The Foundation: Reference Drugs and Test Batches
Before any humans are involved, researchers must establish a baseline. This starts with the Reference Listed Drug (RLD), which is the brand-name product approved by regulators. Agencies like the FDA require that a single batch of the RLD be used for the study to avoid variability between different manufacturing lots. To keep things fair, the test product-the generic version-must be made at a scale that represents actual commercial production. Usually, this means the batch must be at least 1/10th of the full production scale or 100,000 units. If the generic batch is just a tiny lab sample, it might not behave the same way as the medicine that eventually hits the shelves.
Researchers also perform a "pre-test" called comparative dissolution testing. They drop both the brand and generic pills into liquids that mimic the acidity of the stomach (pH 1.2) and the intestines (pH 6.8). If the generic drug doesn't dissolve at a similar rate-measured by a mathematical value called the f2 similarity factor-it likely won't pass the human trials. This step filters out failures early, saving millions of dollars in clinical trial costs.
Designing the Human Trial: The Crossover Method
The gold standard for these studies is the two-period, two-sequence crossover design. Instead of having two separate groups of people, every single volunteer receives both the brand-name drug and the generic drug. This is a brilliant way to reduce "noise" in the data because each person acts as their own control. If Person A processes drugs slowly, they will do so for both the brand and the generic, meaning the only variable being measured is the drug formulation itself.
Here is how a typical crossover study flows:
- Period 1: Group A gets the generic drug; Group B gets the brand drug.
- The Washout: Everyone waits for a set period. This is critical. The washout period must be at least five elimination half-lives of the drug to ensure the first dose is completely gone from the system before the next one starts.
- Period 2: The groups swap. Group A now gets the brand drug; Group B gets the generic.
Usually, these studies involve 24 to 32 healthy volunteers. However, if a drug is "highly variable"-meaning it behaves differently in different people-regulators might require a replicate design with up to 100 subjects to ensure the results aren't just a fluke.
Collecting the Data: Blood Sampling and Analysis
Once the drug is administered, the real work begins. Nurses collect blood samples at precise intervals to map out the drug's journey through the body. They don't just take one or two samples; they need at least seven time points. This includes a pre-dose sample (the "zero" point), several samples around the time the drug hits its peak concentration, and a few more as the drug leaves the system.
The blood samples are then sent to a lab where a process called LC-MS/MS (Liquid Chromatography with tandem Mass Spectrometry) is used. This technology is incredibly sensitive, allowing scientists to detect tiny amounts of the drug in the plasma. For a study to be valid, the analytical method must be precise within ±15%. If the lab's measurements are sloppy, the entire study is thrown out, which is why analytical validation is often the biggest bottleneck in the approval process.
Measuring Success: The PK Parameters
Scientists focus on two primary pharmacokinetic parameters to determine if the drugs are the same. First is Cmax, the maximum concentration the drug reaches in the blood. If the generic hits the blood too fast, it could be toxic; too slow, and it might not work. Second is AUC (Area Under the Curve), which represents the total exposure of the body to the drug over time.
| Parameter | What it Measures | Standard Acceptance Range (90% CI) | Narrow Therapeutic Index Range |
|---|---|---|---|
| Cmax | Peak Blood Concentration | 80.00% - 125.00% | 90.00% - 111.11% |
| AUC | Total Drug Exposure | 80.00% - 125.00% | 90.00% - 111.11% |
To prove equivalence, statisticians calculate a 90% confidence interval for the ratio of the generic's average to the brand's average. If that interval falls entirely between 80% and 125%, the drugs are considered bioequivalent. For drugs where a tiny change in dose can be dangerous (Narrow Therapeutic Index drugs), the window is much tighter-usually 90% to 111%.
Common Pitfalls and Expert Tips
Conducting these studies isn't as simple as following a recipe. Many companies fail because they rush the process. A common disaster is underestimating the washout period. If a drug has a long half-life and the next dose is given too soon, the results become "contaminated," which can cost a company hundreds of thousands of dollars in wasted trial fees. This is why experts strongly recommend running a small pilot study before the main trial. A pilot study helps determine the actual variability of the drug in humans, allowing researchers to pick the right number of subjects and sampling times.
Another frequent error is a poorly planned sampling schedule. If you don't take enough samples around the Cmax, you might miss the peak entirely, leaving you with a gap in your data that regulators will not ignore. Real-time PK sample analysis-where the lab checks the concentrations while the trial is still running-can reduce these protocol deviations by nearly 40%.
Alternatives to Traditional PK Studies
While measuring blood levels (Pharmacokinetics) is the preferred method for 95% of generic submissions, it's not always possible. For some drugs, the concentration in the blood doesn't actually tell you if the drug is working.
- Pharmacodynamic Studies: Instead of measuring the drug, scientists measure the drug's effect. For example, with blood thinners like warfarin, they measure how long it takes for blood to clot.
- Clinical Endpoint Studies: Used primarily for topical creams or inhalers. Since the drug stays in the lung or on the skin, blood levels are irrelevant. Success is measured by the actual therapeutic effect on the patient.
- In Vitro Biowaivers: For certain simple drugs (BCS Class I), the European Medicines Agency or FDA may allow the company to skip human trials entirely if they can prove the drug dissolves perfectly in a lab beaker.
Why is the 80-125% range used instead of a perfect 100%?
Because biological systems are naturally variable. Even if you take the exact same brand-name pill twice, your blood levels won't be identical due to changes in diet, hydration, and metabolism. Scientific data shows that differences within the 80-125% range have no clinical impact on the patient's health or the drug's efficacy.
Can a generic drug fail bioequivalence but still be safe?
Technically, yes, but it cannot be sold as a generic substitute. If a drug fails bioequivalence, it means it doesn't release the active ingredient in the same way as the brand. This could mean the drug is absorbed too slowly to treat the condition or too quickly, potentially causing side effects.
How long does the approval process take after the study is done?
Once the data is submitted via an Abbreviated New Drug Application (ANDA), the median review time for a first-cycle approval is around 10.2 months, depending on the regulatory agency and the complexity of the drug.
What happens if a volunteer drops out of the study?
Dropout rates typically range from 5% to 15%. Because crossover studies rely on each person completing both periods, a dropout can weaken the statistical power of the study. Researchers usually over-recruit (e.g., recruiting 36 people when only 30 are needed) to compensate for this.
Are these studies conducted on patients or healthy people?
Most bioequivalence studies use healthy volunteers because they provide a cleaner baseline and are easier to monitor. However, for certain drugs that are too toxic for healthy people, the studies must be conducted on patients who actually need the medication.
