The impact of donor type on long-term survival outcomes of patients receiving allo-HSCT

Allogeneic hematopoietic stem cell transplantation (allo-HSCT) is an effective therapeutic option for patients with acute myeloid leukemia (AML), and more donor types have become available, including human leukocyte antigen (HLA) 10/10 matched unrelated donor (UD), HLA 9/10 mismatch UD, or cord blood transplant (CBT). In the short term, slow engraftment can create complications for patients receiving CBT; however, this type of donor has been proven to reduce chronic graft-versus-host disease (cGvHD) and improve graft-versus-leukemia (GvL) activity. As new techniques have been developed to improve engraftment in CBT, it is beneficial to identify if there are significant differences in survival between donor types in the long term.¹

Frédéric Baron and colleagues¹ recently published their landmark study in Bone Marrow Transplantation. They compared long-term survival outcomes in patients with AML who were alive 2-years following allo-HSCT, dependent on HLA matched 10/10 UD, HLA, 9/10 mismatch UD, or CBT donor types.¹ We summarize the key results below.

Study design

It was a retrospective study from the Acute Leukaemia Working Party (ALWP) of the European Society for Blood and Marrow Transplantation (EBMT), which includes 600 transplant centers.

Inclusion criteria

• Patients ≥18 years old
• De novo or secondary AML
• First or second allo-HSCT between 2005–2015
• In first complete remission (CR1) or second complete remission (CR2)
• Either received CBT (single or double) or peripheral blood stem cells (PBSC) from HLA-matched UD 10/10 or one locus HLA-mismatched UD 9/10
• No in vitro T-cell depletion of the graft
• Alive without relapse and without needing second transplant at the landmark time of 2 years after transplantation

Primary and secondary endpoint

The primary endpoint was leukemia-free survival (LFS)

The secondary endpoints were relapse incidence, nonrelapse mortality (NRM), and overall survival (OS).

Results

A total of 3,693 patients met the study criteria at the landmark time of 2 years following transplant. Patient characteristics for the overall cohort are summarised in Table 1.

Table 1. Patient characteristics*

AML, acute myeloid leukemia; auto-HSCT, autologous hematopoietic stem cell transplant; CBT, cord blood transplantation; CMV, cytomegalovirus; CR1, first complete remission; CR2, second complete remission; TBI, total body irradiation; UD, unrelated donor. *Data from Baron et al.1
Characteristic	CBT (n = 364)	UD 10/10 (n = 2,648)	UD 9/10 (n = 681)
Age at transplant, median	46	54	51
Patient, Female, %	52.8	49.6	46.6
Donor, Female, %	52.3	27.9	35.7
CMV patient, negative, %	35.9	37.6	37.6
CMV donor, negative, %	63.2	58.4	55.8
Disease status, %
CR1	62.9	79.3	71.1
CR2	37.1	20.7	28.9
Secondary AML, %	11.0	16.1	15.1
Cytogenetic risk, %
Good	9.1	6.9	6.8
Intermediate	44.5	35.3	37.4
Poor	9.6	8.0	8.7
TBI, yes, %	69.8	26.6	23.5
Previous auto-HSCT, yes, %	5.8	2.4	6.0

Primary endpoint

The 5-year LFS was 85.7% (95% CI, 81.4–89.0) in CBT patients, versus 84.0% (95% CI, 82.5–85.5) in UD 10/10 patients (p = 0.36) and 84.0% (95% CI, 80.7–86.7) in UD 9/10 patients (p = 0.86).

On the multivariate analysis, donor type had no significant effect on LFS. Factors that were significantly associated with LFS included age at transplant, sex whether the patient was in CR2 versus CR1, having secondary AML, antecedent of cGvHD, and poor-risk cytogenetics (Table 2).

Table 2. Factors significantly associated with LFS on the multivariate analysis*

AML, acute myeloid leukemia; cGvHD, chronic graft-versus-host disease; CR1, first complete remission; CR2, second complete remission; HSCT, hematopoietic stem cell transplant; LFS, leukemia-free survival. *Data from Baron et al.1
Factor	HR	95% CI	p value
Age at HSCT	1.1	1.1–1.1	<0.001
Sex, female vs male	0.74	0.63–0.87	<0.001
Disease status, CR2 vs CR1	1.2	1–1.5	0.05
cGvHD before 2 years	1.5	1.3–1.8	<0.001
Poor cytogenetic risk	1.8	1.2–2.8	0.01
Secondary AML	1.3	1.1–1.6	0.01

Secondary endpoints

The 5-year relapse incidence from transplant (3 years after the start of landmark analysis) was 9.8% in patients with CBT donors (95% CI, 6.9–13.2), versus 9.3% (95% CI, 8.2–10.6) in UD 10/10 patients (p = 0.85), and finally 11.3% (95% CI, 8.9–14.1) in UD 9/10 patients, (p = 0.93). In a multivariate analysis, there was no significant difference in the relapse incidence between donor types. Greater age at transplantation, in vivo T-cell depletion, disease status (CR2 vs CR1), secondary AML, and poor-risk cytogenetics, were significantly associated with a higher 5-year relapse incidence, and female gender was associated with risk of relapse (Table 3).

At 5 years, the rate of nonrelapse mortality was 4.5% in CBT patients (95% CI, 2.6–7.1), versus 6.6% in UD 10/10 patients (95% CI, 5.6–7.7; p = 0.12), and finally, 4.7% in UD 9/10 patients (95% CI, 3.2–6.6; p = 0.91). As only 15 incidents of NRM were reported for CBT, a multivariate analysis was not performed.

No significant difference was found between donor types for OS at 5 years, which was 88.6% for CBT patients (95% CI, 84.5–91.6), versus 87.9% (95% CI, 86.5–89.2) in UD 10/10 patients (p = 0.58), and 89.9% (95% CI, 87.1–92.1) in UD 9/10 patients (p = 0.58).

Greater age at transplantation, secondary AML, antecedent cGvHD, and poor-risk cytogenetics were significantly associated with worse OS. Conversely, female gender, in vivo T-cell depletion, and myeloablative conditioning were associated with better OS (Table 3).

Table 3. Factors significantly associated with relapse incidence and/or OS on multivariate analysis*

AML, acute myeloid leukemia; CI, confidence interval; cGvHD, chronic graft-versus-host disease; CR1, first complete remission; CR2, second complete remission; HCT, hematopoietic cell transplant; HR, hazard ratio; OS, overall survival. *Data from Baron et al.1
Factor	Relapse incidence			OS
	HR	95% CI	p value	HR	95% CI	p value
Age at HCT	1.1	1.1–1.1	0.01	1.1	1.1–1.2	<0.001
Sex, female vs male	0.64	0.51–0.79	<0.001	0.8	0.67–0.96	0.02
In vivo T-cell depletion	1.4	1.1–1.8	0.02	0.8	0.66–0.98	0.03
Secondary AML	1.4	1.1–1.9	0.01	1.3	1.1–1.6	0.04
Disease status, CR2 vs CR1	1.4	1.1–1.8	0.01	1.2	0.98–1.5	0.08
cGvHD before 2 years	1	0.81–1.2	0.96	1.8	1.5–2.1	<0.001
Myeloablative conditioning	1	0.82–1.3	0.78	0.79	0.64–0.97	0.02
Poor cytogenetic risk	2.8	1.6–5.1	<0.001	2	1.2 – 3.3	0.005

Mortality

Original disease was the main cause of death in 44%, 37%, and 40% of CBT, UD 10/10, and UD 9/10 recipients, respectively. The causes of death were similar between the donor types, except for death related to GvHD which was lower in CBT recipients. Cumulative incidence of death relating to GvHD at 5 years was 0.28% in CBT, 2.66% in UD 10/10, and 2.7% in UD 9/10 (p = 0.004 and p = 0.007, respectively).

Conclusion

This study provided key insights into the effect of donor type on long-term survival of allo-HSCT recipients. CBT was reported to have no effect on the 5-year relapse incidence, contrasting to the investigator’s hypothesis of reduced incidence due to improved GvL activity with this donor type. As a result, there was no significant difference in long-term LFS, OS, and NRM between donor type.