimen and recipients had a higher frequency of complete donor chimerism and faster lymphocyte count recovery, but there was a trend toward more severe acute GvHD and more chronic GvHD. The 5-year survival rates were similar at 62% for mismatched related donor transplant versus 57% for UCB.

In addition, the role of conditioning remains undetermined. Although comprehensive studies are lacking, certain tenets apply and considerations for pretransplant conditioning include the presence of infection or end-organ damage, the molecular diagnosis, the type of donor available, the likelihood of full immune reconstitution, and the risk of short-term and long-term side effects. In a multicenter study between the United States and Europe of 103 patients with SCID, infusion of stem cells from an unrelated donor restored T-lymphocyte immune reconstitution, although the risk of GvHD was significantly higher than when a matched related donor was used. A study of 77 patients with SCID in the United Kingdom, who received stem cell infusions, showed a 90% survival in matched sibling donor/matched family donor transplants but only 60% when alternative donors were used. Infants with NK ⁻ SCID were more likely to survive and had high-level donor T-cell chimerism with superior long-term recovery of CD4 T-cell immunity than NK ⁺ recipients. A third of patients with NK ⁺ SCID required additional transplant procedures.

In a series of 98 patients with SCID, 32% received pretransplant chemotherapy conditioning: the rate of engraftment was significantly higher in these patients than in nonconditioned patients, but overall survival was less, mainly because of infection-related deaths. More recently, European centers have described the outcome of transplantation for 699 patients with SCID, of whom 25% had a T-B- phenotype, and of whom 42% received conditioning. There was no survival advantage for the nonconditioned patients. A small single-center study of 24 patients, of whom at least half had T-B- SCID, showed more favorable outcomes for patients who received conditioning and, in a multicenter study of 178 patients with SCID, the molecular diagnosis significantly affected outcome. Patients with T-B- SCID had worse survival, with diminished rates of stem cell engraftment, and slower T and B lymphocyte immune reconstitution, but there a better overall rate of cure of disease in patients who had received conditioning. These findings were replicated in the most recent European report.

In NK-SCID (IL-2, Jak3, IL7RA, ADA-SCID), prethymic and early intrathymic stromal niches are vacant and thus available for donor T-lymphocyte precursor engraftment, leading to sustained donor-derived thymopoiesis in the absence of myelopoiesis. In contrast, in NK+ SCID caused by VDJ recombination defects, later stages of T-lymphocyte differentiation are blocked; therefore, without conditioning, high numbers of DN2/DN3 cells compete with donor T-lymphocyte progenitors for thymic niches, resulting in a restricted T-cell repertoire and reliance on postthymic T-lymphocyte expansion. Infusion of large doses of stem cells as well as other stromal factors aids engraftment of donor stem cells and improves immune reconstitution. Myeloid engraftment, most often achieved following conditioning, facilitates engraftment of donor stem cells, permitting long-term thymopoiesis, even in NK- SCID phenotypes.

Moreover, although restoration of T lymphocytes occurs following infusion of T-depleted stem cells, many patients with nonconditioned transplants fail to gain functional B lymphocytes, particularly patients with VDJ recombination defects. Recipient B lymphocytes have reduced numbers of memory and isotype-switched B lymphocytes, and poor antibody responses to antigen stimulation, compared with donor-derived B lymphocytes. In contrast, recipient B lymphocytes from patients with IL7RA-deficient SCID are able to function in the presence of donor T lymphocytes and produce antibody to appropriate antigen stimulation. Also, patients with IL-2 receptor gamma chain–deficient or JAK3-deficient SCID without conditioning can develop B lymphocyte function, likely because of donor B-lymphocyte microchimerism. In contrast, not all conditioned transplants achieve donor B lymphocyte engraftment and independence from immunoglobulin replacement.

Patients with ADA-deficient SCID differ from those with other molecular defects because accumulation of toxic metabolites causes common lymphoid precursor toxicity leading to abnormalities of T, B, and NK lymphocyte development and function. In addition, patients often have preexisting inflammatory pneumonitis caused by the metabolic defect. In a retrospective multicenter study of 106 infants with ADA-deficient SCID, outcomes were better for patients who had matched sibling and family donors than matched unrelated or T lymphocyte–depleted donors. Superior survival was seen in patients who received nonconditioned transplants rather than myeloablative procedures. When using haploidentical donors without conditioning, nonengraftment was a major problem, although donor B-lymphocyte engraftment was achieved even after nonconditioned transplants.

In addition, the use of chemotherapeutic agents, even when beneficial, has long-term implications. Patients with Artemis-deficient SCID experience significant long-term sequelae following the use of alkylating agents, including significant growth failure, dental problems, and autoimmunity; findings that are not associated with nonconditioned transplants. Impaired fertility is associated with many of the conditioning agents used, although detailed information from the SCID cohort is not available. Moreover, the long-term risk of malignancy may be increased, at least in some patients who receive irradiation as part of the preconditioning regimen, but the only data are anecdotal. Therefore, more studies are needed to delineate specific long-term immunologic and general health outcomes in large cohorts of patients with the same molecular diagnoses, conditioning regimens, and donors. In addition, more data are needed on the use of nonmyeloablative regimens or RIC so that optimum treatment can be offered to appropriate patient groups.

Non–severe combined immunodeficiencies

In general, patients with non-SCID disorders have at least some residual T lymphocyte–mediated immunity, thus requiring the use of at least some immunosuppression and/or conditioning to ensure engraftment of donor cells. In addition, there is often sufficient time to look for alternative nonfamily adult matched donors or cord blood stem cell units from national and international registries. Moreover, depending on the genetic defect, full donor chimerism may not be necessary to restore immune function. This aspect, combined with improvements in conditioning regimens, has resulted in the use of RIC with good outcomes. In addition, T lymphocyte–depleted haploidentical donors were not used in these conditions because of the risk of nonengraftment or rejection. These patients are often older with preexisting viral infections, and T-lymphocyte depletion prolongs the time to reconstitution, thus increasing the risk of death from disseminated viral infection. However, several groups have been investigating modified grafts depleting specific T-cell subsets, while keeping others to maintain engraftment and antiviral activity, or using posttransplant cyclophosphamide to reduce reactive T cells after the graft infusion.