Clinical Trials Common Terminology Randomization Chance allocation to

Clinical Trials

Common Terminology • Randomization – Chance allocation to one of two or more treatments – To insure comparability of groups on factors (known and unknown)other than the treatment – To control bias – To allow valid application of statistical methods

• Placebo- an inactive substance given to randomly selected patients in a clinical trial so that all patients and investigators believe that they are receiving treatment. • Bias-systematic error introduced into sampling or testing by selecting or encouraging one outcome or answer over others

Blinded Trials • Single Blind- The subject is not told which group he/she is in. • Double Blind- Neither the subject nor the investigator knows which treatment the patient is receiving. • Triple Blind- Neither the patient nor the investigator nor the clinical personnel (monitors, statisticians, etc. ) knows which treatment the patient is getting.

Phase 1 Studies • Exploratory • Primarily concerning safety • Dose-ranging or dose-seeking (route and schedule too) • Define maximum tolerated dose • Define toxicities • Usually small number of people-approximately 5 -15

• Usually performed on healthy individuals • Usually one institution involved

Phase 2 Studies • Confirmatory • Demonstrates safety • Should already have dose, route, schedule from phase 1 trial • Obtain preliminary evidence of efficacy to support phase 3 trial • Performed in sick people • Usually one institution

Phase 2 • Usually more patients than in phase 1 studies. . usually 10 -100 • Meet with FDA at this point to present safety and efficacy data and to present plans for the future definitive Phase 3 trials; FDA gives input. .

Common Problems • Lack of good controls—can weaken study • Poorly defines hypotheses and outcomes • Inadequate sample size which can translate into inadequate statistical “power” and/or precision; more common in Phase 2 trials than in Phase 3 trials • Noncompliance • Results published too early or not published at all (usually negative results)

Phase 3 Studies • Exploratory • Comparative • Required by FDA to demonstrate efficacy-the definitive study or studies upon which the potential approval is based • Almost always randomized • Ideally a double-blind randomized, placebocontrolled trial (some exceptions apply for ethical reasons)

• In almost all cases, larger than phase 1 or phase 2 studies—approximately 50 -2000 subjects • Often conducted at multiple centers • In cases of severe illness, new routes to FDA approval may be instigated (i. e. combinations of phases 2 and 3 trials)

Phase 4 Studies • Post-marketing surveillance • Larger scale-longer term • May pick up ADRs (adverse reactions) not previously seen • Long-term safety. . broader experience Example of drugs taken off the market after in clinical use: DES, Thalidomide, Copper IUD

Choice of Control Groups • Historical Controls- A group “comparable” to current study patients treated or observed in a previous time period (must have very good justification) • No treatment current controls- a group randomized to receive no treatment but are “comparable” to study patients (difficult to justify unless observation only is a standard)

Control Groups • Placebo concurrent control- a group randomized to receive placebo for comparison with study treatment group; in some situations, ethical problems arise • Active-treatment control- a group randomized to receive an active treatment (best current treatment) for comparison with study treatment group

Control Groups – Crossover design- patient acts as his or her own control- patients are randomized to a specific sequence of treatments eliminates many sources of bias, but some difficulties may arise in carryover effects