Table 3.

Liquid biopsies in acute myeloid leukemia.

Target	Methods	Cohort Size/Disease Stage	Evidence: Key Points	Application	Reference
IDH1 and IDH2 genes (R140 and R172 mutations)	Sanger, ddPCR, NGS, and qPCR	n = 60 Diagnosis	Give evidence that the drop-off ddPCR is a valid new molecular tool for detecting IDH2 mutations Techniques such as ddPCR, NGS, and qPCR started to be implemented	MRD	Grassi et al., 2020 [53]
CEBPA mutations and blast cells	RT-qPCR	n = 4 (n = 3 diagnostic, n = 1 relapse)	RT-qPCR can achieve a sensitivity of 10−4–10−6 Describe specific RT-qPCR for CEBPA mutations	Concordance	Smith et al., 2006 [56]
NPM1 mutations	Flow cytometry, RT-qPCR	n = 15 patients n = 45 MRD samples Diagnosis and follow-up	Knowledge of RT-qPCR-based MRD results RT-qPCR has a higher sensitivity than FC (10−4–10−6 vs. 10−3–10−5)	MRD	Pettersson et al., 2016 [51]
Somatic mutations	ddPCR, RT-qPCR	n = 41 patients with AML-M1/M2 n = 20 healthy volunteers	ddPCR can achieve a sensitivity of 10−4–10−6 Identify genes that contribute to leukemogenesis	Concordance	Handschuh et al., 2017 [52]
Residual leukemic cells	Flow cytometry	n = 135 patients with de novo AML (100 achieving CR after intensive chemotherapy)	Flow cytometry achieves a sensitivity of 10−4 MRD patients have a five-year RFS >70%, MRD+ patients have the worst prognosis Incorporate MRD assessment in the protocols for the treatment of AML	MRD Response assessment	Buccisano et al., 2006 [55]
Somatic mutation	NGS	n = 22 (after remission) post-treatment	cfDNA and BM were complementary in the follow-up and monitoring of disease The concordance of the VAF assessment by both methods was high (R2 = 0.849)	Concordance MRD	Short et al., 2020 [67]
Driver mutations	NGS dPCR	n = 53 (n = 37 AML, n = 14 MDS) (after post-alloSCT) Diagnosis and post-treatment	An increase in ctDNA in relapsed patients ctDNA has concordance with BM testing, having a comparable utility	Concordance MRD Response assessment	Nakamura et al., 2019 [68]
Somatic mutation	DDO-ddPCR dPCR qPCR NGS	n = 57 samples (cfDNA), n = 28 (PB), n = 53 (BM) Post-treatment	Demonstrate the application of DDO-ddPCR in comparison to dPCR and qPCR DDO-ddPCR has a sensitivity of 0.037% cfDNA opens new strategies for response assessment, disease monitoring, and molecular profiling of MRD	Concordance MRD	Rausch et al., 2021 [69]
Leukemic cells	Flow cytometry	n = 50 patients with de novo AML	BM and PB samples were significantly concordant (r = 0.86 and 0.82, respectively, p < 0.001) The cut-off value of residual leukemic cells was 1.5 × 10−4 PB MRD was found to have a significant effect on relapse-free survival (p = 0.036).	Concordance MRD	Maurillo et al., 2007 [54]
Primitive blast (CD34+/CD117+/CD133+)	Flow cytometry	n = 114 patients (205 paired BM and PB samples) Diagnosis and post-treatment	Primitive blast frequency was lower in the PB; PB MRD is more specific than BM The role of MRD in the PB may have an essential use in the future clinical management of AML patients Flow cytometry has a sensitivity of 10−3–10−5	Concordance MRD Response assessment	Zeijlemaker et al., 2016 [59]

Ref., reference; ddPCR, droplet digital PCR; NGS, next-generation sequencing; qPCR, quantitative PCR; MRD, minimal residual disease; RT-qPCR, real-time quantitative PCR; FC, flow cytometry; CR, complete response; RFS, relapse-free survival; dPCR, digital PCR; BM, bone marrow; AML, acute myeloid leukemia; MDS, myelodysplastic syndrome; SCT, stem cell transplantation; VAF, variant allele frequency; DDO-ddPCR, D-aspartate oxidase ddPCR, droplet digital PCR; PB, peripheral blood; cfDNA, cell-free DNA.