Hematopoietic stem cell transplantation for activated phosphoinositide 3-kinase δ syndrome: Who, when, and how?

In this issue Okano et al¹ report on the outcome of 23 patients with activated phosphoinositide 3-kinase δ (PI3Kδ) syndrome (APDS), 9 of whom received hematopoietic stem cell transplantation (HSCT) for a severe clinical course. This experience adds to recent reports of the clinical spectrum of the disease in large cohort studies,^2–4 including outcomes of HSCT in 11 patients with APDS.⁵ Altogether, these studies raise the important questions of which patients should be considered for HSCT and when and what would be the optimal preparative regimens to facilitate durable engraftment and reduce the risk of transplant-related morbidity and mortality.

PI3Kδ is predominantly (although not exclusively) expressed in leukocytes and catalyzes conversion of phosphatidylinositol (4,5) bisphosphate to phosphatidylinositol (3,4,5) trisphosphate, which serves as a second messenger to activate the AKT kinase, leading to phosphorylation and inactivation of the transcription factor FOXO1 and activation of the mammalian target of rapamycin (mTOR) and S6 kinase (Table I).⁶ In this manner PI3Kδ plays a critical role in regulating immune responses.

TABLE I.

APDS: Clinical, immunologic, and biochemical features

Activated PI3Kδ syndrome
Clinical features	Immunologic features
Recurrent respiratory tract infections Bronchiectasis Lymphoproliferation Chronic viral infections (EBV and CMV) Autoimmune manifestations Gastrointestinal manifestations ↑ Risk of lymphoma Neurologic manifestations Short stature (APDS2)	Hypogammaglobulinemia Increased IgM levels Impaired vaccination responses Decreased naive T-cell counts Increased effector memory, effector, and exhausted T-cell counts Increased transitional B-cell counts Decreased memory and switched memory B-cell counts
mTOR-dependent and mTOR-independent effects of activated PI3K and increased AKT signaling
mTOR dependent	mTOR independent
Increased glucose uptake Metabolic shift to glycolysis	Increased FOXO1 phosphorylation Increased FOXO1 degradation Decreased FOXO1-dependent transcription
Consequences: ↓ TH cell differentiation ↓ Memory T-cell generation and function ↑ Effector T-cell numbers ↓ Memory CD8+ T-cell responses ↓ Regulatory T-cell homeostasis and function	Consequences: ↓ Naive T cells ↓ B-cell migration during GC reaction Impaired B-cell maturation ↑ Transitional B cells ↓ AID expression ↓ CSR

AID, Activation-induced cytidine deaminase; APDS2, type 2 APDS; CMV, cytomegalovirus; CSR, class-switch recombination; GC, germinal center.

PI3Kδ is a heterodimeric complex that includes a p110δ catalytic molecule and a p85a regulatory subunit encoded by the PIK3CD and PIK3R1 genes, respectively. Heterozygous gain-of-function mutations of p110δ and exon-skipping mutations of p85α result in hyperactivation of the PI3Kδ complex and cause type 1 and type 2 APDS, respectively.⁷ The clinical phenotype of APDS includes recurrent sinopulmonary infections with onset in childhood, often leading to bronchiectasis, acute and chronic viral infections, lymphoproliferative disease, increased risk of lymphoma, and immune dysregulation, including autoimmune cytopenias and gastrointestinal manifestations (Table I).^2–4 Typical immunologic abnormalities include a variable degree of hypogammaglobulinemia, often with increased serum IgM levels, reduced numbers of naive T cells with increased proportion of effector/effector memory T cells, and progressive B-cell lymphopenia with increased proportion of transitional B cells (Table I). These abnormalities are associated with evidence of enhanced PI3Kδ activity in circulating T and B cells, as shown by increased AKT and S6 phosphorylation.⁸

Consistent with these observations, in the series of 23 patients with type 1 APDS reported by Okano et al,¹ 30-year overall survival was 86.1%, but event-free survival was only 39.6%. Severe infections, lymphoproliferative disease, and enteropathy were documented in the vast majority of patients. Inability to control infections despite use of intravenous immunoglobulins and antibiotic prophylaxis and development of life-threatening episodes of lymphoproliferation that did not adequately respond to immunosuppressive agents represented the rationale for attempting HSCT in 9 of the 23 patients. This figure is higher than in previous studies. In particular, HSCT was performed in 5 of 53 patients with APDS reported by Coulter et al² and in 8 of 68 patients included in the European Society for Immunodeficiencies (ESID) APDS Registry.⁴ The higher demand for HSCT in the series reported by Okano et al¹ is likely to reflect a different approach to control immune dysregulation and in particular less common use of mTOR inhibitors (sirolimus and everolimus) in that cohort. The mTOR inhibitor sirolimus was the most frequently used immunosuppressive agent in the ESID-APDS Registry. Of 26 evaluable patients, 19 achieved complete or partial remission; the best responses were observed in patients with lymphoproliferation, whereas less satisfactory results were recorded in those with gastrointestinal disease or cytopenias.⁴

Furthermore, identification of the genetic defect in patients with APDS has prompted development of clinical trials based on PI3Kδ-specific inhibitors. At least in theory, this approach should also affect the mTOR-independent effects of enhanced PI3Kδ activity, such as FOXO1 phosphorylation and inactivation, with potentially important consequences on immune homeostasis (Table I). Two clinical trials based on use of PI3Kδ inhibitors are currently underway, one with oral administration of leniolisib () and the other with inhaled nemiralisib (). Initial results of the first trial, with dose-escalating administration of leniolisib over a period of 12 weeks, have been published recently; the drug was well tolerated, and reduction of lymphadenopathy and splenomegaly and improvement of cytopenias were observed at the end of the treatment.⁹ As to the second trial, it will be interesting to see whether use of inhaled nemiralisib can help reduce the progression and severity of lung disease in patients with APDS.

Although targeted therapy with mTOR inhibitors and selective PI3Kδ inhibitors might be beneficial in patients with APDS, it is unlikely that use of these drugs will eliminate the need for alternative treatment, including HSCT, in particular in patients with treatment-refractory disease or those with drug-related adverse events. Moreover, the long-term safety profile of mTOR and PI3Kδ inhibitors in patients with APDS has yet to be fully defined. Finally, treatment with sirolimus does not eliminate the risk of lymphoma,⁴ and use of PI3Kδ inhibitors might even promote genomic instability by enhancing the off target activity of activation-induced cytidine deaminase.¹⁰ For these reasons, it is important to define optimal strategies for HSCT in patients with APDS.

In this regard the 2 case series reported by Nademi et al⁵ and Okano et al¹ offer important elements for consideration. In the former study⁵ 9 of the 11 patients received transplantation from a matched related or unrelated donor; conditioning was mostly with fludarabine associated with treosulfan (n = 4), melphalan (n = 4) or busulfan (n = 1), whereas 2 patients received busulfan and cyclophosphamide. Nine patients received serotherapy, which consisted of alemtuzumab in 5 of them. With this regimen, robust (>90%) and sustained donor chimerism was observed in 9 patients. Nine of the 11 patients survived; the 2 deaths were due to idiopathic pulmonary syndrome and severe viral infections. Viral reactivation was frequently documented after transplantation; however, survivors attained significant improvement of clinical status and immune function.

On the other hand, in the series of 9 patients reported in this issue by Okano et al,¹ transplants were mostly (n = 6) from HLA-mismatched donors. Conditioning was predominantly (n = 7) with fludarabine and melphalan and associated with low-dose total body irradiation (TBI) in some of them. Eight patients received serotherapy, in all cases with antithymocyte globulin. Primary or secondary graft failure was frequently observed (n = 4), requiring stem cell boosts in 1 patient and second transplantation in 2 other patients. Two patients died: one of them had been previously splenectomized and had fulminant sepsis after transplantation after achieving mixed chimerism, and the other death was observed in a patient with pre-existing veno-occlusive disease who had systemic thrombotic microangiopathy after transplantation. Hemo-phagocytic syndrome at the time of engraftment was observed in 2 patients. Survivors experienced improvement of the clinical manifestations, and none of them required immunoglobulin replacement therapy by day 100.

The survival rate in the 2 series was comparable (9/11 and 7/9, respectively), and in neither was severe graft-versus-host disease recorded. Importantly, significant improvement of humoral immunity was consistently observed. Transplantation from matched donors, use of more intense conditioning, and inclusion of alemtuzumab might have improved engraftment in the study by Nademi et al.⁵ However, such an approach might have also facilitated the higher occurrence of viral reactivation.

Administration of mTOR and PI3Kδ inhibitors might represent a useful bridge therapy to control immune dysregulation and bring patients to transplantation in better clinical status. Inclusion of thiotepa in the conditioning regimen might also prove beneficial. Identifying pre-existing complications associated with increased transplant-related mortality will be very important. Eventually, there is a need for natural history studies that might inform clinical trials aimed at defining optimal management (including HSCT) of patients with APDS. The ESID-APDS Registry in Europe and development of a similar study from the Primary Immune Deficiency Treatment Consortium in North America should help reach these goals.