Feasibility and efficacy of partially replacing post-transplant cyclophosphamide with bendamustine in pediatric and young adult patients undergoing haploidentical bone marrow transplantation

Abstract

Background:

Post-transplant cyclophosphamide (PT-CY) is the most widely applied graft-versus-host disease (GvHD) prophylaxis regimen in T-cell replete haploidentical bone marrow transplantation (haplo-BMT). While PT-CY has been met with great success in the haplo-BMT arena by suppressing GvHD, patients without acute GvHD have high relapse rates.

Objective:

One of the strategies being explored by others to lessen relapse rates is the dose reduction of PT-CY. We have taken a different approach in evaluating whether partially replacing PT-CY with post-transplant bendamustine (PT-BEN) would be advantageous, which is based on our pre-clinical research that delineated several beneficial immunomodulatory properties of BEN.

Study Design:

We therefore initiated and completed a Phase Ia trial which evaluated the progressive substitution of PT-CY with PT-BEN (NCT02996773). Thirteen patients with high-risk hematologic malignancies have received PT-CY/BEN and their outcomes compared to 31 contemporaneous haplo-BMT recipients treated with the same myeloablative conditioning (MAC) regimens but receiving only PT-CY.

Results:

We demonstrate that partial replacement of PT-CY with PT-BEN on day +4 (PT-CY/BEN) is well tolerated and associated with significantly earlier trilineage engraftment. We also show favorable trends to significant improvements in univariate and multivariate analyses, with PT-CY/BEN compared to PT-CY, with respect to chronic GvHD (HR, 0.08; 95% CI, 0.005, 1.11; P=0.06), and GvHD-free-relapse-free survival (GRFS) (HR, 0.22; 95% CI, 0.05, 0.86; P=0.039). Our human trial has transitioned to Phase Ib which will further evaluate the safety and potential benefits of PT-CY/BEN. Herein, we also expand our pediatric, adolescent, and young adult (AYA) experience to 31 patients demonstrating an overall survival (OS), progression-free survival (PFS) and GRFS at 3 years of 85.6%, 76.1% and 58.2%, respectively in a largely racial/ethnic minority cohort.

Conclusions:

PT-CY/BEN appears to be a promising treatment option that requires further evaluation.

Graphical Abstract

Introduction

We have previously reported in preclinical murine haploidentical bone marrow transplantation (haplo-BMT) that replacement of post-transplant cyclophosphamide (PT-CY) with post-transplant bendamustine (PT-BEN) is protective against late graft-versus-host disease (GvHD)¹. We have also documented distinct immunomodulatory properties of BEN in several mismatched murine BMT models^1–6. BEN given pre- or post-transplant and compared to CY, led to significant changes in the proportion, phenotype and function of multiple immune subsets including myeloid derived suppressive cells (MDSCs)² and dendritic cell (DC)^3,5 subsets, yielding tolerant T-cells with a striking absence of GvHD, but preservation of T-cell dependent graft versus leukemia (GvL)⁴. We therefore initiated a Phase Ia (NCT02996773) trial which evaluated the progressive substitution of PT-CY with PT-BEN to determine the maximal tolerated dose. The interim analysis of the Phase Ia demonstrated that partial replacement of PT-CY with PT-BEN was associated with early trilineage engraftment, a low incidence of acute GvHD (aGvHD), chronic GvHD (cGVHD), and reduced cytomegalovirus (CMV) reactivation⁷. We herein report our clinical findings from the completed Phase Ia and compare them with data from contemporaneous haplo-BMT recipients conditioned with the same regimens but receiving only PT-CY. To verify the safety and potential advantages of this platform we have transitioned to a Phase Ib using PT-CY on day +3 with PT-BEN on day +4 (PT-CY/BEN).

Patients and Methods

Patients:

Patients with hematologic malignancies who underwent myeloablative conditioned (MAC) T-replete haplo-BMT between Oct 2015 and June 2021, were included in this study. Eligible patients were between 0 and 44 years, who had no matched related donor, met organ criteria allowing for MAC and had no evidence of active untreated infection. Haplo-BMTs were performed on the pediatric (n=31) or adult (n=13) hematopoietic cell therapy and transplant (HCTT) services at Banner University Medical Center in Tucson, Arizona.

Transplant Procedure:

According to diagnosis and disease characteristics patients received either fractionated total body irradiation (TBI) followed by fludarabine (FLU) or busulfan (BU), fludarabine (FLU) and melphalan (MEL) combination (Table 1)⁸. The first three patients receiving TBI-FLU were conditioned with fractionated TBI of 333 cGy given once daily for three days, while the later twenty patients received 200 cGy twice daily (1200 cGy total dose with lungs shielded to 900 cGy by custom cerrobend blocking) on days −8, −7 and −6, followed by FLU 30 mg/m² on days −5, −4, −3 and −2 ^8,9. Twenty-one patients received BU at 0.8 mg/kg IV every 6 hours for a total of 12 doses (days −8 to −6), targeting an average area under the curve (AUC) of 4.1–4.5 mg x h/L q 6 hours. BU was followed by FLU 30 mg/m² on days −5, −4, −3 and −2, which was increased to 40 mg/m² in the later six patients, and MEL 100 mg/m² on day −2 ^8,10. All patients received bone marrow grafts on day 0.

Table 1.

Patient, disease and transplant characteristics

	PT-CY/BEN	PT-CY	P=
N=	13	31
Age, median yr, (range)	21.4 (9–42)	20.5 (0.6–44)	0.80
Male, n (%)	8 (62)	21 (68)	0.74
Race/Ethnicity, n (%)			0.52
African American	1 (8)	3 (10)
Native American	0 (0)	1 (3)
Asian	0 (0)	1 (3)
White Hispanic	6 (46)	15 (48)
White non-Hispanic	6 (46)	11 (35)
Diagnosis, n (%)			0.72
ALL	6 (46)	16 (52)
AML/MDS	3 (23)	10 (32)
AUL/MLL	1 (8)	1 (3)
CML	1 (8)	2 (6)
NHL	2 (15)	1 (3)
HD	0 (0)	1 (3)
Clinical Service			0.72
Pediatric	10 (77)	21 (68)
Adult	3 (23)	10 (32)
Pretransplant Status, n (%)			0.40
CR1	5 (38)	12 (39)
CR2	3 (23)	11 (35)
>CR2	3 (23)	4 (13)
other	2 (15)	4 (13)
Pretransplant Status, n (%)			0.30
Prior HCT	0 (0)	4 (13)
Prior CAR-T	1 (8)	3 (10)
Disease risk index, n (%)			0.99
Low	1 (8)	3 (10)
Intermediate	9 (69)	20 (65)
High	3 (23)	7 (23)
Very high	0 (0)	1 (3)
Lansky/Karnofsky PS, n (%)
90–100	9 (69)	19 (61)	0.74
≤80	4 (31)	12 (39)
HCT Comorbidity index n (%)			0.20
≤2	10 (77)	17 (55)
≥3	3 (23)	14 (45)
Conditioning, n (%)			0.99
TBI-FLU	7 (54)	16 (52)
BU-FLU-MEL	6 (46)	15 (48)
Donor age, median yr, (range)	27 (15–57)	35.5 (16–62)	0.40
Sex mismatch, n (%)			0.74
Female → Male	3 (23)	9 (29)
Male → Female	1 (8)	4 (13)
None	9 (69)	18 (58)
HLA Match, n (%)			0.16
5/10	11 (84)	18 (58)
6/10	1 (8)	9 (29)
7/10	1 (8)	2 (6)
8/10	0 (0)	1 (6)
9/10	0 (0)	1 (6)
RBC incompatibility, n (%)			0.45
Major	4 (31)	6 (19)
Minor	2 (15)	4 (13)
None	7 (54)	21 (68)

CY, cyclophosphamide; BEN, bendamustine; ALL, acute lymphoblastic leukemia; AML, acute myeloid leukemia;

MDS, myelodysplastic syndrome; AUL/MLL, acute undifferentiated or mixed lineage leukemia;

CML, chronic myeloid leukemia; NHL, non-Hodgkin lymphoma; HD, Hodgkin’s disease; CR, complete remission;

HCT, hematopoietic cell transplant; CAR-T, chimeric antigen receptor T-cells, TBI, total body irradiation;

FLU, fludarabine; BU, busulfan; MEL, melphalan;

GvHD prophylaxis:

Thirty-one patients received PT-CY 50 mg/kg on days +3 and +4 while thirteen were treated as part of our Phase Ia/Ib single-institution trial receiving PT-CY with post-transplant PT-BEN. The Phase Ia/Ib trial was approved by our institutional review board (IRB) and all patients were provided and signed a written informed consent. The study was conducted in accordance with the Declaration of Helsinki. The phase Ia was a standard 3+3 dose escalation design with the first three cohorts receiving on day +4 PT-CY (mg/kg)/PT-BEN (mg/m²): 40/20, 20/60, and 0/90. All patients received PT-CY 50 mg/kg on day +3. Cohort 4 patients received PT-CY/BEN 40/20 on day +3 and only PT-BEN 90 mg/m² on day +4. Cohort 3 dosing (Day +3, CY 50 mg/kg and day +4, BEN 90 mg/m²) was deemed the maximal tolerated dose (MTD) as we transitioned to Phase Ib with patient #13 who received PT-CY on day +3 and PT-BEN on day +4. Additionally, all 44 patients were given mycophenolate mofetil on day +5 through day +28 and tacrolimus starting on day +5 targeting levels of 5 to 8 ng/ml. In the absence of GvHD, tacrolimus was tapered starting day +70 to +90 and discontinued by day +120 to +180. GvHD was graded according to the consensus criteria for grading acute and chronic GvHD ^11,12.

Supportive care:

Antifungal prophylaxis primarily with voriconazole was administered in all patients. Patients received IV pentamidine for pneumocystis jirovecii and acyclovir for herpes simplex and varicella virus prophylaxis. Only two PT-CY patients received letermovir prophylaxis. Bi-weekly polymerase chain reaction (PCR) monitoring for CMV and weekly for adenovirus, Epstein-Barr virus (EBV) and human herpes virus-6 (HHV-6) were performed until discharge from hospital, then subsequently at least every other week during the first 100 days following transplant. All patients were transplanted in positive pressure HEPA filtered rooms on HEPA filtered units.

Donor selection:

Donors were first degree relatives who were HLA-haploidentical based on high-resolution typing at HLA-A, -B, -Cw, -DRB1, and -DQB1. None of the patients had anti-donor HLA antibodies. Major and minor ABO incompatibilities requiring donor RBC reduction using Hespan® (6% hetastarch in 0.9% sodium chloride injection) for RBC sedimentation or plasma reduction are outlined in Table 1.

Engraftment and donor chimerism monitoring:

Granulocyte stimulating factor (G-CSF) was started on day +5 at 5 μg/kg/day until an absolute neutrophil count (ANC) of 2.5 × 10⁹/L was achieved for three consecutive days. Day of myeloid engraftment was defined as the first of three consecutive days with an ANC of ≥0.5 × 10⁹/L. Day of platelet engraftment was considered the first of three consecutive days with platelet counts of ≥20 × 10⁹/L with no platelet transfusions administered in the previous 7 days. Donor chimerism was evaluated on days +28, +100, +180 and +365 by short tandem repeats (STRs) on peripheral blood or bone marrow.

Statistical analysis:

Comparisons of patient characteristics and outcome variables between the PT-CY and PT-CY/BEN groups were performed using Fisher’s Exact Tests for categorical variables and Mann-Whitney/Wilcoxon Rank Sum Tests for continuous variables. Time to event endpoints were estimated using cumulative incidence curves and Kaplan-Meier curves, with comparisons using log-rank tests. The associations between the clinical endpoints and patient characteristics were assessed using a Cox proportional hazards model for both univariate and multivariate analyses. Multivariate analysis first assessed the significance of PT-CY versus PT-CY/BEN adjusted for each covariate individually. Subsequent multivariate analysis was based on the covariates including age, race/ethnicity, clinical service, conditioning, diagnosis, remission status, disease risk index, comorbidity index, performance status, donor-recipient sex mismatch and donor age. For both univariate and multivariate analysis, the hazard ratio (HR), score test p-value, and 95% confidence interval (CI) based on the Wald statistic are presented. Cox proportional hazards models for aGvHD II-IV, aGVHD III-IV, cGVHD and severe cGVHD also were adjusted for competing risks due to disease relapse, second transplants, or death.

Results

Patient, disease and transplant characteristics:

The study included 44 patients (66% male) with a median age of 21 years (0.6–44 years) at the time of transplantation. Thirteen patients received a PT-CY/BEN combination and thirty-one PT-CY alone. Patient and transplant characteristics are summarized in Table 1. Twelve of these patients were part of a 3 + 3 BEN dose escalation with CY de-escalation Phase Ia trial (Table 2). The 13^th patient is part of our Phase Ib trial which will further evaluate the safety and potential benefits of PT-CY/BEN compared to PT-CY alone. Baseline characteristics such as age, sex, race/ethnicity, diagnosis, clinical service, pre-transplant remission status, disease risk index, performance status, comorbidity index, conditioning regimen, donor age, sex mismatch, and HLA match were comparable between the PT-CY/BEN and PT-CY groups (Table 1). Ethnic and/or racial minorities constituted 61% of all patients, the majority of whom were Hispanic (48%). Acute lymphoblastic leukemia (ALL) was the most frequent diagnosis (50%) followed by acute myeloid leukemia and myelodysplastic syndrome (AML/MDS) (30%). Conditioning was almost evenly split with 52% of patients receiving TBI-FLU and 48% BU-FLU-MEL. (Table 1). Univariate analysis of association of outcomes (relapse, NRM, OS, PFS, GRFS) with patient, disease and transplant characteristics showed a significant increase in time to relapse for patients >21 years (HR, 4.69; 95% CI, 1.01, 21.73; P=0.03) and those transplanted on the adult clinical service (HR, 3.11; 95% CI, 0.95, 10.24; P=0.048), shortened PFS in patients receiving haplo-BMT on the adult service (HR, 2.70; 95% CI, 1.01, 7.22; P=0.04) and improved GRFS in cases where donor-recipient sex was matched (HR, 0.42; 95% CI, 0.18, 0.99; P=0.04) (Table 3).

Table 2.

PT-CY/BEN: GvHD, complications, infections

	Phase Ia				Phase Ib
	Cohort 1	Cohort 2	Cohort 3	Cohort 4
Day +3 CY/BEN	50/-	50/-	50/-	40/20	50/-
Day +4 CY/BEN	40/20	20/60	-/90	-/90	-/90
N=	3	3	3	3	1
Graft Failure
Primary	-	-	-	-	-
Secondary	-	-	-	1	-
aGvHD grade
none	1	1	2	1	-
I	-	1	1	-	-
II	2	1	-	1	-
III	-	-	-	-	1
IV	-	-	-	-	-
cGVHD
none	3	3	2	2	1
mild	-	-		-	-
moderate	-	-	1	-	-
severe	-	-	-	-	-
Bacteremia
Gram+	1	2	2	-	-
Gram-	-	-	-	-	-
Fungal	-	-	-	-	-
CMV
R-D+	-	-	-	-	-
R+D+	0/1	1/2	0/1	1/1	-
R+D-	0/1	0/1	0/2	1/1	1/1
BK viruria	2	-	-	1	-
VOD/SOS	-	-	-	-	-
TA-TMA	-	-	-	-	-
Dialysis	-	-	-	-	-
ICU admission	-	-	-	-	-
CRS	-	-	-	1	-
Relapse	1	1	-	1	-
Cause of death
Relapse	1	1	-	-	-

CY mg/kg, BEN mg/m², VOD/SOS; veno-occlusive disease/sinusoidal obstruction syndrome, TA-TMA; transplant associated-thrombotic microangiopathy, CRS; cytokine release syndrome

Table 3.

Univariate analysis of association of outcomes with patient, disease and transplant characteristics

	n	Relapse	P-value	NRM	P-value	OS	P-value	PFS	P-value	GRFS	P-value
Age
<21	22	1	0.03	1	0.98	1	0.17	1	0.10	1	0.21
>21	22	4.69 (1.01, 21.71)		1.02 (0.21, 5.06)		2.47 (0.65, 9.34)		2.36 (0.82, 6.81)		1.74 (0.72, 4.21)
Race
Hispanic	21	1	0.55	1	0.51	1	0.99	1	0.94	1	0.80
White	17	0.66 (0.17, 2.58)		1.71 (0.34, 8.58)		0.99 (0.29, 3.42)		1.04 (0.36, 2.92)		0.89 (0.36, 2.18)
Diagnosis
ALL	22	1	0.26	1	0.06	1	0.12	1	0.80	1	0.55
AML	15	2.3 (0.51, 10.38)		0 (0, 0)		0.22 (0.03, 1.77)		0.86 (0.26, 2.86)		0.74 (0.27, 2.00)
Clinic Service
peds	31	1	0.048	1	0.78	1	0.28	1	0.04	1	0.10
adult	13	3.11 (0.95, 10.24)		1.27 (0.23, 699)		1.89 (0.57, 6.22)		2.70 (1.01, 7.22)		2.05 (0.86, 4.86)
Pretransplant Status
CR1	17	1	0.78	1	0.99	1	0.87	1	0.92	1	0.94
CR2	14	0.77 (0.17, 3.46)		0.93 (0.13, 6.74)		1.31 (0.29, 5.90)		0.84 (0.25, 2.77)		0.88 (0.31, 2.56)
>CR2 or other	13	1.3 (0.32, 5.23)		1.05 (0.15, 7.52)		1.49 (0.33, 6.71)		1.07 (0.33, 3.52)		1.07 (0.39, 2.96)
Disease risk index
L/I	33	1	0.22	1	0.35	1	0.16	1	0.28	1	0.40
H/VH	11	2.13 (0.62, 7.34)		2.19 (0.39, 12.23)		2.38 (0.69, 8.23)		1.78 (0.61, 5.18)		1.47 (0.59, 3.65)
Lansky/Karnofsky PS
90–100	28	1	0.22	1	0.79	1	0.91	1	0.53	1	0.72
≤89	16	0.48 (0.15, 1.58)		1.26 (0.23, 6.87)		1.07 (0.31, 3.67)		0.73 (0.27, 1.96)		0.85 (0.35, 2.06)
HCT Comorbidity index
≤2	27	1	0.24	1	0.77	1	0.45	1	0.14	1	0.13
≥3	17	2.01 (0.61, 6.59)		1.27 (0.25, 6.35)		1.57 (0.48, 5.16)		2.07 (0.77, 5.56)		1.92 (0.82, 4.55)
Conditioning
TBI-FLU	23	1	0.99	1	0.86	1	0.98	1	0.88	1	0.11
BFM	21	1.02 (0.31, 3.36)		1.15 (0.23, 5.70)		1.01 (0.31, 3.32)		0.93 (0.34, 2.50)		0.49 (0.20, 1.22)
Donor age
<40	29	1	0.59	1	0.53	1	0.28	1	0.33	1	0.51
>40	15	1.38 (0.42, 4.52)		1.66 (0.33, 8.26)		1.9 (0.58, 6.25)		1.62 (0.61, 4.34)		1.34 (0.55, 3.24)
Sex mismatch
Yes	17	1	0.60	1	0.11	1	0.21	1	0.29	1	0.04
No	27	0.73 (0.22, 2.4)		0.27 (0.05, 1.51)		0.47 (0.15, 1.56)		0.59 (0.22, 1.58)		0.42 (0.18, 0.99)
Sex mismatch (F->M)
Yes	12	1	0.67	1	0.31	1	0.56	1	0.87	1	0.10
No mismatch	27	1.41 (0.28, 7.01)		0.37 (0.05, 2.68)		0.65 (0.15, 2.74)		0.9 (0.27, 3.01)		0.45 (0.17, 1.19)

Engraftment and chimerism:

PT-CY/BEN patients received a median of 4.05 ×10⁶/kg CD34⁺ cells compared to 3.45 ×10⁶/kg in the PT-CY group (P=0.298) (Figure 1A). Patients treated with PT-CY/BEN had earlier trilineage engraftment, with a median time to an absolute neutrophil count (ANC) of 0.5 × 10⁹/L of 13 days compared to 16 days in those receiving PT-CY (P=0.009) (Figure 1B). Similarly, PT-CY/BEN demonstrated earlier platelet engraftment with median of 22 days compared to 29.5 days in PT-CY patients (P=0.005) (Figure 1C). Consequently, PT-CY/BEN patients required fewer red blood cell (P=0.52) and platelet transfusions (P=0.008) (Figure 1D, 1E). Trilineage engraftment was seen in 95.5% of patients with all demonstrating complete donor chimerism on their day +28 bone marrows and in their peripheral blood studies on days +100 and +180 and +365.

Figure 1.

(A) Number of CD34+ cells x 10⁶/kg infused. (B) Time to an absolute neutrophil count (ANC) of 0.5 × 10⁹/L. (C) Time to a platelet count of 20 × 10⁹/L. (D) Number of units of packed red blood cells (PRBCs) transfused. (E) Number of units of platelets transfused.

Two patients developed graft failure (4.5%), one primary and the other secondary, following early CMV reactivation and ganciclovir treatment. The first patient had received PT-CY following infusion of 4.8 ×10⁶ CD34⁺ cells/kg and the second PT-CY/BEN (Cohort 4, 40/20 day +3 and 0/90 day +4) (Table 2) and 2.2 ×10⁶ CD34⁺ cells/kg. Both patients were male with 6/10 HLA antigen matched maternal donors and both received BU-FLU-MEL (30 mg/m² of fludarabine). They were successfully salvaged after a second haploidentical HCT (30 and 46 days later) using a single day conditioning regimen¹³. The PT-CY treated patient died 42 months following his first haplo-BMT of multi-organ failure while the patient who received PT-CY/BEN is alive and well >20 months following his initial transplant.

Graft versus host disease:

The cumulative incidence of grades II-IV and III-IV aGvHD was 41.3% and 8.3% following PT-CY/BEN compared to 46.2% and 20.8% in patients receiving PT-CY alone (P=0.52, and P=0.29) (Figures 2A, 2B). The cumulative incidence of chronic GvHD (cGvHD) was significantly lower at 9.1% following PT-CY/BEN with 0% of patients developing severe cGvHD compared to 44.1% and 20% in patients treated with PT-CY, respectively (P=0.049 and P=0.12) (Figures 2C, 2D). In multivariate analyses, PT-CY/BEN was not significantly associated with reduced grades II-IV aGvHD (HR, 0.71; 95% CI, 0.24, 2.10; P=0.53), grades III-IV aGvHD (HR, 0.28; 95% CI, 0.03, 2.71; P=0.27), but with a trend toward less cGVHD, overall (HR, 0.08; 95% CI, 0.005, 1.11; P=0.06) (Table 4).

Figure 2.

(A) Cumulative incidence of grades II-IV acute graft-versus-host disease (aGvHD). (B) Cumulative incidence of grades III-IV aGvHD. (C) Cumulative incidence of chronic GvHD. (D) Cumulative incidence of severe GvHD. (E) Cumulative incidence of non-relapse mortality (NRM). (F) Cumulative incidence of relapse. (G) Probability of overall survival (OS). (H) Probability of progression-free survival (PFS). (I) Probability of graft-versus-host-free-relapse-free survival (GRFS).

Table 4.

Univariate and multivariate analysis of outcomes

	Univariate analysis		Multivariate analysis
	HR (95% CI)	p-value	HR (95% CI)	p-value
Grade II-IV aGvHD
PT-CY	1		1
PT-CY/BEN	0.71 (0.25, 2.01)	0.52	0.71 (0.24, 2.10)	0.53
Grade III-IV aGvHD
PT-CY	1		1
PT-CY/BEN	0.34 (0.04, 2.81)	0.29	0.28 (0.03, 2.71)	0.27
cGVhD
PT-CY	1		1
PT-CY/BEN	0.16 (0.02, 1.27)	0.049	0.08 (0.005, 1.11)	0.06
Severe cGVhD
PT-CY	1		1
PT-CY/BEN	0.00 (0.00, 0.00)	0.12	0.00 (0.00, 0.00)	0.99
NRM
PT-CY	1		1
PT-CY/BEN	0.00 (0.00, 0.00)	0.12	0.00 (0.00, 0.00)	0.99
Relapse
PT-CY	1		1
PT-CY/BEN	0.78 (0.21, 2.93)	0.71	1.15 (0.13, 10.03)	0.89
OS
PT-CY	1		1
PT-CY/BEN	0.51 (0.11, 2.37)	0.38	0.25 (0.02, 2.98)	0.27
PFS
PT-CY	1		1
PT-CY/BEN	0.49 (0.13, 1.71)	0.25	0.30 (0.06, 1.39)	0.12
GRFS
PT-CY	1		1
PT-CY/BEN	0.42 (0.14, 1.25)	0.11	0.22 (0.05, 0.86)	0.039

Non-relapse mortality, relapse, survival:

The median follow-up was 32 months (range 8 to 55) in patients treated with PT-CY/BEN and 32 months (range 13–74) in those receiving PT-CY. Non-relapse mortality was absent in patients receiving PT-CY/BEN, however 25% treated with PT-CY died (P=0.12), two from bacterial sepsis, two from complications of cGvHD and one secondary to multi-organ failure (Figure 2E). Relapse rates were comparable with 26.6% in PT-CY/BEN and 31.1% and PT-CY patients (Figure 2F). The overall survival (OS) at 1 and 3 years was similar at 100% and 77.1%, respectively for PT-CY/BEN and at 90.3% and 74.2%, respectively, for PT-CY (P=0.38) (Figure 2G). Progression-free survival (PFS) was also comparable at 83.9% and 73.4% with PT-CY/BEN versus 80.6% and 56.9% with PT-CY, for years 1 and 3, respectively (P=0.25) (Figure 2H). Finally, we observed a trend toward improved GRFS with PT-CY/BEN at 75% and 65.6% versus PT-CY, 53.6% and 40.4% (P=0.11) at 1 and 3 years, respectively (Figure 2I). In multivariate analysis GRFS was significantly improved (HR, 0.22; 95% CI, 0.05, 0.86; P=0.039) (Table 4) in patients treated with PT-CY/BEN compared to PT-CY alone. Similar to the univariate analysis, multivariate analysis did not reveal significant improvement in OS and PFS respectively for patients receiving PT-CY/BEN (HR, 0.25; 95% CI, 0.02, 2.98; P=0.27) and (HR, 0.30; 95% CI, 0.06, 1.39; P=0.12) (Table 4).

Infections:

CMV reactivation was frequently seen with a median time to peak viral load of 40 days post-haplo-BMT (range 8–53 days). The cumulative incidence in at risk patients (seropositive recipients and/or seropositive donors) receiving PT-CY/BEN was 36.4% compared to 57.7% of at-risk PT-CY patients (P=0.31) (Figure 3).

Figure 3.

Cumulative incidence of CMV viremia. Time to peak viral load.

BK viruria was detected in seven (22.6%) PT-CY patients having >5 × 10⁸ viral copies/ml and symptoms of BK hemorrhagic cystitis, compared to three PT-CY/BEN patients (23.1%) (P=0.97) (Table 2). None of the patients in either group had clinically significant reactivation of EBV, HHV-6 or adenovirus warranting therapeutic intervention.

There was no significant difference in the incidence of gram+ bacteremias, between PT-CY and PT-CY/BEN patients in the first two months after haplo-BMT with 29% and 38.5% respectively (P=0.54), developing at least one positive culture (Table 2). Interestingly, gram- bacteremias were not observed in PT-CY/BEN patients while 12.9% of PT-CY had positive blood cultures with all being Klebsiella pneumoniae (P=0.20). Similarly, no fungal infections occurred in PT-CY/BEN patients while one PT-CY patient developed Candida Kruseii fungemia.

Analysis of patients transplanted on the pediatric HCTT service:

Thirty-one pediatric and AYA patients aged 0.6–27.2 years (median 16.8 years) received their haplo-BMT on the pediatric HCTT service. Ten of these patients (32.2%) received PT-CY/BEN. NRM at 3 years was 6.5% and the cumulative incidence of relapse at 1 and 3 years was 6.9% and 18.6% (Figure 4A). The incidence of aGVHD grades II to IV and grades III to IV at 6 months was 43.6% and 14.4%, respectively (Figure 4B), while that of overall and severe cGVHD at 1 year was 23.3% and 7.7% (Figure 4C). OS, PFS and GRFS at 3 years were 85.6%, 76.1% and 58.2%, respectively (Figure 4D).

Figure 4.

Outcomes of pediatric, adolescent and young adult patients receiving haplo-BMT on the pediatric service. (A) Cumulative incidence of non-relapse mortality (NRM) and relapse. (B) Cumulative incidence of grades II-IV and III-IV acute GvHD. (C) Cumulative incidence of chronic and severe chronic GvHD. (D) Probability of overall survival (OS), progression-free survival (PFS), graft-versus-host-free relapse-free survival (GRFS).

Discussion

T-cell-replete haplo-BMT with PT-CY has rapidly become an established alternative to unrelated HCT^14,15,16. PT-CY following the Johns Hopkins reduced intensity conditioning (RIC) regimen has been associated with low rates of severe acute and chronic GvHD^17,18. However, about half of patients without aGvHD ultimately relapse¹⁹. This has prompted the use of MAC regimens^9,20–22, especially in younger patients, to reduce disease recurrence, but at a potential cost of increasing GvHD and non-relapse mortality (NRM). Another strategy being explored to lessen relapse rates is the dose reduction of PT-CY. Preclinical studies in mismatched murine BMT models by Kanakry et al. have demonstrated benefits of reducing the PT-CY dose²³. This led to a clinical trial by the same investigator at the NIH (NCT03983850) evaluating de-escalation of PT-CY to half the dose (25 mg/kg on days +3 and +4) of current standard. Presentation of their early findings indicated that a reduction in PT-CY dose maintained protection against severe acute GVHD while promoting more rapid engraftment and also decreased early post-transplant toxicity. Others have also reported on the feasibility of reducing the dose of PT-CY^24,25.

We have taken a different approach in evaluating whether partially replacing PT-CY with PT-BEN would be advantageous. Our preclinical research in mice supported the replacement of CY with BEN by demonstrating that PT-BEN was equally effective in preventing early GvHD and protecting against late GvHD, while enabling superior GvL when compared to PT-CY¹. BEN is a multipurpose agent, showing promise as a chemotherapeutic for a variety of cancers, as a conditioning regimen component for HCT, and as a lymphodepleting agent given prior to CAR-T cells or donor leukocyte infusions (DLI)^26–31. Since our initial publication and using various preclinical models, our laboratory has delineated several immunomodulatory properties of BEN elucidating its role in modulating GvHD and GvL^1–6. Irrespective of whether BEN is given pre- or post-transplant, we have reliably observed decreased GvHD, increased GvL, and significant changes in the proportion and phenotype of multiple immune cell types⁶. These have included effects of BEN on MDSCs and DC subsets^2,3 and generation of tolerant T-cells with a striking absence of GvHD, while preserving T-cell dependent GvL⁴. Additionally, our in vitro studies have revealed that BEN increases the suppressive functions of MDSCs, skews DC generation towards cDC1s, promotes DC Flt3 expression, increases B-cell production of IL-10, inhibits STAT3 phosphorylation, and suppresses B- and T-cell proliferation^1–6.

This work provided the foundation to further explore the potential of PT-BEN in a first in human Phase Ia clinical trial. The Phase Ia (NCT02996773), was a 3 + 3 dose escalation trial, which accrued pediatric and young adult patients with hematologic malignancies. Our interim analysis included the first three cohorts receiving PT-CY 50 mg/kg on day +3 and on day +4 PT-CY (mg/kg)/PT-BEN (mg/m²): 40/20, 20/60, and 0/90 (Table 2)⁷. PT-CY/BEN was well tolerated with no dose limiting toxicities. Since publication of our interim analysis, we have completed enrollment of Cohort 4. The first patient in Cohort 4 experienced secondary graft failure following early CMV reactivation. The third patient in the same cohort developed cytokine release syndrome (CRS) and responded to a single dose of Tocilizumab. CRS after PT-BEN was recently reported to occur in 70% of patients with advanced primarily refractory hematologic malignancies who received much higher doses of PT-BEN (total dose 140–280 mg/m²)³². Of significance however, is that 44% of the patients in this report did not receive tacrolimus or mycophenolate mofetil and 85% received peripheral blood stem cell (PBSC) rather than bone marrow grafts. CRS was observed in one of our three patients who received a dose of PT-BEN totaling 110 mg/m² but not in any patients in Cohorts 1–3 (PT-BEN 20–90 mg/m²) suggesting a possible increased risk of CRS with escalation of PT-BEN. Based on these two events, the prior dose level used in Cohort 3 (Day +3, CY 50 mg/kg and day +4, BEN 90 mg/m²) was deemed the maximal tolerated dose (MTD) as we transitioned to Phase Ib to further evaluate the safety and potential benefits of PT-CY/BEN compared to PT-CY alone, and provide sample size estimates for the design of a potential Phase II multi-site randomized trial. Our Phase Ia trial accrued patients up 45 years of age. Two standard of care MAC regimens at our institution were used in all patients. TBI-FLU is generally used for acute lymphoblastic leukemia (ALL) and busulfan, fludarabine and melphalan (BFM) for myeloid malignancies and other diagnoses^7,8,33. To include a comparable group in the current report, we have analyzed all patients under 45 years of age undergoing T-replete haplo-BMT with PT-CY and the same two MAC regimens (n=31).

PT-CY has been considered safe against hematopoietic stem cells as these cells express high levels of aldehyde dehydrogenase thus detoxifying CY^34,35. However, BEN has multiple activities as it contains a mechlorethamine group, butyric acid side chain and a benzimidazole ring. The alkylating properties provided by the mechlorethamine group are similar to CY, while the butyric acid increases BEN’s water solubility and the benzimidazole ring is believed to function as a purine analogue, affording antimetabolic characteristics^36–38. We were therefore cautious in designing the Phase Ia trial, only affording gradual progression in de-escalation of CY and escalation of BEN starting with day +4, while leaving the dose of CY on day +3 unchanged. Despite our small numbers, we observed progressively improved trilineage engraftment and decreased transfusion requirements with escalation of PT-BEN and de-escalation of PT-CY (Figure 1). Moreover, we are witnessing favorable trends to significant improvements with PT-CY/BEN compared to PT-CY in univariate and multivariate analyses with respect to cGvHD, and GRFS (Figures 2C, 2D, 2I) and Table 4. We should note however, that the cumulative incidence of cGvHD was higher in our PT-CY patients compared to studies utilizing RIC and even MAC regimens^9,18,39,40.

Viral infections contribute to substantial transplant-related morbidity and mortality in patients undergoing haplo-BMT, with CMV being the leading culprit^41–44. Our previously report of reduced incidence of CMV reactivation found in our interim analysis did not hold true with the additional PT-CY/BEN patients enrolled in the study⁷. While CMV viremia was still less common following PT-CY/BEN, this difference was no longer significant (Figure 3). Two of the 31 patients receiving PT-CY were on letermovir (one who reactivated CMV despite it) while none of the patients treated with PT-CY/BEN received CMV prophylaxis as it was not part of the study protocol at that time. All patients responded to antiviral therapy primarily with ganciclovir or valganciclovir, however, as noted above, CMV contributed to secondary graft failure in a PT-CY/BEN patient.

As BK hemorrhagic cystitis is a recognized complication of high-dose CY, one would expect an increased incidence of BK viruria compared to PT-CY/BEN patients. While we did not observe a significant difference between groups, it was notable that two of the three PT-CY/BEN patients that developed clinically evident BK viruria were from Cohort 1 (with the highest PT-CY dose 50 and 40 mg/kg days +3 and +4 respectively) (Table 2). Our results corroborate that BK is a significant complication of PT-CY which has been reported to occur in a third to half of patients undergoing haplo-BMT with PT-CY⁴⁵. If PT-BEN emerges as a safe alternative to PT-CY, it may have the added advantage of reduced hemorrhagic cystitis and renal complications associated with BK viremia. Regarding other viruses which may complicate transplantation, no patients in either group had clinically significant EBV, adenovirus or HHV6 reactivations. Ongoing immune reconstitution analyses from this trial may shed more light on potential differences in viral control.

Another potential drawback of high dose PT-CY is cardiotoxicity. A recent study found a higher incidence of cardiac events (left ventricular systolic dysfunction, arrhythmias, pulmonary edema, arrhythmias) occurring within the first 100 days post-transplant in patients receiving PT-CY⁴⁶. Not unexpectedly, the younger patients in our study did not reveal evidence of clinically significant cardiac adverse effects of PT-CY. The addition of PT-BEN was well tolerated without any major toxicities such as sinusoidal obstruction syndrome, thrombotic microangiopathy and no admissions to the intensive care unit (ICU).

We recently reported on our haplo-BMT experience in 21 pediatric and young adult patients and the current report expands this number to 31, which to date is the largest study from a single pediatric center in North America. The ten additional mostly high-risk patients, include two infants, and a 2-yr old with KMT2A⁺ AML, a patient with therapy associated AML following treatment of Ewings sarcoma, two patients with FLT3+ AML, a patient with ALL who had relapsed following a matched sibling HCT and CAR-T cell therapy and three patients with ALL in ≥ CR2. Three patients received PT-CY 40 mg/kg and PT-BEN 20 mg/m² on day +3 and PT-BEN 90 mg/m² on day +4 (Phase Ia, Cohort 4), while one received only PT-CY 50 mg/kg on day +3 and only PT-BEN 90 mg/m² on day +4 (Phase Ib). The remaining six patients received PT-CY on both days. The current report has expanded the median follow-up to 32 months from 25 previously. Our patients’ characteristics are comparable to a recent report of 29 pediatric and young adult patients undergoing haplo-BMT following MAC at Johns Hopkins⁴⁷. Engraftment was similar in the two studies as was time to a platelet count of >20,000. However, in our cohort, the time to ANC of >500 was considerably earlier, day +15 compared to +24, despite our patients receiving a slightly lower CD34+ cell dose infusion (4.1 versus 5.4 ×10⁶/kg). Thirty-two percent of our patients received PT-CY/BEN which is associated with an earlier time to an ANC >500 (Figure 1). We observed a higher incidence of aGvHD compared to the Symons study while the incidence of cGvHD appeared to be slightly less in our series. Finally, our OS and PFS at 3 years was 85% and 76%, respectively, compared to 79% and 69% in the Johns Hopkins report confirming that MAC followed by haplo-BMT and PT-CY is an effective approach for pediatric and AYA patients with high-risk hematologic malignancies.

In summary, our group has pioneered the use of PT-BEN, first in an experimental murine haplo-BMT model¹ followed by completion of a Phase Ia trial⁷. Although our clinical findings are preliminary, together with our published preclinical murine bone marrow transplant studies, they provide evidence that partially substituting PT-CY with PT-BEN enhances engraftment and may favor survival with a trend toward reduced cGvHD and consequently improvement in GRFS. With the trial proceeding to Phase Ib we can hopefully provide confirmation of the safety and efficacy of this post-transplant prophylactic platform and its immunomodulatory effects.

Highlights.

• Partial replacement of PT-CY with PT-BEN on day +4 (PT-CY/BEN) is well tolerated

• PT-CY/BEN is associated with earlier trilineage engraftment compared to PT-CY

• A trend toward less cGVHD is seen with PT-Cy/BEN in multivariate analysis (P=0.06)

• Multivariate analysis shows improved GFRS with PT-CY/BEN versus PT-CY (P=0.039)