Toxicity Outcomes in Phase 1 Interventional Studies in Plasma Cell Dyscrasias: Heterogeneous Reporting and High Usage of Minimising Language

Introduction

Toxicity is the main outcome of interest in Phase I trials and informs dose and drug development. Phrases such as ‘tolerable’ or ‘manageable’ are frequently employed to describe toxicity. Reporting is not standardised, with no mandatory minimums of reporting for adverse events in conference presentations or publications, and no uniform criteria for definitions of a ‘tolerable’ therapy stipulated.

Methods

Toxicity reporting was reviewed in published Ph 1 myeloma/AL amyloidosis abstracts from 3 global conferences from 2022-24 (ASH, EHA, ASCO). Details collected included study type (industry vs Investigator-led), toxicity reporting methods, grades, rates and use of descriptive language.

Results

One hundred and twelve abstracts reporting results from Ph I trials were identified. Forty four % were first-in-human, Ph I or Ph Ib; 53% were Ph I/II and 79% were industry sponsored. Abstracts reported preliminary or interim results in 57%. Typical of MM studies, median age was 67 years and 58% were male, with a mean 5 prior lines. One hundred and eight studies enrolled only myeloma patients, 3 AL amyloid, and 1 enrolled both. Most common therapeutic classes were bispecific T-cell engagers (38%) and cellular therapies (16%).

Seventy six trials (68%) reported all-grade toxicity, 88 (79%) >grade 3 and 61 (54%) both. Seventy three reported (65%) adverse events of special interest (AESI), 16 (14%) serious adverse events (SAEs) and 33 (30%) deaths on study. Two trials reported no toxicity outcomes. Eighteen of 76 trials reporting all grade toxicity (24%) included events occurring above a set frequency, 47 (62%) reported the most common events, and in 11 (15%) it was not stated. For 88 of 112 studies reporting ≥grade 3 toxicities, 13 (15%) recorded events above a set frequency, 30 (34%) the most common toxicities and 45 (51%) did not provide this information. Sixty eight (61%) trials specified treatment-emergent or treatment-related nature of AEs, 30 (27%) reported any toxicity-related treatment discontinuations, and 13 (12%) reported dose reductions.

All grade toxicity and >grade 3 toxicity were more likely to be reported in FIH, Ph I and Ib studies compared to Ph I/II studies (All 88% vs. 51%; ≥grade 3 88% vs. 70%; p<0.001). All grade toxicity was reported more frequently in studies of T-cell engagers compared with cellular therapy (93% vs. 28%; p<0.001), and ≥grade 3 toxicity was reported more often in later analyses versus preliminary/interim ones (96% vs. 66%; p<0.001).

AESIs were specified with higher frequency in FIH I and Ib compared to Ph I/II trials (80% vs. 48%, p<0.001), and cellular therapy or T-cell engager trials compared to other agents (92% vs. 32%; p<0.001). AESIs were predominantly cytokine release syndrome (CRS) and immune cell associated neurotoxicity syndrome (ICANS) in 55 studies (75%). A further 11 (15%) and 9 (12%) trials reported keratopathy/ocular symptoms and dysgeusia/nail/skin toxicity, relating to the use of the antibody drug conjugate belantamab mafodotin and GPRC5D-targeting agents respectively.

Eighty one abstracts (72%) used minimising language. The most common terms were ‘well tolerated’, ‘manageable’, ‘safe’ and ‘acceptable’, (37%, 35%, 13% and 12% respectively). Minimising terms were used more frequently in abstracts reporting preliminary/interim analyses (84% vs. 61%, p = 0.01) compared with other timepoints and in investigator-initiated trials versus industry-sponsored trials (92% v 67%, p=0.02). Thirty one studies including minimising language reported all grade toxicity, of which 80% reported AEs in >70% of patients, and 94% in >50%. Forty three studies including minimising terms reported >grade 3 toxicity rates, of which 60% reported >grade 3 AEs in >70% of patients, and 65% in >50%. Treatment discontinuation due to AEs was documented in 25% of studies using minimising terms, and deaths were reported in 20% of these abstracts.

Conclusions

Toxicity reporting in Ph 1 studies demonstrates marked heterogeneity. Serious adverse events and causes of death are reported in a minority of trial abstracts. Minimising language is used in the vast majority despite significant rates of adverse events, treatment discontinuation and even deaths on study. A standardised approach to reporting of safety outcomes at international conferences is desperately needed to improve the transparency of information relayed to the myeloma community regarding novel drugs.