Introduction: Persistent AML is a known risk factor for poor outcomes after allo-HCT. The impact of MRD in patients who achieve complete remission (CR) or CR with incomplete count recovery (CRi) has been less well studied.

Methods: We retrospectively reviewed the records of AML patients who underwent allo-HCT in morphological remission (<5% myeloblasts and normal marrow cellularity) with or without blood count recovery between January, 2000 and January, 2014. Data was collected for variables known to impact the prognosis of AML patients (Table 1). MRD was defined as evidence of abnormalities associated with AML by either flow cytometry, cytogenetics or Fluorescence in situ hybridization (FISH). The impact of MRD identified at the time of allo-HCT on cumulative incidence of relapse (CIR), progression free survival (PFS), and overall survival (OS) was assessed in MRD+ and MRD- patients.

Results: A total of 166 eligible patients were identified. The median follow-up among living patients is 46 months (range, 13-103).Thirty seven (22%) patients had evidence of MRD (13 by flow cytometry only, 17 by cytogenetics/FISH only and 7 by both). MRD was more common in patients with poor risk karyotype at diagnosis and CRi at the time of allo-HCT (Table 1). PFS (P= 0.0016), OS (P=0.002), and CIR (P=0.02) were all significantly worse in MRD+ patients (Figures 1& 2). In univariate analysis, MRD+ patients, assessed by flow cytometry had worse PFS (P=0.0216) and OS (P=0.0314) compared to MRD- patients. Similarly patients with evidence of MRD+ by cytogenetics/FISH had worse PFS (P=0.007) and OS (P=0.0031). In a multivariate cox proportional hazards model 1) any MRD positivity prior to allo-HCT, 2) poor-risk karyotype at diagnosis, and 3) CRi at allo-HCT independently predicted significantly poor PFS and OS. Only poor-risk karyotype was associated with a significant increase in CIR, while MRD positivity showed a trend towards higher CIR.

Conclusion: MRD positivity prior to HCT by either flow cytometry or by cytogenetics/FISH independently predicts adverse AML outcomes.

Table 1.

Comparison of pre-transplant variables

CovariateLabelMRD +
(N=37)
MRD -
(N=129)
P-Value
Age(years) < 40 8 (21%) 20 (16%) 0.708 
40 - 59 20 (53%) 69 (54%) 
≥ 60 10 (26%) 39 (30%) 
Karyotype risk Favorable/
Intermediate 
19 (53%) 95 (74%) 0.011 
Poor 18 (47%) 33 (26%) 
Timing of Allo-HCT 1st remission (CR1) 28 (74%) 97 (76%) 0.792 
> CR1 10 (26%) 31 (24%) 
Allo-HCT after1st relapse(>CR1): duration of CR1 > 12 mo 31 (82%) 113 (88%) 0.285 
≤ 12 mo 7 (18%) 15 (12%) 
Secondary AML No 23 (60%) 78 (61%) 0.964 
Yes 15 (40%) 50 (39%) 
Complete remission vs CRi CR 28 (74%) 110 (86%) 0.077 
CRi 10 (26%) 18 (14%) 
Conditioning Regimen Ablative 24 (63%) 72 (56%) 0.449 
Other 14 (37%) 56 (44%) 
Donor Type Matched
sibling donor 
12 (32%) 42 (33%) 0.887 
Other 26 (68%) 86 (67%) 
Female donor: male recipient (FDMR) Other 28 (80%) 91 (78%) 0.844 
FDMR 7 (20%) 25 (22%) 
CovariateLabelMRD +
(N=37)
MRD -
(N=129)
P-Value
Age(years) < 40 8 (21%) 20 (16%) 0.708 
40 - 59 20 (53%) 69 (54%) 
≥ 60 10 (26%) 39 (30%) 
Karyotype risk Favorable/
Intermediate 
19 (53%) 95 (74%) 0.011 
Poor 18 (47%) 33 (26%) 
Timing of Allo-HCT 1st remission (CR1) 28 (74%) 97 (76%) 0.792 
> CR1 10 (26%) 31 (24%) 
Allo-HCT after1st relapse(>CR1): duration of CR1 > 12 mo 31 (82%) 113 (88%) 0.285 
≤ 12 mo 7 (18%) 15 (12%) 
Secondary AML No 23 (60%) 78 (61%) 0.964 
Yes 15 (40%) 50 (39%) 
Complete remission vs CRi CR 28 (74%) 110 (86%) 0.077 
CRi 10 (26%) 18 (14%) 
Conditioning Regimen Ablative 24 (63%) 72 (56%) 0.449 
Other 14 (37%) 56 (44%) 
Donor Type Matched
sibling donor 
12 (32%) 42 (33%) 0.887 
Other 26 (68%) 86 (67%) 
Female donor: male recipient (FDMR) Other 28 (80%) 91 (78%) 0.844 
FDMR 7 (20%) 25 (22%) 

Disclosures

No relevant conflicts of interest to declare.

Author notes

*

Asterisk with author names denotes non-ASH members.

Sign in via your Institution