New age HCT conditioning regimens: what works and why?

Visual Abstract

Abstract

Conditioning regimens play an essential role in allogeneic transplantation by facilitating engraftment and eradicating malignancies. The landscape of conditioning regimens has undergone an evolution in the concept of intensity. Novel conditioning strategies aim to provide highly efficacious regimens with improved toxicity profiles. Integrating disease- specific chemotherapy and targeted agents into conditioning regimens provides enhanced disease elimination and relapse prevention. Agents like treosulfan provide safer conditioning with a favorable toxicity profile for patients with older age or medical comorbidities and can lower the incidence of long-term complications for younger patients. Additionally, methodologies for precise and targeted radiation delivery with minimal off-target effects are emerging. A promising development is radioimmunotherapy-based regimens that preferentially deplete hematopoietic cells and spare nonhematopoietic tissues. These advancements necessitate reexamination and harmonization of conditioning intensity stratification schemes for a more personalized and selective approach.

Learning Objectives

• Reexamine current conditioning regimen intensity categorization schemes

• Discuss novel conditioning regimens for active or residual disease

• Explore further options to optimize efficacy without increasing toxicity in older and medically infirm younger individuals

Introduction

Conditioning regimens are an integral component of allogeneic hematopoietic cell transplantation (HCT). The primary purpose of conditioning is suppression of the recipient's immune system for the prevention of graft rejection. When HCT is performed in the management of hematologic malignancies, the goal is also to debulk tumor burden or eradicate residual disease. Conditioning regimens usually combine an alkylator and purine analogue or total body irradiation (TBI). These combinations have evolved significantly, progressing from an initial focus to improve safety and toxicity from supralethal TBI-containing myeloablative conditioning (MAC) regimens, broadening to radiation-free regimens that, in conjunction with the recognition of the donor-mediated immunologic graft-vs-tumor (GvT) effect, have led to the development of reduced intensity conditioning (RIC) and nonmyeloablative (NMA) regimens. These have resulted in a remarkable expansion of transplant indications to older or medically infirm patients. With improved supportive care and the advent of less toxic regimens, the current focus is now shifting to develop more efficacious regimens, improving the toxicity profile, along with augmenting disease eradication and relapse reduction by incorporating optimized chemotherapy drugs, monoclonal antibodies, disease-specific therapeutic agents, and innovative radiation delivery techniques.

Overview of conditioning regimen intensity

Traditional classification

Traditionally, conditioning regimens have been classified as MAC, RIC, and NMA, depending on doses of alkylating agents and TBI. These terminologies are somewhat arbitrarily defined, reflecting reversibility of myelosuppression and requirement for stem cell rescue. MAC regimens cause irreversible hematologic toxicity and require stem cells to restore hematopoiesis, while NMA regimens cause reversible cytopenia and do not require stem cell rescue. Regimens that cannot be categorized as MAC or NMA are classified as RIC. In general, the distinction between MAC and RIC is ≥ 30% dose reduction of alkylating agents and TBI.^1,2 The Center for International Blood and Marrow Transplant Research devised an operational definition for RIC regimens. Accordingly, RIC criteria included reversible myelotoxicity without stem cell support, mixed early posttransplant donor chimerism, and low risk of nonhematologic toxicity.³

CLINICAL CASE 1

In this case that integrated MAC and post-transplant cyclophosphamide (PT-Cy) treatments, a 37-year-old man with acute lymphoblastic leukemia with persistent measurable residual disease (MRD), following HyperCVAD (hyperfractionated- cyclophosphamide, vincristine, adriamycin [doxorubicin], and dexamethasone) induction, achieved MRD-negative remission after blinatumomab. He underwent a human leukocyte antigen (HLA)–matched unrelated donor allogeneic transplant following MAC regimen 12 Gy TBI, with PT-Cy and cyclosporine as graft-vs-host disease (GVHD) prophylaxis. He developed mild acute skin GVHD and no chronic GVHD. He is 2 years from HCT and remains in complete remission.

Reexamining conditioning regimen intensity

The classical trichotomized conditioning intensity classification scheme, despite its widespread use in routine practice, clinical trials, and registry studies, poses several limitations. The regimen intensities represent a spectrum rather than discrete categories and vary substantially depending on components and doses of chemotherapeutic agents and radiation. The intensity is defined based on hematologic effect without accounting for nonhematologic toxicities. Additionally, the intensity of purine analogues, etoposide, carmustine, and other agents is not weighed when defining regimen intensities. Consequently, there is a lack of consensus and universal acceptability of these criteria. Terminologies like “reduced-toxicity,” “augmented,” or “intensified” conditioning regimens are frequently encountered in the literature, underscoring the need for redefining conditioning intensity.⁴

In an attempt to update and more precisely define regimen intensity profile contemporaneously, the European Society for Blood and Marrow Transplantation developed a new scoring tool, the transplant conditioning intensity (TCI) score. In this retrospective registry-based study, nonrelapse mortality (NRM) was compared between the TCI score and the binary MAC/RIC classification in patients with acute myeloid leukemia (AML). Scoring was done using established threshold ablative doses or assigning a weighted score based on a known toxicity profile of drugs (Table 1). Regimens were categorized according to composite TCI score as low (1-2), intermediate (2.5-3.5), and high (4-6) intensity. The proposed TCI group stratification was a strong predictive model for early and overall NRM and identified a distinct intermediate group that comprised roughly equal proportions of originally assigned MAC and RIC regimens (Figure 1).⁵ Validation of TCI score using an external, more contemporary AML cohort confirmed the findings.⁶

Table 1.

Intensity weighted scores for common conditioning components

Component	Dose level			Added points for each dose
	Low	Intermediate	High
TBI fractionated (Gy)	≤5	8-9	≥9	1
Busulfan (mg/kg)	≤6.4 IV & ≤8 PO	9.6 IV & 12 PO	12.8 IV & 16 PO	1
Treosulfan (g/m2)	30	36	42	1
Melphalan (mg/m2)	<140	≥140	>200	1
Thiotepa (mg/kg)	<10	≥10	≥20	0.5
Fludarabine (mg/m2)	≤160	>160		0.5
Clofarabine (mg/m2)	≤150	>150		0.5
Cyclophosphamide (mg/kg)	<90	≥90		0.5
Carmustine (mg/m2)	≤250	280-310	≥350	0.5
Cytarabine (g/m2)	<6	≥6		0.5
Etoposide (mg/kg)	<50	≥50		0.5

Reproduced with permission from Springer Nature from https://doi.org/10.1038/s41409-020-0803-y.

IV, intravenous; PO, per os; TBI, total body irradiation.

Figure 1.

Early non-relapse mortality (NRM) according to TCI. (A) Early NRM in the subgroup of patients aged 45-55 years at transplant (n  =  3858). (B) Early NRM in patients aged between 55 and 65 years (n  =  4397). Reproduced with permission from Springer Nature from https://doi.org/10.1038/s41409-020-0803-y.

Reexamination of conditioning regimen intensity in the modern era necessitates compendious assessment in conjunction with GVHD prophylaxis, ex vivo graft manipulation, and posttransplant maintenance therapy as a “transplant package.” Although the TCI score provides a new stratification framework, extrapolating to different disease types, pediatric patients, and alternative donors requires further validation. Additionally, it does not specifically address the impact of PT-Cy on regimen intensity when added for GVHD prophylaxis and is not part of the TCI score, even though the current PT-Cy dose (100  mg/kg) is comparable to that used when cyclophosphamide is a component of MAC regimens (100-120  mg/kg). PT-Cy has been widely adopted as a standard GHVD prophylaxis in the RIC transplant setting based on the Blood and Marrow Transplant Clinical Trials Network (BMT-CTN) 1703 trial,^7,8 and while PT-Cy is increasingly utilized in the MAC transplant setting, the optimal GVHD prophylaxis/MAC platform remains to be elucidated. This is particularly imperative because PT-CY may contribute to cumulative toxicity such as prolonged cytopenias, hemorrhagic cystitis, cardiotoxicity, and increased infection risk in the MAC or even RIC transplant settings. Therefore, this warrants further exploration and provides an impetus for prospective clinical trials to devise ideal MAC regimens, PT-Cy dose, and partner GVHD agents, providing an optimized PT-Cy–MAC platform. While efforts for PT-Cy dose optimization are ongoing to mitigate PT-Cy–associated toxicity, several studies explored the optimal strategy to integrate PT-Cy with MAC. The combination of cyclophosphamide and high-dose TBI (Cy/TBI) is a standard MAC regimen for lymphoid and myeloid malignancies. Sequencing cyclophosphamide following allografting as PT-Cy–based immunosuppression is an alternative and feasible platform. Two prospective trials integrating PT-Cy as GVHD prophylaxis with high-dose TBI (12-13.2 Gy) or fludarabine/busulfan (Flu/Bu₄) MAC regimens demonstrated feasibility without excessive toxicity or increased NRM.^9,10 Prospective trials integrating PT-Cy with MAC are summarized in Table 2. Finally, ideal conditioning regimens in the PT-Cy–based GVHD prophylaxis in the haploidentical transplant are not well defined. There is evidence that certain regimens (eg, Flu/melphalan [Mel]) may have a higher incidence of cytokine release syndrome and NRM.¹¹

Table 2.

Prospective trials of posttransplant cyclophosphamide in MAC

Trial	Study population	MAC regimen	Donor/graft source	GVHD	NRM	Relapse incidence	Survival outcomes
Raiola et al.74	N  =  50 • Adults • Various hematologic malignancies • 27 with active disease	Flu/Bu/TT Flu/9.9 Gy TBI	Haploidentical/BM	Acute GVHD: Grade II-IV  =  12% Chronic GVHD: 26%	18% at 6 months	26%	RFS: 51% at 18 months OS: 62% at 18 months
Kanakry et al.75	N  =  92 • Adults • Various hematologic malignancies • 27 with active disease • 20 with MRD positivity	Flu/Bu4	MSD and MUD/BM	Acute GVHD: Grade II-IV  =  51% Chronic GVHD: 14%	16% at 1 year	22% at 2 years	RFS: 62% at 2 years OS: 67% at 2 years
Mielcarek et al.9	N  =  43 • Adults and pediatrics • Various hematologic malignancies • 21 with active disease or MRD	Flu/Bu4 TBI (12-13.2 Gy)	MSD, MUD, MMUD/PBSC	Acute GVHD: Grade II  =  77% Grade III-IV  =  0% Chronic GVHD: 16%	14% at 2 years	17% at 2 years	RFS: 69% at 2 years OS: 70% at 2 years
Symons et al.76	N  =  96 • Adults and pediatrics • Various hematologic malignancies	Bu/Cy Cy/12 Gy TBI	Haploidentical/BM	Acute GVHD: Grade II-IV  =  11% Chronic GVHD: 15%	11% at 1 year	35% at 1 year	EFS: 49% at 3 years OS: 54% at 3-year
Luznik et al. (PT-Cy arm)77	N  =  114 • Adults and pediatrics • Acute leukemia, MDS, CMML	Bu/Cy Flu/Bu4 Cy/12 Gy TBI	MSD, MUD/BM (90%), PBSC (10%)	Acute GVHD: Grade II-IV  =  38% Chronic GVHD: 27%	16% at 2 years	14% at 2 years	CRFS: 48% at 2 years OS: 76% at 2 years
Popat et al.78	N  =  55 • Adults and pediatrics • Various hematologic malignancies	Fractionated Bu/Flu/TT	MSD, MUD, haploidentical/ PBSC (60%), BM (40%)	Acute GVHD: Grade II-IV  =  38% Chronic GVHD: 11%	22% at 2 years	24% at 2 years	RFS: 54.5% at 2 years OS: 65.5% at 2 years
Fierro-Pineda et al.79	N  =  32 • Pediatric and young adults • Acute leukemia and MDS	Bu/Cy Cy/12 Gy TBI	Haploidentical/BM	Acute GVHD: Grade II  =  13% Grade III-IV  =  0% Chronic GVHD: 4%	0% at 6 months	32% at 1 year	EFS: 64% at 2 years OS: 73% at 2 years
Jurdi et al.10	N  =  125 • Adults and pediatrics • Various hematologic malignancies	Flu/Bu4 13.2 Gy TBI	MSD, MUD, MMUD/PBSC (66%), BM (34%)	Acute GVHD: Grade II-IV  =  17% Chronic GVHD: 5.5%	10% at 2 years	39% at 2 years	GRFS: 52% at 2 years OS: 74% at 2 years

BM, bone marrow; Bu, busulfan; CMML, chronic myelomonocytic leukemia; CRFS, chronic GVHD-free, relapse-free survival; Cy, cyclophosphamide; EFS, event free survival; Flu, fludarabine; GRFS, GVHD-free, relapse-free survival; GVHD, graft vs host disease; MAC, myeloablative conditioning; MDS, myelodysplastic syndrome; MMUD; mismatched unrelated donor; MSD, matched sibling donor; MRD, measurable residual disease; MUD, matched unrelated donor; NRM, non-relapse mortality; RFS, relapse free survival; OS, overall survival; PBSC, peripheral blood stem cell; TBI, total body irradiation; TT, thiotepa.

CLINICAL CASE 2

This case, where we consider conditioning strategies for refractory AML, a 44-year-old woman with AML received 2 cycles of intensive induction chemotherapy with primary refractory disease with 60% bone marrow as well as 25% circulating blasts. She underwent an HLA-compatible unrelated donor allogeneic HCT on a clinical trial using astatine-211 (²¹¹At)–radiolabeled anti-CD45 monoclonal antibody (mAb) combined with Flu/2 Gy TBI as conditioning. She had rapid clearance of circulating blasts and was in complete remission at day 30. She remains in complete remission without significant chronic GVHD 3 years after HCT.

Conditioning regimens for active or residual disease

MAC vs RIC

The crucial role of conditioning in eradicating and mitigating relapse of the underlying malignancy is unquestioned. The fundamental question of MAC vs RIC was evaluated in 3 prospective randomized trials (Table 3). The first trial, by the German AML group, randomized patients with AML to standard Cy/12 Gy TBI (MAC arm) or Flu/8 Gy TBI (RIC arm). Early toxicity and mortality were higher with MAC; however, 3-year outcomes, including NRM, relapse incidence, relapse free survival (RFS), and overall survival (OS), were similar.¹² Ten-year long-term follow-up data continued to show no difference.¹³ One could argue that Flu/8 Gy TBI meets the Center for International Blood and Marrow Transplant Research's MAC definition, and thus there was equipoise with relapse and NRM. In the RICMAC trial, patients with myelodysplastic syndrome (MDS) and secondary AML were randomized to Cy/Bu₄ or Flu/Bu₂ showing equivalent relapse incidence, RFS, and OS, while NRM tended to be higher with MAC.¹⁴ Lastly, the BMT-CTN phase 3 trial randomized patients with AML and MDS in remission to MAC or RIC. Contrary to the European trials, a significantly higher relapse rate was observed. Although NRM was lower with RIC, a disproportionately higher relapse incidence translated into inferior survival, which was not significant in the primary analysis¹⁵ but showed statistical significance in the 4-year follow-up analysis.¹⁶

Table 3.

Randomized trials comparing MAC vs RIC

Trial	Study population	Disease status	Donor/graft source	Conditioning regimen	NRM	Relapse incidence	Survival outcomes
Bornhäuser et al.12 Fasslrinner et al.13	N  =  195 • Adults (18-60 years) • Intermediate/ high-risk AML	Complete remission	MSD, MUD, MMUD/PBSC (92%), BM (8%)	MAC arm: Cy/12 Gy TBI, n  =  96 RIC arm: Flu/8 Gy TBI, n  =  99	17% vs 13% at 3 years (NS)	26% vs 28% at 3 years (NS)	RFS: 56% vs 58% at 3 years (NS) OS: 58% vs 61% at 3 years (NS)
					Long-term follow-up (10 years)
					23% vs 16% (NS)	30% vs 30% (NS)	RFS: 43% vs 55% (NS) OS: 47% vs 60% (NS)
Scott et al.15,16	N  =  272 • Adults (18-65 years) • AML, MDS • HCT-CI ≤4	Complete remission	MSD, MUD/PBSC (92%), BM (8%)	MAC arm: Flu/Bu4, n  =  87 Bu/Cy, n  =  40 Cy/12 Gy TBI , n =  8 RIC arm: Flu/Mel, n  =  21 Flu/Bu2 , n =  89	16% vs 4% at 18 months (P  =  .002)	13.5% vs 48% at 18 months (P  <  .001)	RFS: 68% vs 47% at 18 months (P  <  .01) OS: 77.5% vs 68% at 18 months (NS)
					Long-term follow-up (4 years)
					25% vs 10% (P  <  .001)	20% vs 61% (P  <  .001)	RFS: 58% vs 34% (P  <  .0001) OS: 62% vs 49% (P  =  .022)
Kroger et al.14	N  =  129 • Adults (18-65 years) • MDS, secondary AML with < 20% blasts	48 (37%) with >5% BM blasts	MSD, MUD/ PBSC (93%), BM (7%)	MAC arm: Bu/Cy , n =  64 RIC arm: Flu/Bu2 , n =  65	25% vs 17% at 1 year (NS)	15% vs 17% at 2 years (NS)	RFS: 58% vs 62% at 2 years (NS) OS: 63% vs 76% at 2 years (NS)

AML, acute myeloid leukemia; HCT-CI, hematopoietic cell transplantation-comorbidity index; Mel, melphalan; NS, not significant; RIC, reduced intensity conditioning.

Several caveats limit interpretation of these trials relevant to current practice. All of these trials suffered from premature termination due to slow accrual (the German and RICMAC trials) or strikingly high relapse rates in the RIC arm (BMT-CTN trial). Second, there were inherent differences in biological disease features, primary endpoints, and conditioning regimens. Pretransplant MRD is an established and consistent predictor of poor outcomes in AML. MRD assessment was not integrated in these trials. However, post hoc DNA sequencing analysis of preconditioning blood samples of patients with AML from the BMT-CTN trial showed the impact of regimen intensity on pretransplant MRD. RIC regimen with genomic evidence of disease resulted in inferior survival consequent to exceedingly higher relapse risk, while MAC counterbalanced the adverse risk associated with MRD, resulting in survival equivalent to the MRD- negative group.¹⁷ Early posttransplant MRD eradication is more likely to occur following MAC; however, survival continues to remain inferior relative to MRD negative.^18-20 Finally, a limitation of the BMT-CTN trial was the aggregation of heterogeneous MAC and RIC regimens, whereas the most efficacious MAC and RIC regimens for AML and MDS have yet to be definitively established. In the RIC arm, 20% of patients received Flu/Mel that appears to have superior outcomes compared to Flu/Bu₂, driven primarily by lower relapse rates.¹⁵ Despite these conflicting results, MAC should be preferred for younger (<60 years) and fit individuals. However, the median age of AML diagnosis is late 60s, rendering the majority ineligible for MAC. This underscores the critical need for developing optimal RIC approaches.

Primary refractory or relapsed AML remains a challenging obstacle. Allogeneic transplant is the only potentially curative therapy, but prognosis remains dismal, and an optimal conditioning strategy has not been established. Consequent to improvements in the transplant field, survival of these patients has improved, albeit marginally, with a significant proportion still succumbing to early disease recurrence. There is no satisfactory evidence to support the superiority of intense over less intense regimens in such a scenario, and clinical trials with novel conditioning approaches are urgently required to optimize transplant strategies.

Sequential conditioning regimens

Sequential conditioning regimens provide an alternative transplant strategy for patients with refractory/relapsed leukemia and advanced MDS, where conventional transplant yields a significantly inferior outcome. These regimens involve a short course of intense antileukemia chemotherapy for cytoreduction followed immediately by RIC transplant to introduce GvT at the lowest disease burden state, followed ultimately by prophylactic donor lymphocyte infusion (DLI) for immune reconstitution to facilitate GvT.^21,22 The prototype sequential regimen, introduced in the early 2000s, consisted of fludarabine, amsacrine, and cytarabine (FLAMSA) as cytoreductive chemotherapy, followed by TBI 4 Gy/Cy/antithymocyte globulin (ATG) conditioning prior to allografting. Patients also received prophylactic DLI. The FLAMSA-RIC platform allowed prompt engraftment and sustained donor chimerism in relapsed/refractory AML and progressive MDS.^23,24 Several modifications of the original sequential FLAMSA-RIC protocol have been developed (Table 4).²⁵ While these variations have shown feasibility, there has been minimal improvement in survival (25%-40%).^26-28

Table 4.

Prospective trials summarizing conditioning regimen strategies for active or residual disease

Trial	Study population	Disease status	Conditioning regimen	CR at day 30	NRM	Relapse incidence	Survival outcomes
Sequential conditioning regimens
Schmid et al.23	N  =  75 • AML • MDS	Active AML, n  =  65% Progressive MDS, n  =  13%	FLAMSA→ 4 Gy TBI/Cy/ATG → Prophylactic DLI (n  =  12)	88%	33% at 1 year	19% at median follow-up of 149 days	RFS: 40% at 2 years OS: 42% at 2 years
Schmid et al.24	N  =  103 • AML	Primary induction failure, n  =  37 Early relapse, n  =  53 Refractory, n  =  8 Second relapse, n  =  5	FLAMSA→ 4 Gy TBI/Cy/ATG → Prophylactic DLI (n  =  17)	91%	17% at 1 year	37% at median follow-up of 4 months	RFS: 37% at 2 years OS: 40% at 2 years
Mohty et al.25	N  =  24 • Primary refractory	Active disease	Clofarabine/cladribine → Cy/Bu2/ATG → Prophylactic DLI (n  =  17)	75%	12% at 2 years	54% at 2 years	RFS: 29% at 2 years OS: 38% at 2 years
Craddock et al.29	N  =  244 • High risk AML, MDS	Complete remission 39% MRD positive	FLAMSA-Bu  =  122 Control group (Flu/Bu2/ATG, Flu/Bu2/Al, Flu/Mel/Al (n  =  122)		20.5% vs 17% (NS) at 1 year	27% vs 29.5% (NS) at 2 years	EFS: 54% vs 48% (NS) at 2 years OS: 61% vs 59% (NS) at 2 years
Milano et al.32	N  =  46 • AML, MDS	Active disease  =  37% MRD positive  =  63%	CLAG-M → 3-4 Gy TBI (n  =  36) FLAG-Ida → 4 Gy TBI (n  =  10)	90%	10%	7 relapses at median follow-up of 11 months	RFS: 57% at 1 year OS: 77% at 1 year
Radioimmunotherapy-based conditioning regimens
Vo et al.36	N  =  15 • AML, MDS	Active disease, n  =  9 MRD positive, n  =  6	90Y radiolabeled anti-CD45 mAb/Flu/2 Gy TBI	87%	7% at 100 days	41% at 1 year	OS: 46% at 2 years
Sandmaier et al.37	N  =  20 • AML, MDS, MPAL	Active disease, n  =  10 MRD positive, n  =  6	211At radiolabeled anti-CD45 mAb/Flu/2 Gy TBI	84%	25% at 1 year	40% at 1 year	RFS: 35% at 1 year OS: 43% at 1 year
Gyurkocza et al.35	N  =  153 • Age ≥55 years • Active relapsed or refractory AML	Primary induction failure  =   54% First early relapse  =  25% Relapsed or refractory  =  16% Second or subsequent relapse  =  5%	131I-apamistamab/Flu/2 Gy TBI  (n =  76) Conventional care  (n =  77) (44 crossed over to 131I-apamistamab arm)	60.5% vs 6.5% Durable CR: 17% vs 0% (<0.0001)	26% vs 28% at 1 year (NS)	49% vs 83% (NS) at data cutoff	OS: 16% vs 16% (NS) at 2 years
TMI- and TMLI-based conditioning regimens
Wong et al.39	N   =   32 • AML • ALL	Primary refractory disease, n   =   19 Relapsed/refractory, n   =   13	TMI/VP-16/Cy (Trial 1) TMI/Bu/VP-16 (Trial 2) TMI dose (12-15 Gy)	93%	8% at 100 days (Trial 1) 20% at 100 days (Trial 2)	36% (Trial 1) 35% (Trial 2)	50% (Trial 1) 25% (Trial 2)
Stein et al.41	N  =  51 • AML	Primary induction failure, n  =  34 First relapse, n  =  14 Second relapse, n  =  3	TMLI/VP-16/Cy TMLI dose (12-20 Gy)	88%	8% at 1 year	65% at 2 years	OS: 41.5% at 2 years
Hui et al.42	N  =  12	Active, n  =  4 MRD, n  =  7	Flu/Cy/TMI TMI dose (15 Gy and 18 Gy)		36% at 1 year	42% at 1 year	OS: 42% at 1 year
Ali et al.43	N  =  23 • AML, n  =  15 • MDS, n  =  5	Active disease  =  65% MRD positive  =  8%	TMI/Flu/Bu2 TMI dose (3-6 Gy)		4% at 1 year	43% at 1 year	RFS: 55% at 1 year OS: 65% at 1 year
Al Malki et al.44	N  =  31 • AML, ALL, MDS	Active disease  =  7%	TMLI/Flu/Cy TMLI dose (12-20 Gy)		13% at 2 years	17% at 1 year	RFS: 74% at 1 year OS: 83% at 1 years
Disease-specific drugs as part of conditioning regimens
Xuan et al.50	N  =  202 • MDS, secondary AML	MRD positive , n =  33/57 in CR	G-CSF/decitabine/Bu4/Cy  (n =  101) Bu4/Cy ( n =  101)		19% vs 22% at 2 years (NS)	11% vs 25% at 2 years (P  =  .011)	RFS: 70% vs 53% at 2 years (P  =  .016) OS: 72% vs 57% at 2 years (P  =  .039)
Garcia et al.51	N  =  22 • AML, MDS, myeloproliferative neoplasms	Excess blasts , n =  5 Persistent cytogenetic abnormalities , n =  12 TP53 mutation, n  =  12	Venetoclax/Flu/Bu2		9% at 1 year	37% at 1 year	RFS: 53% at 1 year OS: 67% at 1 years
Popat et al.52	N  =  49 • AML, MDS	Active disease , n =  10 MRD positive , n =  8 TP53 mutation , n =  11	Bu4/Flu/cladribine/TT/venetoclax		22% at 2 years	16% at 2 years	RFS: 61% at 2 years OS: 65% at 2 years

Al, alemtuzumab; ALL, acute lymphoblastic leukemia; ATG, antithymocyte globulin; CLAG-M, cladribine, cytarabine, G-CSF, mitoxantrone; CR, complete remission; DLI, donor lymphocyte infusion; FLAG-Ida, fludarabine, cytarabine, G-CSF, idarubicin; MPAL, mixed phenotype acute leukemia; FLAMSA, fludarabine, amsacrine, cytarabine; G-CSF, granulocyte-colony stimulating factor; mAb, monoclonal antibody; TMLI, total marrow and lymphoid irradiation; TMI, targeted marrow irradiation; VP-16, etoposide.

The FIGARO trial randomized patients with AML and MDS to sequential regimen FLAMSA/Bu or standard regimen Flu/Bu₂ RIC, showing comparable outcomes in both arms. Pretransplant MRD was detectable in 39% of patients while only 5% had active leukemia, rendering limited applicability to active disease status.²⁹ When comparing sequential conditioning with standard MAC in active AML, outcomes are similar, suggesting no benefit of higher intensity.³⁰

This approach could potentially be further optimized by incorporating agents with enhanced antileukemia activity.³¹ Preliminary results of an ongoing prospective trial (NCT04375631) demonstrated the feasibility of sequential CLAG-M (cladribine, cytarabine, granulocyte-colony stimulating factor [G-CSF], and mitoxantrone) and FLAG-Ida (fludarabine, cytarabine, G-CSF, and idarubicin) as cytoreductive chemotherapy, followed by 3 to 4 Gy TBI prior to allogeneic HCT with encouraging RFS and OS rates (Table 4).³²

Radioimmunotherapy-based conditioning regimens

The exquisite radiosensitivity of neoplastic cells has prompted exploration of higher TBI doses; however, reduction in relapse risk is counterbalanced by increased organ toxicity and NRM.³³ Radioimmunotherapy has emerged as an exciting development, enabling higher radiation delivery to hematolymphoid tissue. Fundamentally, it involves conjugating monoclonal antibodies with radionucleotide isotopes, resulting in higher radiation delivery to disease-specific tissues while sparing off-target organs. Theoretically, the biodistribution of radioimmunoconjugates can be enhanced by selecting an antibody target that is highly specific and a radioisotope capable of delivering large and precise radiation payload in target tissues. Several targets have been explored, such as CD20, CD33, and CD66, but the most exploited is CD45 secondary to its relatively high specificity to hematopoietic tissue (Table 4). Anti-CD45 antibodies conjugated with radioisotopes, including iodine-131 (¹³¹I) and yttrium-90 (⁹⁰Y), have demonstrated considerable progress, with ¹³¹I-radiolabeled antibodies achieving 2- to 3-fold higher radiation doses to target organs.

Following demonstration of safety, successful engraftment, and donor chimerism with ¹³¹I-radiolabeled anti-CD45 antibody (¹³¹I-apamistamab) combined with Flu/2 Gy TBI, a randomized phase 3 trial (SIERRA) was conducted to compare the efficacy of ¹³¹I-apamistamab/Flu/2 Gy TBI followed by allografting with the investigator's choice of conventional care in patients with relapsed/refractory AML aged ≥55 years.^34,35 The primary endpoint of the trial was durable complete remission of 180 days, favoring the ¹³¹I-apamistamab arm. Survival was similar between the 2 arms, primarily due to a crossover trial design, with 57% of patients randomized to the conventional arm subsequently receiving ¹³¹I-apamistamab.³⁵ Similarly, the use of yttrium-90 radiolabeled antibody has been explored, demonstrating feasibility and tolerability.³⁶

In contrast to β-particle emitters such as iodine-131 or yttrium-90, α-particle emitters such as astatine-211 (²¹¹At) have a short path length, high energy, and short half-life, properties that enable highly potent, precise, and selective cytotoxicity while minimizing off- target toxicity. In an ongoing phase 1 trial (NCT03128034), ²¹¹At-radiolabeled anti-CD45 antibody/Flu/2 Gy TBI conditioning in advanced leukemia and MDS showed feasibility with successful engraftment and full early donor chimerism in all patients.³⁷ Collectively, anti-CD45 radiolabeled antibodies provide a promising pretransplant conditioning prospect in this high-risk disease population.

Targeted marrow irradiation and total marrow and lymphoid irradiation

Another recent advancement in radiation delivery is targeted radiation to hematopoietic and lymphoid tissues. This strategy utilizes the intensity modulated radiation therapy technique to contour target organs and define avoidance structures, allowing for more precise mapping and dosimetry, conforming higher radiation delivery (up to 20 Gy) to bone marrow, spleen and lymph nodes while minimizing off-target radiation compared to conventional TBI. Targeted marrow irradiation (TMI) and total marrow and lymphoid irradiation (TMLI) have been safely combined with various chemotherapy conditioning mainstays (Table 4).^38-45 In a phase 1 trial, patients with active leukemia received escalating TMLI doses (12-20 Gy) with Cy/etoposide conditioning, reporting successful donor engraftment. The radiation dose to the nontarget organs was reduced by 45% to 85% of the prescribed hematolymphoid dose. NRM was low, and survival was encouraging.⁴¹

Incorporating disease-specific drugs

Leukemia-specific drugs such as venetoclax or hypomethylating agents can enhance the antileukemia potential of conditioning regimens by “priming” leukemia cells, maximizing the cytotoxic activity of chemotherapy. The addition of hypomethylating agents to chemotherapy leads to synergistic killing effects on neoplastic cells, upregulates expression of tumor-associated antigens, and increases regulatory T cells (Treg), thereby leading to enhanced GvT.^46,47 Intensification of conditioning with decitabine in AML and other myeloid neoplasms has been feasible.^48,49 A randomized clinical trial comparing G-CSF/decitabine/Cy/Bu₄ with Cy/Bu₄ in MDS and secondary AML showed significantly improved relapse rates and survival without compromising engraftment, increased toxicity, or NRM.⁵⁰

Venetoclax has also been integrated into Bu-based regimens. The combination of venetoclax with Flu/Bu₂ was feasible in a phase 1 trial without impairment of engraftment or excessive toxicities. Survival was encouraging in patients with high-risk AML, MDS, or myeloproliferative neoplasms, with many harboring TP53 mutation and with active disease at transplant.⁵¹ Exploiting the synergistic killing effect between venetoclax and chemotherapy (busulfan, cladribine, and thiotepa), the so-called “Cladillac” regimen (Bu/Flu/cladribine/thiotepa/ venetoclax) was evaluated in high-risk AML and MDS, showing encouraging outcomes (Table 4).⁵²

CLINICAL CASE 3

A final case is presented here to identify optimal conditioning for older or medically unfit patients: 68-year-old man with high-risk MDS underwent haploidentical transplant from his son. His comorbidities included obesity, hypertension, atrial fibrillation, and diabetes mellitus. HCT/comorbidity index was 3. His conditioning was Flu/Treo/2 Gy TBI. The regimen was well tolerated. He developed mild mucositis and gastrointestinal toxicities. At his 1-year long-term follow-up evaluation, he was noted to have mild GVHD involving his eyes. He was otherwise in complete remission.

Optimizing efficacy without increasing toxicity

RIC vs NMA

Nonmyeloablative transplantation extends the therapeutic option to medically infirm and elderly patients, including septuagenarians. In principle, NMA regimens are minimally toxic but provide sufficient immunoablation to facilitate engraftment, allow varying degrees of mixed-donor chimerism, and rely exclusively on a GvT effect for disease eradication. Although 2 Gy TBI provides sufficient immunosuppression for durable donor chimerism, the addition of fludarabine to 2 Gy TBI is superior in preventing graft rejection. Additionally, fludarabine being a radiosensitizer potentiates the effect of TBI.^53,54 In a randomized trial, NMA Flu/2 Gy TBI had higher relapse rates but lower NRM, resulting in equivalent survival relative to RIC Flu/Bu₂.⁵⁵ Patients up to age 65 were enrolled in this trial, and there are no randomized trials comparing NMA and RIC in patients >65 years. Registry-based analysis comparing Flu/2 Gy TBI with Flu/Bu₂ in patients with AML ≥60 years showed similar outcomes.⁵⁶ For patients >70 years, there has been no proven advantage of RIC over NMA consequent to higher NRM. Conversely, higher intensity does not reduce relapse risk in this population, including pre-HCT MRD positive, suggesting relatively less chemo- or radiosensitive disease.^57,58

Treosulfan-based conditioning regimen

A promising new alkylator addition to the repertoire of conditioning regimens is treosulfan (Treo), a hydrophilic compound with potent myeloablative, immunoablative, and antineoplastic activities. The predictable pharmacokinetic profile and distinct pharmacodynamic properties of treosulfan portend a selective advantage over busulfan with greater antileukemia activity and a favorable extramedullary toxicity profile. Treosulfan in combination with fludarabine has been in use in Europe since the early 2000s. Several studies demonstrated the feasibility of Flu/Treo without increased risk of engraftment failure, nonhematologic toxicity, or NRM.^59-62 A noninferiority randomized phase 3 trial compared Flu/Bu₂ and Flu/Treo in patients with AML and MDS. The trial was stopped early due to superior event-free survival, NRM, and OS favoring Flu/Treo, while relapse rates were similar.⁶³ On the basis of this trial, the US Food and Drug Administration approved Flu/Treo as a conditioning regimen in January 2025. The integration of 2 Gy TBI to Flu/Treo may reduce relapse incidence, particularly in genetically unfavorable malignancies, as reported by prospective trials.^64,65 Treo-based regimens have also been studied in the haploidentical setting, in the pediatric population, and for nonmalignant disorders (Table 5).^66-68

Table 5.

Prospective trials of treosulfan-based conditioning regimens

Trial	Study population	Donor type	Conditioning regimen	NRM	Relapse incidence	Survival outcomes
HLA matched donor transplant
Casper et al.59	N  =  30 • Adults (18-65 years) • Various hematologic malignancies	MSD MUD	Flu/Treo (30 g/m2)	20%	7%	EFS: 49% at a median of 22 months OS: 73% at a median of 22 months
Casper et al.61	N  =  56 • Adults • Various hematologic malignancies • 32 (58%) not in CR	MSD MUD MMUD	Flu/Treo (30-42 g/m2)	20% at 2 years	31% at 2 years	RFS: 49% at 2 years OS: 64% at 2 years
Nemecek et al.62	N  =  60 • Adults and pediatrics • AML, MDS, ALL • 28 with HCT-CI ≥3	MSD MUD	Flu/Treo (36-42 g/m2)	8% at 2 years	33% at 2 years	RFS: 58% at 2 years OS: 65% at 2 years
Ruutu et al.60	N  =  45 • Adults • MDS	MSD MUD	Flu/Treo (30 g/m2)	17% at 2 years	16% at 2 years	RFS: 67% at 2 years OS: 71% at 2 years
Casper et al.80	N  =  75 • Adults • AML	MSD MUD	Flu/Treo (30 g/m2)	11% at 2 years	34% at 2 years	RFS: 55% at 3 years OS: 61% at 2 years
Gyurkocza et al.64	N  =  96 • Age ≤60 years • AML, MDS	MSD MUD	Flu/Treo (42 g/m2)/2 Gy TBI	8% at 2 years	27% at 2 years	RFS: 64% at 2 years OS: 73% at 2 years
Deeg et al.65	N  =  100 • Age ≤70 • KPS ≥70% • AML, MDS, CMML • 64 with HCT-CI ≥3 • 27 with HCT-CI ≥5	MSD MUD MMUD	Flu/Treo (42 g/m2)  =  35 Flu/Treo (42 g/m2)/2 Gy TBI  =  65	9% at 1 years for both arms	22% vs 34% at 1 years (NS)	RFS: 54% vs 68% at 1 year (NS) OS: 69% vs 80% at 1 years (NS)
Beelen et al.63	N  =  476 • Adults • AML, MDS • Age ≥50 • HCT-CI ≥2 • KPS ≥60%	MSD MUD	Flu/Treo (30-42 g/m2)  =  230 Flu/Bu2  =  246	12% vs 28% at 2 years (S)	25% vs 23% at 2 years (NS)	EFS: 64% vs 50% 2 years (S) OS: 71% vs 56% at 2 years (S)
Haploidentical transplant
Peccatori et al.68	N  =  121 • Adults and pediatrics • Various hematologic malignancies • 26 with HCT-CI ≥4	Haploidentical	Flu/Treo (42 g/m2)	30% at 1 year	48% at 3 years	RFS: 20% at 3 years OS: 25% at 3 years
Nonmalignant disorders and pediatric patient population
Burroughs et al.66	N  =  31 • Age <55 years • 9 with HCT-CI ≥3 • Nonmalignant disorders	MSD MUD	Flu/Treo (42 g/m2) ± ATG	0% at 100 days		90% at 2 years
Sykora et al.67	N  =  106 • Pediatrics • Non-malignant disorders	MSD MUD	Flu/Treo (30-42 g/m2) ± TT  =  51 Flu/Bu2 ± TT  =  50	4% vs 12% at 1 year		OS: 96% vs 88% at 1 years

KPS, Karnofsky performance status; S, significant; Treo, treosulfan.

Future directions

Integration of antibody-drug conjugates and naked mAb in conditioning regimens may offer an interesting approach, aiming at further mitigating toxicities. Immunotoxin conjugated mAb, such as anti-CD45 mAb conjugated to saporin with or without 2 Gy TBI, has been shown to specifically deplete hematopoietic precursors in murine models, enabling faster donor engraftment.^69,70 Similarly, anti-CD117 mAb conjugated to saporin or pyrrolobenzodiazepine has been shown to achieve robust multilineage engraftment in mice.^71,72 Anti-CD117 directed chimeric antigen receptor T-cell therapy has been shown to provide selective cytotoxicity to hematopoietic precursors, enabling donor engraftment.⁷³ These antibody- and cell therapy–based conditioning approaches, although preclinical currently, offer a novel framework for a chemotherapy- and radiation-free conditioning approach and may have a selective advantage of preserving thymic function and lymphoid compartment due to lack of CD117 expression on mature lymphocytes, facilitating early immune reconstitution.

Conclusion

The expanding therapeutic arsenal of allogeneic HCT necessitates the development of novel conditioning strategies. Emerging approaches, including optimized chemotherapeutic drugs, disease-specific drugs, radioimmunotherapy, and targeted radiation, provide a foundational framework to further develop highly effective regimens with improved toxicity profiles. Factors such as underlying malignancy, disease status, MRD, and graft- and patient-related factors are critical in decision- making. However, it is prudent to characterize conditioning regimens in tandem with GVHD prophylaxis and posttransplant maintenance therapy as a package—and not an isolation—for a more personalized transplantation approach. With reduced toxicity regimens, one is better able to deliver maintenance therapy.

Conflict-of-interest disclosure

Brenda M. Sandmaier has served on an advisory board for Orca Bio and Actinium Pharmaceuticals.

Naveed Ali: none to disclose.

Off-label drug use

Naveed Ali: Chemotherapy drugs and TBI that are used for conditioning regimens are used off-label.

Brenda M. Sandmaier: Chemotherapy drugs and TBI that are used for conditioning regimens are used off-label.