Before a new therapy is approved for use in routine treatment of cancer, it is required to undergo a series of clinical trials to demonstrate its safety (the amount and type of side effects) and efficacy (how well the treatment works to keep the cancer from growing).  Traditionally, studies are first done to evaluate a safe dose of the chemotherapy, and then studies are done with larger number of patients to evaluate the efficacy of the treatment.  However, even early studies focused on side effects and safety include an emphasis on evaluating preliminary signs of efficacy.  Similarly, later efficacy trials continue to monitor safety to ensure that side effects remain at a reasonably low level.  In order to relay the primary goal of the therapy, the trials are categorized into Phases and numbered as Phase 1, 2, 3 or 4, or their Roman numeral versions.  This terminology is important to understand when evaluating potential clinical trials.

Phase I:   These studies are the first performed with the goal of evaluating safe dose.  In the most common form of this type of study, only a few patients at a time are typically enrolled in order to fully evaluate the possible side effects before moving on to a higher level.  If a predefined treshold of toxicity is exceeded, then no further dose increases are done, and the highest safe dose is used in future studies.  These Phase I can be in several variations:

Phase I, First-in-human:   These are trials using just one treatment with a new therapy not previously administered to patients.  As the safe dose is not known, these typically start at dose levels well below what will eventually be studied in subsequent trials.  However, with many targeted therapies in phase I trials, it has been shown that there is no difference in likelihood of benefit when given a high or low dose.  Animal studies may predict some toxicities, although other unexpected toxicities may be possible.  Typically, patients with many different types of cancer are included in this type of study.

Phase I, one new therapy combined with approved therapies:   In this variant of the phase I study, a new therapy has completed first-in-human phase I testing and the side effects and dose have already been defined, but the new therapy is now being combined with other treatments (usually with agents approved by the FDA for routine use).  Because the safe dose range of each of the agents given separately is know, the starting dose levels are usually close to what will eventually be used in subsequent trials.  Likewise, there is less potential for unanticipated side effects, although this remains a possibility.  These type of phase I study may limit enrollment to patient with a particular type of tumor (usually one where the approved therapies are commonly used).  In a variant of this, phase I studies can be done with therapies which are all approved for use, but for which this combination has not been previously performed.

Phase II:   Once the best dose is identified, a larger study is performed in particular tumor types in order to evaluate the efficacy of the therapy.  All patients are treated with the same dose of the therapy and all patients get the new drug (with the exception noted below).  Typically there is some preliminary evidence of efficacy in Phase I study or a strong laboratory rationale for this particular therapy to be efficacious in this tumor type and more data is required.  Most Phase II studies will have specific requirements for the type of cancer and the number and type of prior chemotherapy regimens.  Many of these studies have minimum levels of efficacy and safety that must be met to complete enrollment of the full planned number of patients (sometimes referred to as interim analyses or "early stopping rules").  In the most common approach, efficacy is evaluated in half of the total number of patients, and enrollment is only continued if there is sufficient signs of benefit to the patients.  It may be useful to know if a Phase II study has passed an interim analysis. 

Phase IIB:   The goal of this study is still to obtain preliminary evidence that the treatment works to control the cancer.  When the novel therapy is given in combination with approved chemotherapy, it can be difficult to know what level of benefit is due to the addition of the novel therapy.  In a typical version of this design, a random half of the patients get the novel therapy and approved chemotherapy and the other half get the approved chemotherapy alone.  This is also called a randomized Phase II.  Unlike the phase III below, a smaller number of patients are enrolled and it is not designed to provide definitive evidence of efficacy for the novel therapy. 

Enrichment Phase II:   In the typical version of this new study design, only patients whose tumors have a particular molecular characteristic (called a "biomarker") are enrolled in the study.  This is usually done for therapies with a well-defined mechanism of action and for which there is strong evidence in the laboratory that the molecular biomarker identifies a type of tumor most likely to benefit.

Phase III:   If there is sufficient evidence of activity in the earlier studies, a definitive study will be performed that, if positive, will lead to approval of the novel therapy.  These studies are randomized by definition, meaning that only a portion of the patients received the novel therapy while the remainder receive an inactive placebo.  The consent documents and study staff will make it very clear that there is a placebo involved in the design of a given trial.  Typically neither you nor your doctor will know whether a patient is receiving the new therapy or a placebo.  These studies are typically performed at many different sites at the same time and involve hundreds or thousands of patients. 

Phase IV:   These are studies done after the approval of the novel therapy to monitor side effects or more robustly evaluate the level of activity of the therapy.  These are uncommon.