Introduction and background

Leukemia encompasses a heterogeneous group of hematologic malignancies characterized by uncontrolled proliferation of abnormal hematopoietic progenitors. Despite the high morphological complete remission rates achieved following standard induction chemotherapy, disease relapse is the primary cause of treatment failure and mortality in both acute lymphoblastic leukemia (ALL) and acute myeloid leukemia (AML) [1,2]. The detection of measurable/minimal residual disease (MRD), defined as the persistence of submicroscopic levels of leukemic cells below the threshold of conventional morphological assessment, has shifted the clinical paradigm in onco-hematology. MRD is established as the most powerful, independent prognostic biomarker for predicting relapse-free and overall survival [1,2], and its robust predictive value has led to its integration into routine clinical practice for risk stratification, guiding crucial therapeutic decisions such as the (de)-intensification of chemotherapy, selection of conditioning regimens for allogeneic hematopoietic stem cell transplantation (HSCT), and deployment of targeted immunotherapies [2,3].

Multiparameter flow cytometry (MFC) and polymerase chain reaction (PCR)-based assays are the principal modalities for MRD monitoring [1]. MFC identifies residual leukemic blasts by detecting leukemia-associated immunophenotypes (LAIPs) or different-from-normal (DfN) antigenic profiles [4,5]. The primary advantages of MFC lie in its broad applicability, being useful in over 90% of patients regardless of their underlying molecular subtype, along with its rapid turnaround time and capacity to assess cell viability [1,6]. However, MFC is limited by a lower analytical sensitivity (10-4 to 10-5), the necessity for fresh marrow samples, and susceptibility to both false-negative results driven by immunophenotypic shifts and false-positive results stemming from regenerating normal hematopoietic progenitors [5,7].

PCR-based techniques, including real-time quantitative PCR (RQ-PCR) and reverse-transcription quantitative PCR (RT-qPCR), offer superior analytical sensitivity, detecting one leukemic cell in 105 to 106 normal cells [1,2]. In ALL, PCR targets patient-specific clonal immunoglobulin (Ig) and T-cell receptor (TCR) gene rearrangements, whereas in AML, it targets recurrent genetic fusions (e.g. RUNX1::RUNX1T1, CBFB::MYH11) or specific mutations (e.g. NPM1 or WT1 overexpression) [1,4,6]. Despite its high sensitivity and rigorous international standardization, PCR-based MRD assessment is time-consuming, expensive, and limited in its applicability. Suitable molecular targets are absent in approximately 50% of AML cases, and Ig/TCR monitoring in ALL requires labour-intensive development of patient-specific primers [1,2]. Furthermore, the clinical specificity of PCR can be confounded by the amplification of DNA from non-viable apoptotic cells post-treatment or the harmless persistence of pre-leukemic clones (e.g. clonal hematopoiesis) that lack immediate relapse potential [2,8].

Given their distinct biological targets and methodological thresholds, MFC and PCR yield discordant results in clinical settings. The concordance between the two assays is variable and often time-point-dependent, influenced by the leukemia subtype, specific molecular targets, and phase of treatment [7,8]. For instance, isolated PCR positivity (MFC-/PCR+) may reflect the superior analytical sensitivity of molecular techniques or the persistence of non-clonogenic transcripts without clinical relevance [3,8], whereas isolated MFC positivity (MFC+/PCR-) may arise from the loss of molecular targets due to clonal evolution or limitations in PCR primer design [3,4,8]. Previous studies have demonstrated that integrating both methods can refine prognostic stratification; for example, combining MFC with WT1 or NPM1 PCR assessment has been shown to more accurately predict post-transplant relapse in AML [3,4]. However, the precise clinical and prognostic significance of these discordant states remains a subject of debate, creating challenges for clinicians attempting to tailor interventions [3,7].

Although numerous primary studies have investigated the use of MFC and PCR in MRD assessment, evidence remains fragmented across disparate clinical settings, pediatric and adult cohorts, and varying therapeutic time points. There is a need to synthesize this literature to determine which modality or combination thereof provides superior diagnostic performance and prognostic information. Therefore, this systematic review and meta-analysis aimed to evaluate and compare the diagnostic accuracy and prognostic value of MFC versus PCR-based MRD assessment in patients with leukemia. This study seeks to inform evidence-based clinical decision-making, optimize MRD-guided management strategies, and highlight areas requiring methodological harmonization in precision hematologic oncology by clarifying the clinical impact of method-specific MRD detection and exploring the implications of inter-method discrepancies.