Bone marrow transplantation for sickle cell disease using posttransplantation cyclophosphamide and 400 cGy TBI

Key Points

• •RIC BMT with 400 cGy provides durable engraftment, low GVHD toxicity, and may preserve fertility.

Visual Abstract

Abstract

Reduced-intensity conditioning (RIC), haploidentical donors, and posttransplantation cyclophosphamide (PTCy) have overcome many barriers associated with bone marrow transplantation (BMT) for sickle cell disease (SCD). However, initial approaches had high graft failure rates. Our preliminary data suggested increasing the total body irradiation (TBI) dose from 200 to 400 cGy could improve engraftment. This study included patients with SCD aged 2 to 70 years undergoing BMT from November 2014 to January 2025. The regimen included antithymocyte globulin, fludarabine, cyclophosphamide, and single-fraction 400 cGy TBI. Graft-versus-host disease (GVHD) prophylaxis included PTCy, mycophenolate mofetil, and sirolimus. Outcomes measured were disease-free survival (DFS), graft failure, overall survival (OS), and GVHD incidence. A total of 43 patients (median age, 23 years) received a transplant; 34 (79%) had ≥2 indications for BMT, and all had ≥2 SCD comorbidities. Five-year OS probability was 95.5% (95% confidence interval [CI], 0.87-1.0) at a median follow-up of 2.43 years. DFS probability at 2 years was 94.5% (95% CI, 0.87-1.0), with only 2 (5%) graft failures. Two patients died late (2.5 and 6 years) after BMT. The cumulative incidence of grade 3/4 acute GVHD was 2.4% (95% CI, 0-0.07), and that of moderate-severe chronic GVHD was 7.3% (95% CI, 0-0.15). Median time to discontinuation of immunosuppression was 354 days. Of 27 female patients, 12 had return of menses and/or normalized gonadal function. RIC haploidentical BMT with 400 cGy maintained a low-toxicity profile, provides high rates of durable engraftment, and may preserve fertility. This regimen expands the availability of curative therapy for severe SCD. This trial was registered at www.ClinicalTrials.gov as NCT00489281.

Introduction

Sickle cell disease (SCD) is the most prevalent inherited hemoglobinopathy globally, affecting ∼100 000 individuals in the United States.¹ SCD is characterized by a single-nucleotide substitution in the β-globin subunit of hemoglobin (Hb), leading to red cell hemolysis and organ dysfunction. Individuals with SCD experience varying severity of chronic, debilitating disease.² Comprehensive care strategies have improved survival³; however, median survival of 48 years has not significantly changed in the last 2 decades.^4,5

Gene therapy and allogeneic bone marrow transplantation (alloBMT) are more available and used with curative intent for SCD. Historically, alloBMT using HLA-matched sibling donors has been the safest option when available.⁶ However, most patients lack an unaffected HLA-identical sibling donor,⁷ and fewer than a third of African American patients find a suitable match in unrelated donor registries.^8,9 Established approaches using myeloablative conditioning are often too toxic for older individuals with SCD given disease-related comorbidities present at the time of referral for curative therapy. Furthermore, permanent gonadal failure is an expected consequence of myeloablation. Reduced-intensity conditioning (RIC) regimens may offer less toxicity but historically have incurred a greater risk of graft rejection.¹⁰

These concerns have been addressed through efforts to expand related RIC haploidentical BMT (haploBMT) approaches for individuals with SCD. We previously showed that posttransplantation cyclophosphamide (PTCy) after haploBMT results in low rates of graft-versus-host disease (GVHD) and nonrelapse mortality; however, a 41% rate of graft failure was observed.¹¹ In a subsequent study, we observed improved engraftment in a cohort of 17 patients with hemoglobinopathies who underwent RIC haploBMT with an increased total body irradiation (TBI) dose of 400 cGy.¹² Here, we present results for a total cohort of 43 patients with SCD.

Methods

Study design

This Johns Hopkins University institutional review board–approved retrospective review included all BMTs for SCD done with TBI 400 cGy. Eight patients were in our prospective protocol (ClinicalTrials.gov identifier: NCT00489281). Once the NCT00489281 trial was completed, this approach¹² became our standard for SCD BMT. The following 35 patients received a transplant with this regimen. Of these, 6 patients also participated in the NCT01166009 trial. All patients in the aforementioned study gave written consent before BMT; age-appropriate assent was obtained from patients aged 7 to 17 years; 5 patients were previously reported.¹²

Participants

Eligibility has been reported previously.¹² Patients were aged 2 to 70 years with at least 1 relative eligible to donate, HLA haploidentical or matched. Genotypes included Hb SS, Hb S/β°-thalassemia, Hb S/β⁺-thalassemia, Hb SC, Hb SE, Hb SD, Hb SO-Arab, or Hb S/hereditary persistence of fetal Hb. Patients were required to have at least 1 severe SCD complication warranting alloBMT, including cerebrovascular accident (CVA), complicated or recurrent acute chest syndrome (ACS), recurrent vaso-occlusive crisis (VOC; >2 per year for the last 2 years), sickle retinopathy, osteonecrosis, red cell alloimmunization (>2 autoantibodies) in the setting of chronic transfusions, or transfusion dependence. The number of eligible patients who declined BMT was not captured.

Treatment

Treatment was as previously described (Figure 1).¹² Conditioning included IV antithymocyte globulin (ATG; rabbit) 0.5 mg/kg (day -9), 2 mg/kg (days -8 and -7), given on day −9 through day −7, IV fludarabine (30 mg/m2/day) given on days −6 to −2, IV cyclophosphamide given at 14.5 mg/kg on days −6 and −5, and TBI 400 cGy in a single fraction on day −1. Steroids were given to prevent reactions to ATG and could be repeated once 3 hours after the first dose if clinically indicated with a subsequent taper.

Figure 1.

Conditioning schema. RIC schema PTCy and 400 cGy TBI. MMF, mycophenolate mofetil; po, by mouth; Rx, regimen; tid, 3 times a day. Adapted from Bolaños-Meade et al.¹¹

GVHD prophylaxis consisted of PTCy 50 mg/kg per day on days +3 and +4, mycophenolate mofetil 15 mg/kg per dose (maximum 1 g) every 8 hours on days +5 to +35, and sirolimus with a target level of 5 to 15 ng/dL from day +5 until day +365 unless medication toxicity was observed, at which point tacrolimus was substituted for sirolimus. Since 2021, some adult patients discontinued sirolimus at day +180 if they had no evidence of GVHD.

Donors and grafts

Donors were relatives in good health with at least 1 matched HLA haplotype. HLA phenotyping was performed as previously described^11,13 and included HLA-A, HLA-B, and HLA-DRB1 loci at intermediate resolution level for screening. Selected donors were further typed to include HLA-C at intermediate resolution and DRB1 and DQB1 alleles at high resolution. Donors with sickle cell trait were allowed. Selection factors also included younger age, ABO compatibility, cytomegalovirus (CMV) serology, and matches without donor anti-HLA antibodies. Donor bone marrow (BM) was collected with a goal yield of 4 × 10⁸ nucleated cells per kg recipient ideal body weight (IBW) and infused on day 0. The marrow was unmanipulated except that major incompatible ABO grafts had red blood cells depleted by buffy coat preparation and minor ABO incompatible grafts had plasma removed.

Peri-BMT supportive care

Although not mandated by protocol, everyone received complete or partial exchange transfusions before BMT (goal Hb S percentage of <30%). Weekly surveillance for CMV reactivation was performed. Additional surveillance for Epstein-Barr virus, herpes simplex virus, and herpes zoster reactivation was done based on patient risk. Seizure prophylaxis with levetiracetam was given for the duration of sirolimus therapy, and platelet threshold for transfusion was 50 × 10⁹ cells per L. Magnesium levels were kept >2 mg/dL. All patients received antiherpetic, Pneumocystis, bacterial, and fungal prophylaxis, which was escalated for infection.¹⁴

Definitions of chimerism, graft failure, GVHD, and neutrophil and platelet recovery

Hematologic, chemistry, and Hb electrophoresis studies were performed at a clinical laboratory improvement amendments–certified laboratory at the Johns Hopkins Hospital. Engraftment of donor cells was evaluated by percentage of donor-host chimerism for recipients and measured by polymerase chain reaction analysis of total peripheral blood and isolated CD3⁺ cells. Unsorted and CD3⁺ chimerism was measured at +30, +60, +180, and +365 days, and then annually as clinically indicated. All available chimerism data were collected. Full donor chimerism meant the presence of >95% donor DNA; mixed chimerism meant that both donor and host DNA were detected. Primary graft failure was defined as undetectable donor DNA on at least 2 occasions (>1 week apart) in the peripheral blood or BM. Secondary graft failure was defined as undetectable donor DNA on at least 2 occasions (>1 week apart) in the peripheral blood or BM after initial donor engraftment and reoccurrence of SCD symptoms requiring SCD care. Acute GVHD and chronic GVHD were graded by established criteria.^15,16 Neutrophil recovery was defined as the 3 consecutive days of neutrophil counts of >0.5 × 10⁹ cells per L. Platelet recovery was determined by last platelet transfusion for a threshold of 50 × 10⁹ cells per L.

Study outcomes

The primary outcome of this study was disease-free survival (DFS; no graft failure, death, or recurrence of SCD). Secondary outcomes were overall survival (OS), incidence of GVHD, and toxicities from BMT. A subanalysis of these outcomes was performed for patients aged <18 years. Long-term health outcomes were reviewed, with a focus on reproductive health, secondary malignancy, and iron overload. Common Terminology Criteria for Adverse Events version 5.0 was used to grade adverse events.¹⁷

Statistical analysis

Baseline patient characteristics are reported as medians with relevant ranges and/or interquartile ranges (IQR) for continuous variables and count and frequencies for categorical variables. The Kaplan-Meier method was used to estimate DFS and OS. Cumulative incidence of acute and chronic GVHD with corresponding 95% confidence intervals (CI) were estimated using the Gray method implemented in R package “cmprsk.”^18,19 Death was the competing risk for acute and chronic GVHD. Statistical analyses were performed with R version 4.3.1.

Results

Patient and BMT characteristics

A total of 43 patients were enrolled between 26 November 2014 and 31 January 2025 and were included in this analysis (36 haploidentical, 7 HLA-matched). One patient (patient 35) had a history of previous failed BMT from graft failure 6 years before. There were 27 adults (median age, 27 yeas [range, 19-41]) and 16 children included in the study (median age, 12 years [range, 5-17]). The median follow-up time was 2.43 years (IQR, 1.04-5.89).

Thirty-four patients had ≥2 indications for BMT (79%), and all patients had ≥2 sickle cell–related comorbidities (Tables 1 and 2). The average hematopoietic cell transplantation–specific comorbidity index score was 3.3 (range, 0-8) for adults and 0.8 (range, 0-4) for children.²⁰ Recipient and donor characteristics are summarized in Table 1., Table 2., Table 3..

Table 1.

Adult recipient demographics

Patient ID	Genotype	Age	Sex	Ethnicity	Performance score	BMT indication∗	Sickle cell complications	Disease-modifying therapy	Chelation therapy	HCT-CI, 1-10†
1	SC	31	M	AA	100	Recurrent VOC, AVN		HU	No	1
2	SS	19	F	ME	70	Recurrent VOC, CVA, AVN	Iron overload, ACS, splenic infarct with splenectomy, bacteremia	HU, RCE	Yes	2
4	SS	30	F	AA	100	Repeat CVA, recurrent VOC	ACS, iron overload, cholelithiasis with cholecystectomy	HU	No	5
5	SS	30	M	AA	70	Recurrent VOC, recurrent ACS, AVN, alloimmunization, sickle retinopathy	DVT, chronic pain, iron overload, priapism	HU, RCE	No	7
6	S/α	26	F	ME	80	Recurrent ACS, recurrent VOC, AVN	Iron overload, PE, splenectomy, chronic pain	HU, RCE	No	0
7	SS	27	F	AA	100	Recurrent VOC, recurrent ACS, AVN		HU	No	1
8	SS	28	F	AA	90	Recurrent VOC, recurrent ACS, TD	Iron overload, endocarditis, urosepsis, meningitis, recurrent Clostridioides difficile colitis	HU, RCE	Yes	6
12	SS	26	F	AA	80	Recurrent ACS, AVN, alloimmunization	Iron overload, PE, life-threatening DVT	HU, RCE	Yes	1
18	SS	23	F	ME	90	Recurrent VOC, AVN	ACS, iron overload	HU, RCE	Yes	4
19	SS	22	F	AA	80	Recurrent VOC, AVN	ACS, chronic pain, iron overload	HU, RCE	Yes	0
20	SS	26	M	AA	90	Recurrent ACS, recurrent VOC, alloimmunization	Splenic sequestration	HU	No	6
22	SS	34	M	AA	80	Recurrent VOC, sickle retinopathy	ACS, splenomegaly, cholelithiasis with cholecystectomy	HU	No	0
23	S/β	19	F	ME	90	Recurrent VOC	Osteomyelitis	HU, RCE	Yes	0
24	SS	25	M	AA	90	ACS, recurrent VOC	Splenic infarct, iron overload, chronic pain	HU, RCE	Yes	5
26	SS	23	F	AA	90	CVA	ACS, VTE, iron overload, chronic pain	HU, RCE	Yes	8
28	SS	29	F	AA	70-80	CVA with Moyamoya disease	Iron overload, chronic pain	HU, RCE	No	7
30	SS	31	M	AA	100	Recurrent VOC, CVA, AVN	Cholecystectomy, hepatitis C infection, osteomyelitis, ACS, retinal hemorrhage, proteinuria	none	No	2
31	S/β	32	F	ME	70	Recurrent VOC	Chronic pain, ACS	HU, RCE	No	4
32	SS	28	F	AA	70	Recurrent VOC, CVA, sickle retinopathy	ACS, cholecystectomy, chronic pain	HU	No	0
33	SS	29	M	AA	100	Recurrent ACS, recurrent VOC, AVN	SCI, recurrent priapism	HU	No	5
34	SS	24	M	AA	100	Recurrent ACS, recurrent VOC	SCI, chronic pain	HU, RCE	No	0
36	SS	30	F	AA	70	Recurrent VOC, alloimmunization	Bacteremia, iron overload	HU, RCE	Yes	5
37	SS	24	M	AA	90	CVA, recurrent ACS, alloimmunization	Splenectomy, iron overload, cholecystectomy	HU, RCE	Yes	Not available
38	SS	25	F	AA	80	Recurrent ACS	AVN, restrictive lung disease, cholecystectomy, VOC	HU	No	3
40	SS	41	F	AA	80	Recurrent VOC, alloimmunization	VOC, iron overload, ACS	HU, RCE	Yes	3
41	SS	22	F	AA	80	Recurrent VOC, alloimmunization	ACS, splenic infarct, cholecystectomy	HU	No	7
43	SS	32	F	AA	80	Recurrent VOC, recurrent ACS, sickle retinopathy, AVN	Leg ulcer, ACS	HU	No	Not available
Median		27								3.3
Range		19-41								0-8

AA, African American; AVN, avascular necrosis; DVT, deep vein thrombosis; F, female; HCT-CI, hematopoietic cell transplantation–specific comorbidity index; HU, hydroxyurea; ID, identity; M, male; ME, Middle Eastern; PE, pulmonary embolism; RCE, chronic exchange transfusion; S/α, Hb S α-thalassemia; S/β, Hb S β-thalassemia; SC, Hb SC; SCI, silent cerebral infant; SS, Hb SS; TD, transfusion dependence.

^∗See “Methods” for study entry requirements.

^†HCT-CI, calculated at the bottom of that column, shows the median and range.

Table 2.

Pediatric recipient demographics

Patient ID	Genotype	Age	Sex	Ethnicity	Performance score	BMT indication∗	Sickle cell complications	Disease-modifying therapy	Chelation therapy	HCT-CI, 1-10†
3	SS	5	F	AA	100	CVA with Moyamoya disease, recurrent VOC		RCE	Yes	0
9	SS	12	M	AA	90	TIA, recurrent VOC, TD, sickle retinopathy	Sickle nephropathy, bacteremia, splenic sequestration, dactylitis	HU, RCE	No	2
10	SS	13	F	AA	80	Recurrent VOC, sickle retinopathy	ACS, aplastic crisis, cholelithiasis with cholecystectomy, splenic sequestration with splenectomy	HU, RCE	Yes	0
11	SS	16	M	AA	90	CVA with Moyamoya disease	Iron overload, VOC	HU, RCE	Yes	3
13	SS	11	M	AA	100	Recurrent VOC	Dactylitis	HU, RCE	No	0
14	SS	13	M	AA	90	CVA, AVN, alloimmunization	Atrial thrombus, VOC, iron overload, ACS	HU, RCE	Yes	0
15	S/β	17	M	ME	100	Recurrent VOC, recurrent ACS, AVN	Iron overload	HU, RCE	Yes	0
16	S/β	12	F	ME	100	CVA, recurrent ACS, AVN		HU	No	4
17	S/β	11	F	ME	60	Recurrent VOC, AVN	ACS, cholelithiasis with cholecystectomy	HU, RCE	No	0
21	SS	15	F	AA	100	CVA	Iron overload	HU, RCE	Yes	0
25	S/D	9	M	ME	90	ACS, recurrent VOC		None	No	0
27	SS	11	F	AA	90	Recurrent VOC	Dactylitis, ACS, cholelithiasis, splenic sequestration, chronic pain	HU, RCE	No	0
29	S/α	16	F	ME	90	Recurrent VOC, AVN, alloimmunization	Splenic sequestration, chronic pain	HU, RCE	Yes	0
35	SS	17	M	ME	100	Recurrent ACS, recurrent VOC, TD	Dactylitis, priapism, splenic enlargement, sickle hepatopathy, iron overload	HU, RCE	No	3
39	SS	8	F	AA	90	Recurrent VOC, recurrent ACS		HU	No	1
42	SS	9	F	AA	90	CVA, recurrent ACS	Iron overload, VOC	HU, RCE	Yes	0
Median		12								0.8
Range		5-17								0-4

AA, African American; AVN, avascular necrosis; CVA, cerebrovascular disease; F, female; HCT-CI, hematopoietic cell transplantation–specific comorbidity index; HU, hydroxyurea; ID, identity; M, male; ME, Middle Eastern; RCE, chronic exchange transfusion; S/α, Hb S α-thalassemia; S/β, Hb S β-thalassemia; S/D, Hb SD; SS, Hb SS; TD, transfusion dependence; TIA, transient ischemic attack.

^∗See “Methods” for study entry requirements.

^†HCT-CI, calculated at the bottom of the column, shows the median and range.

Table 3.

Donor demographics

Recipient ID	Donor ID	Donor ABO/Rh	Recipient ABO/Rh	Donor sex	Donor relationship	Donor age, y	HLA match	Donor/recipient CMV status	CD34 dose infused (cells per kg IBW)	Nucleated cells infused (cells per kg IBW)	CD3 dose infused (cell per kg IBW)
1	1	O+	O+	F	Sibling	34	10/10	−/−	3.31E+06	4.14E+08	3.79E+07
2	2	A+	O+	M	Parent	52	5/10	+/−	6.69E+06	5.77E+08	7.47E+07
3	3	O+	O+	F	Parent	26	5/10	+/+	4.02E+06	5.13E+08	6.77E+07
4	4	O+	O+	M	Parent	60	5/10	+/−	2.92E+06	6.64E+08	4.05E+07
5	5	O+	O+	M	Sibling	23	5/10	+/+	9.74E+05	3.75E+08	2.49E+07
6	6	A+	A+	M	Parent	53	5/10	+/+	2.02E+06	5.94E+08	3.88E+07
7	7	B+	O+	F	Sibling	27	5/10	+/+	3.30E+06	2.20E+08	2.83E+07
8	8	O−	O−	F	Aunt	53	5/10	−/−	3.97E+06	4.10E+08	3.06E+07
9	9	B	AB+	F	Sibling	15	5/10	+/−	4.09E+06	4.30E+08	5.52E+07
10	10	O+	O+	F	Parent	51	5/10	+/−	5.58E+06	5.10E+08	4.51E+07
11	11	O+	A+	M	Sibling	14	5/10	−/−	2.53E+06	4.50E+08	2.88E+07
12	12	O+	O+	F	Parent	57	5/10	+/−	4.67E+06	5.10E+08	4.50E+07
13	13	O+	O+	F	Parent	37	5/10	+/+	4.60E+06	5.40E+08	5.63E+07
14	14	O+	A+	F	Parent	51	5/10	−/−	9.77E+05	3.10E+08	2.99E+07
15	15	O+	O+	F	Parent	39	5/10	+/+	1.08E+07	5.10E+08	4.95E+07
16	16	O+	O+	M	Parent	39	5/10	+/+	8.06E+06	5.90E+08	7.82E+07
17	17	O+	O−	M	Parent	44	5/10	+/+	4.04E+06	5.50E+08	3.83E+07
18	18	AB+	AB+	M	Sibling	35	10/10	+/+	3.15E+06	5.40E+08	2.63E+07
19	19	B+	B+	F	Sibling	13	10/10	+/+	3.52E+06	6.60E+08	2.34E+07
20	20	O+	O+	F	Sibling	26	5/10	−/−	4.66E+06	4.40E+08	3.02E+07
21	21	AB+	B+	F	Parent	45	5/10	−/−	1.02E+07	4.30E+08	1.97E+08
22	22	O+	A+	F	Sibling	19	5/10	+/+	5.93E+06	5.50E+08	3.24E+08
23	23	O−	O+	M	Sibling	14	5/10	+/+	5.50E+06	6.50E+08	3.17E+08
24	24	O+	O−	M	Cousin	25	5/10	−/−	1.25E+07	3.40E+08	1.12E+08
25	25	O+	O+	F	Sibling	12	10/10	+/+	7.05E+06	9.50E+08	4.77E+07
26	26	O+	O+	M	Sibling	27	5/10	+/+	8.98E+06	4.20E+08	3.92E+07
27	27	A+	A+	M	Parent	42	5/10	+/+	1.38E+07	4.20E+08	1.36E+08
28	28	O−	O+	M	Sibling	36	5/10	+/+	3.07E+06	4.00E+08	1.68E+07
29	29	O+	O+	F	Cousin	19	5/10	−/+	3.38E+06	4.00E+08	2.96E+07
30	30	B+	O+	M	Sibling	26	5/10	+/+	4.02E+06	4.79E+08	3.35E+07
31	31	O+	O+	F	Sibling	26	5/10	+/+	4.10E+06	3.87E+08	4.19E+07
32	32	A+	A+	F	Parent	46	5/10	+/−	5.53E+06	4.76E+08	5.35E+07
33	33	A+	A+	M	Sibling	44	10/10	+/−	4.82E+06	3.83E+08	4.37E+07
34	34	B+	B+	F	Cousin	33	5/10	−/+	8.52E+06	4.26E+08	6.32E+07
35	35	A+	A+	M	Cousin	15	5/10	+/−	4.71E+06	4.09E+08	4.48E+07
36	36	O+	O+	M	Brother	25	10/10	−/−	6.19E+06	4.80E+08	5.48E+07
37	37	A+	A+	M	Brother	19	5/10	−/+	7.99E+06	4.90E+08	5.79E+07
38	38	O+	A+	F	Mother	58	5/10	+/+	4.96E+06	3.37E+08	3.52E+07
39	39	O+	O−	F	Mother	37	5/10	+/−	1.14E+07	6.86E+08	5.62E+07
40	40	O−	B+	F	Sister	46	10/10	?/+	6.91E+06	4.73E+08	5.67E+07
41	41	O+	A+	F	Sister	33	5/10	+/−	9.02E+06	5.37E+08	5.34E+07
42	42	O+	O+	M	Brother	19	5/10	−/−	8.59E+06	6.56E+08	9.51E+07
43	43	B+	B+	M	Brother	36	5/10	+/+	6.92E+06	4.38E+08	4.21E+07
Median						34			4.82E+06	4.76E+08	4.50E+07
IQR						24-45			3.75E+06 to 7.76E+06	4.12E+08 to 5.48E+08	3.44E+07 to 5.76E+07

Patient 5 had DQA∗05/DQ7 weakly positive anti–HLA donor-specific antibodies.

F, female; ID, identity; M, male.

There were 3 major and 5 minor ABO-mismatched donors to the recipients (Table 3). Major ABO mismatches were avoided for all but these 3 cases due to limited donor availability. Patient 2 had an HLA-matched sibling donor with insulin-dependent diabetes potentially conferring a >50% risk of insulin-dependent diabetes in the recipient²¹; this patient developed anti-erythroid antibodies requiring lifelong simple transfusions. Patients 7 and 21 had no complications from ABO mismatch.

The median total nucleated cell dose was 4.76 × 10⁸ nucleated cells per kg recipient IBW (IQR, 4.12 × 10⁸ to 5.48 × 10⁸ nucleated cells per kg IBW). Seven patients received <4 × 10⁸ total nucleated cells per kg recipient IBW. All patients received a graft containing >2 × 10⁶ nucleated cells per kg recipient IBW as considered the minimum for a quality BM graft,²² and the CD34 dose for all but 2 patients were >2 × 10⁶ cells per kg. These 2 patients experienced graft failure (patients 5 and 14). Donor-specific anti-HLA antibodies were recognized in 1 patient (patient 5) with low-level DQA∗05/DQ7.

DFS

All patients had chimerism data from day +180 or later, except patient 31 (transitioned back to international provider) and patient 5 (primary graft failure by day +70).¹² The Kaplan-Meier–based probability of OS at 5-years was 95.5% (95% CI, 0.87-1.0; Figure 2). The 2-year probability of DFS was 94.5% (95% CI, 0.87-1.0; Figure 2), and the 5-year probability of DFS was 89.8% (95% CI, 0.79-1.0).

Figure 2.

DFS and OS. Probability by the Kaplan-Meier method of DFS at 2 years of 94.5% (95% CI, 0.87-1.0), represented by the solid line. Probability of OS at 5 years, 95.5% (95% CI, 0.87-1.0), represented by the dashed line.

Two patients (5%) died; patient 1 died at home +885 days after BMT from a suspected overdose, and patient 2 died +2101 days after BMT with an unclear cause of death in the setting of inadequate donor erythropoiesis (Table 4).

Table 4.

Transplantation outcomes

Recipient ID	Neutrophil engraftment, d	60-d chimerism (%)	60-d T-cell chimerism (%)	Chimerism (%)∗	T-cell chimerism (%)∗	Last chimerism, days from BMT	DFS	Acute GVHD, grade	Chronic GVHD, severity†	IST	End of IST, days from BMT§
1	30	100	100	100	100	456	Death at 885 d from suspected overdose	None	None	Sirolimus	365
2	30	100	100	25	79	1681	Death at 2101 d from unknown cause	1	None	Sirolimus	639
3	17	100	100	100	100	371	-	3	Severe; patchy hyperpigmented skin without lichen/sclerosis	Tacrolimus	508
4	25	100	100	100	100	419	-	1	None	Sirolimus	420
5	Did not engraft	0	0	0	0	62	Primary graft failure at 60 d	None	None	Sirolimus	70
6	64	91	60	20	70	772	-	None	None	Sirolimus	Not available; prolonged use for low mixed chimerism
7	30	100	100	100	100	762	-	None	None	Sirolimus	370
8	25	100	100	100	100	760	-	None	None	Sirolimus	346
9	33	100	100	100	100	1182	-	None	None	Sirolimus	372
10	32	100	100	100	100	357	-	None	None	Sirolimus	357
11	23	100	>95	100	100	356	-	None	None	Sirolimus	362
12	21	100	100	100	100	671	-	None	None	Sirolimus	307
13	26	94	86	18.3	29.9	2496	-	None	Mild	Sirolimus	Not available; prolonged use for low mixed chimerism
14	33	78	100	2	22	382	Waning chimerism over 1 y; had recurrent VOC and episode of DVT with PE	None	None	Sirolimus	390
15	23	>95	100	100	100	399	-	None	None	Sirolimus	399
16	35	100	100	100	100	196	-	1	None	Sirolimus	Not available
17	29	100	100	100	100	1456	-	2	Severe; pulmonary FEV1 24%, FVC 36%, ocular	Sirolimus; ibrutinib, ruxolitinib	959
18	20	100	91	100	100	179	-	None	Moderate	Sirolimus	Not available
19	19	100	94	100	>95	727	-	None	Severe; sclerodermatous skin, pulmonary with severe restrictive disease	Sirolimus, tacrolimus; ruxolitinib, steroids	615
20	22	100	100	100	>95	799	-	None	None	Tacrolimus	194
21	31	100	100	100	100	529	-	2	Severe; extensive serositis with skin lichenification and pleural effusions	Sirolimus, tacrolimus; prednisone, MMF, ruxolitinib, ECP	Not available; intermittent use for chronic GVHD flares
22	32	100	100	100	100	767	-	4	None	Sirolimus; tacrolimus	195
23	30	94	100	87	>95	766	-	None	None	Sirolimus	342
24	33	>95	100	100	100%	354	-	None	None	Sirolimus	354
25	27	94	68	>95	92	874	-	None	None	Sirolimus	364
26	22	100	100	100	100	836	-	None	None	Sirolimus	353
27	19	100	>95	100	100	731	-	None	None	Sirolimus	290
28	35	>95	100	>95	100	777	-	1	None	Sirolimus	390
29	28	100	100	100	100	735	-	None	None	Sirolimus	300
30	27	100	100	100	100%	356	-	None	None	Sirolimus	362
31	48	100	>95	100	>95	55	-	None	None	Sirolimus	181
32	34	100	87	100	100	347	-	None	None	Sirolimus	186
33	25	>95	82	100	93	397	-	None	None	Sirolimus	336
34	26	100	100	100	100	168	-	None	None	Sirolimus	156
35	35	94	36	>95	90	377	-	None	None	Sirolimus	391
36	26	100	>95	100	100	168	-	None	None	Sirolimus	174
37	24	>95	92	>95	94	391	-	None	None	Sirolimus	365
38	36	100	100	100	100	175	-	2	None	Sirolimus	165
39	24	100	100	100	>95	367	-	None	None	Sirolimus	367
40	22	100	>95	100	>95	179	-	None	None	Sirolimus	179
41	46	>95	100	100	>95	167	-	2	None	Sirolimus	202
42	24	100	100	100	100	175	-	None	None	Sirolimus	N/A
43	38	100	>95	100	>95	186	-	None	None	Sirolimus	186

All chimera obtained from peripheral blood.

BSA, body surface area; DVT, deep vein thrombosis; ECP, extracorporeal photopheresis; FEV1, forced expiratory volume in 1 second; FVC, forced vital capacity; ID, identity; IST, immunosuppression; MMF, mycophenolate mofetil; PE, pulmonary embolism.

^∗Most recent.

^†Arabic numeral designates chronic GVHD scoring.

^§N/A describes 1 pediatric patient who is less than +365 days and continue on IST.

Graft failure

Two patients (5%), had primary or secondary graft failure with recurrent SCD symptoms (Table 4). Patient 5 had primary graft failure with undetectable donor chimerism at day +60 and received ATG 160 mg/kg salvage, which did not affect outcome. Patient 14 had waning chimerism over 1 year with recurrent VOC and an episode of deep vein thrombosis with pulmonary embolism requiring reinitiation of hydroxyurea. Both patients had full recovery of host hematopoiesis. They continue to be followed up by their hematologist and were alive at last follow-up.

Of patients who achieved engraftment (n = 41), 37 patients (86%) achieved full (≥95%) donor chimerism and 4 patients (9.3%) achieved stable mixed donor chimerism (Table 4). The median time to most recent peripheral blood chimerism for these patients was 399 days (IQR, 272-767). All 7 patients with an HLA-matched donor had 100% donor whole-blood chimerism at day +60. Among patients who achieved engraftment who reached day +365 (n = 35), 30 patients were documented to be off immunosuppression at the day +365 visit. For those who achieved engraftment and had a donor with sickle cell trait (n = 27), Hb S was <50% at first electrophoresis after BMT (typically day +180). Hb S was undetectable at first follow-up electrophoresis (typically day +180) for all others who engrafted (n = 7).

Hb levels improved, with mean values of 13.5 g/dL for male patients (baseline, 10.3 g/dL) and 12.4 g/dL for female patients (baseline, 8.9 g/dL) at day +180 or greater. Hb S concentration and markers of hemolysis all improved from day +180 or greater with median values: lactate dehydrogenase 371 to 267 units/L, absolute reticulocyte count of 249 × 10⁹ to 93 × 10⁹ cells per L, and total bilirubin 1.8 to 0.3 mg/dL (Table 5). Median time to independence of red cell transfusion was 34 days (IQR, 23-46) and median number of red cell units transfused after BMT was 3 units (IQR, 2-5). Patient 2, a recipient of a major ABO-mismatched graft had persistent anti-A antibodies leading to transfusion dependence until the patient’s death at 6 years after BMT. The recipients of the other 2 ABO incompatible transplants both had 100% donor chimerism and achieved transfusion independence by day +36 and +40, respectively.

Table 5.

Hematologic parameters collected at day +180 or later

Recipient ID	Donor genotype	Pre-BMT recipient electrophoresis		Post-BMT electrophoresis, 180 d		Hb (g/dL)		LDH (U/L)		Absolute reticulocyte count (×103/mm3)		Total bilirubin (mg/dL)		Ferritin (ng/mL)		Post-BMT therapeutic phlebotomy	Post-BMT red cells transfused, units	Last RBC transfusion, days after BMT	Last platelet transfusion, days after BMT
		Hb S (%)	Hb F (%)	Hb S (%)	Hb F (%)	Pre-BMT	Post-BMT	Pre-BMT	Post-BMT	Pre-BMT	Post-BMT	Pre-BMT	Post-BMT	Pre-BMT	Post-BMT
1	NA	47.3	0.0	0.0	0.0	12.9	17.3	196	155	122	56	5.9	0.8	948	175	No	4	41	28
2	Hb SA	34.1	12.6	39.2	12.0	11.4	12.5	556	332	81	49	1.2	0.5	1 224	7493	Yes	155	2095	26
3	Hb SA	28.4	3.5	37.2	0.0	9.4	12.4	488	321	468	74	1.6	0.3	2 052	418	Yes	16	92	102
4	Hb SA	46.1	5.4	35.1	1.7	7.8	11.0	495	224	56	81	7.3	0.7	5 449	1814	Yes	6	55	38
5	Hb SA	87.9	4.3	<5	0.0	7.5	10.3	523	436	317	201	2.4	1.5	2 815	977	Yes	12	86	97
6	Hb SA	51.0	4.5	76.2	9.9	9.9	8.7	219	226	67	62	0.6	<0.2	53	39	No	4	417	66
7	Hb SA	51.6	10.7	31.6	0.0	10.5	12.1	314	158	273	88	6.5	0.5	328	97	No	4	40	28
8	NA	67.5	4.4	0.0	0.0	7.3	12.8	632	269	399	191	4.4	<0.2	1 991	952	Yes	3	28	18
9	Hb SA	25.2	5.2	31.5	1.6	10.0	15.0	397	208	227	56	1.4	<0.2	814	264	No	4	33	30
10	Hb SA	88.9	5.0	37.1	0.0	6.5	15.6	346	229	394	126	4.3	<0.2	663	123	No	0	0	19
11	NA	51.0	11.3	0.0	0.0	10.5	14.7	428	419	300	70	7.7	0.3	628	534	No	1	24	31
12	Hb SA	68.1	17.3	38.6	0.0	7.8	13.3	170	169	236	122	1.0	0.3	1 963	788	Yes	3	24	14
13	Hb SA	76.1	11.6	39.4	0.0	7.8	11.7	407	322	235	123	2.2	0.3	1 140	201	No	1	9	13
14	NA	37.9	1.6	77.8	18.9	10.4	12.2	491	199	259	318	4.5	1.8	292	428	No	12	47	71
15	Hb SA	63.9	1.7	26.0	<1	10.4	11.3	266	183	216	108	1.4	0.2	4 893	1484	Yes	6	22	23
16	Hb SA	85.7	5.3	26.7	2.3	7.7	12.8	339	319	315	78	1.4	0.3	1 633	583	Yes	3	20	32
17	Hb SA	76.7	15.3	24.4	1.6	9.4	15.3	409	203	158	84	1.4	<0.2	102	66	No	7	105	199
18	Hb SA	51.3	2.0	37.7	2.7	10.0	12.5	195	289	424	139	2.4	0.3	3 637	2565	Yes	2	22	21
19	Hb SA	39.7	3.7	33.3	<1	9.2	12.4	404	262	144	111	0.4	<0.2	728	93	No	2	582	575
20	Hb SA	70.7	18.9	35.0	<1	10.8	13.6	271	NA	228	98	1.6	0.7	807	375	No	2	20	22
21	Hb SA	23.8	10.4	35.5	7.5	10.2	10.3	279	379	137	138	0.9	0.4	2 589	1773	Yes	4	36	73
22	NA	68.0	10.5	0.0	1.3	7.1	14.4	623	183	102	55	2.2	0.5	753	2172	No	6	47	60
23	Hb SA	60.5	8.7	33.9	2.0	8.1	10.6	263	247	185	82	1.4	0.5	3 604	1550	Yes	3	50	25
24	NA	83.3	10.7	39.8	<1	7.5	12.2	868	307	569	11	6.5	0.4	755	814	No	5	41	42
25	Hb SA	38.5	12.5	38.4	1.2	8.0	11.7	703	302	95	67	2.0	0.3	89	120	No	1	27	38
26	Hb SA	23.2	<1	42.7	<1	8.5	11.0	499	260	442	87	3.2	0.4	14 923	4584	Yes	2	6	13
27	Hb SA	63.6	9.9	38.7	<1	8.0	13.5	299	365	147	201	1.3	0.3	1 557	880	No	4	129	25
28	Hb SA	41.1	<1	39.2	<1	8.2	13.6	309	375	252	86	2.1	0.3	4 801	3563	No	2	23	32
29	Hb SA	28.8	7.0	27.6	1.0	11.7	11.1	225	469	70	103	0.8	0.3	1 180	1179	No	5	22	25
30	Hb SA	88.0	5.0	37.1	0.0	11.0	16.1	326	362	412	118	4.5	0.4	412	82	No	1	27	19
31	Hb SA	30.1	5.7	28.5	0.0	9.8	12.0	404	NA	197	34	1.4	<0.2	197	NA	No	2	40	63
32	Hb SA	70.2	9.1	33.9	1.1	8.8	12.5	249	NA	446	90	1.8	<0.2	446	NA	No	3	34	36
33	Hb SA	48.6	2.9	31.2	1.0	10.2	16.5	402	NA	290	98	2.2	2.0	800	195	No	2	28	28
34	Hb AA	34.7	<1	0.0	<1	10.5	15.3	201	NA	100	144	0.4	0.4	18	17	No	2	31	28
35	Hb SA	35.3	1.5	34.6	1.0	9.8	13.4	428	267	428	102	1.5	1.9	989	1122	Yes	2	36	43
36	Hb SA	49.70	7.2	NA	NA	8.3	14.1	375	NA	286	157.5	3.4	0.6	1 615	8274	Yes	2	21	25
37	Hb AA	34.00	1.1	0.0	<1	10.9	13	167	556	285	83.6	1.2	0.8	14	21	No	0	NA	26
38	Hb SA	67.30	27.8	33.1	2.7	10.3	11	218	NA	264	130.9	1.5	0.2	959	1189	No	3	40	30
39	Hb SA	76.50	16.6	36.9	1.2	9.6	13.9	430	216	166	63.6	1.1	<0.2	251	353	No	2	24	24
40	Hb SA	76.70	18.5	NA	NA	8.9	12.0	275	NA	309	107	3.3	0.4	1 306	649	No	2	23	23
41	Hb AA	80.80	14.1	33.2	1.9	8	12.7	297	NA	274	75.3	2.3	<0.2	1 076	1743	No	6	41	44
42	Hb AA	8.10	<1	NA	NA	8.3	11.3	432	NA	147	68.6	1.8	<0.2	3 706	2933	Yes	1	15	49
43	Hb SA	36.00	1.5	NA	NA	9.3	9.6	371	NA	249	95.8	1.1	<0.2	260	222	No	10	39	42
Median		51.0	7.1	33.9	1.3	9.4	12.5	371	267	249	93	1.8	0.3	959	649		3	34	30
IQR		36-70	3.5-11.3	27.2-38.1	0-2.4	8-10-4	11.5-13.8	271, 432	225-330	147-312	71.3-123	1.4-3.3	0.3-0.5	492-1 977	195-1550		2-5	23-46	25-43

Post-BMT parameters were collected at ≥6 months based on last available laboratory values collected for the patient. Hb F, fetal Hb; Hb SA, sickle cell trait; ID, identity; LDH: lactate dehydrogenase, NA, not available; RBC, red blood cell.

GVHD

The 1-year cumulative incidence grade 2 to 4 GVHD was 11.6% (95% CI, 0.02-0.21). Grade 3 to 4 acute GVHD was 2.4% (95% CI, 0-0.07; Figure 3). All individuals were at risk for chronic GVHD at 2 years after BMT, and the 2-year cumulative incidence of moderate-severe chronic GVHD was 7.3% (95% CI, 0-0.15; Figure 3). Table 4 shows detailed information for all patients with GVHD with description of grade 3 to 4 acute GVHD and severe chronic GVHD.

Figure 3.

GVHD. Cumulative incidence of grade 3/4 aGVHD at 1 year after transplantation was 2.4% (95% CI, 0-0.07; left graph). The cumulative incidence of severe cGVHD at 2-years after transplantation was 7.3% (95% CI, 0-0.15; right graph). aGVHD, acute GVHD; cGVHD, chronic GVHD.

Median time to discontinuation of immunosuppression for patients who reached day +180 was 356 days (IQR, 199-381; Table 4). Ten adult patients who received transplantation after February 2021 discontinued immunosuppression at +180 days and had no subsequent development of chronic GVHD. Delay in discontinuation of immunosuppression was due to persistent GVHD at 1 year after BMT (n = 5) or presence of low mixed chimerism (n = 2). For patients who did not have a date of discontinuation, the reasons for this included transition to another provider (n = 2), lost to follow-up after day +180 (n = 1), or has not reached day +365 (n = 1).

Subanalysis of pediatric patients

Among the pediatric patients (n = 16), we observed higher 5-year Kaplan-Meier–based probability of OS of 100% with no pediatric deaths. Five-year DFS was 92.3% (95% CI, 0.79-1.0; supplemental Table 1). Outcomes for acute and chronic GVHD were also comparable with those of the larger cohort, with no cases of grade 3 to 4 acute GVHD, and a 2-year cumulative incidence of severe chronic GVHD of 6.2% (95% CI, 0-0.19).

Mixed chimerism

For 4 patients with mixed donor chimerism, the median donor whole-blood chimerism was 23.2% (range, 18.3-87) at a median follow-up of 4 years (IQR, 2.2-6.3). Two of these patients remained on prolonged immunosuppression due to low mixed chimerism at day +365 evaluation, the other 2 patients discontinued immunosuppression at +342 days due to stability of chimerism with resolved GVHD, and +639 days due to rising chimerism, respectively. Both of these patients had >95% CD3⁺ chimerism.

There have been no reports of SCD symptoms, late donor lymphocyte infusions, or second transplantations to indicate graft failure. Two of these patients had continued chronic pain, and this outcome did not correlate with the level of mixed chimerism.

Posttransplantation infection and early toxicities

Fifteen patients had CMV viremia without disease; 2 were in the setting of CMV incompatibility (patients 29 and 34). Fifteen patients had bacteremia with positive blood cultures.

There were 2 cases of hemorrhagic cystitis, and no observed cases of sinusoidal obstructive syndrome.

Supplemental Table 2 describes post-BMT infections and BMT complications.

Sickle cell outcomes

All patients with sustained engraftment, including the 4 with mixed donor chimerism, experienced resolution of VOC after BMT. The 2 patients (patients 5 and 14) with graft failure had recurrent VOC within 1 year of BMT.

Twenty-one patients had chronic pain before BMT, requiring chronic opioid therapy. Twelve patients who achieved engraftment experienced persistent chronic pain after BMT, 5 of whom were able to wean fully off opioids by 1 year. The remaining 7 patients had persistent opioid use: 6 patients treated with chronic full- or partial-agonist opioid therapy, and 1 patient treated for chronic pain with opioid use disorder.

Post-BMT–related SCD complications were rare. There were 2 patients with recurrent CVAs with clinical weakness that improved with supportive care: for patient 26 this occurred at +12 months, and for patient 28 at +18 months in the setting of Moyamoya disease. These patients had >95% donor engraftment. Three patients with a history of cholelithiasis before BMT were evaluated for cholelithiasis after BMT. One required cholecystectomy. Six patients had persistent avascular necrosis after BMT, including 2 patients who required opioids for chronic pain, 2 patients who underwent reconstructive joint surgery, and 2 patients who did not require intervention. The remainder of complications were single cases as outlined in supplemental Table 3, and there were no patients with ACS, sickle cell nephropathy, pulmonary hypertension, or changes in cardiac function after BMT.

Pre-BMT iron overload was recognized in almost all patients (median, 959 ng/mL [IQR, 492-1977]), and 20 patients (47%) had ferritin levels of >1000 ng/mL. For patients evaluated at day +180, median ferritin level improved to 649 ng/mL (IQR, 195-1550). Of 17 patients with persistent iron overload, 14 required therapeutic phlebotomy (82%) after BMT. Patient 2 also received post-BMT chelation due to transfusion dependence.

Reproductive outcomes

Of 27 female patients, 22 had reproductive monitoring including evaluation of gonadal hormone levels after BMT (n = 16/27 [59%]), documentation of menses (n = 21/27 [78%]), or both (n = 15/27 [56%]; supplemental Table 4.1). Thirteen postpubescent female patients had return of menses, which included women with normal hormone levels (n = 6; median age, 19.5 years), women without documented hormone levels (n = 6; median age, 29.5 years), and 1 woman with premature ovarian failure, despite return of menses (n = 1; age, 16 years). There were 8 postpubescent female patients who did not have return of menses and had evidence of premature ovarian failure (median age, 22.5 years). There were 2 successful pregnancies within 5 years of each respective patient’s transplantation; 1 of these women had prolonged amenorrhea after BMT.

Of 16 male patients, 9 had evaluation of gonadal hormone levels after BMT. For postpubertal males after BMT (n = 7), 5 had documented normal hormone levels and 2 had abnormal hormone levels (supplemental Table 4.2). No pregnancies were reported for males.

All patients were offered fertility preservation at BMT consent. Only 20 patients (47%) completed this procedure before BMT, 15 of whom were female. The percentage of patients using fertility preservation increased rapidly after 2020 (n = 17, 85% after 2020 vs n = 3, 15% before 2020).

A summary of the long-term health outcomes is provided in supplemental Table 3.

Discussion

We report an expanded cohort of 43 patients who underwent RIC BMT with TBI of 400 cGy. Full donor chimerism was achieved in 37 patients and stable mixed donor chimerism off immunosuppression in 4 patients. These results demonstrate durable engraftment with 400 cGy TBI haploBMT for SCD, with comparable DFS to matched related myeloablative BMT in SCD and no differences in engraftment between children and adults.²³

The 5% graft failure rate matched our previous findings¹² and occurred in patients with low CD34 (<10⁶ CD34 cells per kg) but normal total nucleated cell counts. This was unexpected as BM graft quality is determined by total nucleated cell dose, as also found in aplastic anemia.²⁴ The 7 fully matched patients add to our overall success, given known outcomes for HLA-matched BMT in SCD²⁵; however, prioritization of younger donors, ABO incompatibility, and CMV status is essential for BMT optimization.^26,27 There were 3 patients with low (<50%) donor peripheral blood chimerism, ranging from 18% to 25%. The immune-mediated mechanism for low chimerism is unclear; however, these patients had higher peripheral blood CD3⁺ chimerism levels and no reoccurrence of SCD symptoms, even for 2 patients who were able to come off prolonged immunosuppression. There is likely a range for the minimum donor myeloid chimerism required to reverse the sickle cell phenotype, and the risk of graft failure must be balanced against risk of prolonged immunosuppression.28, 29, 30, 31, 32 Our results challenge previous reports that associate mixed donor chimerism with eventual graft failure and secondary malignancy.³³ Although longer follow-up is needed for all patients, including those with rejection, thus far we have observed no secondary neoplasms or recurrent SCD despite exposure to increased TBI and mixed chimerism, respectively.34, 35, 36

Rates of severe GVHD were very low, there were no BMT-related deaths, and 5-year OS probability was 95.5%. These curative therapy outcomes are encouraging for patients with severe SCD and high mortality risk.^32,37 Hematopoietic cell transplantation–specific comorbidity index scores were high for adults (mean, 3.3), which is consistent with the anticipated evolution of multiorgan dysfunction and chronic morbidity in aging patients with SCD. Earlier transplantation when indicated should be considered because it favorably affects mortality, recovery of organ function, and health care use.^6,25,38 Autologous transplantation with gene therapy was recently approved by the US Food and Drug Administration and is a potentially promising curative therapy option for patients with SCD^39,40; however, most adults in our cohort would not be eligible for this treatment option given severe disease that likely precludes myeloablative conditioning needed for gene therapy.

The addition of thiotepa or increasing the TBI to 400 cGy has increased engraftment to >90% of patients,^12,41, 42, 43 as compared with 60% in our original haploBMT trial for SCD with 200 cGy.¹¹ Engraftment, OS, and GVHD with this regimen are comparable with those in the Vanderbilt Global Haploidentical Learning Collaborative phase 2 multicenter study and the phase 2 multicenter BMT Clinical Trials Network 1507 haploidentical trial using thiotepa. Compared with these trials in which patients were all maintained on immunosuppression until day +360, 10 patients in our study had successful discontinuation of immunosuppression at 6 months after haploBMT with PTCy. This finding speaks to emerging evidence of early immune reconstitution with haploBMT for SCD⁴⁴; we are studying this new standard moving forward for adult patients without GVHD. Of 16 children, 1 experienced graft failure, as opposed to that reported with the addition of thiotepa to 200 cGy TBI, in which all 8 graft failures were in 32 children aged <18 years.⁴² DFS for children in our study was 92.3% at 5 years, which is reassuring, although there were fewer younger patients in our study and caution should be taken with this age group given rejection experience across haploBMT.^41,42

Another major advantage of the TBI-only approach used in this study is fertility preservation. A survey of adults with SCD found that most patients are willing to consider alloBMT even with graft failure rates of >10%, but chronic GVHD and infertility were unacceptable.⁴⁵ Gonadal failure is common after myeloablative conditioning using busulfan,^46,47 and a potential advantage of RIC conditioning is maintenance of fertility.47, 48, 49 The dose of TBI that spares gonadal function is unknown. One study suggests doses up to 600 cGy preserve gonadal function in most patients.⁵⁰ Although not a primary end point, 44% of female patients resumed menses after BMT, and 2 had pregnancies, suggesting possible fertility preservation with 400 cGy TBI. For the 6 women with return of menses and normalization of post-BMT gonadal hormones, the median age was 2 years lower than those with premature ovarian failure without return of menses. This may be explained by low ovarian reserve at a younger age for women and girls with SCD.⁵¹ Hormone monitoring was less common for males, although 5 postpubertal males had return to normal hormone function after BMT.

Post-BMT complications across these studies that were also common to our regimen included infections, specifically viremia and bacterial infections, but resolved with supportive care, and no infection-related deaths or graft failure occurred. As demonstrated in BMT Clinical Trials Network 1703, post-BMT infections are not more severe with PTCy than with other approaches, and prophylaxis and monitoring are necessary to prevent morbidity.⁵² Our study uniquely had a single patient with anti-erythroid antibodies conferring transfusion dependence after BMT; however, this was in the setting of ABO mismatch. All other patients were red cell transfusion independent by +90 days (median, 3 red cell units required after BMT). Our practice is to avoid ABO major mismatched donors for SCD BMT, although it should be noted for patients without alternative donor options that 2 of 3 of such patients in our study became transfusion independent by +90 days.

SCD-related symptoms showed improved acute pain and rare but important post-BMT complications (thrombosis, focal segmental glomerular sclerosis, and recrudescence of CVA), highlighting the need for early intervention before irreversible damage.³⁰ Of 21 patients with pre-BMT opioid use, 67% discontinued opioids by 12 months, suggesting pain improvement after BMT despite persistent chronic pain in some. We attribute these successes to improved provider knowledge and strategic post-BMT opioid tapering and recognize that post-BMT chronic pain in SCD is multifactorial.⁵³ Additionally, specific complications, such as recurrent CVA, are highest immediately after BMT in those with preexisting cerebrovascular disease, as was the case in both patients with this complication.⁵⁴ Consideration of earlier BMT may reduce these types of SCD-related complications after BMT.

Our BMT regimen has multiple advantages compared with current nonmyeloablative and myeloablative BMT regimens and gene therapy and editing trials.⁵⁵ First, we had significant reduction in health care use as demonstrated by a median length of stay of 12 days for adults. Length of stay was longer for pediatric patients, at 45 days, and reflects our institutional pediatric practice of admitting through count recovery. BMT is at least one-fifth the cost of gene therapy and editing trials and remains a more accessible option for patients with severe SCD who cannot tolerate myeloablation.^56,57 Although we did not directly measure quality of life, improved physical health and resolved acute pain positively affect quality of life after BMT for SCD.^58,59 Future studies should incorporate patient-reported outcomes with the expansion of curative therapies.

Limitations of this study include smaller numbers, notably for the pediatric cohort, and single-institution results. Three patients transitioned care before day +365, and further longitudinal follow-up is needed for all patients, limiting our understanding of all potential long-term complications related to BMT and/or SCD. Reproductive outcomes, particularly in males, require further study with monitoring before and after curative SCD therapies.⁶⁰

Conclusion

Our study shows that a RIC regimen with TBI of 400 cGy, and either haploidentical or HLA-matched donors, enables durable engraftment with low toxicity, expanding access to curative therapy for individuals with severe SCD. Promising outcomes in survival and reproductive health merit further long-term study.

Conflict-of-interest disclosure: The authors declare no competing financial interests.