P-475: Are we downplaying toxicity with our choice of words? A systematic review of randomized myeloma trials.

Introduction: Advances in multiple myeloma (MM) treatment have led to improved survival, but most patients still require long-term therapy. Treatment can result in burdensome toxicities and profoundly affect patients' quality of life. In seeking to emphasize the efficacy of tested treatments, reports of clinical trials may use subjective, vague, or otherwise minimizing terms to downplay adverse events (AEs).

To evaluate how AE patterns are reported, we conducted a systematic review of MM randomized controlled trials (RCTs) from 2015-2023. We sought to estimate the prevalence of the use of minimizing terms in MM RCTs and to identify and explore the characteristics of studies that use these terms.

Methods: We searched MEDLINE/PubMed, Embase, and the Cochrane Registry of Controlled Trials to measure the prevalence of minimizing terms in MM RCTs. Our inclusion criteria were all published MM RCTs from 2015-2023. The Chi square test was used to calculate differences in proportions.

“Minimizing terms” were defined as subjective terms used to describe the safety profile of the intervention. Terms searched included “convenient”, “manageable”, “acceptable’, “expected”, “well-tolerated”, “tolerable”, “favorable” and “safe”.

Results: A total of 65 RCTs were included. Among these trials, 86% employed minimizing terms when describing treatment toxicities. The most frequently used terms were "well-tolerated" or "tolerable" in 45% of studies, "manageable" in 28% and "acceptable" in 25% of studies.

The rates of grade 3/4 adverse events in the examined RCTs ranged from 23.1%-94% with a median of 74.7%. Minimizing terms were used in 88% of study reports with more grade 3/4 AEs than the median (i.e., >74.7% high grade AE rates), compared to a similar frequency (85%, X2 (1, N = 65) = .082, p = .774 (>.05)) in studies with fewer grade 3/4 AEs than the median. The rates of grade 5 adverse events ranged from 0%-19% with a median of 3%; minimizing terms were used in 100% of studies with more grade 5 AEs (deaths) than the median (3%), which was similar (82.1%, X2 (1, N = 65) = 3.606, p = .058 (>.05)) in studies with fewer than median grade 5 AEs. This suggests that the use of minimizing terms was not associated with the actual incidence of grade 3/4 toxicities or of treatment-related deaths.

Industry-sponsored studies were more likely to use minimizing terms: 37/39 industry-sponsored studies (95%) vs 19/26 non-industry/cooperative group studies (73%) used minimizing terms, X2 (1, N = 65) = 6.2, p = .013. A minority of studies (25 of 65, 39%) reported patient reported outcomes (PROs). Studies that reported PROs used minimizing terms less often (18 out of 25, 72%) compared to those that did not report PROs (38 out of 40, 95%), X2 (1, N = 65) = 6.8, p = .009.

Conclusions: Trial investigators and sponsors regularly use minimizing terms to describe toxicity in MM trials, especially in industry-sponsored studies, and this descriptive terminology does not reflect the actual rates of AEs or deaths in these studies.