Be Studies _part 3

BE studies- Part 3 : Statistical Phase, Special Situations, Guidelines Dr. Ammar Raza Clinician, Clinical Affairs Manager, Medical Advisor

Introduction  Performance – – –

Two formulations Two batches of the same formulation? Two tablets within a batch?

 Purpose – – –

will never be identical

of bioequivalence (BE)

Demonstrate that performance is not “significantly” different Same therapeutic effect What constitutes a ‘significant’ difference?

Phases of BE studies 

Clinical Phase – – – – – – –

Screening Selection ICF & Recruitment Dosing Sampling Monitoring AEs



Bioanalytical Phase – – – – –

Storage of Samples Method Devpt Method Validation Analysis of samples QC checks

Focus on 

Statistical Phase – – – – –

Data analysis Anova (SAS) PK (WinNonlin)AUC, Cmax etc. T/R ratio 90% CI

BE Studies: Scientific Basis

Two different formulations of a drug resulting in SIMILAR systemic concentration-time profiles will always achieve similar concentration time profiles at the site of efficacy or toxicity

Metrics for BE studies 

Concentration vs. time profiles –

Area under the curve (AUC)

Observable exposure AUC-t (zero to last detectable concn) Overall  Complete exposure AUCinf extent – Maximal concentration (Cmax) Both rate & extent – Time to Cmax (Tmax) 



Statistical measures of BE metrics – –



Mean Variance

Stat BE: comparing the means of two products

PK analysis: Approaches  BE

or rate & extent of BA or exposure can be proven using – –

Compartmental approach - not preferred in BE Non-compartmental approach  Based

on calculation of AUC –body’s exposure  Favored to prove BE- robustness  Min of 15 samples- to calculate AUC

Non-compartmental approach

Determination of BE 

Modern concept of BE is based on a survey of physicians carried out by Westlake in 70s - a 20% diff in dose b/n 2 formulations - no clinical significance for most drugs –



BE limits were set at 80 - 120%.

Pl concn. dependent measures - Cmax or AUC are not normally distributed – –

Are log normal, BE limits became 80 - 125% (or ± 0.225 on natural log scale)

Statistical Determination of BE 

Past method –

tested null hypothesis- no difference between means 



Not adequate in 1980s

Current method – –

proves similarity between two products BE- diff <20% 

–

Avg rate & extent of BA of T within ±20% that of R

Log transformed scale- limits of ratio b/n 0.8 and 1.25

The statistical procedure .. 

‘two one-sided test’ –



introduced by Hauck

Method – defines error α- probability of concluding BE when in reality it is not true – – – – –

Usually fixed to a min- 5% BE concluded if 90% CI of the ratio is within 80-125 Power of study (regulatory)=80% 20% probability of not demonstrating BE even if they are truly BE No of sub: based on variability of metric that study must pass on 

Cmax: most variable metrics

Sample size

Analysis of Variance 

ANOVA –



Most common technique of analysis and estimation

Lognormal distribution – – – –

Raw data must be log transformed Comparison of means & variances of transformed data Geometric mean (GM) Results reported in original scale

Confidence Intervals (CI)  Inference

from study to wider world  Range of values within which we can have a chosen confidence that the population value will be found  Study findings expressed in scale of original data measurement

Confidence Intervals (cont.)  Width

of CI indication of (im) precision of sample estimates  Width partially dependent on: – –

Sample size Variability of characteristic being measured  Between

subjects  Within subjects  Measurement error  Other error

Confidence Intervals cont. 

Degree of confidence required –





More confidence = wider interval

Width of equivalence limits represents allowable boundary for ratio (or difference) of means b/n products in comparison In other words, width of CI dependent on: –

–

Standard error (SE)  Standard deviation, sample size Degree of confidence required

Statistical Analysis (Two One-sided Tests Procedure)  Statistical

analysis of pharmacokinetic

measures – –

Confidence intervals Two one-sided tests

 AUC –

and Cmax

90% Confidence Intervals (CI) must fit between 80%-125%

Typical BE Assessment Criteria  90%

confidence interval  Ratio of geometric means  Acceptance criteria: 80 – 125%  Log transformed AUCT & Cmax

Statistical Approaches for BE  Average

bioequivalence  Population bioequivalence  Individual bioequivalence Average BE  Conventional method  Compares only population averages  Does not compare products variances  Does not assess subject x formulation interaction

Statistical Approaches for BE  Population –

Include comparisons of means and variances

 Population –

–

BE

Assesses total variability of the measure in the population

 Individual –

and individual BE

BE

Assesses within subject variability Assesses subject x formulation interaction

Statistical effects in model  Sequence

effect  Subject (SEQ) effect  Formulation effect  Period effect  Carryover effect  Residual

Statistical Analysis  Bioequivalence criteria –

Two one-sided tests procedure  Test

(T) is not significantly less than reference  Reference (R) is not significantly less than test  Significant

difference is 20% (α = 0.05 significance

level) – –

T/R = 80/100 = 80% R/T = 80% (all data expressed as T/R so this becomes 100/80 = 125%)

Special Situations  Highly

variable drugs  Endogenous substances  Parent/ metabolite issues  Long half-life drugs

Highly Variable Drugs (HVDs)  Intrasubject –

variability (CV%) ≥30%

Significant first pass metab or to a poor or erratic absorption process

 Sample

size in BE studies is determined -by BA parameter with highest variability –

most cases, Cmax has higher variability than AUC

 May

not pass even when the reference product is tested against itself

Factors Contributing to the Variability 

Related to Formulation – – –

Disintegration Dissolution Permeability

 

NON-related to Formulation Absorption: – –

 

Pancreatic or bile acid secretion Drug metabolism – – –



Rate of GI transit: Stomach to the colon Transport through GI mucosa

induction inhibition Liver blood flow

Excretion –

Renal blood flow

HVDP 

highly variable drug product (HVDP) - formulation of poor pharmaceutical quality - drug itself is not highly variable- big component of within formulation variability (WFV) – –

cannot be detected in traditional 2-treatment, 2-period, 2sequence cross-over design studies Replicate designs- facilitate their detection - within-subject variabilities of test & reference formulations can be estimated separately 

When they are v. different - one of the formulations is a HVDP

Approaches 

Evaluate bioequivalence at steady-state – – –



Assessment of BE on the metabolite – –

  

Variability expected at steady-state is < that after single dose Always true? Can not be applied to all HVD/HVDP When parent undetectable- metabolite is less variable smaller sample size - BE studies for HVD if based on the metabolite

Add-on Individual BE- may help overcome existing problems ABE Average BE with scaling approach & widen

CI

IBE Vs. ABE  Study – –

periods are duplicated 4 vs. 2

Bad: duration & cost x 2 Good: may reduce pool size of volunteers - HVD

 Has

not found unanimous consensus in the scientific community – –

Remains under investigation Subject of discussion in future

Wider CI  Major

regulatory agencies have provisions can accommodate effect of higher variability associated with Cmax on design of BE studies  EMEA -expanded limits (e.g., 75-133%) for Cmax in certain cases -NO safety or efficacy concerns  MCC, SA - allow for expanded limits for Cmax in certain cases

Example 

2 BE studies on formulations of drugs A & B – – –

 

same no. of sub in each study GMR is the same in both Two One-Sided Test - only difference b/n 2 studies is magnitude of CV

Drug A - low within-subject variability (ANOVACV 15%) - 90% CI falls comfortably within BE limits Drug B is highly variable - ANOVA-CV of 35% –

Study on drug B was underpowered- simple remedy repeat study with a greater no. of sub

Progesterone

The Poster Drug for High Variability A

repeat measures study of Prometrium® 2x200 mg caps in 12 healthy PM females yielded: Intrasubject CV for AUC of 61% Intrasubject CV for Cmax of 98%

 Generic

company calculated that a 2 period crossover BE study - require dosing in 300 PM women to achieve adequate statistical power

Endogenous substances  Pose

a major problem  Baseline levels present – – –

Administration of drug can alter levels / feedback mechanism Oral admin- frequently produces only a negligible inc in baseline; wide variability Baseline should be measured throughout the day before dosing

Endogenous substances Issues  100s or 1000s of vols to operate with net post-dose values –





Lesser no of vols- post dose values without baseline subtraction Steady state studies –



Not acceptable from ethical or financial point of view

preferred design when possible

Assay sensitivity issues

Predominating active metabolite Parent/ metabolite issues  Parent more variable  Difficult to get detectable concn. (absent or marginally present)  Rate of abs- adequately evaluated only assaying parent  Measure metabolite – –

Less of subs reqd Easier to prove BE

Allopurinol, flutamide, terfenadine

Long half-life drugs  

Crossover- adequate washout to avoid carryover – study lasts 4-6 m or more Parallel design –

More no. of sub (n=18 in crossover is stat equivalent to ~n=50 in parallel) 



Costly

Approaches: steady state or truncated AUC (stopping at 24 or 48 h) – –

crossover -washout cannot be shortened, duration of study partially reduced Parallel design- markedly reduce duration of study

Topical application Three classes  Administered topically for absorption into sys circulation e.g. patches –



Designed to exert topical activity only – absorption is negligible e.g. ointments, creams –



can use usual BE

PD study or clinical efficacy study

Designed to exert local activity – absorbed to a certain extent only e.g. vaginal prep etc. –

Considered individually

BE vs. Clinical Trial: Differences Clinical Trial  Multicentric  Subjects: mostly patients (except Ph I)  Multiple doses  Costly and time consuming

BE study  Single centre  Subjects: Mostly healthy vol; rarely pts.  Single dose; sometimes multiple dose  Cheaper and require less time

Focus on FDA Guidance 

Two main guidances –

–

–

General : “Bioavailability and Bioequivalence Studies for Orally Administered Drug Products — General Considerations” http://www.fda.gov/cder/guidance/5356fnl.pdf “Food effect bioavailability studies and fed Bioequivalence studies” http://www.fda.gov/cder/guidance/5194fnl.pdf Drug specific guidances  

– –

Levothyroxine Potassium hydrochloride

Biowaiver Retention samples

Hatch-Waxman Amendments to FFD&C Act - 1984  Considered

one of the most successful pieces of legislation ever passed

 Created

the generic drug industry

 Increased  1984

availability of generics

12% prescriptions were generic

 2000

44% prescriptions were generic - yet only revenue for prescription drugs

 Compromise

8% of

legislation to benefit both brand and generic firms

Hatch-Waxman Amendments to FFD&C Act - 1984  Allowed

generic firms to rely on findings of safety and efficacy of innovator drug after expiration of patents and exclusivities (do not have to repeat expensive clinical and preclinical trials)  Allowed patent extensions and exclusivities to innovator firms

Requirements for generic drugs •

Labeling

•

Chemistry/Microbiology

•

Bioequivalence

•

Legal

Labeling •

“Same” as brand name labeling

•

May delete portions of labeling protected by patent or exclusivity

•

May differ in excipients, PK data and how supplied

Chemistry • • • • • • •

Components and composition Manufacturing and controls Batch formulation and records Description of facilities Specs and tests Packaging Stability

Manufacturing Compliance Programs  Purpose

- To assure quality of marketed drug products  Mechanisms - Product Testing – – –

Surveillance Manufacturing/Testing plant inspections Assess firm’s compliance with good manufacturing processes

Guidance for CROs Scope: Guidance to organizations involved in the conduct and analysis of in vivo bioequivalence (BE) studies Note: BE studies should be performed in compliance with: • General regulatory requirements • Good clinical practice (GCP) • Good laboratory practices (GLP)

Guidelines Guideline provides information on: - organization and management - study protocols - clinical phase of a study - bio-analytical phase of a study - pharmacokinetic & statistical analysis - study report

Comparison of guidelines

Importance  Understanding

the generic drug approval process and the issues surrounding BE is of paramount importance to both clinicians and scientists

Welage LS, Kirking DM, Ascione FJ, Gaither CA.J Am Pharm Assoc (Wash). 2001 NovDec;41(6):856-67

Resources

Text book of pharmacokinetics http://pharmacy.creighton.edu/pha443/pdf/pkin08.pd

Summary 

Planning is important –



Study design, sample size, sampling schedule, incl & excl criteria

Conduct –

–

Clinical & ethical: Protocol approval, selection of volunteers, housing, dosing, sampling, AE recording and reporting, ambulatory samples, Bioanalytical 

–

PK & Statistical 



assay method, equipment (HPLC / LC MS/MS), SOPs Software (WinNonlin, SAS etc.)

Reporting –

3 Parts, CRFs, TMF, Chromatograms

Thank You [email protected]

Biowaiver 

Recommended for a solid oral Test product that exhibit rapid (85% in 30 min) and similar in vitro dissolution under specified conditions to an approved Reference product when the following conditions are satisfied: – Products are pharmaceutical equivalent – Drug substance is highly soluble and highly permeable and is not considered have a narrow therapeutic range – Excipients used are not likely to effect drug absorption