Transparent Trade-Offs - Pharmaceutical Executive


Transparent Trade-Offs
A clinical utility index (CUI) openly evaluates a product's attributes—and chance of success

Pharmaceutical Executive

Information Flow
Metrics and methods When modeling prescribing physician behavior, a CUI is similar to models used in later clinical development (Phases III and IV). Both CUI and conjoint analysis attempt to weigh the relative value of treatment attributes and to capture that information in a single metric for comparison of competing treatment options. A CUI differs from conjoint and discrete choice analyses in its source of information and cost. Tools such as discrete choice analysis actually survey a large number of physicians asking them to choose between hypothetical product profiles. The trade-offs among elements of the competing product profiles are inferred from physician responses. A similar tool goes by the name of "conjoint analysis." Both tools use slightly different methodologies to arrive at the same empirical result and a large number of physicians must be interviewed to obtain a meaningful model, leading to high cost.

Product position CUI, in addition to lower cost, offers specific benefits when used in lieu of or as an early measure of physician prescribing behavior. Construction of a CUI leads the development team to identify those product attributes critical to the patient. A CUI focuses development team efforts on exploring the most valuable region of treatment—not necessarily the most efficacious—and the best market position. Further, CUI can be used to examine differences between internal team opinion and outside physician input. Multiple CUIs can explore the potential influence of non-prescriber decision makers, such as the FDA or third-party payers.

Clinical Impact
Practical Tool In clinical practice, the patient's decision maker is generally the prescribing physician. Ideally, the development process will result in drugs that are just what the prescribing physician ordered. The framework for the CUI is elicited from the project team. Attribute utilities and weights are determined based on physician preference data, if available, or internal expert opinion. When combined with models of what the body does to the drug (pharmacokinetics) and what the drug does to the body (pharmacodynamics), the CUI is a tool for understanding the benefit of the drug to the patient, relative to competing therapies.

Because these models also include the uncertainty in the drugs effects, they can also be used to predict the outcomes of proposed development trials. As a development program progresses, ongoing or completed trial results provide learning and models are updated. Model output can be broken into categories that show the chance of achieving a given product profile and CUI, given what is known about the drug. The chances of different product profiles can, in turn, be linked to financial models for decision making.

Such models have been used successfully to support better choices in clinical development, as illustrated in the following case studies. Each is derived from a real project conducted for a client in the pharmaceutical industry. Indications and other information may have been blinded to protect confidentiality.

Case Study: Efficacy vs Side Effects CUIs needn't be complex to prove useful. A simple index, constructed using a penalized dose-response model, was used to explicitly value trade-offs between increased efficacy and side effects for a new chemical entity (NCE) with a novel mechanism of action. This index formed the basis for the choice of best dose, which was in turn used in simulations to optimize trial strategy for a phase II dose-ranging trial.

The compound was a psychiatric medication with a novel mechanism of action and little prior information was available to inform expectations for efficacy or side effects. Phase I safety trials and one inconclusive Phase IIa trial had been completed, leaving high uncertainty in efficacy and dose-response for the NCE.


blog comments powered by Disqus

Source: Pharmaceutical Executive,
Click here