Nonmyeloablative Haploidentical Bone Marrow Transplantation with Post-Transplant Cyclophosphamide for Pediatric and Young Adult Patients with High-Risk Hematologic Malignancies

Abstract

Lower intensity conditioning regimens for haploidentical blood or marrow transplantation (BMT) are safe and efficacious for adult patients with hematologic malignancies. We report data for pediatric/young adult patients with high-risk hematologic malignancies (n=40) treated with nonmyeloablative haploidentical BMT with post-transplantation cyclophosphamide (PT/Cy) from 2003–2015. Patients received a preparative regimen of fludarabine, cyclophosphamide, and total body irradiation. Post-transplant immunosuppression consisted of cyclophosphamide, mycophenolate mofetil, and tacrolimus. Donor engraftment occurred in 29/32 (91%), with median time to engraftment of neutrophils >500/μL of 16 days (range 13–22) and platelets >20,000/μL without transfusion of 18 days (range 12–62). Cumulative incidences of acute GVHD grades II–IV and grades III–IV at day 100 were 33% and 5%, respectively. Cumulative incidence of chronic GVHD was 23%, with 7% moderate-severe chronic GVHD according to NIH consensus criteria. Transplant related mortality (TRM) at 1 year was 13%. The cumulative incidence of relapse at 2 years was 52%. With a median follow-up of 20 months (range 3–148), 1-year actuarial overall and event-free survival are 56% and 43%, respectively. Thus, we demonstrate excellent rates of engraftment, GVHD, and TRM in pediatric/young adult patients treated with this regimen. This approach is a widely-available, safe, and feasible option for pediatric and young adult patients with high risk hematologic malignancies, including those with a prior history of myeloablative BMT and/or those with co-morbidities or organ dysfunction that preclude eligibility for myeloablative BMT.

Introduction

Allogeneic blood or marrow transplantation (alloBMT) is the only curative option for many pediatric and young adult patients with relapsed or refractory hematologic malignancies. Typically, high-risk patients have undergone several rounds of intensive chemotherapy, or have relapsed after prior BMT, increasing the likelihood that organ toxicity and/or reduced performance status would preclude a myeloablative conditioning regimen. Additionally, only 50% of those in need of an alloBMT have a suitable HLA-matched unrelated donor, and this number is as low as 20% in some minority populations[1]. On the other hand, haploidentical related donors are almost always identifiable. Combining a lower intensity conditioning regimen with partially-matched, haploidentical, related donor BMT (haploBMT) is therefore crucial for expanding the availability of BMT to this high-risk group of young patients.

Post-transplantation cyclophosphamide (PT/Cy) has gained worldwide acceptance as a highly effective method of graft versus host disease (GVHD) prophylaxis [2–10]. When given on days +3 and +4 after BMT, high-dose PT/Cy selectively depletes alloreactive T cells while preserving hematopoietic stem cells and quiescent memory T cells responsible for protection against common pathogens that are encountered post-transplant[11–14]. In adults, rates of engraftment, GVHD, and transplant related mortality (TRM) after nonmyeloablative haploBMT with PT/Cy are similar to those seen with HLA-matched related (MRD) and HLA-matched unrelated (MUD) donors[2, 3, 15–17]. There has been success using reduced intensity conditioning (RIC) BMT in pediatric patients using HLA-matched related, HLA-matched unrelated, and HLA-mismatched unrelated donors, with bone marrow, cord blood, or peripheral blood stem cells[18–21]. Herein, we report outcomes using nonmyeloablative (NMA) haploBMT with PT/Cy looking exclusively at pediatric and young adult patients with high-risk hematologic malignancies, transplanted at our institution from 2003–2015.

Materials and Methods

Patients

This study was approved by the Institutional Review Board (IRB) of The Johns Hopkins Hospital. We retrospectively reviewed all pediatric and young adult patients (ages 1 month to 25 years) with hematologic malignancies who underwent NMA alloBMT using a haploidentical related donor at The Johns Hopkins Hospital from January 1, 2003 through June 30, 2015. All patients age 18 years or older, and all guardians of patients age 17 or younger, gave consent for transplantation. Patients were either treated on a prospective IRB-approved institutional clinical trial (n=28, 70%), or following the current open study (n=12, 30%) if they were unable to be enrolled secondary to insurance coverage limitations, or as standard of care following study closure if they otherwise met eligibility criteria. Sixteen of these patients have been previously reported in other manuscripts describing outcomes in our adult population[2, 3].

Eligibility criteria included patients with high-risk leukemias and lymphomas as previously published[2, 3, 15], Eastern Cooperative Oncology Group performance status ≤2 or Lansky/Karnofsky score ≥60%, left ventricular ejection fraction ≥35%, forced expiratory volume in the first second and functional vital capacity ≥40% of predicted (≥60% of predicted after thoracic or mantle radiation), not on dialysis, and absence of uncontrolled infection. Morphologic complete remission (CR) was required for acute leukemias and partial remission (PR) or better for aggressive lymphomas. Donors were first-degree relatives or half-siblings whom were HLA-haploidentical based on high resolution typing at HLA-A, -B, -Cw, -DRB1, and -DQB1, as previously described[2]. Donor selection criteria, in order of priority, included: medical fitness, no anti-donor HLA antibody, no major ABO incompatibility, matched cytomegalovirus (CMV) immunoglobulin G serostatus, no minor ABO compatibility, and sex (male donor preferred for male patient).

Treatment

The majority of the patients (n=36, 90%) received a preparative regimen of fludarabine (30 mg/m² IV, days −6 to −2), cyclophosphamide (Cy, 14.5 mg/kg IV, days −6 and −5), and total body irradiation (TBI, 200 cGy, day −1) (Figure 1). Two patients received IV busulfan (0.8mg/kg IV every 12 hours) on days −6 to −3 instead of low-dose Cy because they were enrolled on a clinical trial investigating this alternative NMA preparative regimen. One patient received fludarabine and cyclophosphamide at the dosing described above along with alemtuzumab (test dose of 3 mg IV on day-14 followed by a dose escalation schedule of 10 mg/15 mg/20 mg IV on days −14, −13, −12) and melphalan (100 mg/m² IV on day −2) after failing to engraft after two prior cord blood transplants. One patient received alemtuzumab (20 mg IV on days −6 through −2) and fludarabine at the dosing described above, as per the previously published regimen for patients failing to engraft after myeloablative alloBMT[22]. All patients received either T-cell-replete bone marrow (n=38, 95%) or peripheral blood stem cells (n=2, 5%) from haploidentical related donors on day 0. All patients received Cy 50 mg/kg/dose IV on days +3 and +4, followed by mycophenolate mofetil (MMF) 15mg/kg/dose PO TID (maximum daily dose 3 gm/d) from days +5 through +35, and tacrolimus 0.015mg/kg/dose IV every 12 hours from day +5 through either day +60 or +90 (n=13, 32%) or day +180 (n=27, 68%), according to the clinical trial on which the patient was enrolled or following. The tacrolimus was transitioned to oral as tolerated, and the dose was adjusted to maintain a trough level between 5 and 15 ng/mL. Filgrastim was administered starting on day +5 until neutrophil recovery to ≥ 1.0 × 10⁹/L.

Figure 1.

Preparative regimen and graft-versus-host disease prophylaxis. Fludarabine 30 mg/m2 IV days -6 through −2, cyclophosphamide (Cy) 14.5 mg/kg IV days −6 and −5, total body irradiation (TBI) 200 cGy day −1, graft infusion day 0, Cy 50 mg/kg/dose IV days +3 and +4, mycophenolate mofetil (MMF) 15mg/kg/dose PO TID (maximum daily dose 3 gm/d) days +5 through +35, tacrolimus 0.015 mg/kg/dose IV every 12 hours day +5 through either day +60, +90, or day +180, and filgrastim 5 μg/kg/day day +5 through neutrophil recovery.

Routine supportive care measures were followed according to institutional standards, as previously described[2, 4]. Antimicrobial prophylaxis for pneumocystic jirovecii and fungus were given to all patients, as per institutional guidelines. Patients whom were at risk of cytomegalovirus (CMV) reactivation, defined as either the donor or recipient having positive CMV immunoglobulin G (IgG), received ganciclovir (500mg/m2) from Day 0 through engraftment and CMV polymerase chain reaction (PCR) levels were measured weekly until day +100. Preemptive therapy with either IV gancyclovir or oral valganciclovir was initiated when CMV reactivation was detected at more than 500 copies CMV/mL. Patients with a history of varicella zoster virus infection received acyclovir prophylaxis for 1 year post-transplantation.

Definitions of Disease Status and Clinical Outcomes

Neutrophil recovery time, or engraftment, was defined as the number of days from BMT to the first of 3 consecutive days with an absolute neutrophil count at or above 0.5 × 10⁹/L. Platelet recovery time was defined as platelet count greater than 20 × 10⁹/L without platelet transfusion in the preceding 7 days. Donor chimerism analysis was performed on days +30, +60, +90, +180, and +365 after BMT on peripheral blood, and on bone marrow as clinically indicated or per protocol. Mixed donor chimerism was defined as >5% and <95% donor chimerism, and full chimerism as ≥95% donor chimerism, in whole blood or bone marrow. Primary graft failure was defined as <5% donor chimerism in bone marrow by day +60. Secondary graft failure was defined as loss of donor engraftment (<5% donor chimerism) after achieving neutrophil recovery. Acute GVHD (aGVHD) was graded per standard criteria[23], and chronic GVHD (cGVHD) was graded per the 2005 National Institutes of Health Consensus Criteria[24]. Overall survival (OS) was defined as the time from BMT to death from any cause. Event-free survival (EFS) was defined as the time from BMT to death or relapse. Transplant-related mortality (TRM) was defined as death without disease relapse. Minimal residual disease (MRD) before transplant and relapse after transplant were defined as any disease detectable by flow cytometry, molecular, fluorescent in-situ hybridization (FISH), or cytogenetic analysis. For patients with acute lymphoblastic leukemia (ALL) the MRD cutoff used was 0.01%, and for patients with acute myeloid leukemia (AML) the cutoff used was 0.1%. The level of sensitivity of MRD detection for acute leukemia was variable based on the year of BMT and the type of leukemia, ranging in earlier years from 1/500 for AML and T-ALL and 1/1000 for pre-B ALL, to 1/10,000 for AML, T-ALL, and pre-B ALL in later years.

Statistical analysis

Descriptive statistics were used to summarize baseline patient and transplant characteristics. The probability of OS and EFS were estimated using the Kaplan-Meier method with 95% confidence intervals (CIs). Median follow-up time was calculated using the reverse Kaplan-Meier method. Cumulative incidences (CuI) of relapse, TRM, and GVHD were estimated by competing-risk analysis using Gray's method. Relapse and TRM were competing risks for each other. Death, relapse, and graft failure were competing risks for GVHD. Data were analyzed with the R program, version 2.12 (R Core Development Team, Vienna, Austria) and Prism version 7.0a (Graphpad software, La Jolla, CA).

Transplant outcomes were calculated for the population as a whole and also by subcategory including year of transplant (2003–2010 versus 2011–2015), age (<18 versus ≥18y), whether a patient was eligible for myeloablative conditioning or not, whether a patient had a prior BMT or not, and MRD positive versus negative. Unadjusted comparisons of EFS and OS between patient groups were performed using the log rank test[25].

Results

Patient, Donor, and Graft Characteristics

From January 1, 2003 until June 30, 2015, forty pediatric and young adult patients received a NMA haploBMT. Patient characteristics are summarized in Table 1. The median age at transplant was 20 years, ranging from 1 to 25 years. Diagnoses and disease status are listed in Table 1. There were no statistically significant differences in the number of acute leukemia patients between age groups <18 years and ≥18 years. Hodgkin lymphoma (HL) was the diagnosis in 9 patients ≥18 years and in 5 patients <18 years. Of the acute leukemia patients, 4 patients had detectable MRD. One primary refractory AML patient in CR1 transplanted in 2004 had MRD detected at <0.5%, which was the lower limit of detection at the time. Three ALL patients had detectable MRD: one in CR3 transplanted in 2005 at 4%, one in CR2 transplanted in 2008 at 3%, and one in CR2 transplanted in 2011 at 0.27%. Additionally, one patient with hepatosplenic T-cell lymphoma (HSTCL) transplanted in 2009 had MRD detectable at 0.2%, and one patient with CML in CP2 transplanted in 2015 had BCR-ABL1 detectable by PCR at 0.18%. Of the HL patients, one patient had primary refractory disease with bulk disease at the time of transplant, and the remainder of the patients had chemo-responsive disease without bulk.

Table 1.

Patient Baseline Characteristics

Characteristic	Number
Patient Median Age at BMT (range)	20 years (1–25 years)
Female gender	12 (30%)
Diagnosis
AML, CR1	4 (10%)
AML, CR≥2	5 (12.5%)
ALL, CR1	2 (5%)
ALL, CR≥2	7 (17.5%)
MDS	5 (12.5%)
HL, PR	11 (27.5%)
HL, CR	3 (7.5%)
Other*	3 (7.5%)
MRD status (for leukemias, CML, and NHL, n=21)
Pos	6 (29%)
Neg	15 (71%)
Prior MA BMT
Autologous	9 (22.5%)
Allogeneic	8 (20%)
Total	17 (42.5%)
CMV at risk	29 (72.5%)

^*Other = 1 each of mixed-lineage leukemia, CML, NHL

ALL, acute lymphoblastic leukemia; AML, acute myeloid leukemia; BMT, blood and marrow transplantation; CMV, cytomegalovirus; CR, complete remission/response; HL, Hodgkin lymphoma; PR, partial response; MA, myeloablative; MDS, myelodysplastic syndrome.

Donor and graft characteristics are summarized in Table 2. The median donor age was 40 years, ranging from 18 to 58 years. The donor source was bone marrow for 38 (95%) patients and peripheral blood stem cells (PBSCs) for 2 (5%) patients, both of whom had relapsed after prior myeloablative (MA) alloBMT with bone marrow as the donor source. There were 29 patients (72.5%) whom were at risk for CMV.

Table 2.

Donor and Graft Characteristics

Characteristic	Number
Donor Median Age at BMT (range)	40 years (18–58 years)
Female gender	20 (50%)
Female-into-male	11 (27.5%)
Mother-into-child	10 (25%)
Relationship of Donor
Parent	30 (75%)
Sibling	10 (25%)
ABO mismatch
Compatible	30 (75%)
Major	6 (15%)
Minor	4 (10%)
Graft source
Bone Marrow	37 (92.5%)
Peripheral Blood	3 (7.5%)
TNC/kg BM, median (range)	4.3 × 108 (0.07 – 15.7 × 108)
CD34+/kg, median (range)	4.39 × 106 (2.24 – 19.6 × 106)

ABO, blood type designation; BM, bone marrow; BMT, blood and marrow transplantation; kg, kilogram; TNC, total nucleated cell.

Eight patients (20%) had undergone a prior myeloablative alloBMT, and nine (23%) had undergone a prior autologous BMT (autoBMT), for a total of 17 patients (43%) whom had undergone prior myeloablation. Twenty five (62%) patients would have been eligible for a myeloablative alloBMT given that they were >6 months post prior myeloablative BMT and met standard MA organ function eligibility and performance status; however, they had no available HLA-matched donor, and thus they were enrolled on an open NMA haploBMT clinical trial. At the time of their enrollment, a MA haploBMT protocol was not available. Fifteen (38%) of patients were not eligible for MA conditioning based on organ function (n=12), relapsing <6 months from prior myeloablation (n=2), or performance status (n=1).

Outcomes

Engraftment/Chimerism

Neutrophil engraftment occurred at a median of 16 days (range 13–32 days). Platelet engraftment occurred at a median of 18 days (range 12–62 days). On competing-risk analysis, the cumulative incidence (CuI) of engraftment by day +60 was 94%, or 30 out of 32 evaluable patients. Of these thirty patients, all but two are >95% donor. The remaining two patients, both with Hodgkin lymphoma, have 80% donor chimerism, and are now each 2- and 3-years post-transplant, respectively, without any evidence of lymphoma.

Regarding the patients with graft failure, one patient received NMA haploBMT for a therapy-related neoplasm that developed after autoBMT for diffuse large B cell lymphoma. This patient went onto a second NMA haploBMT from a different donor four months later, and died from Stenotrophomonas sp sepsis at day +14, before engrafting. The second patient with gamma-delta T cell lymphoma in CR achieved 93% donor chimerism on day +30. The patient subsequently developed CMV viremia, was treated with valganciclovir, and by day +60 had no detectable donor cells in the bone marrow, as well as a small population of cells which were suspicious for relapse. He went on to unrelated cord blood transplant at an outside institution, and died from unknown causes 17 months after his initial BMT.

GVHD, Relapse, and TRM

The cumulative incidences (CuI) of grade 2 to 4 and grade 3 to 4 acute GVHD (aGVHD) at 6 months were 33% and 13%, respectively (Figure 2A). The estimated CuI at 2 years of any chronic GVHD (cGVHD) was 24%, and of moderate-severe cGVHD was 7% (Figure 2B). The estimated CuI of relapse at 2 and 3 years was 52% (Figure 2C). The estimated CuI of relapse in acute leukemias and HL at 2 years was 35% and 34%, respectively. The median time to relapse was 154 days, ranging from 31 to 720 days. Twelve of seventeen relapses occurred in patients ≥18 years old. The estimated CuI of TRM at 1 year was 13% (Figure 2D). Causes of TRM included non-infectious diffuse alveolar hemorrhage (n=3) and infection (n=2) (Table 3). None of the cases of TRM were related to GVHD. All TRMs occurred in patients aged ≥18 years.

Figure 2.

Cumulative incidences of A. Grades 2 to 4 and grades 3 to 4 acute GVHD; B. Overall chronic GVHD and moderate-severe GVHD; C. relapse; D. transplant-related mortality.

Table 3.

Causes of TRM

Cause of Death	Number (Percent)	Number associated with GVHD (percent)
Diffuse Alveolar Hemorrhage	3 (7.5%)	0 (0%)
Infection	2 (5%)	0 (0%)

Survival

The median follow up was 21 months, ranging from 4 to 148 months. The 1-year, 2-year-, and 3-year OS probabilities were 56%, 52%, and 52%, respectively, and the 1-year, 2-year-, and 3-year EFS probabilities were 43%, 32%, and 32%. The overall OS and EFS were 46% and 32% respectively (Figure 3).

Figure 3.

Cumulative incidence of Overall Survival and Event-Free Survival.

Sub-group analyses for OS were performed (Figure 4). The only factor significantly associated with OS outcome was age at BMT. Patients <18 years at the time of transplant had an OS of 72%, and those ≥18 years had an OS of 25%, p=0.02. There were several other factors associated with trends towards improved OS, though none reached significance with our smaller patient numbers. Patients without prior MA BMT, patients whom were MRD negative (MRD analysis included all acute and chronic leukemia patients as well as the patient with HSTCL, which universally involves the bone marrow[26]), and patients whom were transplanted in later years all trended towards improved overall survival. There was no difference in OS for patients who met eligibility for MA transplant versus those who did not.

Figure 4.

Cumulative incidence of Overall Survival, analyzed by sub-groups. A. Age at the time of BMT; B. Prior myeloablative alloBMT; C. Year of BMT; D. MRD status; E. Met eligibility for myeloablation.

Hospitalizations and Complications

Patients treated by the pediatric BMT service (n=24, 60%) were hospitalized starting from the beginning of the preparative regimen through engraftment and recovery from the acute side effects of BMT. The median time of hospitalization for the initial BMT was 28 days (range 10–60 days). Patients treated by the adult oncology service (n=16, 40%) received their transplants in the outpatient setting. Among all patients, there were 54 hospitalizations after initial discharge or after outpatient BMT, in 32 patients. The median number of hospitalizations or re-hospitalizations after initial discharge through day +100 was 1, ranging from 0 to 5. The median number of inpatient days after initial discharge or after outpatient BMT through day +100 was 11.5 days, ranging from 0 to 59. The most common cause of hospitalization was fever, with (33%) and without (25%) a documented source. Additional causes included acute GVHD (15%), relapse (7%), and other (20%).

There was one case of veno-occlusive disease without multi-organ dysfunction, as defined by the Baltimore criteria[27], in a patient with therapy-related MDS, which resolved with supportive care alone.

There were 7 total cases of hemorrhagic cystitis (HC) that developed at a median of 19 days post BMT (range 6–68 days): grade I (n=4), grade II (n=2), grade III (n=1), and no cases of grade IV, graded per standard criteria[28, 29]. All were positive for BK virus, and one patient also had adenovirus detected. Four resolved with supportive care, two were treated with intravenous cidofovir, including the patient with both BK virus and adenovirus, and one with norfloxacin.

There were three cases of DAH. One patient was admitted with acute respiratory decompensation and had bloody secretions lavaged during a bronchoscopy. All viral, fungal, and bacterial cultures from the bronchoalveolar lavage (BAL) and blood were negative. A second patient whom had been treated for aGVHD developed acute respiratory decompensation three months later. The autopsy and BAL cultures were all negative. The third patient had a history of CMV viremia which was successfully treated and was negative for a month. She developed acute respiratory decompensation, for which she was intubated and then was successfully extubated, then two weeks later had frank hemoptysis and rapidly decompensated. All blood and sputum cultures were negative, including for CMV.

Infections

All cases of infection are listed in Table 4. There were 23 episodes of documented bacterial infections in the first 100 days post-transplant. Most (n=14, 61%) were coagulase-negative staphylococcus bacteremia. There were three documented fungal infections. Two of these patients died, one from Candida krusei fungemia and one from mucormycosis infection of the sinuses and orbit with intracranial extension. The other was successfully treated for fungal pneumonia. There were 25 documented cases of viral infections, including 14 cases of CMV reactivation (48% of at-risk patients). Thirteen of the fourteen cases resolved with therapy, and one patient died from relapse of his underlying malignancy after presenting on day +33 with circulating blasts.

Table 4.

Infections

Infection	Number	Timing of Infection
Bacterial		Median day +20 (Range day −6 to +99)
Coagulase-negative staphylococcus bacteremia	14
Escherichia coli bacteremia	2
Lactobacillus bacteremia	2
Escherichia coli diarrhea	1
Other bacteremia*	4
Fungal
Candida krusei fungemia/pneumonia	1	Day +15
β-D-glucan positive pneumonia	1	Day +34
Invasive Mucormycosis	1	Day +84
Viral		Median day +33 (Range day +10 to +66).
CMV reactivation	14
BK virus HC	6
BK virus/adenovirus HC + viremia	1
Parainfluenza URI	3
RSV URI/LRTI	1

^*Corynebacterium diphtheriae, viridans streptococci, Enterococcus faecium, Fusobacterium nucleatum

CMV, cytomegalovirus; HC, hemorrhagic cystitis; LRTI, lower respiratory tract infection; RSV, respiratory syncytial virus; URI, upper respiratory infection.

Discussion

We report one of the first studies of NMA haploBMT with PT/Cy for pediatric and young adult patients with high-risk hematologic malignancies[30, 31], and the first using this preparative regimen. The current study builds upon the previously reported outcomes of 372 adult patients using NMA haploBMT and PT/Cy[3, 30, 31]. There is merit to independently describing the outcomes of younger patients after NMA haploBMT with PT/Cy. First, children, adolescents, and young adults (AYA) generally have better organ function and higher performance status at the time of BMT as compared to their older adult counterparts, and are often treated with similar if not the same intensive chemotherapy regimens at diagnosis and relapse. Additionally, children and AYA patients may have different diagnoses and disease risk factors as well as unique short- and long-term toxicities that only become apparent when studied separately from older adult cohorts.

Younger patients treated with a range of lower intensity BMT regimens, for both malignant and nonmalignant disorders, have been reported, with rates of TRM of 11% to 40%, relapse rates of 21% to 77%, and overall survival at 2, 3, and 5 years of 45% to 73% (Table 5) [18–21, 30]. We used a truly nonmyeloablative conditioning regimen, and our results are comparable to the published literature.

Table 5.

Comparison of results to other published studies

	This study (n=40)	Institutional, Adult1 (n=372)	National, Adult2 (n=50)	National, Pediatric3,4,5,6 (n=224)
OS (%)	1-year: 58 3-year: 52	3-year: 50	1-year: 62	2–5 years: 45–73
EFS (%)	1-year: 43 3-year: 32	3-year: 40	1-year: 48	2–5 years: 15–60
TRM (CuI %)	180 days: 13 1-year: 13	180 days: 8	1 year: 7	Overall: 13–32%
Relapse (CuI %)	1-year: 41 3-year: 52	3-year: 46	1-year: 45	1–5 years: 21–77
Acute GVHD II–IV (CuI %)	100 days: 33 180 days: 33	180 days: 32	100 days: 32	100 days: 20–37
Acute GVHD III–IV (CuI %)	100 days: 5 180 days: 5	180 days: 4	100 days: 0	100 days: 2
Chronic GVHD (CuI %)	1-year: 16 2-year: 24	2-year: 13	1-year: 13	Overall: 13–46

¹McCurdy et al, Blood 2015, 3024–3031

²Brunstein et al, Blood 2011, 282–288

³Bitan et al, Blood 2014, 1615–20

⁴Pulsipher et al, Blood 2009, 1429–36

⁵Satwani et al, BBMT 2013, 552–61

⁶Verneris et al, BBMT 2010, 1237–44

CuI, cumulative incidence; EFS, event-free survival; GVHD, graft-versus-host disease; OS, overall survival; TRM, transplant-related mortality.

In our group of high-risk patients, we observed a cumulative incidence of relapse at 2 years of 52%. Though this is higher than some other reports with different conditioning regimens and/or donor sources [19, 21, 31], our TRM is lower, placing more patients at risk for the competing event of relapse. Compared directly to a reduced intensity haploBMT with PT/Cy regimen for patients with active disease[30], our relapse rate is lower and our overall survival higher. Additional contributors to our incidence of relapse may include the presence of significant MRD or active disease (MDS) in some patients at the time of transplant, a well-recognized risk factor[32–34], and heavy pretreatment (43% with prior MA BMT). However, smaller numbers do not allow us to conclusively make these attributions. Upon further analysis, our CuIs of relapse for acute leukemia and HL patients were 35% and 34%, respectively, and four out of five of our MDS patients relapsed. Furthermore, we observed 12/17 relapses in patients ≥18 years, despite similar numbers of acute leukemia diagnosis and more HL patients ≥18 years. Potential biological and cancer specific differences between young adults and pediatric patients have been reported and need further exploration[35–37].

Satwani et al demonstrated that patients whom are MRD negative prior to RIC alloBMT may benefit from the lower toxicity of RIC regimens, while maintaining disease-free survival of 76%[20]; this remains to be tested in the haploidentical PT/Cy setting. Recently, the issue of regimen intensity was studied in a randomized controlled phase III BMT-CTN trial of RIC versus MA HLA-matched related or 7–8/8 unrelated donor BMT for adults ≥18 years with MDS and AML; this study was closed early due to lower PFS in the RIC group[38]. Relapse after alloBMT remains a significant issue in high-risk patients, and the role of conditioning regimen intensity for different disease groups needs to be further investigated prospectively in this age group.

Although our cohort is small and we do not have enough patients to power a direct comparison between our reported adult data with the pediatric/young adult data, we do report a similar CuI of TRM (13% in this study versus 8% in the reported adult data) and a higher CuI of chronic GVHD (24% versus 13%). Historically, adults have higher rates of TRM after alloBMT; however, with a TRM of only 8% in the adult population [3], it is possible that this low intensity regimen is equally safe in all age groups. It is also possible that selection bias increases our CuI of TRM and/or cGVHD, as many of our patients were highly pretreated (43% with prior MA BMT in this study versus 19% with prior autoBMT in the reported adult data) and/or had preexisting comorbidities.

This study has several limitations. The total number of patients is small, heterogeneous in terms of disease type, status, and prior therapies, and from a single institution. Though this study includes patients of all pediatric and AYA ages, there is a predominance of older patients, and hence the diseases being treated are not all representative of the typical pediatric malignancies. However, they are representative of the pediatric and AYA population as a whole[39]. Eighty-three percent of the reported patients in this study are those with standard diagnoses eligible for BMT, such as acute leukemia, chronic leukemia beyond CP1, relapsed NHL, and HL patients who relapsed after autologous BMT. Seven patients with HL had not undergone prior autoBMT and are thus more atypical, yet were eligible and received NMA haploBMT given high risk features such as primary refractory disease, persistent positron emission tomography (PET) scan positivity, and/or extranodal disease, and are transplanted with precedent[6, 40, 41]. It will be important in the future to identify HL patients that are likely to relapse after autoBMT and therefore might be upfront alloBMT candidates, even with the addition of brentuximab vedotin[42, 43].

We do not have a concurrent MA haploBMT control cohort since our group did not start performing MA haploBMT until 2009. Currently, MA conditioning is still considered the gold standard for the pediatric age group. Unlike the adult population, the TRM associated with MA BMT for pediatric and young adults is already quite low[44, 45]. Thus, selection bias is usually an issue with NMA regimens for younger patients as those receiving NMA conditioning usually do not qualify for MA alloBMT because of poor organ function, performance status, and/or closely timed prior myeloablation. Comparing outcomes of patients who are eligible for MA yet were treated with NMA haploBMT with those who received MA haploBMT will be an important future focus as more patients are accrued. Additionally, reducing the duration of post-BMT immunosuppression is likely an important strategy for decreasing morbidity and mortality, optimizing immune reconstitution, and incorporating strategies to prevent and/or treat relapse; we will continue to optimize the cessation of post-transplant immunosuppression in future trials.

PT/Cy is inexpensive, easy to administer, and easily transportable, making this a feasible regimen in parts of the world with limited resources. PT/Cy allows for the utilization of haploidentical related donors, which nearly guarantees a rapidly available bone marrow donor for every patient in need, as well as donor lymphocyte infusion (DLI) or antigen specific T cell infusion post-transplant[46]. This regimen can be administered in an outpatient infusion clinic, further lowering cost and improving quality of life.

In conclusion, NMA haploBMT with PT/Cy for children and young adults with high-risk hematologic malignancies is a safe, feasible, and potentially efficacious therapy. We have demonstrated acceptable rates of OS, TRM, and GVHD, with minimal complications in patients with prior MA BMT, as well as in those who were ineligible for MA BMT due to organ function and/or performance status. Relapse rates are comparable to those in the published literature with similar regimens. Further prospective trials with larger patient numbers investigating the intensity of preparative regimen in pediatric and young adult patients with high-risk hematologic malignancies are warranted, particularly for those with varying degree of MRD and/or active disease.

Key Points.

• Nonmyeloablative haplo-BMT with post-transplant cyclophosphamide in pediatric patients with high-risk heme malignancies is a safe and effective treatment.

• Outcomes for pediatric and young adult patients are comparable to published adult outcomes.

Highlights.

• Nonmyeloablative haplo-BMT with post-transplant cyclophosphamide in pediatric patients with high-risk heme malignancies is a safe and effective treatment.

• Outcomes for pediatric and young adult patients are comparable to published adult outcomes.