IMW 2019 | MRD in myeloma

During the XVII International Myeloma Workshop (IMW) in Boston, US, Dr Bruno Paiva¹ and Dr Francesca Gay² both presented on the topic of measurable residual disease (MRD) in multiple myeloma (MM) in a session entitled “MM: diagnosis and management”. This article provides a summary of these talks, covering when and how to measure MRD, and investigating what role MRD may play in the future of regulatory approvals. Additionally, during a subsequent session entitled: “Great debates in myeloma”, Prof. Jesus San Miguel and Prof. Sundar Jagannath debated whether MRD should be used as an endpoint to therapy, in clinical practice.³ The main arguments made by both speakers are summarized in this article (Table 2).

When and how should MRD be evaluated in MM?¹

There is extensive data showing that MRD evaluation is prognostic at any timepoint during the MM treatment pathway including; post-induction, maintenance, and after salvage treatment. Therefore, Dr Paiva suggested that MRD should be measured to define the quality of complete response (CR) and efficacy of treatments given after CR. Additionally, MRD could be used to identify high-risk patients, and ultimately assist treatment-making decisions.

In a study published in the Journal of Clinical Oncology in 2017 by Juan-Jose Lahuerta and Bruno Paiva et al., it was shown that achieving MRD-negativity prolonged survival. Patients in CR, who were MRD-positive, had the same progression-free survival (PFS) and overall survival (OS) as patients in near CR (nCR) or partial response (PR), highlighting that the value of CR lies in the MRD status. Achieving a CR led to a 33% reduction in the risk of progression and/or death, whilst achieving MRD-negativity almost doubled this reduction to 58%. The reduction in risk with MRD negativity was pronounced in patients with high-risk cytogenetics for whom achieving CR had no impact on survival. The benefit of MRD-negativity was seen regardless of; eligibility for transplant, International Staging System (ISS) stage I-III disease, and standard- or high-risk cytogenetics.⁴

Methods for evaluating MRD have improved significantly over recent years. New methods such as next-generation sequencing (NGS) and next-generation flow (NGF) have been able to detect MRD up to x10^-6, which has been shown to be clinically relevant. Aurore Perrot and colleagues recently published results that showed patients who were MRD-negative (at the level of x10^-6 by NGS) had a higher probability of prolonged PFS compared to those with detectable disease. This was true regardless of treatment arm, cytogenetic risk or ISS stage at diagnosis.⁵ Studies utilizing NGS and NGF are also expected to demonstrate the superiority of MRD-negative responses more significantly, compared to studies using older, more outdated methods of measurement which may have underestimated the benefit.

Since MM is a heterogeneous disease, imaging is often used in combination with MRD assessment. In a manuscript currently under review for publication by Dr Paiva and colleagues, it was reported that only 7% of patients with undetectable MRD progressed at a median follow-up of 40 months and half of the patient population had extraosseous plasmacytomas at diagnosis and/or relapse.¹ Another study by Leo Rasche and colleagues, published in Leukemia in 2018, investigated whether positron emission tomography (PET) and diffusion-weighted magnetic resonance imaging (DW MRI) results could complement MRD detection at a single site. The study was conducted in patients with newly diagnosed MM (NDMM, n= 83). By combining MRD and functional imaging results, it was shown that patients with the best outcome were MRD-negative and imaging-negative whilst the worst outcomes were seen in patients who were MRD-positive and imaging-positive. In general, MRD-positivity was associated with inferior outcomes (p= 0.03).⁶ Therefore, there is value in combining cellular assessment of MRD, with functional imaging techniques.

The 2016 International Myeloma Working Group (IMWG) updated criteria for measuring MRD, identifies methods that should be used for MRD assessment and criteria for; ‘sustained MRD-negative’, ‘imaging MRD-negative’, ‘flow MRD-negative’ and ‘sequencing MRD-negative’.⁷ Dr Paiva encouraged attendees to use the recommended methods and guidelines, so that studies can be compared to one another based on the consensus criteria.

Dr Paiva added that one of the most important factors when discussing MRD, is how well MRD-negativity is sustained. The impact of sustained MRD-negativity has been demonstrated, even in the complex setting of relapsed/refractory MM (RRMM). In an analysis presented during the annual American Society of Hematology (ASH) meeting in 2018, undertaken by Herve Avet-Loiseau and colleagues, patients with RRMM treated in the POLLUX and CASTOR trials were evaluated for the impact of sustained MRD-negativity. The authors found that patients who sustained MRD-negativity had a longer PFS and OS compared to patients who did not achieve MRD-negativity. The authors of this study recommended durable MRD-negativity should be a treatment goal for patients with RRMM.⁸

Additionally, Dr Paiva noted that since the risk is dynamic, patients can shift from adverse to favorable risk if they can achieve deep responses to treatment (MRD-negative to x10^-6). However, this is also true of patients with favorable risk who can shift to adverse risk. For example, patients with ISS stage I or II at diagnosis may have a PFS similar to that of patients with ISS stage III disease if they are MRD-positive.

Dr Paiva concluded that MRD-negativity should be a new endpoint of trials based on the data that has been presented in the last few years. However, he also cautioned that MRD-negativity must be sustained to have an effect on prolonging PFS.

Should MRD status guide drug development and approval?²

OS is the most clinically meaningful endpoint for clinical trials and drug approvals, however, it requires long follow-up. As such, several other endpoints have been evaluated as surrogates for OS, and are considered for drug approvals. PFS is commonly used as an alternative endpoint since it is appropriate for therapies with a curative intent, and for those which aim to control disease. Response rates such as CR are also considered for treatments with curative intent that involve a sequential approach. Other endpoints considered by the United States (US) Food & Drug Administration (FDA) are disease-free or event-free survival (DFS or EFS) and time-to-progression (TTP). Dr Gay highlighted that the focus of these assessments must be on safety and patient-reported outcomes (PROs), as well as efficacy. Each endpoint has advantages and disadvantages, as shown in Table 1.

Table 1. Pros and cons of different surrogate endpoints

The MM Hub’s coverage of Dr Rajkumar and Prof. Kumar’s article from the Lancet on the use of surrogate endpoints is available here and a summary of the BELLINI trial here.

PFS is widely accepted as a surrogate endpoint for trials and drug approvals, however, with newer regimens and novel treatments, PFS rates are becoming longer, leading to the same issue as observed with OS. For example, in the CASSIOPEIA trial, patients with NDMM treated with daratumumab, bortezomib, thalidomide and dexamethasone (dara-VTd) had an estimated 18-month PFS of 93%.⁹ Therefore, in order to demonstrate an advantage, a long follow-up may be required, particularly with new frontline therapies.

Many studies have shown high-quality response correlates with MRD-negativity, the latter being prognostic of PFS and OS. However, MRD is not currently universally accepted as a surrogate endpoint since it is not necessarily a substitute for patient survival. Dr Gay explained that in order to be a surrogate endpoint, MRD must prove; biologically plausibility (sensitive to tumor cell clearance), specificity (to the treatment being evaluated), proportionality (change in MRD is proportional to change in PFS/OS), universal action (applicable across different treatments and patient populations). To identify whether MRD can be used as a surrogate endpoint, an international, independent team combining MM research groups in Europe and the US, pharmaceutical companies and independent statisticians is conducting the i²TEAMM trial. The group are currently working on >4,000 pieces of patient-level data.

Dr Gay acknowledged several questions that remain open before MRD can be used as an endpoint:

• Which patients should be assessed for MRD status?
• What is the optimal cut-off?
• Which techniques should we use to measure MRD?
• Should we perform imaging in all patients?
• What is the optimal timing for measurement?

Dr Gay concluded that before using MRD routinely, it will be important to standardize measurement methods and definitions as much as possible. Additionally, MRD-negativity alone is not sufficient, rather it is important for MRD-negativity to be sustained. As a compromise, Dr Gay suggested MRD be used as an intermediate endpoint, with this data supplemented by PFS/OS and safety data as confirmatory endpoints. Alternatively, MRD may be used as part of a composite endpoint combining MRD status with safety data.

Should minimal residual disease (MRD) negativity be the endpoint of therapy, in clinical practice?³

Table 2. Summary of debate between Prof. San Miguel and Prof. Jagannath on MRD as an endpoint in clinical practice

On the MM Hub Twitter poll, conducted just after the debate concluded, 60% of voters agreed with Sundar Jagannath, that we should not use MRD negativity as the end point of therapy.

Another poll conducted after a similar debate at the Society of Hematologic Oncology (SOHO) showed 80% of voters agreed that we can use MRD for treatment decisions.

The MM Hub were pleased to conduct an expert debate between Prof. Jeff Wolf and Prof. Tom Martin at the SOHO meeting. The MM Hub asked: “Are we ready to use MRD for treatment decisions in patients with MM?”  This debate is available to watch below.