In this article I am going to detail in broad strokes the necessary steps to getting marketing approval to sell a drug in the US. It's a complex process, but the various steps are easy to follow. It's also critical to understand this process if you wish to profit from investing in biotech and pharma.

To sell a drug in the United States, a company must get permission from the Food and Drug Administration, an entity which is part of the Department of Health and Human Services.  The FDA enjoys authority over a large part of the economy, but exercises it most clearly when it comes to the marketing and sale of medications and medical devices.

Contrary to what is written in many investing publications, the FDA is actually a very easy entity to navigate.  I am not saying that it’s an agency without problems.  It has a clear process for drug approval and, if these are met, a company should expect permission to sell its product in the United States.

In Europe there are similar entities, but on the national and EU level: Europe is moving towards a centralized system of drug approval.  In fact, the EMA, formerly called the EMEA, can grant approval for marketing in all EU member states under certain conditions; otherwise, the EMA seems to function more as a coordinating body.  It has similar characteristics to the FDA.  Japan has its own approval process and is, in many important respects, similar to the US.

The FDA focuses on two things during the approval process: is the drug safe and is it effective.  Companies spend hundreds of millions of dollars trying to answer these questions before they receive marketing approval.  In most cases, companies do not get permission to sell their drug; if they do, there’s only a 30% chance that they will recoup their development expenses.  For those that follow biotech companies, most companies fail because most development candidates fail.  The chance of success is dependent on a number of factors, but speaking very generally, for ten to fifteen drugs that enter a phase 1 clinical trial, only 1 makes it to the market.  Remember, that drug has only a 30% chance of recouping its development costs.

Later today, I will add a number of posts that detail the approval process in the US.  This process is mirrored pretty closely in other countries.

I am going to skip over the early discovery efforts as they’re not really germane to the conversation and focus instead on the necessary steps to entering the clinic.  Entering the clinic means administering an investigational drug to humans—you’re essentially using humans as test subjects.

As said previously, the FDA wants to see that a drug is both safe and effective.  The company tests the drug in many biological and model systems to demonstrate efficacy.  Often animals, mammals, but usually mice or rats, are specifically bred with a condition: a type of cancer or other disease.  The drug is tested against these animals and if the experiment works, the drug is pushed along towards the clinic.  These animals or other model systems are administered escalating doses of the drug and the response is monitored.  A therapeutic range is determined: how many milligrams of the drug is required per kilogram of body mass of the test subject.

In parallel companies carry out experiments on other animals to determine if the drugs are safe.  The animals are given escalating doses of the drug and their main organs and systems are examined for signs of toxicity.  The dose that most interests investigators is in the vicinity of the therapeutic range and higher.  How the drug is absorbed, distributes itself around the body, is metabolized (consumed by various biological processes) and is disposed of is also examine.  Animals are “sacrificed.”  It’s a part of science that makes many uncomfortable and scores of companies are investigation different ways of accomplishing he same results without the need for animal sacrifice; however, to date, no satisfactory alternative has been found.

The reason animals are used is that their DNA and, consequently, their biological systems resemble those of humans.  Yeast has 50% of our genes; higher primates are closer to 97 or more percent.

If the drug is found to be effective in some model systems and safe in lower life forms, the safety is tested in other animals: the experiment moves up the food chain.  Work carried out in rats and mice continues into dogs and, eventually, into primates.  There are chimpanzees who have “worked” in the pharmaceutical industry for thirty to forty years.  Primates aren’t sacrificed in the discovery/development process, but other animals are.  Typically, animals are administered

Once the company believes that the drug works effectively and that asks the FDA for permission to administer the drug in humans by submitting an IND or Investigational New Drug application.  The company shows that the results in animal and model systems and gives a proposal for how this drug will be administered in humans.  The goal is to demonstrated to the agency that the company knows what it is doing; it has developed meaningful scientific data in its preclinical studies and that it will administer this novel drug in humans in a responsible way under the guidance of professionals, ensuring that any problems with toxicity (side effects) are found quickly.

If the agency is happy with the IND and approval is granted, the company can begin a Phase 1 clinical trial.

The clinical trial process begins with successful submission of an NDA.  The company then conducts what are called clinical trials: experiments where the drug is administered to humans.  There are three main types of trials: a phase I where safety is evaluated; a larger phase II trial where the first determination of efficacy is made and a much, much larger phase III trial where efficacy is determined in a group that mirrors the target population.  Once the trials are completed, a New Drug Application, NDA, is submitted to the FDA.  If the FDA sees problems, it can send a “refusal to file” notice within two months; otherwise, it issues a PDUFA date, the date on which is will make its decision.  About 55 days or more before the PDUFA date, the FDA can announce an advisory panel.  This panel gives its recommendation which the FDA can decide to follow, though sometimes it doesn’t.  If the FDA is happy with the submission, it can grant marketing approval to the company, after which the company is free to sell the drug.

The next few posts will provide more details about this process.

Phase I clinical trials

Once the company receives approval from the FDA for its Investigational New Drug application, the company is free to administer the drug to humans using the protocols it proposed to the agency.  A pure phase I clinical trial is used to determine the safety of the drug—it doesn’t look at efficacy.  The goal is to ensure that the drug is safe to administer to humans in the proposed therapeutic window for the active ingredient: the molecule that does the work.  The therapeutic window is the concentration of the drug that’s anticipated to measure its effectiveness.

There are different types of Phase I trials.  However, the main goal is to determine safety in the therapeutic window.  The subjects selected for the trial are almost always healthy volunteers without the target indication.  Often escalating doses are administered to a small group of patients, around 20 to 50 people, and the response measured.  The trial usually occurs in a hospital setting and major systems and organs are monitored until the drug is safely expelled from the body.  How the body absorbs the drug, distributes it around the various organs, metabolizes it (destroys it through metabolic processes in the body) and gets rid of it are monitored carefully.  The dose may then be slowly escalated in other groups of subjects up to and above the effective level.  In no cases is the dose escalated to the level where it was found to be toxic in any animal studies.

For small organic molecules, about thirty percent or so of programs fail because of toxicity issues; the take home lesson here is that, while in many cases animal studies show the drug to be safe, it’s only in an actual human trial that this statement can be verified.  Suffice it to say that if investigators could find a better option than animal models, they would use it.  The success rate for large molecules, proteins (which includes monoclonal antibodies, etc.) is higher.  Over ninety percent of such drugs are successful in a phase 1 trial.

In rare instances a Phase I trial is carried out with subjects who have a serious condition.  This option is sometimes deemed appropriate for patients with, say, a fatal cancer were few if any other therapeutic options exist.

While healthy volunteers are used in the study, the company could conduct two or three clinical trials to ensure the safety of the drug at different doses and when administered under different conditions.  A complete phase I trial set can last from six months to, in some cases, two years.

The company presents its results to the FDA and if the drug is deemed safe, the company can then proceed to a phase II trial.

Phase II Clinical Trials

The combined efforts of preclinical and Phase I clinical activities convince the researchers that the drug candidate is safe: a determination is now made if the drug is effective.  A phase II clinical trial begins.

A Phase II trial typically consists of about 50 to 200 patients with the target indication, the disease or ailment, and few if any other problems.  The goal is to answer questions about the indication in question.  Some clinical trials can be larger; in a few instances they can be smaller.  A Phase II trial is larger than a Phase I with an emphasis on efficacy, but there is also an overriding emphasis on safety.

Nowadays, you’ll almost always hear of a company having “a Phase II trial” and a “Phase IIb trial.”  There are differences between the two.  A IIb is sometimes called a pivotal study.  Since the subjects in a Phase II trial are stricken with the indication of interest, dosing experiments are carried out to confirm or determine the therapeutic window: basically, now much drug to give a patient.  Typically a Phase II trial is the first time that a patient with the indication is administered the drug.

A Phase IIb trial is the one to watch, because here the efficacy of the drug is assessed.  The gold standard for such a trial, and one that the FDA requires for approval, is a double-blind, randomized study.  In this study, the patient group is divided in two, though not necessarily half; one group is administered the investigational drug, the other group is administered a placebo.  A placebo is a tablet or other drug form that has no therapeutic value but that, in its make-up, is indistinguishable from the real investigational drug.  Neither the medical practitioners administering the drug nor the patients receiving it know which group they are in.  The drugs, patients and practitioners are coded by the investigators who, at the end of the study, “unblind” the results and analyze the data.  Once analyzed, the company publishes the results at conventions or in referred journals and communicates the news to the FDA.  Successful completion of a Phase IIb clinical trial is necessary if the drug is to progress further.

It should be noted that in many instances a traditional placebo is not used.  When there are ethical implications—for example, if the indication is life threatening—the new therapy is compared against the current standard of care in a controlled environment.  It makes the statistical analysis of the results potentially more challenging, as the company must show that the new therapy is at the very least as good as what is currently offered in the market place.

The stock price of a public company general moves up after successful completion of a Phase IIb trial.  It is a major milestone for the company and greatly reduces, but doesn’t eliminate, the risk of failure of the company.

Small molecules have a higher rate of failure in Phase II than do large molecules such as proteins and related compounds.  Approximately of third of small molecule drugs are successful in a Phase II study; around two-thirds of large molecule drugs move from Phase II to III.  Phase II clinical trials can last for 6 months, in exceptional circumstances, to three years.

Once the trial is over, the company usually meets with the FDA to decide on next steps.

In some circumstances, the FDA can order the trial to halt if there are safety concerns.  If a death occurs, the company itself acts quickly and stops the trial, but there have been examples where the drug appears to be so effective that the FDA orders the study to be unblinded.  If it’s clear that the drug works as well as or better than expected, the agency usually tells the company to proceed immediately into Phase III.

A discussion on Phase III follows next.

Phase III Clinical Trials

A phase III clinical trial is where the proverbial rubber meets the road.  Without at least two successful phase III trials, a drug is unlikely to be approved by the FDA in the United States or the EMA (or national agencies) in Europe.  A phase III trial can, in rare instances, consist of as few as 300 enrollees to as many as 3,000.  Typical trials contain about 1,000 people.  The goal of the trial is to show that the treatment works in a typical cross section of the population that would take the medicine.  A phase II trial is much more controlled in that the patient population usually has the condition of interest and nothing else.  In a phase III trial, the goal is to mimic a typical user, so patients may be taking other mediations; they may have strange diet habits; they could have other ailments, etc.  One of the key challenges in constructing any clinical trial is compliance: that patients will continue to take the medication as prescribed.  If many patients fail to complete the course of therapy, it can affect the outcome of the results.  Therefore, companies usually enroll many more patients in the trial than are strictly required, since they’re aware that many will simply not complete the treatment.

A phase III clinical trial consists of a group of patients with the target ailment; the study is randomized: patients are divided into different groups for reasons discussed later; these selections are carried out randomly.  The study is double-blind: that is neither the clinician involved in the study nor the patients know which medicine, if any, they are receiving.  Finally, the investigational new drug is compared against a current therapy in a control group—often the “gold standard” or best therapy for the indication.  If the drug is demonstrated to be safe and more effective than the current, best therapy then it is almost certainly assured of receiving marketing approval from the FDA.

Trials are also usually conducted at multiple sites—five or more is not uncommon—so you’ll often hear of a “multicenter trial.”  The site is normally a hospital where a physician who is a leading practitioner in area heads the study.  To get approval, a company generally needs at least two successful, phase III, multicenter, double blind trials.  The trials must meet their endpoints.  The endpoint is a result that either (a) the FDA has as a standard or (b) that the company and the FDA have agreed together prior to the start of the trial.

The study is conducted in the hospital center.  Patients are carefully monitored for what are termed “adverse events” (AEs): side effects or signs of toxicity.  The worst of all AEs is the death of a patient: unless the trial is being carried out for an indication where the mortality rate is high, such an AE can have a serious impact on the ability of the company to continue the trial.  If an autopsy suggests that the medication in the trial was in any way responsible for the death of the subject, it is likely that the trial would be suspended or even ended.

In some cases where an AE occurs, if the reason isn’t clear, the result can be sent to an independent third-party panel for adjudication.  If the panel determines that there are no concerns, the trial could continue.

Pharmaceutical companies are all about label claims.  A good rule of thumb is that a company needs at least one phase III clinical trial for each label claim that is made.  When reading a medication, you’ll see often many claims made by the company.  For example, Lipitor says things like “LIPITOR, along with diet, is clinically proven to reduce the risk of heart attack, stroke, certain kinds of heart surgeries, and chest pain in patients with heart disease or several common risk factors for heart disease. Common risk factors include family history of early heart disease, high blood pressure, age, low HDL ("good") cholesterol, and smoking.”  Each of these claims must be demonstrated by a clinical trial for the FDA to grant permission to use the assertion.

Once the trial is completed, the results are unblinded and statisticians examine and analyze the results.  A comparison is made between the group or groups receiving the medication and those in the control group.  The goal is to demonstrate that the difference, if one exists, couldn’t occur randomly.  The assumption then becomes that the drug has beneficial effects.  If both clinical trials demonstrate efficacy, then it is in a good position to be approved by the FDA.

The success rate for small molecules in a phase III clinical trial is 60%, higher than in either phase I or II; for a large molecule (a macromolecule such as a protein), the success rate is closer to 45%.

The results are compiled into a submission called a “New Drug Application” (an NDA) and submitted to the FDA for review.  That process is described next.

Requesting Marketing Approval from the FDA

Once all the clinical trials are completed, the drug company combines all its preclinical and clinical results and combines them into a document known as a New Drug Application or NDA.  The form and format of the NDA is determined by the Food and Drug Administration.  While in the past, a paper submission could consists of tens of thousands of documents occupying millions of pages—enough to fill a semi—today, the FDA favors electronic submission and almost three-quarters of submissions are in electronic form.

The FDA reviews the submission to determine if the drug is safe and effective; if the manufacturing process for the drug is in control and if the labeling proposed by the company is accurate and appropriate.

It’s important to realize that the FDA does not conduct any clinical trials on the drug itself; they analyze those presented by the company.  The FDA may also take the raw data submitted by the company and subject it to an independent or different statistical analysis to determine efficacy.  The FDA may also visit both the manufacturer of the drug itself and the producer of the final medication (the tablet, patch, injectable form, etc.) to ensure that production is in control.

The submitting company is required to present a detail, unvarnished history of the development of the drug from its earliest stages in discover through its most recent clinical trial.  The company is not allowed to omit information that it believes is detrimental to the successful approval of the drug.  The review process has become pretty predictable following the enactment of the Prescription Drug User Fee Act of 1992, (PDUFA).  With this law, companies can pay fees which cover the costs that the FDA incurs in hiring, training and using its employees to review submissions.  Some believe that these PDUFA fees, which can total $1 million to $2 million per drug make the agency too close to the drug companies; other argue that a better analogy might be a toll collector and a driver.  The PDFUA act and changes made a few years later have really increased the speed at which drugs are reviewed.

Once the submission is received, the FDA makes an evaluation of the package within 60 days and issues a response within 74 days.  A reviewer will examine the package and the agency will issue either an acceptance letter or a “Refusal to File” letter.  This refusal letter is public and the agency will state why: the submission may not be in the correct form or the agency may have found a problem with some of the results.  If the application is accepted the agency will inform the company if it will receive standard (10 month) or accelerated (6 month) review.  A PDUFA date—the date on which a decision will be made—is most likely given.

A refusal to file letter is not a death sentence.  If the reason is trivial, the company can quickly correct the problem and submit amendments to the application.  Overreaction to such a response in the market could present a buying opportunity for an astute investor.

The FDA can continue to make more requests for information or clarification during the review process.  Unless these requests are deemed substantive by the company, it’s unlikely that any public disclosure will be made.

At least fifty-five days before the PDFUA date, the FDA can inform the company that the application will be referred to an advisory committee, often called a panel.  The FDA has forty to fifty standard committees made up of experts in different fields of medicine, biology, etc.  The committee reviews an abbreviated document of around 200 pages that describes the drug, its development, clinical trials, other therapies, etc.  The committee, which can consist of a few dozen experts, will vote on efficacy and safety.  The FDA has no obligation to follow the advice of the committee.  About 20 to 30% of applications are referred to advisory committees.  FDA guidance documents are published two days before the committee meets and provide good insight into the process.  Involvement of an advisory committee or panel doesn’t appear to have any impact on the ultimate success of the drug in the approval process.

The final step is a decision on marketing approval by the FDA.  In about 1 in 5 cases, the FDA has failed to meet its PDUFA date obligation and has delayed a decision by up to six months.  The FDA will publish its decision called an approval letter if the result is positive.  If the FDA declines marketing approval, it will state the reason or reasons in a non-approval letter.  The agency will communicate to the comapny any actions it needs to take to correct defects in the original submission.

Other Things that Happen during the Approval Process

The approval process starts when a company files its NDA with the FDA.  At that point a clock starts ticking leading to drug approval ideally twelve months after the NDA is received by the FDA.  During that time a number of things might happen.

Inspection of Manufacturing Sites

Prior to receiving marketing approval, the FDA has the right to inspect the manufacturing site where the active ingredient is made (the molecule that does the work) and where the drug is formulated and packaged (where the active ingredient is formulated into a tablet or other end product).  These inspections are called Pre-Approval Inspections or PAIs.  In some cases, where the FDA has a good relationship with the manufacturing or formulation site, the agency might waive this inspection.

Companies manufacture and formulate according to a quality philosophy called current Good Manufacturing Practices or cGMP.  It is a philosophy as the FDA insists that each manufacturing site keep abreast of standards of excellence within the industry and to incorporate them into the manufacturing processes for a particular site.  Each manufacturing site develops its own cGMP system run by standard operating procedures (SOPs).  An FDA inspector will examine these guiding documents to ensure they meet current standards.  SOPs drive things like Master Batch Records, MBRs, the “recipe” by which a chemical is produced.  The FDA will look at executed MBRs, which are documents which chemical operators have followed in carrying out a chemical step in making the active ingredient.  The executed MBR will contain appropriate details of what the operator did and when he or she did it.  If problems occurred, these will be noted and addressed following the appropriate SOP in the GMP system.  Chemicals will be tested as required following steps outlined in the appropriate document.  It’s a complex document-driven system that details, step-by-step, what must be done.  It is designed to ensure the highest quality of the end product and complete repeatability in the manufacturing process.  The process for formulating the API into a tablet or other deliverable uses a similar, cGMP system; an FDA inspector reviews a final dosage form manufacturing site in a similar fashion.

If an FDA inspector notes problems in the process, he or she can issue an observation on form 483, detailing what was observed and what corrective actions are necessary.  If the problems are serious, the matter can be escalated to the point of shutting the plant down.  If the manufacturing site does not pass the PAI, the NDA will be rejected or, more likely, delayed.  Problems in the manufacturing process, problems in carrying out a long-term stability study, or problems in the formulation or packaging are sufficient to delay marketing approval for a drug.

Large pharmaceutical companies often carry out chemical manufacturing, formulation and packaging in their own facilities.  However, almost all emerging pharmaceutical companies outsource these activities to one of scores of Contract Manufacturing Organizations (CMOs) located in the US, Europe and Asia.