Abstract

Toxicity is a major cause of attrition in therapeutic drug development and is a key factor in decision-making around the advancement of candidate drugs through the development pipeline. The expansion of preclinical testing to incorporate assays that better predict potential toxicities earlier in the drug development process has obvious advantages for the selection of successful lead candidates. In vitro testing on primary cells can allow investigators to preview in vivo responses, thus facilitating design of better dosing strategies and optimization of animal and Phase I clinical studies. Primary hematopoietic stem and progenitor cells are routinely utilized in the colony-forming unit (CFU) assay to predict in vivo marrow toxicity and clinical cytopenia. This review summarizes the application of the CFU assay for predicting hematopoietic toxicity and summarizes some key studies that illustrate its utility in this setting.

Introduction

During the standard lead optimization, the drug development process shifts from the utilization of high throughput to lower throughput, more clinically predictive screening assays. Testing for undesirable effects, both in vitro and in vivo, at the earlier stages of drug development helps to identify compounds with unacceptable levels of toxicity. A major challenge in the advancement of candidate drugs from preclinical development into Phase I trials is converting in vivo toxicology results drawn from relevant animal models into meaningful data to predict the response in humans.^1,7

The colony-forming unit (CFU) assay is a predictive test that uses stem and progenitor cells from humans and traditional model organisms such as mice, rats and dogs. This in vitro assay offers a tool to investigate potential hematological disturbances seen in animal models and to design dosing regimens for human studies. The use of primary human cells that exhibit in vivo functionalities in culture allows the collection of information that has a greater likelihood of being clinically relevant.² By utilizing progenitor cells of the blood-forming (hematopoietic) system, the CFU assay can be used to investigate the toxic effects of small molecule compounds and biologics on specific subpopulations of blood cells. This assay can be used to accurately predict drug doses that may induce anemia, neutropenia and/or thrombocytopenia in animal models and in humans.

The CFU Assay in Toxicology

The CFU assay was first developed in the 1960's to identify and quantitate primitive cells at different stages of the hematopoietic cell hierarchy.³ Primary hematopoietic progenitor cells obtained from bone marrow, peripheral blood or cord blood are cultured in a semi-solid medium that restricts their movement, and which supports proliferation and differentiation in situ. Individual clonogenic cells form distinct colonies that can be counted and "scored" on the basis of the type (i.e. lineage) of the cells they contain. In this way, the CFU assay is both quantitative and qualitative in nature.

Drugs of interest are tested in this system by addition to the culture medium at a range of concentrations to determine effects on the in vitro proliferation and differentiation of hematopoietic progenitor cells. By generating a dose-response curve of the total number of colonies as a function of drug concentration (normalized to solvent controls), it is possible to determine in vitro inhibitory concentrations, expressed as IC₅₀ or IC₉₀ values. In addition, changes seen in colony morphology (e.g. size and cellular composition) can be used to predict lineage-specific effects of the drug under investigation. Finally, IC values can be correlated to in vivo measures of toxicity (e.g., maximum tolerated doses, C_max in serum, etc.), to provide critical information for establishing effective dosing strategies both for animal and clinical studies.^2,4

Many drug candidates being developed for oncology, antiviral and other therapeutic indications specifically target rapidly dividing cells, including normal hematopoietic progenitor cells, and consequently induce cytopenia. By investigating the acute effects of agents on progenitor cells of the granulocyte-macrophage (CFU-GM), erythroid (BFU-E, CFU-E), and megakaryocyte (CFU-Mk) lineages in vitro, it is often possible to predict dose-limiting hematotoxicity that may lead to cytopenia in vivo.

The CFU Assay as an In Vitro Predictive Tool

In selecting an appropriate in vitro assay to assist in the prediction of in vivo drug toxicity, several assay attributes should be considered:⁵

The CFU assay can be performed with primary hematopoietic cells from human tissues as well as from other species commonly used in toxicity testing. All of these different cell types exhibit well-defined and reproducible proliferation and differentiation potentials in vitro. The use of qualified media components, extensively screened and validated by STEMCELL Technologies, further ensures the reproducibility of results that can then be compared against benchmark compounds that have been selected on the basis of their recognized hematotoxicity.^2,4

The CFU assay has been shown in several studies to be a reliable in vitro predictive tool for clinical cytopenia. Both the CFU-GM assay and CFU-Mk assay have been independently validated by the European Centre for the Validation of Alternative Methods (ECVAM) for the prediction of drug-induced neutropenia² and predicting the potential for thrombocytotoxicity.^4,8

Table 4. Suggested Prediction Model Based on Comparison Between the C_max and IC Values

The authors of the above studies have concluded that mathematical modeling of IC values derived from CFU assays can be:

1. Used to refine in vivo dosing regimens, thereby reducing the number of animals required for preclinical toxicology studies and informing the establishment of appropriate starting doses for phase I clinical trials.
2. Considered as a validated in vitro model for neutropenia and thrombocytopenia that can be utilized to address a number of preclinical drug development considerations.²

The CFU assay can provide accurate answers to questions that are crucial to the development of effective drug dosing strategies:

• Is the compound toxic to the bone marrow in patients (e.g., does it induce cytopenia)?
• If so, at which dose is toxicity seen and to what extent (e.g., what are the IC₉₀, IC₅₀ values)?
• Is the drug toxic only to specific lineages of blood cells (e.g. does it cause neutropenia, thrombocytopenia and/or anemia)?