Significance of measurable residual disease in adults with secondary acute myeloid leukemia undergoing allogeneic hematopoietic cell transplantation

TO THE EDITOR:

Many studies, including from our institution, have demonstrated an association between measurable residual disease (MRD) before allogeneic hematopoietic cell transplantation (HCT) and increased relapse risk/shorter survival in adults with acute myeloid leukemia (AML) [1-3]. While many of these studies have included patients with secondary AML, most have not focused on potential differences in the prognostic significance of MRD for patients with de novo AML relative to those with secondary disease. However, in a recent analysis from the Acute Leukemia Working Party of the European Society of Blood and Marrow Transplantation that encompassed 318 adults age 18–75 years with secondary AML, Maffini and colleagues reported no difference in post-HCT outcomes between patients with and without MRD at the time of allografting [4, 5]. An acknowledged limitation of such registry-based data is the heterogeneity in the MRD testing across the study cohort with lacking information regarding quality and performance characteristics of the MRD assays [4]. We therefore sought to determine the relative prognostic significance of MRD in adults with de novo or secondary AML who had pre-HCT MRD assessed uniformly.

To this end, we studied all adults ≥18 years with AML who underwent allografting while in first or second remission between 4/2006 and 5/2021 and had bone marrow MRD testing by multiparameter flow cytometry (MFC) before HCT [6]. The MRD assay was essentially unchanged throughout the study period, detecting MRD in most cases to a level of 0.1% and in progressively smaller subsets of patients as the level of MRD decreases below that level [7]. Following our previous approach, any detectable level of MRD was considered positive [6, 7]. The refined MRC/NCRI criteria [8] were used to assign cytogenetic risk at diagnosis. Secondary AML was defined as disease following an antecedent hematologic disorder or treatment with systemic chemotherapy and/or radio-therapy for a different disorder [6, 7]. Treatment response was categorized via 2017 European LeukemiaNet (ELN) criteria [9] except that relapse was defined as emergence of >5% blasts in blood or marrow, emergence of cytogenetic abnormalities seen previously, or presence/emergence of any level of disease if therapeutically intervened on [6, 7]. Data follow-up was current as of February 10, 2022.

Overall survival (OS) and relapse-free survival (RFS) were estimated using the Kaplan-Meier method. Probabilities of relapse and non-relapse mortality (NRM) were summarized using cumulative incidence estimates, with death without prior relapse considered a competing risk for relapse and relapse being a competing risk for NRM. Associations with RFS and OS were assessed using Cox regression; cause-specific regression models were used for relapse and NRM. Statistical analyses were performed using R (http://www.r-project.org).

We identified 979 patients who met study inclusion criteria and agreed to their data being used for research. Among these, 257 (26%) had secondary AML, of whom 217 (84%), 16 (6%), and 24 (9%) received peripheral blood, bone marrow, or umbilical cord blood as stem cell source, respectively. Of the non-cord blood allografts, 196 (84%) were from HLA-matched related or unrelated donors, 29 (12%) from 1–2 allele/antigen-mismatched donors, and 8 (3%) from haploidentical donors. There were expected differences between patients with de novo and secondary AML, including older age at HCT (median [range]: 59.6 [18.2–77.7] vs. 53.1 [18.0–80.9] years; P < 0.001), lower white blood cell (WBC) count at diagnosis (3.8 [0.3–295] vs. 9.6 [0.1–348] x 10³μL; P < 0.001), higher cytogenetic disease risk (e.g., adverse risk: 29% vs. 20%; P = 0.024), lower proportion of second rather than first remission (10% vs. 29%; P < 0.001), higher proportion of cases with MRD (29% vs. 16%; P < 0.001), lower likelihood of blood count recovery before HCT (64% vs. 72%; P = 0.022), higher HCT-Comorbidity Index (HCT-CI; P < 0.001), and lower proportion of patients receiving myeloablative conditioning (MAC) for those with secondary AML (46% vs. 64%; P < 0.001).

There were 460 deaths, 308 relapses, and 193 NRM events contributing to the estimates for relapse, OS, RFS, and NRM with a median (range) follow-up after HCT among survivors of 62 (3–182) months. Across all patients with secondary AML, the risk of relapse was 34% (95% confidence interval: 28–39%) at 3-years, with 3-year RFS and OS estimates of 48% (42–54%) and 54% (47–60%). The risks of NRM at 100 days and 3 years were 6% (3–9%) and 18% (14–23%), respectively. Relative to patients with de novo AML, estimated outcomes were worse for RFS (P = 0.008), OS (P = 0.022), and NRM (P = 0.024) but not relapse risk (P = 0.12) for individuals with secondary AML.

As shown in Fig. 1, pre-HCT MRD was associated with higher risk of relapse and inferior survival in both patients with de novo or secondary AML. For the latter, the 3-year risk of relapse was 27% (20–32%) for those without pre-HCT MRD vs. 50% (38–60%) for those with it. Corresponding 3-year RFS and OS were 56% (48–63%) and 61% (54–68%) for those without vs. 30% (20–41%) and 36% (25–47%) for those with pre-HCT MRD. For comparison, in patients with de novo AML, the 3-year risk of relapse was 21% (18–25%) for those without pre-HCT MRD vs. 71% (62–79%) for those with MRD, whereas 3-year RFS and OS were 62% (58–66%) and 67% (63–70%) for patients without MRD vs. 15% (9–23%) and 28% (20–37%) for those with it.

Fig. 1. Post-HCT outcomes for 257 adults with secondary AML and 722 adults with de novo AML undergoing allogeneic HCT, stratified by pre-HCT MRD status.

a Risk of relapse, b relapse-free survival, c overall survival, and d risk of non-relapse mortality.

To study the relationship between pre-HCT MRD status and post-HCT outcomes in more detail, we then evaluated multivariable regression models for the endpoints of relapse, RFS, and OS, accounting for age at HCT, cytogenetic risk at AML diagnosis (favorable/intermediate vs. adverse), WBC at diagnosis, cytogenetics at time of HCT (normalized vs. not normalized for patients presenting with abnormal karyotypes), pre-HCT peripheral blood counts (recovered vs. not), first vs. second remission at HCT, conditioning intensity (MAC vs. non-MAC), and stem cell source. Potential interactions between de novo/secondary AML status and pre-HCT MRD were evaluated in these multivariable models. Notably, a significant interaction was found for relapse (P = 0.001) and RFS (P = 0.002), but not OS (P = 0.14). The interaction models indicated that the hazard ratios (HRs) for the impact of pre-HCT MRD were less extreme among secondary AML compared to de novo AML patients. For the relapse endpoint, the HR for pre-HCT MRD was 5.35 (95% confidence interval: 3.99–7.16) among individuals with de novo AML, and 2.34 (1.52–3.62) among those with secondary disease. Likewise, for RFS, the HRs for pre-HCT MRD were 3.66 (2.85–4.69) and 1.94 (1.38–2.72), respectively, for patients with de novo and secondary AML. Lastly, corresponding HRs for OS were 2.62 (2.02–3.40) and 1.92 (1.35–2.73).

In this large study with uniform MRD testing, pre-HCT MRD status was prognostically informative of post-HCT outcomes in patients with secondary AML, albeit to a lesser degree than in patients with de novo disease. As one limitation, data from molecular testing was not available for most patients and could not be included in our analyses. Thus, it remains unknown whether the prognostic significance of MRD is similarly maintained across molecular subclasses, and to what extent the differential impact of MRD in de novo vs. secondary AML might be accounted for by differences in the underlying distribution of mutational profiles [10, 11]. Nonetheless, our data validate MFC MRD testing as a useful biomarker for risk stratification in patients with secondary AML undergoing HCT. Taken together with the results of Maffini and colleagues [4], our analyses highlight the need for standardization in the implementation and reporting of MRD testing in AML [3].

DATA AVAILABILITY

The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.