The International Council on Harmonisation (ICH) needs to further clarify its addendum on estimands and sensitivity analysis in clinical trials, according to comments released this week from Pfizer, Teva, Alexion and Boehringer Ingelheim.
According to ICH, the addendum, known as ICH E9(R1)
, presents a structured framework to link trial objectives to a suitable trial design and tools for estimation and hypothesis testing. The framework introduces the concept of an estimand, which ICH defines as “the target of estimation to address the scientific question of interest posed by the trial objective,” in addition to aiming to create a dialogue between disciplines involved in clinical trial planning, conduct, analysis and interpretation, as well as between sponsor and regulator, “regarding the treatment effects of interest that a clinical trial should address.”
ICH explains in the addendum how the statistical analysis, “aligned to the estimand, will be associated with assumptions and data limitations, the impact of which can be investigated through sensitivity analysis.”
Boehringer Ingelheim notes in its comment dated 30 April that in general, this estimands framework will add complexity to trial planning and documentation, meaning it’s “important that the material benefits from the framework are sufficient to outweigh the additional burdens.
“Since estimands affect all clinical trials, the estimands framework should be mandatory for all trials, including single-arm and pragmatic trials. We believe that this is the intention of ICH E9(R1), but it is not clear from how it is written; we request that this is clarified. Applying the estimands framework to all trials simplifies implementation, increases compliance and avoids ambiguity around whether it applies,” Joanne Palmisano, VP of regulatory affairs at Boehringer, wrote.
She also raised concerns about how, “No attention is paid to choice of estimand in early-phase trials, and the focus of estimand choice/strategy within the document appears to only be from a regulatory perspective on pivotal trials.”
Both Boehringer and Pfizer also recommended that ICH use the term “post-randomization” rather than “intercurrent” to define events that occur in trials that complicate the description and interpretation of treatment effects.
Pfizer also said that the operational definition of the intent-to-treat principle is unclear in the addendum and the original E9 guidance.
Alexion, meanwhile, stressed that the addendum “provides much needed direction that the statistical analysis will depend on the definition of estimand. It will be important to emphasize that the definition of estimand itself has to be meaningful and interpretable especially in the context of hypothetical strategy.”
And Teva called for more discussion of estimand considerations and complications associated with adaptive trial designs, “for example, when the adaptation is to a sub-population.”