ALBERT

Description: Introduction: Recessive dystrophic epidermolysis bullosa (RDEB) is a severe, life-limiting systemic genodermatosis, characterized by COL7A1 mutation(s) yielding inadequate type VII collagen to maintain the integrity of the cutaneous basement and mucosal membranes. The mainstay of therapy for RDEB is supportive care to minimize the daily morbidity of blistering/scarring, pain/pruritis, systemic inflammation, and high metabolic demand. Leveraging the pluripotency of bone marrow cells, investigations of BMT as a systemic therapy for RDEB showed promise in pre-clinical murine (Tolar J, et al., Blood 2009, 113(5): 1167-74) as well as early phase human clinical trials utilizing fully myeloablative conditioning (Wagner J, et al., NEJM 2010, 363(7): 629-39). However, regimen-rated toxicity limited dampened enthusiasm for the clinical approach. Methods: In an effort to modulate disease activity with decreased toxicity, we investigated the safety of and preliminary responses to BMT with a reduced-intensity conditioning regimen (rabbit anti-thymocyte globulin 2.5 mg/kg with a methylprednisolone taper, cyclophosphamide 28 mg/kg, fludarabine 150 mg/m2, and low dose total body irradiation 300-400 cGy). Graft-versus-host disease (GvHD) prophylaxis included post-transplant cyclophosphamide (PTCy, 50 mg/kg recipient weight/dose on days +3 and 4) and mycophenylate mofetil from day +5 to 35. All except HLA-matched related donor recipients received tacrolimus from day +5 until tapered at day +100. Immunomodulatory donor-derived mesenchymal stromal cells (MSCs, 2 x 106 cells/kg recipient weight/dose) were infused at 60, 100 and 180 days post-BMT. Skin biopsies, medical photography, and dynamic assessments of RDEB disease activity by providers and parents were completed at baseline, day +100, +180 and 1 year post-BMT. Results: Ten RDEB patients were transplanted at a median age of 9.9 years (range 1.8 to 22.1), with a median follow-up of 16 months (1 year evaluations available for only 4 of 10). Donors were haploidentical related (n=6), HLA-matched related (n=3), and HLA-matched unrelated (n=1). BMT complications included graft failure in 3 patients (2 elected to pursue a 2nd PTCy BMT), veno-occlusive disease in 2, posterior reversible encephalopathy in 1, and chronic GvHD in 1, the latter deceased. Infectious complications in the first 100 days were limited to 3 viremia, 7 bacteremia, and 0 fungemia episodes. No serious adverse events were observed with MSC infusions. In the 9 ultimately engrafted patients, median donor chimerism at day +180 was 100% in both myeloid and lymphoid fractions of peripheral blood and 27% in skin. Skin biopsies at day +180 show stable (n=7) to improved (n=2) type VII collagen protein expression by immunofluorescence (IF) and gain of anchoring fibril components by transmission electron microscopy in 4 of 9 patients. Early clinical response included a trend toward reduced body surface area of blisters/erosions from a median of 45.5% at baseline to 27.5% at day +100 (p=0.05), with parental measures indicating stable quality of life. Select medical photography and skin biopsy results are shown for Patient 1 [IF: 40x merged dapi (blue) and C7 collagen antibody (red); immunoelectron microscopy with C7 collagen-directed immunostain (black)]. Conclusions: BMT using reduced-intensity conditioning, PTCy and donor-derived MSC infusion for RDEB was largely well tolerated with low rates of GvHD and death from regimen-related toxicity in 1 of 10 patients undergoing 12 total BMTs. Early follow-up suggests this treatment modulates RDEB disease activity but requires longer follow-up for evaluation of efficacy. Importantly, the PTCy BMT platform provides a means of attaining immunotolerance for future donor-derived cellular grafts. Figure Figure. Disclosures Wagner: Magenta Therapeutics: Consultancy, Research Funding; Novartis: Research Funding.

Description: Notch signaling is a critical regulator of T cell effector functions during acute graft-versus-host disease (GVHD). Pan-Notch inhibition in donor-derived T cells or systemic antibody-mediated blockade of Delta-like1 (Dll1) and Delta-like4 (Dll4) Notch ligands results in near-complete protection from acute GVHD in mouse models of allogeneic bone marrow transplantation. Notch-deprived alloreactive T cells proliferate and accumulate in vivo, but produce dramatically reduced levels of the proinflammatory cytokines IFNγ, TNFα and interleukin-2 (IL-2) (Zhang et al., Blood 2011; Sandy et al., J Immunol 2013; Tran et al., J Clin Invest 2013). In this study, we sought to: 1) determine the kinetic requirements for Notch signaling in the pathogenesis of acute GVHD; 2) identify the essential cellular compartment that delivers Dll1 and/or Dll4 ligands to incoming alloreactive T cells. In the B6 anti-BALB/c major histocompatibility complex-mismatched model, a single dose of Dll1 and Dll4 blocking antibodies at the time of transplantation abolished alloreactive T cell production of IFNγ, TNFα, and IL-2, increased regulatory T cell numbers (as assessed at day 10), and conferred long-term protection from GVHD. Conversely, delaying antibody administration by only two days after transplantation resulted in persistent T cell cytokine production, no changes in regulatory T cell numbers, and loss of long-term protection from GVHD. These findings identify a critical early window of Notch activity that promotes the pathogenesis of acute GVHD. To identify the dominant cellular source of Dll1 and Dll4, we assessed the impact of Cre-mediated Dll1 and Dll4 inactivation within host hematopoietic, donor hematopoietic, or host non-hematopoietic tissues. Bone marrow chimeras that lacked Dll1 and Dll4 solely within the host hematopoietic system were generated from poly(I:C)-induced Mx1-Cre;Dll1fl/fl;Dll4fl/fl donor mice. Both donor chimerism and Cre-mediated excision efficiency were 〉97%. Unlike systemic Dll1/4 blockade, Dll1 and Dll4 inactivation within the host hematopoietic system failed to decrease GVHD mortality or severity. Likewise, Mx1-Cre-mediated deletion of Dll1 and Dll4 within the donor hematopoietic system had minimal effects on T cell proinflammatory cytokines. In contrast, Ccl19-Cre-mediated Dll1 and Dll4 inactivation within host stromal cells profoundly impaired donor T cell production of IFNγ, TNFα, and IL-2, and resulted in long-term protection from GVHD. Lineage tracing in Ccl19-Cre x ROSA26-YFP mice revealed Cre activity within a small subset of CD45-negative lymph node and spleen stromal cells, but not in professional hematopoietic antigen-presenting cells. These data suggest that a specialized subset of non-hematopoietic stromal cells delivers an early pulse of Notch signaling to alloreactive T cells during acute GVHD. To our knowledge, these results provide the first in vivo evidence for non-motile secondary lymphoid-resident stromal cells as critical drivers of T cell-mediated immune pathology, with a central role for Notch signaling in this process. Transient interference with Notch ligand function or with their expression by the stromal cell niche in the peri-transplant period could serve as a novel therapeutic strategy for GVHD. Disclosures Yan: Genentech: Employment, Equity Ownership. Siebel:Genentech: Employment, Equity Ownership.

Description: Background: Immune mediated cytopenias (IMC) - isolated or combined hemolytic anemia, thrombocytopenia, and neutropenia - are increasingly recognized as potentially serious complications following allogeneic hematopoietic cell transplantation (HCT) for non-malignant disorders (NMD). However, IMC incidence, severity, response to therapy, and associated risk factors are not well defined. Methods: Pediatric patients undergoing HCT for NMD between 2010 and 2017 at a single institution were identified, excluding those experiencing graft failure or death prior to day +100. Demographics and HCT characteristics were obtained from a prospectively-maintained, institutional database. A retrospective chart review identified cases of IMC with positive direct antiglobulin test, anti-platelet antibody, and/or anti-granulocyte antibody; and documented treatment(s) and response. Composite IMC severity, incorporating peak acuity and response to therapy, was defined as mild (outpatient management with supportive care or a single course of therapy), moderate (outpatient management with 2 or more courses of therapy), or severe (inpatient management with 2 or more courses of therapy). Factors hypothesized to be associated with IMC included younger age, reduced intensity conditioning, mixed or declining donor chimerism, and inherited metabolic disorder as indication for HCT. For the first year after transplant, peripheral blood donor chimerism, B-cell reconstitution, and IgG and IgM levels were obtained. Results: A total of 271 NMD HCT recipients were included, with 53 (20%) exhibiting IMC, 20 (38% of cases) with multiple cell lines affected, diagnosed at a median of 131 days post-HCT (range, 30-264 days). IMC disease was mild, moderate, or severe in 56%, 32%, and 11%, respectively. For those with moderate or severe IMC, all received steroid therapy, with a response rate of 13%. Eighty-seven percent received additional therapy including IVIG, rituximab, bortezomib, and/or vincristine. Resolution of moderate IMC required an average of 2 individualized treatment courses, while severe IMC responded after an average of 6. Notably, no patients died of IMC. Univariate analysis revealed statistically significant association between IMC and younger age at HCT (median 3.1 vs. 7.1 years, p

Description: Allogeneic bone marrow transplantation (allo-BMT) is limited by graft-versus-host disease (GVHD). We previously reported an essential role for Notch signaling in alloreactive T cells mediating GVHD after allo-BMT, demonstrating a profound decrease in GVHD incidence and severity with genetic Notch inhibition in donor T cells or systemic antibody-mediated blockade of Delta-like1 (Dll1) and Delta-like4 (Dll4) Notch ligands (Zhang et al., Blood 117(1), 2011; Sandy et al., J Immunol 190(11), 2013; Tran et al., JCI123(4), 2013). However, the cellular source of these critical Notch ligands remains unknown. While host hematopoietic antigen-presenting cells (APCs) seem a likely source, recent evidence indicates that these cells are not solely responsible for donor T cell activation in GVHD. We considered three alternative sources of Notch ligands: donor-derived hematopoietic cells, host hematopoietic APCs surviving lethal irradiation, and host non-hematopoietic radioresistant cells. To test these possibilities, we used complementary genetic and biochemical approaches to inactivate Dll1/Dll4 in specific compartments, or to provide ubiquitous systemic blockade of these ligands. Bone marrow (BM) chimeras were created by transplanting BM from poly(I:C)-induced Mx-Cre+ x Dll1f/fDll4f/f B6-CD45.2 mice into lethally irradiated B6-CD45.1 mice, generating chimeric mice without Dll1 and Dll4 only in the host hematopoietic compartment. After 14 weeks of reconstitution, CD45.2 → CD45.1 turnover was near complete in hematopoietic progenitors and professional APCs. Dll1/Dll4 excision was 〉99% by qPCR. WT (wild-type) BM chimeras subjected to MHC-mismatched allo-BMT (BALB/c → [Mx-Cre- x Dll1f/fDll4f/f (B6-CD45.2) → B6-SJL CD45.1]) had evidence of severe aGVHD and poor survival, as expected. Interestingly, BM chimera mice lacking Dll1 and Dll4 expression in the host hematopoietic compartment had equally severe GVHD and impaired survival following allo-BMT (BALB/c → [Mx-Cre+ x Dll1f/fDll4f/f (B6-CD45.2) → B6-SJL CD45.1]). In contrast, both WT and Dll1/Dll4-deficient BM chimeras treated with anti-Dll1/Dll4 antibodies (i.p. x4 over days 0-10 of transplant) had markedly decreased GVHD and improved survival (log rank Χ2=30.6, p

Description: Prolonged neutropenia increases risk for lethal invasive fungal infection (IFI) such as Rhizopus. Isavuconazonium sulfate is a new triazole that lacks pediatric dosing recommendations. Clinical courses of 4 pediatric patients with IFI in the peri-allogeneic hematopoietic cell transplantation (alloHCT) period between 2015 and 2017 were reviewed, including previously unreported pharmacokinetic and safety data. IFI included Rhizopus and presumed fungal meningitis. Isavuconazonium sulfate was initiated with a loading dose followed by daily dosing, adjusted to a goal trough 〉3 ug/mL based on adult literature. This target was achieved at a median of 7 days, demonstrating varying rates of metabolism. Renal insufficiency, electrolyte disturbances, and transaminitis were noted, though attribution was confounded by other alloHCT complications. 1 patient survived infection-free to hospital discharge and 1 of 3 deceased patients had evidence of unresolved IFI (Case 2). Case 2 was subtherapeutic for 39% of the duration of treatment, compared to others at an average of 29%, suggesting this target trough to be clinically relevant. We recommend initiation of isavuconazonium sulfate at 10 mg/kg with a max dose of 372 mg. A loading dose of 10 mg/kg is utilized every 8 hours for 6 doses followed by 10 mg/kg dosing every 24 hours. Monitoring must continue beyond steady state. If early monitoring is not possible, we recommend a first drug level at week 3. If dose increases are required, a partial reload has been more successful instead of increasing daily doses. Further larger studies are needed to demonstrate optimum dosing in pediatric patients. Figure 1 Disclosures No relevant conflicts of interest to declare.

Description: Background Hemophagocytic lymphohistiocytosis (HLH) is an increasingly recognized life-threatening syndrome characterized by excessive inflammation and immune dysregulation. However, little is known about HLH presenting in neonates. Methods To describe the phenotype, etiology, and efficacy of therapy for infants presenting within 28 days of birth and diagnosed with HLH prior to 90 days of life, a retrospective multi-center study was conducted identifying 13 previously unreported cases from 8 centers. A systematic literature review identified 58 additional cases that were sufficiently similar to allow pooling for meta-analysis (n=71). Results Presenting symptoms and laboratory findings of HLH were non-specific and indistinguishable from other neonatal critical illnesses. More than 85% of patients demonstrated hepatomegaly, splenomegaly, hyperferritinemia, transaminitis, hypoalbuminemia, and severe thrombocytopenia. Notably, liver dysfunction appears more common in neonatal HLH compared to published reviews of HLH cases across the age spectrum. Genetic testing was pursued in 60% of the cohort with pathogenic variants identified in a third of patients (PRF1 in 15, UNC13D in 8). Ten additional cases occurred with a family history of HLH, for a total of 46% (33/71) of cases of presumed familial HLH. Seventeen cases (24%) represented secondary HLH with an identified trigger for immune dysregulation (e.g. infection or non-HLH disease) in the absence of an HLH-associated genetic mutation or family history. Twenty-one (30%) neonatal HLH cases were of unknown etiology. Two-year survival from presentation was low at 27%, with a trend toward superior 2-year overall survival for those with secondary HLH at 65% (unknown etiology 21%, familial HLH 18%, p=0.06; Figure). Ten patients (n=6 familial HLH, n=4 unknown etiology) underwent allogeneic hematopoietic cell transplantation (HCT) at a median of 4.8 months of age. HCT was associated with improved 2-year overall survival (OS) from HLH diagnosis (78% with HCT versus 11% without HCT, p