Next-Generation Sequencing-Based Measurable Residual Disease (MRD) Monitoring Is Highly Predictive for Outcome after Allogeneic Stem Cell Transplantation

Background: Molecular predictors for relapse following allogeneic haematopoietic stem cell transplantation (alloHSCT) are urgently needed as 30-40% of AML patients experience recurrence of disease after HSCT. Measurable residual disease monitoring (MRD) for nucleophosmin 1 ( NPM1 ) mutations has proven to be highly predictive for relapse in AML patients treated with or without alloHSCT. However, the majority of AML patients undergoing alloHSCT do not have the favorable NPM1 marker so that alternative markers and techniques are required. We studied a next-generation sequencing (NGS)-based error-corrected sequencing approach that is applicable to all mutations for its ability to prognosticate relapse and survival following alloHSCT.

Aim: To evaluate the prognostic impact of MRD before alloHSCT in AML patients in morphologic complete remission (CR) using a novel approach of error-corrected sequencing applicable to the majority of AML patients.

Patients and Methods: We studied 75 patients who underwent myeloablative (MA) (n=36) or reduced-intensity conditioned (RIC) (n=39) alloHSCT for AML in morphologic CR for the presence of MRD shortly before alloHSCT. All patients had at least one mutation in genes other than NPM1 or DNMT3A at the time of diagnosis that was identified by NGS with a myeloid panel on the Illumina platform. Amplicon-based error-corrected sequencing and bioinformatics was developed and applied to CR samples before alloHSCT analyzing 1-3 diagnostic mutations. Genomic DNA from peripheral blood (n=46) or bone marrow (n=29) was used for MRD assessment. Mutations in the following genes were used for MRD analysis in the indicated number of patients: IDH2 (n=12), FLT3 (n=10), RUNX1 (n=7), SF3B1 (n=6), NRAS (n=5), IDH1 and TP53 (n=4 each), KRAS and WT1 (n=3 each), CBL, ETV6, EZH2, STAG2 (n=2 each), BCOR, BCORL1, DDX41, PHF6, PTPN11, RAD21 , and SMC3 (n=1 each). 55 mutations were single nucleotide variants and 14 mutations were indels. The sensitivity of the assay was determined for each analysis and ranged from 10^-4 to 10^-5. Six patients (8%) had molecular persistence at high copy number in CR (variant allele frequency, VAF 31-51%) and were excluded from further analysis.

Results: 31 of 69 patients (45%) were found MRD positive before alloHSCT with a median VAF of 0.19% (range 0.012% to 14.5%). Clinical and transplantation-associated characteristics were similarly distributed between MRD positive and negative patients. Cytogenetic risk according to MRC, ELN risk groups and molecular aberrations showed no significant differences between MRD positive and negative patients apart from more MRD positive patients being SF3B1 mutated. The median follow up was 6.1 years. 19 of 31 MRD positive patients (61%) and 7 of 38 MRD negative patients (18%) relapsed after alloHSCT. In MRD positive patients the MRD marker was also found in the majority of available relapse samples (9 of 11 patients, 82%). MRD negative patients who relapsed had lost the MRD marker in 4 of 5 available relapse samples (80%), suggesting high accuracy of our MRD assay. By competing risk analysis for cumulative incidence of relapse (CIR) and non-relapse mortality (NRM) patients with positive MRD had a significantly higher CIR than MRD negative patients (P<.001; HR 4.85, 95% CI 1.95-12.1; 5 year CIR 67.7 vs 19%, Figure 1A), while NRM was not significantly different (P=.31; HR 0.44, 95% CI 0.09-2.14). Overall survival (OS) was significantly shorter in MRD positive compared to MRD negative patients (P=.029; HR 2.26, 95% CI 1.07-4.79; 5 year OS 38 vs 70%, Figure 1B). In multivariate analysis MRD status was an independent predictor of CIR (P<.001, HR 5.80, 95% CI 2.18-15.48) besides cytogenetic risk, and of OS (P=.013, HR 2.57, 95%CI 1.22-5.38) besides cytogenetic risk and WBC count.

Conclusion: We provide evidence that NGS-based MRD monitoring can be applied to a large proportion of AML patients using almost any available molecular aberration. This approach requires morphologic confirmation of CR status in order to exclude patients with molecular persistence at high copy number, likely originating from preleukemic clones or germline mutations. NGS-based MRD is highly predictive of relapse and survival and may help refining transplant and posttransplant management in AML patients.

View largeDownload slide

View largeDownload slide

Close modal