How Bioequivalence Studies Are Conducted: Step-by-Step Process

When a generic drug hits the shelf, you might assume it’s just a cheaper copy. But behind every generic pill is a rigorous scientific process designed to prove it works exactly like the brand-name version. That process is called a bioequivalence study. It’s not guesswork. It’s not marketing. It’s hard science-measured in blood samples, statistical models, and tightly controlled clinical trials. And if it fails, the drug doesn’t get approved. Here’s exactly how it’s done.

Why Bioequivalence Matters

Before a generic drug can be sold, regulators like the FDA, EMA, and PMDA need proof that it behaves the same way in the body as the original. This isn’t about ingredients alone. Two pills can have identical active ingredients but different shapes, coatings, or fillers. These differences can change how fast or how much of the drug enters your bloodstream. That’s dangerous if you’re taking something like warfarin or lithium, where tiny changes can cause serious side effects.

Bioequivalence studies answer one question: Does the generic deliver the same amount of drug, at the same speed, as the brand? If yes, it’s considered therapeutically equivalent. The FDA estimates that since 2010, generic drugs saved the U.S. healthcare system over $1.68 trillion. None of that would be possible without these studies.

The Gold Standard: Crossover Design

Most bioequivalence studies use a two-period, two-sequence crossover design. That sounds complicated, but here’s what it actually means:

• 24 to 32 healthy volunteers (sometimes more, sometimes fewer) sign up.
• Half get the generic drug first, then the brand-name drug after a break.
• The other half get the brand-name drug first, then the generic.
• There’s a washout period between doses-usually five times the drug’s elimination half-life. For a drug that clears in 12 hours, that’s 60 hours. For a long-acting drug like dulaglutide, it could be weeks.

This design controls for individual differences. If one person naturally absorbs drugs slowly, they’ll still be compared to themselves-just under two different conditions. That’s more accurate than comparing two separate groups of people.

Collecting the Data: Blood Samples and Timing

After each dose, volunteers give blood samples at specific times. The schedule isn’t random. It’s calculated to capture the full journey of the drug in the body.

• Before dosing (time zero) - baseline level.
• Just before the expected peak (Cmax) - when the drug hits its highest concentration.
• Two samples around that peak - to map the rise and fall.
• Three or more samples during elimination - to track how the body clears it.

Sampling continues until the area under the curve (AUC) reaches at least 80% of the total possible exposure (AUC∞). That usually means collecting samples for 3 to 5 half-lives. For most drugs, that’s 12 to 48 hours. For slow-clearing drugs, it can stretch to days.

The blood is spun down to plasma or serum. Then, it’s analyzed using liquid chromatography-tandem mass spectrometry (LC-MS/MS)-the most precise method available. The lab must prove its method is accurate within ±15% (±20% at very low levels). If the assay isn’t validated, the whole study is thrown out.

The Numbers That Matter: Cmax and AUC

Two numbers determine success:

• Cmax: The highest concentration of drug in the blood.
• AUC(0-t): The total exposure over time, from dosing to the last measurable point.

Sometimes AUC(0-∞) is used too, if the full elimination curve is captured.

These values are log-transformed because drug concentrations follow a logarithmic scale. Then, statisticians run an ANOVA model with factors for sequence, period, treatment, and subject. The goal? To calculate the 90% confidence interval (CI) for the geometric mean ratio of test (generic) to reference (brand).

The Pass/Fail Line: 80%-125%

Here’s the rule: For the study to succeed, the 90% CI for both Cmax and AUC must fall between 80.00% and 125.00%. That means the generic delivers between 80% and 125% of the brand’s exposure.

For drugs with a narrow therapeutic index-like phenytoin, digoxin, or cyclosporine-the window tightens to 90.00%-111.11%. These drugs have little room for error. Even a 10% difference could be dangerous.

If the CI slips outside those limits, the study fails. No approval. No market. That’s why companies run pilot studies first. According to FDA data, pilot studies cut failure rates from 35% down to under 10%.

What Happens When the Drug Is Highly Variable?

Some drugs vary wildly from person to person. Their within-subject coefficient of variation (CV) is over 30%. For these, the standard 80-125% rule doesn’t work well.

The EMA requires replicate crossover designs-often four periods with multiple doses of both products. This gives more data to estimate variability. The FDA allows reference-scaled average bioequivalence (RSABE), which adjusts the acceptance range based on how variable the drug is. Both approaches are scientifically valid but require more subjects-sometimes 50 to 100.

Other Study Designs

Crossover isn’t always possible. For drugs with half-lives longer than two weeks, a parallel design is used: one group gets the generic, another gets the brand. No washout needed. But you need more people to compensate for the lack of within-subject comparison.

For extended-release products, multiple-dose studies are required. You can’t just give one dose and assume it behaves the same over time. The body builds up the drug, and the release profile must be consistent.

In rare cases, pharmacodynamic or clinical endpoint studies are used. For example, with anticoagulants, you measure INR levels. For inhalers or topical creams, you might measure skin absorption or lung deposition. The FDA requires this for certain complex products.

Dissolution Testing: The Silent Partner

Even if a drug passes the blood test, regulators check how it dissolves. A tablet must release its drug at a similar rate across different pH levels (1.2 to 6.8), mimicking the stomach and intestines.

The f2 similarity factor must be above 50. That’s calculated by comparing the dissolution profiles of the generic and brand over time. At least 12 units are tested per condition. If the dissolution curves don’t match, the study fails-even if the blood levels look perfect.

What Goes Wrong?

Bioequivalence studies are expensive and complex. A single failure can cost $250,000 and delay approval by months.

Common reasons for failure:

• Washout period too short (45% of deficient studies).
• Wrong sampling times (30%).
• Statistical errors or wrong models (25%).
• Unvalidated analytical methods (22% of studies face delays).

One company, Alembic Pharmaceuticals, had its generic version of Trulicity rejected in 2022 because Cmax values were inconsistent across multiple studies. Another, Teva, got Januvio approved after just one successful study with 36 subjects. The difference? Preparation.

Who Runs These Studies?

Most are done by Contract Research Organizations (CROs). These are specialized labs with clinical units, bioanalytical labs, and statisticians trained in bioequivalence models. The FDA processes about 2,500 bioequivalence submissions each year. The average review time is just over 10 months.

The people involved need real expertise:

• Clinical staff: 6-12 months of BE study experience.
• Biostatisticians: Must know BE-specific ANOVA models.
• Bioanalytical scientists: Must master LC-MS/MS validation.

The Future: Modeling, Simulation, and Waivers

The field is evolving. More studies now use physiologically based pharmacokinetic (PBPK) modeling to predict how a drug will behave-cutting down on human testing.

The FDA also grants biowaivers for BCS Class I drugs-those that are highly soluble and highly permeable. If the dissolution profile matches and the drug meets these criteria, no blood study is needed. In 2022, 27% of approvals used this route.

The FDA’s 2024-2028 plan aims to reduce study requirements by 30% using real-world evidence. But for now, the blood test remains king.

Final Thought: It’s Not Magic, It’s Measurement

Bioequivalence studies aren’t about proving a generic is ‘good enough.’ They’re about proving it’s identical in how it works inside the body. Every drop of blood, every hour of sampling, every statistical calculation is there to protect patients. That’s why a generic drug approved in the U.S. works the same as the brand-because science demanded it.