Hematopoietic Cell Transplantation in Aplastic Anemia

Immune Suppression Therapy or Hematopoietic Cell Transplant: Which Should Come First?

High transplant-related mortality resulting largely from GVHD and graft failure following serologically typed, unrelated donor (UD) transplants in the 1980s led to HCT being reserved as a first-line treatment option only for young patients (≤ 40 years old) with matched sibling donors (MSD). For all other patients, IST with hATG and CSA is the typical upfront treatment modality. In these settings, allogeneic HCT is reserved for IST-refractory patients or after first relapse.

Response to IST in the first line is slow with median onset at 3 months and more often results in incomplete hematologic recovery after 6 months of IST. Addition of eltrombopag to IST resulted in up to 94% overall response (recovery of at least one cell line) at 6 months; however, complete hematologic response was seen only in 58% of patients. Five-year OS following IST ranges from 60% to 85% in more recent reports. Therapeutic failures following IST also include disease relapse in 30% to 44% of patients and clonal evolution into myeloid neoplasms, which occurs in 10% to 20% of patients. As discussed later, long-term event-free survival (EFS) tends to be lower following IST compared to after MSD HCT. The survival benefit of HCT is most evident in younger patients. While a second course of IST may be considered before HCT, particularly in older patients, EFS following salvage IST is worse than after upfront HCT.

Points in favor of considering upfront HCT in the absence of MSD include the aforementioned complications of IST, refinements to HCT in SAA that have improved posttransplant outcomes, and the observation that longer duration from diagnosis to HCT negatively affects outcome.

Comparison of IST to HCT in SAA Favors Upfront HCT from Both Matched Related and Unrelated Donors for Children

An analysis by the European Bone Marrow Transplant (EBMT) SAA Working Party of 304 patients under the age of 15 years who received either IST or HCT between 1970 and 1988 revealed an estimated 10-year survival of 48% (IST) and 63% (HCT), respectively. Factors associated with improved survival in the HCT cohort were receiving HCT as first line, the use of CSA, and shorter time to transplant (< 90 days). Neutropenia categorized as very SAA (absolute neutrophil counts [ANC] < 0.2 × 10 ⁹ /L) at time of diagnosis was associated with poorer outcomes in patients receiving IST. This study and other reviews led to recommendations that HCT even with UDs should be considered for young children with SAA.

A pediatric study in Japan conducted between 1984 and 1998 similarly showed inferior 10-year OS of 55% for upfront IST (n = 63) compared to the 97% OS seen in 37 HCT patients, 26 (70%) of whom received upfront HCT from MSD. In addition, 10-year failure-free survival (FFS) in this study was superior in upfront HCT (97%) versus 40% in the IST cohort. The HCT cohort analyzed included 11 patients who received matched unrelated donor (MUD) transplants for SAA refractory to IST. A subsequent study from the same population evaluated patients aged 0 to 16 years, treated between 1992 and 2009 for SAA with upfront IST (n = 386) or upfront HCT from MSD (n = 213). Ten-year OS in this time period was not statistically different at 88% for IST and 92% for HCT, while 10-year FFS was again superior for HCT at 87% versus 56% after IST. Of the 113 patients that failed or relapsed after IST (29% of cohort) and subsequently had HCT from non-MSD, 10-year OS was 79%. This latter finding of poorer outcomes following second-line HCT also highlighted the impact of longer duration from SAA diagnosis.

In a time series comparing upfront IST and HCT in multiple European centers before year 1999, 10-year OS was similar for patients aged 1 to 20 years (bone marrow transplant 67%, IST 60%; P = .09) but slightly lower for those aged 21 to 40 years receiving HCT (HCT 58%, IST 62%; P = .002). This EBMT SAA Working Party study included both MSD and MUD transplants in the HCT cohort. The analysis also showed that for patients aged 1 to 20 years who had ANC < 200/μL at the time of diagnosis (i.e., very severe AA), the benefit of HCT increased with time compared to IST. This late benefit was attributed to the development of late clonal disorders in the IST patients.

Dufour et al. evaluated outcomes in a series of 29 pediatric (< 20 years) SAA patients who received upfront MUD with fludarabine (FLU), CY, and alemtuzumab (CAMP) as conditioning regimen. Outcomes from this cohort were compared to matched historical controls from a similar population who received upfront MSD HCT, upfront IST, or second-line MUD HCT after failure of IST. Although five of the upfront MUD cohort (17%) were single-allele mismatches, 2-year OS and EFS were not significantly different from the MSD cohort (OS, 96% MUD, 91% MSD; EFS, 92% MUD, 87% MSD). Low rates of aGVHD were seen (10% grade II–IV, 3.5% grade II–IV) and no chronic GVHD (cGVHD) was observed at median of 1.7 years follow-up. Patients receiving upfront IST had similar 2-year OS (94%) with recipients of upfront MUD; however, 2-year EFS was significantly lower following IST (40% vs. 92% after MUD HCT, P = .0001).

In addition to validating earlier reports showing the noninferiority of MUD to MSD for SAA patients younger than 20 years, this study also underscored the importance of prioritizing this approach in first-line approaches. The 24 patients analyzed in the matched historical cohort who received salvage MUD after IST failure had significantly lower 2-year OS and EFS (both 74%, P = .02) compared to upfront MUD. The median time from diagnosis to second-line MUD was 1.1 years (range 0.3–4.1 years).

From the EBMT SAA Working Party, an analysis of SAA patients managed from 1999 onward showed that 44% of patients receiving upfront IST went on to undergo HCT as salvage for refractory disease. In this study, there was no statistical difference in 10-year survival between upfront HCT and IST for patients aged 21 to 40 years (76% HCT, 65% IST; P = .6). Factors associated with improved survival for HCT were younger age and duration of less than 100 days from SAA diagnosis to start of treatment. Patients older than 40 years fared worse with either treatment approach, resulting in 10-year OS of 56% (HCT) and 57% (IST). The survival curves were more significantly in favor of IST until after 5 years from treatment when both curves began merging toward the 10-year time point.

In aggregate, these studies indicate that HCT should be first-line management of SAA even in the absence of an MSD if an MUD is available for patients ≤ 20 years old and may be considered between ages 20 and 30 years. An ongoing multicenter study (NCT02845596) randomizes between UD HCT and IST as upfront therapy for SAA ≤ 25-year-old patients without MSD. Along with treatment outcomes, this study is designed to test feasibility of randomization and timeliness to transplant.

When Should Hematopoietic Cell Transplantation Be Considered in Patients Older Than 40 Years?

Because disease relapse and clonal evolution to myeloid neoplasm following IST are common occurrences, there is continued interest in expanding HCT to more SAA patients across the age spectrum even in the absence of MSD. This is even more pertinent as transplant outcomes in general improve because of better supportive treatments, infection management, and optimization of conditioning regimen to minimize toxicity.

Over the past three decades, older age has remained a poor prognostic factor for outcomes following HCT in SAA for both MSD and other donors, with the worst outcomes seen in patients 60 years and older. This is also true of IST as seen in a prospective EBMT study of patients aged > 60 years receiving first-line IST resulting in 3-year OS of 65%. A review of 439 patients transplanted between 2010 and 2015 at median age of 52 (range 40–77) years from MSD (48%) and UD (52%) showed 5-year OS in the entire cohort of 58%. Subset analysis demonstrated 67%, 58%, and 45% survival for patients aged 40 to 49 years, 50 to 59 years, and > 60 years, respectively. Factors associated with improved outcomes were use of ATG or alemtuzumab and treatment in HCT centers that transplanted more than three patients in the time period covered. Combined primary and secondary graft failure (10.7%) and grade II-IV aGVHD rates (11%) did not appear different from HCT done on the same age group in the prior decade, while cGVHD at 25% was reduced. Donor type did not significantly influence outcomes with composite 5-year OS of 54.5% (MUD) and 63.5% (MSD).

Reducing the conditioning CY dose from 200 mg/kg used in MSD to 50 mg/kg by including FLU along with in vivo T-cell depletion for MUD HCT may be associated with improved outcomes (reduced toxicity and higher OS) in older patients. Low-dose TBI facilitates engraftment when in vivo T-cell depletion uses ATG. Substituting ATG with alemtuzumab allows for the omission of TBI and results in rates as low as 13.7% acute and 2.7% chronic GVHD. A subsequent review of older patients undergoing FLU, CY, and Alemtuzumab conditioning before largely MUD grafts showed similar GVHD-free, relapse-free survival of 86% for patients > 50 years, and those ≤ 50 years. Rather than age, an HCT-specific comorbidity index (CI) less than 3 was more predictive of superior survival (OS 98% vs. 76% with HCT-CI score ≥ 3). This finding was confirmed in a large registry (combined EBMT and Center for International Blood and Marrow Transplant Research [CIBMTR]) study of patients aged 50 to 77 years (median 57.8 years) who received MSD (n = 275) or MUD (n = 187) for IST-refractory SAA in at least 90% of patients. An analysis of the 56% 3-year OS probability seen (59% in MSD, 52% in MUD) showed that Karnofsky performance scores (KPS) 90 to 100 correlated with improved survival (66% for MSD, 57% for MUD) compared to KPS < 90 (57% for MSD, 48% for MUD). No other variable, including age at transplant, correlated with significantly with survival.

Findings from these studies suggest that functional status and comorbid conditions should trump absolute age when evaluating SAA patients older than 40 years for MSD and MUD HCT. However, the consensus remains that this discussion is best reserved for patients who are refractory to upfront IST.

Haploidentical (Haplo) Donor HCT is a Viable Option for Second-Line Management of Severe Aplastic Anemia

An analysis of 11 studies reporting haplo-HCT in pediatric and young adults (median ages 13 and 27 years, respectively) found comparable survival in the 80% to 90% range between both groups. There was no appreciable difference in engraftment rates between reduced-intensity conditioning and nonmyeloablative and myeloablative conditioning regimens.

The Chinese Bone Marrow Transplant Registry reported on 89 haplo-HCT done upfront for SAA in patients aged 4 to 51 (median 22) years showed 98% engraftment, 86% 3-year OS, and 85% 3-year FFS. There was no significant difference in these outcomes compared to a cohort from the same registry that received MSD HCT first line. Marked differences were observed in the rates of aGVHD, which were 30.3%, 10.1%, and 30.6% for grade II to IV aGVHD, grade III to IV aGVHD, and cGVHD, respectively, in the haplo cohort compared to < 5% for all three outcomes in the MSD cohort. Of note, granulocyte colony stimulating factor primed bone marrow combined with peripheral blood mononuclear cells were the grafts used for haplo-HCT.

DeZern et al. showed no graft failures, minimal aGVHD, and satisfactory survival outcome following haplo-HCT as management of IST-refractory SAA at Johns Hopkins. The “Baltimore conditioning regimen” in these patients included rATG (0.5 mg/kg on T-9 and 2 mg/kg on T-8 and -7), FLU (30 mg/m ² /day on T-6 to -2), CY (14.5 mg/kg/day on T-6 & -5), and TBI 2 to 4 Gy on T-1. GVHD prophylaxis was with posttransplant CY (PTCY) at a dose of 50 mg/kg/day on T+3 and +4 in combination with ATG, MMF, and a calcineurin inhibitor (CNI). The initial cohort of patients experienced low rates and severity of aGVHD (12% grade II–IV, 2.5% grade II–IV), favorable engraftment (< 10% engraftment failure), and survival rates of 80% to 100%. A follow-up study in a larger cohort of 37 patients using this regimen that also included 17 treatment-naïve patients showed 2-year OS of 94% and cumulative incidences of grade II to IV aGVHD and cGVHD of 11% and 8%, respectively. A TBI dose of 4 Gy was associated with enhanced engraftment in treatment-naïve patients.

In an EBMT review of 33 SAA patients ranging from 2 to 45 years (median 20.4 years) undergoing HCT, 48% received the Baltimore conditioning regimen. The entire cohort had median time to neutrophil engraftment of 22 days, seven engraftment failures, 23% aGVHD II to III (no grade IV), 10% cGVHD at 2 years, and 2-year GVHD/relapse-free survival (GRFS) of 63%. At a median follow-up of 35.9 months, 1- and 2-year OS were 82% and 78%, respectively. The use of the Baltimore conditioning regimen was associated with a higher 2-year OS of 93% compared with 64% in the other regimens. Age group, stem cell source, and ATG use did not influence OS.

Favorable outcomes from haplo-HCT in the era of PTCY makes this donor source a viable approach for second-line management of SAA in the absence of a MSD and MUD. Haplo-HCT with PTCY as salvage after failure of IST is the subject of an ongoing multicenter prospective study by the Blood and Marrow Transplant Clinical Trials Network (BMT CTN) along with collaborators at the National Cancer Institute, National Heart, Lung, and Blood Institute, and National Marrow Donor Program (BMT CTN 1502; NCT02918292).