The major challenge in entering clinical development is translating in vitro and preclinical data into a clinical regimen. Knowing a clear scientific profile from preclinical data is crucial to design an adequate first trial. Preclinical models include tumor models, transgenic models, tissue data, and cell investigations. The goal is to know how a drug is acting: what are the primary targets and what may be secondary targets that could become safety concerns or produce inadvertent effects.

Develop a solid trial design. In a Phase 1 study, the primary endpoint will be safety. Secondary endpoints might include measures to evaluate effectiveness or narrow an indication or target population. Clinically meaningful endpoints such as progression-free survival take a long time to measure and are rarely used in Phase 1 studies. Shorter-term endpoints can include surrogate parameters of efficacy, such as activity of a kinase or amount of drug binding to excised tumor cells. Other diagnostic variables may indicate something about the activity of the drug before clinical effects are seen. For example, in the case of tumor vaccines, immune responses deemed to be a mandatory response for any later clinical response can be detected as a valid surrogate endpoint far before an effect on a tumor can be discerned.

Although later phase trials must have strong power analysis to scientifically prove the value of a therapy, early phase trials, especially phase 1 studies, do not have such a consistent need for power analysis or justification of sample size.

Dosing is an important consideration in study design. The goal is to start at lowest therapeutic dose possible to avoid having cohorts that will not yield useful information. Furthermore, because of regulatory and scientific considerations it is useful to determine the minimal effective dose of a new agent.