Outcomes following allo-HSCT in patients with BP-MPN

Blast phase (BP) transformation of myeloproliferative neoplasms (BP-MPN) is characterized by ≥20% blasts in the peripheral blood and bone marrow and generally confers an adverse prognosis with a median overall survival (OS) of 3–5 months.¹ While intensive chemotherapy is the current standard of care, many patients with BP-MPN are unfit for such treatment.¹

The MPN Hub previously reported on outcomes in patients with BP-MPN treated with venetoclax combinations. Despite initial reposes following chemotherapy, allogeneic hematopoietic stem cell transplantation (allo-HSCT) is necessary to translate these responses into long-term survival benefits and remains the only potentially curative option for patients with BP-MPN.¹ Donor availability is a limiting factor of allo-HSCT and further analysis is required to determine the optimal patient selection strategy.¹

Ortí et al.¹ recently published an article in the American Journal of Hematology on a retrospective, multicenter analysis of patients with BP-MPN who underwent allo-HSCT, and we are pleased to summarize the key findings below.

Study design and patient characteristics

A retrospective analysis was performed with patient data from 169 institutions using the European Society for Blood and Marrow Transplantation (EMBT) registry. The analysis included 663 patients aged ≥18 years with BP-MPN treated with allo-HSCT between 2005 and 2019.

The primary endpoint was OS, with progression-free survival (PFS), non-relapse mortality (NRM), cumulative incidence of relapse (CIR), and cumulative incidences of graft-versus-host-disease (GvHD) and GvHD relapse-free survival after allo-HSCT as secondary endpoints. The median age was 60 years with a median follow-up of 5.2 years.

Results

Engraftment response

Graft failure was reported in 5.5% of patients. Response data 100 days post allo-HSCT were available for 528 patients, with:

• 76% of patients in complete remission (CR);
• 9% of patients relapsed or had progressive disease; and
• 16% who were never in CR.

The 3- and 5-year OS was 36% and 32%, respectively, with a median OS of 13.9 months. The 5-year OS probability was significantly impacted by Karnofsky prognostic score, active disease, and the year of allo-HSCT, while the type of donor trended towards significance (Table 1).

Table 1. 5-year OS probability*

Allo-HSCT, allogeneic hematopoietic stem cell transplantation; CR, complete remission; MMRD, mismatch related donor; MMUD, mismatch unrelated donor; MSD, matched sibling donor; MUD, matched unrelated donor. *Adapted from Ortí, et al.1
Characteristic, %	5-year OS probability	p value
Karnofsky prognostic score
≥90 (n = 367)	36	<0.001
<90 (n = 240)	26
Year of allo-HSCT
<2010 (n = 132)	23	0.02
2010–2014 (n = 292)	34
2015–2019 (n = 239)	34
Donor type
MSD (n = 188)	32	0.06
MMRD (n = 58)	25
MUD (n = 231)	37
MMUD (n = 121)	24
Disease stage at allo-HSCT
CR (n = 291)	37	<0.001
Active disease (n = 360)	27

Cause of death

In total, 436 patients died, with cause of death data available for 421 patients. The leading causes of death included:

• Disease relapse or progression (45%)
• Infection (30%)
• GvHD (14%)
• Organ damage or organ failure (5%)
• Other transplant-related causes (3%)
• Other causes (2%)
• Secondary malignancies (1%)

Relapse and progression were the most common causes of death among patients who received allo-HSCT from mismatch sibling donor (MSD), mismatch unrelated donor (MMUD), matched unrelated donor (MUD), and matched unrelated donor, while infection was more common in patients who received allo-HSCT from mismatch related donor (MMRD).

Non-relapse mortality

Overall, the estimated 3- and 5-year NRM was 24% and 24%, respectively. Subgroup analysis revealed estimated 5-year NRM rates were higher in patients:

• Who received allo-HSCT from a MMRD (41%) versus a MSD (17%), MUD (22%), and MMUD (28%; p < 0.001)
• Who received post-transplant cyclophosphamide (38%) versus post-transplant calcineurin inhibitors (22%) or any other form of GvHD prophylaxis (31%; p = 0.01)
• With a Karnofsky prognostic score <90 (33%) versus ≥90 (19%; p < 0.001)
• Undergoing allo-HSCT with active disease (29%) versus patients in CR (19%; p < 0.001)
• Who received a graft from other graft sources (34%) versus peripheral blood (23%; p = 0.04)

Cumulative incidences of relapse

The 3- and 5-year CIR was 48% and 51% respectively. The 5-year CIR was higher in patients receiving post-transplant calcineurin inhibitors (53%) compared with patients receiving post-transplant cyclophosphamide (38%; p = 0.04).

Progression-free survival

The estimated 3- and 5-year PFS was 28% and 25%, respectively, with a median PFS of 7.8 months. The 5-year PFS was higher in patients:

• Who received allo-HSCT between 2015–2019 (29%) and 2010–2014 (27%) versus 2005–2010 (16%; p = 0.01)
• With a Karnofsky prognostic score ≥90 (27%) versus <90 (22%; p = 0.01)
• Who received allo-HSCT in CR (29%) versus those who received allo-HSCT with active disease (22%; p < 0.001)

Multivariate analysis

Multivariate analysis using multivariable Cox proportional hazard models determined the following:

• Karnofsky performance score <90 (p < 0.0001), year of allo-HSCT (p = 0.004), and active disease at transplant (p = 0.0007) influenced OS
• Karnofsky performance score <90 (p = 0.003), year of allo-HSCT (p = 0.006), and active disease at transplant (p = 0.0005) influenced PFS
• Active disease at transplant (p = 0.02) influenced relapse rates
• Karnofsky performance score <90 (p = 0.0002), overall donor type (p = 0.01), MMRD (p = 0.005), MMUD (p = 0.05), and active disease at transplant (p = 0.004) influenced NRM
• Age at allo-HSCT (per 10-year increase), sex, interval diagnosis-acute myeloid leukemia, and stem cell source were not significant factors in terms of OS, PFS, relapse, and NRM

GvHD

The 100-day cumulative incidences of GvHD Grade II–IV and III–IV acute GvHD were 29% and 13%, respectively. The 100-day cumulative incidences of acute GvHD were higher in patients who received total body irradiation-based conditioning (38%) than in patients who received non-total body irradiation-based conditioning (27%; p = 0.02).

Post-transplant treatment

Post allo-HSCT data were available for 285 patients, of which 62% received disease-related treatment and 20% received a donor lymphocyte infusion.

Conclusion

This large, retrospective analysis showed that survival outcomes are poor in patients with BP-MPN who underwent allo-HSCT. However, outcomes have improved in recent years, with significantly improved outcomes from 2010 onward. Patients with a Karnofsky prognostic score ≥90 and who underwent allo-HSCT in remission also experienced improved survival outcomes. Prospective studies are warranted and may improve patient selection for allo-HSCT.