GI Drug Bioequivalence: Testing Locally Acting Oral Drugs
For oral solid dosage forms that act locally in the gastrointestinal (GI) tract, plasma concentration is an unreliable GI drug bioequivalence measure because systemic exposure does not track local drug concentration at the site of action. Regulators accept pharmacokinetic, pharmaceutic, pharmacodynamic, or clinical endpoint methods, selected by drug, release site, and absorption pattern.
Why plasma levels fail for GI locally acting drugs
When a drug acts throughout the body, it enters the bloodstream and reaches the target through circulation, so BE is assessed by measuring drug concentration in body fluid, mainly plasma. For drugs that act locally in the GI tract and are not meant to enter the bloodstream, systemic exposure does not track local drug concentration in the GI tract. Therefore BE evaluation is more complex. A generic does not have to match the reference product in blood concentration to be equivalent at the local site.
Acceptable BE methods: a comparison table
A BE method for GI locally acting oral solid dosage forms must accurately and sensitively measure and compare the rate and extent to which the active ingredient reaches the GI site of action, and it must be reproducible. The main options are pharmacokinetic (PK), pharmaceutic, pharmacodynamic (PD), and clinical endpoint methods.
BE method | Best suited to | What it measures | Representative example |
Pharmacokinetic (PK) | Drugs with measurable systemic absorption; lower GI release | Cmax, AUC, partial AUC | Mesalamine modified release |
In vitro dissolution | Highly soluble immediate-release drugs, same excipients as reference | Dissolution profile across GI-relevant pH range | Acarbose tablets |
In vitro binding | Phosphate, potassium, or bile acid binding drugs | Binding rate, affinity, and capacity | Lanthanum carbonate, sevelamer |
Pharmacodynamic (PD) | Minimal systemic absorption, low water solubility | Percent fecal fat excretion, blood glucose change | Orlistat capsules |
Clinical endpoint | Only when no other method is feasible | Clinical outcome | Rarely used |
Method 1: Pharmacokinetic (PK) studies
PK studies can still reveal product differences even when blood level does not track efficacy. For lower GI locally acting drugs, the release site is often the main systemic absorption site, so PK features such as Cmax and AUC can partly reflect differences in release across GI regions. For modified-release products, if part of the exposure correlates with a specific absorption site, partial AUC can distinguish release at the target site from release elsewhere.
Example: mesalamine oral modified-release products are absorbed in the colon, the site of action in ulcerative colitis, so Cmax, AUC, and partial AUC can be used to evaluate BE. Because total AUC and Cmax cannot distinguish release in different GI regions, partial AUC is needed to reflect colon-only release. Mesalamine has a short half-life of about 42 minutes, so systemic exposure mainly comes from absorption at the action site, and PK features can partly reflect local concentration there. Mesalamine blood concentration varies widely; scaling average bioequivalence to the reference product handles this and reduces the number of subjects needed.
Safety caveat: if systemic exposure creates a safety risk, an additional PK study is needed to prove generic systemic exposure is not higher than reference. Example: potassium bismuth citrate is a gastric mucosal protector but absorbs a small amount of bismuth. When blood bismuth exceeds 0.1 microgram per milliliter, bismuth encephalopathy may occur. Therefore granules, capsules, and tablets all require PK comparison to prove generic bismuth absorption is not higher than reference.
Method 2: Pharmaceutic methods
In vitro dissolution. For lower GI locally acting immediate-release products with high drug solubility and excipients (qualitative and quantitative) identical to the reference, dissolution testing alone may be sufficient to evaluate BE. High solubility across the GI pH range ensures most drug is in solution when it reaches the lower GI site, so in vitro dissolution predicts in vivo release. Dissolution comparison at different pH values proves similar in vivo release in each GI region. If excipients match, their effect on GI transit is also expected to match.
Example: acarbose dissolves well in water but is almost not absorbed after oral intake, with bioavailability of just 1 percent to 2 percent. If an acarbose tablet generic and the reference have identical Q1 and Q2, a dissolution study in multiple media covering the GI pH range can be used to evaluate BE.
For complex dosage forms such as modified-release products, dissolution alone is limited because GI transit and the GI environment are complex. Do not rely on dissolution alone for BE. If the active ingredient dissolves in aqueous media, dissolution can predict in vivo performance and systemic exposure is measurable, then combined dissolution in multiple media representing GI conditions plus PK can prove BE. Example: mesalamine modified-release products. For poorly soluble drugs or drugs acting in the mouth and throat, routine dissolution cannot reliably predict in vivo performance and is not recommended as a BE method.
In vitro binding. For some GI locally acting oral solid dosage forms, such as binding agents, an in vitro binding study that reflects the mechanism of action and has clear clinical relevance and sensitivity can evaluate BE. This method has relatively low variability and is easier to control, so it is more likely to detect product differences. Phosphate binding drugs such as lanthanum carbonate chewable tablets, calcium acetate tablets, and sevelamer carbonate tablets bind phosphate in the upper GI tract to form insoluble complexes that are excreted in feces. In vitro phosphate binding is directly related to the mechanism of action and reflects local bioavailability at the site of action. Therefore these drugs should use in vitro binding to evaluate BE. Kinetic binding compares the binding rate and time to reach equilibrium between products; equilibrium binding compares binding affinity and binding capacity.
Method 3: Pharmacodynamic (PD) studies
PD endpoints usually vary more than PK endpoints and pharmaceutic methods. Use PD studies when PK and pharmaceutic methods are not applicable. PD effect versus dose typically follows an S-shaped curve. Conduct a pilot study to determine a suitable dose in the linear range of the dose-response curve and a sample size with enough statistical power for the key study.
Example: orlistat capsules have minimal systemic absorption and low water solubility, so PK and pharmaceutic methods are not suitable. Use percent fecal fat excretion (%FFE), relative to daily fat intake, as the endpoint for a PD study to evaluate BE. Orlistat has a dose-response relationship with %FFE, and an Emax model can characterize the curve. The labeled minimum dose is 60 mg three times daily (180 mg per day), which is already in the nonlinear range of the dose-response curve. At this dose, even a large difference in the amount of drug reaching the action site produces only a small difference in PD effect, so the effect dimension cannot accurately reflect relative bioavailability differences. A dose-scaling method is therefore needed. Fit the dose-response curve (Emax model) to reference %FFE data at different doses (baseline, 60 mg, 120 mg), then determine the reference dose corresponding to the generic average effect. Evaluate BE using the ratio F of the reference dose corresponding to the generic to the actual generic dose (60 mg three times daily). The 90 percent confidence interval of F should be within 80.00 percent to 125.00 percent.
Method 4: Clinical endpoint studies
Clinical endpoint studies are generally not recommended for BE. Doses are usually on the plateau of the dose-response curve, so even large dose differences produce only small effect differences; this is the least sensitive method for detecting product differences. These studies also need many subjects and recruitment can be difficult. Use this method only in special cases when no other feasible method exists. Domestic experience with clinical endpoint BE studies for GI locally acting drugs is relatively limited.
A counter-intuitive conclusion
Low or zero systemic exposure is not, by itself, enough reason to exempt a drug from BE testing. When a drug has no systemic absorption and PK is not feasible, pharmaceutic, PD, or even clinical endpoint methods can still evaluate BE. The appropriate method must be built around the drug, the formulation, the mechanism of action, and the target site.
Practical recommendations
- Choose the BE method by drug, release site, indication, route, whether systemic absorption exists, physicochemical properties, formulation design, and safety and efficacy features.
- For lower GI locally acting immediate-release products with high solubility and matching excipients, consider dissolution testing across GI-relevant pH media first.
- For modified-release products, combine dissolution and PK; use partial AUC when release site matters.
- For phosphate, potassium, or bile acid binders, use in vitro binding with dissolution support.
- For minimally absorbed, poorly soluble drugs such as orlistat, use a PD endpoint such as %FFE with dose scaling.
- Involve pharmaceutics, biopharmaceutics, clinical pharmacology, and clinical medicine experts to select a method that fits the drug.
Limitations
BE method selection depends on product-specific factors, and the examples here illustrate principles rather than serve as a validated protocol for any single product. Some drugs allow more than one acceptable method; the choice should match the actual situation. Not every numerical indicator in this article applies to every formulation.
Bottom line: GI drug bioequivalence cannot rely on blood concentration alone
GI drug bioequivalence for locally acting oral solid dosage forms cannot rely on blood concentration alone. Match the method to the release site and absorption profile, and use pharmacokinetic, pharmaceutic, pharmacodynamic, or clinical endpoint methods alone or in combination as appropriate.
FAQ
Why is plasma concentration a poor measure of GI drug bioequivalence?
Plasma concentration reflects systemic exposure, but for GI locally acting drugs systemic exposure does not correlate with local drug concentration at the site of action. A product can be equivalent in the gut while showing different blood levels, so plasma data alone cannot establish GI drug bioequivalence.
When can dissolution testing alone support a BE decision?
Dissolution alone may suffice for lower GI locally acting immediate-release products with high drug solubility and excipients identical to the reference, because high solubility across the GI pH range keeps most drug in solution when it reaches the target site.
What role does partial AUC play in mesalamine BE studies?
Total AUC and Cmax cannot distinguish release in different GI regions. Partial AUC reflects release limited to the colon, the site of action, and therefore supports BE for mesalamine modified-release products.
Why is an in vitro binding study used for phosphate binder BE?
Phosphate binders such as lanthanum carbonate act by binding phosphate in the upper GI tract. In vitro binding is directly tied to the mechanism of action and reflects local bioavailability, so it is suitable for BE of these products.
What acceptance range applies to the lanthanum binding constant ratio?
For lanthanum carbonate, the 90 percent confidence interval of the equilibrium binding capacity constant ratio (k2) between products should be within 80.00 percent to 120.00 percent under GI-relevant pH conditions.
How is fecal fat excretion used as a PD endpoint for orlistat?
Orlistat has minimal absorption, so percent fecal fat excretion relative to daily fat intake serves as a quantitative pharmacodynamic indicator of local bioavailability at the GI site of action.
Why does orlistat need a dose-scaling approach?
The labeled 180 mg per day dose sits on the nonlinear part of the dose-response curve, so a large difference in drug reaching the site may produce only a small effect difference. Dose-scaling converts the insensitive effect difference into a more sensitive dose difference using the F ratio.
When is a clinical endpoint study acceptable?
Only in special cases where no pharmacokinetic, pharmaceutic, or pharmacodynamic method is feasible. Clinical endpoint studies are the least sensitive option and usually require many subjects.
Does lack of systemic absorption exempt a drug from BE testing?
No. Low or zero systemic exposure is not enough reason for exemption. When PK is not feasible, pharmaceutic, PD, or clinical endpoint methods can still establish BE.
Which products need an extra safety-focused PK comparison?
Products whose systemic exposure creates a safety risk, such as potassium bismuth citrate where blood bismuth above 0.1 microgram per milliliter may cause bismuth encephalopathy, need PK comparison to prove generic absorption is not higher than reference.
How do dissolution media pH conditions support BE?
Comparing dissolution at multiple pH values across the GI range proves similar in vivo release in each GI region, which supports local bioavailability and BE.
Which team should select the BE method for a product?
A multidisciplinary team including pharmaceutics, biopharmaceutics, clinical pharmacology, and clinical medicine experts should select the method based on the drug, formulation, mechanism of action, and target site.