Allogeneic Hematopoietic Cell Transplantation (HCT) Yields Lower Relapse Rates but No Overall Survival Benefit for Adults with Acute Lymphoblastic Leukemia (ALL) in First Minimal Residual Disease (MRD)-Negative Remission

BACKGROUND: Allogeneic HCT is used commonly in adults with ALL.  MRD is an accepted prognostic factor in ALL.  Being MRD-negative (MRD^Neg) overcomes age and cytogenetics (CG; Bassan, Blood, 2009), and it is predictive of outcome without (No) HCT and prior to myeloablative (MAC) and reduced-intensity HCT (RIC).  However, comparisons of these treatments in this favorable-risk subgroup are limited.  We sought to examine outcomes with these 3 approaches exclusively in pts in first MRD^Neg remission (MRD^Neg CR1).

METHODS: We performed a retrospective analysis of pts that were diagnosed or referred for HCT at our Center since 2005.  MRD^Neg was defined as undetectable by all methods used: Multiparameter flow cytometry was used in all pts, though other measures (PCR, FISH, and/or CG) were considered if performed.  Clinical risk factors at diagnosis (dx) included age ≥ 35, high white blood count (WBC; > 30 in B-ALL, > 100 in T-ALL), and adverse CG (Ph+, low hypodiploid/near triploid, complex [≥ 5 abnormalities], MLL rearranged, t[8;14], -7, and +8).  Group characteristics were compared with Fisher's exact test.  Kaplan-Meier plots and Cox proportional hazard models were used to investigate associations between variables. 

RESULTS: Of 127 pts identified, 53 (42%) received MAC, 19 (15%) received RIC, and 55 (43%) No HCT while in MRD^Neg CR1.  RIC pts were significantly more likely than MAC and No HCT pts to be ≥ 35 years old at dx (89% vs 60% vs 51%, respectively [resp.]; P = 0.0084) and have adverse CG (83% vs 53% vs 30%, resp.; P < 0.001).  There were no significant differences in WBC at dx, initial therapy received, or (among the HCT groups) stem-cell donor or product.  All Ph+ pts received tyrosine kinase inhibitors.

Overall survival (OS) from dx in these 3 groups was similar (P = 0.35; Fig 1), though relapse was significantly less with MAC (P = 0.01; Fig 2).  3-year (yr) estimates of OS with MAC, RIC, and No HCT were 71%, 73%, and 75% (resp.), and 30%, 51%, and 60% (resp.) for relapse.  3-yr non-relapse mortality was 17% with MAC vs 5% with RIC (P = 0.15).  Compared to No HCT, the risk of death was not different with MAC (hazard ratio [HR] 1.6, P = 0.19) or RIC (HR 1.0, P = 1.0), adjusted for WBC.  However, relapse was lower with HCT, adjusted for WBC and (due to more events) adverse CG: significantly so after MAC (HR 0.35, P = 0.0014) but not with RIC (HR 0.57, P = 0.18).  OS was also similar with Ph+ and Ph- studied separately (P = 0.99 and 0.14, resp.).

In a distinct analysis of 23 pts who underwent HCT while MRD^Neg beyond CR1 (CR2+), OS from HCT was comparable for CR1 and CR2+ (P = 0.46).  3-yr estimates of OS were 69% and 62% (resp.).

CONCLUSIONS: Among contemporary adults with ALL in MRD^Neg CR1, RIC pts were more commonly older with adverse CG.  MAC yielded significantly less relapse, but OS was no different than with RIC or No HCT.  HCT while MRD^Neg yields similar outcomes in CR1 and CR2+.  For pts in MRD^Neg CR1, the role of up-front HCT is less clear.