Kerala Heart Journal -Tiny Nair

Back to Basics

Kerala Heart J  2016; 6(2):xx-xx.  


Clinical trials

 

Corresponding author


Tiny Nair, MD, DM, FACC, FRCP(E)

Head, Dept Of cardiology

PRS Hospital

Trivandrum 695002

Kerala, India

Email:tinynair@gmail.com

 

Clinical Trials

 

Clinical Trials can be either observational (we observe) or assign an intervention. In the ‘observational’ trials there are several types like cohort study, case controlled study, or cross sectional study. Interventional trials allocate an intervention (usually a new drug) to one arm while the other arm acts as placebo for comparison of outcome. Randomized controlled trials(RCT) are example of this type of trials.

 

Cohort Study

(From Exposure to Outcome)

 

In these trails, a group of people are studied after exposure to an agent. For example, a group of smokers and non smokers followed up for 10 years to see effects of smoking (bronchitis and lung cancer) in each group.

 

Case controlled study

(From Outcome to Exposure)

 

Here specific outcomes are taken and the history traced back to see for exposure in the past. Cases of lung cancer may be tracked back to see how many people smoked.

 

Cross Sectional Study

(Exposure and Outcome)

 

At one given time data is collected from a population on exposure and outcome. Long term effects may not be obvious in such studies.

 

Human Clinical Trials have multiple phases in their process of development

 

Phase I

(Is It Safe?)

 

The drug is tested in various dosage in small number of healthy volunteers, roughly in about 10 to 20 in number. The idea is to check for safety. There may be chance of unexpected side effects in this phase.

 

Phase II

(Is it effective?)

 

The drug is tested in around 200 people for its efficacy (and safety). A anti-hypertensive drug is tested for its BP lowering effect while an antidiabetic drug tested for its effect on lowering blood sugar and A1c. Most new drugs fail at this stage because of not being able to deliver its anticipated effect.

 

Phase III

(Is it better and safer than the available drugs?)

 

Confirmation of safety and efficacy in large population of around 1000 people. Rare side effects tend to come out during this phase. Most large scale clinical trials belong to this phase.

 

Phase IV

(Post marketing, Real Life.)

 

These trials keep a track on effects and side effects of drugs in much larger perspectives in extended indications and use. They give a ground reality of the drug usage. Despite the notion that these trials are ‘pharma’ sponsored commercially oriented projects, many new side effects tend to show up in these trials.

 

 

 

End point

 

Trials have to have a goal and that is called end point. The most reliable primary end point (PEP) of trials are hard end points like Death, MI, Stroke, Heart failure. The secondary end points (SEP) include ‘softer’ criteria like symptoms, hospitalization surrogate end points (like BNP in heart failure) and combination of multiple endpoints clubbed together. A trial that shows positive PEP are the most relevant ones. Even if a drug shows improvement in SEP

While Absolute number of Hard CV endpoints in many clinical situations like advanced CAD and HF are high, they may be quite low in situations of hypertension or dyslipidemia. In such situations a surrogate marker like lowering of blood pressure or attaining lipid goal may suffice. More challenging are the new surrogates like CRP, lowering of which might or might not indicate reduction of hard end points.

End points that are decided after the trial is over post hoc are considered inferior markers and much less robust statistically.

 

 

Randomization

 

A test drug is tried in a random fashion to avoid bias to prevent milder ‘better’ patients getting active drugs. Such randomizations may be done by using a random allocation by a computer algorithm, or a mathematical pattern.

 

Placebo controlled

 

New drugs are needed to be compared to comparator (often placebos). But today ethical practice dictates that placebo groups also receive ‘standard therapy’ and the active drug or the placebo may be added on top of that therapy.

 

Multicentric

 

Data collected for multiple centers spanning all over a country or multiple countries allow a fair mix of different ethnic population on one side as well as different hospital protocols and treatment algorithm making the final data more robust and bias proof.

 

Blinding

 

If the physician knows as to who is getting the study drug, he may be biased in interpreting the result. So most large scale trials are double blind, meaning neither the physician nor the patient is aware whether he is on a placebo or active drug.

 

Randomized Controlled Trials (RCT)

 

A properly randomized double blind multicentric trial is considered a gold standard today. Thses trials help us in identifying a better drug with more efficacy and less side effects. The advantage is that even a slightly better drug may be picked up by such trials.

The downside is that, theses trials are strictly monitored on count of compliance and follow up. An antiplatelet drug showing improvement in RCT (fully compliant patients) may show higher incidence of stent thrombosis on inadvertently stopping the drug for a few days (which might happen commonly in real life scenario). Such findings may often be picked up in real world registries and not show up in RCTs

 

PROBE Design

 

RCTs involve huge cost and manpower. An open trial with endpoints blinded are becoming popular, though considered statistically less robust by many. Such Prospective Open-labelled Blinded End-point trials go by the name PROBE.

 

 

 

 

 

Refbacks

  • There are currently no refbacks.