Bioequivalence Overview Pillars Substrate History · Evolution Current 2026 Future 2026-2035 AI lens Flow of trials Players & stakeholders Notes
Library/Bioequivalence Trials/Pillars
chapter 01 · the regulatory spine

The eight BE pillars · six regulators.

Quick-reference grid of the eight pillars that every bioequivalence study must satisfy · then the cross-regulator comparison drilldown for each. FDA · EMA · ICH M13A · ANVISA · WHO · CDSCO. The flagship chapter for the BE regulatory spine.

/ 00

The deep reference: iFeed.bioequivalence.

Cross-regulator · live

iFeed maintains a deep reference for bioequivalence under the sub-brand iFeed.bioequivalence. The foundation document is a cross-regulator comparison across six regulators (FDA · EMA · ICH M13A · ANVISA · WHO · CDSCO/PMDA) and the eight BE pillars that form the regulatory spine. Built on the official BE guidelines, the 50-year arc from Lindenbaum 1971, and operational substrate from sponsor-CRO contract studies in fed/fasted designs, replicate crossovers, and biowaiver filings.

/ Foundation document

6 regulators · 8 BE pillars · one reference.

Pick a pillar. Read the convergence/divergence summary. Compare 6 regulators side-by-side. Each acceptance criterion highlighted. Designed for the BE study director · the QA auditor · the cross-region sponsor preparing simultaneous filings.

FDA 21 CFR 320EMA CPMP/EWP/QWP/1401/98ICH M13AANVISA RDC 742WHO TRS 992CDSCO · PMDA
/ Universal acceptance bound

The 80–125% TOST envelope.

Geometric mean ratio of test/reference for AUC and Cmax must fall within 80.00–125.00% (90% CI), Schuirmann 1987 two one-sided tests. Universal across FDA, EMA, ICH M13A, ANVISA, WHO, CDSCO. The single most-converged number in pharmaceutical regulation.

80–125
/ 00b

The eight BE pillars.

The regulatory spine

These eight pillars are what every bioequivalence package must satisfy to be regulator-complete. Below: a quick-reference grid · then the cross-regulator comparison drilldown for each. Reference-product handling is where most non-US filings fail first; HV drug scaling (★) is where FDA-EMA divergence runs deepest.

Quick reference · the eight pillars.

/ 4.1
Acceptance bound · 80–125% TOST.

Geometric mean ratio of AUC and Cmax inside 80.00–125.00% (90% CI). Schuirmann two one-sided tests. The single universal number across all six regulators.

/ 4.2
Study design.

2×2 crossover (default) · replicate (HV drugs / RSABE) · parallel (long half-life). Sample-size driven by within-subject CV and expected GMR.

/ 4.3
Reference product handling.★

Where regulator divergence runs deepest. FDA Reference Listed Drug (RLD) · EMA reference Member State product · ANVISA medicamento de referência · WHO comparator product list. Sourcing chain documentation mandatory.

/ 4.4
Biowaivers · BCS Class I & III.

ICH M9 (Step 4 Nov 2019) harmonised biowaiver scope to BCS Class I (high solubility, high permeability) and Class III (high solubility, low permeability) under specific dissolution criteria.

/ 4.5
HV drug scaling · RSABE.★

Reference-scaled average BE for highly variable drugs (within-subject CV > 30%). FDA: AUC and Cmax both scalable. EMA: Cmax-only scaling. The persistent post-2010 FDA-EMA divergence.

/ 4.6
Multiple-dose · steady-state.

Required for non-linear PK, time-dependent kinetics, controlled-release formulations, drugs with long half-life. Cmin,ss / AUCτ / fluctuation. Specific FDA product-specific guidance per product.

/ 4.7
Fed / fasted conditions.

Fasted study default. Fed study mandatory for: modified-release, food-effect labelled, BCS Class II (low solubility) products. FDA high-fat high-calorie meal, EMA standardised meal protocol.

/ 4.8
NTI drugs · narrow therapeutic index.

FDA 90.00–111.11% bounds, full replicate design, warfarin-anchored. EMA list-based with tightened 90.00–111.11% Cmax bound. ANVISA NTI list updated under RDC 742/2022.

Cross-regulator comparison · FDA · EMA · ICH M13A · ANVISA · WHO · CDSCO.

FDA EMA ICH M13A ANVISA WHO CDSCO · PMDA
/ 4.1 Acceptance bound · 80–125% TOST.90% CI of GMR (test/reference) inside 80.00–125.00% · Schuirmann TOST. +
At a glance · convergence vs divergence
CONVERGE
All six regulators converge on 80.00–125.00% as the default acceptance bound for AUC and Cmax. 90% CI of geometric mean ratio (log-transformed). Schuirmann's TOST is the universal statistical engine. The single most-harmonised pillar in BE.
DIVERGE
NTI overrides: FDA & EMA tighten to 90.00–111.11%. HV scaling widens both bounds (see 4.5). Point-estimate constraint: FDA, ICH M13A, EMA require GMR fall inside 80–125% regardless of CI width.
FDA21 CFR 320
1992 amendment
Log-transformed 90% CI · both AUC0-t and Cmax inside 80.00–125.00%. Point-estimate constraint codified in product-specific guidance.
Acceptance criteria80.00–125.00% · 90% CI · AUC + Cmax
EMACPMP Rev 1
2010 BE Guideline
Same numerical bound. Point-estimate constraint reinforced in 2010 revision · tightened 0.90-1.11 for Cmax of NTI drugs.
Acceptance criteria80.00–125.00% · AUC + Cmax
ICH M13AStep 4 · Jul 2024
Harmonised IR oral BE
Single-dose IR oral BE. M13A §4.2.6: 80.00–125.00% 90% CI on log-transformed AUC0-t and Cmax. Major harmonisation milestone.
Acceptance criteria80.00–125.00% · harmonised default
ANVISARDC 742/2022
Effective 1 Mar 2023
RDC 742 §15: same numerical bound. Replaced RDC 31/2010. Brazilian-distinctive: stronger CRO inspection regime around the test.
Acceptance criteria80–125% · same TOST
WHOTRS 992
Annex 7 · multisource
WHO Prequalification programme. Same TOST bound. Explicitly references 80–125% in Annex 7 §9.4.
Acceptance criteria80–125% · PQ pathway
CDSCOD&C Rules
India / PMDA aligned
CDSCO Guideline on BE Studies (2005, revisions 2018+). PMDA aligned via ICH M13A. Same numerical bound.
Acceptance criteria80–125%
Inspector's eye
Inspectors flag two failure modes here: GMR-only reporting without CI bounds, and CI bounds reported on linear (untransformed) scale. M13A §4.2.6 explicitly requires log-transformation before TOST. Pre-M13A submissions sometimes used reverse arithmetic conversion · flagged as out-of-compliance from July 2024.
/ 4.2 Study design.2×2 crossover default · replicate for HV drugs · parallel for long t½. +
At a glance · convergence vs divergence
CONVERGE
All converge on 2×2 crossover as the default for IR oral BE. Replicate (3-period or 4-period) is the canonical design for HV drugs across FDA, EMA, ICH M13A, ANVISA, WHO. Parallel design accepted when half-life precludes crossover.
DIVERGE
FDA partial-replicate (3-period) permits Cmax-AUC scaling. EMA full-replicate (4-period) required for ABEL (average bioequivalence with expanding limits), Cmax only. ANVISA RDC 742 follows EMA. WHO TRS 992 accepts both.
FDA2003 + PSGs
Partial-replicate · 3-period
Default 2×2 crossover. Partial-replicate (RTR or RTRR) acceptable for RSABE per product-specific guidance. Parallel for half-life > 24h.
Acceptance criteria2×2 default · 3-period RSABE
EMA2010 GL §4.1.5
Full-replicate · 4-period
Default 2×2. Full-replicate (RTRT/TRTR or RRTT/TTRR) required for ABEL. Within-subject variability of reference must be calculated.
Acceptance criteria2×2 · 4-period ABEL
ICH M13A2024
§5.1 design framework
M13A §5.1: harmonised crossover and parallel design framework. Replicate designs deferred to product-specific.
Acceptance criteria2×2 default · replicate per PSG
ANVISARDC 742/2022
EMA-aligned
RDC 742 follows EMA on full-replicate for HV drugs. Brazilian healthy-volunteer registry constrains overlapping recruitment.
Acceptance criteria2×2 + 4-period replicate
WHOTRS 992 §9.3
Both replicates accepted
PQ studies accept both 3- and 4-period replicate designs. WHO comparator product source guidance applies.
Acceptance criteria2×2 + replicate flexible
CDSCO2005 + 2018
2×2 standard
CDSCO BE Guideline. Default 2×2 crossover. Replicate accepted per individual sponsor justification.
Acceptance criteria2×2 default
/ 4.3 Reference product handling.★RLD vs reference Member State vs Brazilian RP vs WHO comparator · sourcing divergence. +
At a glance · convergence vs divergence · the deepest pillar
CONVERGE
All converge on the principle: a defined reference product, sourced from regulated supply, with full chain-of-custody documentation through dosing.
DIVERGE
FDA: Reference Listed Drug per Orange Book · US-sourced. EMA: any innovator product authorised in any EU/EEA Member State, sourced from that Member State. ANVISA: medicamento de referência from RDC list, Brazilian-sourced. WHO: comparator product list (CPL). CDSCO: Indian-marketed innovator. Cross-region dossiers must run separate BE studies against separate reference products.
FDAOrange Book
★ RLD · US-sourced only
Reference Listed Drug per Orange Book Pharmacology Class. US-sourced. Sponsor cannot use foreign-sourced RLD even if same NDA holder.
Acceptance criteriaRLD · US source
EMA2010 §3.5
★ Member-State-sourced
Innovator product authorised in any EU/EEA Member State, sourced from that Member State. Multi-state references require justification.
Acceptance criteriaEU/EEA-sourced innovator
ICH M13A2024
§5.2 sourcing principles
M13A §5.2 harmonises sourcing principles but defers to regional regulator on the actual reference product. Mutual acceptance NOT achieved here.
Acceptance criteriaRegional regulator · sourcing chain
ANVISARDC 742 §III
★ Brazilian RP list
Medicamento de referência from ANVISA published list. Brazilian-marketed. Importation chain documentation mandatory under RDC 742.
Acceptance criteriaBrazilian RP only
WHOTRS 992 §6
CPL · comparator list
WHO Comparator Product List. Updated periodically by Expert Committee on Specifications. WHO-PQ-bound studies must use CPL.
Acceptance criteriaCPL-listed comparator
CDSCO2005 + 2018
India-marketed innovator
India-marketed innovator product. Where no innovator marketed in India, foreign-sourced product accepted with justification.
Acceptance criteriaIndian-marketed RP
Inspector's eye
Reference-product sourcing chain is the most-flagged finding in cross-region BE submissions. Sponsors filing simultaneously in US + EU + Brazil run three separate BE studies against three separate reference products. Single-study cross-region filings (using one reference product) are routinely rejected at the second-region review stage.
/ 4.4 Biowaivers · BCS Class I & III.ICH M9 harmonised BCS biowaiver scope across ICH regions. +
At a glance · convergence vs divergence
CONVERGE
ICH M9 (Step 4 Nov 2019) harmonised BCS-based biowaiver scope to Class I (high solubility, high permeability) and Class III (high solubility, low permeability) under in vitro dissolution criteria. Adopted by FDA, EMA, PMDA, ANVISA (RDC 742/2022 alignment).
DIVERGE
FDA Class III historically slower to grant biowaivers; M9 harmonisation closed most of this gap post-2020. WHO TRS 992 retains additional Class II biowaiver scope for weak acids. CDSCO follows ICH M9 default.
FDA2017 BCS GL
M9-aligned 2020+
2017 FDA BCS biowaiver guidance. Updated to M9-aligned scope post-2020. Class I + III biowaivers granted under specific dissolution criteria.
Acceptance criteriaBCS I + III · M9-aligned
EMA2010 GL Annex III
Class I + III earlier adopter
EMA accepted Class III biowaivers earlier than FDA (2010 BE Guideline Annex III). M9 harmonisation aligned EMA scope from 2020.
Acceptance criteriaBCS I + III
ICH M9Step 4 · Nov 2019
★ Harmonisation milestone
ICH M9 BCS-Based Biowaivers. Harmonised criteria across ICH regions. Pivotal convergence event. Excipient effect assessment newly mandatory.
Acceptance criteriaBCS I + III · harmonised
ANVISARDC 742/2022
M9-aligned 2023+
RDC 742/2022 brought ANVISA biowaiver scope to ICH M9 alignment from 1 March 2023. Replaced RDC 37/2011.
Acceptance criteriaBCS I + III
WHOTRS 992 §10
Class II weak-acid scope
Annex 7 §10. Retains some Class II biowaiver scope for weak acids (LMIC public-health context). Broader than ICH M9.
Acceptance criteriaBCS I + III + some II
CDSCO2018 revision
M9-aligned
CDSCO 2018 BE Guideline revision adopted ICH M9 scope.
Acceptance criteriaBCS I + III
/ 4.5 HV drug scaling · RSABE.★FDA Cmax+AUC scaling vs EMA Cmax-only ABEL · the persistent divergence. +
At a glance · convergence vs divergence
CONVERGE
All converge on the principle: highly variable drugs (within-subject CV > 30%) need a scaled approach. Reference within-subject standard deviation (SwR) is the universal scaling parameter.
DIVERGE
FDA RSABE: AUC and Cmax both scalable for HV drugs (per product-specific guidance). EMA ABEL: Cmax-only widening (up to 0.6984–1.4319 if SwR > 0.294). AUC always strict 80–125%. ANVISA: EMA-aligned post-RDC 742. ICH M13A: defers HV scaling to M13B (in development). The single most-persistent post-2010 divergence.
FDA2010 + PSGs
★ AUC + Cmax scalable
RSABE per product-specific guidance. Both AUC and Cmax can be widened beyond 80–125% based on SwR. Different from EMA approach.
Acceptance criteriaAUC + Cmax scalable
EMA2010 GL §4.1.10
★ Cmax-only ABEL
ABEL: Cmax-only widening up to 0.6984–1.4319 when SwR > 0.294. AUC remains strict 80–125%. Justification clinically.
Acceptance criteriaCmax only · AUC strict
ICH M13A2024
Defers to M13B
M13A explicitly leaves HV scaling to M13B (in development). M13A §1.2 scope excludes RSABE.
Acceptance criteriaPer region until M13B
ANVISARDC 742/2022
EMA-aligned
RDC 742 follows EMA Cmax-only ABEL. Closing the historical FDA-aligned approach pre-2022.
Acceptance criteriaEMA-aligned Cmax ABEL
WHOTRS 992 §9.4
Either approach accepted
PQ programme accepts both FDA RSABE and EMA ABEL approaches with justification.
Acceptance criteriaFDA or EMA approach
CDSCO2018
EMA-aligned default
CDSCO follows EMA-style ABEL approach by default. Replicate full-period preferred.
Acceptance criteriaEMA-aligned
Inspector's eye
For HV drugs in cross-region filings (US + EU): two parallel statistical analyses required · one with FDA RSABE (both parameters scaled), one with EMA ABEL (Cmax-only). Same raw data, two SAP versions. Sponsors who attempt single-SAP filings see deficiency letters from one or both regions.
/ 4.6 Multiple-dose · steady-state.When single-dose BE is insufficient · non-linear / time-dependent / MR. +
At a glance · convergence vs divergence
CONVERGE
All require multiple-dose / steady-state BE for: non-linear PK, time-dependent kinetics, controlled-release formulations, drugs with active metabolites contributing to effect, drugs with long half-life unsuited to single-dose closure.
DIVERGE
FDA: per product-specific guidance, often steady-state required for MR. EMA: steady-state Cmin,ss + AUCτ for MR (2010 GL Annex II). ANVISA: aligned with EMA. WHO: per CPL product entry.
FDAPSG-driven
Per product-specific
Per product-specific guidance. Single-dose default; steady-state required where PK justifies.
Acceptance criteriaPSG-defined
EMA2010 Annex II
Cmin,ss + AUCτ for MR
Steady-state required for MR. Cmin,ss + AUCτ + fluctuation index. Multiple-dose protocol per Annex II.
Acceptance criteriaCmin,ss + AUCτ
ICH M13A2024
Single-dose IR scope
M13A §1.2 limits scope to single-dose IR oral BE. Steady-state and MR designs deferred to M13B and M13C.
Acceptance criteriaOut of M13A scope
ANVISARDC 742/2022
EMA-aligned
EMA-aligned multiple-dose framework. Brazilian healthy-volunteer registry adds operational constraint.
Acceptance criteriaEMA-aligned
WHOTRS 992 Annex 7
CPL-driven
Per WHO CPL product entry. Many anti-TB and anti-HIV drugs require steady-state.
Acceptance criteriaCPL-defined
CDSCO2018
EMA-aligned
Multiple-dose framework follows EMA pattern by default.
Acceptance criteriaEMA-aligned
/ 4.7 Fed / fasted conditions.Fasted default · fed mandatory for MR / food-effect / BCS II. +
At a glance · convergence vs divergence
CONVERGE
All require: fasted study as default; fed study mandatory for modified-release, food-effect labelled, BCS Class II products. Same 80–125% TOST applied separately to fed and fasted arms.
DIVERGE
FDA high-fat high-calorie: 800–1000 kcal, 50% fat (per FDA 2002 Food-Effect guidance). EMA standardised meal: high-fat (~50% caloric), high-calorie (~800–1000 kcal), aligned with FDA but described in 2010 BE GL Annex IV. ANVISA: EMA meal protocol. CDSCO: India-specific meal composition guidance under review.
FDA2002 Food-Effect
High-fat high-calorie
FDA 2002 Food-Effect guidance: 800–1000 kcal, ~50% fat caloric content. Specific test meal composition published.
Acceptance criteria800–1000 kcal · 50% fat
EMA2010 GL Annex IV
FDA-aligned meal
Annex IV: high-fat high-calorie meal aligned with FDA composition. Fed study mandatory for MR and food-effect labelled.
Acceptance criteriaFDA-aligned meal
ICH M13A2024 §5.4
Harmonised meal description
M13A §5.4: harmonised meal composition (high-fat, high-calorie). Fed study triggers harmonised across ICH regions.
Acceptance criteriaHarmonised meal · M13A
ANVISARDC 742/2022
EMA / M13A meal
RDC 742 adopted EMA meal composition. M13A alignment expected at next revision.
Acceptance criteriaEMA meal
WHOTRS 992
CPL-driven
Per CPL product entry. Generally FDA/EMA-aligned meal composition.
Acceptance criteriaCPL-defined
CDSCO2018 + 2024
India meal under review
India-specific meal composition pilot. Generally aligned to FDA/EMA composition for cross-region filings.
Acceptance criteriaFDA-aligned default
/ 4.8 NTI drugs · narrow therapeutic index.90.00–111.11% bounds · full replicate · warfarin-anchored. +
At a glance · convergence vs divergence
CONVERGE
All converge on tighter bounds for NTI drugs (90.00–111.11% as the standard tightening). Full-replicate design preferred. Within-subject variability of reference (SwR) becomes a regulatory parameter.
DIVERGE
FDA NTI list: warfarin-anchored, drug-by-drug PSG. EMA NTI list: tighter Cmax bound published in 2010 GL §4.1.9. ANVISA: NTI list updated in RDC 742/2022. WHO: Annex 7 §9.4.4 lists NTI drugs distinctly. CDSCO: NTI list under continuous revision.
FDA2010 + PSG
★ Warfarin-anchored
90.00–111.11% bounds, full-replicate design. Drug-by-drug PSG. Warfarin (2009) the canonical NTI BE precedent.
Acceptance criteria90.00–111.11% · replicate
EMA2010 §4.1.9
List-based tightening
EMA NTI list. Cmax tightened to 90.00–111.11%. AUC also tightened. Full-replicate preferred for SwR characterisation.
Acceptance criteria90.00–111.11% Cmax
ICH M13A2024
Defers to M13B/M13C
M13A scope excludes NTI tightening. Future M13B may harmonise the NTI list and bounds.
Acceptance criteriaOut of M13A scope
ANVISARDC 742/2022
Updated NTI list
RDC 742 updated NTI list, EMA-aligned bounds. Reference: RDC 742/2022 Anexo II.
Acceptance criteriaEMA-aligned bounds
WHOTRS 992 §9.4.4
Distinct NTI handling
Annex 7 §9.4.4 explicitly lists NTI drugs. PQ programme aligns to FDA / EMA bound.
Acceptance criteria90.00–111.11% NTI
CDSCO2018 + revisions
List under revision
CDSCO NTI list under continuous revision. Generally adopts FDA list with India-specific additions.
Acceptance criteriaFDA-aligned bounds
Inspector's eye
NTI status of a drug is determined by the regulator's published list, not by sponsor judgement. Sponsors who file at standard 80–125% bounds for a listed NTI drug receive deficiency letters at first review · this is a recurring root cause of resubmission cycles in cross-region BE programmes.