Types of Clinical Trials

Clinical trials are experiments. Medical researchers and drug developers conduct clinical trials on real patients to find new and better treatments for cancer. Before the Food and Drug Administration (FDA) approves an oncology drug for market release and general use, the new medicine must go through a series of tests or clinical trials.  The clinical trials process is well developed and follows established guidelines for patient safety and notification.

Cancer clinical trials investigate new surgical techniques and radiation therapy methods, as well as chemotherapy drugs. Combinations of treatment methods are also tested, sometimes for FDA approval and sometimes just to see if they work. Many chemotherapy trials are in support of regulatory approval of extension of the drug’s use for a different form of cancer from the ones it has previously been approved for. For instance, a chemotherapy agent that is FDA-approved for kidney cancer might be tried in lung cancer patients to see if it has enough efficacy to merit FDA approval for widespread use in lung cancer.

It is expensive to carry out clinical trials.  The bill is paid by a sponsor, who is usually the pharmaceutical company that developed the new drug.  Sometimes more than one pharmaceutical company will team up, or other investors will become part of sponsor syndicates if they reach an agreement about money splits following an FDA approval.  Even so, the number of clinical trial starts for oncology drugs increased 60 percent between 2015 and 2022, as pharmaceutical companies eye the huge revenues a new drug can bring.

Do new drugs ever not pass the clinical trials process?  Yes, all the time.  Clinical trials are experiments, so there is no success or failure for the trial per se, but the medicine may prove ineffective or unsafe. Most cancer drugs that enter the clinical trials process are never approved by the FDA.  Indeed, it is common for sponsors to cancel drug development following negative results in a trial.  An article in Nature stated that “clinical trials in oncology have the highest failure rate compared with other therapeutic areas.” The authors speculated that the reason was the high number of new therapies submitted for trials.  A separate article in Nature estimated that over 19,000 clinical trials are going on for cancer drugs at any one time.

New drugs not previously used for cancer treatment are referred to as investigational medicines in the clinical trial process.  New oncology drugs are first tested in a laboratory (in vitro) and in animals (in vivo). This preclinical testing determines toxicity of the drug at least in animals. Knowing the toxicity level of the oncology drug provides the sponsor with an upper limit to the amount of the drug that can safely be used in human subject clinical trials.

Once the oncology drug has passed the preclinical phase with an acceptable safety profile, the sponsor moves to human subject clinical trials. There are three phases (I-III) of human subject clinical trials prior to FDA approval for market release and one phase (IV) aftermarket release.

The US Federal Government’s website clinicaltrials.gov says over 400,000 clinical trials are ongoing as of March 2022.  This includes other fields in addition to oncology, other treatments in addition to drugs, and includes many outside the US.  Some trials accept patients from both the US and other countries.  32 percent take patients from only the United States.  The trials include both observational and interventional trials, and the interventional trials encompass investigations of surgical procedures, medical devices, and drugs.  There were a reported 169,000 drug trials happening in March 2022.

Phase I Clinical Trials

Phase I clinical trials are known as safety trials. The sponsor submits documentation for an investigational new drug (IND) to the FDA and an investigational plan to the institutional review board (IRB) to receive approval prior to starting a Phase I clinical trial. This is the first trial of the new drug on humans, so the FDA does not give approval to just any drug.  Regulators look at results of tests in animals and at the process the sponsor uses to make the drug.  Phase I clinical trials usually have twenty to eighty healthy volunteers who sign an informed consent allowing a clinical investigator (typically a physician/scientist) to administer the new oncology drug. These human subjects do not have cancer.  The clinical investigator records data in a case report. The data collected includes side effects and anything that can be ascertained about how the investigational drug is eliminated from the body. The goal of the Phase I clinical trial is to determine if the medicine is safe to use in subjects with cancer by assuring that the data from the animal studies (level of toxicity) has not changed (i.e., harmed the subjects) during its use in the healthy volunteers.

Phase I trials typically have fewer than 50 participants.  The small size lowers the financial risk and the risk of widespread serious detrimental effects.  The main information desired at this point relates to safety/toxicity. If the drug appears to work, great, but the sample size is too small at this point to prove efficacy.  The idea is that is the drug is strongly unsafe or toxic, it will show in even a small number of users and not too many people are put at risk to find out this information. Researchers may also learn about the medicine’s efficacy and side effects in Phase I, and that may help them to design protocols for Phase II.  

Phase II Clinical Trials

Phase II clinical trials can begin if the Phase I clinical trial shows the drug is safe. A Phase II clinical trial typically has 30-300 subjects who have the type of cancer specifically to be treated by the new medicine.  These trials are explicitly intended to determine efficacy and the best dosage that strikes a balance between side effects and curative action.  Phase II trials more often include people who do not have the disease the researchers are targeting. Phase I trials may also include non-patients, but the inclusion of these participants tends to increase in later stages of trials.  Why might non-patients be recruited? For one thing, the researchers may want to find out if the drug has negative effects on healthy people. Also, the non-patients may have a related disease or one with a common etiology and the scientists might want to see what the medicine does in those people.

Phase II clinical trial data is intended to demonstrate the investigational oncology drug’s safety as well as the effectiveness (change in illness/clinical benefit). Most often, Phase II clinical trials are controlled trials where the drug is compared to an existing, FDA-approved oncology drug or to a placebo. The subjects are randomized and given one of the drugs (or placebo) during the treatment period. Investigators keep an eye out for side effects that appear during the trial.  If they are too bad and too prevalent, investigators may pause or cancel the trial.

The goals of Phase II clinical trials are usually to collect data on endpoints, such as time to tumor progression (TTP), progression free survival (PFS), and time to treatment failure (TTF). Analysis of data from these endpoints helps the sponsor and FDA determine if the investigational oncology drug is safe and effective for a larger population. At the completion of the Phase II clinical trial, FDA and the sponsor meet to determine if the investigational oncology drug has met the safety and effectiveness hypothesis in the investigational plan and how many subjects will be needed in a Phase III clinical trial.

Phase III Clinical Trials

Phase III clinical trials are also known as approval trials. These clinical trials typically have 300-3000 subjects. These clinical trials continue to review the investigational oncology drug for safety and effectiveness but in a larger population and with different dosages. Additionally, a Phase III clinical trial may compare one or more approved drugs (cocktail) to the investigational drug.  Regulators will take this data into consideration: if the new medicine is better than accepted standard-of-care therapy, it is more likely to be approved.

Phase III trials typically have thousands of patients and protocols are established only after the sponsor has a firm idea of what they are going to claim in their application.  So things like dosages (or range of dosages) and method and timing of administration are close to what is eventually approved. The large number of people taking the medicine in Phase III helps generate high quality data.  A higher number of people and more data allow researchers to get statistically significant results. Phase III trials are trying to prove efficacy, of course, but also to establish the best dose, the nature, prevalence, and severity of side effects, and also to find interactions with other common medicines.   It is pretty common to see news stories about Phase III trials being cancelled for one of two different outcomes: (1) safety if the medicine seems to produce beyond-acceptable adverse side effects or (2) happy news if the patients taking the medicine seem to be benefiting so much that it is considered cruel to deny if to the patients taking the placebo.

You might sometimes see the term “pivotal study” or “pivotal trial” to refer to large Phase III trials with participants of the target population for the drug under consideration.  The researchers go out of their way to keep tight protocols and do detailed statistical analysis. The regulators likewise scrutinize these studies closely. The pivotal studies ideally provide strong evidence of efficacy and safety and are important in gaining approval for NMEs.

What if the drug is shown to be effective (reduce mortality or extends survival period) and reasonably safe?  Then the sponsor is free to file a new drug application (NDA) with FDA. This application includes data collected on preclinical and human subject clinical trials, manufacturing processes, and drug labeling. The application is reviewed by a team of FDA experts including chemists, statisticians, doctors, and pharmacologists. If the reviewers agree with the sponsor’s results and conclusion, the FDA approves the oncology drug for market release. If there is any question regarding the results or the conclusion, the FDA sends a letter to the sponsor who responds as applicable and may change their application or continue trials.

Phase IV Clinical Trials

The FDA may require an oncology drug sponsor to have a post-approval or Phase IV clinical trial after the drug has been approved for market release. Phase IV clinical trials look at long-term safety (risks) and effectiveness (benefits) in the general population. Phase IV clinical trials typically have several thousand subjects.

Protocols and Experiment Design

Double-arm or multiple-arm studies organize patients into cohorts in which everyone in the cohort receives that same treatment.  The differences between the treatments on the arms are what drives the experiments. One arm is the experimental one, where patients gets the drug under investigation.  The regimens used in other arms are called comparators. At least one arm has a placebo comparator (an inactive therapy). (Some researchers designate the placebo as a “sham competitor”, meaning the inactive therapy is made to look like an active therapy.)  Some clinical trials include one or more arms with an active comparator – another treatment, perhaps one widely used for the illness the new drug targets.

A non-randomized trial is one in which all participants receive the therapy under investigation.  There is no control group. These also called single-arm trials. These happen frequently in Phase I and Phase II trials, in which less precise and convincing data on safety and efficacy is sought. The endpoint in these trials is often the response rate – what percentage of patients show an improvement in disease status.

Phase III trials, which are done to prove to the regulatory agency the drug should be approved, include a control arm.  The patients are assigned randomly to arms so as to maintain clean unbiased data.

Almost always, the patients do not know which arm they are in, in order to control for placebo or psychological effects.  This is called a blind study. Phase III trials are always blind. Often the researchers who personally interact with the patients also are kept in the dark about which patients are getting which treatment.  This is an even tighter control because it eliminates the possibility that human researchers will treat patients differently based on what kind of treatment (or placebo) they are receiving. This is called a double-blind study.  A randomized double-blind study is the gold standard of drug testing.

The number of patients in the study (in all arms) is called the sample size in statistical parlance.  Generally speaking, a bigger sample size means the scientists (and their regulatory reviewers) will have more confidence in the results of the experiment.  But it is expensive to run bigger clinical trials with more patients. And the protocols that the researchers set (e.g. age, sex, general health, presence of disease of interest) often limits the percentage of the population that is eligible to participate.  So getting a high enough number of participants to get reliable data within a budget it often tough.  It was reported in 2020 that there were approximately 100 Phase III trials on therapies for non small-cell lung cancer happening in the US.  A good percentage of patients with that type of cancer are in a trial.

Data Privacy and Publication Bias

What if the clinical trial doesn’t produce the results its sponsors hope it will?  This happens a lot and everyone in the research community knows the desirability of results affect whether those results are published.  This is called publication bias.  Are bad results intentionally covered up?  Maybe.  Some suspect they are.  Of course, corporations spend millions of dollars on clinical trials and own the data produced.

Some voices call for all research data to be public.  This practice is said to enhance confidence in the integrity of the trial process, and encourage general sharing of scientific information.  Further, it allows others who were not involved with the trial attempt to squeeze out insight from statistical analysis of the data.

Power of a Trial

Often you will see a clinical trial’s statistical power noted.  This is an indication of the probability that the drug will be shown to be effective if it is truly effective.  Strictly speaking the power is the probability that the test of the hypothesis rejects a false null hypothesis.  The false error rate is lower in trials with higher statistical power.

Basket Trials

A basket trial is a clinical trial that tests how well a specific drug works in many different cancer types at the same time, all of which have the same genomic alteration.

Also called a “bucket trial,” this type of trial allows a researcher to find out the efficacy of a drug in multiple cancers.   Cancers arising in different organs, which are technically different diseases, can be tested in this type of trial.  Traditional trials test patients with narrow disease criteria – one type of cancer

The way most traditional clinical trials are designed, a large number of patients must be recruited.  That’s feasible and possible if the trial focuses on one disease that a lot of people have – so that there are enough enrollees in the trial to produce a meaningful result.  But when drug developers are looking at finding treatments for cancers with specific genetic mutations the pool of potential trial participants is small if the trial is limited to patients with one type of cancer.  Basket trials are a way to find out if a drug works in this situation.

Unlike conventional trials, the scientists must know in advance that the target they are aiming at with the drug is one that tumor growth depends on.  Basket trials are particularly useful for testing tumor-agnostic chemotherapy regimens and medicines. They often rest on the idea that a molecular marker in the bloodstream is a reliable indicator of anti-tumor efficiency.

This all gets complicated and even more than with conventional trials, basket trials require close planning and adherence to the plan.  Writing in the Journal of Clinical Oncology in 2017, oncologists noted a “trade-off between complexity and efficiency.”

Smaller and smarter trials are better because they allow researchers to figure out what works faster.

“Big Data” and faster trials

Another method of speeding up trials involves reuse of data from other trials.  While traditional trial design divided participants into two “arms” – one group that got the new treatment and one that did not – a new idea is to employ information gathered from previous studies to be the de facto control group.  This use of existing data and sophisticated statistical methods may allow all participants in a trial to receive the new treatment and produce faster results from the study.

Data and Proof of Efficacy

Data for clinical trials comes from all over the world.  The patients do not live in the United States, for the most part.  A report issued by the FDA stated 105,808 people were involved in pivotal clinical trials worldwide in 2015, with 29,775 of those people in the United States.  A 2022 article in Nature stated there are 19,500 cancer clinical trials underway worldwide.

The globalization of clinical trials makes sense.  Data gathered from trials may be used in support of drug approval applications to agencies in different countries and regions.  And the large pharmaceutical companies are global in scope – the intended to sell or license medicines in many countries.

When clinical trials are planned, their designers designate “endpoints” they will look for.  This is how good experiments are made: decide in advance what hypothesis you are testing and what constitutes confirmation of your hypothesis.  Along the way you might find interesting things (good, bad, or otherwise) by serendipity. But you still have planned endpoints.

More on endpoints.

Participating in Trials

Clinical trials are set up to answer specific questions and they therefore have to function along narrow criteria. The investigators – the people who design and run the trials – establish the rules before they ever recruit any patients. They define the treatment regimen they will test, the type and dosage and duration of treatment, and the characteristics of the patients in whom they want to test it. Once these criteria are in place and trial recruitment is underway, there are no changes. Don’t take it personally if you are turned down for a clinical trial because you don’t meet the criteria.

The National Cancer Institute has a search engine to help patients and doctors find trials at https://www.cancer.gov/about-cancer/treatment/clinical-trials/search

The American Society of Clinical Oncology has a webpage with resources for finding trials here.

Do I have to pay to be in a trial?

No.  There is no law that says the trial cannot charge you or your insurance company, but patients are rarely, if ever, asked to pay.  As a patient, you are valuable to the experiment, so they should be glad to have you.  In fact there is currently in the US a shortage of cancer patients willing to be in trials.

Some trials require regular check-ins to the local treatment center, and participants should live near the hospital or clinic. Others are not so restrictive and patients can be recruited from across the country.

The teams that run clinical trials include doctors, nurses, administrators, statisticians, and scientists. As a participant, you will probably get some diagnostic tests and free treatment. In return you agree to allow the trial investigators to collect information about you. Sometimes participants in completed trials are asked to take place in new trials.

You will probably never meet or learn the names of the other patients in the trial. Some trials go on for years, with early patients finishing the trial before later patients even start.  The National Institutes of Health has a good webpage answering common questions.

Helping out medical science

If you are interested and qualify and it is feasible for you to enroll in a study, your participation can help the medical community move forward with new treatments. Even if it turns out the treatment is ineffective, a well constructed clinical trial can be invaluable to science by pointing out that a given treatment doesn’t work, or it can give clues as to new research paths to pursue.  There is so much interest from pharmaceutical companies in developing new immunotherapy treatments and personalized therapy that experts worry there aren’t enough cancer patients to fill up the openings in clinical trials needed to test the new therapies.

Access to Clinical Trial Results

The US federal government’s clinical trials website can often point you to reports of results: https://clinicaltrials.gov/ct2/help/how-find/find-study-results  If the results have been published in a journal, you can usually find the abstract, and sometimes more, on the internet.  Google’s academic search engine scholar.google.com is useful.  You can ask for results from Pfizer-sponsored tests directly from that company, although they do not give the results out to just anyone.  Novartis also has a webpage on clinical trial results.

Trialtrove is an online service that sells clinical trial data.  ClinicalTrialsArena sells reports.  We have no experience with them so cannot say how useful or cost-effective their products are.

More information can be found at the US government website ClinicalTrials.gov and the European Medicine Agency’s Clinicaltrialsregister.eu