Investigators have used a population-based cohort of patients with presumed monoclonal gammopathy of undetermined significance (MGUS) to create a prediction model for smoldering multiple myeloma (SMM) or worse that accurately identifies those patients who can reasonably defer bone marrow sampling and referral to hematology. Their study provides an evidence-based approach to selecting patients with presumed MGUS for whom bone marrow sampling may be deferred, potentially decreasing the number of individuals exposed to the invasive procedure. Elias Eythorsson, MD, PhD, of the University of Iceland in Reykjavik, and colleagues published their findings in Annals of Internal Medicine.
“A diagnosis of MGUS represents a dilemma,” Dr. Eythorsson said. “It is an asymptomatic precursor to MM and related diseases, but the majority of individuals with MGUS will never progress to these diseases. In this sense, MGUS is not itself a disease but rather a risk factor for disease.” However, MGUS has a high risk of progression and requires specialist management, and a bone marrow sample is required to distinguish between MGUS and SMM, which is determined by 10% or greater plasma cells.
To address this problem, the investigators created a predictive model using data from 1,043 patients with MGUS in Iceland. The model leveraged laboratory parameters commonly available at the time of diagnosis of presumed MGUS: MGUS isotype, myeloma (M) protein, free light chain ratio, and total concentrations of immunoglobulin (Ig) G, IgA, and IgM. The researchers calculated that the c-statistic for SMM or worse was 0.85 (95% CI 0.82-0.88), and calibration was excellent (intercept, -0.07; slope, 0.95). At a threshold of 10% bone marrow plasma cells (BMPC) predicted risk for SMM or worse, sensitivity was 86%, specificity 67%, positive predictive value 32%, and negative predictive value 96%.
When the researchers compared the predictive capability of the new model to the predictive capability of the Mayo Clinic model, which is currently used to predict the progression of MGUS, they found that the net benefit for sampling was between 0.13 and 0.30 higher over a range of plausible low-risk thresholds — or the equivalent of avoiding between 130 and 300 additional procedures per 1,000 patients. The researchers’ calculator thus accurately predicted the presence of 10% or more BMPCs, the criteria for diagnosing SMM or MM with bone marrow samples taken from individuals with presumed MGUS. The calculator also outperformed the current standard method of determining whether a patient would benefit from bone marrow sampling. Because the model was based on patients in Iceland, the authors noted that it will need to be externally validated in other screened and clinical cohorts to ensure generalizability.
In their paper, the authors recommended that instead of basing bone marrow sampling recommendations on risk groups, recommendations should be based on each individual’s predicted risks as determined by the multivariable model. Some individuals, Dr. Eythorsson explained, might be willing to undergo the procedure even though the probability of finding at least 10% BMPCs is very low. Others would be willing to do so only if the probability is high.
Although it is safe, bone marrow sampling can be painful, is associated with rare complications, and is only performed in specialized centers, Dr. Eythorsson said. Moreover, the samples must be examined by a hematopathologist. The model could facilitate shared decision-making and incorporate an individual’s risk tolerance when deciding whether to undergo bone marrow sampling. Dr. Eythorsson explained that the calculator is beneficial not only for the individual but also on the societal level, where it could be used to prioritize waiting lists. He suggested that professional societies may also want to use it to provide guidance on which predicted risks and benefits of bone marrow sampling might outweigh costs.
Any conflicts of interest declared by the authors can be found in the original article.
Reference
Eythorsson E, Rognvaldsson S, Thorsteinsdottir S, et al. Development of a multivariable model to predict the need for bone marrow sampling in persons with monoclonal gammopathy of undetermined significance: a cohort study nested in a clinical trial. Ann Intern Med. 2024;177(4):449-457.
