Researchers reviewed reasons for treatment discontinuation among multiple myeloma (MM) patients enrolled in randomized trials and found that the most common reasons for discontinuation were disease progression and toxicity.
However, roughly a quarter of the trials analyzed had imbalances in reasons for discontinuation between the intervention and control arms.
These differences — particularly those seen in trials that met their primary endpoint — raise questions about the possibility of informative censoring and bias, according to researchers.
Continue Reading
The researchers reported these and other findings from the review in the European Journal of Haematology.
The review included 45 randomized controlled trials (RCTs) that encompassed 21,236 patients with MM. About half of these patients (47.8%) had discontinued study treatment by the time of endpoint assessment. Reasons for discontinuation included disease progression (22.6%), toxicity (12.1%), patient/physician withdrawal (5.7%), and death (2.3%).
Discontinuation due to progression occurred in 18.9% of patients assigned to intervention arms and 22.8% of those assigned to control arms. Discontinuation due to toxicity occurred in 9.5% and 10.6%, respectively, and discontinuation due to death occurred in 2.4% and 1.9%, respectively.
About a quarter (24.4%) of the trials had a greater than 5% absolute difference in discontinuation rate for reasons other than death, progression, and toxicity between patients in the intervention and control arms.
As an example, the researchers highlighted the BOSTON study, which was designed to compare bortezomib, selinexor, and dexamethasone (XVd) to bortezomib plus dexamethasone (Vd). Patients in the XVd arm were more likely than patients in the Vd arm to stop treatment due to patient or physician withdrawal (22.1% and 12.5%, respectively).
“Within RCTs, if patients are withdrawn, they are assumed to have a similar risk of progression or death as the patients who continue on the trial for Kaplan-Maier survival analysis,” the researchers wrote. “However, this assumption may be incorrect, as patients withdrawing may be more likely to progress or die due to various reasons such as poor tolerability of treatments or high perceived risk of staying on trial treatment. This phenomenon, known as informative censoring, is prevalent in oncology RCTs and needs further evaluation in MM RCTs, given our findings.”
“Although publication bias may affect the results, the demonstration of a trend where these phenomena were present mostly in trials that met their primary endpoint is important and needs further investigation,” the researchers added. “We also observed that, in these trials, attrition was more common in the control arm rather than the intervention arm. This is a finding worthy of further investigation, as it may significantly influence results in either direction.”
Disclosures: Some study authors declared affiliations with biotech, pharmaceutical, and/or device companies. Please see the original reference for a full list of disclosures.
Reference
Hentzen S, Meirson T, Koehn K, et al. Attrition and withdrawal in multiple myeloma randomized controlled trials: A systematic review. Eur J Haematol. Published online June 29, 2023. doi:10.1111/ejh.14032
This article originally appeared on Cancer Therapy Advisor
