INTRODUCTION — Induction therapy in the acute leukemias aims to reduce the total body leukemic cell population from approximately 1012 to below the cytologically detectable level of about 109 cells. It is generally assumed, however, that a substantial burden of leukemia cells persists undetected (ie, "measurable residual disease," formerly referred to as "minimal residual disease"), leading to relapse within a few weeks or months if no further post-remission therapy (ie, additional consolidation chemotherapy) were administered. However, even when an adult patient is in complete remission (CR) following additional induction and/or consolidation therapy, the majority will ultimately relapse, indicating that attainment of CR, as defined below, is not sufficient to guarantee long-term remission and/or "cure." (See "Induction therapy for acute myeloid leukemia in medically-fit adults".)
This review will discuss the subject of measurable residual disease (MRD) in patients who have been treated for acute myeloid leukemia (AML) and are in complete remission. Standard morphologic and cytologic methods for diagnosing de novo or relapsed leukemia are discussed separately, although their limitations are briefly reviewed here. (See "Clinical manifestations, pathologic features, and diagnosis of acute myeloid leukemia".)
DEFINITION OF COMPLETE REMISSION (CR) — CR in AML has been defined using the following criteria developed by an International Working Group (table 1) [1-3]:
●Normal values for absolute neutrophil count (>1000/microL) and platelet count (>100,000/microL), and independence from red cell transfusion.
●A bone marrow biopsy that reveals no clusters or collections of blast cells. Extramedullary leukemia (eg, central nervous system or soft tissue involvement) must be absent.
●A bone marrow aspiration reveals normal maturation of all cellular components (ie, erythrocytic, granulocytic, and megakaryocytic series). There is no requirement for bone marrow cellularity.
●Less than 5 percent blast cells are present in the bone marrow, and none can have a leukemic phenotype (eg, Auer rods). The persistence of dysplasia is worrisome as an indicator of residual AML but has not been validated as a criterion for remission status.
●The absence of a previously detected clonal cytogenetic abnormality (ie, complete cytogenetic remission, CRc) confirms the morphologic diagnosis of CR but is not currently a required criterion. However, conversion from an abnormal to a normal karyotype at the time of first CR is an important prognostic indicator, supporting the use of CRc as a criterion for CR in AML [2,4,5].
Other responses — Some patients may fulfill all of the above criteria for CR but may not recover peripheral blood counts to the required level. These are denoted as CRi, or CR with insufficient hematological recovery (platelets or neutrophils). CRp describes a subset of patients with CRi, wherein patients fulfill all criteria for CR except that platelet counts are <100,000/microL [2,6]. The survival and relapse rates for patients with CRp appear to be worse than those with a CR, but better than those with a partial remission (PR) [7].
Patients who fail to achieve CR or CRi may experience a PR, defined as a ≥50 percent decrease in bone marrow blasts with normalization of peripheral blood counts, or some other measure of hematologic improvement. A PR in AML is generally expected to be of short duration, and, in most circumstances, is unlikely to serve as a surrogate reasonably likely to predict for clinical benefit [6].
As will be seen below, techniques employing the polymerase chain reaction (PCR) and multiparametric flow cytometry are significantly more sensitive for detecting the presence or absence of residual disease (ie, complete molecular remission, CRm). However, with the exception of acute promyelocytic leukemia, the clinical utility of this assay is uncertain, since not all patients with PCR evidence for residual disease will relapse. (See "Clinical manifestations, pathologic features, and diagnosis of acute promyelocytic leukemia in adults", section on 'Genetic features'.)
CR AS DEFINED BY LONG-TERM SURVIVAL — There is no clear guideline for how often bone marrow biopsies should be performed in patients who obtain a complete clinical remission (CR) after therapy. Many patients treated on protocol have a bone marrow aspiration repeated every two to three months during the first year after attainment of CR and every three to six months for the next two years. In fact, most relapses are signaled by a decline in the platelet count to abnormal levels or the appearance of leukemic blasts in the peripheral blood smear. In general, for non-protocol patients, a single bone marrow exam performed two to three months after the completion of treatment is likely sufficient as long as the complete blood count remains normal.
The risk of relapse is highest during the first two years after the end of consolidation chemotherapy [8-11]. It is relatively uncommon to have a late relapse.
●Using data from nine trials initiated by the Cancer and Leukemia Group B (CALGB) between 1974 and 1992, hazard rates for death and relapse were found to be greatest in the first year, decreased substantially between years 1 and 2, and decreased further between years 2 and 3 [9]. Rates of death and relapse were quite low after 3 to 4 years. It was concluded that patients with AML who are in complete remission for 3 to 4 years can be reassured that late relapse and death are relatively uncommon events.
●A single-institution retrospective study looked at the long-term outcomes in 1069 consecutive patients with AML in first complete remission from 1991 to 2003 [10]. Relapse rates decreased as the patients were further from treatment with rates of 69, 38, 17, 8, and 7 percent during years one through five, respectively.
●Among patients with AML who received hematopoietic cell transplantation and were in CR two years later, the overall chance of being alive in CR at nine years was 82 percent [11]. The latest relapses in this group occurred at four to seven years.
DIFFICULTIES IN DEFINING CR — Several observations illustrate the problems involved in ascertaining whether a patient in complete remission (CR) is destined to remain clinically disease-free:
Sampling error — The sampling error of a single bone marrow examination is immense, as the examined specimen represents only a very small percentage of the total bone marrow cellular population. Many documented cases exist in which a bone marrow aspiration was normal at one site and showed leukemia at another [12].
Sensitivity of routine tests — Standard definitions of CR in acute leukemia require that blasts constitute less than 5 percent of the total nonerythroid cells within the bone marrow. Therefore, up to 5000 blasts/100,000 nonerythroid cells may still be present in a patient in clinical CR. Other testing limitations include:
●The sensitivity of cytogenetics for detecting residual disease is limited to about 5 percent when only 20 to 40 metaphase cells are examined.
●In one study, the median lower limit of detection of leukemic blasts by conventional microscopic examination of bone marrow by two experienced cytologists was greater than 1 percent. That is, even experienced observers have difficulty detecting fewer than 1000 leukemic blasts/100,000 nucleated cells [13].
●It is common experience that seeing a single myeloblast with an Auer rod in a remission bone marrow examination is a harbinger of persistent AML. Since most bone marrow examinations review fewer than 1000 nucleated cells, even a single observable blast cell represents approximately 100 blasts/100,000 nucleated cells.
IDENTIFYING MEASURABLE RESIDUAL DISEASE — Methods for evaluating measurable residual disease (MRD) in AML are evolving and await standardization [14]. The most commonly used methods for monitoring measurable residual disease (MRD) in AML include quantitative polymerase chain reaction (Q-PCR) and multiparameter flow cytometry (MFC); some studies have used next-generation sequencing (NGS) to detect gene mutations for MRD analysis [14-17]. Each method has associated benefits and complications [14,18-21], and with each technique the detection of abnormal cells is not, in itself, an indication that relapse is imminent [22,23]. As examples, such cells might be incapable of division or proliferation, held in check by the patient's immune system, or may reflect another disorder, such as clonal hematopoiesis of indeterminate potential (CHIP) [14,24,25].
Numerous technical and practical areas of uncertainty remain regarding MRD monitoring in AML [21]:
●Type of sample – Peripheral blood is easy to obtain and lacks immature normal populations of cells that may interfere with analysis by MFC. However, the sensitivity of MRD analysis of the peripheral blood is lower than that of bone marrow and requires a lower MRD threshold to be prognostic.
●Sample size
●Timing (eg, during induction, after induction, after consolidation)
●Sampling interval (eg, every three months)
●Length of screening
Of importance, variables that affect the ability of MRD monitoring to have an impact on patient outcome differ by AML genetic subtype [26]. The ideal timing as well as sampling interval for MRD monitoring and the length of screening for relapse is unknown and likely differs depending upon the molecular marker used. As an example, AML clones with certain molecular lesions are known to replicate at faster or slower rates than others, resulting in a wide range of elapsed time between the emergence of a positive MRD sample and hematologic relapse. To be effective, a genetic subtype with a short time period between the emergence of MRD and hematologic relapse may require more frequent MRD monitoring than a subtype with a slower progression. Similarly, a genetic subtype that progresses more slowly may require more prolonged MRD monitoring. In addition, the rate of fusion transcript degradation varies by molecular marker.
While MRD assessment appears to have prognostic value, it is unknown whether MRD assessment will have therapeutic consequences that will improve long-term outcomes. These and other logistical aspects to MRD monitoring must be clarified and validated in rigorous clinical trials before MRD monitoring can become part of the routine follow-up of all patients with AML.
Polymerase chain reaction — Real time quantitative polymerase chain reaction (Q-PCR) can be used to look for fusion gene transcripts (eg, BCR-ABL1), gene mutations (eg, FLT3), or aberrantly expressed genes (eg, WT1). The leukemic cells must express (or overexpress) a gene not seen in normal cells. Q-PCR is then used to compare mRNA expression of these targeted genes with that of reference housekeeping genes. Q-PCR is a potential technique for measuring residual disease in more than half of patients with AML.
Currently available PCR techniques are able to detect approximately one leukemia cell diluted 105 to 106 times, or one blast or less per 100,000 nucleated cells [6,13,27,28]. As an example, one study that used PCR to estimate the numbers of residual leukemic cells found that the relapse rates for children with <15 or >15 blasts/100,000 mononuclear cells after the end of induction and consolidation therapy were 4 and 47 percent, respectively [13]. Details on the procedure of PCR are discussed separately. (See "General aspects of cytogenetic analysis in hematologic malignancies".)
The above calculations indicate that routine microscopic determination of clinical complete remission status (<5000 blasts/100,000 cells) allows for a residual blast count up to 300 times greater than the number associated with leukemic relapse (>15 blasts/100,000 cells). Even the chance observation of a single blast cell containing an Auer rod (see above) indicates a residual blast count approximately seven times that associated with ultimate relapse. These observations would lead one to conclude that "true" complete remissions must be determined by techniques at least as sensitive as PCR (table 2). (See "Detection of measurable residual disease in acute lymphoblastic leukemia/lymphoblastic lymphoma".)
The following are some known limitations to the use of PCR for the detection of MRD [18,29]:
●There is no standardized method for Q-PCR measurement of MRD across laboratories and, as such, values cannot be compared between different laboratories.
●PCR techniques require that the genome of the leukemic clone is sufficiently different from the normal clone in order to prepare the appropriate primers and probes. This may be relatively straightforward for common recurring translocations in AML, such as t(15;17), and inv(16) or, in ALL, clonal rearrangements of the T-cell receptor or immunoglobulin genes [13,27,28,30,31], but would not be possible in the absence of detectable genetic alterations unless the leukemic cells express (or overexpress) a gene not seen in normal cells [32,33].
●As a result of chemotherapy or clonal evolution, a genetic subclone might develop with a genetic makeup different from the primary leukemic clone [34]. This clone might expand, leading to a leukemic "relapse" but would not be detected with the same PCR primer pairs and probes used for the original clone.
These markers require rigorous clinical validation before they are used clinically. This validation would help assure that marker positivity would truly reflect an increased risk of relapse.
Multiparameter flow cytometry — Multiparameter flow cytometry (MFC) can be used at diagnosis to identify leukemia-associated aberrant immunophenotypes (LAIPs) which are present on malignant cells in the bone marrow but not found on the patient's normal cells. MFC can be performed in about 80 percent of cases of AML, with a sensitivity of about one blast cell in 103 to 104 nucleated cells, depending on the antigenic marker [6].
The following are some limitations to the use of MFC for detecting MRD [29]:
●Multiple LAIPs have been identified in patients with AML, but each individual LAIP may only be present in <5 percent of patients thereby requiring an extensive panel at diagnosis in order to identify an appropriate LAIP for the patient.
●Malignant cells within an individual patient may display phenotypic heterogeneity such that some leukemic cells may not express the LAIP and therefore go undetected by this method.
●In a minority of cases, the LAIP expressed by the malignant cells may change from diagnosis to relapse.
SIGNIFICANCE OF RESIDUAL DISEASE
Detected by PCR — The long-term outcome of patients with continuous morphologic complete remission but who still have tumor cells, as determined by reverse transcriptase PCR (RT-PCR), is not known. Serial quantitative measurements may be helpful; finding increasing values for measurable residual disease (MRD) at various time points in the course of the disease may identify patients at higher risk of relapse [13,27,28,35-40].
Use in APL — The prognostic importance of MRD has been noted in studies that performed serial RT-PCR studies in patients undergoing treatment for newly-diagnosed acute promyelocytic leukemia (APL) [41-44]. As examples:
●The largest study followed serial real time quantitative PCR (Q-PCR) assays every three months from peripheral blood and bone marrow samples of 406 patients with newly diagnosed APL undergoing induction therapy [41]. APL persisted or recurred at the molecular or clinical level in 30 patients. Of the 20 patients with frank relapse, 11 demonstrated Q-PCR positivity in samples collected a median of 74 days before. Of the nine patients with frank relapse not detected by Q-PCR, seven had been non-compliant with their Q-PCR monitoring schedule.
●In another study, 40 of 47 patients (85 percent) who were induced by ATRA alone had residual disease detected by RT-PCR [42]. However, after three cycles of consolidation therapy with idarubicin and cytarabine, residual disease was found in only 10 percent. APL relapsed in only 3 of 41 patients (7 percent) who had two or more negative RT-PCR assays for MRD compared with all four patients with two or more positive results.
●A third study of 123 patients found that patients having a ratio of PML-RARA mRNA to that of the housekeeping enzyme GAPDH greater than 10-5 following consolidation therapy were at a fourfold increased risk for relapse compared with those with a ratio of less than 10-5 [43]. However, 73 percent of patients who experienced relapse had ratios less than 10-5.
Monitoring of MRD is routinely incorporated into the post-consolidation care of patients with APL. Such monitoring is only useful if done frequently with a planned intervention as soon as a positive signal is seen [45]. (See "Initial treatment of acute promyelocytic leukemia in adults", section on 'Monitoring response during maintenance'.)
Use in AML — The role of MRD monitoring in AML is evolving. Detection of MRD has prognostic value for certain categories of AML (eg, mutated NPM1 or RUNX) and prior to allogeneic transplantation. However, because of current challenges in methodology and standardization, at present we suggest not using MRD assessment as the sole determinant of treatment decisions for AML.
Use of MRD for treatment decisions in AML is currently limited because of several issues. Although MRD positivity at the time of clinical complete remission (CR) is associated with higher relapse rates and inferior survival, not all patients with MRD positivity will relapse clinically and some patients will relapse despite negative MRD results. There is uncertainty regarding the best time to measure MRD and the preferred source material (ie, blood versus bone marrow). Measurement immediately after achieving CR can provide information about the bulk of malignant cells, but may not detect minor subclones of potentially treatment-resistant residual leukemic cells. There are other practical issues that complicate use of MRD for patients with AML. MRD assays for each individual marker used for MRD assessment needs rigorous validation and standardization. MRD analysis may also be confounded by the presence of mutations (eg, DNMT3A, TET2, ASXL1) that are associated with age-related clonal hematopoiesis, rather than residual leukemic cells [46-48]. Use of MRD prior to hematopoietic cell transplantation (HCT) in adults with AML is discussed separately. (See "Post-remission therapy for acute myeloid leukemia in younger adults", section on 'NMA/RIC versus MAC regimens'.)
Informative studies of MRD assessment of NPM1 include:
●RUNX1-RUNX1T1 – MRD monitoring by Q-PCR was useful for predicting relapse and overall survival (OS) in a multicenter study of 155 patients with RUNX1-RUNX1T1 rearranged AML [49]. After treatment with intensive remission induction therapy, achievement of MR2.5 (>2.5 log reduction) after cycle 1 and MR3.0 (>3.0 log reduction) after cycle 2 was associated with a reduced risk of relapse. Achievement of MRD negativity (>6.0 log reduction) in either peripheral blood or bone marrow was an independent favorable prognostic factor for OS and cumulative rate of relapse at four years. Serial MRD assessment from peripheral blood identified more than three-quarters of the patients who ultimately relapsed.
●NPM1-mutated AML – Among patients with NPM1-mutated AML treated as part of the National Cancer Research Institute AML17 trial, when compared with MRD negative cases, those with persistence of MRD at the time of first CR had higher relapse rates (82 versus 30 percent at three years; hazard ratio for relapse 4.8) and inferior OS (24 versus 75 percent at three years; hazard ratio for death 4.4) [50]. Assessment of MRD at other time points did not provide additional prognostic value above that seen with MRD assessment at first CR, nor did the coexistence of mutations usually associated with poor prognosis (eg, FLT3-ITD and DNMT3A).
Similarly, among 152 patients with NPM1-mutated AML who were treated in the Acute Leukemia French Association 0702 trial, those who did not achieve a four-log reduction in NPM1 had a higher cumulative incidence of relapse (hazard ratio 5.8) and shorter OS (hazard ratio 11) compared with those who did reach that benchmark [51].
The following examples illustrate the complexity of incorporating MRD assessment into practice:
●Some patients with t(8;21) who have been treated for AML have had positive evidence for the AML1-ETO (AML1-MTG8) fusion gene by RT-PCR for as long as eight years without relapsing [23,52]. On the other hand, patients who become RT-PCR negative have a low rate of relapse and are probably cured [22,53,54].
●In a study of 45 patients with DNMT3A positive AML, 14 patients had detectable DNMT3A mutations in remission samples up to eight years after the initial AML diagnosis, without other molecular markers of AML [55]. This suggests that the presence of DNMT3A positive clones does not eliminate the possibility of cure.
●In a study of 424 patients with AML undergoing transplantation, 93 patients (22 percent) had informative discordant results from concurrent cytogenetics and flow cytometry testing (ie, negative cytogenetics with positive flow or positive cytogenetics with negative flow) either before or after transplant [56]. Detection of residual disease by either method at either time was associated with a higher relapse rate and worse survival.
Next-generation sequencing (NGS) has been used for detection of MRD, but its variable sensitivity adds further questions regarding interpretation of findings. Examples include:
●In a study of nearly 500 patients with newly diagnosed AML, NGS detected at least one mutation in >89 percent of patients at diagnosis, with allele frequencies that ranged from 0.02 to 47 percent [48]. After treatment, detection of mutations associated with clonal hematopoiesis (eg, DNMT3A, TET2, ASXL1; collectively referred to as DTA mutations) did not correlate with an increased relapse rate. However, after excluding DTA mutations, multivariate analysis indicated that detection of MRD (versus no detectable MRD) was associated with a higher rate of relapse (55 versus 32 percent, respectively) and lower rates of relapse-free survival (37 versus 58 percent) and OS (42 versus 66 percent). Detection of MRD by NGS added significant prognostic value versus flow cytometry.
●Another study used targeted deep sequencing to assess mutation patterns after induction therapy and found that the detection of persistent leukemia-associated mutations in at least 5 percent of bone marrow cells in day 30 remission samples was associated with an increased risk of relapse and reduced overall survival [57].
Detected by flow cytometry — As discussed above, it is also possible to monitor MRD in children and adults with AML through multiparameter flow cytometry (MFC) [29,58-68]. The cut-off values in these studies have been derived from retrospective analysis, and appropriate prospective validation to confirm their results is in early stages.
The following studies illustrate the clinical relevance of this approach:
●In one report, the sensitivity of this technique was similar to that noted for PCR, namely one leukemic cell diluted 104 to 105 times [61]. Using this methodology in 34 patients with AML, a level of MRD greater than or less than 3.5 x 10-4 following consolidation therapy was associated with relapse rates of 77 and 17 percent, respectively (p<0.001). MRD above this cut-off value was also significantly associated with a multidrug resistance 1 (MDR1) phenotype and intermediate or unfavorable cytogenetics [58]. (See "Clinical manifestations, pathologic features, and diagnosis of acute myeloid leukemia".)
●A post-hoc analysis applied an MRD cutoff value of >0.1 percent to a prospective cohort of 241 patients with AML in first morphologic complete remission enrolled on the HOVON/SAKK AML 42A study [69]. Among the 183 patients with MRD assessment after induction cycle 2, 42 (23 percent) were MRD positive by flow cytometry. When compared with those in whom MRD was ≤0.1 percent, patients with MRD had a higher cumulative incidence of relapse (72 versus 42 percent; hazard ratio 2.60; 95% CI 1.49-4.55) and lower relapse free survival (23 versus 52 percent) at four years.
●In another study, MRD was evaluated by flow cytometry in the first bone marrow of 126 patients with AML and aberrant phenotypes following attainment of complete remission (CR) after induction therapy [62]. The cumulative rates of relapse were 0, 14, 50, and 84 percent for those with levels of MRD less than 10-4, between 10-4 and 10-3, between 10-3 and 10-2, and >10-2, respectively. Multivariate analysis indicated that MRD was the most powerful independent prognostic factor, followed by cytogenetics and the number of cycles needed to achieve CR.
RELAPSED DISEASE — We consider patients to have relapsed disease if, after attainment of complete molecular remission, subsequent analyses confirm the loss of such molecular remission. Relapse occurs in 5 to 10 percent of patients with APL and in about 10 to 15 percent of those with high-risk APL. Attempts have been made to identify patients during remission who are at highest risk of relapse using reverse transcriptase-polymerase chain reaction (PCR) analysis of the PML/RAR-alpha fusion gene derived from the t(15;17) translocation. (See "Treatment of relapsed or refractory acute myeloid leukemia", section on 'Definitions of refractory disease and relapse'.)
A multicenter Italian trial prospectively monitored PCR positivity in 163 patients with APL [70]. All patients were in hematologic remission and tested PCR-negative at the end of consolidation. Of 21 who later converted to PCR-positivity, 20 underwent relapse at a median time of three months from the first PCR-positive result. Of 142 who were persistently PCR-negative, only eight had hematologic relapse after a median follow-up of 18 months. It is not yet known if early detection will translate into improved treatment of relapsed disease.
The mechanisms responsible for relapse are unclear, but at least two causes of acquired ATRA resistance have been proposed:
●Insufficient plasma ATRA concentration may be caused by increased oxidative catabolism of ATRA by cytochrome P450 enzyme activity, increases in cellular retinoic acid binding protein, or the multidrug resistance (MDR) glycoprotein [71-74].
●In a subset of patients with relapse and ATRA resistance, missense mutations have been found in the RAR-alpha region of the PML/RAR-alpha gene, which may interfere with the molecular interaction with ATRA [74,75].
SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Acute myeloid leukemia".)
INFORMATION FOR PATIENTS — UpToDate offers two types of patient education materials, "The Basics" and "Beyond the Basics." The Basics patient education pieces are written in plain language, at the 5th to 6th grade reading level, and they answer the four or five key questions a patient might have about a given condition. These articles are best for patients who want a general overview and who prefer short, easy-to-read materials. Beyond the Basics patient education pieces are longer, more sophisticated, and more detailed. These articles are written at the 10th to 12th grade reading level and are best for patients who want in-depth information and are comfortable with some medical jargon.
Here are the patient education articles that are relevant to this topic. We encourage you to print or e-mail these topics to your patients. (You can also locate patient education articles on a variety of subjects by searching on "patient education" and the keyword(s) of interest.)
●Beyond the Basics topics (see "Patient education: Acute myeloid leukemia (AML) treatment in adults (Beyond the Basics)")
SUMMARY
●Induction therapy for acute myeloid leukemia (AML) aims to achieve a complete remission (CR) in which the total body leukemic cell population is reduced from approximately 1012 to below the cytologically detectable level of about 109 cells. Thus, it is generally assumed that a substantial burden of leukemia cells persists undetected (ie, "measurable residual disease"; MRD; formerly called "minimal residual disease"), leading to relapse within a few weeks or months if no further post-remission therapy (ie, additional consolidation chemotherapy) were administered.
●CR is defined by International Working Group criteria using morphologic and clinical data (table 1). (See 'Definition of complete remission (CR)' above.)
●Ascertaining whether a patient in CR is destined to remain clinically disease-free after post-remission therapy is limited by the inherent insensitivity of routine tests on the bone marrow for detecting residual leukemia and the likelihood that the small area of bone marrow examined does not reflect the much larger bone marrow compartment. (See 'Difficulties in defining CR' above.)
●Techniques employing quantitative polymerase chain reaction (PCR) and multiparameter flow cytometry techniques are significantly more sensitive for detecting MRD than morphology or cytogenetics. Next-generation sequencing (NGS) of DNA is an emerging technology for detection of MRD. Each method has associated benefits and complications. It is not yet clear how best to intervene with alternative therapy to improve outcomes in those patients with positive MRD. (See 'Identifying measurable residual disease' above.)
●PCR is routinely used to confirm a CR in patients with acute promyelocytic leukemia (APL). Once a molecular CR is achieved, many clinicians follow patients periodically with PCR for the PML-RARA fusion transcript in order to monitor for relapse. Such monitoring is likely only to be useful if done frequently with a planned intervention as soon as a positive signal is seen. For APL, we consider patients to have relapsed if, after attainment of complete molecular remission, subsequent analyses confirm the loss of such molecular remission. (See 'Use in APL' above and "Initial treatment of acute promyelocytic leukemia in adults", section on 'Monitoring response during maintenance'.)
●The use of MRD monitoring in patients with AML is less clear. At present, decisions to continue, alter, or resume therapy cannot be based solely upon a positive signal with RT-PCR or NGS. The cut-off values examined in the published literature have been derived from retrospective analyses, and appropriate prospective validation is needed to confirm these results. The detection of cells from the leukemic clone is not, in itself, an indication that relapse is imminent. Such cells might be incapable of division or proliferation, held in check by the patient's immune system, or may reflect another disorder, such as clonal hematopoiesis of indeterminate potential (CHIP). (See 'Use in AML' above.)
