HLA-Haploidentical Hematopoietic Cell Transplantation for Treatment of Non-Malignant Diseases Using Nonmyeloablative Conditioning and Post-Transplant Cyclophosphamide

Abstract

Allogeneic hematopoietic cell transplant (HCT) is often the only curative therapy for patients with non-malignant diseases; however, many patients do not have a human leukocyte antigen (HLA)-matched donor. Historically, poor survival was seen following HLA-haploidentical HCT due to poor immune reconstitution, increased infections, graft-versus-host disease (GVHD), and graft failure. Encouraging results have been reported using a nonmyeloablative T-cell replete HLA-haploidentical transplant approach in patients with hematologic malignancies. Here we report the outcomes of 23 patients with various non-malignant diseases using a similar approach. Patients received HLA-haploidentical bone marrow (n=17) or G-CSF mobilized peripheral blood stem cell (n=6) grafts following conditioning with cyclophosphamide (CY) 50 mg/kg, fludarabine 150 mg/m², and 2 Gy or 4 Gy total body irradiation (TBI). Post-grafting immunosuppression consisted of CY, mycophenolate mofetil, tacrolimus, ± sirolimus. The median patient age at HCT was 10.8 years. Day 100 transplant-related mortality (TRM) was 0%. Two patients died at later time points, 1 from intracranial hemorrhage/disseminated fungal infection in the setting of graft failure and 1 from infection/GVHD. The estimated probabilities of grades II–IV and III–IV acute GVHD at day 100 and 2-year NIH-consensus chronic GVHD were 78%, 26% and 42%, respectively. With a median follow-up of 2.5 years, the 2-year overall and event-free rates of survival were 91% and 78%, respectively. These results are encouraging and demonstrate favorable disease specific lineage engraftment with low TRM in patients with non-malignant diseases using nonmyeloablative conditioning followed by T-replete HLA-haploidentical grafts. However, additional strategies are needed for GVHD prevention in order to make this a viable treatment approach for patients with non-malignant diseases.

INTRODUCTION

Allogeneic hematopoietic cell transplantation (HCT) offers curative therapy for patients with non-malignant diseases, including primary immunodeficiencies, bone marrow failure syndromes, hemophagocytic disorders, metabolic disorders and hemoglobinopathies. Historically, myeloablative regimens have been used; however, this has been associated with increased risk for morbidity and mortality due to comorbidities as a result of their underlying disease. Therefore, less intense regimens are required. However, many non-malignant diseases have increased barriers to engraftment that make it difficult to lower the intensity of the conditioning regimen. These barriers may include normocellular marrows, a history of being heavily transfused which may result in the patient being highly sensitized to HLA and minor histocompatibility antigens, as well as competent immune systems (1, 2). Therefore, less toxic conditioning regimens that are effective at establishing engraftment are needed.

Finding a donor can be another barrier for patients with non-malignant diseases. Many patients do not have a suitable HLA-matched sibling because the sibling may be affected by the same inherited disease. In addition, many of the non-malignant diseases such as sickle cell disease and thalassemia affect ethnic minorities, making it even more difficult to find a suitable HLA-matched unrelated donor (3). Specifically, the probability of finding an 8/8 HLA-matched unrelated donor for black Americans of all ethnic backgrounds is 16-19%, and for Asians range between 27-42% (3, 4). Historically, HLA-haploidentical HCT required myeloablative conditioning and T-cell depletion in order to decrease the risk for rejection and graft-vs-host disease (GVHD). However, this resulted in inferior survival due to increased infections, poor immune reconstitution, and graft failure (5-7). Therefore, it is imperative to develop safe and effective approaches using less intense conditioning regimens for patients receiving HLA-haploidentical grafts. An HLA-haploidentical related donor is more likely to be available to such patients and has the advantage of having one completely matched HLA-haplotype. Numerous studies have now established that nonmyeloablative conditioning followed by post-HCT cyclophosphamide (PT-CY) results in low transplant-related mortality (TRM) for patients with hematologic malignancies given T-replete HLA-haploidentical hematopoietic stem cell grafts (8-11). The approach relies on CY to eliminate both donor and host allo-activated T cells to reduce the risk for GVHD and graft rejection (12). Studies in adult patients with hematologic malignancies demonstrated graft failure rates of approximately 13%, while acute grades II-IV and III-IV GVHD, and chronic GVHD was approximately 34%, 6%, and 5-25%, respectively (8). We reasoned that the regimen reported by Luznik/O’Donnell et al. (8, 13) consisting of CY (14.5 mg/kg/day IV on days −6 and −5), fludarabine (30 mg/m²/day IV on days −6 to −5) and 200 cGy of total body irradiation (TBI) on day −1, could be modified for treatment of patients with non-malignant diseases. First, to overcome engraftment barriers, both the total dose of CY and the dose of TBI were increased for some disorders. Second, to reduce the risk for GVHD, sirolimus was later added to the post-grafting immune suppression regimen. This change was based on the results of two separate clinical trials using nonmyeloablative conditioning in adults with hematologic malignancies given HLA matched or mismatched unrelated peripheral blood stem cell (PBSC) grafts that showed decreased rates of GVHD with the addition of sirolimus (14, 15). Here we report the outcomes of the modified nonmyeloablative regimen for treatment of patients with non-malignant diseases given HLA-haploidentical T-replete grafts.

METHODS

Patients included in this analysis had a non-malignant disorder (excluding Fanconi Anemia) and underwent a nonmyeloablative T-cell replete HLA-haploidentical HCT between May 24, 2006 and July 31, 2018. Twenty-three patients were included in the analysis performed on March 22, 2019. Patients or their legal guardians provided written consent.

The modified nonmyeloablative conditioning regimen consisted of CY 25 mg/kg/day given IV on days −6 and −5 (total dose 50 mg/kg), fludarabine (FLU) 30 mg/m²/day given IV on days −6 through −2 (total dose 150 mg/m²), followed by 200 or 400 cGy TBI on day −1. In general, patients with primary immunodeficiencies received 200 cGy TBI, while patients with severe aplastic anemia (SAA), sickle cell disease (SCD), or other non-malignant diseases considered at increased risk for graft rejection received 400 cGy TBI (2 fractions of 200 cGy). Of the 23 patients treated, 17 received bone marrow (BM) as the stem cell source, and 6 received unmodified granulocyte colony stimulating factor (G-CSF) mobilized PBSC grafts.

GVHD prophylaxis was consistent with that described for adult patients with hematologic malignancies (16). The initial four patients were given CY 50 mg/kg I.V. on day +3, and subsequent patients were given an additional dose of CY 50 mg/kg I.V. on day +4. The addition of a second dose of PT-CY was based on concurrent modifications to the adult trial for hematological malignancies, which noted a lower rate of chronic GVHD in patients who received two doses of PT-Cy versus a single dose (8). All patients received tacrolimus (TAC) and mycophenolate mofetil (MMF) beginning on day +5, as previously described. However, even with the addition of a second dose of CY for the next three patients, the incidence of acute GVHD was higher than observed for the concurrent adult cohort. Therefore, sirolimus (2 mg (>1.5 m²) or 1 mg/m2 (≤1.5 m²) orally per day) was added, starting on day +5. The addition of sirolimus was based on the results of two separate clinical trials using nonmyeloablative conditioning in adults with hematologic malignancies given HLA matched or mismatched unrelated PBSC grafts showing decreased rates of GVHD with the addition of sirolimus (14, 15). All patients received supportive care with prophylactic antimicrobials, intravenous immunoglobulin (IVIG), and weekly PCR monitoring for reactivation of cytomegalovirus (CMV), Epstein-Barr virus (EBV), and adenovirus, according to institutional standard practices. Except for patients with SCD, G-CSF (5 μg/kg/day) IV was initiated on day +5 and continued until the patient had an absolute neutrophil count (ANC) ≥500/μL for 3 consecutive days.

The pre-HCT co-morbidity score was assessed by the HCT augmented co-morbidity index (HCT-CI) (17). Diagnosis, clinical grading, and treatment of acute and chronic GVHD were performed according to established criteria (18-20). NIH chronic GVHD scoring and grading was performed according to established guidelines (21, 22). Toxicities were defined by the National Cancer Institute’s Common Toxicity Criteria, version 2.0, excluding hematologic toxicities (23). Procedures for immunophenotype analyses and sorting of peripheral blood cell subsets by flow cytometry have been described previously (24). Donor chimerism levels were assessed in flow cytometry sorted CD33+, CD3+, CD19+, and CD56+ subsets by PCR-based analyses of polymorphic microsatellite regions, using methods previously described (25-27).

Disease response was evaluated by disease-specific clinical parameters and cellular functional assays. Specifically, in patients with certain primary immune deficiency diseases, immune reconstitution was assessed by lymphocyte immunophenotype analysis, lymphocyte proliferation to mitogen and anti-CD3 antibody, and quantitative serum immune globulin levels (28, 29). Additional immunologic assays included neutrophil oxidative burst in patients with chronic granulomatous disease (CGD), and natural killer (NK) cell function in patients with hemophagocytic lymphohistiocytosis (HLH). In patients with SAA, hematopoietic recovery was assessed by peripheral blood counts and/or bone marrow aspiration or biopsy. Hemoglobin gel electrophoresis and transfusion independence was used to assess SCD.

Study Endpoints and Statistical Analysis

Engraftment was defined as achieving an ANC >500/μL for 3 consecutive days and documentation of >5% donor CD3 T-cell chimerism. Platelet engraftment was defined as achieving a platelet count >20,000/μL without a platelet transfusion in the preceding 7 days. TRM and incidence of GVHD were calculated using cumulative incidence estimates, treating death due to disease as a competing risk event for TRM, and graft rejection with disease recurrence and death as a competing risk for GVHD. The method of Kaplan-Meier was used to estimate overall survival (OS), which was defined as the duration from date of HCT to the date of death due to any cause. Event free survival (EFS) was calculated using the Kaplan-Meier method, with an event defined as disease recurrence, graft rejection and death. Patients last known to be alive and event free were censored at their date of last follow up. Patients who received PT-CY, TAC and MMF were assigned to the 3-drug immune suppression group, and those who received PT-CY, TAC, MMF and sirolimus were included in the 4-drug immune suppression group. Grey’s test was used to compare differences between groups. Donor chimerism comparisons were analyzed with the Wilcoxon rank-sum test. Graphic representations of donor chimerism values across time were created by taking the chimerism value closest to days +28, day +80, day +180, and yearly thereafter up to 5 years after HCT.

RESULTS

Patient Characteristics

Patient characteristics at the time of HCT are shown in Table 1. Our cohort included primarily pediatric and young adult patients, with a median age at HCT was 10.8 (range, 0.6-47.3) years. Demographic makeup consisted of African American (n=9), Caucasian (n=5), Asian (n=4), Hispanic/Latino (n=4) and Native American/Alaskan Native (n=1) patients. The majority of patients had one or more comorbidities, which placed them at increased risk for TRM (17, 30, 31). The median augmented HCT-comorbidity index (HCT-CI) was 4 (range, 0 to 10). This was the second HCT on this protocol for three patients (#5, #7, #13) who had graft rejection and recurrence of their original disease after their initial HCT. They underwent a second HCT within 55-78 days of their initial HCT. The median total nucleated cell count was 4.3×10⁸/kg (range, 1.7–63.7×10⁸/kg); the median CD34 count was 5.8×10⁶/kg (range, 2.6–22.7×10⁶/kg), and the median CD3 count was 0.5×10⁸/kg (range, 0.01–13.7×10⁸/kg). Seventy-four percent of patients were ABO matched with their donors (n=17), while the remainder had either a major ABO mismatch (n=5) or a minor ABO mismatch (n=1). Eighty-seven percent of donors were parents (n=20), and the remainder were siblings (n=2) or offspring (n=1).

Table 1.

Patient and pre-HCT characteristics.

Patient no.	Age at HCT (years) / gender	Diagnosis	CMV donor / recipient serology	Lansky/ Karnofsky score	Augmented HCT-CI score	Pre-HCT co-morbidities	TBI dose (cGy)	Cell source	GVHD prophylaxis
Severe aplastic anemia/marrow failure
1	6 / F	SAA	D+ / R−	100	2	Iron overload	400	BM	PT-CY (x2), MMF, TAC
2	21 / F	SAA	D+ / R+	90	8	Pulmonary aspergillus, iron overload, cardiomyopathy	400	BM	PT-CY (x2), MMF, TAC, Sirolimus
3	12 / M	SAA	D+ / R+	70	10	Candida fungemia, intracranial hemorrhage, pulmonary hemorrhage	400	BM	PT-CY (x2), MMF, TAC, Sirolimus
4	16 / M	SAA	D+ / R+	100	1	CNI induced AKI	400	BM	PT-CY (x2), MMF, TAC, Sirolimus
5	7 / F	SAA	D− / R+	N/A	4	Rejected HLA-mismatched DUCB HCT 55 days before haploidentical HCT	400	BM	PT-CY (x1), MMF, TAC
6	10 / M	SAA	D+ / R+	90	2	CKD stage II	400	BM	PT-CY (x2), MMF, TAC
7	7 / M	SAA	D+ / R+	100	6	Rejected HLA-mismatched URD HCT 78 days before haploidentical HCT	400	PBSC	PT-CY (x2), MMF, TAC, Sirolimus
8	10 / M	DKC	D− / R−	100	1		200	BM	PT-CY (x2), MMF, TAC, Sirolimus
Hemoglobinopathies
9	11 / M	SCD	D+ / R−	80	6	Stroke, acute chest syndrome, iron overload	400	BM	PT-CY (x2), MMF, TAC, Sirolimus
10	13 / M	SCD	D+ / R−	90	2	Acute chest syndrome, splenic sequestration, iron overload	400	BM	PT-CY (x2), MMF, TAC, Sirolimus
11	15/ M	SCD	D− / R−	100	3	Acute chest syndrome	400	BM	PT-CY (x2), MMF, TAC, Sirolimus
12	47 / M	SCD	D+ / R+	70	6	Stroke, Pulmonary embolism	400	PBSC	PT-CY (x2), MMF, TAC, Sirolimus
13	13 / M	SCD	D+ / R+	100	7	Rejected HLA-haploidentical HCT 65 days before haploidentical HCT from same donor. Stroke with moyamoya, iron overload	400	PBSC	PT-CY (x2), MMF, TAC, Sirolimus
Hemophagocytic disorders
14	13 / F	HLH	D+ / R−	50	7	Chronic inflammatory demyelinating polyneuropathy, left pontine hemorrhage	200	BM	PT-CY (x1), MMF, TAC
15	10 / F	HLH	D+ / R−	70	2	CNS HLH, history of ECMO, CRRT at diagnosis	400	BM	PT-CY (x2), MMF, TAC, Sirolimus
16	16 / M	HLH	D+ / R−	80	5	CKD stage III	400	PBSC	PT-CY (x2), MMF, TAC, Sirolimus
17	32 / M	HLH	D+ / R+	100	4	Candida fungemia	400	PBSC	PT-CY (x2), MMF, TAC, Sirolimus
18	1 / F	HLH	D+ / R+	90	2	EBV driven HLH, CNS disease, Cytogenetic abnormality	400	PBSC	PT-CY (x2), MMF, TAC, Sirolimus
Primary immunodeficiency diseases
19	0.8 / M	X-SCID	D+ / R+	100	4	CMV pneumonia, PJP pneumonia	200	BM	PT-CY (x1), MMF, TAC
20	1 / M	X-SCID	D+ / R−	100	1	Metapneumovir us infection	200	BM	PT-CY (x1), MMF, TAC
21	0.6 / M	X-SCID	D+ / R−	100	3	PJP pneumonia	200	BM	PT-CY (x2), MMF, TAC
22	5 / F	CGD	D+ / R+	80	6	Pulmonary aspergillus, invasive Geosmithia argillacea chest wall fungal infection received pre-HCT granulocyte infusions	200	BM	PT-CY (x2), MMF, TAC, Sirolimus
Glanzmann thrombasthenia
23	2 / M	GT	D+ / R+	100	0		200	BM	PT-CY (x2), MMF, TAC, Sirolimus

Abbreviations: AKI, acute kidney injury; CGD, chronic granulomatous disease; CKD, chronic kidney disease; CMV, cytomegalovirus; CNS, central nervous system; CRRT, continuous renal replacement therapy; D, donor; DKC, dyskeratosis congenita; DUCB, double umbilical cord blood; EBV, Epstein-Barr virus; ECMO, extracorporeal membrane oxygenation; GT, Glanzmann Thrombasthenia; HCT, hematopoietic cell transplantation; HLH, hemophagocytic lymphohistiocytosis; PJP, N/A, not available; Pneumocystis jiroveci pneumonia; PT-CY, post-transplant cyclophosphamide; R, recipient; SAA, severe aplastic anemia; SCD, sickle cell disease; TAC, tacrolimus; TBI, total body irradiation; URD, unrelated donor; X-SCID, X-linked severe combined immunodeficiency disorder.

Engraftment and Chimerism

Donor engraftment (defined as achieving an ANC >500/μL for 3 consecutive days and >5% donor CD3 T-cell chimerism) was established in 22 patients (95%) including the 3 patients who had a rejected a previous HCT. The median time to neutrophil recovery was 16.5 (range, 10-25) days. Following exclusion of the patients with sickle cell disease who received regular platelet transfusions to maintain a platelet count >50,000/μL, the median time to platelet recovery of 20,000 or 50,000/μL × 7 days was 29 (range, 0–36) and 29 (range, 11–41) days, respectively. Primary graft rejection with subsequent autologous recovery was observed in one patient with CGD who received 200 cGy TBI as part of the conditioning.

Donor chimerism levels of CD3, CD33, and CD56 subsets at the most recent evaluation are shown in Table 2. The majority of patients had full donor CD3 ≥95% (n=19) or CD33 ≥95% (n=15) chimerism. Specifically, full donor CD3, CD33, and CD56 chimerism was established in all patients treated for SAA and HLH. In addition, full donor (n=3) or high-level mixed (50%–94%; n=2) CD3, CD33, and CD56 engraftment was established in all five patients with SCD. Similar to other studies (32-34), split donor/host chimerism was observed in the three patients treated for severe combined immunodeficiency disorder (SCID). Specifically, all three patients with SCID had 100% donor CD3 chimerism but lost the myeloid graft. Donor B-cell chimerism was variable, with one patient with SCID having low-level (29%) donor B-cell engraftment and one patient having complete loss of the B cell graft. B cell chimerism was not quantifiable in the third patient with SCID who was receiving several immunosuppressive agents for GVHD treatment. The patient with Glanzmann thrombasthenia had low-level CD3 and CD56 engraftment but had complete loss of the myeloid graft.

Table 2.

Transplant Outcomes.

Patient no.	Diagnosis	Non- hematological toxicities within the first 30 days post- HCT	Viral reactivation (treatment ) within day +100	Acute GVHD grade	Chronic GVHD	Off immune- suppression (months post- HCT)	Donor Chimerism (%)*			Outcomes (time of last follow up), disease response
							CD 3	CD3 3	CD5 6
Severe Aplastic Anemia/Marrow Failure
1	SAA	—	—	—	—	5	100	100	100	Alive (2.5 yrs), normal blood counts-transfusion independent
2	SAA	Grade 3 pericardial effusion	CMV (ganciclovir) BK viruria/viremia (none)	Grade II	Yes	23	100	100	100	Alive (2.1 yrs), mild thrombocytopen ia- transfusion independent
3	SAA	Grade 3 Pancreatitis, typhlitis Grade 4 hypoxia (ventilation)	Adenovire mia (brincidofo vir), EBV (rituximab), BK viruria (none), HHV6 viremia (none)	Grade III	Yes	ongoing	100	100	100	Alive (2.7 yrs), poor graft function with ongoing cytopenias on eltrombopag, transfusion-dependent thrombocytopenia
4	SAA	—	CMV (ganciclovir)	Grade II	Yes	21	100	100	99	Alive (2.7 yrs), normal blood counts-transfusion independent
5	SAA	Grade 3 Pancreatitis, typhlitis, pneumonitis (non-invasive ventilation)	CMV (ganciclovir), HSV nasal lesion (acyclovir), BK viruria/viremia (none)	Grade III	Yes	24	100	100	100	Alive (5.7 yrs), normal blood counts-transfusion independent
6	SAA	—	CMV (foscarnet)	Grade II	Yes	41	100	100	100	Alive (3.8 yrs), normal blood counts-transfusion independent
7	SAA	—	Adenovirus (cidofovir), EBV (none), HSV stomatitis (acyclovir)	Grade II	Yes	ongoing	100	100	100	Alive (1.6 yrs), normal blood counts-transfusion independent
8	DKC	—	—	—	—	11.5	100	100	100	Alive (1.6 yrs), normal blood counts - transfusion independent, 40-60% marrow cellularity at 1 year
Hemoglobinopathies
9	SCD	—	CMV (ganciclovir) BK viruria (none),	Grade II	Yes	ongoing	67	68	70	Alive (3.1 yrs), Hb S 34% (donor with Hb S trait), resolution of symptoms
10	SCD	—	—	Grade II	Yes	ongoing	100	100	100	Alive (2.5 yrs), Hb S 24% (donor with Hb S trait), resolution of symptoms
11	SCD	—	—	—	—	ongoing	51	87	56	Alive (1.7 yrs), Hb S 34% (donor with Hb S trait), resolution of symptoms
12	SCD	—	CMV (ganciclovir)	Grade II	—	29	100	100	100	Alive (2.4 yrs), Hb S 25% (donor with Hb S trait), resolution of symptoms
13	SCD	—	CMV (ganciclovir) HHV6 viremia (none), BK viremia/vir uria (none)	Grade III	—	29	100	100	100	Alive (2.7 yrs), Hb S 26% (donor with Hb S trait), resolution of symptoms
Hemophagocytic Disorders
14	HLH	Grade 3 CNS (seizure), typhlitis, renal failure (reversible); Grade 4 septic shock, hypoxia (ventilation), pancreatitis	EBV (none)	Grade II	Yes	28	100	92	64	Alive (10.1 yrs), normal NK function, clinical remission
15	HLH	—		Grade II	—	ongoing	100	100	100	Alive (0.6 yrs), normal NK function, clinical remission
16	HLH	—	CMV (ganciclovir)	Grade II	—	13	100	100	100	Alive (2.2 yrs), normal NK function, clinical remission
17	HLH	—	CMV (ganciclovir)	Grade II	Yes	ongoing	100	100	100	Alive (2.8 yrs), normal NK function, clinical remission
18	HLH	—	CMV (foscarnet), EBV (none)	—	—	ongoing	100	100	100	Alive (1 yr), normal NK function, clinical remission, resolution of cytogenetic abnormality
Primary Immunodeficiency Diseases
19#	X-SCID	—	CMV (foscarnet)	Grade II	Yes	56	100	0	100	Alive (11 yrs), normal absolute CD3, CD4, CD8 and CD56 lymphocyte numbers, T-cell proliferation to PHA and anti-CD3 within normal range compared to control, remains on immunoglobulin replacement
20#	X-SCID	—	CMV (ganciclovir) EBV (none),	Grade IV	Yes	at time of death	100	0	95	Died (1.4 yrs), cause of death related to chronic extensive skin GVHD and disseminated candidemia, severe persistent lymphopenia, no significant T-cell proliferation in response to PHA and anti-CD3, on immunosuppression
21#	X-SCID	—		Grade III	Yes	42.5	100	0	24	Alive (7.1 yrs), normal absolute CD3, CD4, CD8 and CD56 lymphocyte numbers, T-cell proliferation to PHA and anti-CD3 within normal range compared to control, remains on immunoglobulin replacement
22**	CGD	Grade 3 pneumatosis, AKI, hypoxia (non-invasive ventilation)	CMV (foscarnet), BK viruria (none)	—	—	2.5	0**	0	0	Died (0.5 yrs), cause of death related to pre-HCT fungal infection and continued CGD in setting of myeloid graft loss
Glanzmann Thrombasthenia
23	GT	—	CMV (ganciclovir) adenovirus (cidofovir)	Grade III	Yes	14	21	0	10	Alive (2.2 yrs) continued disease: gpIIb and gpIIa antibodies still present post-HCT

^*At time of last follow-up

^**Primary graft rejection

^#B cell chimerism at last follow up- patient 19: 29% donor, patient 20: not quantifiable, and patient 21: 0% donor

Abbreviations: AKI, acute kidney injury; CMV, cytomegalovirus; EBV, Epstein-Barr virus; gp, glycoprotein; GVHD, graft-vs-host disease; Hb, hemoglobin; HCT, hematopoietic cell transplantation; HHV6, human herpesvirus 6; HSV, herpes simplex virus; yrs, years;

Graft-Versus-Host Disease

The cumulative incidences (CI) of grade II-IV and III-IV GVHD in patients who received 3-drug vs. 4-drug immune suppression were 86% vs. 75% (p=0.33) and 43% vs. 19% respectively (p=0.24; Figure 1). The 2-year CI of NIH-consensus chronic GVHD was 43% in the 3-drug group compared to 38% in the 4-drug group (p=0.93). The 2-year CI of extensive chronic GVHD in patients who received 3-drugs vs. 4-drugs was 86% vs. 52%, respectively (p=0.07). Although there appeared to be lower incidences of acute and chronic GVHD in the cohort given four drugs, the differences were not statistically significant, likely due to small patient numbers. There was no statistically significant difference in the cumulative incidence of grades II-IV and III-IV acute GVHD or NIH chronic and chronic extensive GVHD between recipients of BM versus PBSC grafts (data not shown).

Figure 1. Acute and chronic graft-vs.-host disease (GVHD).

Shown is the cumulative incidence of acute GVHD grades II-IV (panel A), grades III-IV (panel B), and extensive chronic GVHD (Panel C) according to 3 versus 4-drug immune suppression.

Disease Response

The most recent disease response assessments for each patient are shown in Table 2. Of the 23 patients, 19 had improvement in or resolution of their disease. Of the three patients with X-SCID, two demonstrated clinical improvement without significant infections since 1-year post-HCT. In addition, both patients had a CD3 count >300 cell/μl and a CD8 count >50 cells/μL by day+100, which has recently been shown to be prognostic of survival in patients with SCID (28). Both patients had evidence of normal T-cell proliferation in response to both PHA and anti-CD3 at 5 years post-HCT. However, both patients remained on IVIG replacement at time of last follow-up likely secondary to low-level or complete loss of the donor B-cell graft. One patient with SCID continued to have poor immune reconstitution 1.4 years post-HCT due to chronic extensive skin GVHD requiring continued immune suppression at time of death and was therefore unevaluable for disease response. In addition, one patient with CGD had disease recurrence due to loss of the myeloid graft.

Normal NK function was established in all 5 patients with HLH. Six of seven patients with SAA maintained normal blood counts and transfusion-independence. One patient with SAA had poor graft function with continued transfusion-dependent thrombocytopenia and ongoing eltrombopag therapy despite having 100% donor myeloid engraftment. All patients with SCD had improved hemoglobin levels (median 11.2 g/dL; range 9.0–14.9 g/dL), became transfusion independent, had hemoglobin S percentage levels consistent with the donor having sickle cell trait (24%-34%), and had no further episodes of vaso-occlusive crisis. The patient with Glanzmann thrombasthenia also had disease recurrence following complete loss of the myeloid graft.

Transplant-Related Complications

Non-hematologic toxicities possibly related to the conditioning regimen are summarized in Table 2. Three patients developed one or more toxicities within the first 30 days post-HCT [grade 3 (n=3) or grade 4 (n=2)]. None of the patients developed veno-occlusive disease, including the three patients with a history of iron overload pre-HCT, and none developed mucositis. One patient with HLH and a history of stroke pre-HCT had a single focal seizure (grade 3) on day +5 post-HCT in the setting of tacrolimus and sepsis.

Sixteen patients developed one or more viral reactivations or infections requiring treatment within the first 100 days post-HCT (Table 2). EBV reactivation was detected in 5 patients, of which one required rituximab. CMV reactivation was detected in 11 of the 13 patients who were CMV seropositive pre-HCT, and in 3 of the 8 patients who were CMV seronegative pre-HCT but received grafts from CMV positive donors. All patients received pre-emptive antiviral treatment and had complete resolution of CMV reactivation. Adenovirus was detected in three patients requiring treatment with complete resolution. Additional viral infections included BK virus (n=6), human herpes virus 6 (n=2), and herpes simplex virus (n=2).

Overall Survival and Transplant-Related Mortality

With a median follow up of 2.5 (range 0.5-11) years, 21 patients are alive with a 2-year OS of 91% (95% CI; 68%–98%; Figure 2). Day +100 and 1-year TRM were 0% and 4%, respectively. The 2-year event free survival (EFS) was 78% (95% CI; 54%–90%). Two patients died. One patient with CGD and disseminated mold infection present at time of conditioning died of progressive mold infection in the setting of graft rejection at 6 months post-HCT. In addition, one patient with SCID died of chronic extensive skin GVHD and disseminated candidemia at 16 months post-HCT.

Figure 2.

Shown are the Kaplan Meir estimates of overall survival (solid line), event-free survival (dotted line), and the cumulative incidence of transplant-related mortality (dashed line).

DISCUSSION

This current study found that HLA-haploidentical HCT, originally developed to treat adult patients with hematological malignancies by Luznik et al. could be modified and applied successfully to treat patients with various non-malignant disorders (16). The two-year OS for patients with a variety of non-malignant diseases, including aplastic anemia, immunodeficiency disorders, sickle cell disease and hemophagocytic disorders, was 91%. Evaluation of all 23 patients revealed the two-year EFS was 78%, with events defined as disease recurrence, graft rejection and death. Day 100 and 1-year TRM were low at 0% and 4%, respectively, which compares favorably to the adult trial (16). Our results showed that patients with high augmented HCT-CI scores can tolerate a moderately more intensive nonmyeloablative regimen in the setting of PT-CY. Bolaños-Meade et al., DeZern et al., Thakar et al. and Bonfim et al. have published similar outcomes using nonmyeloablative conditioning regimens and PT-CY in patients with SCD, SAA and Fanconi anemia, respectively (35-38).

For patients with non-malignant diseases, we presumed that the barrier to engraftment would be higher because of the risk of transfusion-induced sensitization to the HLA-haploidentical graft; therefore, we intensified the regimen with the goal of maintaining engraftment without increasing toxicity or TRM. The total dose of CY given during conditioning was increased slightly from 28 mg/kg to 50 mg/kg, bringing the overall dose of CY given pre- and post-HCT to 150 mg/kg. Patients considered at increased risk for graft rejection, specifically those with intact immune systems and/or robust hematopoiesis, received an additional 200 cGy TBI, for a total nonmyeloablative dose of 400 cGy TBI. Since many patients with these diseases undergo HCT in childhood, we did not wish to mandate that all patients be exposed to 400 cGy TBI, but only those who were considered at increased risk for graft rejection. The modified nonmyeloablative regimen resulted in favorable disease-specific lineage engraftment for most of our patients. However, graft rejection occurred in one patient treated for CGD. This patient received granulocyte transfusions from unrelated donors before HCT secondary to an invasive pulmonary and chest wall mold infection. Although T cell cytotoxic crossmatch between the patient and donor was negative one month prior to HCT, this patient continued to receive granulocyte transfusions until the start of conditioning and after HCT. Therefore, this patient may have become sensitized to major and minor histocompatibility antigens and, as a result, may have developed reactive class I and class II panel-reactive antibodies (PRAs), thereby increasing their risk for rejection (39-41). In comparison, Bolaños-Meade reported their experience using CY 28 mg/kg, FLU 150 mg/kg and 200 cGy TBI with the addition of antithymocyte globulin (ATG) in adult patients with SCD given HLA-haploidentical grafts and noted a higher graft failure rate of 43% (35). They subsequently intensified the regimen by increasing the TBI dose to 400 cGy and showed a decrease in the incidence of graft failure to 6%, with 76% of patients achieving full donor chimerism (42).

Alloreactivity of the donor graft plays an important role in establishing donor hematopoiesis after nonmyeloablative HCT, and for patients with hematologic malignancies also contributes an important graft-vs-leukemia (GVL) effect to prevent relapse (43-45). The original Luznik/O’Donnell study found PT-CY could control the risk for acute GVHD, reporting the incidence of day +200 grade II–IV and III–IV GVHD of 34% and 6%, respectively (16). Subsequent retrospective studies in adult patients treated for hematologic malignancies have reported similar incidences (46, 47). Since patients with non-malignant diseases do not benefit from GVL, we intensified the post-HCT immune suppression to include sirolimus, based on promising results from randomized trials in adults with hematologic malignancies given nonmyeloablative conditioning and unrelated donor matched or mismatched PBSC grafts (14, 15). Although we found a lower incidence of acute grade III-IV GVHD in recipients of 4-drug versus 3-drug immune suppression, this difference was not significant likely due to small patient numbers. Importantly, the addition of sirolimus did not appear to result in increased toxicities. Nonetheless, despite the addition of sirolimus to post-grafting immune suppression, the rates of acute and chronic GVHD were higher compared to that reported in adult patients with hematologic malignancies (16).

It is not intuitively obvious why the incidence of acute and chronic GVHD was increased in our patients with non-malignant diseases. Several studies using haploidentical bone marrow HCT followed by PT-CY in patients with non-malignant diseases have reported lower rates of GVHD (36, 48-52). Several of these studies included serotherapy, which may in part explain the lower incidence of GVHD. Although PBSC has been associated with an increased incidence of chronic GVHD (53, 54), there was still a high rate of acute and chronic GVHD in our BM recipients. Therefore, we do not think the higher rate of GVHD seen was solely due to the use of PBSC grafts. Our study population also included much younger patients; therefore, it is plausible that age-associated differences in CY metabolism may have contributed to less stringent control of alloreactive T cells. In a recent study of HLA-haploidentical HCT with PT-CY in pediatric patients with acute leukemia, the cumulative incidence of grade II-IV acute GVHD was 40.3% (55). Furthermore, a subgroup analysis found that acute GVHD developed in 80% of patients <10 years of age compared to 13.3% in those patients between 10-20 years of age at time of HCT (P = 0.001) (55). Similar to known variability in metabolism of other chemotherapy agents such as busulfan, studies have demonstrated that CY metabolism in pediatric patients is associated with age and body-surface area (56). These insights suggest that future studies utilizing PT-CY in pediatric patients should incorporate assessments of CY pharmacokinetic and pharmacodynamics and aim to establish optimal dosing strategies of PT-CY to reduce GVHD. We observed a high rate of viral reactivation within the first 100 days post-HCT, with 16 of 23 patients (70%) requiring treatment for CMV, adenovirus, EBV and HSV reactivation post-HCT. Importantly, none of the patients died from viral infections. Others have report similar incidences of viral reactivation (35%-84%) in patients undergoing haploidentical T-cell replete HCT utilizing PT-CY (57-59). Contributing factors to viral reactivation requiring treatment in our study may be due to a high number of patients with pre-HCT viral exposures and a high rate of GVHD requiring additional immunosuppression. Due to limited numbers we were unable to evaluate if the addition of sirolimus contributed to viral reactivation requiring treatment post-HCT.

Although this study included a heterogeneous group of diagnoses, the individuals in our study cohort share some of the common problems of chronic infections, inflammation, and shortened life expectancy; therefore, the goals of reducing the intensity of the conditioning regimen, reducing rates of GVHD and TRM, and maximizing sustained donor engraftment and cure are the same. The modified nonmyeloablative regimen for HLA-haploidentical HCT appears to be a reasonable treatment approach for high-risk patients with life threatening comorbidities; however, additional strategies are needed for GVHD prevention in order to optimize this approach for patients with non-malignant diseases.

HIGHLIGHTS.

• Modified nonmyeloablative conditioning has a low toxicity profile

• Overall survival of 91% and event-free survival of 78%

• High incidence of acute and chronic GHVD