Haploidentical and Mismatched Unrelated Donor Allogeneic Hematopoietic Cell Transplantation in Patients with Myelofibrosis: A Retrospective Cohort Study

Background:

Allogeneic hematopoietic stem cell transplantation (allo-HCT) using Haploidentical (HID) and mismatched unrelated donors (MMUD) are viable options for hematological malignancies but data in patients with myelofibrosis (MF) are limited. This study examines the outcomes of MF patients receiving MMUD or HID transplants.

Methods:

We conducted a retrospective study of patients with MF undergoing first allo-HCT from MMUD or HID donors between January 2015 and January 2023. Disease, treatment, and transplant characteristics were collected. Outcomes included Overall Survival (OS), Non-relapse Mortality (NRM), graft versus host disease (GVHD), relapse, GVHD-free/relapse-free survival (GRFS), engraftment kinetics, and graft failure.

OS and GRFS were estimated using the Kaplan-Meier method. Cumulative incidences of relapse (CIR), NRM, and acute and chronic GVHD were analyzed using competing risk models. Cox regression was used for OS & GRFS, and Fine-Gray regression for NRM, CIR, and graft failure. A p-value of <0.05 was considered statistically significant and analyses were performed using EZR (version 1.62).

Results:

Thirty-four patients with MF received MMUD or HID allo-HCT. The majority (n=21, 61.8%) had primary MF and were in the chronic phase (n=31, 91.2%). The median age at allo-HCT was 60 years (27 – 73), and 20 (59%) were female. The DIPSS+ score (n=30) was intermediate 2 – high in 28 (93.3%) patients. Twenty-one patients (61.8%) received JAK inhibitors before allo-HCT, the median interval from diagnosis to allo-HCT was 15.8 months (5 – 21). The median spleen size at allo-HCT was 19 cm (13 – 35), and 13 (38.2%) had a size >22 cm. Four (12%) had a splenectomy before transplant.

Most (70.6%) received reduced-intensity conditioning. Twenty-six (76%) received total body irradiation [median, 200cGy (200-400)]. All patients received peripheral blood stem cells with a median CD34+ cell dose of 8.3 x10⁶/kg (3.4 – 11.7). Eleven (32.3%) received cryopreserved grafts. Anti-thymocyte globulin (ATG), post-transplantation cyclophosphamide (PTCy), and cyclosporine was used as GVHD prophylaxis in 27 patients (79%).

Twenty-six (76.4%) patients engrafted. The median time for neutrophil and platelet engraftment was 20 days (14–29) and 29 days (12–152), respectively. The D+100 cumulative incidence of platelet count >50 x10⁹/L was 85.4% (52.5 – 97). The cumulative incidence of primary graft failure (PGF) at D+100 was 23.5% (10.9 - 38.8) and secondary graft failure at D+365 was 6.3% (1.1-18.6).Patients receiving cryopreserved grafts had a D+100 cumulative incidence of PGF of 63.6% (26.6–85.7) vs 4.3% (3.1–18.7) for fresh grafts (p<0.001). On MVA, the use of cryopreserved grafts was associated with an increased risk of PGF [HR 22.8 (3.3 – 91) p<0.001].

At D+100, the cumulative incidence of all grades aGVHD was 29.4% (15 – 45%), and 6.6% (1.1 – 19.3) for grade III-IV aGVHD. Chronic GVHD (cGVHD) at 24 months was 42.6% (23–61) for all grades and 28.3% (13.8–44.8) for moderate-severe cGVHD.

Ten (29.4%) patients relapsed, with a median time to relapse of 20.9 months (2 – 67.2). The CIR at 12 and 24 months was 6.3% (1.1–18.5) and 21.6% (8–38), respectively. The 12 and 24-month NRM rates were 24.2% (11.2–39.9) and 31% (16–48), respectively. Infections were the most common cause of NRM (n=5), followed by aGVDH (n=3), and graft failure (n=2).

The cumulative incidence of EBV reactivation at 100 days was 39.1% (22–55), and only one patient developed post-transplant lymphoproliferative disorder. CMV reactivation at D+100 was noted in 21.2% (9.1–36.5). None had CMV-related end-organ damage.

The 12 & 24 months OS was 72.5% (53.7–84.7) and 62% (42–76), respectively, with a median OS of 42 months (13–NR). GRFS at 12 months was 28.7% (14.1–45.1), significantly higher in patients receiving ATG + PTCy prophylaxis (37% vs NR; p<0.001).

There were no significant differences in OS, CIR, NRM, aGVHD, cGVHD, EBV reactivation, or CMV reactivation between HID/MMUD and MSD/MUD cohorts (n=144). On MVA, HID/MMUD, and cryopreserved grafts were associated with a significantly higher incidence of PGF.

Conclusion:

HID and MMUD grafts demonstrate comparable OS, NRM, GVHD, and relapse rates to MSD/MUD grafts. However, significantly higher rates of PGF, particularly with cryopreserved stem cells, highlight the need for careful consideration of graft type. Prospective studies comparing different donor options are warranted.