Home » News

Minimal Residual Disease, Deep Sequencing, And Prognosis In Multiple Myeloma

11 Comments By and
Published: Apr 8, 2014 7:41 am; Updated: Apr 8, 2014 4:35 pm

A recent Spanish study adds important findings to the growing literature about the measurement and significance of minimal residual disease in multiple myeloma.

Myeloma patients are said to have minimal residual disease if, after having responded well to treatment, they nevertheless continue to have myeloma cells in their bodies.

In their study, the Spanish researchers used a sensitive new technique known as deep sequencing to conduct minimal residual disease testing. They used the technique to test for residual disease in bone marrow samples from myeloma patients who had completed their first line of therapy and achieved at least a very good partial response.

Patients who had no signs of residual disease based on deep sequencing had markedly longer pro­gres­sion-free and overall survival than other patients.

In addition, among patients who had residual disease after initial treatment, the extent of the residual disease had a noticeable impact on survival. Patients with lower levels of residual disease had longer progression-free and overall survival than those with higher levels of residual disease.

The investigators also demonstrated that testing using deep sequencing provides results that, in a large majority of cases, are the same as those of more common, but less sensitive, testing methods such as multiparameter flow cytometry.

However, in cases where the results from deep sequencing and multiparameter flow cytometry differed, the deep sequencing results provided a better guide to patient prognosis.

Although their results underscore the potential value of deep sequencing for minimal residual disease testing, the Spanish investigators recognize that the technique has its limitations.  Myeloma cells, for example, are not uniformly distributed in the bone marrow.  Thus, even when sensitive tests fail to detect residual disease in a bone marrow sample, there may still be myeloma cells elsewhere in the patient's body.

The researchers therefore suggest that additional methods, such as imaging techniques and monitoring of circulating myeloma cells, be used to assess disease status and improve estimates of the risk of pro­gres­sion.

They add that the detection of minimal residual disease by deep sequencing may also contribute to the design of patient-specific treatment approaches, such as discontinuation of treatment for patients without residual disease, or intensification of treatment for patients with minimal residual disease


Due to the progress that has been made in the treatment of multiple myeloma in the last decades, the share of newly diagnosed patients who achieve a complete response following treatment has increased sig­nif­i­cant­ly.

However, almost all myeloma patients eventually relapse after initial treatment.  Most of these relapses, the Spanish researchers say, "can be attributed to the persistence of undetectable minimal residual disease."

The term "minimal residual disease" (MRD) is used to describe the presence of small numbers of myeloma cells in a patient's body even after the patient has achieved a complete (or very good partial) response to a treatment.

Traditionally, MRD testing in multiple myeloma patients has been carried out using cells from bone marrow samples.  It is also possible, however, to conduct such testing using cells from the blood.

Patients who have no detectable myeloma cells in their cell samples are described as being MRD-negative. Patients with at least some myeloma cells in their samples are categorized as MRD-positive.

One established method for measuring MRD in myeloma patients is a technique called multiparameter flow cytometry.

However, according to the Spanish researchers, there is still room for improvement in the detection of min­i­mal residual disease.

They therefore evaluated the prognostic value of a newer, more sensitive, method of minimal residual disease detection called deep sequencing.  While multiparameter flow cytometry currently has a sensitivity of between 1 out of 10,000 and 1 out of 100,000 cells tested, deep sequencing is sensitive to levels up to 1 out of 1,000,000 cells tested.

(In the literature about MRD testing, a sensitivity of 1 out of 10,000 is often described using the shorthand 10‑4.  A sensitivity of 1 out of 1,000,000, for example, would be labeled 10-6.)

The deep sequencing approach to MRD testing gets its name from the fact that it involves sequencing, which is the process of determining the genetic makeup of a cell.  Deep sequencing is when the sequencing process is carried out multiple times for each single cell tested, which increases the accuracy of the testing.

When deep sequencing is used for MRD testing for myeloma, pre-treatment samples are first analyzed to identify a genetic sequence characteristic of an individual patient's disease. This sequence is then looked for in post-treatment samples from the patient to differentiate between myeloma cells and normal cells.

Study Design

In the current study, the Spanish investigators deep sequenced bone marrow samples from 133 multiple myeloma patients who had participated in several different clinical trials in Spain.

The samples used for most of the analyses during the study were obtained after the patients had completed their first line of therapy.  For patients under the age of 65, the first line of therapy included induction therapy followed by a stem cell transplant.  For patients 65 and older, the first line of therapy included only induction therapy without a stem cell transplant.

Only patients who achieved at least a very good partial response following their first line of therapy were included.  The mean patient age was 62 years old.

All patients also had their MRD levels measured through multiparameter flow cytometry.

Initial findings from the current study were presented at the ASCO 2013 meeting (see related Beacon news).

Study Results

The investigators first deep sequenced bone marrow cells that were taken from the patients prior to their starting treatment for their myeloma. They were able to identify myeloma cell-specific genetic sequences in 91 percent of patients, making them eligible for later MRD testing using deep sequencing.

During follow-up, the majority of patients (73 percent) remained MRD-positive, where MRD-positive was defined as having more than 1 myeloma cell out of 100,000 cells tested.

MRD Status And Patient Prognosis

Patients who were MRD-negative after their first line of therapy had a significantly longer median time to disease pro­gres­sion (80 months) than patients who were MRD-positive (31 months).

Median overall survival was also significantly longer in MRD-negative patients (not yet reached), compared to MRD-positive patients (81 months).

Patients were subsequently divided into three groups according to their MRD levels identified by deep sequencing: MRD levels of greater than 1 myeloma cell out of 1,000 cells; MRD levels between 1 out of 1,000 and 1 out of 100,000; and MRD levels of less than 1 out of 100,000 (MRD-negative).

The investigators found that the median time to pro­gres­sion varied significantly among the three groups (see graph on the left in the figure below).  Patients who were MRD-negative (less than 1 out of 100,000 cells tested) had the longest median time to pro­gres­sion (80 months), followed by patients with MRD levels between 1 out of 1,000 and 1 out of 100,000 cells (48 months), and patients with MRD levels greater than 1 out of 1,000.

The investigators made similar observations for overall survival (see graph on the right in the figure below). Median overall survival was not yet reached for MRD-negative patients (less than 1 out of 100,000) and for patients with MRD levels between 1 out of 1,000 and 1 out of 100,000, compared to 55 months for patients with MRD levels greater than 1 out of 1,000 (27 months).

Time to Progression (left) and Overall Survival (right)
(click on image to view a larger version of it)

When the investigators limited the analysis to patients who had achieved a complete response to first-line therapy, they found that patients who were MRD-negative had a significantly longer median time to pro­gres­sion compared to MRD-positive patients (131 months versus 35 months). The median overall survival was not reached in either group.

Comparisons With Other MRD Testing Techniques

The researchers also compared MRD information obtained by deep sequencing with the results from multiparameter flow cytometry testing.  They found that 83 percent of the samples yielded results that were the same using the two methods: 61 percent of samples were MRD-positive, and 22 percent were MRD-negative.

They also found an overlap in 85 percent of the patients when comparing MRD results based on deep sequencing with MRD results obtained using a third type of MRD testing, known as allele specific oligonucleotide - polymerase chain reaction (ASO-PCR).

For patients whose samples resulted in different MRD results using deep sequencing and multiparameter flow cytometry, the patients who were MRD-negative by deep sequencing had a longer time to disease progression than those who were MRD-negative by multiparameter flow cytometry but MRD-positive by deep sequencing (median not reached versus 50 months).

According to the Spanish investigators, their results suggest that low-level MRD results detected by deep sequencing are clinically significant.

For more information, please refer to the study by Martinez-Lopez, J. et al., “Prognostic value of deep sequencing method for minimal residual disease detection in multiple myeloma,” in Blood, March 11, 2014 (accepted) (abstract).

Update (Apr 8, 2014 - 4:35 pm) - This article has been revised slightly from its original version. The primary change was to the description of how deep sequencing detects myeloma cells during MRD testing. The original version of the article suggested that the testing is done by looking in cells for myeloma-related genetic patterns that are common in myeloma patients in general. That is not correct. Instead, the patterns that are used for the detection of myeloma cells are patient-specific. For each patient, the myeloma-related genetic patterns are established by analysis of a patient's cells from a pre-treatment sample. The Beacon thanks a representative from Sequenta, the company that developed the deep sequencing assay that was used in the Spanish study, for reaching out to The Beacon to clarify this point.

Photo by Bobjgalindo on Wikipedia – some rights reserved.
Tags: , , , , , ,

Related Articles:


  • multibilly said:

    One thing that has always puzzled me about deep MRD testing is whether our doctors would know what to do with the MRD results if they found some specific, small population of cancerous cells as a result of deep MRD testing. Let's say you just went through a cycle of VRd and it was found that you had some small, minimal amount of disease left that hadn't responded to the VRd. Would our doctors then know what to do next (increase or tweak the VRd dose, switch to a different drug or cocktail, not do anything and write it off?). Or, would it pretty much be hit and miss at that point?

    There is already so much uncertainty and debate about what the appropriate front-line or maintenance therapy should be for a given patient that it strikes me that this level of treatment tweaking is beyond the current capabilities and insights of the mm medical community.

    Am I off base here?

  • Karen Morrow said:

    After my last bone marrow I was told I was told I was still in remission but have clone cells. Is MRD the same thing? All my blood work has been coming back good and am on R/D.

  • multibilly said:

    Hi Karen,

    Whether you had MRD performed on as part of your BMB is really something you need to discuss with your doctor. Most BMB procedures involve some level of MRD using a technique that is know as parametric flow cytometry to help identify cancerous cells based on unique proteins that are found on the cell's surface. However, these tests and the way the tests are set up vary widely from institution to institution, as does the "deepness" of their testing.

    The Spanish study is using a different approach to MRD testing that involves what is known as molecular "deep sequencing". It sounds cool, but the technique is expensive and is not widely available.

    There's a lot of investigation going on right now regarding how to best do highly accurate MRD testing in a cost-effective manner that can hopefully be standardized across various institutions. This is one of the chief goals of an effort you may have heard of called the Black Swan Research Initiative.

    Hope this helps.

  • Alice said:

    I had a follow-up bone marrow biopsy nearly a year after I had apparently reached 'complete response' with initial therapy. I was told I had 5% plasma cells that were not light chain restricted ie. a 'normal bone marrow'. I have no idea if I have MRD or not. That biopsy was 2 years ago and I am still in 'compete response'.

  • Steve said:

    Still, since mm bone involvement is EXTREMELY patchy, any type of flow cytometry/deep sequencing using bone marrow for analysis is going to be limited by that very fact....it's patchy!

    However, from a practical application perspective, does this process of flow/sequencing which is dependent upon bone marrow for analysis, suggest that those of us with patchy bone involvement have CT-guided BMX?

    Thanks for the article!


  • multibilly said:

    Regarding Steve's comments, I am guessing that the inherent "patchiness" of BMB samples is not going to affect the flow cytometry results.

    It is is true that BMB samples are patchy when it comes to picking up the total population of plasma cells in a given BMB sample. But it occurs to me that it is unlikely that you would completely miss getting at least a few cells of a given subpopulation of clonal cells in any given BMB sample. That is, you may come up with different percentages of a given subpopulation between different BMB samples, but what are the chances that you wouldn't catch any of a given subpopulation of clonal cells in a given BMB sample?

    Again, this is just wild-ass speculation on my part, so I'd love to hear a more educated comment on this topic.

    Karen and Alice: If you request a copy of your BMB pathology report , you will very likely find a section in there labeled "Flow Cytometry". Within the flow cytometry section of report, you will find a summary on various percentages of cells that express (or don't express)specific "clusters of differentiation". The report will identify percentages of cells that have specific markers that will be labeled as CD38, CD56, CD138, etc. The "CD" prefix simply means "cluster of differentation". A cluster of differentiation is simply a fancy name for an antigen, which is a protein that is found on all cells that you immune system in turn uses to identify friends and foes. There are over 300 different kinds of CDs.

    The flow cytometry test is in fact an MRD test.

    The Black Swan Research Initiative is currently developing an improved and (hopefully) standardized test based on flow cytometry that will be more sensitive and widely used than most of the current flow cytometry tests used today at most labs.

  • Mark said:

    Hi Multibilly,

    Good first post and I do not think you are off base at all. I basically agree with everything you wrote but I do think patients will find knowing if they are MRD positive or negative can make them more confident in the therapy decisions they will need to make.

    If you look at the TTP MRD negative graph the patients on it have a small chance of relapse for the first 4 years or so. Lets say multiple future studies confirm that being MRD negative via 8 color flow cytometry has a TTP curve that looks like that one. If a patient achieves MRD negative status you could choose to go drug free for say 3 years or so to enjoy an excellent QOL knowing you have a low risk of relapse during that time period. You could than make a decision to start treating again at that time knowing your risk of relapse is rising. If you are MRD positive that would likely influence your decision in favor of using maintenance or consolidation therapy. Here is another point in a study of PCR negative patients showing how this information could be used similar to what I was describing above. MR means molecular response (PCR negative) and note points 2 and 3.

    "Besides confirming the strong prognostic value of PCR-based MRD monitoring in MM, our long-term results indicate the following: 1) the 42 months TNT of patients achieving MR underlines the excellent disease control of MM patients once obtained MR; 2) the occurrence of MR loss heralds relapse, with a TNT from MR loss comparable to TNT of patients not achieving MR; 3) there is a 9 months lag between MR loss and need for salvage treatment. These observations will have increasing relevance considering that ongoing methodological developments will allow effective MRD monitoring in the vast majority of MM patients."

    My therapy was based on achieving MRD negative status. My reward for getting it is that I have been myeloma therapy free for 3 years now. I sometimes use the term "mental QOL" and what I mean by that is that I am satisfied based on the small studies of allo patients that achieve MRD negative status that I was not "over treated" or "under treated". I am 100% satisfied with my therapy decisions based on statistics as opposed to what my doctors or my own "treatment philosophy" is or my favorite is when patients say things like "my doctor is a fan of maintenance" or whatever therapy they are discussing. I really do not care what my or any other doctor is a "fan" of or not a "fan" of. I want statistics to base my treatment decisions off of. I think this will be a step forward toward better decision making for patients/doctors. Hopefully knowing if they are MRD negative or positive will help make patients more confident in the therapy decisions they are making.


  • Steve said:

    Found a recent study, published in the journal Blood, that actually used peripheral blood to detect MRD:

    "Samples were subjected to deep sequencing using the LymphoSIGHT™ platform, which has a sensitivity to detect one cancer cell per million leukocytes in peripheral blood (Faham et al, Blood 2012)."

    That could be a real game changer if it can be standardized and made broadly available(read: affordable).

    Who knew? :)

  • Myeloma Beacon Staff said:

    Great discussion everyone. Thanks for all the comments and further insights related to the article.

    Please note that we have updated the article slightly with a revised description of how deep sequencing detects myeloma cells. The change is described in an "update" included at the end of the article. We apologize if the original description was confusing to those of our readers who are more knowledgeable about how the different MRD testing methods work.

  • Steve said:

    In regards to multibilly's comment,

    " I am guessing that the inherent “patchiness” of BMB samples is not going to affect the flow cytometry results. "

    I contacted the lead investigator of this study, Dr. Joaquin Martinez-Lopez,M.D., Phd., who indicated in his return email that the patchiness of myeloma in the bone marrow does in fact pose a limitation on multi-parameter flow cytometry for purposes of MRD detection.

    However, he also stated, "We are now working in detection of MRD in peripheral blood, this could resolve the problem of patchy, in the coming months we are going to have the first results."

    Freakin' awesome!! :)

  • multibilly said:

    Steve: Thanks for clarifying the patchiness issue with respect to flow cytometry. Serum-based flow cytometry would indeed be great.