Clinical Relevance of Morphologic and Molecular Complete Remission in AML Patients Undergoing Reduced Intensity or Non-Myeloablative Conditioning

Introduction: For most acute myeloid leukemia (AML) patients (pts) an allogeneic stem cell transplantation (HSCT) offers the best chance for relapse-free long-term survival. Evaluation of measurable residual disease (MRD) at HSCT allows risk stratification additionally to genetic risk at diagnosis. Pts with active AML or with positive MRD status (MRD+) pre-HSCT have similar and dismal outcomes following myeloablative conditioning. Reduced intensity (ric) or non-myeloablative (nma) conditioning regimes have been developed to allow HSCT also in AML pts with higher age or comorbidities. Here, we analyzed and compared the clinical relevance of morphologic and MRD-based remission status in AML pts prior to nma- or ric-HSCT.

Methods: We analyzed 345 AML pts who received an allogeneic peripheral blood HSCT at a median age of 63 (range 21-77) years with active disease (34%), or in first (49%) or second (17%) complete remission (CR, 87% of pts in remission) or CR with incomplete peripheral recovery (CRi, 13% of pts in remission) after nma (75%) or ric (25%) conditioning. Donors were HLA matched related (11%), matched unrelated (67%), antigen mismatched unrelated (21%) or haploidentical related (1%). At diagnosis, cytogenetics and the mutation status of CEBPA, NPM1 and presence of FLT3-ITD were assessed. Using a next-generation targeted amplicon sequencing approach we analyzed a panel comprising 54 recurrently mutated genes in myeloid malignancies on the MiSeq platform (Illumina). Pre-HSCT morphologic remission status as well as MRD status in pts in morphologic CR/CRi based on NPM1 mutations, BAALC, MN1 and WT1 expression were evaluated. MRD+ pts were defined by being positive for any of the analyzed markers. Median follow up after HSCT was 2.2 years.

Results: Pts transplanted with active disease differed from pts in remission with or without MRD pre-HSCT: they were less likely to have de novo AML (P=.02 & P=.09, respectively) and had higher genetic risk including a higher frequency of an abnormal (P=.001 & P<.001, respectively), a complex (P=.06 & P=.04, respectively) or a monosomal karyotype (P<.001 & P=.003, respectively), a lower frequency of NPM1 mutations (P<.001 & P<.001, respectively) and worse ELN genetic risk (P<.001 & P<.001, respectively). They were also more likely to receive ric-HSCT (P<.001 & P<.001) because pts with active AML were frequently transplanted after FLAMSA conditioning. Pre-HSCT MRD- pts only differed from pre-HSCT MRD+ pts regarding a lower white blood count (P=.006) and lower circulating blasts at diagnosis (P=.05) while the proportion of pts transplanted in CR or CRi did not differ between MRD- and MRD+ AML pts. Also the number of applied chemotherapy cycles prior to HSCT did not differ between the three pts groups. Pre-HSCT MRD- pts had a significantly lower cumulative incidence of relapse/progression (CIR) compared to both MRD+ pts (P<.001) and pts transplanted with active disease (P<.001) while CIR did not differ between MRD+ pts and pts transplanted with active disease (P=.24, Figure 1A). Pre-HSCT MRD- pts had longer overall survival (OS) than pre-HSCT+ pts (P=.04) who again had longer OS than pts transplanted with active disease (P=.01, Figure 1B). In multivariate analyses, the MRD corrected remission status prior to HSCT remained a significant factor for CIR (Hazard Ratio 1.65, Confidence interval [CI] 1.31-2.06) after adjustment for ELN risk and for OS (Odds Ratio 0.63, CI 0.49-0.84) after adjustment for ELN risk, hemoglobin and platelet count at diagnosis.

Conclusion: AML pts transplanted with active disease showed a variety of high-risk diagnostic parameters compared to pre-HSCT MRD- or MRD+ pts, as secondary disease and adverse genetic risk. In contrast, pre-HSCT MRD- and MRD+ pts could not be discriminated by high risk factors at diagnosis, underlining the importance of a dynamic risk stratification during remission. Both pts with active disease or a MRD+ status in CR/CRi prior to nma- or ric-HSCT showed dismal outcomes with higher CIR and shorter OS than MRD- pts. However, while CIR was comparable in pts with molecular or morphologic detectable disease, OS was worst in pts transplanted with active disease, indicating that MRD+ pts might still be salvageable after suffering relapse and able to achieve long-term outcomes.