A framework for setting enrollment goals to ensure participant diversity in sponsored clinical trials in the United States.

BACKGROUND: Diversity in clinical trials (CTs) has the potential to improve health equity and close health disparities. Underrepresentation of historically underserved groups compromises the generalizability of trial findings to the target population, hinders innovation, and contributes to low accrual. The aim of this study was to establish a transparent and reproducible process for setting trial diversity enrollment goals informed by the disease epidemiology. METHOD: An advisory board of epidemiologists with expertise in health disparities, equity, diversity, and social determinants of health was convened to evaluate and strengthen the initial goal-setting framework. Data sources used were the epidemiologic literature, US Census, and real-world data (RWD); limitations were considered and addressed where appropriate. A framework was designed to safeguard against the underrepresentation of historically medically underserved groups. A stepwise approach was created with Y/N decisions based on empirical data. RESULTS: We compared race and ethnicity distributions in the RWD of six diseases from Pfizer's portfolio chosen to represent different therapeutic areas (multiple myeloma, fungal infections, Crohn's disease, Gaucher disease, COVID-19, and Lyme disease) to the distributions in the US Census and established trial enrollment goals. Enrollment goals for potential CTs were based on RWD for multiple myeloma, Gaucher disease, and COVID-19; enrollment goals were based on the Census for fungal infections, Crohn's disease, and Lyme disease. CONCLUSIONS: We developed a transparent and reproducible framework for setting CT diversity enrollment goals. We note how limitations due to data sources can be mitigated and consider several ethical decisions in setting equitable enrollment goals.

Analysis of drug-induced effect patterns to link structure and side effects of medicines.

The high failure rate of experimental medicines in clinical trials accentuates inefficiencies of current drug discovery processes caused by a lack of tools for translating the information exchange between protein and organ system networks. Recently, we reported that biological activity spectra (biospectra), derived from in vitro protein binding assays, provide a mechanism for assessing a molecule's capacity to modulate the function of protein-network components. Herein we describe the translation of adverse effect data derived from 1,045 prescription drug labels into effect spectra and show their utility for diagnosing drug-induced effects of medicines. In addition, notwithstanding the limitation imposed by the quality of drug label information, we show that biospectrum analysis, in concert with effect spectrum analysis, provides an alignment between preclinical and clinical drug-induced effects. The identification of this alignment provides a mechanism for forecasting clinical effect profiles of medicines.

Biospectra analysis: model proteome characterizations for linking molecular structure and biological response.

Establishing quantitative relationships between molecular structure and broad biological effects has been a long-standing goal in drug discovery. Evaluation of the capacity of molecules to modulate protein functions is a prerequisite for understanding the relationship between molecular structure and in vivo biological response. A particular challenge in these investigations is to derive quantitative measurements of a molecule's functional activity pattern across different proteins. Herein we describe an operationally simple probabilistic structure-activity relationship (SAR) approach, termed biospectra analysis, for identifying agonist and antagonist effect profiles of medicinal agents by using pattern similarity between biological activity spectra (biospectra) of molecules as the determinant. Accordingly, in vitro binding data (percent inhibition values of molecules determined at single high drug concentration in a battery of assays representing a cross section of the proteome) are useful for identifying functional effect profile similarity between medicinal agents. To illustrate this finding, the relationship between biospectra similarity of 24 molecules, identified by hierarchical clustering of a 1567 molecule dataset as being most closely aligned with the neurotransmitter dopamine, and their agonist or antagonist properties was probed. Distinguishing the results described in this study from those obtained with affinity-based methods, the observed association between biospectra and biological response profile similarity remains intact even upon removal of putative drug targets from the dataset (four dopaminergic [D1/D2/D3/D4] and two adrenergic [alpha1 and alpha2] receptors). These findings indicate that biospectra analysis provides an unbiased new tool for forecasting structure-response relationships and for translating broad biological effect information into chemical structure design.

Biological spectra analysis: Linking biological activity profiles to molecular structure.

Establishing quantitative relationships between molecular structure and broad biological effects has been a longstanding challenge in science. Currently, no method exists for forecasting broad biological activity profiles of medicinal agents even within narrow boundaries of structurally similar molecules. Starting from the premise that biological activity results from the capacity of small organic molecules to modulate the activity of the proteome, we set out to investigate whether descriptor sets could be developed for measuring and quantifying this molecular property. Using a 1,567-compound database, we show that percent inhibition values, determined at single high drug concentration in a battery of in vitro assays representing a cross section of the proteome, provide precise molecular property descriptors that identify the structure of molecules. When broad biological activity of molecules is represented in spectra form, organic molecules can be sorted by quantifying differences between biological spectra. Unlike traditional structure-activity relationship methods, sorting of molecules by using biospectra comparisons does not require knowledge of a molecule's putative drug targets. To illustrate this finding, we selected as starting point the biological activity spectra of clotrimazole and tioconazole because their putative target, lanosterol demethylase (CYP51), was not included in the bioassay array. Spectra similarity obtained through profile similarity measurements and hierarchical clustering provided an unbiased means for establishing quantitative relationships between chemical structures and biological activity spectra. This methodology, which we have termed biological spectra analysis, provides the capability not only of sorting molecules on the basis of biospectra similarity but also of predicting simultaneous interactions of new molecules with multiple proteins.