CD3⁺/CD19⁺ depleted matched and mismatched unrelated donor hematopoietic stem cell transplant with targeted T cell addback is associated with excellent outcomes in pediatric patients with non-malignant hematologic disorders

Abstract

Unrelated donor (URD) hematopoietic stem cell transplantation (HSCT) is increasingly being utilized to cure non-malignant hematologic diseases (NMHD) in patients who lack HLA matched related donors (MRD). Both graft rejection and graft vs host disease (GVHD) remain major barriers to safe and effective transplant for these patients requiring URDs. Partial T cell depletion combined with peripheral stem cell transplantation (pTCD-PSCT), has the potential advantages of providing a high stem cell dose to facilitate rapid engraftment, maintaining cells that may facilitate engraftment, and decreasing GVHD risk compared to T replete HSCT. Here, we report a single institution, retrospective experience of URD pTCD-PSCT for pediatric patients with NMHD. From 2014 to 2017, 12 pediatric patients with transfusion-dependent NMHD underwent matched (MUD) or mismatched (MMUD) unrelated donor pTCD HSCT in our center using disease-specific conditioning. Donor peripheral stem cells underwent CD3⁺ T cell and CD19⁺ B cell depletion using CliniMACS, followed by a targeted addback of 1 × 10⁵ CD3⁺ T cells/kg to the graft prior to infusion. All 12 patients demonstrated rapid trilinear engraftment. At a median follow-up of 740 days (range 279–1466), all patients are alive with over 92% total peripheral blood donor chimerism, and without transfusion dependence or recurrence of their underlying hematologic disease. Immune reconstitution was rapid and comparable to T replete HSCT. No patients developed severe acute GVHD (Grade III-IV) or chronic extensive GVHD, and all patients have discontinued systemic immune suppression. Viral reactivations were common, but no patient developed symptoms of life-threatening infectious disease. Our data indicate that MUD and MMUD pTCD-PSCT are safe and effective approaches that enable rapid engraftment and immune reconstitution, prevent severe GVHD, and expand availability of HSCT to any patients with NMHD who have closely matched unrelated donors.

Introduction

Matched related donor bone marrow transplantation (MRD-BMT) has long been a standard treatment option for pediatric patients with bone marrow failure syndromes (BMF) or hemoglobinopathies (1–4). Historically, in cases where a MRD was not identified, first line therapy typically consisted of medical management with immune suppression, growth factors or chronic transfusion protocols (5–8). Because of the historically high rates of morbidity and treatment failure associated with alternative donor hematopoietic stem cell transplant (HSCT), these approaches were reserved for patients that failed medical therapy or developed significant morbidities.

Recent advances in alternative donor HSCT strategies for non-malignant hematologic disorders (NMHD), including matched (MUD) and mismatched (MMUD) unrelated donor HSCT and haploidentical related donor HSCT, have now led to survival and engraftment outcomes that in many cases are comparable with MRD-BMT (9–14). However, incidence of severe acute and chronic graft versus host disease (GVHD) when T-replete BM grafts are used for MUD or MMUD HSCT remain significant. Thus, novel MMUD or MUD HSCT approaches geared toward trying to minimize rates of GVHD while still preventing graft rejection in multiply transfused patients with NMHD are currently under development (15–16).

We have used various methods of Ex vivo partial T cell depletion in unrelated donor HSCT for nearly two decades (17–19). Partial T cell depletion (pTCD) enables reduction in rates of severe GVHD, while avoiding the higher rates of graft rejection and infections seen with complete T cell depletion strategies (20–21). These pTCD strategies are particularly beneficial in reducing risk of GVHD in patients in whom a fully matched unrelated donor cannot be identified and consequently a 7/8 or 9/10 match is the best available donor, a scenario common for patients of Non-European ancestry who lack sibling donors (22).

Combining pTCD with use of mobilized peripheral stem cells as a graft source has the added benefit of providing a much higher donor CD34⁺ stem cell dose than present in most BM grafts, which in turn improves engraftment efficiency (23). For patients with severe aplastic anemia, T replete PSCT versus BMT is associated with higher rates of severe acute GVHD and consequently mortality (24). However, combining PSCT with pTCD would be expected to mitigate the GVHD risk associated with PSCT. Whether higher CD34⁺ doses and more rapid engraftment in MUD and MMUD pTCD-PSCT will have a positive impact on survival and morbidity rates for patients with NMHD compared to T replete BMT has yet to be studied (25).

Here, we present a retrospective analysis of 12 pediatric patients with non-malignant hematologic disorders that received MUD or MMUD pTCD-PSCT through an expanded access study (ExpMACS) of CD3⁺ T cell/CD19⁺ B cell depletion using CliniMACS. A targeted dose of 1 × 10⁵ CD3⁺ T cells/kg was added back to the graft prior to infusion, a dose selected based on our prior CHP 834 protocol (NCT00579124) for patients with hematologic malignancies (26). Our findings indicate that MUD and MMUD pTCD-PSCT for patients with NMHD is an effective strategy for achieving rapid and sustained donor engraftment while preventing severe and chronic GVHD.

Methods

Patients and Study Information

For this analysis, we included patients who received pTCD-PSCT as a first allograft for NMHD at the Children’s Hospital of Philadelphia (CHOP) between January 2014 and December 2017 using an expanded access protocol (ExpMACs, >NCT02356653) for CD3⁺/CD19⁺ depletion approved by the CHOP Internal Review Board (IRB) and by the United States Food and Drug Administration (FDA). Patients with severe aplastic anemia (SAA) without concurrent paroxysmal nocturnal hemoglobinuria (PNH) were considered eligible for pTCD-PSCT if they had failed one or more lines of immune suppression therapy (IST). Patients with SAA and either concurrent hemolytic PNH disease or with large PNH clones (> 30% in granulocyte lineage) were eligible for pTCD-PSCt without prior IST. Patients with Diamond-Blackfan Anemia (DBA) and Beta Thalassemia Major (BTM) were eligible for pTCD-PSCT if they exhibited persistent transfusion-dependence. The cohort also includes a patient with congenital transfusion-dependent pancytopenia, for which a genetic etiology could not be identified despite extensive testing including whole exome sequencing. Patients were excluded from this analysis if they had cytopenias associated with underlying multisystem metabolic disorders, including one patient with MIRAGE syndrome that is reported elsewhere (27). All patients and/or their legal guardians signed written informed consent. Organ criteria for eligibility included cardiac shortening fraction ≥27%, creatinine ≤1.5x upper limit of normal, alanine and aspartate aminotransferases <5x upper limit of normal, and no evidence of active untreated infection.

HLA typing, Donor Selection, and Stem Cell Collection

HLA typing was performed using the the Holotype HLA kit by Omixon (Budapest, Hungary), a paired-end next generation sequencing (NGS) approach using the Illumina MiSeq platform. Sequence alignment and HLA allele assignment were performed using Target™ (Omixon; Budapest, Hungary) and NGSengine™ (GenDx; Utrecht, Netherlands) software. Unrelated donor searches were performed through the National Marrow Donor Program (NMDP), optimized to match at HLA-A, -B, -C, -DRB1, and -DQB1. 10/10 MUD were first preference, followed by one-antigen mismatched donors, provided that the mismatch was not at -DRB1. 8/10 MMUD’s were accepted, provided that one of the mismatches was at -DQB1. Donor PSCs were mobilized with G-CSF according to collection center and NMDP guidelines.

Partial T Cell Depletion

Depletion of CD3⁺ and CD19⁺ target cells was performed in the CHOP Good Manufacturing Practice (GMP)-compliant Stem Cell Processing Laboratory. CD19⁺ B cell depletion was included in this protocol to prevent increased risk of EBV reactivation associated with pTCD. Donor hematopoietic progenitor cells collected by apheresis [HPC(A)] were incubated with CD3- and CD19-depleting antibody-conjugated iron dextran beads and passed through the ClinMACS device in a closed system according to manufacturer protocol. Post-depletion cells were suspended in 0.9% sodium chloride with 25% human albumin for infusion. Residual doses of CD3⁺ and CD19⁺ cells were determined by flow cytometry, followed by an addback of CD3⁺ cells to achieve the target CD3⁺ cell dose of 1 × 10⁵/kg. Standard release testing included viability assay, Gram stain with sterility testing, and endotoxin quantitation.

Transplant Procedure

Patients received previously established disease specific conditioning regimens (1,2,28,29) as shown in Table 2. Patients with SAA without PNH received thymoglobulin 3 mg/kg/day x 3 days (starting day ranged from day −11 to day −7); fludarabine 30 mg/m2/day x 5 days (starting day −7 or −6), cyclophosphamide 50 mg/kg/day x 2 days (starting day −5 or −4), and 200 cGy total body irradiation (TBI) in a single fraction on Day −7. Patients with SAA and PNH received an identical regimen, except that TBI dose was increased to 300 cGy, given in 2 divided fractions on Day −7. Patients with DBA or BMT received busulfan x 4 days at an initial dose of 3.2 mg/kg/day (either as daily or q6h dosing) with pharmacokinetics-guided dose adjustments to maintain optimal steady state concentration; thymoglobulin 3 mg/kg/day x 3 days (starting day −10 or −9); fludarabine 30 mg/m2/day x 5 days (starting day −7 or −6), and thiotepa 5 mg/kg/day x 2 days (starting day −4 or −3). Due to young age and weight < 10 kg, the patient with congenital BMF received thymoglobulin 3 mg/kg/day x 3 days (starting day −9), fludarabine 1 mg/kg/day x 5 days (starting day −6), thiotepa 5 mg/kg/day x 2 days (starting day −5), and cyclophosphamide 50 mg/kg/day x 2 days (starting day −3). Unlike TCRαβ T cell depletion which entirely removes this GVHD causing T cell subset, the CD3⁺ addback we used did contain TCRαβ T cells. Therefore, GVHD and graft rejection prophylaxis was used per institutional standard, consisting of calcineurin inhibitor (cyclosporine infusion transitioned to oral tacrolimus) in all patients for a minimum of 3 months, and mycofenalate mofetil (MMF) in 11 of 12 patients until Day 45. In patients with acquired aplastic anemia tacrolimus was continued for a minimum of 6 months. Growth factors were not routinely used.

Table 2:

Graft Characteristics and Transplant Outcomes

					Engraftment (days)*		Days of Followup	GVHD (outcome)
Subject#	Age at SCT, Diagnosis	Conditioning	CD34+ x106/kg	CD3+x105/kg	Neutrophil	Platelet	(clinicalstatus)	Acute	Chronic	Organ Toxicity (outcome)
1	2y, SAA	rATG, Flu, Cy, 200 cGy TBI	19.3	1	19	13	1466 (alive, engrafte d)	-	Limited, skin (resolved)	-
2	9y, SAA	rATG, Flu, Cy 200 cGy TBI	6.2	1	14	15	1077 (alive, engrafte d)	-	-	TA-TMA (self-limited, resolved), Hyperthyroidism (treatment ongoing)
3	16y, SAA	rATG, Flu, Cy 200 cGy TBI	2.43	1	18	16	1465 (alive, engrafte d)	-	-	-
4	16y, SAA	rATG, Flu, Cy 200 cGy TBI	7.56	1	13	16	475 (alive, engrafte d)	Stage 3 Skin (Gra de II)	-	-
5	15y, SAA	rATG, Flu, Cy 200 cGy TBI	5.3	1	15	15	322 (alive, engrafted)	-	-	-
6	14y, SAA/PN H	rATG, Flu, Cy 300 cGy TBI	12.9	1	13	15	748 (alive, engrafte d)	-	-	-
7	22y, SAA/PN H	rATG, Flu, Cy 300 cGy TBI	5.77	1	13	14	746 (alive, engrafte d)	-	Limited, skin (resolved)	Bell’s Palsy (resolved with short steroid course)
8	7y, DBA	rATG, Bu, Flu, TT	16.8	1	11#	31	1111 (alive, engrafte d)	-	-	SOS/VOD, (resolved with defibrotide)
9	5y, DBA	rATG, Bu, Flu, TT	13.2	1	13	16	734 (alive, engrafte d)	-	-	-
10	12y, Thal	rATG, Bu, Flu, TT	14.9	1	11	14	306 (alive, engrafte d)	-	-	-
11	9y, Thal	rATG, Bu, Flu, TT	11.1	1	18	17	279 (alive, engrafte d)	-	-	-
12	0.7y, Congenit al BMF	rATG, Cy, Flu, TT	14.2	1.98^	20	15	280 (alive, engrafte d)	-	-	-
		Median (Range):	12.0 (2.4–19.3)	1 (1–1.98)	13.5 (11–20)	15 (13–31)	740 (279–1466)

All data are represented as median (total range).

^*Engraftment defined per CIBMTR criteria.

^#Only subject to receive filgrastim post-SCT.

^{^}CD3⁺ cell number for subject 12 included a 1 × 10⁵ CD3⁺ cell/kg addback plus a cell population with low CD3 expression of unknown significance. rATG, rabbit anti-thymocyte globulin (total dose 9 mg/kg); Flu, fludarabine (total dose 150 mg/ml except patient 12 who received 5 mg/kg TBI based on young age); Cy, cyclophosphamide (total dose 100 mg/kg); Bu, busulfan (3.2 mg/kg/day starting dose, then PK adjusted for total of 4 days); TT, thiotepa (Total dose 10 mg/kg). GVHD, graft versus host disease; TA-TMA, transplant-associated thrombotic microangiopathy; SOS, sinusoidal obstruction syndrome; VOD, veno-occlusive disease.

All patients were transplanted in HEPA filtered rooms. Supportive care included pneumocystis and antifungal prophylaxis, as well as prophylactic acyclovir if seropositive for herpes simplex or varicella virus. Weekly monitoring for CMV, adenovirus, and Epstein-Barr virus by polymerase chain reaction was performed until Day +100. Patients with CMV positive serology and negative serology donors were considered at high risk for CMV reactivation and received prophylactic foscarnet or valgancyclovir until day +100. Standard consensus criteria were used for grading acute and chronic GVHD (30).

Engraftment and Immune Reconstitution Monitoring

Time to neutrophil and platelet engraftment was based on Center for International Blood and Marrow Transplant Research (CIBMTR) criteria. Donor chimerism was evaluated at a minimum at 30, 120, 240, 365, and 720 days post-SCT, with some patients receiving additional assessments per clinician preference. Immune reconstitution analysis was performed per institutional standard on days 60, 120, 240, 365 and 720 post-transplant, with additional testing performed per clinician preference.

Results

Patient, Disease, and Donor Characteristics

Patient and disease characteristics are outlined in Table 1. For the 12 subject cohort receiving pTCD-PSCT, the median age at HSCT was 11 years (range 0.7–22.9), with the median time from diagnosis to HSCT being 4.1 years (range 0.4–13.5). Diagnoses included SAA (5), SAA + PNH (2), DBA (2), BTM (2) and congenital BMF (1). All but one had received greater than 10 lifetime packed red blood cell (pRBC) transfusions. All patients with SAA alone had received one or more prior cycles of IST. One patient with hemolytic PNH treated with eculizumab had concurrent transfusion dependent thrombocytopenia and severe neutropenia requiring G-CSF, and was classified as SAA + PNH. The other patient classified as SAA + PNH met Camitta criteria for SAA and had a PNH clone in granulocytes > 30%, resulting in our recommendation to proceed to pTCD-PSCT without prior IST. Both patients with DBA had failed prior steroid therapy, and all patients with SAA alone had failed immune suppression therapy. Sixty-seven percent of patients were classified as an ethnicity other than white, non-Hispanic. The majority of subjects (67%) had MMUD (8/10 or 9/10) as the best available donor options. HLA mismatches were HLA-A (5), HLA-B (1), HLA-DQB1 (1) and HLA-B and -DQB1 (1).

Table 1:

Patient Demographics and Disease/Donor Characteristics

Total Cohort	N = 12
Age at transplant (years), median (range)	11.0 (0.7–22.9)
Gender, # of female (%)	7 (58.3)
Ethnicity, n (%) White, Non-Hispanic African-American Hispanic Middle-Eastern Asian Multiple	4 (33.3) 3 (25.0) 1 (8.3) 1 (16.7) 1 (8.3) 1 (8.3)
Disease, n (%) SAA SAA + PNH (>30% clone) DBA Thalassemia Major Congenital BMF	5 (41.7) 2 (16.7) 2 (16.7) 2 (16.7) 1 (8.3)
Time from diagnosis to SCT (years), median (range) Greater than 10 lifetime PRBC transfusions pre-SCT, n (%) Prior Treatment for SAA or SAA+PNH hATG/CSA x 1 only hATG/CSA x 2 only hATG/CSA x 2, HDCy, Eltrombopag Eculizumab None	4.1 (0.4 to 13.5) 11 (91.7) n=7 3 1 1 1 1
Allele Level URD Donor Match, n (%) 10/10 9/10	4 (33.3) 7 (58.3) 1 (8.3)
8/10

All data are represented as number (%) or median (range). SAA, severe aplastic anemia; SAA+PNH, severe aplastic anemia and paroxysmal nocturnal hemoglobinuria defined as a >30% PNH clone in the granulocyte fraction; DBA, Diamond-Blackfan Anemia; BMF, bone marrow failure; SCT, stem cell transplantation, PRBC, packed red blood cells; hATG, horse (equine) anti-thymocyte globulin; CSA, cyclosporine; HDCy, high dose cyclophosphamide (200 mg/kg total dose); URD, unrelated donor

Graft Characteristics and Engraftment

As shown in Table 2, the median donor CD34⁺ stem cell dose received was 12 × 10⁶ cells/kg (range 2.4–19.3). Of note, the subject who received the lowest CD34⁺ cell dose per kg ( 2.4 × 10⁶) was an adolescent with one of the highest weights of the cohort, but was able to achieve trilinear engraftment at a rate comparable to other patients in the cohort. The target CD3⁺ T cell addback of 1 × 10⁵ cells/kg was achieved in 11 of 12 patients. Subject 12 had a CD3^lo expressing cell population remaining in the graft after initial processing comprising 0.98 × 10⁵/kg, and the clinical decision was made to add back an additional 1 × 10⁵ CD3⁺ cells/kg to ensure adequate T cell content.

All subjects exhibited rapid trilinear engraftment. Median time to neutrophil engraftment was 13.5 days (11–20), despite only 1 of 12 patients receiving post-transplant granulocyte colony stimulating factor (G-CSF). Median time to platelet engraftment was 15 days with 11 of 12 patients exhibiting platelet engraftment by Day +17. Due to defibrotide treatment and a consequent need to maintain platelet count at a higher threshold, subject #8 did not meet CIBMTR criteria for platelet engraftment until Day +31.

Survival and Donor Chimerism

At a median follow-up of 740 days (range 279 to 1466), overall and event-free survival are 100%. All 12 subjects remain in excellent health with performance status of 100% and remain stably engrafted. No patients experienced either primary or secondary graft failure. All patients are transfusion independent with hemoglobin > 10 g/dL, Platelet > 100 × 10³/μL, and absolute neutrophil count > 1000/μL. At last follow-up, all patients exhibited peripheral blood total donor chimerism of 93% or greater (Figure 1). For the 10 patients in whom sorted myeloid chimerism was assessed, all demonstrated 100% myeloid chimerism, beginning with the first assessment at approximately day 30 and continuing through the most recent follow-up point. In contrast, T cell donor chimerism was as low as 6–10% during the first 60 days post-PSCT. T cell chimerism subsequently rose steadily in all patients out to 1 year post-PSCT. This low initial donor T cell chimerism occurred in the absence of any clinical evidence of immunologic graft rejection. Only 3 of 11 patients achieved complete (100%) T cell donor chimerism by last follow-up.

Figure 1: Donor Chimerism after partial CD3 depleted HSCT.

Total (top), myeloid (middle), and T cell (bottom) peripheral blood donor chimerism over the first 800 days after partial T cell depleted (pTCD) HSCT, assessed by cell sorting and short tandem repeat methodologies. Each line represents an individual subject, and each marker represents chimerism at a single time point of assessment. Total donor chimerism data was available for all 12 subjects, whereas myeloid and T cell chimerism data were available for 10 and 11 subjects, respectively.

GVHD and Immune Reconstitution

No subject developed severe grade III-IV acute GVHD or chronic extensive GVHD. One subject developed Grade II, stage III skin GVHD, yielding a rate of grade II-IV acute GVHD of 8.3%. In this case, symptoms resolved with a course of prednisone, and did not recur after prednisone was discontinued, which occurred at less than 4 months post-PSCT. Two other patients developed late-onset eczematous skin rashes primarily on the face and neck, consistent with chronic limited skin GVHD, giving a rate of chronic limited GVHD of 16.7%. Both required resumption of oral tacrolimus after initial taper but were able to stop tacrolimus by 16 months post-transplant. Neither has exhibited GVHD recurrence and both remain off immune suppression. Excluding these two subjects, all other patients discontinued the immune suppression used as GVHD prophylaxis at a median time of 198 days (range 59–340 days).

T cell immune reconstitution was rapid (Figure 2A). All 12 subjects demonstrated recovery of CD3⁺ T cell counts to greater than 500/μL by 8 months post-PSCT, with kinetics similar to that seen after T replete BMT. Eleven of 12 also demonstrated recovery of CD4⁺ T cell counts to greater than 200/μL by 8 months, with each demonstrating recovery of both naïve (CD45RA) and memory T cell (CD45RO) populations. The one subject with slow CD4 recovery, and specifically slow CD45RA recovery, was a patient with DBA who had a history of severe congenital heart disease, raising the possibility that the slow T cell recovery in this patient may have in part been linked to poor thymic function. B cell numbers (Figure 2B) and recovery of IgM and IgG production was robust in all patients, except for one subject (subject #1) who required 3 doses of rituximab for EBV reactivation, and continued to exhibit low IgM production by 8 months post-PSCT. This subject ultimately recovered B cell numbers and Ig production between 1 and 2 years post-PSCT.

Figure 2: Immune Reconstitution after partial CD3 Depleted HSCT.

Total CD3⁺ and CD3⁺CD4⁺ T cell recovery over the first 800 days after pTCD HSCT, as measured by white blood cell count and flow cytometry-based immunophenotyping (A, top panels). Each line represents an individual subject, with flexion points denoting assessment time points. Hatched horizontal lines represent standard thresholds above which reconstitution is generally thought to be adequate. Recovery of CD4+ T cells is further broken down into naïve CD45RA as well as memory CD45RO subsets (A, bottom panels). Recovery of total CD19+ B cells out to 800 days post-SCT (B) as well as functional recovery of IgM and IgG immune globulin production by 8 months post-HSCT (C) are also shown. Each panel shows data from all 12 subjects, with the exception of IgG, where data from 11 subjects is shown (one subject excluded as they remained on IVIg at 8 months post-SCT).

Viral Reactivation and Organ Toxicity

Of the 10 patients at risk for CMV reactivation (donor and/or recipient CMV serology positive), 7 developed reactivation detected by serum PCR (median of 35 days, range 11–42 days), though the maximal viral load was low in each case (less than or equal to log 3.66 for the 6 patients in whom testing was performed using the standardized IU/ml CMV quantification system). Notably, CMV reactivation requiring treatment dosing of antiviral agents occurred in only 1 of 4 subjects already on CMV prophylaxis and in 5 of 6 patients not receiving antiviral prophylaxis. No subjects developed symptoms of CMV disease, and CMV viremia resolved in all cases. Two subjects developed EBV reactivation, which resolved following rituximab (1 dose in subject #4, 3 weekly doses for subject #1). Three developed mild symptoms of BK cystitis, each of which were given one dose of cidofovir before symptoms resolved. Two subjects developed varicella zoster reactivation greater than 1 year post-PSCT, which manifested as shingles and was successfully treated in each case with acyclovir.

Major organ toxicities included: subject #8 who had DBA and a history of severe transfusional iron overload developed hepatic sinusoidal obstruction syndrome (veno-occlusive disease, VOD) that resolved with defibrotide treatment. Subject #2, who had SAA and over 5 years of preceding cyclosporine therapy, developed transplant-associated thrombotic microangiopathy (TA-TMA) that was managed with supportive care only. TA-TMA eventually resolved without requiring complement inhibition therapy. This subject later developed hyperthyroidism over 2 years out from PSCT, for which she continues to receive therapy. Subject #7 developed Bell’s Palsy more than one year removed from PSCT, which was successfully treated with one week of prednisone, and has not since recurred.

Discussion

This analysis demonstrates durable engraftment with limited GVHD for MUD or MMUD PSCT using CD3⁺/CD19⁺ depletion and targeted CD3⁺ addback for pediatric patients with NMHD. All patients received robust CD34⁺ donor stem cell doses and achieved rapid trilinear engraftment. Importantly, there were no cases of grade III-IV acute or extensive chronic GVHD. All demonstrated stable total donor chimerism above 92%, with 100% myeloid chimerism. Despite low initial T cell chimerism in many patients, there were no instances of graft rejection. This low initial T cell chimerism likely reflected initially low total T cell numbers, and both T cell numbers and T cell chimerism improved in temporal correlation typically 4–6 months post-transplant. Use of graft rejection prophylaxis with CNI/MMF may play a role in preventing graft rejection in this early period of low T cell chimerism.

Immune reconstitution was comparable to that seen with T replete BMT. Viral reactivation, and particularly CMV reactivation, was common suggesting that all patients at risk for CMV (serology positive patient and/or recipient) who receive pTCD-PSCT using this strategy may benefit from CMV prophylaxis All viral reactivations responded to appropriate therapy, and no patient developed life-threatening viral disease.

For pediatric patients with acquired BMF who lack matched sibling donors, unrelated donor pTCD-PSCT is just one of several emerging treatment options that are demonstrating promising efficacy, including T-replete MUD/MMUD-BMT, haploidentical related donor HSCT, and eltrombopag used in combination with immune suppression therapy (5, 9, 10, 33). In a recent study from Europe, 29 patients with SAA received T-replete unrelated donor BMT or PSCT, with an event-free survival rate of 92% and only one patient developing severe acute GVHD (9). It should be noted, however, that 82.8% of this cohort had a 10/10 MUD, reflecting a much higher likelihood of having a 10/10 MUD than was seen in the more ethnically diverse cohort we have presented, where 67% of subjects required a 1–2 antigen mismatched donor. In an analysis of 79 patients receiving MUD/MMUD T replete BMT from the BMT CTN 0301 study where 27% of the patients had a MMUD donor, disease-free survival was 89%, with a cumulative incidence of severe Grade III-IV acute GVHD of 8.9% and a 25% incidence of Grade II-IV GVHD (33). Our experience, in which we saw no patients with Grade III-IV acute GVHD and only 1 patient (8.3%) with Grade II acute GVHD albeit in a much smaller cohort, suggests that pTCD-PSCT may compare favorably to T-replete unrelated donor BMT, particularly for patients with MMUD.

Recent reports suggest that haploidentical HSCT using either post-transplant cyclophosphamide or ex vivo T cell depletion can be a highly effective strategy for patients with NMHD lacking matched sibling donors (10, 31, 34). However, these studies and our unpublished experience suggest that graft rejection remains a challenge with these approaches, particularly in pediatric patients with SAA. A recent report using TCRαβ⁺ T cell/CD19⁺ depletion in the haploidentical setting had a 50% (2/4) rate of immunologic graft rejection (31). While numbers of patients in both series are low, our approach using matched and mismatched unrelated donors with pTCD PSCT, with which we saw no graft rejection, may have an advantage over use of haploidentical donors with similar Ex vivo pTCD strategies, at least for patients with SAA. A critical advantage of haploidentical HSCT is that the majority of patients have available donors. Likewise, one of the advantages of our pTCD-PSCT approach is that pTCD enables patients with 1–2 antigen MMUD to have similar outcomes as MUD-PSCT in terms of survival, GVHD, and rejection risk. Thus, pTCD-PSCT enables expansion of the unrelated donor pool such that the majority of pediatric patients have a viable unrelated donor option that may carry a lower rejection risk than that seen with haploidentical HSCT.

Gene therapy approaches are now showing promise in providing disease-free survival for genetically-driven NMHD (35). To date, these strategies still require myeloablative conditioning, and have been limited to monogenic disorders such as BTM. Conditions such as DBA that are caused by mutations in over 20 genes (36), have thus far been less amenable to these approaches. If pTCD-PSCT can eliminate severe GVHD risk as indicated by our preliminary findings, MUD/MMUD pTCD-PSCT would provide an alternative to these gene therapy strategies with comparable safety and favorable efficacy depending on an individual patient’s genotype.

The success of this CD3⁺ T cell depletion strategy has led us to open a prospective clinical trial of a similar pTCD strategy, using TCRαβ/CD19 depletion for patients with bone marrow failure syndromes receiving MUD, MMUD, or haploidentical-related PSCT (>NCT03047746). In contrast to our partial CD3⁺ depletion strategy, TCRαβ depletion specifically targets GVHD-causing TCRαβ T cells and does not require an addback step, thus shortening processing time. The TCRαβ depletion strategy has been explored in Europe for haploidentical transplantation in both malignant and non-malignant diseases (31–32). However, its use in MUD and MMUD PSCT and its comparative effectiveness versus the partial CD3⁺ depletion strategy we have reported here have yet to be fully explored.

Highlights.

Study of partial CD3+ T Cell/CD19 depletion in closely matched URD PSCT for Non-malignant disorders.

URD-PSCT with pTCD resulted in 100% disease free survival with no graft rejection and no severe GVHD

URD pTCD PSCT is efficacious alternative to haploidentical HSCT for patients lacking matched donors