Sickle cell disease (SCD) is currently the abnormality detected most frequently in newborn screening tests in several countries, including Spain. In the Autonomous Community of Madrid, where screening has been performed since May 2003, this is corroborated with an incidence of 1 in every 5000 live births.1 Sickle cell disease is a significant source of multisystemic morbidity and carries an increased risk of early death.2 In the past three decades, the introduction of several measures, such as penicillin prophylaxis, vaccination against Streptococcus pneumoniae, the use of hydroxyurea, have been implemented leading to a decrease in morbidity and mortality in patients with SCD.3–5 There is evidence that hydroxyurea considerably reduces the incidence of vaso-occlusive events and acute chest syndrome6,7; however, this treatment does not reverse organ damage or reduce the risk of stroke. The use of transcranial Doppler ultrasonography (TCD) in the followup of these patients allows the identification of those that are at high risk of experiencing an acute stroke, while of brain magnetic resonance imaging (MRI) allows the diagnosis of silent strokes.8 Transfusion therapy has proven to be useful in the primary prevention of stroke, but it has the disadvantage of secondary iron overload and poor long-term adherence.9

At present, haematopoietic stem cell transplantation (HSCT) is the only available curative treatment. The first successful HSCT was performed in 1984 in the Untied States in a patient with SCD and acute myeloid leukaemia.7 Subsequent multicentre studies have reported good outcomes, with an overall survival of 93–94% and an event-free survival of 82–86%, while the most frequent complications are graft failure and graft-versus-host disease (GVHD).10–14

The aim of our study was to describe the clinical and analytical outcomes of patients with SCD that underwent allogeneic HSCT from an HLA-identical sibling donor in our hospital.

We conducted a retrospective descriptive single-centre study. We collected data for paediatric patients with homozygous (HbSS) or double heterozygous (HbSβ0) SCD that underwent bone marrow transplantation from an HLA-identical sibling donor (Hb AA or Hb AS phenotype) between January 2010 and December 2014, with followup until December 2015. We studied demographic variables (age, sex) and variables concerning the patients’ history of complications associated with the disease: acute chest syndrome (ACT), vaso-occlusive crises (VOCs), dactylitis, priapism, stroke, splenic sequestration and need for splenectomy. As regards treatment prior to HSCT, we documented whether patients received hydroxyurea (HU), transfusion therapy or iron chelation therapy. Other variables under study associated with HSCT were donor-recipient ABO compatibility, stem cell infusion, day of platelet and neutrophil engraftment and chimerism outcomes in bone marrow at day +30 post transplant and in peripheral blood later in time. We collected the following laboratory parameters measured at baseline and at 3, 6, 9 and 12 months post transplantation: haemoglobin, reticulocytes, lactase dehydrogenase (LDH), bilirubin, ferritin, glomerular filtration rate based on cystatin C, lymphocyte differential count (B and T lymphocytes, NK cells), and immunoglobulins (Igs). Last of all, we gathered data for post-transplant complications: acute GVHD, neurologic complications (seizures, subarachnoid haemorrhage [SAH], neurotoxicity), acute renal failure, hypertension (HTN), cytomegalovirus (CMV) reactivation and cause of death.

Despite the small sample size (n=11), the results of our study were consistent with those of multicentre studies8,9,15 conducted in countries with more experience in the treatment of children with SCD. Performance of HSCT in these patients requires not only knowledge of transplantation technique, but also a thorough understanding of the overall treatment of haemoglobinopathies. In Spain, access to cohorts of children with SCD is a relatively new phenomenon, so this is the largest series studied to date.

One patient died of acute GVHD, which is also one of the main causes of death in other series (Table 3). The incidence of acute GVHD (grade>II) was 36%, higher than the incidence reported in other series despite similar prophylaxis regimens with ciclosporin and methotrexate, although this result may have been different with a larger sample size. A possible cause is that the source of the stem cells was bone marrow in all cases, whereas in other studies the source is umbilical cord blood, which carries a lower risk of acute GVHD, in up to 14% of cases.14,16,17 At any rate, none of the patients in our series developed chronic GVHD.

Cerebrovascular problems are a significant cause of neurologic morbidity associated with transplantation in patients with SCD. Walter et al.18 reported an incidence of neurologic complications post HSCT of 50% in patients with a prior history of cerebral infarction and of 23% in the rest of the patients; however, a French series13 did not find a significant increase in risk in patients with a past history of cerebrovascular disease. In our series, the two patients with evidence of stroke in the MRI scan performed before HSCT did not develop acute neurologic complications. Hypertension developed in seven out of the eleven patients and caused significant morbidity, as despite the measures taken for its management, it was probably related to SAH in three patients. The incidence of seizures in the French and North American series was 24% and 28%, respectively, and seizures are the most frequent complication post HSCT1; they usually develop in the first few days after transplantation in association with hypertension and with corticosteroid therapy.13,18 In our study, and despite prophylaxis with phenytoin, three of the eleven patients developed seizures, all of them in association with hypertension. Patient 1 had a new episode of seizures on day +27, this time due to ciclosporin toxicity, and leading to mild cognitive impairment. In our series, patients did not die from neurologic complications nor develop severe neurologic sequelae.

Graft failure and autologous reconstitution also continue to be among the major problems that follow HSCT in patients with SCD, with an incidence that ranges between 7% and 13% (Table 3) and of 9% in our study. Among its causes are previous alloimmunisation due to multiple transfusions, bone marrow hyper-regeneration, blood group incompatibility, cell dose (CD34+) and viral infection. The French series demonstrated that intensification of immunosuppression with addition of rabbit anti-thymocyte globulin to the conditioning regimen reduces the probability of graft rejection from 22.6% to 3%.13 All our patients received alemtuzumab to induce lymphocyte depletion. Patient 9 experienced autologous reconstitution with recurrence of SCD symptoms on day +180, with the main risk factors for it being a cell dose near the lower limit, major blood group incompatibility and reactivation of CMV (Tables 1 and 2). Discontinuation of immunosuppression and infusion of donor T lymphocytes did not succeed in improving chimerism in our patient, who had full donor chimerism in the overall bone marrow at day +30, but with 51% of T lymphocytes already of recipient origin. Similarly, patient 11 had mixed chimerism in the T lymphocyte population at day +30, which led to a change in management, intensifying immunosuppression with the addition of mycophenolate to ciclosporin; this had achieved a stable chimerism at one year of followup in the absence of clinical manifestations of SCD (Table 2). Mixed chimerism can be an acceptable option in nonmalignant blood disorders, and although the main risk it carries is autologous reconstitution in the first year, if it remains stable it can prevent the development of clinical manifestations of the disease19 and improve haemolysis parameters.20

The findings of multicentre studies have led to the development of new management strategies with the purpose of reducing transplant-related mortality, graft failure and treatment toxicity. Thus, a study conducted in Belgium found a significant improvement in event-free survival (no cases of graft failure or autologous reconstitution) in the group treated with hydroxyurea in the months preceding HSCT, compared to the group treated with the standard conditioning regimen with busulfan, cyclophosphamide and anti-thymocyte globulin and no hydroxyurea (97.4% vs. 66.7%; P=.006).21

Recently, based on the results of studies in adults treated with reduced-toxicity conditioning regimens and one study in children with thalassaemia major,22 a study in 15 children with SCD used a conditioning regimen consisting of treosulfan, thiotepa and fludarabin. Treosulfan has a structure analogous to that of busulfan and exhibits a potent myeloablative activity combined with minimal extramedullary toxicity and a more linear pharmacokinetic profile. This study obtained excellent results, with an overall survival of 100%, an event-free survival of 93% and a decreased treatment-related toxicity for HSCT, encouraging a change in the conditioning of these patients.23

In our study, HSCT was indicated in all patients based on the clinical practice guidelines of SCD of the SEHOP1; however, patients that receive the transplant in the early stages of the disease have the best outcomes, as was observed in the Belgian series, in which overall survival and event-free survival of asymptomatic patients were higher than those of patients with severe forms of disease (100% vs. 88% and 93% vs. 76%, respectively).12 This is why there is an ongoing debate on whether HSCT from an HLA-identical sibling donor should be performed in asymptomatic patients or its indications should be expanded.24

Haematopoietic stem cell transplantation is the only curative treatment available to patients with SCD, and our results are consistent with those of the international literature (overall survival, 90.9%, event-free survival, 81.82%). However, it still carries a risk of death (9%), with the most frequent reason being GVHD, and of significant comorbidities, such as graft failure (9%) and neurologic complications (36%). In light of these results and the reviewed literature, we decided to make changes to the conditioning regimen (treosulfan, fludarabin, thiotepa), the suppression of erythropoiesis in the two months prior to HSCT and pre-transplantation ambulatory blood pressure monitoring, and are contemplating a change in GVHD prophylaxis. T lymphocyte chimerism analysis helps the early detection of mixed chimerism and, in our experience, leads to better outcomes through the subsequent intensification of immunosuppression to prevent graft failure.

The authors have no conflict of interests to declare.