Luciano Costa: Some considerations on long-term follow-up of the CASSIOPEIA trial
Luciano Costa, a Professor at the University of Alabama at Birmingham, recently shared a post on X:
“Some considerations on long-term follow-up of the CASSIOPEIA randomized controlled phase 3 trial.
1st things 1st: Celebrate! It is not every day that a drug changes OS in NDMM, particularly for young TE patients. It only happened once for Len (in the CALGB maintenance trial) and once for Bortezomib ( GIMEMA-MMY-3006, 15 years after Bort availability).
Daratumumab reduced the risk of Death by 45%, Bortexomib reduced by 32%, and Len ( OS in maintenance) by 38% for this pop. Factually, anti-CD38mAb has the highest known impact in OS. This adds to MAIA and ALCYONE on OS advantage for adding Dara.
IMHO design was clean. Anti CD38 was added for a fixed duration of time, even if at maintenance. More than two/third of patients not receiving anti-CD38 did so at first progression. Best drugs first.
It helps us understand the duration of anti-CD38 mAb. 2 years of maintenance reduced death or progression by 24% when the anti-CD38 was used in induction-consolidation. This is the first time any drug proves on a clean design to impact PFS in maintenance in the setting of QUADs.
Apples with Apples: Len reduced the risk of progression or death by 63%. but in a population where only 34% of patients had received Len (CALGB study). The impact of Dara Mx on non-data-exposed patients was a whopping 66% reduction in risk.
The greatest caveat is that we don’t have Len on Mx. Yet the benefit of indefinite Len maintenance was established at the time there was much more territory to conquer, mostly doublet induction. Its benefit in the QUAD era is a reasonable extrapolation with an unknown effect size.
How would Len alone perform after quad induction + cons? We don’t know. Would Len add anything to Dara in maintenance? Likely, but how much?
Granular answers will come from SWOG1803 and GMMG-HD7 trials. For disclosure, until proven otherwise I believe this to be a class-effect, not unique to Dara. We don’t know the optimal duration for everyone, 2 years reasonable guess based on data.”
Read further
Source: Luciano Costa/X