ABSTRACT
Background and Aims
Haploidentical hematopoietic stem cell transplantation (haplo‐HSCT) provides curative therapy for patients lacking HLA‐matched donors, yet optimal donor characteristics remain uncertain. This study assessed the effect of donor relationship, age, sex, and ABO compatibility on outcomes of haplo‐HSCT for acute leukemia under a uniform ATG/post‐transplant cyclophosphamide (PTCy)‐based protocol.
Methods
In this retrospective single‐center cohort, 159 adults with acute myeloid or lymphoblastic leukemia received peripheral blood haplo‐HSCT following myeloablative busulfan/cyclophosphamide plus ATG and PTCy (2015–2022). Overall survival (OS), disease‐free survival (DFS), GVHD‐free/relapse‐free survival (GRFS), and relapse incidence were analyzed using Kaplan–Meier and Fine–Gray methods; predictors were evaluated by multivariable Cox regression.
Results
Of 159 patients, 96/159 (60.4%) had AML and 63/159 (39.6%) had ALL. Donors were siblings in 90/159 (56.6%), parents in 48/159 (30.2%), and offspring in 21/159 (13.2%). Median recipient age was 27 years (range, 15–59), and median donor age was 31 years (range, 2–65). At 36 months, unadjusted outcomes varied across donor groups. In AML, OS was 52.7% (95% CI, 38.8–64.9) with sibling donors, 52.3% (95% CI, 29.4–71.0) with parent donors, and 32.1% (95% CI, 9.5–57.9) with offspring donors. In ALL, offspring donors were associated with numerically lower survival and higher relapse. In multivariable analysis, ALL (HR, 1.99; 95% CI, 1.25–3.17; p = 0.004), advanced remission status (HR, 1.79; 95% CI, 1.16–2.77; p = 0.01), and older recipient age (HR, 1.03 per year; 95% CI, 1.01–1.05; p < 0.001) were independently associated with worse OS. Donor's age, sex, relationship, and ABO compatibility were not independently associated with major outcomes.
Conclusions
In this single‐center cohort of adults undergoing haploidentical HSCT with a uniform ATG/PTCy‐based platform, donor relationship was associated with descriptive differences in unadjusted outcomes, but it was not an independent predictor in multivariable models. The inferior outcomes observed with offspring donors appeared to be influenced in part by older recipient age.
Trial Registration
Not applicable (retrospective study).
Keywords: acute lymphoblastic leukemia, acute myeloid leukemia, donor selection, graft‐versus‐host disease, haploidentical stem cell transplantation, survival analysis
1. Introduction
Allogeneic hematopoietic stem cell transplantation (HSCT) remains a potentially curative option for many hematologic malignancies, including acute myeloid leukemia (AML) and acute lymphoblastic leukemia (ALL). However, a significant number of patients do not have access to an HLA‐matched sibling or unrelated donor, which makes haploidentical HSCT (haplo‐HSCT) an increasingly vital alternative [1, 2].
The introduction of post‐transplant cyclophosphamide (PTCy) has significantly reduced graft‐versus‐host disease (GVHD) and improved survival in haplo‐HSCT [1, 3, 4]. Despite these advances, donor selection remains challenging, as factors beyond HLA compatibility, such as age, sex, and biological relationship, influence transplant outcomes.
Haploidentical donors typically include first‐degree relatives such as parents, siblings, or children. In rare cases, more distantly related family members may also be considered [2, 5]. When multiple potential donors are available, selecting the optimal donor requires careful consideration of characteristics that may affect survival, GVHD, and relapse risk [6, 7].
For instance, donor age may influence outcomes. Some studies suggest younger donors reduce the risk of GVHD, lower non‐relapse mortality (NRM), and improve survival [3, 8], while other reports have found no significant association between donor age and overall transplant outcomes [3, 4, 6, 9]. In female donors, parity may increase rates of chronic and acute GVHD [10]. However, this risk may be counterbalanced by stronger graft‐versus‐leukemia effects and lower relapse rates. These benefits do not consistently improve survival [7, 11].
The biological relationship between donor and recipient also plays a critical role. Some evidence suggests that paternal donors may be associated with lower rates of GVHD and NRM compared to maternal donors [12]. While sibling donors are often preferred, their impact on outcomes varies across studies [8, 13].
Sex mismatch, particularly female‐to‐male (FtoM) transplants, has been linked to higher risks of acute and chronic GVHD and increased NRM in certain settings [14, 15, 16]. However, findings on donor sex and sex mismatch vary across cohorts, and their overall impact remains under debate [17, 18].
Altogether, while haplo‐HSCT has become a vital alternative for patients lacking matched donors, selecting the optimal donor remains a clinical challenge. In this single‐center retrospective cohort study, we evaluated how donor‐related factors, including donor age, sex, and biological relationship, affect key outcomes such as overall survival (OS), disease‐free survival (DFS), GVHD, and relapse after haploidentical HSCT in adults with acute leukemia. Our aim was to provide practical evidence to guide donor selection and improve transplantation outcomes.
2. Materials and Methods
2.1. Study Design and Setting
This was a single‐center retrospective cohort study conducted at Shariati Hospital, Tehran University of Medical Sciences, Tehran, Iran. The primary objective was to evaluate whether donor‐recipient biological relationship was associated with outcomes after haploidentical HSCT in patients with acute leukemia. Secondary objectives were to examine the associations of donor age, donor sex, and ABO blood group compatibility with transplantation outcomes. Analyses comparing outcomes by disease subtype and the assessment of recipient age as a confounder were considered exploratory.
2.2. Patients
We included consecutive patients aged 15 years or older with AML or ALL who underwent haploidentical HSCT between January 2015 and December 2022. Eligible patients were in complete remission at transplantation, including first complete remission (CR1) or second or subsequent complete remission (CR2+), and received peripheral blood stem cell grafts. Patients with prior allogeneic HSCT, refractory disease at transplantation, or major comorbidities precluding myeloablative conditioning were excluded. Follow‐up data were updated through July 2024.
2.3. Data Collection
Clinical and transplant‐related data were obtained from institutional electronic medical records, transplant charts, outpatient follow‐up records, and structured telephone follow‐up when necessary. Collected variables included recipient age and sex, diagnosis, remission status at transplantation, donor age and sex, donor‐recipient biological relationship, ABO compatibility, and graft cell doses.
2.4. Ethical Approval
The study was conducted in accordance with the Declaration of Helsinki and approved by the Ethics Committee of the Research Center for Oncology, Hematology, and Cell Therapy (HORCSCT), Tehran University of Medical Sciences (IR.TUMS.HORCSCT.REC.1402.060). Written informed consent had been obtained from patients for the use of anonymized clinical data for research purposes.
2.5. Donor Selection and HLA Considerations
All donor‐recipient pairs fulfilled institutional criteria for haploidentical transplantation according to routine clinical practice during the study period. Donors were first‐degree relatives, including siblings, parents, or offspring. HLA typing was performed according to institutional standards; however, detailed locus‐level mismatch data (e.g., HLA‐A, ‐B, ‐C, ‐DRB1, and ‐DQB1), typing resolution (high vs. low), and donor‐specific anti‐HLA antibodies (DSAs) were not consistently retrievable in the retrospective dataset.
2.6. Transplant Procedure
A total of 159 patients underwent haploidentical HSCT using a uniform myeloablative conditioning platform. Conditioning consisted of busulfan administered orally at 4 mg/kg/day or intravenously at 3.2 mg/kg/day from Day −6 to Day −3, combined with cyclophosphamide 60 mg/kg/day on Days −3 and −2.
GVHD prophylaxis consisted of a modified combination of rabbit anti‐thymocyte globulin (ATG) and PTCy. Rabbit ATG (Thymoglobulin; Genzyme, Lyon, France) was administered at 2.5 mg/kg/day on Days −3, −2, and −1. PTCy was administered intravenously at 40 mg/kg/day on Days +3 and +4. Cyclosporine A was started intravenously at 1.5 mg/kg/day on Day −2, increased to 3 mg/kg/day by Day +7, and then converted to oral administration when tolerated.
Supportive care was delivered according to institutional standards. Infection prophylaxis included acyclovir, fluconazole, and trimethoprim/sulfamethoxazole. Cytomegalovirus was monitored twice weekly by polymerase chain reaction, and preemptive antiviral treatment was initiated when indicated. Routine post‐transplant maintenance therapy was not administered. In selected AML patients, preemptive interventions such as hypomethylating agents, FLT3 inhibitors, or donor lymphocyte infusion were considered in the setting of measurable residual disease positivity or impaired donor chimerism.
2.7. Outcomes and Definitions
All outcomes were measured from the date of transplantation. The primary outcome was OS, defined as the time from transplantation to death from any cause. DFS was defined as the time from transplantation to relapse or death, whichever occurred first. GVHD‐free/relapse‐free survival (GRFS) was defined as survival without Grade III–IV acute GVHD, extensive chronic GVHD requiring systemic immunosuppression, relapse, or death.
Relapse incidence was analyzed as a competing‐risk endpoint. NRM was defined as death without prior relapse and was reported descriptively because the number of events was limited. Neutrophil engraftment was defined as the first of 3 consecutive days with an absolute neutrophil count of at least 0.5 × 109/L. Platelet engraftment was defined as the first of 7 consecutive days with a platelet count of at least 20 × 109/L without transfusion support.
Acute GVHD was graded according to the modified Glucksberg criteria, and chronic GVHD was classified according to the 2014 NIH consensus criteria [19, 20]. Complete remission status was confirmed by bone marrow evaluation and flow cytometry according to institutional practice. ABO compatibility was classified as matched, minor mismatch, major mismatch, or bidirectional mismatch [21].
2.8. Statistical Analysis
Continuous variables are presented as median (range), and categorical variables as number/total (%). Baseline characteristics were compared using the Mann–Whitney U test or Kruskal–Wallis test for continuous variables and the chi‐square test or Fisher's exact test for categorical variables, as appropriate.
Median follow‐up was estimated using the reverse Kaplan–Meier method. OS, DFS, and GRFS were estimated using the Kaplan–Meier method and compared with the log‐rank test. Engraftment, relapse, NRM, and GVHD were analyzed using competing‐risk methods, and between‐group comparisons were performed using Fine–Gray subdistribution hazard models.
Univariable and multivariable Cox proportional hazards regression models were used to evaluate factors associated with OS. Variables with p < 0.20 in univariable analysis and variables considered clinically relevant were entered into multivariable models. Because recipients of offspring grafts were older than recipients of sibling or parental grafts, we additionally performed exploratory analyses to assess whether recipient age confounded the association between donor relationship and major outcomes. All tests were 2‐sided, and p < 0.05 was considered statistically significant. Statistical analyses were performed using Stata version 11.2 (StataCorp, College Station, TX, USA) and R version 4.4.2.
3. Results
3.1. Patient and Donor Characteristics
A total of 159 patients underwent haploidentical HSCT. The median recipient age was 27 years (range, 15–59), and the median donor age was 31 years (range, 2–65). The median follow‐up among surviving patients was 40.4 months (95% CI, 31.9–52.5).
Donors were siblings in 90/159 patients (56.6%), parents in 48/159 (30.2%), and offspring in 21/159 (13.2%). Male donors accounted for 97/159 (61.0%) and female donors for 62/159 (39.0%). Among recipients, 101/159 (63.5%) were male, and 58/159 (36.5%) were female. Underlying diagnoses included AML in 96/159 patients (60.4%) and ALL in 63/159 (39.6%). At transplantation, 64/159 patients (40.3%) were in CR1, and 95/159 (59.7%) were in CR2+.
ABO compatibility was present in 99/159 (62.3%) transplants, whereas 25/159 (15.7%) had a minor mismatch, 27/159 (17.0%) a major mismatch, and 8/159 (5.0%) a bidirectional mismatch. The median CD34+ cell dose was 8.0 × 106/kg (range, 2.04–16.14), and the median CD3+ cell dose was 306 × 106/kg (range, 103–704.29).
3.2. Engraftment and OS
Neutrophil engraftment occurred rapidly, with a 30‐day cumulative incidence of 96.2% (95% CI 92.0–98.6) among haploidentical recipients. Platelet engraftment was achieved in 94.6% (95% CI 90.0–97.5) by Day +30.
In AML patients, the probability of 36‐month OS demonstrated 52.7% (95% CI, 38.8–64.9), 52.3% (95% CI, 29.4–71.0), and 32.1% (95% CI, 9.5–57.9) for sibling donors, parent, and offspring donors, respectively. In ALL patients, 36‐month OS was 37.8% (95% CI, 19.5–56.1) with sibling donors, 28.4% (95% CI, 11.4–48.4) with parent donors, and 14.3% (95% CI, 0.7–46.5) with offspring donors. Detailed OS probabilities are shown in Tables 1 and 2. Kaplan–Meier survival curves by donor–recipient relationship for AML and ALL are also shown in Figure 1.
Table 1.
Estimated probabilities of OS, DFS, GRFS, and cumulative incidence of relapse at 36 months in patients with AML according to donor–recipient relationship.
| Donor relation | OS % (95% CI) | p value | DFS % (95% CI) | p value | GRFS % (95% CI) | p value | Relapse % (95% CI) | p value |
|---|---|---|---|---|---|---|---|---|
| Sibling | 52.73 (38.75–64.92) | Ref | 48.92 (35.01–61.45) | Ref | 32.58 (20.56–45.16) | Ref | 27.81 (16.12–45.33) | Ref |
| Parent | 52.32 (29.35–71.02) | p = 0.96 | 44.32 (23.16–63.58) | p = 0.50 | 29.46 (11.91–49.57) | p = 0.53 | 31.93 (15.36–58.81) | p = 0.22 |
| Offspring | 32.14 (9.49–57.86) | p = 0.18 | 32.14 (9.49–57.86) | p = 0.25 | 19.05 (3.61–43.69) | p = 0.30 | 30.77 (7.18–83.72) | p = 0.98 |
Note: Values are presented as estimated probability or cumulative incidence with 95% confidence intervals (CIs). p values are shown relative to the sibling donor group as the reference category. OS, DFS, and GRFS were compared using Cox proportional hazards models; relapse was analyzed using competing‐risk methods with Fine and Gray's subdistribution hazard model.
Abbreviations: AML, acute myeloid leukemia; CI, confidence interval; DFS, disease‐free survival; GRFS, GVHD‐free/relapse‐free survival; OS, overall survival.
Table 2.
Estimated probabilities of OS, DFS, GRFS, and cumulative incidence of relapse at 36 months in patients with ALL according to donor–recipient relationship.
| Donor relation | OS % (95% CI) | p value | DFS % (95% CI) | p value | GRFS % (95% CI) | p value | Relapse % (95% CI) | p value |
|---|---|---|---|---|---|---|---|---|
| Sibling | 37.83 (19.54–56.06) | Ref | 33.41 (15.99–51.91) | Ref | 27.60 (12.07–45.66) | Ref | 21.07 (7.10–53.26) | Ref |
| Parent | 28.44 (11.35–48.37) | p = 0.41 | 31.50 (14.72–49.84) | p = 0.40 | 30.77 (14.63–48.55) | p = 0.31 | 31.47 (14.26–60.46) | p = 0.21 |
| Offspring | 14.29 (0.71–46.49) | p = 0.20 | 14.29 (0.71–46.49) | p = 0.18 | 14.29 (0.71–46.49) | p = 0.39 | 73.33 (31.39–99.03) | p = 0.01 |
Note: Values are presented as estimated probability or cumulative incidence with 95% confidence intervals (CIs). p values are shown relative to the sibling donor group as the reference category. OS, DFS, and GRFS were compared using Cox proportional hazards models; relapse was analyzed using competing‐risk methods with Fine and Gray's subdistribution hazard model.
Abbreviations: ALL, acute lymphoblastic leukemia; CI, confidence interval; DFS, disease‐free survival; GRFS, GVHD‐free/relapse‐free survival; OS, overall survival.
Figure 1.
Overall survival according to donor–recipient relationship in haploidentical HSCT for acute leukemia. Kaplan–Meier curves show overall survival after transplantation in patients with acute myeloid leukemia (AML) (A) and acute lymphoblastic leukemia (ALL) (B), stratified by donor–recipient biological relationship: sibling, parent, and offspring. Time is shown from transplantation in months. Tick marks indicate censored observations. If displayed, shaded areas represent 95% confidence intervals.
3.3. Disease‐Free Survival and Graft‐Versus‐Host Disease‐Free/Relapse Survival
As shown in Table 1, the 36‐month probability of DFS was 48.9% (95% CI, 35.0–61.5) with sibling donors, 44.3% (95% CI, 23.2–63.6) with parent donors, and 32.1% (95% CI, 9.5–57.9) with offspring donors in AML patients. GRFS at 36 months for AML patients was 32.6% (95% CI, 20.6–45.2) with sibling donors, 29.5% (95% CI, 11.9–49.6) with parent donors, and 19.1% (95% CI, 3.6–43.7) with offspring donors.
In ALL, 36‐month DFS was 33.4% (95% CI, 16.0–51.9), 31.5% (95% CI, 14.7–49.8), and 14.3% (95% CI, 0.7–46.5) for sibling, parent, and offspring donors, respectively. GRFS was 27.6% (95% CI, 12.1–45.7), 30.8% (95% CI, 14.6–48.6), and 14.3% (95% CI, 0.7–46.5), respectively (Table 2).
3.4. Relapse Incidence
In AML, the cumulative incidence of relapse at 36 months was 27.8% (95% CI, 16.1–45.3) with sibling donors, 31.9% (95% CI, 15.4–58.8) with parent donors, and 30.8% (95% CI, 7.2–83.7) with offspring donors. For ALL, offspring donors had the highest cumulative incidence of relapse, estimated at 73.3% (95% CI, 31.4%–99.0%) at 36 months. By comparison, the 36‐month cumulative incidence of relapse was 21.1% (95% CI, 7.1%–53.3%) for sibling donors and 31.5% (95% CI, 14.3%–60.5%) for parent donors. Because the offspring donor subgroup was small, especially among patients with ALL, confidence intervals were wide, and the precision of these estimates was limited. Detailed relapse incidences at 36 months are shown in Tables 1 and 2. Cumulative incidence curves for relapse by donor–recipient relationship in AML and ALL are shown in Figure 2.
Figure 2.
Cumulative incidence of relapse according to donor–recipient relationship in haploidentical HSCT for acute leukemia. Competing‐risk curves show the cumulative incidence of relapse after transplantation in patients with AML (A) and ALL (B), stratified by donor–recipient biological relationship: sibling, parent, and offspring. Death without relapse was treated as a competing event. Time is shown from transplantation in months. If displayed, shaded areas represent 95% confidence intervals.
3.5. Acute Graft‐Versus‐Host Disease, Chronic Graft‐Versus‐Host Disease, and Non‐Relapse Mortality
The cumulative incidence of Grade III–IV acute GVHD at Day +90 and extensive chronic GVHD at 12 months is presented in Table S1. Due to the limited number of events, donor‐related predictors of acute and chronic GVHD were not included in multivariable models; therefore, these analyses remain descriptive.
In AML patients, the 90‐day cumulative incidence of Grade III–IV acute GVHD was 9.2% (95% CI, 3.9–20.8) for sibling donors, 14.3% (95% CI, 4.8–38.0) for parent donors, and 23.1% (95% CI, 8.1–55.8) for offspring donors. In ALL patients, the corresponding cumulative incidence was 8.0% (95% CI, 2.0–28.5) for sibling donors and 22.4% (95% CI, 9.9–45.8) for parent donors. No events were observed in recipients of offspring grafts. Overall, although numerically higher rates were observed in offspring donors in AML and in parent donors in ALL, these estimates were associated with wide confidence intervals, reflecting limited sample size, and should be interpreted with caution.
In AML patients, the 12‐month cumulative incidence of extensive chronic GVHD was 19.1% (95% CI, 10.0–34.6) for sibling donors, 10.0% (95% CI, 2.6–34.4) for parent donors, and 40.0% (95% CI, 14.5–81.0) for offspring donors. In ALL patients, the cumulative incidence of extensive chronic GVHD was 21.6% (95% CI, 8.6–48.3) for sibling donors and 11.3% (95% CI, 2.9–38.6) for parent donors, while no cases were observed in offspring donors. Given the limited number of events in several subgroups, particularly among offspring donors, these estimates are associated with substantial uncertainty and should be interpreted as exploratory.
The cumulative incidence of NRM at 36 months varied across donor types and primary disease. In AML patients, the 36‐month cumulative incidence of NRM was 31.8% (95% CI, 21.1–46.0) for sibling donors, 28.6% (95% CI, 13.9–53.1) for parent donors, and 53.3% (95% CI, 29.8–80.4) for offspring donors. In ALL patients, the corresponding 36‐month cumulative incidence of NRM was 56.2% (95% CI, 37.7–76.3) for sibling donors, 48.1% (95% CI, 30.4–69.5) for parent donors, and 46.4% (95% CI, 17.5–86.8) for offspring donors. Although numerically higher NRM was observed in offspring donors, particularly in AML, confidence intervals were wide due to small subgroup sizes, and these differences should be interpreted cautiously. NRM results are provided in Table S1.
3.6. Univariable and Multivariable Cox Regression Analysis
Univariable Cox regression identified factors associated with OS. Patients with ALL had worse OS compared to patients with AML (HR = 1.57, 95% CI, 1.03–2.37, p = 0.03). Advanced disease stage (≥ CR2) was also associated with worse OS (HR = 1.67, 95% CI 1.08–2.58, p = 0.02). Increasing patient age negatively impacted OS (HR = 1.02 per year, 95% CI 1.00–1.04, p = 0.02).
No significant associations were observed for ABO matching, donor age, donor sex, or FtoM transplants in the univariable analysis. Donor relationship (other than sibling) showed a non‐significant trend toward worse OS (HR 1.29, 95% CI 0.98–1.71, p = 0.07).
In multivariable Cox regression analysis, ALL diagnosis (HR = 1.99, 95% CI 1.25–3.17, p = 0.004), advanced disease stage (≥ CR2; HR = 1.79, 95% CI 1.16–2.77, p = 0.01), and increasing patient age (HR = 1.03 per year, 95% CI 1.01–1.05, p < 0.001) remained independently associated with worse OS. Donor age, donor sex, FtoM transplants, and donor relationship did not significantly influence OS in the multivariable model. Detailed univariable and multivariable results for OS are shown in Table 3. Sensitivity analysis using donor age categories (≤ 35 vs. > 35 years) did not show a significant association with OS, DFS, or relapse (data not shown).
Table 3.
Univariable and multivariable Cox proportional hazards regression analyses for overall survival in patients with acute leukemia undergoing haploidentical HSCT.
| Variable | Univariable HR (95% CI) | p value | Multivariable HR (95% CI) | p value |
|---|---|---|---|---|
| ALL vs. AML | 1.57 (1.03–2.37) | p = 0.03 | 1.99 (1.25–3.17) | p = 0.004 |
| CR2+ vs. CR1 | 1.67 (1.08–2.58) | p = 0.02 | 1.79 (1.16–2.77) | p = 0.01 |
| Patient age (per year) | 1.02 (1.00–1.04) | p = 0.02 | 1.03 (1.01–1.05) | p < 0.001 |
| Donor age | 1.01 (1.00–1.02) | p = 0.18 | 1.01 (1.00–1.02) | p = 0.18 |
| Donor sex (M vs. F) | 1.23 (0.80–1.90) | p = 0.34 | — | — |
| F → M transplant | 1.47 (0.89–2.45) | p = 0.14 | 1.26 (0.75–2.12) | p = 0.37 |
| ABO mismatch | 1.00 (0.80–1.25) | p > 0.99 | — | — |
| Donor relation | 1.29 (0.98–1.71) | p = 0.07 | 1.05 (0.78–1.42) | p = 0.75 |
Note: HRs, 95% CIs, and p values are shown for the association between clinical and transplant‐related variables and overall survival. Reference categories were AML for disease type, female for sex, CR1 for remission status, and sibling donor for donor relationship. Variables entered into the multivariable model were selected based on clinical relevance and/or p < 0.20 in univariable analysis.
Abbreviations: ALL, acute lymphoblastic leukemia; AML, acute myeloid leukemia; CI, confidence interval; HR, hazard ratio.
3.7. Impact of Recipient Age on Donor‐Relationship Comparisons
Recipients of offspring grafts were significantly older than recipients of sibling or parental grafts in both disease groups (AML: median age, 47 vs. 28 vs. 24 years, respectively; p < 0.001; ALL: 40 vs. 23 vs. 20 years, respectively; p < 0.001). Because recipient age differed substantially across donor groups, we explored its potential confounding effect on the association between donor relationship and transplantation outcomes. After adjustment for recipient age, hazard ratios for offspring versus sibling donors were attenuated for OS, DFS, and GRFS, particularly in ALL, while confidence intervals remained wide because of the limited sample size of the offspring subgroup. These findings suggest that recipient age contributed importantly to the unadjusted differences observed across donor‐recipient relationship groups. Adjusted model results are shown in Table 4.
Table 4.
Association between donor–recipient relationship and outcomes before and after adjustment for recipient age.
| Disease | Outcome | Comparison | Unadjusted HR (95% CI) | p value | Adjusted HR (95% CI) | p value | % change |
|---|---|---|---|---|---|---|---|
| ALL | DFS | Offspring vs. sibling | 1.88 (0.74–4.77) | p = 0.18 | 0.46 (0.13–1.66) | p = 0.23 | −75.7 |
| ALL | DFS | Parent vs. sibling | 1.33 (0.68–2.58) | p = 0.40 | 1.81 (0.89–3.68) | p = 0.10 | 36.4 |
| ALL | GRFS | Offspring vs. sibling | 1.51 (0.60–3.78) | p = 0.38 | 0.48 (0.14–1.64) | p = 0.24 | −68.5 |
| ALL | GRFS | Parent vs. sibling | 1.34 (0.71–2.52) | p = 0.37 | 1.69 (0.87–3.28) | p = 0.12 | 26.3 |
| ALL | OS | Offspring vs. sibling | 1.76 (0.69–4.46) | p = 0.23 | 0.45 (0.13–1.62) | p = 0.22 | −74.2 |
| ALL | OS | Parent vs. sibling | 1.23 (0.63–2.39) | p = 0.54 | 1.62 (0.81–3.25) | p = 0.18 | 31.6 |
| ALL | Relapse | Offspring vs. sibling | 5.57 (1.15–27.02) | p = 0.033 | 10.19 (0.80–129.66) | p = 0.074 | 83.0 |
| ALL | Relapse | Parent vs. sibling | 2.12 (0.53–8.52) | p = 0.29 | 1.92 (0.42–8.70) | p = 0.40 | −9.5 |
| AML | DFS | Offspring vs. sibling | 1.55 (0.73–3.29) | p = 0.25 | 1.17 (0.50–2.73) | p = 0.72 | −24.8 |
| AML | DFS | Parent vs. sibling | 1.26 (0.64–2.48) | p = 0.50 | 1.33 (0.67–2.64) | p = 0.41 | 5.7 |
| AML | GRFS | Offspring vs. sibling | 1.43 (0.73–2.79) | p = 0.30 | 0.89 (0.41–1.93) | p = 0.77 | −37.8 |
| AML | GRFS | Parent vs. sibling | 1.21 (0.66–2.20) | p = 0.53 | 1.28 (0.70–2.34) | p = 0.41 | 6.2 |
| AML | OS | Offspring vs. sibling | 1.66 (0.78–3.52) | p = 0.19 | 1.22 (0.52–2.88) | p = 0.65 | −26.4 |
| AML | OS | Parent vs. sibling | 1.07 (0.53–2.15) | p = 0.85 | 1.12 (0.56–2.26) | p = 0.75 | 4.9 |
| AML | Relapse | Offspring vs. sibling | 0.81 (0.18–3.61) | p = 0.78 | 0.78 (0.19–3.27) | p = 0.74 | −2.8 |
| AML | Relapse | Parent vs. sibling | 1.86 (0.69–5.01) | p = 0.22 | 1.86 (0.69–5.06) | p = 0.22 | 0.4 |
Note: HRs or sHRs, 95% CIs, and p values are shown for comparisons between donor relationship groups before and after adjustment for recipient age. Percentage change was calculated as the relative change in the point estimate after age adjustment. Large changes in HR/sHR after adjustment were interpreted as suggestive of confounding by recipient age.
Abbreviations: CI, confidence interval; DFS, disease‐free survival; GRFS, GVHD‐free/relapse‐free survival; HR, hazard ratio; OS, overall survival; sHR, subdistribution hazard ratio.
4. Discussion
Haplo‐HSCT has become an increasingly important option for patients who do not have matched donors. Recent studies show that its outcomes are now comparable to HLA‐matched transplantation [22, 23, 24, 25, 26, 27]. Optimizing donor selection remains critical to maximizing survival while minimizing complications.
Among donor‐related variables, biological relationship showed the clearest unadjusted separation across outcomes, particularly in the disease‐specific analyses. Sibling donors were associated with more favorable unadjusted OS and DFS estimates than parent or offspring donors, whereas offspring donors showed numerically poorer outcomes, especially in ALL. These observations are directionally consistent with several prior reports suggesting that donor kinship may influence transplantation outcomes in haploidentical settings [8, 12, 28]. At the same time, our data do not support donor relationship as an independently validated prognostic factor, since this association did not persist in multivariable models. The results therefore suggest that donor relationship may reflect underlying differences in patient or donor context rather than acting as a strong independent determinant on its own.
A key finding of this study was the strong imbalance in recipient age across donor groups. Patients receiving grafts from offspring donors were substantially older than those receiving grafts from siblings or parents, a pattern that is biologically plausible in routine donor selection practice. When recipient age was taken into account, the apparent disadvantage associated with offspring donors was attenuated, particularly in ALL. Although some reports have shown comparable results between different donor types [13], others have highlighted biological factors such as maternal microchimerism, fetal antigen exposure, or differences in immune tolerance, which may underlie these discrepancies [28, 29]. This finding suggests that at least part of the poorer outcome observed in the offspring‐donor group was attributable to recipient age rather than the donor relationship itself. Because older recipient age is an established adverse factor in allogeneic transplantation, this confounding effect is clinically important and underscores the need for caution when interpreting donor‐group comparisons in retrospective studies.
We did not observe an independent association between donor age and major transplantation outcomes. This finding differs from some reports that have identified older donor age as a risk factor for NRM, GVHD, or inferior survival [3, 6, 10], but it is consistent with other studies showing limited effect after adjustment for clinical covariates [4, 9, 12]. Several factors may explain this result. First, the donor population in our cohort was relatively young overall, which may have reduced the ability to detect age‐related gradients. Second, all patients were treated with the same transplantation platform, including uniform peripheral blood grafts and combined ATG/PTCy‐based GVHD prophylaxis, which may have mitigated the clinical impact of donor‐age differences.
Taken together, our findings suggest that donor age alone should not be overemphasized in donor prioritization within similar clinical settings, although it remains reasonable to consider in conjunction with other established factors. Several studies have also shown that the effect of donor age may be modified by recipient age and donor relation. Mariotti et al. [10] demonstrated a continuous adverse effect of increasing donor age on acute graft‐versus‐host disease (aGVHD) and NRM, independent of recipient age, while Ciurea et al. [30] found reduced OS with older donors (> 40 years) in patients aged ≥ 55. These findings highlight the complexity of interpreting donor age effects in clinical practice.
Similarly, we did not observe a significant effect of the donor's gender or sex mismatch on survival, consistent with some previous studies [3, 28]. Although FtoM transplants have historically been associated with higher GVHD risk [14, 15], recent data suggest that enhanced GVHD prophylaxis strategies may mitigate these risks [28]. Nevertheless, specific subgroups, such as multiparous female donors, may still confer elevated GVHD risk and warrant further investigation [7, 11]. While a study from our center previously reported higher grade III–IV aGVHD and extensive chronic graft‐versus‐host disease (cGVHD) in recipients of grafts from parous female donors, it also noted a reduced relapse incidence in this group [7]. Notably, donor parity was not captured in our dataset; therefore, parity‐related mechanisms could not be assessed in this cohort.
Our analysis confirmed that patient‐related factors, including older age and advanced disease stage at transplant, remain strong predictors of inferior survival, reinforcing the importance of timely referral and transplantation in early disease remission [31, 32]. ABO incompatibility did not significantly affect survival in our cohort, which was consistent with findings from previous studies [8]. However, major ABO mismatches are typically avoided in clinical practice due to their known risks, such as hemolysis during graft infusion, delayed engraftment, and pure red cell aplasia (PRCA) [21].
Taken together, these results emphasize the complexity of donor selection in haploidentical transplantation. While improvements in conditioning regimens and supportive care have enhanced overall outcomes, donor characteristics may contribute to outcome variability, although they were not independently predictive in this study.
5. Strengths and Limitations
The strengths of this study include a relatively homogeneous transplantation platform, the exclusive use of peripheral blood grafts, and uniform GVHD prophylaxis with a modified ATG/PTCy‐based regimen, which reduced treatment heterogeneity across donor groups. The study also reflects real‐world donor selection in a clinically relevant population of patients with acute leukemia.
Several limitations should be acknowledged. The retrospective single‐center design limits causal inference and generalizability. The offspring donor subgroup was small, particularly in ALL, resulting in wide confidence intervals and limited precision for some estimates. Importantly, lack of detailed HLA mismatch burden and absence of DSA assessment represent major limitations, as both factors are well‐established determinants of graft failure, GVHD, and survival in haploidentical HSCT [30]. In addition, the limited number of severe GVHD and NRM events precluded robust multivariable modeling for those endpoints [12].
6. Conclusion
In this single‐center retrospective cohort of adults with acute leukemia undergoing haploidentical HSCT with a uniform ATG/PTCy‐based platform, donor relationship was associated with differences in unadjusted outcomes, but it was not independently associated with survival or relapse in multivariable analysis. The poorer outcomes observed among recipients of offspring grafts appeared to be influenced in part by older recipient age, particularly in ALL. Donor age, donor sex, and ABO compatibility were likewise not independently associated with major transplantation endpoints. These findings support a balanced and individualized donor‐selection strategy in which donor relationship is interpreted alongside recipient age, remission status, donor fitness, and immunologic factors not captured in the present dataset.
Author Contributions
Mahtab Mashayekhi: data curation ‐ writing. Sahar Saeidi: data curation ‐ writing. Ghazaleh Kheiri: writing ‐ methodology. Fatemeh Salami: data curation ‐ writing. Marjan Masoud: investigation, resources. Sahar Tavakoli: writing – review and editing, methodology. Hanieh Rouzbahani: investigation, resources. Mohammad Vaezi: methodology, validation, writing – review and editing. Mohammadreza Ostadali Dehaghi: writing – review and editing, methodology. Mojtaba Azari: data curation. Ghasem Janbabai: writing – review and editing. Maryam Barkhordar: conceptualization, methodology, supervision, project administration, writing – review and editing, visualization.
AI Use Declaration
Portions of this manuscript were prepared with assistance from an artificial intelligence large language model (ChatGPT, GPT‐5, and OpenAI). The tool was used solely to enhance clarity, grammar, and structure of the text, and to help refine phrasing. No part of the research design, data analysis, interpretation of results, or generation of original scientific content was performed by the AI system. All scientific content, conclusions, and interpretations reflect the authors' own expertise and independent review. The authors carefully verified all AI‐assisted text to ensure accuracy, originality, and compliance with Wiley's ethical and editorial standards.
Funding
The authors have nothing to report.
Ethics Statement
This study was approved by the Ethics Committee of the Research Institute for Oncology, Hematology and Cell Therapy, Tehran University of Medical Sciences, Tehran, Iran [insert approval code].
Consent
All participants provided written informed consent for the use of their clinical data in accordance with institutional regulations and the Declaration of Helsinki.
Conflicts of Interest
The authors declare no conflicts of interest.
Data Responsibility Statement
All authors have read and approved the final version of the manuscript. Maryam Barkhordar had full access to all the data in this study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Transparency Statement
Maryam Barkhordar affirms that this manuscript is an honest, accurate, and transparent account of the study being reported; that no important aspects of the study have been omitted; and that any discrepancies from the study as planned have been explained.
Supporting information
Table S1: Cumulative incidence of acute (grade III–IV), Ext chronic GVHD, and NRM according to donor–recipient relationship.
Acknowledgments
The authors would like to thank the patients and their families, as well as the physicians, nurses, and staff of the Hematology, Oncology and Stem Cell Transplantation Research Center at Shariati Hospital, Tehran, Iran, for their contribution to patient care and clinical follow‐up.
Mashayekhi M., Saeidi S., Kheiri G., et al., “Donor‐Related Factors Affecting Haploidentical HSCT Outcomes in Acute Leukemia: A Single‐Center Retrospective Cohort Study of a Modified Combination of Anti‐Thymocyte Globulin and Post‐Transplant Cyclophosphamide,” Health Science Reports 9 (2026): e72631. 10.1002/hsr2.72631.
Mahtab Mashayekhi and Sahar Saeidi contributed equally to this work and share first authorship.
Data Availability Statement
The datasets generated and analyzed during the current study are available from the corresponding author on reasonable request.
References
- 1. Chang Y.‐J. and Huang X.‐J., “Is Human Leukocyte Antigen‐Matched Sibling Donor Transplant Always Better Than Haploidentical Allograft?,” in Seminars in Hematology (Elsevier, 2019). [DOI] [PubMed] [Google Scholar]
- 2. Fabricius W. A. and Ramanathan M., “Review on Haploidentical Hematopoietic Cell Transplantation in Patients With Hematologic Malignancies,” Advances in Hematology 2016 (2016): 1–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Wang Y., Chang Y. J., Xu L. P., et al., “Who Is the Best Donor for a Related HLA Haplotype‐Mismatched Transplant?,” Blood 124, no. 6 (2014): 843–850. [DOI] [PubMed] [Google Scholar]
- 4. Ciurea S. O., Cao K., Fernandez‐Vina M., et al., “The European Society for Blood and Marrow Transplantation (EBMT) Consensus Guidelines for the Detection and Treatment of Donor‐Specific Anti‐HLA Antibodies (DSA) in Haploidentical Hematopoietic Cell Transplantation,” Bone Marrow Transplantation 53, no. 5 (2018): 521–534. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Ye Y., Labopin M., Chen J., et al., “Similar Outcomes Following Non‐First‐Degree and First‐Degree Related Donor Haploidentical Hematopoietic Cell Transplantation for Acute Leukemia Patients in Complete Remission: A Study From the Global Committee and the Acute Leukemia Working Party of the European Society for Blood and Marrow Transplantation,” Journal of Hematology & Oncology 16, no. 1 (2023): 1–5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Pruitt A., Gao F., De Togni E., et al., “Impact of Donor Age and Relationship on Outcomes of Peripheral Blood Haploidentical Hematopoietic Cell Transplantation,” Bone Marrow Transplantation 58, no. 8 (2023): 855–862. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Azari M., Barkhordar M., Bahri T., et al., “Determining the Predictive Impact of Donor Parity on the Outcomes of Human Leukocyte Antigen Matched Hematopoietic Stem Cell Transplants: A Retrospective, Single‐Center Study,” Frontiers in Oncology 14 (2024): 1339605. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. DeZern A. E., Franklin C., Tsai H. L., et al., “Relationship of Donor Age and Relationship to Outcomes of Haploidentical Transplantation With Posttransplant Cyclophosphamide,” Blood Advances 5, no. 5 (2021): 1360–1368. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Mehta R. S., Marin D., Alousi A., et al., “Haploidentical vs Matched Unrelated Donors for Patients With All: Donor Age Matters More Than Donor Type,” Blood Advances 7, no. 8 (2023): 1594–1603. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Mariotti J., Raiola A. M., Evangelista A., et al., “Impact of Donor Age and Kinship on Clinical Outcomes After T‐Cell‐Replete Haploidentical Transplantation With PT‐Cy,” Blood Advances 4, no. 16 (2020): 3900–3912. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Loren A. W., Bunin G. R., Boudreau C., et al., “Impact of Donor and Recipient Sex and Parity on Outcomes of HLA‐Identical Sibling Allogeneic Hematopoietic Stem Cell Transplantation,” Biology of Blood and Marrow Transplantation 12, no. 7 (2006): 758–769. [DOI] [PubMed] [Google Scholar]
- 12. Solomon S. R., Aubrey M. T., Zhang X., et al., “Selecting the Best Donor for Haploidentical Transplant: Impact of HLA, Killer Cell Immunoglobulin‐Like Receptor Genotyping, and Other Clinical Variables,” Biology of Blood and Marrow Transplantation 24, no. 4 (2018): 789–798. [DOI] [PubMed] [Google Scholar]
- 13. Danylesko I., Peczynski C., Labopin M., et al., “Impact of Donor Kinship on Non‐T‐Cell Depleted Haploidentical Stem Cell Transplantation With Post Transplantation Cyclophosphamide for Acute Leukemia: From the ALWP of the EBMT,” Bone Marrow Transplantation 57, no. 8 (2022): 1260–1268. [DOI] [PubMed] [Google Scholar]
- 14. Randolph S. S. B., Gooley T. A., Warren E. H., Appelbaum F. R., and Riddell S. R., “Female Donors Contribute to a Selective Graft‐Versus‐Leukemia Effect in Male Recipients of HLA‐Matched, Related Hematopoietic Stem Cell Transplants,” Blood 103, no. 1 (2004): 347–352. [DOI] [PubMed] [Google Scholar]
- 15. du Toit J. R., Mcdonald A., Brittain D., et al., “Is Haploidentical Hematopoietic Cell Transplantation Using Post‐Transplantation Cyclophosphamide Feasible in Sub‐Saharan Africa?,” Transplantation and Cellular Therapy 27, no. 12 (2021): 1002.e1–1002.e8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Wu Y., Lai X., Shi J., et al., “Effect of Donor Characteristics on T Cell‐Replete Haploidentical Stem Cell Transplantation Over the Last 10 Years at a Single Institution,” British Journal of Haematology 196, no. 5 (2022): 1225–1238. [DOI] [PubMed] [Google Scholar]
- 17. Kim H. T., Zhang M. J., Woolfrey A. E., et al., “Donor and Recipient Sex in Allogeneic Stem Cell Transplantation: What Really Matters,” Haematologica 101, no. 10 (2016): 1260–1266. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Kumar A. J., Kim S., Hemmer M. T., et al., “Graft‐Versus‐Host Disease in Recipients of Male Unrelated Donor Compared With Parous Female Sibling Donor Transplants,” Blood Advances 2, no. 9 (2018): 1022–1031. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Jagasia M. H., Greinix H. T., Arora M., et al., “National Institutes of Health Consensus Development Project on Criteria for Clinical Trials in Chronic Graft‐Versus‐Host Disease: I. The 2014 Diagnosis and Staging Working Group Report,” Biology of Blood and Marrow Transplantation 21, no. 3 (2015): 389–401.e1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Przepiorka D., Weisdorf D., Martin P., et al., “1994 Consensus Conference on Acute GVHD Grading,” Bone Marrow Transplantation, 1995 15, no. 6 (1995): 825–828. [PubMed] [Google Scholar]
- 21. Booth G. S., Gehrie E. A., Bolan C. D., and Savani B. N., “Clinical Guide to ABO‐Incompatible Allogeneic Stem Cell Transplantation,” Biology of Blood and Marrow Transplantation 19, no. 8 (2013): 1152–1158. [DOI] [PubMed] [Google Scholar]
- 22. D'Souza A., Fretham C., Lee S. J., et al., “Current Use of and Trends in Hematopoietic Cell Transplantation in the United States,” Biology of Blood and Marrow Transplantation 26, no. 8 (2020): e177–e182. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Haebe S., Fraccaroli A., Stauffer E., et al., “PTCY‐Based Haploidentical Donor Transplantation Versus HLA‐Matched Related and Unrelated Donor Transplantations in Patients With Refractory or Relapsed Lymphoma‐A Matched‐Pair Analysis,” Cancers 15, no. 21 (2023): 5246. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Passweg J. R., Baldomero H., Bader P., et al., “Use of Haploidentical Stem Cell Transplantation Continues to Increase: The 2015 European Society for Blood and Marrow Transplant Activity Survey Report,” Bone Marrow Transplantation 52, no. 6 (2017): 811–817. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Barkhordar M., Kasaeian A., Janbabai G., et al., “Modified Combination of Anti‐Thymocyte Globulin (ATG) and Post‐Transplant Cyclophosphamide (PTCy) as Compared With Standard ATG Protocol in Haploidentical Peripheral Blood Stem Cell Transplantation for Acute Leukemia,” Frontiers in Immunology 13 (2022): 921293. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Barkhordar M., Kasaeian A., Janbabai G., et al., “Outcomes of Haploidentical Peripheral Stem Cell Transplantation With Combination of Post‐Transplant Cyclophosphamide (PTCy) and Anti‐Thymocyte Globulin (ATG) Compared to Unrelated Donor Transplantation in Acute Myeloid Leukemia: A Retrospective 10‐Year Experience,” Leukemia Research 120 (2022): 106918. [DOI] [PubMed] [Google Scholar]
- 27. Barkhordar M., Kasaeian A., Tavakoli S., et al., “Selection of Suitable Alternative Donor in the Absence of Matched Sibling Donor: A Retrospective Single‐Center Study to Compare Between Haploidentical, 10/10 and 9/10 Unrelated Donor Transplantation,” International Journal of Hematology‐Oncology and Stem Cell Research 15, no. 1 (2021): 51–60. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28. McCurdy S. R., Zhang M. J., St. Martin A., et al., “Effect of Donor Characteristics on Haploidentical Transplantation With Posttransplantation Cyclophosphamide,” Blood Advances 2, no. 3 (2018): 299–307. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. de Lima V. J. T., da Silva A. F., Kerbauy L. N., et al., “Risk Factors for Adverse Outcomes Following Haploidentical Hematopoietic Cell Transplantation With Posttransplant Cyclophosphamide: A Two‐Center Analysis,” Annals of Hematology 101, no. 8 (2022): 1795–1802. [DOI] [PubMed] [Google Scholar]
- 30. Ciurea S. O., Al Malki M. M., Kongtim P., et al., “The European Society for Blood and Marrow Transplantation (EBMT) Consensus Recommendations for Donor Selection in Haploidentical Hematopoietic Cell Transplantation,” Bone Marrow Transplantation 55, no. 1 (2020): 12–24. [DOI] [PubMed] [Google Scholar]
- 31. Aoki J., Kanamori H., Tanaka M., et al., “Impact of Age on Outcomes of Allogeneic Hematopoietic Stem Cell Transplantation With Reduced Intensity Conditioning in Elderly Patients With Acute Myeloid Leukemia,” American Journal of Hematology 91, no. 3 (2016): 302–307. [DOI] [PubMed] [Google Scholar]
- 32. Arab Borzu Z., Khadembashi N., Hajifathali A., Baghestani A., and Roshandel E.. “Survival Rate and Prognostic Factors Among Patients Undergoing Hematopoietic Stem Cell Transplantation: Using the Joint Model,” International Journal of Cancer Management 14, no. 3 (2021): e106846, 10.5812/ijcm.106846. [DOI] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Table S1: Cumulative incidence of acute (grade III–IV), Ext chronic GVHD, and NRM according to donor–recipient relationship.
Data Availability Statement
The datasets generated and analyzed during the current study are available from the corresponding author on reasonable request.