HLA‐matched related peripheral blood stem cell and bone marrow transplantation with RIC regimens yield comparable outcomes for adult AML

Abstract

Introduction

Understanding differences in clinical outcomes between PBSCT and BMT is important, and this study compared outcomes of HLA‐matched related PBSCT and BMT using reduced‐intensity conditioning (RIC) in adult acute myeloid leukemia (AML) patients.

Methods

Data from 402 patients who underwent either PBSCT (n = 294) or BMT (n = 108) between 2000 and 2022 were analyzed using the Japanese nationwide registry database. The primary endpoint was overall survival (OS), and secondary endpoints included disease‐free survival (DFS), non‐relapse mortality (NRM), and GVHD.

Results

Results indicated no significant difference in 3‐year OS (44.6% for PBSCT vs. 46.9% for BMT, HR 1.173, P = 0.299) and DFS (42.1% vs. 41.8%, HR 1.073, P = 0.639). PBSCT was more beneficial for avoiding relapse (20.3% vs. 12.4%, HR, 0.715, P = 0.059). However, PBSCT was associated with higher NRM (20.3% vs. 12.4%, HR 1.801, P = 0.025) due to more frequent, chronic GVHD (HR 1.889, P = 0.035). Subgroup analysis did not reveal specific patient groups that benefited more from PBSCT or BMT. Incidence of extensive chronic GVHD and NRM has improved in PBSCT recipients in recent years (2014–2022).

Conclusions

We conclude that related PBSCT with RIC regimens offers comparable prognosis to BMT for adult AML patients. Further optimization of prophylactic strategies for chronic GVHD is required to improve outcomes after PBSCT.

1. INTRODUCTION

Allogeneic hematopoietic stem cell transplantation (HSCT) is a curative treatment for hematologic malignancies, including acute myeloid leukemia (AML) [1]. Recently, peripheral blood stem cell transplantation (PBSCT) has begun to see more frequent use than bone marrow transplantation (BMT) due to the reduced burden on donors. PBSC harvesting is less invasive, and it is simpler to coordinate harvesting than with BM [2, 3]. Moreover, engraftment of PBSCs after transplantation is more rapid [4, 5]. A study analyzing stem cell source data up to 2022 found that peripheral blood was the most common source. [4]

Thus, PBSCT is increasingly being employed and has almost overtaken BMT, especially in the United States. The choice of BM or PBSC should be carefully considered because the two treatments can lead to different clinical outcomes, such as graft‐versus‐host disease (GVHD) or engraftment. PBSCT is preferred due to faster engraftment, especially in cases with a higher stem cell yield, but it is also associated with higher incidence of chronic GVHD compared to BM. [6, 7] Another study found no significant differences between PBSCT and BMT in terms of non‐relapse mortality (NRM) and overall survival (OS) [8]. However, most studies included patients transplanted with myeloablative conditioning regimens (MAC), rather than reduced‐intensity conditioning (RIC). Intensity of conditioning regimens and related organ failure immediately after conditioning, i.e., around the time of HSCT, are closely related to the incidence of later acute or chronic GVHD [9, 10]. Therefore, comparison of PBSCT and BMT performed exclusively in an RIC setting [11], with or without total body irradiation (TBI) [12] is needed. AML should be analyzed, as it is the most prevalent disease with an indication for related HSCT. [5]

Accordingly, we performed a retrospective cohort study using the Japanese national registry to compare outcomes of adult patients with AML who underwent related PBSCT or BMT with RIC regimens. In this study, we evaluated post‐transplant complications, including engraftment, GVHD, and infections, and we performed subgroup analysis to identify populations that may benefit from related PBSCT versus BMT.

2. METHODS

2.1. Patients

Data on adult patients (aged 16 years or older) with AML who had undergone their first allogeneic PBSCT or BMT from HLA‐matched, related donors with an RIC regimen between 2000 and 2022 were identified through the Japanese registry (Transplant Registry Unified Management Program) sponsored by the Japanese Society for Transplantation and Cellular Therapy [13, 14]. We excluded patients without survival data. The study was approved by the data management committee of the Transplant Registry Unified Management Program and Kyoto University Ethics Committee and conducted in accordance with the Declaration of Helsinki.

2.2. Study endpoints and definitions of each variable

The primary endpoint was overall survival (OS) after HSCT. Death, regardless of cause, was considered an event. Secondary endpoints were disease‐free survival (DFS), cumulative incidence of relapse, NRM, neutrophil and platelet engraftment, aGVHD, cGVHD, GVHD relapse‐free survival (GRFS), and viral, bacterial and fungal infections. DFS was defined as survival without disease progression or relapse. Both aGVHD and cGVHD were defined and re‐evaluated according to standard criteria [15, 16]. GRFS was defined as survival without death, relapse, development of grade III–IV aGVHD or development of cGVHD that required systemic treatment [17]. Neutrophil and platelet engraftment were defined as the first three consecutive measures with a neutrophil count > 0.5 × 10⁹/L and a platelet count > 20 × 10⁹/L without platelet transfusion after transplantation. Viral infections included cytomegalovirus, Epstein‐Barr virus (including Epstein‐Barr virus post‐transplant lymphoproliferative disorder), and human herpesvirus 6. Bacterial infections included any bacterial infection, excluding febrile neutropenia without proven infection. Fungal infections included candidiasis, proven, probable or possible aspergillosis with previously reported criteria [18] and other proven fungal infections.

Cytogenetic risk was classified in accordance with criteria specified by the National Comprehensive Cancer Network guidelines, which have been described in detail elsewhere [19]. Conditioning intensity was defined according to operational definitions of the National Marrow Donor Program/Center for International Blood and Marrow Transplant Research [20]. HLA matching was assessed using allele data for the HLA‐A, ‐B, ‐C and ‐DRB1 loci [21].

2.3. Statistical analysis

Categorical variables and continuous variables were compared between groups with Fisher's exact test and two‐tailed, unpaired, Student's t‐tests, respectively. Probabilities of OS, DFS and GRFS were estimated using the Kaplan–Meier method and compared among groups with the Cox proportional hazards model. Probabilities of NRM, relapse, engraftment, aGVHD, cGVHD, viral, bacterial or fungal infection, and infection‐related mortality were estimated on the basis of cumulative incidence methods and compared among groups with the Fine–Gray proportional hazards model, considering death without relapse as a competing event for relapse, relapse as a competing event for NRM, death without engraftment as a competing event for neutrophil and platelet engraftment, and death or relapse without GVHD as a competing event for aGVHD and cGVHD. The following variables were considered in multivariate analyses: patient age at the time of HSCT, sex, ECOG PS, cytogenetic risk, disease status at the time of HSCT, donor age, donor‐sex mismatch, usage of TBI, GVHD prophylaxis, addition of ATG, and year of HSCT. All tests were two‐sided, and p < 0.05 was considered statistically significant. All analyses were performed with EZR 1.65 software [22].

3. RESULTS

3.1. Baseline characteristics

A total of 402 patients were eligible for analysis (Table 1). Of these, 294 underwent PBSCT, and 108 received BMT. The median patient age was 58 years (range, 17–70) for PBSCT and 59 years (range, 16–74) for BMT. ECOG PS at time of HSCT was equivalent between the groups. Cytogenetic risk at initial diagnosis was 8.5% favorable, 56.8% intermediate, and 29.3% poor among PBSCT recipients, and 9.3% favorable, 59.3% intermediate, and 22.2% poor for BMT recipients (p = 0.337). Disease status at the time of HSCT was CR in 57.0% of the whole cohort (55.8% in PBSCT and 60.2% in BMT, p = 0.496). Regarding donors, distribution of donor age and recipient–donor sex disparities were comparable between PBSCT and BMT. TBI was used in 65.4% of the whole cohort. Regarding GVHD prophylaxis, in PBSCT, tacrolimus‐based prophylaxis tended to be used (33.0% vs. 20.4%, p = 0.014) than in BMT, and ATG was used only in PBSCT (5.0% vs. 0.0%, p < 0.001). Median follow‐up of survivors in the PBSCT and BMT groups was 5.1 years and 4.2 years, respectively.

TABLE 1.

Patient characteristics.

	Total (n = 402)	PBSCT (n = 294)	BMT (n = 108)	p
Patient age, years, median (range)	58(16‐74)	58(17‐70)	59(16‐74)	0.368
Patient age, years, n (%)				0.349
< 50	60 (14.9)	47 (16.0)	13 (12.0)
≥ 50	342 (85.1)	247 (84.0)	95 (88.0)
Patient sex, n (%)				0.357
Male	246 (61.2)	184 (62.6)	62 (57.4)
Female	156 (38.8)	110 (37.4)	46 (42.6)
ECOG PS, n (%)				0.496
0–1	352 (87.6)	255 (86.7)	97 (89.8)
2–4	50 (12.4)	39 (13.3)	11 (10.2)
Cytogenetic risk, n (%)	,
Favorable	35 (8.7)	25 (8.5)	10 (9.3)	0.842
Intermediate	231 (57.5)	167 (56.8)	64 (59.3)	0.733
Poor	110 (27.4)	86 (29.3)	24 (22.2)	0.168
Unevaluable	26 (6.5)	16 (5.4)	10 (9.3)	0.175
Disease status, n (%)				0.496
CR	229 (57.0)	164 (55.8)	65 (60.2)
Non‐CR	173 (43.0)	130 (44.2)	43 (39.8)
Donor age
years, median (range)	55 (13–84)	54 (15–84)	57 (13–79)	0.147
< 45	89 (22.1)	68 (23.1)	21 (19.4)	0.499
≥ 45	313 (77.9)	226 (76.9)	87 (80.6)
Sex mismatch, n (%)
Match	206 (51.2)	146 (49.7)	60 (55.6)	0.313
Male to Female	85 (21.1)	64 (21.8)	21 (19.4)	0.683
Female to Male	111 (27.6)	84 (28.6)	27 (25.0)	0.530
TBI, n (%)				0.813
No	263 (65.4)	191 (65.0)	72 (66.7)
Yes	139 (34.6)	103 (35.0)	36 (33.3)
Total dose of TBI, Gy, median (range)	4 (2‐8)	4 (2–8)	3 (2–4)	0.521
GVHD prophylaxis, n (%)				0.014
CyA‐based	283 (70.4)	197 (67.0)	86 (79.6)
TAC‐based	119 (29.6)	97 (33.0)	22 (20.4)
Addition of ATG to conditioning, n (%)
No	382 (95.0)	274 (93.2)	108 (100.0)	0.003
Yes	20 (5.0)	20 (6.8)	0 (0.0)
Year of Transplantation, median (range)	2014 (2000–2022)	2013 (2000–2020)	2015 (2004–2022)	0.004
Year of Transplantation, n (%)				0.055
2000–2013	229 (57.0)	176 (59.9)	53 (49.1)
2014–2022	173 (43.0)	118 (40.1)	55 (50.9)

Abbreviations: CyA, cyclosporin A; TAC, tacrolimus.

3.2. Comparisons of gross outcomes between PBSCT and BMT

3‐year OS as the primary endpoint of this study was comparable between PBSCT and BMT (44.6% vs. 46.9%) (Figure 1A), and DFS was also similar (42.1% vs. 41.8%) (Figure 1B). Multivariate analyses revealed that OS (hazard ratio [HR] 1.173, 95% confidence interval [CI] 0.868–1.587, p = 0.299) and DFS (HR, 1.073, 95%CI 0.800–1.438, p = 0.639) did not differ significantly between PBSCT and BMT (Table 2 and Table S1 for univariate analyses). 3‐year GRFS also showed similar prognosis (32.1% vs. 33.0%) (Figure 1C).

FIGURE 1.

Comparison of outcomes in peripheral blood stem cell transplantation (PBSCT) and bone marrow transplantation (BMT) from related human leukocyte antigen (HLA)‐identical donors in the entire study population. (A) overall survival (OS), (B) disease‐free survival (DFS), (C) GVHD relapse‐free survival (GRFS), and cumulative incidence of (D) non‐relapse mortality (NRM) and (E) relapse. Adjusted HRs and p‐values were calculated using the Cox proportional hazards model (A, B, C) and Fine–Gray tests (D, E, F) after adjustment for confounding factors.

TABLE 2.

Multivariate analysis for overall survival (OS), disease‐free survival (DFS), non‐relapse mortality (NRM), and relapse in the entire study population.

	OS			DFS			NRM			Relapse
	HR	95%CI	P	HR	95%CI	P	HR	95%CI	P	HR	95%CI	P
Donor source
PB vs. BM	1.173	0.868–1.587	0.299	1.073	0.800–1.438	0.639	1.801	1.075–3.017	0.025	0.715	0.504–1.013	0.059
Patient age
≥ 50 vs. < 50	1.631	1.063–2.503	0.025	1.607	1.053–2.454	0.028	1.905	0.834–4.355	0.130	1.213	0.761–1.934	0.420
Patient sex
Male vs. Female	0.855	0.574‐1.272	0.439	0.926	0.624–1.375	0.703	0.670	0.333–1.347	0.260	1.201	0.730–1.975	0.470
ECOG PS
2‐4 vs. 0–1	1.776	1.225–2.576	0.002	1.422	0.985‐2.055	0.060	1.107	0.524–2.336	0.790	1.287	0.849–1.95	0.230
Cytogenetic risk
Intermediate vs. favorable	0.743	0.448–1.231	0.248	0.808	0.490–1.333	0.403	1.005	0.492–2.051	0.990	0.839	0.425–1.659	0.610
Poor vs. favorable	1.367	0.803–2.329	0.249	1.466	0.866–2.485	0.155	0.836	0.361–1.936	0.680	1.904	0.942–3.848	0.073
Unevaluable vs. favorable	0.992	0.496–1.984	0.982	1.119	0.562‐2.23	0.749	0.480	0.121‐1.910	0.300	1.614	0.728–3.581	0.240
Disease status
CR vs. non‐CR	2.532	1.897–3.378	0.000	2.710	2.042–3.597	0.000	1.184	0.706‐1.985	0.520	2.793	1.996–3.908	0.000
Donor age
≥ 45 vs. < 45	0.977	0.674–1.415	0.901	0.943	0.657–1.352	0.748	1.495	0.743–3.009	0.260	0.849	0.562–1.283	0.440
Sex mismatch
Female to Male vs. match	1.299	0.933–1.81	0.122	1.207	0.873–1.669	0.256	1.920	1.095–3.368	0.023	0.807	0.535–1.219	0.310
Male to Female vs. match	0.932	0.605–1.434	0.748	1.000	0.654‐1.53	0.998	0.815	0.388–1.714	0.590	1.246	0.732–2.121	0.420
TBI
No vs. Yes	1.070	0.808–1.415	0.637	1.060	0.807‐1.393	0.677	1.353	0.877–2.088	0.170	0.870	0.613–1.237	0.440
GVHD prophylaxis
TAC‐based vs CyA‐based	0.813	0.595–1.110	0.193	0.861	0.638–1.160	0.325	0.795	0.471–1.339	0.390	0.841	0.578–1.224	0.370
Addition of ATG to conditioning
No vs. Yes	0.943	0.469–1.898	0.870	0.921	0.475–1.789	0.809	0.909	0.312–2.647	0.860	0.847	0.421–1.703	0.640
Year of Transplantation
	0.986	0.953‐1.02	0.415	0.993	0.960‐1.027	0.678	0.952	0.900‐1.007	0.085	1.010	0.969‐1.052	0.650

Abbreviations: CI, confidence interval; CyA, cyclosporin A; and TAC, tacrolimus.

On the other hand, NRM and relapse did differ between PBSCT and BMT. 3‐year cumulative incidences of NRM were 20.3% versus 12.4% for PBSCT and BM (Figure 1D), while cumulative incidences of relapse were 37.8% versus 48.0% at 3 years (Figure 1E). Multivariate analyses revealed a higher incidence of NRM in PBSCT (HR, 1.801, 95%CI, 1.075–3.017, p = 0.025), but a lower incidence of relapse (HR, 0.715, 95%CI 0.504–1.013, p = 0.059) (Table 2 and Table S1 for univariate analyses).

3.3. PBSCT showed higher risk of extensive cGVHD

PBSCT thus induced significantly higher incidence of NRM, so we performed detailed analyses for engraftment and post‐HSCT complications. The cumulative incidence of neutrophil engraftment was significantly higher in PBSCT than in BMT (HR, 1.456, p < 0.001) (Figure 2A), while platelet engraftment was comparable (HR, 1.164, p = 0.160) (Figure 2B). Median days after HSCT of neutrophil and platelet engraftment were fewer in PBSCT than BMT (14 vs. 16 days for neutrophils, 20 vs. 23 days for platelets). Among patients who achieved engraftment, respective incidences of secondary graft failure were 0.50% and 0.84% (p = 0.124).

FIGURE 2.

Comparison of engraftment and post‐transplantation complications in peripheral blood stem cell transplantation (PBSCT) and bone marrow transplantation (BMT) from related human leukocyte antigen (HLA)‐identical donors in the entire study population. Cumulative incidence of (A) neutrophil engraftment, (B) platelet engraftment, (C) acute GVHD relapse‐free survival (GVHD) grade II–IV and (D) grade III–IV, and (E) total and (F) extensive chronic GVHD. Adjusted HRs and p‐values were calculated using the Fine–Gray test after adjustment for confounding factors.

Regarding GVHD, the respective cumulative incidence of grade II–IV aGVHD, grade III–IV aGVHD at 100 days and total cGVHD in PBSCT were comparable to those in BMT (21.2% vs. 26.9%, HR, 0.926, 95%CI 0.580–1.479, p = 0.750 for aGVHD grade II–IV [Figure 2C and Table S2], 8.6% versus 4.6%, HR, 2.255, 95%CI 0.871–5.837, p = 0.094 for aGVHD grade III–IV [Figure 2D and Table S2], and 32.0% versus 30.9%, HR, 1.178, 95%CI 0.773–1.797, p = 0.450 for total cGHVD [Figure 2E and Table S2]). On the other hand, the cumulative incidence of extensive cGVHD at 2 years in PBSCT was significantly higher than in BMT (21.6% vs. 14.0%, HR, 1.889, 95%CI 1.046‐3.412, p = 0.035) (Figure 2F and Table S2).

Respective cumulative incidences of bacterial infection at 100 days, and fungal and viral infections at 3 years in PBSCT were comparable to those in BMT (16.9% vs. 19.4%, p = 0.940 for bacterial infection, 8.9% vs. 6.5%, p = 0.450 for fungal infection, and 0.0% vs. 2.2%, p = 0.850 for viral infection) (Figure S1).

3.4. Gross outcomes of NRM and severe GVHD incidence improved over the last decade in PBSCT

Having confirmed that OS is similar between PBSCT and BMT in the whole cohort, we then performed subgroup analyses of OS in order to identify situations in which PBSCT was more beneficial or adverse than BMT (Figure 3). HRs for BMT/PBSCT are basically comparable, indicating that there are no strongly benefited or disadvantaged groups for PBSCT. Among these subgroups, we focused on the year of HSCT subgroup (subgroup of 2000–2013 vs. 2014–2022) because multivariate analyses showed that HCST from 2000 to 2013 presented a higher risk of NRM and cGVHD than HSCT from 2014 to 2022 (Table 2 and Table S2), and that differences of NRM and cGVHD may be the result of donor source selection. Therefore, we first performed multivariate analyses for the following outcomes in each patient subgroup (HSCT in 2000–2013 and those in 2014–2022), respectively: OS, DFS, NRM, relapse, aGVHD grade II–IV, aGVHD grade III–IV, cGVHD (all grade), and extensive cGVHD. 3‐year OS in the 2000–2013 subgroup was 40.8% versus 46.8% (PBSCT vs. BMT, HR 1.391, p = 0.123, Figure 4A), whereas DFS was 37.5% versus 42.1% (HR, 1.273, p = 0.259, Figure 4B). OS in the 2014–2022 subgroup was also comparable between PBSCT and BMT (47.1% vs. 47.4%, HR 0.859, p = 0.499, Figure 4A), as was DFS (45.0% vs. 41.6%, HR 0.793, p = 0.278, Figure 4B). On the other hand, NRM at 3 years was higher in PBSCT than BMT from 2000 to 2013 (29.4% vs. 14.5%, HR, 2.111, P = 0.034), while it was comparable from 2014–2022 (19.2% vs. 15.8%, HR, 1.392, p = 0.420) (Figure 4C). Regarding cumulative incidence of relapse, PBSCT tended to be more beneficial than BMT throughout the year, although the difference was not significant (HR 0.725, p = 0.210, in 2000–2013 subgroup, and HR 0.649, p = 0.096 in 2013–2022 subgroup) (Figure 4D). Then, we analyzed the incidence of GVHD in each subgroup per HSCT year and found a positive trend for PBSCT. Multivariate analyses indicated that in PBSCT compared with BMT, higher cumulative incidence of aGVHD grade III–IV and extensive cGVHD was mitigated between 2000–2013 and 2014–2022 (PBSCT vs. BMT, HR 3.584, p = 0.021 in 2000–2013 and HR 2.279, p = 0.220 in 2014–2022 for aGVHD grade III–IV [Figure 4F], and HR 2.106, p = 0.046 in 2000–2013 and HR 1.528, p = 0.380 in 2014–2022 for extensive cGVHD [Figure 4H]). Cumulative incidence of aGVHD grade II–IV and total cGVHD did not change significantly during this period (Figure 4E,G).

FIGURE 3.

Subgroup analysis. Table and forest plot of subgroup analysis.

FIGURE 4.

Comparison of outcomes and post‐transplantation complications in peripheral blood stem cell transplantation (PBSCT) and bone marrow transplantation (BMT) from 2000 to 2013 and from 2014 to 2022. (A) overall survival (OS), (B) disease‐free survival (DFS), cumulative incidence of (C) non‐relapse mortality (NRM), (D) relapse, acute GVHD relapse‐free survival (GVHD) (E) grade II–IV and (F) grade III–IV, and (G) total and (H) extensive chronic GVHD. Adjusted HRs and p‐values were calculated using the Cox proportional hazards model (A, B) and the Fine–Gray test (C, D, E, F, G, H) after adjustment for confounding factors. *p < 0.05.

4. DISCUSSION

In this retrospective cohort study using the Japanese nationwide registry database, we analyzed outcome differences between PBSCT and BMT from related donors with RIC regimens in adult AML patients. There were two major findings. First, OS and DFS in PBSCT were not inferior to those in BMT. Second, higher incidence of extensive cGVHD resulted in significantly higher NRM in PBSCT compared to BMT, although recent years have shown improvements in the incidence of aGVHD grade III–IV and extensive cGVHD in PBSCT.

First, we confirmed that OS and DFS after related PBSCT with RIC regimens are not inferior to those after related BMT. These results are consistent with previous reports, shown mainly with MAC regimens [23, 24, 25, 26, 27]. Previous reports have shown that PBSCT from unrelated donors is associated with more beneficial effects than BMT [28, 29], whereas our results are similar between PBSCT and BMT in related HSCT. These differences may be explained by discrepancies in patient background between unrelated and related HSCTs. Cryopreservable PBSCs are often chosen in cases of uncontrollable infections and chemotherapy refractory conditions [30, 31], and cases in which such benefits of PBSCT should be applied may result in selection of related donors rather than unrelated donors because of the shorter coordination duration. Therefore, we speculate that patient cohorts using related PBSCT may be at some disadvantage in terms of infection or disease control compared with unrelated PBSCT. Such discrepancies are not often observed in BMT, in which cryopreservation of harvested BM is not routinely recommended. These speculations may explain the comparable outcomes in PBSCT and BMT, i.e., not superior in PBSCT, in related settings. In our cohort, there was no difference of cytogenetic risk and disease status at HSCT (Table 1), but other unmeasurable factors, including comorbidity index [32] and past chemotherapy history data, should be included in future models in order to prove the foregoing hypothesis.

Second, regarding post‐transplant complications, we found that PBSCT has a higher risk of extensive cGVHD and NRM than BMT. This result is consistent with previous reports mainly including MAC‐HSCT [33, 34], and may be associated with abundant effector and naïve T cells in PBSC grafts [35, 36]. It is important to quantify the relatively higher incidence of cGVHD treated with RIC regimens because incidence of GVHD induced by RIC‐HSCT is not widely discussed. Extensive cGVHD can trigger inflammation for certain periods, inducing various types of organ failure and associated poor prognosis [37].

In our sensitivity analyses to identify those subcohorts most likely to benefit from PBSCT, we found mitigation of cGVHD and related‐NRM in the cohort since 2014. Development of relatively novel immunosuppressants, including mycophenolate mofetil (MMF) and mesenchymal stromal cells (MSC) [38] may mitigate cGVHD in PBSCT patients, although non‐epidemiological or statistical confirmation were not available. More recently developed GVHD prophylaxis or treatment, including ibrutinib [39], ruxolitinib [40], extracorporeal photopheresis [41], and belumosudil [42]can be more effective for treatment of GVHD, and results of PBSCT are expected to improve further in the next decade.

This study has shown the usefulness of PBSC from HLA‐matched, related donors in adult AML patients conditioned with RIC regimens, but several limitations need to be mentioned. First, there are missing data regarding usage or dosages of GVHD prophylactic agents such as MTX, MMF, and antithymocyte globulin. For instance, several previous reports indicated that omitting the day 11 dose of MTX was a risk factor for both acute and chronic GVHD [44, 45]. Additionally, the combination of MMF and MTX with CyA was reported to be beneficial for aGVHD and overall survival [46]. Furthermore, serum concentrations of Tac and CyA were not included in our dataset, although deviation may be directly associated with a higher risk of aGVHD and subsequent cGVHD [47, 48, 49].

We hypothesize that individual adjustments made by transplant physicians, such as avoiding omitting Day 11, in addition to increasing the dosage of MTX, may have improved cGVHD in recent years. In Japan, there is no clear guidance on MTX dosage, and the actual doses were 15 mg/m²‐10 mg/m²‐10 mg/m²‐(10 mg/m²), 10 mg/m²‐7 mg/m²‐7 mg/m²‐(7 mg/m²), 5 mg/m²‐5 mg/m²‐5 mg/m²‐(5 mg/m²). Due to missing values, the analysis was not possible, but there was no use of 5 mg/m²‐5 mg/m²‐5 mg/m²‐(5 mg/m²) dosage in PBSCT since 2014. Similarly, there was no omitting of Day 11 in PBSCT after 2014, while BMT was omitted in all cases. Therefore, to reduce the risk of GVHD and to improve the safety of PBSCT, a detailed evaluation is necessary to optimize GVHD prevention strategy in a larger cohort including MTX dosage and whether omitting Day 11 of MTX or not.

In conclusion, outcomes after PBSCT and BMT with RIC for adult AML patients from related HLA identical donors were comparable in terms of OS and DFS. NRM in PBSCT was higher than that in BMT because of higher risk of extensive chronic GVHD. However, this high risk of GVHD may have been overcome in recent years, although the responsible factor is unknown. Therefore, our findings justify the donor selection of either PB or BM depending on logistic circumstances and donor preferences in adult AML patients conditioned with RIC regimens. Further optimization of the prophylactic strategy for cGVHD is required to improve outcomes after PBSCT.

AUTHOR CONTRIBUTIONS

Takaya Mitsuyoshi and Yasuyuki Arai planned this study, analyzed the data, and wrote the manuscript. Tadakazu Kondo, Takahito Kawata, Shigeki Hirabayashi, Junya Kanda, and Masamitsu Yanada advised in the analysis. Masatsugu Tanaka, Yasuo Mori, Noriko Doki, Tetsuya Nishida, Takeharu Kotani, Masao Ogata, Takayuki Tabayashi, Tetsuya Eto, Masashi Sawa, Kazunori Imada, Tatsuo Ichinohe, and Yoshiko Atsuta reviewed and provided critiques on the manuscript.

CONFLICT OF INTEREST STATEMENT

The authors declare no conflicts of interest.

FUNDING INFORMATION

The authors did not receive support from any organization for the submitted work.

ETHICS STATEMENT

The study was approved by Kyoto University Ethics Committee and conducted in accordance with the Declaration of Helsinki.

CLINICAL TRIAL REGISTRATION

The authors have confirmed clinical trial registration is not needed for this submission.

PATIENT CONSENT STATEMENT

The authors have confirmed all patients consented to be enrolled in this study.

Supporting information

Supporting information

Mitsuyoshi T, Arai Y, Kondo T, Kawata T, Hirabayashi S, Tanaka M, et al. HLA‐matched related peripheral blood stem cell and bone marrow transplantation with RIC regimens yield comparable outcomes for adult AML. eJHaem. 2025;6:e21088. 10.1002/jha2.1088

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.