The use of overall survival (OS) as a primary endpoint in clinical trials is important to estimate the efficacy of a drug. However, this endpoint is inevitably influenced by the choice of subsequent treatments, and the salvage of patients in later lines of therapy can mean that determining the true efficacy of an experimental treatment is complicated. Alternative endpoints are needed to better compare experimental treatments, and can also speed up approval of effective drugs, by acting as a proxy for OS or progression-free survival (PFS).

Measurement of minimal residual disease (MRD) has been performed for many years in a number of hematological cancers, including acute myeloid leukemia (AML), acute lymphocytic leukemia (ALL), chronic myeloid leukemia (CML), and multiple myeloma, to assess patients’ prognoses and guide treatment decisions. MRD is now increasingly being utilised in other indications, including in patients with chronic lymphocytic leukemia (CLL) and some types of non-Hodgkin’s lymphomas. MRD positivity is associated with poorer patient outcomes, including an increased likelihood of relapse. Patients who are MRD positive following treatment with curative intent will typically receive more aggressive therapy. However, there is still a higher likelihood that these patients will eventually relapse compared to MRD-negative patients.

A meta-analysis conducted by Avet-Loiseau and colleagues looked at six randomised studies, encompassing 3,283 newly diagnosed multiple myeloma patients, and correlated MRD negativity with PFS. In a larger study, presented at the European Hematology Association’s annual meeting in July of this year by Mamolo and colleagues, MRD was found to strongly correlate with median PFS in a large sample of newly diagnosed multiple myeloma patients, supporting the use of this endpoint as a proxy for PFS. In another study, Short and colleagues found MRD negativity was correlated with better survival outcomes, as measured by OS and disease-free survival, in a meta-analysis of 81 publications, corresponding to 11,151 patients with AML. Earlier this year the FDA agreed with Kronos Bio that an upcoming Phase III trial of entospletinib in newly diagnosed AML patients with NPM1 mutations can use MRD negativity as a primary endpoint. This will be the first Phase III trial to utilise this endpoint and could represent a major shift in how clinical endpoints are assessed.

While there is increasing use of MRD in clinical trials, this is not without issue. The testing for MRD varies between treatment centres, and particularly between academic and community centres, both in terms of the types of test performed as well as sensitivity, and it will be important to standardise these testing regimens in the future. This is also not a perfect predictor, as there will be a proportion of patients who are MRD negative after treatment but who will then go on to relapse. This may be due to testing being insufficiently sensitive to detect residual disease. It is also possible that certain genetic changes predispose these patients to poorer outcomes. More research is needed on the latter in order to improve predictive patient outcomes based on the use of MRD alone. While OS is the most important outcome to many physicians and patients alike, the assessment of OS and PFS takes five to ten years or more and large sample sizes, which ultimately means these measures are not always practical. The use of MRD negativity as a surrogate endpoint has the potential to speed up the approval of new therapies, and is likely to be used increasingly in the future.