Safety Assessment
Lauren Black


An Investigational New Drug application (IND), the first step toward clinical testing of a new medicine, will contain data from numerous preclinical studies to support a sponsor’s plan to test the drug in humans. It is important that the species selected and the doses used for each of these studies make scientific sense. This requires that the studies be conducted with a programmatic, mindful objective guiding the research plan from the outset.

Mindful pharmacology (drug action) and toxicology (hazard assessment) studies fit together like a complete puzzle – and like basic research, later studies rely on earlier data gathered on the drug’s nature and mechanism. At the end of a preclinical program, the data should fit together and the drug’s actions make sense. There should be a clear list of doses and what happens at each dose. In each species that mimics human response, there should be a clear correlation between dose and drug concentration in the blood, expected activity and toxic effects.  All of this information will allow an initial human dose to be selected that will not cause harm to the patient and subsequent doses to be evaluated for efficacy with minimal and/or controllable toxicity (side effects).

Reducing Studies to the Ones that Count in a Program
With an objective-driven, mindful effort in the first stages of preclinical research – pharmacology — we can focus the toxicology portion of the program’s scope of work specifically on the studies that will really count in risk assessment. This allows a reviewer to clearly see the dose levels that are toxic and dose levels that appear to be safe.

Right Species: The R of human “Relevance:”

  • Make sure the drug’s action is expressed in the animal. This is a major concern for the manufacture of biologics, but is also becoming increasingly important for all drugs. In guidances regulators state that the species chosen for toxicology should be “relevant” – meaning predictive of human responses to a drug.

The choice of this species for toxicology studies begins with the planning for pharmacology studies.   A limited number of animals is needed for pharmacology studies, but these studies have to be specifically designed to look for drug action at receptors, examine where receptors are expressed in tissues, and (in the case of in vitro studies) show what cellular actions are caused by the receptors across all species.

If pharmacology studies are not properly designed, not only would the animal studies provide misleading data about how a drug works, but animal lives would also be wasted in toxicology research that shouldn’t have been conducted in the first place.  Relevance of the species used in pharmacology studies is also necessary because without it, toxicology studies could find that a drug is safe, when in actuality it just has no place to act in the animal. This terrible situation has actually happened, with disastrous clinical trials results (internet search “Tegenero 2006”):

  • For small molecule programs, we need to know from the beginning that the drug’s human metabolites are made by the animal. Without that knowledge, the species could be chosen incorrectly and, when caught by regulators, the sponsor would be placed on clinical hold – needing to restart the program again in the “right” species (one with the metabolite). When FDA requires a clinical hold with a new species, the sponsor’s time and money are wasted, along with the lives of the animals used in the first round of studies. By avoiding these oversights before a GLP program starts, animal lives – program after program – are saved.

The R of “Right Doses”
Any drug can be toxic. Drugs are designed to cause structural or functional changes in the body. They have their good dose levels and also dose levels that can be harmful (usually those that cause high blood levels).

Toxicity can be caused in two different ways. It can arise from a harmful degree of the action that is actually wanted (pharmacologic toxicity) or from side-effects. The term “side-effects” means that we don’t understand the mechanisms that led to functional and structural harm (or in some cases, mortality). Many of us have experienced these harmful actions– one time I took a double dose of Sudafed ® for a cold by accident and got the well-known heart palpitations and dizziness. Quite scary. This is direct action of the drug — too much stimulant action at its receptors on nerves.

FDA doesn’t need animal deaths to approve an IND and select the safe range of human doses for first-in-human trial; they need only to know about the side effects and doses that appear to be safe. For example, they need to know the lowest dose at which we start to see glucose dropping in the blood after taking a drug for diabetes, or where we start to see an anticoagulant action cause a prolonged bleed at an injection site.

When studies are designed with prior knowledge of earlier studies, it helps reduce animal use. The earlier studies needed (especially for all new drugs and novel formulations) of are termed “pilot,” or “dose-ranging” studies. In these small program starter studies, acute and repeat doses of a drug might be studied in between two and 10 animals before larger GLP studies start. These animals are treated like patients going for a physical – blood chemistries and observations are used as important indicators. These animals help assure that the dose choices are right for the later and larger GLP studies. In fact, three animals studied carefully can save several later on.

Conserving animals through proper reduction strategies, while also protecting humans in the long run, can be a difficult balance. However, through a strong commitment to the 3Rs, and by practicing some of these tips in early planning, we can work as a team – sponsors and lab scientists –  to help reduce animal use by discussing drug action at receptors or through other mechanisms, and then designing stepwise, mindful preclinical programs.

We can often reduce animals in a program at least 30% by incorporating endpoints into ongoing studies (like combined safety pharmacology, genotoxicity, and immunotoxicology), and by conducting appropriate dose-ranging pilots that employ enhanced clinical pathology and observations to identify doses that are clearly toxic and/or identifying biomarkers that could be monitored in humans. This can reduce the cost of research and/or allow us to learn more about the drug. In both cases, the number of animals’ lives involved is reduced while at the same time increasing drug development efficiency.