Study endpoints

Additional objectives of clinical trials conducted in Phase II may include evaluation of potential study endpoints, therapeutic regimens (including concomitant medications) and target populations (e.g., mild versus severe disease) for further study in Phase II or III. These objectives may be served by exploratory analyses, examining subsets of data and by including multiple endpoints in trials. 

The experienced reproductive toxicologist understands the normal variability of the various study endpoints, e.g., the prevalence of clusters of cleft palate in the control population of mice. He is also familiar with the relative sensitivity of the various parameters to toxicological insult, e.g., frequent reactive increases in liver weight due to metabolic activation versus the relative stability of brain weight, which is rarely influenced by systemic toxicity. An understanding of the relationship between the endpoints is also important, e.g., an increase in the incidence of morphological abnormalities is more likely to be... 

Independent blinded raters are health professionals from a different department or from outside, with no information about details of the treatment studied. Their involvement in the trial is limited to providing ratings for the primary study endpoint(s). 

Well-designed CRFs capture all essential scientific and regulatory information and do not capture information that is not needed. Collected data include those related to study endpoints, adverse events (AEs), potential confounding influences, and protocol compliance. 

Assumption 5 states that the endpoints measured in the toxicity tests on which the SSD is based must be ecologically relevant. Mortality is the most frequently studied endpoint in laboratory tests. In chronic tests, endpoints such as reproduction and inhibition of growth are also studied. Forbes et al. (2001a) argued that individual-level endpoints like survival, fecundity, and growth may not reflect effects at the population level (Assumption 5). They recommended that additional consideration be given to the relative frequency of different life cycle types, to the proportion of sensitive and insensitive taxonomic groups in communities, and to the role of density-dependent influences on population dynamics (see also Forbes and Calow 2002). 

The following should be kept in mind when a treatment population and the study endpoints are defined ... 

The flaws in this study are that only a small group of patients has been studied, and that pregnancy has not been taken into account as a study endpoint. 

In selecting molecular descriptors upon which to base a QSAR model, one needs to know the role that these parameters play, either the way the chemical behaves or the way the studied endpoint is expressed [4], While 3D molecular descriptors are now widely used in drug design [12], most QSAR models in environmental sciences are derived from 2D molecular descriptors that are broadly classified into three main types (Figure 23.2). 

Study (Statin) N FU (years) Polymorphism Studied Endpoint Placebo Event Rate Statin Event Rate P... 

The algorithm of an acceptable QSAR model, that correlates the studied endpoint with chemical structure using molecular descriptors, must be reproducible and easily applicable, even by non-experts. For this reason, there is a... 

It has been demonstrated that none of the descriptors can independently explain the observed distribution of the biological data, it is only the combination of selected descriptor sets that allows the modelling of the studied endpoint. Describing the different mechanisms simultaneously makes it difficult to interpret their role in the model. Some descriptors are more obvious and easily interpretable, while others are more difficult to interpret, but their role is in ensuring the high overall predictive power of the models. ... 

Assay sensitivity is the ability of a trial to detect differences between treatments if they exist. There are many potential reasons to be concerned about assuming assay sensitivity, such as lack of compliance with the study medications, inclusion of inappropriate patients in the trial, and poor assessment of the study endpoints. For example, if a large fraction of subjects in each treatment arm failed to take their study medication, then the results for the two treatments might appear similar, resulting in a conclusion of noninferiority, even if the experimental treatment were in fact inferior to the control. This is just as much a concern in safety trials as in efficacy trials. 

To calculate QALY, we need to estimate the life years of the patient population based on clinical trial data and the health states and their durations during the life years. Health states are usually related to the study endpoints, for example, disease progression or deaths. Multistate analysis can be used to make these estimates. Once the health states are determined, the value or QoL (Q) associated with each state can be established using expert opinion, QoL data collected in the clinical studies, or direct or indirect research. The most frequently used method in practice is direct or indirect research by way of preference survey of health providers and patients. Time trade-off, standard gamble, and rating scale are among the ways to assess the preference of specific health state. As an alternative, indirect research uses questionnaires for health state in several health domains or attributes (e.g., EQ5D) and then to construct a multiattribute utility as a summary measure that reflects preferences both within and across health domains. 

The size and hence geographical distribution of investigational sites needed for the large cardiovascular safety outcome trials discussed in Chap. 13 mean that there can be a considerable degree of variability in the identification of cardiovascular endpoints of interest since classification of events as actual study endpoints is a partially subjective process based on the application of a complex set of medical endpoint criteria to an often complex clinical event. Regulatory agencies therefore require the centralized adjudication of these events to control for the impact of this variability in identification and thereby to generate data for use in statistical analyses that are as standardized as possible. 

---