Bioequivalence Evidence for Generic Nasal Sprays and Orally Inhaled Products: FDA vs. EMA

Bioequivalence for generic nasal sprays and orally inhaled products: how FDA and EMA diverge

Locally acting nasal sprays and orally inhaled drug products (OINDPs) are among the hardest generics to bring to market, and the reason is physiological. The drug acts where it lands, on the nasal mucosa or in the lung, and the plasma concentration that ordinary oral generics rely on generally does not reflect how much drug reached that site of action. FDA and EMA both accept that a single pharmacokinetic (PK) crossover study cannot answer the bioequivalence (BE) question for these products, and both assemble evidence from in vitro performance, PK, and clinical or pharmacodynamic (PD) data. The difference lies in how the two agencies combine that evidence. FDA builds an aggregate "weight-of-evidence" package in which the components sit alongside one another, while EMA runs a stepwise hierarchy in which an applicant climbs to the next tier of evidence only when the previous one fails.

Why these products need more than a PK study

For a locally acting nasal or inhaled product, systemic plasma concentrations generally do not reflect the amount of drug delivered to the nasal site of action, so BE cannot rest on PK alone ⁶. FDA therefore assesses both local delivery and systemic exposure, treating in vitro tests as central because they are less variable, more controllable, and can be more sensitive than clinical endpoints for detecting product differences ²⁶. The same logic applies to the lung: FDA notes that for products where the systemic circulation does not adequately represent local exposure, systemic PK may not by itself establish BE, and characterises metered-dose inhalers (MDIs) as complex dosage forms and complex drug-device combination products whose lung delivery depends on patient-device interaction, inspiratory flow, and peak inspiratory flow ^76837577. EMA reaches the same conclusion from the opposite direction, treating PK endpoints as surrogate markers for the pattern and extent of lung deposition and for systemic exposure, and hence as surrogates for efficacy and safety, rather than as direct measures of the clinical effect ³³.

The FDA approach: an aggregate weight of evidence

FDA's framework, articulated for nasal products in its 2003 draft BA/BE guidance and operationalised product by product through Product-Specific Guidances (PSGs), splits along the solution-versus-suspension line and, for OINDPs, offers menued options that trade off in vitro, PK, and clinical work.

Nasal solutions versus nasal suspensions

For a locally acting nasal solution spray or aerosol, FDA's recommended pathway is essentially in vitro. BE is supported by qualitative and quantitative (Q1/Q2) formulation sameness, a comparable container-closure/delivery system, and in vitro evidence of equivalent performance, on the premise that in vitro tests are more sensitive indicators of delivery to nasal sites of action than clinical studies for solution formulations ².

For a nasal suspension, FDA requires in vivo evidence in addition to in vitro testing, because the drug particle-size distribution in the emitted spray cannot be fully characterised and particle size affects availability both locally and systemically ². The aggregate showing for a suspension ANDA is formulation Q1/Q2 sameness, a comparable container-closure system, equivalent in vitro tests, equivalent systemic exposure or absorption, and an equivalent local-delivery study ⁵. Where systemic concentrations are too low to measure, FDA recommends a comparative clinical-endpoint study for local delivery plus a PD or clinical-endpoint study to address systemic absorption ³.

The fluticasone propionate nasal spray PSG (RLD NDA 020121) illustrates how this plays out. FDA offers two routes: either eight in vitro BE studies, or six in vitro studies plus one PK study and one comparative clinical-endpoint study ¹³¹. The eight in vitro tests are single actuation content, droplet-size distribution by laser diffraction, spray pattern, plume geometry, priming/repriming, drug in small particles/droplets, drug particle-size distribution, and dissolution ^116124126. The PK arm is a fasting, single-dose, two-way crossover measuring plasma fluticasone propionate, with an analytical limit of quantitation FDA suggests should sit below 1 pg/mL ¹¹⁵. The clinical arm is a randomised, double-blind, three-arm, placebo-controlled, parallel-group study in seasonal allergic rhinitis, with a 7-day placebo run-in and 14-day treatment, whose primary endpoint is the change from baseline in reflective total nasal symptom score (rTNSS) ¹¹⁴¹¹⁵. Notably, FDA does not want a double-dummy design here, because the extra nasal fluid volume could wash drug off deposition sites ¹¹⁴. For the all-in-vitro option, the generic must have no formulation differences that could significantly affect local or systemic availability, with Q1/Q2 sameness given as the example of how to satisfy that expectation ¹³¹. FDA also expects the generic device design to be informed by the RLD's size, shape, external critical design attributes, and operating principles ¹²³.

Orally inhaled products: menued options

For OINDPs, FDA's PSGs again present alternative packages that always combine in vitro BE with in vivo work. For albuterol sulfate MDI (RLD-referenced), three options are offered: three in vitro studies plus one PK study; three in vitro studies plus two PK studies (the second a charcoal-block study); or two in vitro studies plus one PK study plus a pharmacodynamic bronchoprovocation study ⁶⁷. The PK study is designed to capture early lung exposure, collecting samples within 10, 20, and 30 minutes post-dose and using partial AUCs (pAUC0-10, pAUC0-20) as supportive lung-exposure data, with BE judged on 90% confidence intervals for pAUC0-30, AUC0-t, AUC0-inf, and Cmax ⁵⁵. The bronchoprovocation option uses methacholine PC20/PD20 as the PD endpoint, analysed on the dose scale with an acceptance interval for relative bioavailability of 67.00% to 150.00% ⁶¹⁶². Q1/Q2 sameness applies, with Q2 defined as inactive-ingredient concentrations within +/-5% of the reference ⁶⁷.

For budesonide inhalation powder (DPI, RLD-referenced), FDA again offers two packages: four in vitro studies plus one comparative-characterization study plus two PK studies; or two in vitro studies plus one PK study plus one comparative clinical-endpoint study ⁹⁴⁹⁰. The in vitro battery includes single actuation content and aerodynamic particle-size distribution (APSD) tested across 30, 60, and 90 L/min flow rates, dissolution, and a particle-morphology/agglomeration comparison ¹⁰¹⁹⁴. One of the PK studies is a charcoal-block study, and FDA encourages a pre-ANDA meeting to align on the charcoal dose and overall strategy before it is run ⁹⁰. FDA analyses in vitro metrics such as single actuation content and APSD by population bioequivalence (PBE) ¹⁰¹⁹⁴. Later inhalation PSGs cross-reference the older budesonide inhalation suspension guidance for the in vitro PBE method and acceptance criteria ^9099101.

Across these PSGs, the recurring FDA elements are Q1/Q2 formulation sameness, device design comparability against the RLD, a defined in vitro test battery evaluated by PBE, PK studies (often including a charcoal-block arm to isolate lung deposition), and, where needed, a comparative clinical-endpoint or PD study.

The EMA approach: a stepwise therapeutic-equivalence hierarchy

EMA frames the question as demonstrating therapeutic equivalence (TE): test and reference must be comparable enough in efficacy and safety that clinically relevant differences can be reliably excluded ⁴⁴. Its OIP guideline lays out an explicit three-step ladder, and the applicant stops as soon as a step succeeds.

Step 1, comparative in vitro data. In vitro comparability data must always be submitted, even when they are not sufficient on their own ³³. TE can be concluded from in vitro data alone only if every one of a defined set of conditions is met: the same active substance including the same salt/ester/hydrate/solvate; an identical dosage form; no solid-state differences affecting performance; justified excipient differences; similar device handling; device airflow resistance within +/-15%; target delivered dose within +/-15%; and similar APSD ⁵⁰. The APSD comparison must use the complete individual-stage profile from a validated multistage impactor, with stage grouping allowed only under justified conditions ⁵⁰. If all criteria are satisfied, no in vivo data are needed ⁴⁴.

Step 2, PK studies, if in vitro evidence is insufficient. PK studies address two questions in parallel: efficacy, through lung deposition/absorption, and safety, through total systemic exposure ³⁴. For the efficacy read-out, where gastrointestinal absorption contributes 5% or more of systemic exposure it may be blocked with activated charcoal, or a partial-AUC approach may be used, so that AUC0-t reflects the amount reaching the lung while Cmax flags differences in intrapulmonary deposition pattern ³³³⁴³⁵. For the safety read-out, total systemic exposure is compared without charcoal so both lung and GI absorption are captured, and equivalence is met if the test AUC0-t and Cmax are equivalent to or lower than the reference ³⁵. If PK demonstrates TE for every active substance, no PD or clinical confirmation is required ³⁴.

Step 3, PD and, where necessary, clinical studies. If PK fails for an active substance, EMA's first recommendation is to consider reformulating. A sensitive PD model may then be used to confirm equivalent efficacy or safety, but PD and clinical endpoints are explicitly regarded as insensitive and are not the preferred route; stand-alone clinical data are required only when PK plus PD together still cannot establish TE ³⁴³³.

EMA and nasal sprays

For locally acting nasal sprays, EMA works from its guideline on the pharmaceutical quality of inhalation and nasal products together with the BE and locally-applied/locally-acting clinical guidance ¹⁵. The core expectation is a comprehensive, protocol-driven in vitro comparison against the reference, with pre-specified methods and acceptance criteria and a justification for every parameter selected or omitted ¹⁶. Relevant parameters include the complete droplet-size distribution (typically by laser diffraction), the mass fraction of droplets below 10 microns, delivered dose, spray pattern and plume geometry, rheology (thixotropy and viscosity), surface tension, pH, density, osmolality, and buffer capacity ¹⁶¹¹. Because nasal products should deposit locally, EMA wants evidence that the great majority of droplets exceed 10 microns, with the sub-10-micron fraction assessed to minimise unintended lung delivery ¹⁹²³. Where systemic exposure from a locally acting product could cause systemic adverse reactions, EMA requires it to be measured, and the test product must not exceed the reference: the upper bound of the 90% confidence interval must not exceed 125% ³⁰.

EMA also offers a solution waiver analogous to FDA's in vitro route: comparative equivalence data may be waived for a locally acting nasal-spray solution when the test product is the same type of solution (aqueous or oily), at the same concentration of the same active substance, with the same method of administration and identical or essentially similar pharmaceutical properties; only minor, justified excipient differences are tolerated ³⁰. Suspensions cannot use this waiver and must address the additional solid-state, delivery-performance, and deposition variables, with residual uncertainty resolved through justified PK and/or comparative clinical or PD evidence ¹⁶²³.

Where the two frameworks diverge

The practical consequence is a real difference in study burden and sequencing. Under EMA, a well-matched OINDP that clears the Step 1 in vitro gate can, in principle, avoid human studies entirely, and even when it cannot, a successful PK package ends the exercise without any clinical endpoint study ⁵⁰³⁴. Under FDA, the picture is split: a nasal suspension whose PSG offers an in vitro-only option can rely on that route when Q1/Q2 sameness and device sameness are shown, whereas OINDPs more consistently still need an in vivo arm chosen from the PSG menu, on top of in vitro BE evaluated by PBE, one or more PK studies, and, depending on the option chosen, a comparative clinical-endpoint or PD study, all supported by Q1/Q2 sameness and device comparability ⁵⁹⁴⁶⁷. FDA's clinical-endpoint studies (rTNSS in allergic rhinitis, bronchoprovocation in asthma) are a standard fixture of its PSGs, whereas EMA actively steers applicants away from clinical endpoints on the grounds of insensitivity and toward reformulation if PK fails ^1146234. Because FDA's expectations are set product by product in PSGs and EMA's flow from a general OIP TE framework plus quality guidance, an applicant developing for both markets should read the specific FDA PSG and the EMA OIP guideline side by side early, since the in vitro test lists, the device-sameness metrics, and above all the point at which human data become unavoidable differ in ways that shape the entire development program.