Detection of measurable residual disease in acute lymphoblastic leukemia/lymphoblastic lymphoma

INTRODUCTION — Most patients with acute lymphoblastic leukemia (ALL)/lymphoblastic lymphoma (LBL) achieve complete remission (CR; ie, <5 percent blasts) and many experience prolonged disease-free survival/cure. Nevertheless, a minority of patients experience a recurrence and die from the disease. Relapse usually results from residual leukemic cells that remain following the achievement of CR, but that are below the limits of detection using conventional morphologic assessment. These subclinical levels of residual leukemia are termed measurable residual disease (MRD; formerly called minimal residual disease) and can be evaluated using techniques that are more sensitive than conventional microscopy.

MRD measurements during and after induction therapy and/or after consolidation therapy are highly prognostic, correlate with relapse rates, and are routinely used in the clinical care of patients with ALL/LBL.

This topic reviews methods for MRD monitoring in ALL/LBL.

The clinical significance of MRD detection in ALL/LBL and the use of MRD testing in acute myeloid leukemia and chronic myeloid leukemia are discussed separately. (See "Clinical use of measurable residual disease detection in acute lymphoblastic leukemia" and "Overview of the treatment of chronic myeloid leukemia", section on 'Monitoring response' and "Acute myeloid leukemia: Induction therapy in medically fit adults", section on 'Introduction'.)

DIFFICULTIES IN DEFINING COMPLETE REMISSION — The primary goal of induction therapy for acute lymphoblastic leukemia (ALL)/lymphoblastic lymphoma (LBL) is achievement of an initial complete remission (CR), defined as the eradication of visible leukemia cells (at least to <5 percent blasts) from the bone marrow and blood by microscopic review and the restoration of normal hematopoiesis (typically with >25 percent cellularity and normal peripheral blood counts).

While CR has historically been defined based on morphologic criteria, an assessment of MRD can define a more stringent response that is better able to predict prognosis. This approach is supported by several observations that illustrate the difficulty in ascertaining whether a patient with ALL/LBL in morphologic CR is likely to remain disease-free:

●Hematogones – Hematogones are benign lymphoid precursors. Morphologic evaluation may be unable to distinguish between ALL/LBL blast cells, activated mature lymphocytes, and hematogones. This distinction is particularly difficult in samples of bone marrow recovering from chemotherapy or transplantation, where hematogones may account for 10 percent of the lymphoid cells. Most of the time, hematogones can be distinguished from leukemia cells by their immunophenotype [1,2].

●Sampling error – A single bone marrow specimen represents only a very small percentage of the total bone marrow cellular population and thereby presents a potential for sampling error. Documented cases exist in which a bone marrow aspiration was normal at one site and showed leukemia at another. While this may occasionally be true early in the course of the disease, it is less of a problem after chemotherapy has been given. (See 'Source of cells' below.)

●Detection limits – The detection of blasts by morphologic review or conventional cytogenetics is limited by operator error and the number of metaphases evaluated. 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 [3]. The sensitivity of cytogenetics for detecting residual disease is limited to approximately 5 percent when only 20 to 40 metaphase cells are examined.

Accordingly, a patient in clinical remission can harbor as many as 10¹⁰ leukemic cells, equivalent to a detection limit of about one malignant cell for every 20 to 100 normal cells [4]. On the other hand, more sensitive assays, such as polymerase chain reaction (PCR), may detect one leukemic cell diluted 10⁵ times, or approximately one to five blasts per 100,000 nucleated cells [3].

The problem of defining CR is best exemplified by a study in which residual leukemic cells were monitored by PCR [3]. The risk of relapse of ALL/LBL for children with <15 or >15 blasts per 100,000 mononuclear cells after both induction therapy and consolidation therapy was 4 and 47 percent, respectively. These calculations indicate that routine microscopic and cytogenetic determination of CR (ie, one blast in 20 normal cells) allows for a residual blast count up to 300 times greater than the number capable of causing leukemic relapse (ie, >15 blasts/100,000 cells). This observation suggests that "true" remission must be determined by means at least as sensitive as PCR (table 1).

SOURCE OF CELLS — MRD analyses can be performed on bone marrow aspirate specimens or peripheral blood. Bone marrow is preferred, especially in cases of B-lineage ALL/LBL. The amount of tissue needed differs according to the tissue source and testing used, and clinicians should contact the referral laboratory for details. A 2 to 5 mL sample from the first bone marrow aspirate usually provides sufficient numbers of bone marrow cells to achieve adequate sensitivity [5]. Larger samples and those from subsequent aspirates at the same spot may result in erroneously low MRD measurement due to dilution by peripheral blood. MRD analyses are performed using a single bone marrow aspirate, as blasts are thought to be homogeneously distributed throughout the marrow [6].

For T lineage ALL, peripheral blood and bone marrow may yield similar MRD assessments. Peripheral blood is not a sensitive measure of MRD in B-lineage ALL. In studies using paired blood and bone marrow samples, MRD levels were higher in the bone marrow and the difference between bone marrow and blood MRD levels was greater for B lineage ALL/LBL than for T lineage ALL/LBL [7-9]:

●T lineage ALL/LBL – Blood MRD levels were comparable to or up to 1 log lower than bone marrow MRD levels.

●B lineage ALL/LBL – Blood MRD levels were 1 to 3 logs lower than bone marrow MRD levels.

As an example, flow cytometry was used to measure MRD in paired samples of bone marrow and peripheral blood collected from 226 children at several time points during treatment for newly diagnosed ALL/LBL [7]. MRD was detected in marrow and blood in 72 pairs, and in marrow but not in blood in 67 pairs; it was undetectable in the remaining 579 pairs. Findings in marrow and blood were completely concordant in the 150 paired samples from patients with T-lineage ALL. In B-lineage ALL/LBL, only 37 of 104 positive marrow samples had a corresponding positive blood sample. The four-year cumulative incidence of relapse in B-lineage ALL/LBL was 80 percent for those who had peripheral blood MRD at the end of remission induction therapy, but only 13 percent for those with MRD confined to the marrow.

METHODS FOR DETECTING MRD — A variety of techniques have been studied for the detection of residual disease, including cytogenetics, cell culture systems, fluorescence in situ hybridization (FISH), multicolor flow cytometry, polymerase chain reaction (PCR), and deep sequencing [10-12]. (See "Tools for genetics and genomics: Cytogenetics and molecular genetics".)

In order to apply these techniques for MRD detection, they must demonstrate three salient characteristics [13]:

●Specificity – The ability to discriminate between malignant and normal cells

●Sensitivity – The ability to detect one malignant cell in a background of at least 1000 normal cells

●Reproducibility and applicability – The techniques must be standardized and reproducible, and results must be available in a timely manner

PCR-based techniques, multicolor flow cytometry, and deep sequencing are able to fulfill the above criteria for detection of MRD for most patients (table 1). In Europe, many centers use an approach to flow cytometry and PCR that has been standardized and validated allowing for more uniform assessment of MRD and of reporting results. While flow cytometry has been the technique most commonly used for pediatric ALL/LBL in the United States, there has been no standardization of this technique in most clinical laboratories in the United States. The percentage of patients eligible for monitoring using each technique depends upon the phenotype and genotype of the leukemia.

The preferred method depends at least in part on the impact the results will have on patient care. When MRD analysis is used to identify patients at high risk of relapse with plans to escalate care, some clinicians will favor flow cytometry over PCR. While slightly less sensitive, results from flow cytometry are available more rapidly, and the likelihood of relapse following a positive result is high. In contrast, PCR is often preferred to identify patients with a low risk of relapse for whom a reduction in treatment intensity may be an option. With either method, the sample and technique used are critical for ensuring a reliable result. As an example, analysis of a small sample risks a false-negative interpretation.

The following studies have compared the sensitivity of different techniques:

A multicenter study compared the ability of real-time quantitative PCR and multicolor flow cytometry to predict bone marrow relapse using 726 surveillance samples from 228 children with ALL/LBL [14]. Concordant results were obtained with these two techniques in 84 percent of day 29 samples using an MRD threshold of 0.1 percent. For patients with B lineage ALL/LBL, PCR and flow cytometry were equally able to predict bone marrow relapse. For patients with T lineage ALL/LBL, PCR was superior to flow cytometry in predicting bone marrow relapse. Neither method was able to predict isolated extramedullary relapse.

In another study, diagnostic and follow-up samples from 106 patients with ALL/LBL were evaluated with deep sequencing, multiparameter flow cytometry, and allele-specific oligonucleotide PCR [12]. Sequencing detected MRD in all 28 samples shown to be positive by flow cytometry and in 35 of the 36 shown to be positive by PCR and revealed MRD in 10 and 3 additional samples that were negative by flow cytometry and PCR, respectively. These results suggest that deep sequencing is more sensitive than PCR and flow cytometry for detecting MRD. While deep sequencing is available commercially, the clinical utility of such assays has not yet been proven and prospective studies are needed to assess the correlation of MRD detected using this method with clinical outcomes.

POLYMERASE CHAIN REACTION (PCR)

Real-time quantitative PCR — Real-time quantitative polymerase chain reaction (RQ-PCR) can amplify a DNA or cDNA sequence unique to the leukemic clone resulting in the detection of one malignant cell among 10⁴ to 10⁵ normal cells [15]. (See "Tools for genetics and genomics: Cytogenetics and molecular genetics" and "Polymerase chain reaction (PCR)", section on 'Measurable residual disease'.)

There are two types of targets used to detect MRD in patients with ALL/LBL:

●Antigen receptor genes – Immunoglobulin (Ig) or T cell receptor (TCR) gene rearrangements

●Gene fusions – Leukemia-specific chromosomal rearrangements (fusion gene transcripts)

Most leukemia clones contain Ig or TCR gene rearrangements that can be used as targets. In comparison, leukemia-specific chromosomal rearrangements are found in a minority of cases. Two or more independent PCR targets are commonly monitored to prevent false-negative results due to evolution of the dominant clone.

Targets of RQ-PCR

Ig and TCR gene rearrangements — Random deletions and insertions of nucleotides that occur during Ig and TCR rearrangement result in the generation of unique junctional sequences that can serve as leukemic clone-specific markers. At the time of diagnosis, the precise nucleotide sequence of the junctional region of the Ig or TCR genes in the leukemic cells can be identified. Junctional region-specific oligonucleotides can then be designed to use as patient-specific probes (or primers) for MRD detection [16]. This approach to MRD analysis is applicable to at least 95 percent of ALL/LBL cases (both B cell lineage and T cell lineage).

Chromosomal rearrangements — RQ-PCR can also be used to identify and quantify leukemia-specific breakpoint fusion regions resulting from chromosomal rearrangements. This approach to MRD analysis is applicable to approximately 40 percent of B lineage ALL/LBL and 25 percent of T cell lineage ALL/LBL. (See "Chromosomal translocations, deletions, and inversions" and "Classification, cytogenetics, and molecular genetics of acute lymphoblastic leukemia/lymphoma".)

Oligonucleotide primers can be designed at opposite ends of the DNA breakpoint fusion region such that the PCR product contains the tumor-specific fusion sequences. Since PCR products should not exceed approximately 2 kilobases (kb) in routine MRD studies, DNA PCR amplification can only be used when breakpoints cluster in a small area (preferably <2 kb). In ALL/LBL, DNA-based PCR for MRD has been standardized in Europe for [13,17]:

●t(1;19) with TCF3::PBX1 (formerly called E2A::PBX1)

●t(4;11) with KMT2A (formerly called MLL)::AFF1

●t(9;22) with BCR::ABL1 p190 and BCR::ABL1 p210

●t(12;21) with ETV6::RUNX1 (formerly called TEL::AML1)

●Microdeletion 1p32 with SIL::TAL1

Advantages of RQ-PCR — The advantages of RQ-PCR for MRD monitoring include [18]:

●Sensitivity – While the sensitivity for each junctional region target varies, the sensitivity of PCR is typically 0.5 to 1.0 log higher than that obtained with standard flow cytometry. PCR for Ig/TCR clones usually reaches sensitivities of 10^-4 to 10^-5.

●Wide applicability – PCR testing is applicable to most cases of ALL, given the high frequency of Ig and TCR gene rearrangements in ALL.

●Sample stability – The DNA samples tested with PCR are very stable during transport.

●Minimal tissue requirements

●Standardized method

Limitations of RQ-PCR — The disadvantages of RQ-PCR for MRD monitoring encompass a variety of technical and biological issues including [19,20]:

●Contamination – Avoidance of contamination of the reaction product requires strict quality control.

●Reproducibility – Poor reproducibility can occur when small numbers of transcripts are present.

●Initial testing – Initial testing to determine the unique clonal rearrangement is expensive and time-consuming (results may take up to four to five weeks).

●Clonal evolution – Evolution of the leukemic clone, subclone formation, and/or oligoclonal populations can cause both false-negative and false-positive results [21,22].

●Need for diagnostic sample – Since the junctional region rearrangements are unique to the leukemia clone, a sample from the time of diagnosis must be available to determine the appropriate primers.

●Availability – PCR is restricted to experienced molecular hematology laboratories.

Oligoclonality of Ig and TCR gene rearrangements at diagnosis is a relatively frequent phenomenon [23,24]. Primary and secondary rearrangements occurring during the course of the disease and during remission might result in the loss of specific junctional regions initially identified at diagnosis; therefore, it appears prudent to monitor ALL/LBL patients with two or more independent PCR targets in order to prevent false-negative results during follow-up [13,25,26].

Reverse transcriptase PCR — Reverse transcriptase polymerase chain reaction (RT-PCR) can be used to identify leukemia-specific fusion mRNA created by DNA breakpoints that span areas larger than 2 kb, such as BCR/ABL1 resulting from t(9;22).

In RT-PCR, messenger RNA (consisting of transcribed coding exons with the non-coding introns excised) is converted into complementary DNA (cDNA) by an RNA-dependent DNA polymerase, termed reverse transcriptase. The cDNA complex is then changed into double-stranded DNA, becoming the template for a subsequent PCR reaction. Results of RT-PCR are qualitative rather than quantitative (figure 1).

In ALL, RT-PCR has been used to detect the following transcripts:

●BCR::ABL1 resulting from t(9;22) [27,28]

●E2A::PBX1 in cases precursor-B ALL/LBL with t(1;19)

●KMT2A::AFF1 in t(4;11) leukemia

●ETV6::RUNX1 resulting from t(12;21) in precursor B ALL/LBL [29-32]

Advantages of RT-PCR — Like PCR, RT-PCR is extraordinarily sensitive and rapid. Unlike Ig and TCR gene rearrangements, the leukemia-specific fusion mRNA is not specific to an individual and remains stable throughout the course of disease.

Limitations of RT-PCR — RT-PCR is applicable to a minority of cases. The most common RT-PCR target used in ALL/LBL is BCR::/ABL1, which can be found in approximately 25 and 5 percent of adult and pediatric cases, respectively. Other limitations include:

●Degradation – RNA degradation and inefficient conversion of mRNA to cDNA

●Varying methods – Methods for RQ-PCR vary among centers

●Cross-contamination – Risk of cross-contamination, requiring strict quality control

DNA SEQUENCING — Next generation sequencing (NGS) is the most sensitive method for detection of MRD in ALL/LBL, with sensitivity of one leukemic cell among 1 million leukocytes (ie, 10^-6) [12,33]. NGS is available commercially.

NGS uses multiplex PCR to amplify several different (eg, immunoglobulin [Ig] or T cell receptor [TCR]) gene rearrangements and then sequences the DNA with high depth (coverage) to identify clonal gene rearrangements (ie, leukemia-specific sequences) to serve as a genetic "barcode" with which to quantitate MRD.

One study reported that NGS was more sensitive for detecting MRD than multiparameter flow cytometry or allele-specific oligonucleotide PCR [12]. In one study, among 106 patients with ALL/LBL, NGS detected MRD in all 28 samples shown to be positive by flow cytometry and in 35 of the 36 shown to be positive by PCR; in addition, NGS detected MRD in 10 and 3 additional samples that were negative by flow cytometry and PCR, respectively. In another study, results from paired bone marrow and peripheral blood specimens in adults with ALL/LBL were highly concordant for MRD; sensitivity and specificity of MRD detection in blood was 87 and 90 percent, respectively, relative to MRD in marrow. [34]. Among the patients who later relapsed following allogeneic hematopoietic cell transplantation (HCT) or chimeric antigen receptor (CAR)-T cell therapy, MRD was detected two to three months before clinical relapse.

FLOW CYTOMETRY — Multicolor flow cytometry uses a laser to determine specific immunophenotypic features of up to thousands of cells per second. MRD can be identified using flow cytometry based on the aberrant expression of antigens [4,35-39]. This method of MRD assessment is applicable to approximately 85 percent of B cell lineage ALL/LBL and 90 percent of T cell lineage ALL/LBL.

The sensitivity varies with the number of simultaneous colors used. Current flow cytometry techniques use six to eight colors to assess MRD with a sensitivity which is approximately 10^-4, or about 0.5 to 1 log lower than that of polymerase chain reaction (PCR). Investigational next-generation flow cytometry techniques using eight or more colors (eg, EuroFlow-based algorithms) have reported sensitivities of <10^-4 to 10^-5 [40-42].

Two examples illustrate the power of multicolor flow cytometry:

●B lineage – In B cell ALL/LBL, leukemic blasts may co-express T cell antigens (eg, CD5 and CD7) or myeloid antigens (eg, CD13 and CD33), have asynchronous expression of antigens, or delete or overexpress a specific antigen (particularly CD10) [43,44]. Using two four-color combinations of antibodies (CD19, CD45, CD20, CD10; and CD19, CD45, CD9, CD34), aberrant leukemic blasts could be identified in 81 of 82 pediatric B lineage ALL/LBL cases [45].

●T lineage – In T cell ALL/LBL, leukemic blasts nearly always co-express terminal transferase, CD2, cytoplasmic CD3, CD5, and CD7 in addition to other T cell antigens [46]. Cells with this immunophenotypic pattern are absent or rare in normal bone marrow or blood [13].

Sequential MRD monitoring using multiparameter flow cytometry has been shown to be a valuable predictor of relapse in children and adults with ALL/LBL [47,48]. As an example, in one study, the cumulative rate of relapse for those negative for MRD by flow cytometry was 10 percent [49], whereas it was 23, 43, and 72 percent for those with MRD of <0.1, 0.1 to <1.0, and ≥1.0 percent, respectively.

Advantages of flow cytometry — Advantages to flow cytometry for the detection of MRD include:

●Wide applicability – Immunophenotyping with flow cytometry can be applied to 80 to 95 percent of patients with ALL/LBL [18].

●Speed – Same day reporting of results

●Quantitation – Although not standardized, results from flow cytometry are quantitative rather than qualitative. Actual cells are counted and not just numbers of transcripts or DNA molecules. Although investigational, standardized techniques are available for next-generation flow cytometry (eg, EuroFlow-based algorithms).

●Samples – Most published data use bone marrow as the sample source, but whole blood may be used. (See 'Source of cells' above.)

Limitations of flow cytometry — Limitations to the use of flow cytometry for MRD assessment include:

●Immunophenotypic shifts – There is also the potential for a change in immunophenotypic expression of the leukemic cells during disease progression, resulting in a false-negative result [50,51]. The likelihood of a false-negative result is less when next-generation flow cytometry techniques are used.

●Samples – Low cellularity of bone marrow samples during and after therapy can be a limitation of the assay.

●Need for diagnostic sample – Although not strictly required for all cases, evaluation of the immunophenotype of the leukemic clone at the time of diagnosis is preferred.

In Europe, many centers use a technique for flow cytometry that has been standardized and validated. There has been no standardization of this technique in most clinical laboratories in the United States.

RESPONSE DEFINED — In addition to standardization of the techniques used to assess and quantify MRD, it is essential that clinicians and researchers use common terminology to report individual patient results and allow for the comparison of trial outcomes. As such, a panel of representatives of the major European study groups on childhood and adult ALL/LBL has proposed the following definitions [18]:

●Complete MRD response – No MRD is detected with assessment that complies with a set of minimal technical requirements for the method used.

●MRD persistence – Presence of a continuously quantifiable MRD positivity measurable at at least two time points with at least one relevant treatment element in between.

●MRD reappearance – Conversion from MRD negativity to quantifiable MRD positivity, with confirmation from a second sample prior to a change in treatment.

The terminology "MRD negative" cannot be applied simply because an MRD assessment does not identify residual disease below the level of detection of a given technique. As commonly used, the term "MRD negative" implies that a technique with adequate sensitivity (≤10^-4) has been used with proper technique on an adequate sample and that no residual disease was found. High certainty regarding the MRD assessment is necessary if the results will impact treatment decisions. The clinical use of MRD assessment in ALL/LBL is discussed in more detail separately. (See "Clinical use of measurable residual disease detection in acute lymphoblastic 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 5^th to 6^th 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 10^th to 12^th 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 lymphoblastic leukemia (ALL) treatment in adults (Beyond the Basics)")

SUMMARY

●MRD in acute lymphoblastic leukemia (ALL)/lymphoblastic lymphoma (LBL) – Measurable residual disease (MRD; previously called minimal residual disease) refers to the presence of leukemic cells that remain after achievement of hematologic complete remission, but below the limits of detection by morphology alone. Such subclinical levels of residual blasts are responsible for most relapses of ALL/LBL after initial disease response.

●Role of MRD in ALL/LBL – Measurement of MRD after remission induction and/or consolidation therapy is highly prognostic for relapse, prognosis, and is routinely used in clinical care of patients with ALL/LBL, as described separately. (See "Clinical use of measurable residual disease detection in acute lymphoblastic leukemia".)

●Source of cells – MRD analysis can be performed on bone marrow aspirate specimens or blood. Bone marrow is preferred, especially in cases of B cell lineage blasts. (See 'Source of cells' above.)

●Choice of method – The preferred approach for detecting MRD varies according to immunophenotype and molecular features of the blasts and available resources. Sensitivity for detection of blasts varies with the technique (table 1). (See 'Methods for detecting MRD' above.)

Methods include:

•Flow cytometry – Multicolor flow cytometry, which identifies blasts based on their aberrant immunophenotype, can be used for most patients with ALL/LBL and is rapid, but it is less sensitive that molecular techniques and there can be false-negative results if the immunophenotype changes. (See 'Flow cytometry' above.)

•Polymerase chain reaction (PCR) – PCR can quantify MRD in most cases of ALL/LBL. PCR detects blasts by immunoglobulin (Ig) or T cell receptor (TCR) gene rearrangements or leukemia-specific chromosomal rearrangements. PCR is more sensitive for detecting MRD than flow cytometry, but it requires a sample obtained at the time of diagnosis; false-positives can arise from sample contamination and false-negatives can be due to the evolution of the leukemic clone. (See 'Polymerase chain reaction (PCR)' above.)

•Reverse transcriptase PCR (RT-PCR) – RT-PCR can identify leukemia-specific fusion mRNA, such as the BCR::ABL1 rearrangement associated with t(9;22). RT-PCR is very sensitive and remains stable throughout the disease course, but it is applicable for only a minority of patients and the mRNA target is more easily degraded during sample handling/transport, compared with DNA. (See 'Reverse transcriptase PCR' above.)

•DNA sequencing – Next generation sequencing (NGS) uses multiplex PCR to amplify Ig or TCR gene rearrangements and then sequences the DNA with high depth (coverage). NGS is the most sensitive method for detecting MRD and it enables monitoring of MRD and it can identify clonal evolution during therapy. The clinical utility of its greater sensitivity has not yet been proven. (See 'DNA sequencing' above.)

●MRD responses – Clinical application of MRD assessment requires proper technique with adequate sensitivity (≤10^-4) if the results will impact treatment decisions. Descriptions of complete MRD-negativity, MRD persistence, MRD reappearance are presented above. (See 'Response defined' above.)