Tuesday, April 3, 2007

Clinical Trials

What is a Phase I Clinical Trial?

An experimental study that is designed to evaluate the safety of a treatment.


For pharmacotherapeutics, this often is a trial that escalates a drug to determine dose-limiting toxicities.


What is a Phase II Clinical Trial?

An experimental study that evaluates the biologic response of a treatment.


The results of this kind of trial are often compared to historical controls.


How are responses often categorized for cancer therapeutics?

CR - Complete Response
PR - Partial Response
SD - Stable Disease
PD - Progressive Disease


What is a Phase III Clinical Trial?

An experimental study that evaluates the efficacy of a treatment.


This is typically a randomized, controlled trial.


What is blinding?

A method to minimize bias from the investigators and/or the participants where the individuals are not made aware of treatment assignments.


What are eligibility criteria?

These may be inclusion or exclusion criteria that specify the type of patients that can or cannot be enrolled in the study.


Studies typically exclude minors and pregnant women for ethical reasons.
Studies may also limit participation to people without certain medical problems (i.e., coagulopathy, renal insufficiency).
Studies may try to include only certain patients (i.e. Blunt Trauma patients with a GCS less than 8)


What are some clinical endpoints commonly studied in oncology Phase III trials?

Overall survival
Disease-free survival


Distant disease-free survival may describe the proportion of patients who were alive and did not develop metastatic disease, while disease-free survival describes patients who were alive and did not develop any recurrence (locoregional/distant).


What are methods typically used for survival analysis?

1. Kaplan-Meier Analysis
2. Cox Proportional Hazard Models


Kaplan-Meier curves provide estimates of survival based on the length of follow-up for each individual patient. Patients who are lost to follow-up are censored, while patients who do not survive will cause the curve to go down in proportion to the total number of patients followed at that time interval.


A Cox Proportional Hazard Model is a form of multivariate analysis that allows the investigator to control for known variables (i.e. age, gender,..) that might influence an outcome.


What is the Intention-to-Treat Principle?

The investigators analyze the data keeping patients in their original treatment assignment groups, regardless if patients crossed-over to the other treatment group.


If you designed a study that randomized patients to stenting versus surgical revascularization, and some patients in the stent group ended up having surgery, their outcomes would still be analyzed in the stent group. By not following the intention-to-treat principle, the outcomes of the stent group may be biased against stent failures.


What is clinical equipoise?

The balance between evidence of an experimental therapeutic's efficacy and the degree of uncertainty surrounding the evidence.


This allows for an ethical deliberation leading an investigator to decide to start a study, or to stop a study early.


What is an interim analysis?

A planned preliminary assessment of data to allow for monitoring of various outcome measures.


Stopping rules may terminate a trial early if the data show harm to patients.


What is the Hawthorne effect?

The concept that the knowledge of being observed will affect the behavior of the participants.


While this was initially described in reference to studies on worker productivity under different lighting conditions, patients may avoid unhealthy behaviors if they know they are participating in a clinical trial.


What is a Phase IV Clinical Trial?

After FDA approval of a therapeutic, this is a study that evaluates its effectiveness in a wider population, as well as follows for late adverse events.


So to recap the types of clinical trials:

I - Safety
II - Response
III - Efficacy (RCT)
IV - Effectiveness


Late... SG

Source:
Essentials of Surgical Oncology, Mosby (2007).

No comments: