The Measurable Residual Disease Status of Acute Myeloid Leukemia in First Complete Remission at Allogeneic Hematopoietic Cell Transplantation Interacts with Conditioning Intensity to Predict Outcome

BACKGROUND: Allogeneic hematopoietic cell transplant (allo-HCT) is offered to selected patients with acute myeloid leukemia (AML) in first complete remission (CR1). Despite overall survival (OS) benefits compared to non-HCT therapy, transplant related mortality (TRM) & relapse incidence (RI) remain obstacles. Reduced intensity (RIC) & non-myeloablative (NMA) conditioning regimens have reduced early toxicity & permit extension of allo-HCT to older or infirm patients but RI offsets this benefit. Disease status at allo-HCT predicts subsequent relapse. Patients with AML CR1 & measurable residual disease (MRD POS) have a higher RI than those who are MRD NEG. We hypothesised that in AML CR1, MRD status would interact with conditioning intensity to predict survival & toxicity in patients <50y & ≥50y. METHODS: This was a multicentre, retrospective registry study by the Acute Leukemia Working Party of the European Society for Blood & Bone Marrow Transplantation. Eligibility: Age ≥18y, first allo-HCT 2000-15, diagnosis of AML with a molecular marker, CR1 & availability of MRD status (polymerase chain reaction / immunophenotyping as per center policy), peripheral blood stem cells (PBSC) or bone marrow (BM) from a matched related (MRD), volunteer unrelated (VUD) or haplo-identical donor.Univariate & Cox Regression multivariate analyses (MVA) were applied to data. Measured outcomes included 2y OS, leukemia free survival (LFS), non-relapse mortality (NRM), graft vs host disease (GVHD) & GVHD-free/relapse-free survival (GRFS).

RESULTS: A total of 2292 patients, 1192 male, were eligible. Conditioning intensity was high intensity "myeloablative" (MAC) in 1254 (55%), RIC in 867 (38%) & NMA in 171 (7%). Patients were defined by age, MRD status & conditioning. This yielded 4 paired groups of patients: <50y MRD POS MAC (N=240) vs RIC/NMA (N=58); <50y MRD NEG MAC (N=665) vs RIC/NMA (N=195); ≥50y MRD POS MAC (N=126) vs RIC/NMA (N=230) & ≥50y MRD NEG MAC (N=223) vs RIC/NMA (N=555). Groups were comparable for the following parameters: Cytogenetic risk groups (intermediate in 71% & high risk in 29%); FLT3 or NPM1 mutation status (reported in 55% & 45%); CR rate with first induction (69% & reported in 1706 [74%]). While donor source & gender were also comparable, there was significantly greater use of PBSC than BM within all groups. In vivo T-cell depletion (TCD) was used in 46% & 59% of MAC & RIC allo-HCT overall & retained significance within all paired groups.

Univariate analysis of all patients confirmed the predictive effect of MRD status on allo-HCT outcomes. Patients <50y & MRD POS benefited from MAC vs NMA/RIC allo-HCT with RI 36.6% vs 50.7% (P=0.013), LFS 49% vs 36.5% (P=0.004), OS 59.4% vs 43.7% (P=0.008) & GRFS 35.3% vs 28.2% (P=0.025). In contrast, patients who were <50y MRD NEG, ≥50y MRD POS or ≥50y MRD NEG derived no survival benefits from MAC compared to RIC/NMA allo-HCT.

Next, Cox regression analysis was performed, adjusting for age by decade, cytogenetic risk, stem cell source, donor & TCD. In <50y MRD POS, RIC/NMAwas inferior to MAC for RI (HR 1.7 CI 1.05-2.79) & LFS (HR 1.6 CI 1.01-2.40) with a trend to lower OS. Use of TCD reduced acute GVHD grades III-IV (HR 0.32 CI 0.12-0.843).

In <50y MRD NEG, RIC/NMA & MAC allo-HCT had equivalent RI, NRM, LFS & OS but less chronic GVHD following RIC/NMA regimens (HR 0.7 CI 0.52- 0.99). Adverse factors in this group were poor risk cytogenetics (impacting RI, LFS & OS) & use of VUD vs MRD, associated with higher aGVHD & cGVHD rates. TCD conferred reduced GVHD & LFS without significant effect on OS. Use of PBSC increased NRM & GVHD toxicity without significant effect on LFS or OS.

In ≥50y MRD POS patients, RI, NRM, LFS & OS were unaffected by conditioning intensity. Poor risk cytogenetics exerted adverse effects on RI (HR 1.8 CI 1.26-2.57), LFS (HR 1.7 CI 1.28-2.31) & OS (HR 1.8 CI 1.32-2.47). Increasing recipient age associated with increased risk of aGVHD (III-IV) (HR 3.2 CI 1.36 - 7.61) but no other parameter altered outcome.

In ≥50y MRD NEG patients, no benefits accrued to MAC for RI, NRM, LFS & OS & there was a trend to lower NRM in RIC/NMA (HR 0.7 CI 0.48-1.03). TCD increased RI & reduced NRM & GRFS without effect on LFS or OS.

CONCLUSION: MAC allo-HCT benefits <50y MRD POS patients. RIC allo-HCT may be sufficient for other groups. Procedural toxicity may be reduced further by in vivo TCD. MRD impact on allo-HCT needs prospective evaluation & standardization of molecular & immunophenotyping assays.