Comparison of posttransplant outcomes between haploidentical and matched unrelated donor recipients receiving PTCy prophylaxis

Triple posttransplant cyclophosphamide (PTCy)-based graft-versus-host disease (GvHD) prophylaxis has improved survival rates in patients receiving haploidentical (haplo) human leukocyte antigen (HLA) hematopoietic stem cell transplantation (HSCT), likely due to tolerogenic effects on donor T cells. Incidences of acute and chronic GvHD are also reduced to a level comparable to transplantation with matched unrelated donors (MUD) following traditional prophylaxis. Furthermore, the use of PTCy has since been expanded to include MUD recipients following clinical trials demonstrating superiority over traditional prophylaxis.¹ However, it is currently uncertain whether conditioning regimens and posttransplant prophylaxis methods are contributory factors in these results.

Here, we present results from a recent observational study, published by Gooptu et al.¹ in Blood, which compared posttransplant outcomes following MUD or haplo transplantation with PTCy GvHD prophylaxis. The study separately analyzed patients by conditioning regimen (reduced intensity conditioning [RIC] and myeloablative conditioning [MAC] regimens).

Study design

A controlled observational study using registry data from patients transplanted between 2011 and 2018 at 111 centers in the US, contributing to the Center for International Blood and Marrow Transplant Research (CIBMTR).

Inclusion criteria

• Age ≥18
• Diagnosis of acute myeloid leukemia (AML), acute lymphoblastic leukemia (ALL) in first or second complete remission, or myelodysplastic syndromes (MDS)
• Haplo donors: mismatched at ≥2 HLA loci
• MUD donors: matched at the allele level at HLA-A, -B, -C, and -DRB1

Patients who received PTCy/calcineurin inhibitor prophylaxis without mycophenolate mofetil, were transplanted in the third complete remission or relapse, had AML transformed from MDS, or received CD341-selected peripheral blood grafts following in vivo T-cell depletion were excluded.

Endpoints

• The primary endpoint was overall survival.
• Secondary endpoints included hematopoietic recovery, acute and chronic GvHD, relapse, non-relapse mortality, and disease-free survival.

Results

Patient characteristics

Patient characteristics by conditioning regimen are summarized in Table 1.

Table 1. Patient characteristics*

ALL, acute lymphoblastic leukemia; AML, acute myeloid leukemia; CMV, cytomegalovirus; haplo, haploidentical; MDS, myelodysplastic syndromes; MUD, matched unrelated donor; TBI, total body irradiation. *Data from Gooptu, et al.1
Characteristic	Reduced-intensity conditioning		Myeloablative conditioning
	Haplo donor (n = 1,211)	MUD (n = 187)	Haplo donor (n = 825)	MUD (n = 97)
Median age, years (range)	62 (18–81)	65 (20–80)	45 (18–75)	50 (18–71)
White, %	72	95	69	88
Female, %	41	44	44	47
Performance score
90–100	53	59	56	59
≤80	45	39	42	38
Not reported	2	2	2	3
Comorbidity score ≥3, %	50	54	45	48
Positive CMV serostatus, %	69	63	70	58
Disease
AML	60	60	55	48
ALL	18	16	32	28
MDS	23	24	13	24
Disease risk
Low/intermediate	79	80	76	79
High/very high	18	19	20	20
Not reported	3	1	4	1
TBI conditioning regimen, %	94	47	50	34
Graft type
Bone marrow	44	16	26	12
Peripheral blood	56	84	74	88
Transplant period
2011–2014	21	13	19	7
2015–2018	79	87	81	93

Reduced-intensity conditioning regimen

For patients who received RIC, hematopoietic recovery, including both neutrophil and platelet recovery rates, was lower for haplo HSCT vs MUD HSCT. The incidence of graft failure at 2 years was higher for haplo HSCT compared with MUD HSCT (11% vs 3%; p < 0.001). Multivariate analysis revealed that the incidence of Grade 2–4 acute GvHD and non-relapse mortality risk were both lower in patients receiving MUD HSCT (Table 2). As such, disease-free and overall survival rates were higher for this donor type. The incidence of chronic GvHD and relapse risk did not differ significantly between donor types.

Table 2. Multivariate analysis for posttransplant outcomes in patients receiving reduced-intensity conditioning*

CI, confidence interval; GvHD, graft-versus-host disease; haplo, haploidentical; HR, hazard ratio; MUD, matched unrelated donor. *Data from Gooptu, et al.1
Outcome	Number of events/evaluable	HR (95% CI)	p value
Grade 2–4 acute GvHD
Haplo	389/1,171	1.00
MUD	49/180	0.70 (0.52–0.95)	0.022
Grade 3–4 acute GvHD
Haplo	118/1,167	1.00
MUD	8/179	0.41 (0.20–0.85)	0.016
Non-relapse morality
Haplo	205/1,205	1.00
MUD	18/187	0.33 (0.19–0.57)	<0.0001
Disease-free survival
Haplo	171/1,205	1.00
MUD	90/187	0.74 (0.60–0.93)	0.008
Overall survival
Haplo	568/1,211	1.00
MUD	65/187	0.65 (0.50–0.84)	0.001

Furthermore:

• Recipients of MUD HSCT who received peripheral blood grafts had a higher incidence of Grade 2–4 chronic GvHD (p = 0.07), as well as higher non-relapse mortality, independent of donor type.
• In centers that performed both types of donor transplants, survival was also higher after MUD HSCT (HR, 0.75; 95% CI, 0.57–0.98; p = 0.03) compared with haplo HSCT.
• In the 25% of the cohort with available infection data, a higher 6-month incidence of fungal infections was observed in haplo HSCT recipients compared with those receiving MUD HSCT (10% vs 1%, respectively; 95% CI, 0–5; p < 0.001), with no significant difference between bacterial or viral causes.
• There was no significant difference between the proportion of patients who received donor leukocyte infusions or underwent a second HSCT: haplo HSCT, 13% (159/1211); MUD HSCT, 9% (16/187).

Finally, when comparing mortality, 47% (568/1,211) of patients receiving haplo HSCT died compared with 35% (65/187) of patients receiving MUD HSCT (Table 3). Disease reoccurrence was the most common cause, although this accounted for fewer deaths after haplo HSCT compared with MUD HSCT (55% vs 71%, respectively; p = 0.02).

Table 3. Causes of deaths by donor type for patients who received reduced-intensity conditioning*

GvHD, graft-versus-host disease; haplo, haploidentical; MUD, matched unrelated donor. *Data from Gooptu, et al.1
Cause of death, %	Haplo (n = 568)	MUD (n = 65)
Disease reoccurrence	55	71
Graft failure	2	—
GvHD	7	2
Infections	15	6
Interstitial pneumonitis	3	2
Organ failure	8	14
Malignancy excluding primary diagnosis	2	0
Other causes	3	3

Myeloablative conditioning

When comparing the same outcomes for MAC, neutrophil recovery rates did not differ by donor type; however, platelet recovery rates were significantly lower following haplo HSCT compared with MUD (87% vs 93%, p < 0.001). The incidence of graft failure after 1 year did not differ by donor type. Unlike RIC, the incidence of Grade 2–4 acute GvHD was similar between donor types; however, incidences of Grade 3–4 acute GvHD (HR, 0.37; 95% CI, 0.14–1.00; p = 0.05) and chronic GvHD were lower for MUD HSCT (HR, 0.66; 95% CI, 0.43–1.01; p = 0.053). No differences in relapse risk, non-relapse mortality, or disease-free and overall survival were observed.

Additionally:

• Peripheral blood grafts were associated with increased risk of chronic GvHD (p = 0.22) in both haplo and MUD HSCT.
• No significant differences in survival after MUD HSCT vs haplo HSCT were seen in centers undertaking both (HR, 0.82; 95% CI, 0.50–1.33; p = 0.41).
• The 6-month incidence of bacterial infections was higher in patients receiving haplo HSCT compared with MUD HSCT (56% vs 27%; p = 0.003); however, incidences of fungal and viral infections were similar.
• A total of 10% (82/825) of haplo HSCT recipients received donor leukocyte infusions or a second HSCT, comparable to 8% (6/79) of MUD HSCT recipients.

Overall, 33% (275/825) of haplo HSCT recipients died compared with 27% (48/176) of MUD HSCT recipients. No significant differences in the causes of death were observed between donor types, with recurrent disease the most common cause, as for RIC (Table 4).

Table 4. Causes of death by donor type for patients who received myeloablative conditioning*

GvHD, graft-versus-host disease; haplo, haploidentical; MUD, matched unrelated donor. *Data from Gooptu, et al.1
Cause of death, %	Haplo (n = 275)	MUD (n = 48)
Disease reoccurrence	50	47
Graft failure	3	2
GvHD	9	10
Infections	12	10
Interstitial pneumonitis	7	2
Organ failure	11	13
Other causes	7	8

Subset analysis

The majority of MUD HSCTs were carried out between 2016 and 2018, with survival outcomes consistent with the overall time period for both RIC and MAC regimens. Within the RIC group, recipients of MUD HSCT had a lower risk of non-relapse mortality (HR, 0.34; 95% CI, 0.15–0.73, p = 0.006) and, by extension, higher disease-free (HR, 0.69; 95% CI, 0.50–0.94; p = 0.018) and overall survival (HR, 0.57; 95% CI, 0.39–0.84, p = 0.004) rates than haplo HSCT recipients. However, within the MAC arm, non-relapse mortality risk, disease-free surivial, and overall survival were comparable between donor types. The incidence of Grade 3–4 acute GvHD was lower for MUD HSCT compared with haplo HSCT in both the RIC (HR, 0.38; 95% CI, 0.15–0.97, p = 0.04) and MAC (HR, 0.33; 95% CI, 0.10–1.07, p = 0.06) setting.

Future considerations

In a recent journal club held at the International Academy for Clinical Hematology, Mary Eapen and Arnon Nagler discussed the impact of these registry data.² The key point was the need for randomized studies comparing donor types, to eliminate registry study bias and variance in transplant numbers that occur over a large data collection period. Additionally, Eapen highlighted a lack of data on the long-term effects of PTCy-based prophylaxis in this study, which included only a 2-year follow-up. The impact of socioeconomic status on access to MUD HSCT was considered, given the differences in outcomes depending on donor type. Notably, there is a higher incidence of GvHD in African Americans, along with a higher mortality rate associated with transplantation. As socioeconomic status is known to be a barrier to transplantation, it is therefore important to reduce the disparity in access to MUD HSCT and expand donor pools, especially for minority groups.

Conclusion

Registry data from this study support the preferred use of MUD HSCT, particularly in patients who received RIC. In these patients, a higher incidence of graft failure and acute GvHD was noted, which translated into lower disease-free and overall survival following haplo HSCT. In patients treated with MAC, while posttransplant survival differences were reduced with PTCy, there remained a higher incidence of Grade 3–4 acute GvHD. As access to HSCT, independent of donor type, provides survival benefits for patients with hematologic malignances, haplo HSCT should still be offered if a MUD donor is not available.