The Evidence for Allogeneic Hematopoietic Stem Cell Transplantation for Congenital Neutrophil Disorders: A Comprehensive Review by the Inborn Errors Working Party Group of the EBMT

Congenital disorders of the immune system affecting maturation and/or function of phagocytic leucocytes can result in severe infectious and inflammatory complications with high mortality and morbidity. Further complications include progression to MDS/AML in some cases. Allogeneic stem cell transplantation is the only curative treatment for most patients with these diseases. In this review, we provide a detailed update on indications and outcomes of alloHSCT for congenital neutrophil disorders, based on data from the available literature.


INTRODUCTION
Congenital neutrophil disorders as a category of primary immunodeficiency (PID) can be classified in many ways, but a key point of distinction is whether the disorder is quantitative, or qualitative (1). The 2017 International Union of Immunological Societies (IUIS) Phenotypic Classification for Primary Immunodeficiencies divides neutrophil disorders into four broad categories: congenital neutropenia associated with or without syndromic disease, and functional neutrophil defects with or without syndromic disease (2). Affected patients can present with variable symptoms including recurrent infections, failure to thrive, and overwhelming septic episodes leading to high morbidity and mortality. Early and severe respiratory infections (e.g., Burkholderia cepacia, Aspergillus spp.), visceral abscesses, cellulitis, lymphadenitis, and granulomatous lesions are observed in patients suffering from CGD (3,4). Some patients develop severe autoinflammatory complications underlining the role of neutrophils in autoinflammatory processes beyond microbial defense (5,6). In many of these diseases there is a recognized risk of progression to myelodysplastic syndrome (MDS) and acute myeloid leukemia (AML) (1,2,7,8). Treatment for neutrophil disorders classically comprises anti-microbial therapy, granulocyte-colony stimulating factor (G-CSF), and allogeneic hematopoietic stem cell transplantation (alloHSCT) (8).
In this review we provide an update on the evidence, indications and modalities of alloHSCT for the various congenital neutrophil disorders based on data from the available literature, excluding CGD which will be discussed elsewhere in this special edition. Also beyond the scope of this manuscript is neutropenia that features in predominantly lymphocyte immune deficiencies (e.g., some forms of severe combined immune deficiency, CD40L deficiency, etc.).

MATERIALS AND METHODS
We used the 2017 IUIS Phenotypic Classification for Primary Immunodeficiencies as the basis for this review (2). Data were gathered via an English-language Pubmed literature search whereby the name of each disorder in figure five of the IUIS classification was searched individually, and together with the terms "alloHSCT" and "transplantation." We searched for case reports and case series on each disorder, focusing on the question of treatability of the disease by alloHSCT.

Group 1: Syndrome-Associated Neutropenia
This group of diseases is dominated by conditions defined by bone marrow failure (BMF), predisposition to myelodysplastic syndrome (MDS) and acute myeloid leukemia (AML), as well as additional non-hematological manifestations such as neurodevelopmental delay (Table 1). Disease control can be achieved for some patients with G-CSF in combination with antibiotic prophylaxis (40,41). However, the refractoriness of the disease and the predisposition to myeloid malignancies raise the question of if and when curative treatment by alloHSCT is indicated. A lack of evidence regarding longterm outcomes both with and without transplantation, the lack of genotype-phenotype correlation, and the persistence of neurodevelopmental disorders, make recommendations difficult for many of these diseases ( Table 1). In some cases only case reports are available, such as in the case of successful transplant of Glycogen Storage Disease type 1b using reduced intensity conditioning (18,42). On the contrary, more data is available on Schwachman Diamond Syndrome and deficiency of VPS45 protein, and is presented here (43).

Schwachman Diamond Syndrome (SDS)
Schwachman Diamond Syndrome is caused by mutations in the SBDS gene at chromosome 7q11.21. Mutations in this gene lead to impaired RNA metabolism and ribosomal function, with clinical features as described in Table 1 (10). One third of patients develop major hematological complications (9), and alloHSCT is indicated in cases of bone marrow failure (with subsequent transfusion-dependent anemia, bleeding and severe, recurrent infections), MDS, or AML (11, 12,44,45). Numerous studies have shown that alloHSCT is able to correct hematological abnormalities in patients with SDS, however most studies are limited by small sample numbers and/or short follow-up time. A 2005 study by Cesaro et al. (44) reviewed 26 patients with SDS who underwent alloHSCT and demonstrated overall survival of 64.5% at 1 year follow-up. Fifty-six percent of patients who were transplanted following the development of MDS/AML died posttransplant, compared to 19% who were transplanted for BMF, although the difference was not statistically significant. Patients transplanted with total body irradiation (TBI) appeared to do worse than those transplanted with busulfan and fludarabine conditioning, with 67 vs. 20% mortality (P = 0.03) (44). In 2002 Hsu reviewed 15 cases of SDS alloHSCT patients and found that transplantation following the development of MDS/AML was associated with a worse outcome; overall survival in this series was 40% (12).
A 2005 review of French registry data found 10 patients who had been transplanted for SDS, five because of bone marrow failure (BMF) and five following development of MDS/AML. They found a 5 years event-free survival of 60% with two patients failing to engraft, one dying 17 months post-HSCT from a respiratory illness, and one death from relapsed MDS. This study noted that patients transplanted prior to the development of malignancy had improved transplant outcomes, and that patients transplanted for BMF demonstrated sustained engraftment with myeloablative regimens based on busulfan and cyclophosphamide. The authors felt that more intense conditioning for BMF-SDS is not warranted, but reduced intensity conditioning for patients with MDS/AML would be insufficient (45). In contrast, Burroughs et al. document three SDS patients transplanted from matched unrelated donors (MUD) with treosulfan and fludarabine conditioning regimens and no serotherapy; these patients underwent transplantation for BMF and achieved sustained engraftment with acute graft vs. host disease (GvHD) developing in two of the three (13).
Thus, for patients with SDS, alloHSCT can correct the hematopoietic aspects of the disease with improved outcomes if transplantation is undertaken prior to development of MDS/AML. Regular monitoring for cytogenetic abnormalities is thus recommended, although some appear to be indolent or transient and are not an automatic indication for HSCT [e.g., i7(q10) and del(20q)]. Larger studies are required before recommendations can be made regarding ideal conditioning regimens (treosulfan or busulfan) and need for serotherapy (particularly when using MUDs) (12,13,44,45).

Deficiency of VPS45 Protein
Deficiency of VPS45 protein leads to impaired trafficking of endosomes and lysosomes, with impaired degranulation, release of inflammatory mediators, and neutrophil migration. Patients present very early in life (before 1 year of age) with severe, recurrent, deep-seated bacterial and fungal infections, and a severe neutropenia unresponsive to G-CSF therapy. Bone marrow biopsy classically demonstrates primary myelofibrosis with a dry tap (26)(27)(28)(29). HSCT is indicated for severe neutropenia unresponsive to G-CSF and for recurrent, severe infections (27,30). alloHSCT should be undertaken as early as possible with a myeloablative regimen including busulfan (27). A total of nine transplants have been performed for children with biallelic mutations in the VPS45 gene, with three deaths and six patients surviving. Surviving patients were transfusion-independent with resolution of the extramedullary hematopoiesis if full donor chimerism was achieved (27).

Group 2: Neutropenia Without Syndromic Disease
This group includes congenital neutropenia caused by mutations in ELANE, GFI1, HAX1, WAS (X-linked neutropenia), CSF3R, and SRP54 genes ( Table 2) (40,43,46). Several reports on successful alloHSCT are available for both reduced intensity conditioning (RIC), and myeloablative conditioning (MAC) transplantation for this group of diseases ( Table 2) and patients younger than 10 years of age appear to have favorable outcomes (7). Severe infections are seen in ELANE and HAX1 mutations, thus alloHSCT is indicated in these patients, particularly if they require high-dose G-CSF to maintain their neutrophil count, or progress to MDS/AML (it has been found that patients who require more than 8 mcg/kg/day of G-CSF to maintain a neutrophil count above 0.5 × 10 9 /L have an increased risk of sepsis and MDS/AML) (43,46,55). With regard to HAX1, patients with mutations in exon 2 encoding isoform A develop isolated congenital neutropenia, whereas other mutations encoding both isoform A and B cause hematological and neurological manifestations (neurological delay, epilepsy). Thus, the degree of neurological impairment should be taken into account when considering alloHSCT (48,49). Patients with GFI1 and WAS gain of function (GOF) mutations are very rare and seem to present with mild to moderate neutropenia, and there are no reports on alloHSCT in these patients (40,47). WAS GOF (X-linked neutropenia) is a distinct clinical entity from Wiskott Aldrich Syndrome (caused by WAS loss of function) with male patients exhibiting variable neutropenia, recurrent infections, and lymphopenia, with normal platelet counts and  (54) no eczema. Transient myelodysplastic changes have been seen in the bone marrow of some patients, resolving in most cases but progressing to AML at an elderly age in one case (51)(52)(53). In contrast, patients with acquired mutations in CSF3R present very often with MDS/AML, therefore yearly screening for receptor mutations is indicated, as is pre-emptive alloHSCT ( Table 2) (43).

Group 3: Phagocyte Function With Syndromic Disease
This group of disease is characterized by defective neutrophil function with normal or elevated neutrophil numbers that is part of a broader syndrome (  (61)(62)(63)(64)(65). However, a recent study of the EBMT/IEWP on 83 transplanted LAD patients shows no significant benefit in patients receiving MAC vs. non-myeloablation. Furthermore, regardless of the conditioning regimen, a relatively high frequency of severe inflammatory complications (graft rejection and severe aGVHD) during alloHSCT was observed. This cohort included a few LAD III patients, and their outcome was not significantly different from that of the LAD I patients. Early transplantation using anti-inflammtory treatment pre-alloHSCT and the additional use of thiotepa in the conditioning protocol might be beneficial [unpublished data of author SB, (66)]. Following on from successful gene therapy trials for other primary immune deficiencies, a gene therapy phase I/II trial for LAD-1 has been announced for the end of 2019 whereby autologous CD34+ stem cells of patients with LAD1 will be transduced using a lentiviral vector; stem cells successfully transduced will be transplanted following conditioning with a low dose of busulfan (Rocket Pharma, USA; https://clinicaltrials. gov/ct2/show/NCT03812263, https://clinicaltrials.gov/ct2/show/ NCT03825783) (67).

Group 4: Phagocyte Dysfunction Without Syndromic Disease
Excluding CGD, the only disease in this group with evidence of successful bone marrow transplantation is MonoMAC (monocytopenia and mycobacterial infections syndrome) ( Table 4). This disease is characterized by profoundly decreased or absent monocytes, B lymphocytes, natural killer (NK) cells, and circulating and tissue dendritic cells (DCs), with little, or no effect on T-lymphocyte numbers. Patients are susceptible to mycobacterial, viral and opportunistic fungal infections. Bone marrow dysfunction is prominent and variable with progressive aplastic and dysplastic changes. Stem cell transplantation is curative and can be performed pre-emptively in cases where a matched donor is available (68). Anti-inflammatory pre-treatment and proper infection control might result in lower transplant related mortality in these patients.

DISCUSSION
The list of neutrophil disorders is varied and expanding rapidly with the increasing availability of next generation gene sequencing. As our ability to establish a genetic diagnosis is improved, the challenge of linking genotype to phenotype arises, a challenge that is made difficult by the small numbers of patients diagnosed with each individual disease. There is much we still do not understand about the pathomechanism of these diseases, and this makes prognostication and treatment decisions difficult. Serious infectious complications and neutropenia are life threatening but can be treated with anti-microbial therapy and G-CSF, respectively, however inflammatory complications of these diseases are likely underappreciated and undertreated (5,6). Furthermore, the decision to undertake alloHSCT, with all its inheritent risks and potential complications, is made difficult by the lack of published data for most diseases. In 2015, the EBMT and SCETIDE released the findings of the largest retrospective cohort of patients with severe congenital neutropenia to undergo alloHSCT. The 136 patients in that study demonstrated an overall 3 years survival of 82%, with transplant related mortality at 17%. It concluded that transplantation should be considered in patients with severe infections or unresponsiveness to G-CSF or requiring high doses of G-CSF (over 8 mcg/kg/day to maintain an absolute neutrophil count over 0.5 × 10 9 /L). Both MAC and RIC conditioning was effective and transplant outcomes were improved if patients were transplanted before 10 years of age and before the development of MDS/AML (7). Despite the wide range of disease we have covered in this review, those findings appear to hold true, although in diseases where evidence is lacking family choice and centerspecific expertise must also be taken into account.
In diseases where the need to transplant has been established (e.g., deficiency of VPS45 protein, LAD, SDS), there is little to guide decision making as to the method of transplantation. The need to achieve stable myeloid engraftment, often into an impaired bone marrow environment, may well call for the use of toxicity reduced MAC over the RIC regimens that have become established practice for PIDs characterized by loss-of-function lymphoid deficiencies (e.g., severe combined immunodeficiency, or SCID) (27). Again, evidence is lacking regarding side effects, long-term outcomes, and the degree of chimerism required to maintain cure. Secondary graft failure or graft insufficiency in the case of mixed chimerism with a reappearance of posttransplant neutropenia has been reported in the literature and observed by the authors in some of the quantitative phagocytic disorders (73)(74)(75).
We recommend that all physicians treating patients with neutrophil disorders submit their data to the EBMT and SCETIDE, as it is only through the ordered collection of data that clinical experience can be gathered, studied and shared. It is our hope that in the coming years many disease mechanisms will be uncovered and long-term treatment data accumulated, which will be of great benefit to physicians, patients, and their families.

AUTHOR CONTRIBUTIONS
SB, BS, and PS designed the study, collected data, performed review, wrote the manuscript.