Improved Outcomes in Paroxysmal Nocturnal Hemoglobinuria Patients Undergoing Allogeneic Hematopoietic Stem Cell Transplantation in 2011-2020: A Retrospective EBMT-Saawp Study

Introduction

Allogeneic Hematopoietic Stem Cell Transplantation (HSCT) in Paroxysmal Nocturnal Hemoglobinuria (PNH) patients has not been investigated in the era of complement inhibitors and available data demonstrated survival rate below 70% (Peffault de Latour et al., Haematologica 2012). Here we describe a retrospective study conducted by the Severe Aplastic Anemia Working Party (SAAWP) looking at the outcomes of PNH patients undergoing HSCT between 2011 and 2020.

Methods

All consecutive first HSCT for PNH performed at EBMT centers were evaluated in retrospective analysis. Data were retrieved from the EBMT registry Promise (looking specifically for the diagnosis of PNH), which includes detailed information about the underlying disease, the HSCT procedure, and treatment outcomes, including engraftment, graft failure, acute and chronic graft versus host disease (GVHD), overall survival (OS), as well as event-free survival (EFS; events: primary and secondary graft failure, second transplant, relapse and death) and GvHD/relapse-free survival (GRFS; events: aGvHD III-IV, extensive cGvHD, relapse, graft failure (any)/no engraftment and death). Death was a competing event for engraftment, graft failure and GVHD.

Results

A total of 240 PNH patients underwent HSCT in the decade 2011-2020 in 125 EBMT transplant centers, with a slight imbalance between the two quinquennia (60% vs 40%). Median age at HSCT was 26.4 years (IQR, 18.6 to 37.2 years). The median time from diagnosis to allo-HSCT was 10.3 months (IQR, 4.8 to 34.7 months). Median follow up was 2.9 years. Patients were transplanted either from an HLA-matched sibling (MSD; 46.7%), matched unrelated (MUD; 37.1%), mismatched-unrelated (MMUD; 12.9%) or mismatched haplo-identical (Haplo; 3.3%) donor. The stem cell source was peripheral blood (PB) in 52.9% and bone marrow (BM) in 45.8%; 76.6% of patients received reduced-intensity conditioning (RIC) and 23.4% myeloablative conditioning (MAC) regimen. GvHD prophylaxis was based on calcineurin inhibitors alone (CNI; 23.7%), or associated to serotherapy (anti-thymocyte globulin 67.8%, alemtuzumab 3.4%) or to post-transplant cyclophosphamide (5.1%). The cumulative incidence of engraftment by day 28 was 87% for neutrophils (median time 18 days (16-19) and 85% by day 60 for platelets (median time 22 days (18-24). The 3-years OS was 79%, with infections and GVHD as main causes of death. Strikingly, donor type and patient age were associated with survival with 3-years estimates of OS 86% for MSD, 78% for MUD, 62% for MMUD (p=0.003). Three-year OS was 83% in patients <20 years, 82% 20-40 years, 67% >40 years (p=0.047). Interval between diagnosis and transplant (>1 years) was also associated with OS (73% vs 84% <1 year; p=0.036), while conditioning regimen, comorbidity index (HCT-CI), stem cell source and GVHD prophylaxis (CNI vs serotherapy) was not.

Rate of grade II-IV acute GVHD was 14% (MSD 12%; MUD 10%; MMUD 28%; p=0.046), and chronic GVHD was 13% (MSD 15%; MUD 5%; MMUD 25%; p=0.06). As a result, 3-years GRFS was 66%, with an impact of donor type (MSD 74%; MUD 65%; MMUD 46%; p=0.025) and of interval between diagnosis and transplant (>1 year 57% vs 73% £1 year; p=0.02). Primary (at 30 days) and secondary (at 3 years) graft failure were 8% and 6%, respectively. Donor type impacted primary graft (MSD 2%; MUD 14%, MMUD 3%; p<0.001) but not secondary graft failure, whereas stem cell source impacted only secondary graft failure (BM 10%, PB 3%; p=0.008).

The 3-years EFS was 69%. Here, donor type (MSD 80%, MUD 65%, MMUD 54% (p=0.006) and interval between diagnosis and transplant (>1 year 61% vs 76% <1 year; p=0.015) impacted EFS; conditioning regimen, patient’s age, HCT-CI, stem cell source and GvHD prophylaxis did not.

Out of the 8 patients transplanted from an Haplo donor, 4 experienced graft failure (1 died, 2 rescued by second HSCT), 2/8 suffered from aGVHD and 1/7 from cGVHD (1 fatal), with 3/8 being alive free of GVHD and relapse.

Conclusions

The outcome of PNH patients receiving allogeneic HSCT has meaningfully improved in the last decade with increasing OS rate in HLA-identical donors (both related and unrelated), especially in young patients; a RIC regimen was as good as a MAC regimen. Complement inhibitors remain the standard of care for hemolytic PNH patients, however allogeneic HSCT can be considered an alternative option in case of non-availability of complement inhibitors, or in specific clinical settings.