Long-Term Immune Recovery After Hematopoietic Stem Cell Transplantation for ADA Deficiency: a Single-Center Experience

Adenosine deaminase (ADA) deficiency results in one of the more common forms of autosomal recessive severe combined immunodeficiency (SCID) [1, 2]. ADA is an enzyme in the purine metabolism. More than 70 disease-causing mutations have been described, leading to an impaired purine salvage pathway with accumulation of toxic metabolites (adenosine, 2’deoxyadenosine, and deoxyadenosinetriphosphate (dATP)), as well as inactivation of S-adenosylhomocysteine hydrolase (SAHase) [3]. ADA expression is particularly high in lymphoid tissues, and therefore, ADA deficiency leads to severe abnormalities in lymphoid development [4‐6]. It is typically associated with lymphopenia with abnormal T, B, and natural killer (NK) cell numbers and function. Patients usually present with failure to thrive and severe or recurrent opportunistic infections. Additionally, patients display typical non-immunological features, including sensorineural hearing loss [7, 8], non-infectious pneumonitis, and pulmonary alveolar proteinosis [9, 10], skeletal dysplasias [11‐14], urogenital abnormalities [15], hepatic dysfunction [16], cognitive impairment, and behavioral abnormalities [17, 18]. This variety of clinical manifestations can be explained by the metabolic pathogenesis affecting multiple organs and tissues [3]. The immune system is the most severely affected, and without treatment, ADA-deficient individuals die in their first or second year of life from infective causes. Early intervention is thus critical. Unlike other forms of SCID, a number of options are available for the treatment of ADA-SCID [1, 19, 20], namely hematopoietic stem cell transplantation (HSCT) [21], enzyme replacement therapy (ERT) with PEG-ADA [22], and more recently gene therapy (GT) where both licensed and experimental therapies are now available [23‐26]. Importantly, the survival associated with GT, which is currently 100%, invites comparison with HSCT outcomes.

Several single-center [27‐30] and multicenter studies [21, 31, 32] have highlighted the efficiency of HSCT, in particular from HLA-matched sibling donors (MSDs) and HLA-matched family donors (MFDs), providing long-term correction of the immune and metabolic abnormalities in ADA-deficient patients. The largest study to date is a multicenter study investigating the outcome of HSCT in 106 ADA-deficient SCID patients [21]. Fifty-four patients included in this study received a MSD or MFD HSCT between 1981 and 2009. These patients showed a better overall survival (OS) (86% and 81%, respectively) than patients who underwent a transplant procedure from HLA-matched unrelated donors (MUDs), haploidentical donors, or HLA-mismatched unrelated donors (MMUD) (67%, 43%, and 29%, respectively). Most HSCT procedures with MSD/MFD were unconditioned, i.e., without reduced intensity conditioning (RIC) or myeloablative conditioning (MAC). However, these data and data from other ADA specific studies were collected from patients as far back as 1981. As such, these data are not reflective of current HSCT practices and outcomes and may not be an appropriate comparison for the more recently conducted GT studies. Additionally, data on immune reconstitution and other parameters of immune recovery, such as discontinuation of immunoglobulin replacement treatment (IgRT), were not available for many patients included in these studies. With this in mind, we sought to collect data from a more contemporaneous ADA patient cohort, specifically from a center undertaking both HSCT and GT with similar standards of patient care and monitoring. This report therefore documents the outcome of all patients with ADA SCID who have undergone HSCT from 2000 onwards at a single center. To our knowledge, this is the largest single-center cohort to date.

We undertook a retrospective analysis of the outcome of all HSCT procedures performed for ADA-deficient SCID at Great Ormond Street Hospital (GOSH), London, since 2000. Thirty-one HSCT procedures were performed in 28 patients between September 2001 and January 2019. Transplant procedures in 12 patients happened before March 2009, and analysis of their outcome was included in a previous multicenter study [21]. We report more detailed outcomes for these patients together with the first description of 16 additional patients, who have been transplanted after March 2009. ADA deficiency was diagnosed by impaired or absent ADA activity and was confirmed by genetic testing. Before HSCT, all patients received supportive care, including antimicrobial prophylaxis, IgRT, and in most cases ERT with PEG-ADA. Upon referral, all patients were assessed for HSCT at GOSH and were managed there immediately after HSCT. Long-term follow-up was shared with local centers in the UK and abroad. Each patient included in this study had at least 2 years follow-up at GOSH after transplantation, except two, of whom one was last assessed from a laboratory point of view at GOSH at 0.9 years post-transplantation and one was transplanted in January 2019. This second patient had 1 year follow-up up to the endpoint for data collection. Data was retrieved from center-based electronic patient records and was stored anonymously. Pre-transplant data, such as gender, ADA activity pre-ERT when available, initial clinical presentation, and comorbidities at the time of transplantation, were collected for all patients. Characteristics of the transplant procedures were gathered including patient age at the time of transplantation, donor source, tissue-typing results, stem cell source, and conditioning details. Matched donors were defined as displaying a 10/10 match in tissue type with the recipient. We assessed survival time after HSCT up to January 2020 and, where applicable, cause of death. Kaplan–Meier curves were used to analyze overall and event-free survival (OS and EFS) in SPSS (events are defined as death, administration of an additional cell infusion from the same donor or the need for a second HSCT procedure using a different donor). Three patients received a second transplant, and one patient underwent HSCT after previously having received HSC-GT. For these four patients, survival was calculated from the last treatment received. For the two patients who received a second infusion from the same donor, survival was calculated from the date of the initial procedure. The log rank test was used to compare the survival distributions between subgroups differentiating between either donor source or type of conditioning. Transplant-related morbidity, in particular the incidence of acute and chronic graft-versus-host disease (GVHD), was analyzed. Laboratory results to evaluate immune reconstitution post-HSCT were collected at 6 months (± 1 month) and 12 months (± 1 month) after transplantation, as well as at the time of the last follow-up. Given the small cohort size, comparison of outcomes between subgroups was analyzed using descriptive statistics, with results plotted in Python using box plots. We also collected information on donor engraftment and, when available, ADA activity at the time of the last evaluation, as well as information regarding interruption of IgRT and vaccine responses at 2 years post-HSCT. For continuous variables, Kruskal–Wallis and Mann–Whitney tests were used to compare outcomes between subgroups. For categorical variables, chi-square and Fisher’s exact tests were used to assess differences in outcomes between subgroups.

A total of 28 ADA-deficient SCID patients (17 males and 11 females) underwent 31 transplant procedures at GOSH between 2000 and 2019. Patient characteristics pre-HSCT are summarized in Table 1. Twelve patients suffered from infections before transplant, of which half were receiving treatment for active infections at the time of transplantation. The median age at the time of transplantation was 12.5 months. The youngest patient was under a month old when transplanted, and the oldest patient, who had previously been treated by GT, was 119.6 months old. At least 7 patients (25%) did not receive ERT pre-HSCT. There was no significant difference in the age at transplant of those receiving ERT. Post-HSCT data was collected over a cumulative total of 209.9 years of follow-up after the 31 transplant procedures. The median follow-up after HSCT was 6.3 years with the longest follow-up period covering 17.4 years. The follow-up period for each procedure was determined from the time of transplantation to the time of the most recent laboratory assessment at GOSH up to 01/2020, the time of a second HSCT procedure, or the time of death. For this study, five patients, who continue to be monitored abroad by their referring centers, have been considered lost to follow-up after the last assessment of their immune reconstitution at GOSH.

Thanks to optimized supportive care over the past years, outcome following allogeneic HSCT has improved, as patients enter transplant in a better state of health. Further improvements are expected, thanks to the wider implementation of universal newborn screening (NBS) for SCID, allowing for earlier diagnosis and initiation of protective measures, including prophylactic antimicrobials, IgRT, and isolation [33]. Recently updated guidelines for the management of ADA-SCID recommend that all patients should also receive ERT upon diagnosis [20]. This should be followed as soon as feasible by definitive treatment with either of 2 available options: allogeneic HSCT or ex vivo-corrected autologous HSC-GT. In particular, the most recent consensus statement recommends that HSC-GT and MSD/MFD HSCT (without RIC) are seen as equal therapeutic options [20]. Previous studies have reported only smaller numbers of patients who underwent unconditioned infusions [27‐30, 32], but this recommendation is supported by the findings from the largest multicenter study to date, which in addition to better OS reports faster T cell immune recovery after MSD/MFD transplant procedures, as well as good humoral recovery even though most were unconditioned procedures [21]. Overall, detailed analysis of the long-term outcome after HSCT in ADA deficiency is still limited. Moreover, there are no reports focused on cohorts of ADA-deficient patients who have been transplanted more recently, which would constitute a more appropriate comparison for current GT studies. GOSH is one of the biggest pediatric bone marrow transplantation centers worldwide with one of the largest cohorts of transplanted ADA-deficient SCID patients. We thus undertook a single-center study reviewing the immunological outcome of all HSCT procedures performed in ADA-deficient patients since 2000. To our knowledge, this is the largest single-center cohort, with 28 ADA-deficient patients, who collectively underwent 31 HSCT procedures.

More than half of these procedures were unconditioned MSD/MFD transplants. GVHD complicated 45% of all procedures and, similarly, 47% of the procedures using MSDs and MFDs. In some patients, GVHD was less than grade II, but any degree of GVHD is undesirable in this population of patients. As in previous studies, we observed a trend of improved OS after unconditioned procedures (Fig. 1). However, four of these patients (22%) required a second intervention to achieve long-term engraftment and immune reconstitution. Specifically, two patients received a second unconditioned infusion of whole bone marrow from the same donor, and two patients underwent second myeloablative HSCT procedures from MSDs, after first unconditioned procedures with a different MSD and a MFD, respectively. The reasons for this higher failure rate after unconditioned MSD/MFD remain poorly understood. These patients received ERT prior to HSCT, and one hypothesis is that the level of immune function established by ERT at the time of HSCT may contribute to rejection or non-engraftment of donor cells. Standard practice in our center is to initiate ERT at diagnosis and, in general, stop ERT 1 month prior to proceeding to HSCT. However, detailed data on when exactly ERT was discontinued is not available, and therefore, we could not address whether the timing of ERT discontinuation impacts hematopoietic engraftment after HSCT. Overall, our data shows that, alternatively, the use of RIC, even in MSD/MFD procedures, could be used to deplete the cells resulting from the immune reconstitution due to ERT.

Independently of the type of transplant procedure, whether unconditioned MSD/MFD procedures or not, we observe similar absolute T and B cell counts at > 2 years post-HSCT (Figs. 3 and 4 and data not shown). However, long-term follow-up of our patient cohort highlights that, whereas high levels of lymphoid donor engraftment are achieved in the CD3⁺ compartment even in the absence of conditioning, donor engraftment in CD19⁺ and CD15⁺ compartments is significantly better in conditioned procedures than in unconditioned ones (Fig. 2 and data not shown). Previous observations suggested that there may be a selective advantage for donor B cell engraftment [21], but in more than half the patients in our cohort, we observed a good correlation between B cell and myeloid engraftment when paired results were available. Interestingly, approximately 30% of the patients in our cohort had not yet discontinued IgRT at 2 years post-HSCT. All of these patients, except for one, had undergone an unconditioned procedure and were found to have absent myeloid engraftment. These patients also show poor metabolic correction with lower ADA activity at their last follow-up appointment than the other patients in the cohort. All patients with absent ADA activity and with elevated dATP levels after HSCT (all unconditioned with low myeloid chimerism) remained on IgRT at 2 years post-HSCT, suggesting that insufficient myeloid engraftment is associated with inadequate metabolic correction post-HSCT. The lack of conditioning thus comes at a significant cost.

Previous studies have already reported that lymphoid and humoral immune reconstitution in SCID patients post-HSCT are improved using conditioning procedures [34, 35]. Additionally, a recent study showed that the use of a RIC regimen for non-malignant HSCT procedures using HLA-identical donors, including MSDs and MFDs, can achieve sustained myeloid engraftment with a low incidence of GVHD and conditioning-related toxicities [36]. A case report of 2 ADA-deficient siblings treated by MSD HSCT, one unconditioned and one after RIC, already proposed the need to explore the use of RIC in HSCT procedures from HLA-matched donors to improve long-term immune reconstitution with better myeloid engraftment and metabolic recovery in ADA SCID specifically as well [37]. In our cohort, all first HSCT procedures using a MSD/MFD were unconditioned. Nevertheless, we postulate that routine use of RIC in ADA SCID patients receiving a MSD/MFD transplant may improve the long-term outcome of these patients by achieving better engraftment.

ADA-deficient SCID is rare, and the cohort size of our study remains small. To further confirm the potential role for RIC in the long-term outcome for ADA SCID patients after MSD and MFD transplants, this type of procedures must continue to be performed in the context of multicenter clinical trials. This is particularly important for ADA-deficiency given that, unlike other forms of SCID, its management includes multiple treatment options. Specifically, more than 100 ADA-deficient patients have been treated with HSC-GT and the overall survival is 100% [23‐26, 38] (unpublished data and personal communications), including from the use of an approved treatment in Europe using a gammaretroviral vector licensed as Strimvelis® which is available in a single center [38, 39]. This treatment requires RIC with single agent low-dose busulfan (dose range ~ 4–5 mg/kg for a target AUC of 20 mg/L × hour) to achieve engraftment of gene-corrected cells, immune recovery, and metabolic detoxification [23‐25, 40, 41]. This conditioning regimen is typically well tolerated and is milder than RIC regimens used in allo-HSCT. Until recently, no serious adverse events, nor genotoxic insertional mutagenesis, had been reported in the largest cohort of ADA-deficient patients treated with HSC-GT [42]. A patient with Strimvelis® has very recently been diagnosed with T cell leukemia [43]. Causality is under investigation. Global clinical trials using a lentiviral approach have demonstrated extremely promising results ([44], unpublished data and personal communication) and may offer a further treatment option in the near future. No insertional oncogenesis has been reported after lentiviral vector-based GT, in any indication [45].

In conclusion, long-term outcome data is evolving, and further monitoring is required. Longitudinal data collection in appropriately powered prospective trials will inform treatment optimization for future patients and will be the basis for updates to treatment guidelines. This will become increasingly important as newborn screening programs are introduced more widely, given several therapeutic approaches exist to treat ADA SCID. We propose that, if accessible, HSC-GT should possibly be considered a preferred first-line treatment to HSCT, even when a MSD or MFD is available, as the autologous nature of this procedure abrogates any risk of alloreactivity, in particular if further studies support the recommendation for routine use of RIC for all HSCT procedures using HLA-identical donors in order to achieve better myeloid engraftment, humoral immune recovery, and metabolic correction.