Minimal Residual Disease: ALL It’s Cracked Up to Be?

Monitoring minimal residual disease (MRD) has become the standard of care in pediatric patients with acute lymphocytic leukemia (ALL) based on evidence that it is a strong prognostic factor for patient outcomes – patients who test negative for MRD have better outcomes than those who test positive.

However, MRD isn’t a panacea. There are still many unanswered questions about what MRD is able to tell us, and what it isn’t – especially in adult patients with ALL. The use of MRD monitoring in adult patients is much less prevalent due to a lack of clear evidence and inconsistencies among the labs that conduct the tests.

Despite some lingering questions, there is no doubt that the movement toward MRD monitoring is in full swing. ASH Clinical News spoke with several experts about MRD monitoring in pediatric and adult ALL, its role in signaling a relapse, and whether its results can help doctors avoid overtreating low-risk patients and identifying high-risk patients who would benefit from more intensive therapy.

Measuring MRD

In the pediatric realm, MRD measurement is replacing routine bone marrow tests as the standard in assessing persistent disease, even as oncologists place less emphasis on the results from what has in the past been the standard of bone marrow biopsies.

“I don’t trust it,” Ching-Hon Pui, MD, chair of the department of oncology at St. Jude Children’s Research Hospital in Memphis, Tennessee, told ASH Clinical News. “I don’t rely on my eye to tell me whether or not a patient has a leukemic cell.”

Although he still performs morphologic bone marrow tests, Dr. Pui no longer depends on them. MRD monitoring is able to provide greater sensitivity and identify disease not found on bone marrow tests, without adding significant costs.

There are several ways to test for MRD, and often the technique chosen depends on the lab where the test is being conducted.

In the United States, the most common method for detecting MRD is flow cytometry: leukemia-associated immunophenotypes that are not found on normal hematopoietic cells are identified through a process of reacting live cells with inked antibodies that fluoresce when illuminated with a laser.

In Europe, the common approach is using polymerase chain reaction (PCR) to screen and amplify a DNA in the immuglobulin gene or T-cell receptor to identify a clone associated with leukemia. Next, specific primers or molecules that will only bind to this clone are made to test later samples for the presence of this clone. So, the test is only looking for that specific clone present in that particular leukemia.

Each method has its own set of advantages and disadvantages. Flow cytometry, for instance, is less expensive, can often report quantitative results within a day, and has a larger evidence base (having been used in most U.S.-based trials). PCR is said to be more sensitive, although researchers say it isn’t clear yet how critical it is to have that lower detection.

“Studies have compared these two methods head-to-head, and they are quite concordant, particularly in higher levels of MRD,” said Stephen Hunger, MD, chief of the division of pediatric oncology and director of the center for childhood cancer research at Children’s Hospital of Philadelphia.

There is an emerging technique on the horizon: next-generation sequencing. Essentially, it works on the same principle as PCR but, instead of only amplifying one sequence and making a specific primer, next-generation sequencing amplifies all possible sequences in the gene region of interest. At diagnosis, patients get a signature for that particular leukemia that can be screened for in subsequent tests.

“In standard PCR testing, we have to do a special test for each patient. In next-generation sequencing we only do one test for all possible sequences and then look for that particular signature,” Michael J. Borowitz, MD, PhD, professor of pathology and oncology at Johns Hopkins Medical Institutions, explained.

Next-generation sequencing has a greater sensitivity and, according to Dr. Hunger, is able to “get down to one cell of a 100,000 and, potentially, to one cell in a million.”

This emerging technology is predicted to become the future of MRD detection for ALL, but there are still hurdles to overcome – including a need for large-scale clinical trials – before it is used widely in practice.

Powerful Prognostics

MRD monitoring aids in risk stratification, can be used to determine the efficacy of treatment, identify the presence of tumor cells after therapy, or determine the best timing for a stem cell transplantation. MRD may also forecast the possibility of a relapse – although those in the field caution that MRD does not define relapse.

Ample research has shown that MRD is a valuable risk-prediction tool in pediatric patients with ALL. In a recent study, Dr. Pui and colleagues found that patients who had MRD on day 19 of remission induction therapy had significantly worse 10-year event-free survival (EFS) rates than those who tested negative – regardless of their initial risk stratification.1 Patients who were categorized as being low-risk but who had MRD levels ≥1 percent had 10-year EFS rates of just 69.2 percent, while patients who were initially categorized as standard-risk but who had MRD levels <1 percent had 10-year EFS rates of 82.9 percent.

Due to its highly prognostic nature, hematologists and oncologists often use MRD monitoring to tailor therapies for pediatric patients based on their risk level.

For instance, Dr. Borowitz explained, patients who test positive for MRD or those who have the highest levels of residual disease at the end of the first round of treatment are good candidates for intensified treatment or different drug therapies to more aggressively combat the disease. “If patients are MRD-positive, no matter what other risk factors are present, they get more intensive therapy than they would have otherwise,” he said.

But whether intervening and intensifying therapy for patients who are MRD-positive produces better outcomes has never been formally studied in a large randomized trial. Once MRD was found to be such a strong prognostic factor in patient outcomes, Dr. Borowitz noted, it introduced an ethical dilemma: Many hematologists and oncologists did not want to hold back on the intensity of treatment if an ALL patient was found to be MRD-positive, given that these patients had worse overall outcomes, even though continuing or introducing more intense therapy has never been shown prospectively to improve on those outcomes.

“I think everybody is comfortable with the idea that MRD-positive patients must get different treatment,” he said, “but there are no data from a phase III, randomized trial that definitively states that giving these patients more therapy improves their outcome compared to the same cohort. Those data are not easy to find.”

While MRD is used to justify intensifying therapy for patients who are positive for MRD, Dr. Borowitz commented that perhaps a more difficult question is whether to reduce therapy for those who test negative.

“Reduction of therapy is always a little bit trickier, because an MRD-negative patient may have had a good outcome based on the therapy that he or she received,” he said. Patients also tend to be more cautious about reducing than intensifying therapy, he added.

Despite the strong evidence for MRD’s powerful prognostic ability, it’s not time to throw out other testing methods when making decisions about altering therapy. For example, in a study of 33 pediatric patients with T-cell ALL, MRD levels alone, measured at the end of induction therapy, were not sufficient to justify altering therapy.2 After a median follow-up of four years, 32 patients remained in continuous complete remission, despite 19 patients testing positive for MRD; all 19 of those MRD-positive patients remained in remission at the time of the study.

The Role of MRD in Adult Settings

While the data on the prognostic value of MRD are strong in pediatric ALL, the research in the adult setting is less extensive, and the knowledge of MRD’s role in care is more fragmented. This is in part due to a lower number of adults with ALL enrolled in clinical trials compared with their pediatric counterparts. Low enrollment makes it difficult to compare outcomes, as the lack of standardization leads to patients being given different medications and having MRD levels measured at different time points.

So, is MRD monitoring ready to graduate to adult ALL?

“Right now, I would say that the vast majority of physicians in the ALL community do not use MRD in adult patients because they don’t know what the levels mean,” said Jerry Radich, MD, member of the Clinical Research Division at Fred Hutchinson Cancer Research Center in Seattle.

Compared with pediatrics, the testing process in adult ALL is not as structured or streamlined. With recent comparative analyses of ALL MRD testing laboratories, the National Cancer Institute (NCI) discovered that, although there was high concordance between two primary reference laboratories used by the Children’s Oncology Group (COG), the adult reference labs that participated in the pilot study were not at all concordant.3

“As a result, there was enthusiasm among the adult reference laboratories to participate in a voluntary standardization approach and adopt the COG six-color panel as the starting point for that standardization,” said J. Milburn Jessup, MD, chief of the Diagnostics Evaluation Branch at the Cancer Diagnosis Program in the Division of Cancer Treatment and Diagnosis at NCI.

As the industry moves toward precision medicine, Dr. Jessup said it is essential to have accurate assays to provide accurate results that can help guide physicians’ decision making.

To address some of these concerns, NCI is sponsoring a meta-analysis of all the available data in pediatrics and adults to confirm MRD’s prognostic abilities, which will also hopefully guide the use of MRD in adults. As part of NCI’s efforts, adult reference laboratories are also voluntarily participating in a standardization process to improve consistency between laboratories. However, these efforts are just the beginning.

Ultimately, NCI would like to see MRD used as “somewhat of a surrogate” in adult clinical trials to reduce the time of the trial, and to identify which arms of clinical trials are working well at earlier timepoints, Dr. Radich noted.

To that end, NCI is conducting another meta-analysis to compare and contrast ALL in adults and pediatrics to determine what MRD levels matter at what time points for adults. “Those levels are pretty much nailed down in pediatrics, but I think there are still very, very loose variables in adults,” Dr. Radich said.

Is Timing Everything?

Identifying patients who have a worse prognosis is critical in determining best treatment options, but just how early can the prognosis be determined?

Most agree that assessing patients’ bone marrows after induction therapy at about four or five weeks can evaluate the effectiveness of treatment and identifying patients who are at the greatest risk for a poor outcome.

“It’s very clear that pediatric ALL patients who have high levels of MRD at the end of the first month of therapy – using 10-4 or 0.01 percent as the common definition of ‘high’ MRD – do significantly worse than those that are negative at that point,” Dr. Hunger said. “Most groups and physicians will intensify treatment after the first month for those who are MRD-positive.”

But, some believe even earlier testing could be a benefit, as Dr. Pui demonstrated in his recent study in The Lancet Oncology, where patients who tested MRD-positive (>1%) after just 19 days of remission induction therapy had significantly worse 10-year EFS.1

“For most centers, I think the best timing is really two weeks into induction,” Dr. Pui stated. At this point, he explained, false-positives are unlikely and oncologists can make decisions about whether to back off or intensify treatment even earlier.

But measuring MRD at only one time point may not be enough – especially considering the heterogeneity among ALL subtypes. For instance, T-cell patients have a slower response than patients with precursor B-cell ALL, according to Paul Gaynon, MD, professor of pediatrics at Children’s Hospital Los Angeles. “In T-cell ALL, many patients were slow responders by day 29 but then rallied by day 85,” he said. “They had good responses and did quite well.”2

According to Dr. Hunger, there is also significant evidence that testing again at later time points could be valuable – particularly for patients who test MRD-positive at the one-month mark. “Research has shown that people who test positive after induction, but test negative at three months do substantially better than those who are still positive at three months,” he said.4

Many physicians also rely on MRD measurements to evaluate the best timing to perform a stem cell transplantation. If patients are MRD-positive going into transplant, Dr. Borowitz said, data have shown that their outcomes are much worse than patients who are MRD-negative before transplant.5 For this reason, oncologists may want to take MRD into consideration when deciding on whether transplant is needed, as well as thinking about using additional therapy to get patients to an MRD-negative state if transplant is deemed necessary.

The Search for  a New Surrogate Endpoint Continues

Patients who have MRD do worse than those who don’t – that much is clear. But despite its significant prognostic properties, using MRD as a clinical endpoint is still under dispute.

“There are some well-known examples of cases where therapy has improved outcomes but didn’t change MRD,” Dr. Hunger said. “It makes people a little leery about this idea.”

For example, an open-label randomized trial comparing the effect of mitoxantrone with idarubicin in children with first relapse of ALL found that, although the mitoxantrone-treated patients had a lower relapse rate, there was no apparent difference in MRD between the two drugs in the intermediate-risk group – leading researchers to believe that the decrease in relapse was unrelated to the kinetics of disease clearance.6

“To enable the quick assessment of the number of new drugs now in the pipeline, study designs are incorporating the use of MRD as a surrogate marker of outcome,” lead author Catriona Parker, PhD, and co-authors wrote. “If we had opted to use such a study design, mitoxantrone would have been discarded.”

The hope, Dr. Gaynon said, was that if there was no difference in MRD levels after the one-month time point, there would also be no difference in EFS in the long run, but evidence has failed to support this hypothesis.

“Sadly, MRD is not a surrogate for efficacy,” he said.

An Early Indicator for Relapse?

While researchers tend to agree that MRD has clear prognostic properties, its role after remission in predicting a relapse is less clear. “This is an ongoing research question,” Dr. Pui cautioned. “Nothing has been definitively proven.”

MRD appears to be a significant risk factor for relapse in adult patients: A 2009 study published in Blood found that MRD analysis during early post-remission therapy improves risk definitions and helps improve risk-oriented treatment strategies. Patients who were MRD-negative had five-year overall survival and disease-free survival rates of 0.75 and 0.72, respectively, compared with rates of just 0.33 and 0.14 in MRD-positive patients. Presence of MRD, then, was the most significant risk factor for relapse, with a hazard ratio of 5.22.7

According to Dr. Hunger, MRD is far from a perfect predictor of relapse: Some patients who are MRD-negative will still relapse, and others who are MRD-positive will not. The main reason for the hesitation to adopt rising MRD levels as an equivalent of relapse (before a morphologic relapse is determined) is that physicians do not want to expose patients to unnecessary morbidity and mortality associated with additional therapy or stem cell transplants.

“Until there is more proof that persistent or recurrent MRD during treatment is predictive of impending relapse – highly predictive, not just ‘suggested’ – people are not really willing to act on that information by and large,” Dr. Hunger said.

What We Still Don’t Know

So, is there a “safe” level of MRD? That is still unclear, but newer technology such as next-generation sequencing could provide more answers in the near future. Dr. Hunger, however, cautioned that immunoglobular T-cell receptor-based technologies get down to such a low level of measurement that it is not always possible to distinguish between a pre-leukemic ancestral clone and a fully leukemic clone.

“We really do not know what some of these very low levels mean yet,” he said.

While we do know that ALL patients who don’t have MRD at earlier points in their therapy tend to have the best prognosis, researchers are still trying to determine the best way to treat patients who appear to be in remission based on morphologic examinations but who still have MRD. Consensus about whether MRD can be used as an endpoint in clinical trials, or whether oncologists should be monitoring MRD to signal relapse, have yet to be reached.

Currently, it seems that the unanswered questions surrounding MRD in ALL outweigh the answered ones.

Is MRD monitoring in ALL all it’s cracked up to be? There was a common refrain among all of the experts we spoke with: Despite previous and ongoing research, it’s still too soon to call.

As research forges ahead, finding the answers to these questions will be the key in ultimately understanding the true power of MRD.—By Jill Sederstrom


  1. Pui CH, Pei D, Coustan-Smith E, et al. Clinical utility of sequential minimal residual disease measurements in the context of risk-based therapy in childhood acute lymphoblastic leukaemia: a prospective study. Lancet Oncol. 2015;16:465-74.
  2. Parekh C, Gaynon PS, Abdel-Azim H. End of induction minimal residual disease alone is not a useful determinant for risk stratified therapy in pediatric T-cell acute lymphoblastic leukemia. Pediatr Blood Cancer. 2015 May 14. [Epub ahead of print]
  3. National Cancer Institute. Comparison of laboratory test results of minimal-residual disease in samples from patients with acute myeloid leukemia (NCT01498302). Accessed August 14, 2015 from
  4. van Dongen JJM, van der Velden VHJ, Bruggeman M, Orfao A. Minimal residual disease diagnostics in acute lymphoblastic leukemia: need for sensitive, fast, and standardized technologies. Blood. 2015;125:3996-4009.
  5. Eckert C, Biondi A, Seeger K, et al. Prognostic value of minimal residual disease in relapsed childhood acute lymphoblastic leukaemia. Lancet. 2001;358:1239-41.
  6. Parker C, Waters R, Leighton C, et al. Effect of mitoxantrone on outcome of children with first relapse of acute lymphoblastic leukaemia (ALL R3): an open-label randomised trial. Lancet. 2010;376:2009-17.
  7. 7. Bassan R, Spinelli O, Oldani E, et al. Improved risk classification for risk-specific therapy based on the molecular study of minimal residual disease (MRD) in adult acute lymphoblastic leukemia (ALL). Blood. 2009;113:4153-62.