How to use MRD status to treat high-risk multiple myeloma

This month, the Multiple Myeloma Hub is exploring the theme of high-risk multiple myeloma (MM). The first article on this subject dealt with what you need to know about high-risk MM, and can be found here. We now wish to focus on the treatment of patients with high-risk MM. Two recent studies discuss the use of measurable (minimal) residual disease (MRD) status to treat patients with high-risk disease. The first, by Leire Burgos and colleagues, is a review published in the Journal of Haematology & Oncology regarding the application of MRD in clinical practice.¹

Patients with high-risk MM have a poorer prognosis than groups classified as standard-risk. They also have different treatment requirements. When considering the treatment of high-risk patients, one key factor is how to measure response, whether post-induction, in between rounds of therapy, or during follow-up.

Traditionally, complete response (CR), defined as < 5% plasma cells in the bone marrow, has been used as an endpoint. However, patients with high-risk cytogenetic abnormalities (CAs) have a poorer prognosis than patients with standard-risk MM, even when achieving CR. This has aroused some doubt over the validity of CR as a clinical endpoint for this patient subgroup.

MRD assessment has received a considerable amount of attention in recent years, and many studies have pointed to its use as an alternative endpoint. Patients with high- and standard-risk disease who achieve MRD negativity show more favorable outcomes in terms of progression free survival (PFS) and overall survival (OS).

Key points

The traditional endpoint CR is often assessed using serum immunofixation and counting plasma cells by morphology, which do not differentiate between cancer cells and normal cells. New techniques of increased sensitivity are required to assess the effect of treatment with the advent of new therapeutics. For example, in patients who have achieved CR, how do you monitor the effectiveness of subsequent therapies?

On the other hand, MRD is an option used in clinical trials but has yet to translate consistently into clinical practice. There are multiple techniques for measuring MRD, which can be split up into two categories according to their ability to identify extramedullary disease:

• Positron emission tomography/computed tomography (PET/CT) can detect lesions within the bone marrow along with extramedullary disease.
• Flow cytometry and polymerase chain reaction (PCR) are used for intramedullary disease detection. They have been optimized to include the use of next-generation flow (NGF) cytometry and next-generation sequencing (NGS) techniques, both of which yield comparable clinical results.

For MRD to be a useful factor for guiding treatment, it must fulfill certain criteria:

• It must be more sensitive than CR in predicting prognosis.
• Results must be reproducible regardless of methodology and setting.
• Finally, it must be a useful assessment in all patients, not just for specific subgroups.

Both NGS and NGF have been able to discriminate between patients with MRD and undetectable disease levels in a reproducible fashion between centers. Data is available for transplant-eligible patient groups as well as for ineligible groups. A threshold of at least 10^-5 is recommended for MRD negativity by the European Medicines Agency and the International Myeloma Working Group. It is recommended that all studies state the method and MRD threshold used within the study.

Please note that throughout this article MRD negativity is defined according to the validated threshold used in each individual study.

MRD supersedes CR

Significant differences in survival have been reported in patients that have achieved CR but have MRD compared with MRD-negative groups. MRD negativity is associated with longer PFS and OS. Patients who achieve CR but have persistent MRD have similar prognoses to patients in very good partial response or even partial response.

This improvement in OS and PFS is also reported in patients with high-risk disease. High-risk patients may still have poor outcomes, even if they reach CR. However, if they reach MRD negativity, they can overcome their risk predictions and achieve PFS and OS in line with standard-risk patients.

Caveats of using MRD

• To obtain accurate flow cytometry or PCR MRD results, the samples must be representative of the bone marrow in terms of cellular populations. As MM tends to show patchy intramedullary infiltration, this can be difficult. In addition, hemodilution can occur and reduce the number of cells obtained.
• MM can be present in focal lesions or paramedullary sites, which are not identifiable by NGS or NGF. To comprehensively assess MRD, PET/CT imaging is advised in combination with these other methods. Patients who achieve MRD-negative status both in the bone marrow and at extramedullary sites have the best prognosis.
• Achieving MRD-negative status once is not enough to achieve the best outcome. The duration of MRD negativity is also important. Patients who sustain this deep remission status for 6–12 months do significantly better in survival indices than those who do not.

Application of MRD in clinical practice

How and when to use MRD

Assessing MRD can be done wherever results can help to guide treatment, for example, in between therapy stages, or periodically, e.g., yearly as part of the follow-up. Notably, the presence of residual disease is always an adverse prognostic factor, while the few issues raised previously indicate that MRD negativity has some elements that require clinical validation. Therefore, it is better to make clinical decisions based on the presence of persistent MRD.

MRD measurement can be useful for assessing patients’ responses to subsequent therapies once CR is reached. MRD could potentially be a guide to indicate how long a patient should be kept in continuous treatment. Moreover, identifying which patients may benefit from an increased intensity therapy can be aided by MRD assessment; differentiation between novel therapies may be assessed by their ability to achieve MRD-negative status. As a result of these reasons, many trials are now using MRD as a surrogate endpoint.

Impact of the depth of MRD response and resistance profiling

The second study, published in Blood by Ibai Goicoechea and colleagues, investigated the advantages of MRD assessment using the PETHEMA/GEM2012MENOS65 (NCT01916252) and PETHEMA/GEM2014MAIN (NCT02406144) clinical trials and found that patients who achieve MRD-negative status can overcome their initial risk categorization and demonstrate improved survival. In addition, different modes of treatment resistance were analyzed and identified in MM cells.² The results of this trial were recently covered by the Multiple Myeloma Hub; full details can be found here.³

MRD negativity can overcome high-risk prognosis

The PETHEMA/GEM2012MENOS65 trial used NGF to measure MRD in 390 patients with high- and standard-risk MM. Patients were treated with:

• Six cycles of bortezomib (V), lenalidomide (R), and dexamethasone (d),
• Bul-Mel or Mel-200 conditioning,
• Autologous stem cell transplant, and
• Two cycles of VRd consolidation.

Following this, patients were included in the PETHEMA/GEM2014MAIN trial and treated with either randomized Rd maintenance or Rd and ixazomib for 2 years. If patients were MRD negative at this point, therapy was stopped. If they were MRD positive, Rd treatment was continued for 3 more years.

In terms of CR, the difference between patients with standard-risk (62%) and high-risk (57%) CAs was not statistically significant. However, significantly more patients with standard-risk MM (49%) reached MRD negativity than those with high-risk MM (37%) (p = 0.04). So, despite high-intensity therapy, MRD negativity rates are still inferior in patients with high-risk CAs.

At 36 months, the PFS and OS were > 90% in MRD-negative patients, and of note, there was no significant difference in outcome between patients with high-risk CAs and standard-risk cytogenetics. Thus, patients with high-risk CAs that achieve MRD negativity can overcome the poor prognosis traditionally assigned to them. In patients with detectable residual disease, the 36-month PFS rate decreased to 60% in standard-risk and 37% in high-risk groups (p < 0.001). Similar results were obtained for OS.

Resistance profiling of cancer cells

Whole exome sequencing was performed on 14 sets of matched samples at diagnosis and remaining tumor cells following induction with VRd. At both time points, 40% of mutations and 72% of copy number alterations were present. In patients with standard-risk MM, it was interesting to note that most mutations and copy number alterations became undetectable in the remaining tumor cells, whereas in the high-risk group, there was an increase, indicating potentially a greater genetic instability in the latter.

Analysis of gene set pathway enrichment indicated differential expression between the two risk groups with reactive oxygen species pathways being upregulated in patients with high-risk CAs. Further analysis showed that the expression levels of sodium oxide dismutase 1 (SOD1) and peroxiredoxin 6 (PRDX6) were significantly associated with PFS. Patients with high-risk CAs that upregulated expression of SOD1 or PRDX6 had a significantly worse PFS (these results were later confirmed in an external dataset). By single-cell RNA sequencing, the investigators did not find any clustering of SOD1 and PRDX6 in any particular clone tumor cells with high-risk CAs. Additionally, there were no common lost or acquired genetic changes in the MRD tumor cells.

Overall, these results indicate that instead of certain clones with predefined genetic abnormalities being selected for, resistance evolves transcriptionally within a selected cell population during treatment.

Conclusion

MRD shows great promise as a surrogate biomarker of remission in MM. Clinical trials are already using MRD as a primary endpoint, potentially allowing for more rapid drug development. The reproducibility of results between techniques and centers highlights the value of this measurement and indicates its potential importance for use in clinical practice. For this reason, there is a global effort to standardize when and how MRD is assessed in clinical trials in MM.

In many ways, MRD supersedes CR as a treatment goal. MRD is associated with improved long-term survival outcomes for patients achieving undetectable MRD, even in patients with high-risk CAs, who traditionally fare poorly compared with patients with standard-risk disease. This illustrates the concept that the treatment approach should be focused on achieving MRD-negative status, and not on a specific therapy or combination regimen, since these protocols have shown heterogeneity amongst high-risk populations.

There is still further work required in this field to aid the transition of MRD from clinical trials to general clinical practice. As previously mentioned, the longer the MRD negativity is sustained, the greater its impact on outcomes. This requires periodic bone marrow sampling, which is invasive and painful for patients. Therefore, further investigation into developing methods to measure MRD in the blood is needed.

The Multiple Myeloma Hub will be covering regimens reporting higher rates of MRD negativity in the following weeks, so please check back regularly to see the latest updates.