Introduction

In 2014, biallelic deleterious mutations in the cat eye chromosome region 1 gene (CECR1, subsequently renamed ADA2), encoding adenosine deaminase 2 (ADA2), were reported as the cause of a monogenic inborn error of immunity disease, deficiency of ADA2 (DADA2) (OMIM # 615,688) [1, 2]. The phenotype comprises recurrent fever and vasculopathy, ranging from livedo racemosa and polyarteritis nodosa (PAN) to intracranial vasculopathy with lacunar strokes and hemorrhages [1,2,3]. Cytopenias, either autoimmune or due to bone marrow failure (BMF), occur in 50% of patients and present as congenital pure red cell aplasia (PRCA), neutropenia, thrombocytopenia, or pancytopenia [4,5,6]. Immunodeficiency with hypogammaglobulinemia and recurrent viral and bacterial infections and malignant lymphoproliferation (T-large granular lymphocyte leukemia (T-LGL leukemia) and lymphoma) have also been described [5, 7,8,9]. DADA2 diagnosis is based on an absence or low levels of plasma ADA2 enzymatic activity and the demonstration of biallelic loss-of-function mutations of ADA2 [10]. The pathophysiology of DADA2 remains unclear. A picture emerges where ADA2 deficiency results in skewing of macrophage differentiation towards inflammatory M1 macrophages [11], leading to endothelial instability, as shown in a zebrafish model and in endothelial cell coculture systems [1]. Recent findings have revealed an even more complex interplay between endothelial cells and monocytes and macrophages with marked ADA2 secretion by endothelial cell lines [12].

Treatment of DADA2 is challenging and case mortality is estimated to be around 8%, mostly in childhood and related to vasculopathy-associated complications and infections [13,14,15]. None of the classical immunosuppressive drugs are an option for the long-term treatment of DADA2, because their efficacy is temporary, especially for the DADA2 related cytopenia, or because of the toxicity associated with long-term use. Anti-TNF agents, etanercept in particular, are the mainstay of treatment for the inflammatory and vasculopathy phenotypes [16]. However, anti-TNF agents do not cure the hematological phenotype and in a proportion of patients, vasculopathy persists despite anti-TNF treatment [17, 18]. Finally, the cost of life-long TNF-inhibition is a limitation for some patients. Hematopoietic cell transplantation (HCT) has been reported to result in a rapid and sustained resolution of the systemic inflammation and hematological phenotype, with all patients from a cohort of 14 patients surviving after HCT [5, 19,20,21]. We report here the results of a multinational study of a cohort of 30 patients with DADA2 undergoing HCT, including the previously reported cases.

Methods

Overview of the Study

We conducted an investigator-driven retrospective international non-interventional multicenter study on HCT for DADA2. Invitations to participate were sent to the physicians allied to the DADA2 Foundation, the European Group for Blood and Marrow Transplantation (EBMT), and the European Society for Immunodeficiencies (ESID). We also invited all authors of published single case reports on HCT in DADA2 to participate in the study. Data collection began after the Second Inaugural International Conference on DADA2 hosted by the DADA2 Foundation on November 9, 2018. The study was approved by the Ethics Committee of Leuven University Hospitals (study number S63982). The study was performed in accordance with the principles of the Declaration of Helsinki. The authors assume responsibility for the accuracy and completeness of the data and analyses and for fidelity to the study protocol.

Patients

The criteria for patient inclusion in the study were as follows: (1) genetic diagnosis of DADA2 and/or clinical findings consistent with DADA2 and plasma ADA2 activity level in the deficient range and (2) HCT performed with a follow-up time for survivors of at least 3 months after HCT. All participating physicians completed a questionnaire. All patients or their guardians gave written informed consent for data collection. Patients for whom incomplete data were obtained (indication for HCT, age at HCT, total nucleated cell dose or CD34 + stem cell dose, stem cell donor, conditioning regimen, graft-versus-host disease (GvHD) prophylaxis, time to engraftment, graft failure, conditioning for subsequent HCTs, chimerism) were excluded from the study.

HCT Data

Neutrophil engraftment was defined as the first of three consecutive days with a neutrophil count ≥ 0.5 × 109/L and platelet engraftment as the first of seven consecutive days with a platelet count ≥ 20 × 109/L without platelet transfusion in the prior 7 days. Full donor chimerism was defined as ≥ 95% donor cells in myeloid or whole-blood fractions. The type of test was at the discretion of the transplant center. Primary and secondary graft failures (GF) were defined according to EBMT guidelines. Second (or third) HCT was defined as the infusion of hematopoietic progenitor cell containing product, according to CIBMTR, regardless of conditioning regimen. The diagnosis and grading of acute and chronic GvHD were based on international standard criteria [22]. Transplant regimen, GvHD prophylaxis, antimicrobial prophylaxis, and pre-emptive treatment were chosen according to center preferences. Preparative regimens were classified as reduced intensity conditioning (RIC) if the dose of alkylating agents or TBI is reduced by at least 30% from a myeloablative conditioning (MAC) approach. A total dose of treosulfan > 30 g/m2 was considered MAC whether or not combined with another alkylator, whereas a total busulfan dose < 8 mg/kg and fludarabine-melphalan regimens were considered RIC.

Kaplan–Meier curves were plotted for overall survival (OS) and GvHD-free relapse-free survival (GRFS), and p values were obtained for Mantel-Cox log-rank tests performed with Graph-Pad Prism Software version 9. Values of p < 0.05 were considered statistically significant. GvHD relapse-free survival was calculated as the time from first HCT until the first occurrence of any of the following events: grades 3–4 aGvHD or moderate/severe cGvHD GF, disease relapse (poor graft function/graft failure with DADA2 disease relapse requiring repeat transplant), or death. HCTs inadvertently using affected donors were excluded. The cumulative incidence of GvHD and GF were also calculated using competing risk analysis, using R, for all HCT procedures, but excluding HCTs from affected donors.

Results

Patient Characteristics and Diagnosis of DADA2

We included 30 DADA2 patients undergoing HCT between 2000 and 2020 in this study. Four other patients were excluded due to incomplete data sets. Patients underwent HCT at 21 different centers from 12 countries in Europe and North America. Twenty of the patients have been reported before [4, 5, 7, 9, 19,20,21, 23,24,25,26,27]. Median age at disease onset was 2.25 years (range: birth to 16 years). Median age at genetic diagnosis was 12 years (range: 2–28 years) (Table 1). DADA2 diagnosis was confirmed at the molecular level in all patients, by demonstrating the presence of biallelic pathogenic ADA2 variants. Plasma ADA2 activity was assessed before HCT in 18 patients and was low in all cases. Twenty-six patients harbored known pathogenic ADA2 mutations. The R169Q variant was the most common mutation, found in 15 patients. Four patients harbored novel mutations, all with combined annotation-dependent depletion (CADD) scores above the mutation significance cutoff (MSC) for this gene; all these variants were private or had a MAF < 10–6 [28], strongly suggestive of a deleterious effect. Three of these patients were tested for ADA2 enzyme activity, which was found to be low or absent (Table 2). Patient and HCT characteristics are summarized in Table S1 in the Online Supplement.

Table 1 Demographic and clinical features of the 30 DADA2 patients before HCT
Full size table
Table 2 Genetics and ADA2 enzymatic activity for the 30 DADA2 patients
Full size table

Hematological Phenotype Pre-HCT

PRCA was documented in 8/30 patients, isolated neutropenia in 6/30, combined RCA and neutropenia in 5/30, severe aplastic anemia in 1/30, severe lymphopenia in 1/30, anemia and neutropenia in 2/30, autoimmune hemolytic anemia (AIHA) in 2/30, and pancytopenia in 5/30 patients, at presentation. Six patients had a hematological malignancy or myelodysplasia and received HCT as part of the therapeutic approach (P4, P9, P17, P21, P26, and P27). Twenty-nine patients had received at least one immunosuppressive treatment prior to HCT including 13 patients who received at least 3 lines of immunosuppressive medications. Fourteen patients had received anti-TNF agents before HCT, without effect on cytopenias. P6 received single agent etanercept, which failed to reverse neutropenia; a combination of adalimumab, cyclosporine, and low-dose prednisone resulted in normal neutrophil counts for 6 months prior to HCT. Two patients received pre-HCT Interleukin-1 receptor antagonist (anakinra), without amelioration of cytopenias or immune dysregulation. Seven patients received granulocyte colony stimulating factor (G-CSF) for neutropenia, with no response.

Immunological and Vasculitis Phenotype Pre-HCT

IgG levels were low in 12/27 tested patients, IgA levels were low in 13/26, and IgM levels were low in 15/27 tested patients. Recurrent infections were reported in 17 of the 30 patients, mostly viral infections (in 14/17). Herpesvirus infections predominated, with three patients suffering from recurrent herpes zoster, one having protracted CMV infection, one with severe chicken pox, one with recurrent cutaneous HSV-1, two with HHV6 viremia, and four with EBV viremia (2 transient, 1 chronic, and 1 in the context of lymphoproliferative disease). Warts (n = 4), and mollusca contagiosum (n = 4) were also reported. Immunoglobulin substitution treatment was administered to 15 of the 30 patients before HCT, and splenomegaly was reported in 23/30 patients. Fifteen of 30 patients had reported vasculitis prior to HCT (Table 1): 9 had livedo racemosa, 3 had polyarteritis nodosa (PAN) (P6 with livedo, P7 with ICH and livedo, P11 isolated), three patients had ischemic stroke (P2, 26, 27), and four had intracranial hemorrhage (P7,10, 27, P12).

Transplant Characteristics

The indications for HCT were cytopenia with or without immunodeficiency and/or lymphoproliferation or malignancy (Table 3). The median age at HCT was 9 years (range: 2–28). Six of the 30 patients had received HCT before the description of DADA2 in 2014. Two patients (P16, P29) were inadvertently transplanted using affected siblings as donors and received salvage second HCT from unrelated donors. A total of 38 HCTs were performed for 30 patients. Twenty patients received MAC (P12 with two subsequent HCTs), eight received RIC, and two received NMA conditioning for the final curative transplant (Tables 3 and S1). The most commonly used regimen (in 11 patients) was treosulfan/fludarabine ± thiotepa with antithymocyte globulin (ATG) or alemtuzumab. Serotherapy was used in 25/30 patients: ATG in 10 (rabbit ATG in all except horse ATG in P6 and P18) and alemtuzumab in 15 patients. The source of the stem cells for the final transplant was peripheral blood (PB) for 10 and bone marrow (BM) for 20 patients. Donors were HLA-matched related (n = 4), HLA-haploidentical sibling (n = 2), 10/10 HLA-matched unrelated (n = 16), and 9/10 HLA-mismatched unrelated (n = 8) for the final transplant (Tables 3 and S1). Two of the six related donors carried heterozygous ADA2 mutations (P3, P22), three were healthy (P1, P2, P21), and one donor was of unknown status (P4).

Table 3 HCT data and post-HCT complications for the 30 DADA2 patients
Full size table

Engraftment, Graft Failure, Transplant-Related Morbidity, Survival

Median engraftment was d + 20 for neutrophils and d + 23 for platelets. In 25/28 patients receiving grafts from unaffected donors, full donor chimerism was achieved by d + 30. Overall survival at 2 years was 97%, with a median follow-up of survivors of 2 years (range: 0.5–16), accounting for 1545 patient-months post-HCT (Fig. 1A). P24 passed away 2 months post-HCT due to respiratory failure secondary to parainfluenza pneumonia despite full donor chimerism and treatment with steroids and etanercept for the suspicion of immune reconstitution inflammatory syndrome. Viral reactivation occurred in 17/30 patients (56%). Adenovirus, CMV, and BK were most frequent involved, each observed in six patients. GRFS was 73% at 2 years, with all events occurring within the first year post-HCT. Of those transplanted with affected donors (n = 2), P16 had primary graft failure requiring salvage HCT, and P29 had obtained near-full donor chimerism but failed erythroid line engraftment and remained with PRCA. Three patients experienced graft failure. P18 required two and P19 one subsequent HCT for secondary GF. The latter two patients were found to have aggregates of CD8 + T cells in their BM with low donor T cell chimerism (9% and 0%) (Fig. 1B, Fig. 2B). P28 required a second HCT likely due to low stem cell dose (1.4 × 106/kg). P12 originally received MAC, but subsequently required two unconditioned HCTs due to drop in whole blood donor chimerism to 30% and new-onset RCA and agranulocytosis unresponsive to G-CSF. Cumulative incidence of aGvHD grades 2–4 was 20% at 1 year. Moderate-severe chronic GvHD developed in 2/30 patients (7%) at 1 year (Fig. 2A). P1, P19, and P24 developed sinusoidal obstruction syndrome (SOS), which responded to fluid restriction and diuresis in P1 and P19 and to defibrotide in P24. All three patients with SOS received either high-dose cyclophosphamide or myeloablative busulfan.

Fig. 1
figure 1

Kaplan–Meier curves representing A overall survival, B GvHD-free, relapse-free survival (GRFS). GvHD relapse-free survival was calculated as the time from first HCT until the first occurrence of any of the following events: grades 3–4 aGvHD or moderate/severe cGvHD GF, disease relapse (poor graft function/graft failure with DADA2 disease relapse requiring repeat transplant), or death. Overall survival is calculated on total number of patients (n = 30); GRFS is calculated on total number of HCT procedures, excluding the two procedures performed with stem cells from an affected sibling (n = 34). C Effect of HCT on clinical features resolution. Black squares indicate death post-HCT. Dark gray squares represent the presence of a clinical feature/phenotype. Light gray squares represent major improvement in clinical features. White squares represent complete resolution of clinical features. Each patient is presented by 2 attached columns (before and after HCT) for comparison. Follow-up time post-HCT for each patient is shown in months (second row). Severe infections represent any viral, bacterial, or fungal infection that required antiviral or antifungal treatment or led to sepsis. FTT, failure to thrive; HSM, hepatosplenomegaly; LAP, lymphadenopathy

Full size image
Fig. 2
figure 2

Kaplan–Meier curves representing A cumulative incidence of GvHD (a, acute grade 2 or higher; c, chronic moderate or severe); B cumulative incidence of graft failure. Cumulative incidence of GvHD and cumulative incidence of graft failure are calculated on total number of HCT procedures, excluding the two procedures performed with stem cells from an affected sibling (n = 34)

Full size image

Cure of DADA2

HCT cured the hematological phenotype in all patients, as confirmed at the most recent follow-up visit (Fig. 1C). No central vascular events were reported after engraftment (Fig. 1C). ADA2 plasma enzyme activity normalized in 16/17 patients tested post-HCT (Table 2). This normalization occurred as early as d + 12, coinciding with the reappearance of monocytes in the PB, as demonstrated by the prospective monitoring of plasma ADA2 enzyme activity in one patient [29]. At last follow-up, 29 patients were still alive, and 28 of these patients displayed full donor chimerism. Donor chimerism fell to 55% in P25 but remained stable over several months, with no evidence of disease. Transient hematological autoimmunity post-HCT was reported in three patients: ITP in two and AIHA in one. This autoimmunity responded to various treatment regimens (intravenous immunoglobulins (IVIG)/steroids/sirolimus/rituximab/bortezomib/romiplostim). Five patients are still on IVIG (one less than a year after HCT, one for mild bronchiectasis, and three post-rituximab treatment).

Discussion

We show here that HCT for DADA2 cures the hematological and immunological phenotypes of DADA2, with no new vascular events, with excellent survival, after a median follow-up of 2 years. Outside this study, two additional patients with DADA2 have received HCT and are reported to be alive and well [30, 31]. Another two patients have also undergone HCT, one of whom died after receiving a graft from a donor heterozygous for the pathogenic mutation in ADA2 (P. Stepensky, personal communication). Adding these patients to the current report, 32 of the 34 DADA2 patients who have undergone HCT were cured. Despite the temporary resolution of neutropenia by a multidrug approach (cyclosporine, steroids, adalimumab) in one patient prior to HCT, this option is not feasible in the long term, and HCT is, therefore, a valuable alternative. HCT also resolved the vascular phenotype in all 15 patients with vasculitis. No additional central nervous system vascular events were reported post-engraftment. Overall, the available data show an absence of new vascular events after engraftment post-HCT, demonstrating that hematopoietic cell-derived ADA2 plays a non-redundant role in restoring monocyte-endothelium interactions [12].

In the presence of this vascular phenotype, despite the impossibility of formal comparisons with other transplant indications, it is advisable to avoid high dose or untargeted busulfan and/or cyclophosphamide or high-dose radiation and to consider preventive measures for SOS. Liver disease was prominent and had a multifactorial etiology, directly related to DADA2 in some cases, but not in others (vasculitis, SOS, inflammation (DADA2/GvHD), iron overload, drug toxicity, GvHD). We therefore advise involving hepatologists in the care of DADA2 patients’ right from diagnosis, with imaging, functional assessment, liver biopsy, and iron chelation if indicated, and opting for the least hepatotoxic conditioning regimen. The evaluation of pre/post-HCT renal disease, although less well described, requires a similarly high level of attention.

Three patients who were grafted with a non-diseased donor, suffered GF. Low CD34+ stem cell dose probably contributed to GF of P28. The choice of conditioning may have contributed to secondary GF in P18 and P19, both of whom had low donor T cell chimerism preceding GF. Our data thus suggest that robust host lymphodepletion, more than myeloablation, is essential in preventing GF. It remains a matter of debate whether a related donor with a single deleterious allele of ADA2 can be considered a suitable donor. Two of the six related donors carried heterozygous ADA2 mutations (P3, P22: both HCT were successful), three were healthy (P1, P2, P21: all HCTs were successful), and one donor was of unknown status (P4). P25 has suffered no disease relapse, with a whole blood donor chimerism of 55%. In P12, whole blood donor chimerism fell to 30%, resulting in disease relapse, suggesting that there is a minimum required level of donor chimerism in DADA2. Indeed, patients carrying only a single pathogenic allele, with intermediate levels of ADA2 enzyme activity, have been reported to manifest DADA2 symptoms [10, 32]. In contrast, 9/10 MMUD and MUD may be considered suitable options for donors: There were 16 MUD and 8 MMUD transplants in this cohort.

The indication for HCT in this cohort was cytopenia and/or malignancy and immunodeficiency, not responding to treatment with TNF inhibitors. DADA2 patients with refractory BMF or immune cytopenia should be referred early on for HCT evaluation, given the morbidity and mortality due to hemorrhage, iron overload, infection, and long-term treatment with multiple immunosuppressive agents [18]. HCT is also a treatment option for patients who do not have access to anti-TNF agents even in the absence of immune cytopenia, BMF, or immunodeficiency. Theoretically, HCT could be a treatment option for patients on long-term treatment with anti-TNF inhibitors who develop neutralizing anti-drug antibodies [33]. Gene therapy (GT) is a promising option for the future, but is unlikely to be available to all DADA2 patients worldwide. In addition, GT requires conditioning and the presence of a sufficient number of autologous hematopoietic stem cells and may fail to reverse immune cytopenias if residual host T cells remain. Thus, GT is only feasible in patients without refractory cytopenia/BMF/malignancy, without prohibitive organ dysfunction, and without host T cell-mediated cytopenias. In the latter, HCT is likely the only potential curative treatment. For both HCT and GT, it emerges that full replacement of the host T cell compartment, and at least partial replacement of the myeloid compartment with ADA2 sufficient cells, is required.

In conclusion, we report here experience with the treatment of 30 DADA2 patients with cytopenia, BMF and immunodeficiency by HCT. All but one of the patients are alive and well and are cured at a median follow-up of 2 years. This successful treatment of an auto-inflammatory condition paves the way for application of HCT in other auto-inflammatory conditions refractory to classic immunosuppressive approaches.