Swissmedic & TFDA Generic Drug Regulator Workshop for WHO PQ

Swissmedic • Schweizerisches Heilmittelinstitut • Hallerstrasse 7 • 3000 Bern 9 • Schweiz • www.swissmedic.ch

Swissmedic & TFDA Generic Drug Regulator

Workshop for WHO PQ

Current Swissmedic Perspectives on

Bioequivalence and Cases Arno Nolting, Senior Pharmacokineticist, Swissmedic

Program

2

1. Introduction to Swissmedic

2. Regulatory framework

3. Worksharing and reliance a. ACSS GMWST

b. Swissmedic Article 13

c. GCP violations by CROs

4. Bioequivalence topics of interest a. BCS-based biowaivers

b. Narrow therapeutic index drugs (NTID)

5. Common deficiencies in bioequivalence studies

6. Conclusions

1. Introduction to Swissmedic

3

Geographic location of Switzerland – in the heart of Europe

4

Switzerland and the world Switzerland and the EU

o Switzerland is not a member state of the EU.

o Swissmedic does not participate in EMA

procedures such as CP, DCP or MRP.

Swissmedic - the agency

5

Source: p. 19 of the Swissmedic Annual Report 2017

Swissmedic – the organziational matrix

6

Review of

generics

applications

o Submission of application

o Validation period

o Review is conducted by peer system with separate clinical and quality reports

o Upon completion of the review, a List of Questions (LoQ) is issued by Swissmedic.

o Following the review of applicant`s responses to LoQ a pre-decision notification is issued.

o The applicant has the opportunity to respond to pre-decision notification.

o Swissmedic decision (approval / rejection / approval with conditions or requirements)

o Swissmedic makes evaluation reports available to applicant only; there are no Swiss public

assessment reports (no «SWISSPAR»).

o Note: This presentation will focus on immediate release oral dosage forms.

Swissmedic Review Stages

7

Origin of bioequivalence (BE) studies submitted to Swissmedic

● Canada (40%)

● India (40%)

● Eastern Europe (15%)

● Other countries (5%)

o Switzerland < 1%

o => The vast majority of all BE studies submitted to Swissmedic use

a foreign-sourced comparator product.

8

2. Regulatory framework

9

Regulatory documents relevant for the Swissmedic review (not exhaustive)

● Swissmedic Guidelines

o Authorisation of human medicine with known active pharmaceutical ingredient

o Authorisation of human medicine under Art. 13

● EMA Guidelines

o Guideline on the investigation of bioequivalence

o Guideline on bioanalytical method validation

o Clinical pharmacology and pharmacokinetics: questions and answers document of the Pharmacokinetic

working party

o Product-specific bioequivalence guidances (52 x final; 6 x in public consultation)

● FDA Guidelines

o Bioanalytical Method Validation

o Waiver of In Vivo Bioavailability and Bioequivalence Studies for Immediate-Release Solid Oral Dosage

Forms Based on a Biopharmaceutics Classification System (new version in Dec. 2017)

o Product-specific bioequivalence guidances (too many to count)

● Additional documents for BCS classification

o WHO Technical Report Series No.937, Annex 8: BCS classification of essential medicines

o International Pharmaceutical Federation (FIP) Monographs

o Not yet finalized: ICH M9: Biopharmaceutics Classification System-based biowaivers (Step 3: released for

public consultation

10

Recommendations for industry

● Swissmedic is in alignment with EMA requirements regarding the

bioequivalence issues.

● Product-specific guidelines by EMA and FDA or Q&A documents are

reflective of the current regulatory thinking and need to be considered

carefully when designing BE studies intended for submission in

Switzerland.

● When in doubt about the appropriate BE study strategy, it is advisable

to seek clarification from a regulatory agency like Swissmedic.

11

3. Worksharing and Reliance

12

● ICH Working Groups

o (M09 – BCS-based biowaivers, M10 – bioanalytical methods validation)

● ACSS (Australia, Canada, Singapore, Switzerland) - consortium

● IPRP – International Pharmaceutical Regulators Program (formerly known as IGDRP – International Generic Drug Regulators Program)

● WHO – PQP Programme, Review Sessions in Copenhagen

● Bill and Melinda Gates Foundation (BMGF)

o EAC – East African Community, Joint assessment sessions

o ECOWAS – Economic Cooperation of West African States, Joint

assessment sessions

Swissmedic multi-lateral international collaborations with respect to generics

13

Objectives for international collaborations

● Networking

● Building trust and confidence through worksharing and subsequent

reliance

● Efficient and circumspect use of the available Swissmedic resources

● Work sharing / reliance for generics can serve as a proof-of-concept

● Involvement in shaping the future regulatory landscape regarding BE

study / generics.

● Striving towards harmonizing regulatory requirements benefits patients,

industry and regulators.

14

ACSS Consortium

15

Working Groups

• New Chemical Entities

• Complementary Herbal Products

• Information Technology/Communications

• Pharmacovigilance

• Compliance and Enforcement (GMP)

• Generic Medicines

ACSS Consortium – Generic Medicines WG

The Generic Medicines Working Group (GMWG) was established in 2012 with a

specific focus on issues relating to generic medicines to:

• Create opportunities and benefits for regulatory programs through:

• Greater alignment of regulatory approaches and technical requirement.

• More efficient use of resources through information and work sharing.

• Establishment of an effective network among trusted, like-minded regulatory

authorities;

• Produce immediate and ongoing results in priority work areas

• Serve as a “proof of concept” for other international regulatory cooperation

initiatives such as the IGDRP/IPRP.

To this end the Generic Medicines Working Group has focused on Quality,

Bioequivalence, and multi-disciplinary projects on issues associated with generic

applications.

16

Generic medicines worksharing trial (GMWST «Jim West»)

• An innovative review model with the coordinated assessment of a application submitted

simultaneously to the ACSS members:

– one agency acts as the “Reference Regulatory Agency” - RRA

– other agencies act as the “Concerned Regulatory Agencies” - CRA

• Potential benefits of the work sharing model:

– an opportunity for a unique global collaboration between regulatory authorities and

the pharmaceutical industry

– a model that could be adopted on an even larger scale

– a reduction in regulatory burden (e.g., the filing of common dossiers)

– concurrent marketing authorisations to multiple markets

• First application has gone through the pilot:

– filed in June 2016, assessments completed in April 2017

– valuable experience has been gained that will inform internal procedures for the use

of foreign assessment reports as well as international collaborative work.

17

GMWST «Jim West»

● Teva Pharmaceuticals participated in the first trial.

● The application was submitted to the participating regulatory agencies of

Australia, Canada, and Switzerland, while Singapore’s HSA served as an

observer.

● Australia served as the RRA with Canada and Switzerland as the CRAs.

● The application concerned an oral, immediate-release tablet and was

supported by two bioequivalence (BE) studies: a fed BE study against the EU

reference product (RP) and a three-arm fasting BE study against the Canadian

and EU RPs.

● Health Canada and Swissmedic based their reviews on the TGA evaluation

reports.

18

Outcome

● The application was approved in Switzerland in late March 2017, and

in both Australia and Canada in early April 2017

o Time to this decision in all countries was ~9 months due to

ambitious trial timelines and deviations from standard national

procedures

● A reduction in regulatory burden was realized for the applicant.

● A strengthening of the collaboration among regulatory authorities,

greater understanding of technical requirements, and promotion of

regulatory convergence was achieved.

● Efforts are now directed towards transitioning the model into “business

as usual”.

19

Art. 13 - Reliance

20

Art. 13 application

● Companies can submit generics applications to Swissmedic referring

to approvals in other jurisdictions with comparable control systems for

medicinal products by submitting the identical documentation to

Swissmedic (plus agencies`evaluation reports).

● For most of these applications (approx. 90%), reference is made to EU

approvals (CP or DCP).

● Swissmedic (CR and QR) will make an initial assessment, based on

the evaluation reports, to determine if an in-depth review by

Swissmedic is justified.

● If there are no reasons to justify an in-depth review by Swissmedic,

Swissmedic will not issue an LoQ but will issue the pre-decision

notification.

21

Art. 13 statistics

22

Source: p. 27 of the Swissmedic Annual Report 2017

Comment: The number of Art. 13 generics applications is steadily increasing with benefits for companies

(reduced time to approval) and for Swissmedic (reduced need for resources); in 2018 approx. 1/3 of all generics

applications were Art. 13; numbers for 2018 are expected to rise further.

GCP inspections - Reliance

23

● Clinical and bioanalytical CROs are checked against an internal list of

CRO with GCP-findings (applied also to Art. 13 applications)

● Request of tabular listings of inspections (including possible findings) for

clinical and bioanalyical CROs (usually for 10 years prior to study

conduct).

● Request of sponsor-generated monitoring reports

● PK results are compared with relevant literature results and internal data

(from other applications); substantial differences may raise concern

(Swissmedic does not recalculate PK or statistical analyses)

● The absence of any AEs (GCP issue: subject safety) and an unusually

high number of drop-outs will raise concern.

Swissmedic measures to assess data integrity in a BE study

24

25

Application-Inspections Overseen by OSI/OSIS (CDER, FY 2007 – FY 2016)

OSI = Office of Scientific Investigation; OSIS = Office of Study Integrity and Surveillance; slide taken from FDA presentation on BIMO

metrics, December 2016, slide 5

Link: https://www.fda.gov/downloads/AboutFDA/CentersOffices/OfficeofMedicalProductsandTobacco/CDER/UCM438250.pdf

Recent GCP violations by CROs

● Cetero Research, Houston, USA, 2011: significant instances of

misconduct and violations of federal regulations, including falsification

of documents and manipulation of samples.

● GVK Biosciences, Hyderabad, India, 2015: GCP violation / data

manipulations regarding ECGs during the conduct of BE studies

● Semler Research Center, Bangolore, India, 2016: Evidence of

concentration data manipulations

● Alkem Laboratories, Inc., Mumbai, India, 2016: Data falsification

● Micro Therapeutics Research, India, 2017: misrepresentation of study

data and deficiencies in documentation and data handling.

26

Semler case

● WHO issued Notice of Concern on 12 April 2016.

FDA issued warning letter on 19 April 2016.

● Major findings: a spreadsheet file was found on the company server

that contained an overview of manipulations of bioanalytical samples

and instructions useful for the purposes of manipulating samples

● Subsequent further investigations by inspectors corroborated the

manipulations described in the spreadsheet.

● FDA inspection lasted from 29 Sep to 9 Oct 2015 involving 4 FDA

inspectors

27

Information from the FDA Letter, dated 19 April 2016

28

Healthy volunteers Cmax Geometric

Mean Ratio (%)

90 %

Confidence interval

Full analysis (as included in the study report)

All volunteers 101.43 89.07 – 108.92

Sub-group analysis (as requested by FDA)

Volunteers

1 – 20

137.69 109.81 – 172.66

Volunteers

21 - 40

128.95 101.41 – 163.96

Volunteers

41 – 60

62.18 51.51 – 75.06

Graphical representation – Individual ratios Cmax

A = Test B = Comparator

29

Reassignment of samples

Volunteer Switched with samples

of volunteer

41 10

42 16

43 23

44 9

45 28

… …

60 17

30

Comment: Such a table was found during the FDA inspection in an xls spreadsheet on the CRO data server.

Graphical representation – Cumulative ratios Cmax

A = Test B = Comparator

31

BE data with no obvious trend

● Graphical representation – Individual ratios Cmax

32

BE data with no obvious trend

● Graphical representation – Cumulative ratios Cmax

33

Recommendations for industry ● Worksharing and reliance is an integral part of the Swissmedic assessment of generics

applications as it allows for the efficient use of the available resources. Making use of the tools

Swissmedic has in place (GMWST, Art. 13) companies can realize substantial advantages.

● Ensuring GCP-compliance and data integrity is a critically important part of the clinical study

conduct and is vital to the approval process.

● The selection of the clinical and bioanalytical CRO should be conducted with great care

(reducing time lines and costs should not go at the expense of data quality).

● Adequate monitoring activities need to be implemented prior to and during study conduct.

● CROs need to have a consistent record of inspections by major regulatory bodies without

major findings.

● Results of the bioequivalence study need to be in agreement with information from publicly

available sources (label information, literature, evaluation reports).

● Swissmedic will utilize additional sources of information (bioequivalence studies from other

applicants with identical designs) for comparative purposes. 34

4. Bioequivalence topics of interest

35

Swissmedic approach to BCS-based biowaivers

● The concept of BCS-based biowaivers was introduced in Switzerland in 2010

● Swissmedic requirements are in line with EMA Guideline on the Investigation of

Bioequivalence, Appendix III – BCS based biowaivers

o Drug solubility is based on the highest therapeutic dose.

o Permeability: human data is considered pivotal, in vitro data is considered

supportive.

o Dissolution: rapid for BCS class 1 and very rapid for BCS class 3

● Swissmedic has, like most agencies, no list of substances and BCS classifications.

● Swissmedic takes into consideration: o FDA guideline on BCS-based biowaivers (issued in Dec. 2017)

o FDA and EMA recommendations on individual products (where it is indicated if as an alternative to

BE studies BCS-based approaches may be possible).

o WHO BCS classifications

o FIP (International Pharmaceutical Federation) monographs

36

BCS classifications according to EMA product specific recommendations

Compound in EMA guideline BCS class 1 BCS class 3

Agomelantine X

Emtricitabine X

Tenofovir disoproxil x

Lenalidomide BCS class I is possible; adequate

solubility data needed

Memantine X

Miglustat If the Applicant generates the solubility data and classifies the drug according to the BCS

criteria as highly soluble, a BCS biowaiver could be applicable.

Oseltamivir BCS class III is possible; adequate solubility data

needed

Paracetamol x

Sitagliptin x

Sunitinib malate x

Telithromycin A BCS biowaiver could be applicable if the applicant generates data according to the BCS

criteria to support its classification as BCS class I or III.

37

Comment: This is not an official EMA list, but a compilation based on the presenter`s review of the EMA product specific recommendations.

BCS- based biowaivers possible according to FDA product specific recommendations

38

Oxycodone hydrochloride

Palonosetron hydrochloride

Pramipexole dihydrochloride

Primaquine phoshpate

Proproanolol hydrochloride

Protriptyline hydrochloride

Ramelteon

Rivastigmine tartrate

Ruxolitinib phosphate

Sotalol hydrochloride

Temozolomide

Tiagabine hydrochloride

Varenicline tartrate

Venlafaxine hydrochloride

Comment: This is not an official FDA list, but a compilation based on the presenter`s review of the FDA product specific recommendations. FDA

until recently had exclusively granted BCS class 1 biowaivers; as of Dec. 2017, FDA will also consider BCS class 3 biowaiver applications.

Brivaracetam

Buspirone

Capecitabine

Cefadroxil; Cefadroxil hemihydrate

Cyclophosphamide

Emtricitabine

Galantamine hydrobromide

Granisetron hydrochloride

Lacosamide

Levetiracetam

Levofloxacin

Levorphanol tartrate

Linezolid

Memantine hydrochloride

Metoprolol tartrate

Milnacipran hydrochloride

ICH M09 BCS based biowaivers

● ICH expert working group (EWG) M09 – BCS-based biowaivers

o Current status: Draft Guideline released for public consultation (ICH

Step 3)

● ICH M09 EWG in existence since 2016.

● Working group meetings in Osaka (2016), Montreal (2017), Geneva

(2017) and Kobe (2018)

● Compromise solutions were reached on a number of topics.

● BCS-based biowaivers for classes 1 and 3 will be considered.

39

Examples for compromise solutions in M09 guideline

● Highest strength vs. highest dose

o Some members supported the highest therapeutic dose for the

assessment of solubility. Other members supported the highest

strength.

● Proposed compromise

• A drug substance is classified as highly soluble if the highest

single therapeutic dose is completely soluble in 250 ml or less of

aqueous media over the pH range of 1.2 – 6.8 at 37 ± 1°C. In

cases where the highest single therapeutic dose does not meet

this criterion but the highest strength of the reference product is

soluble under the aforementioned conditions, additional data

should be submitted to justify the BCS-based biowaiver

approach.”

40

Examples for compromise solutions in M09 guideline

● Acceptability of CaCo2 cell results as pivotal

o Members resolved differences of opinion on the acceptability of the

pivotal nature of CaCo-2 for demonstrating permeability / absorption.

● Proposed compromise

o An annex in the ICH M9 guideline provides literature references,

data justification & model examples to demonstrate viability of

CaCo-2 results as primary justification for permeability.

41

Dose strength biowaivers

● Swissmedic applies the EMA guideline, section 4.1.6 «Strengths to be

investigated».

● Linear PK:

• 1 BE study with the highest strength is preferred.

● Non-linear PK: • less than proportional increase: 2 BE studies (lowest and highest

strengths)

• more than proportional increase: 1 BE study (highest strength)

● Demonstration of linear PK: • difference in dose-adjusted AUC < 25%, when comparing studied

strength and strength for which a biowaiver is sought.

• Swissmedic will also accept adequate data regarding linear PK

contained in the label.

• Company should consider all data available in public domain.

42

Recommendations for industry

● The ICH guideline on BCS-based biowaiver has been released for

commenting.

● The regulatory framework regarding BCS biowaivers is undergoing substantial

changes. ICH-efforts are underway towards harmonization.

● Swissmedic currently applies the same criteria as the EMA (i.e. solubility

based on highest therapeutic dose, human permeability data is considered

pivotal).

● However, Swissmedic will consider the totality of the evidence (in vitro

permeability, human absorption data, literature) to reach a conclusion on the

BCS classifcation.

● If the permeability data is not deemed convincing by Swissmedic, the drug will

be considered BCS class 3 (if solubility is sufficiently high).

43

Narrow therapeutic index drugs (NTID) ● Swissmedic will decide on a case-by-case basis if a drug is NTID in accordance with

EMA Guideline on investigation of bioequivalence, section, 4.1.9 «Narrow therapeutic

index drugs»: «It is not possible to define a set of criteria to categorise drugs as narrow

therapeutic index drugs (NTIDs) and it must be decided case by case if an active

substance is an NTID based on clinical considerations.”

● Swissmedic, like FDA and EMA, does not have a list of NTID.

● Swissmedic will consider the following documents when deciding on the NTID status of

a drug and the applicable acceptance criteria:

o FDA and EMA issue product-specific bioequivalence guidances, which indicate if drug

is considered NTID and how to set acceptance limits.

o The EMA Q&A document of the PK working group also contains information on

NTIDs.

44

Examples of narrow therapeutic index drugs

● EMA o Cyclosporine (Cmax: 90 - 111; AUC: 90 – 111)

o Everolimus (transplant setting; Cmax: 80 – 125; AUC: 90 – 111)

o Sirolimus (Cmax: 80 – 125; AUC: 90 – 111)

o Tacrolimus (Cmax: 80 – 125; AUC: 90 – 111)

● FDA (“reference-scaled ABE”) Canada

(Cmax: 80 – 125; AUC: 90 – 112)

o Carbamazepine

o Cyclosporine Cyclosporine

o Digoxin Digoxin

o Everolimus Flecainide

o Levothyroxine

o Lithium Lithium

o Phenytoin Phenytoin

o Sirolimus Sirolimus

o Tacrolimus Tacrolimus

o Theophylline Theophylline

o Warfarin Warfarin

45

FDA reference-scaled ABE

● BE limits will change as a function of the within-subject variability of

the reference product (reference-scaled average bioequivalence

(“reference-scaled ABE”))

● If reference variability is ≤10%, then BE limits are reference-scaled

and are narrower than 90-111.11%

● If reference variability is > 10%, then BE limits are reference-scaled

and wider than 90-111.11%, but are capped at 80-125% limits

● BE study design needs to be fully-replicative.

46

NTID Euthyrox – Problems despite bioequivalence

● Euthyrox: Levothyroxine

● At the request of the French agency, the innovator Merck made changes to the

formulation to increase stability and to tighten specifications (95 – 105% content during

entire stability).

● Changes in new formulation: mannitol instead of lactose; addition of citric acid.

● 2 studies were conducted (new vs. old formulation; dose proportionality).

● Following introduction of the new formulation in France, there were substantial patient

protests who experienced increases in adverse events indicative of loss of titration

(fatigue, perspiration, tremor, palpitations, insomnia, dizziness and headaches).

● The French agency made the two bioequivalence study reports (including all

appendices) publicly available.

● Swissmedic approval in April 2018

47

Results of Merck BE study (new vs. old formulation)

48

Primary endpoint: Baseline-adjusted T4, following single oral dose

of 600 µg Levothyroxine (given as 3 x 200 µg tablets)

Source: http://ansm.sante.fr/Mediatheque/Publications/Rapports-Syntheses-Medicaments#folder_110079

Results of Merck dose proportionality study

49

Primary PK

parameters for T4

Source: http://ansm.sante.fr/Mediatheque/Publications/Rapports-Syntheses-Medicaments#folder_110079

Recommendations for industry

● Decisions regarding the NTID status of a drug will be made by

Swissmedic on a case-by-case basis.

● Drugs identified by other agencies as NTID will be considered as such

by Swissmedic.

● Swissmedic is open to the FDA reference-scaled ABE and the EMA

approach for setting acceptance criteria.

● Note: so far, no BE study was submitted to Swissmedic using the FDA

approach for calculating acceptance limits.

● Even with demonstrated bioequivalence for NTIDs, problems might

arise.

50

5. Common deficiencies in BE studies

(not exhaustive)

51

● GCP violations of participating CRO (clinical and bioanalytical)

● Incomplete study documentation

(i.e. study report incomplete or additional BE studies were submitted in response to list of questions)

● BE study design not in line with current requirements, e.g. widening of acceptance criteria without replicate design, fed

study without recommended high-fat, high calorie breakfast.

● PK data in BE study is not in agreement with the known PK data (i.e. in public domain, Swissmedic internal data)

● Demonstration of linear pharmacokinetics is not in line with current requirements.

● Parent compound vs. metabolite measurement

● Inappropriate determination of truncated AUCs

● Lack of any reported adverse events (AE); all subjects with AE are discontinued from study, regardless of nature or

timing of AE

● Exclusion of subjects or profiles or time points from PK and statistical analysis

● Bioanalytical reassays due to PK reasons

● Lack of stability data covering the storage time for samples prior to measurement

● Lack of assessment of potential for PK or bioanalytical interference due to concomitant medications

● Lack of incurred sample reanalysis

52

GCP violation: GVK – CRO in Hyderabad, India

● At an inspection by the French regulatory agency in July 2014 serious GCP-violations

were detected at the CRO.

● As a consequence several national authorities (e. g. France and Germany) suspended

the marketing authorization for a number of generics in December 2014.

● In January 2015 EMA issued a statement (excerpt): o “The inspection of GVK that led to the CHMP’s recommendation was carried out by the French

medicines agency (ANSM). The inspection revealed data manipulations of electrocardiograms

(ECGs) during the conduct of some studies of generic medicines. These manipulations appeared

to have taken place over a period of at least five years. Their systematic nature, the extended

period of time during which they took place and the number of members of staff involved cast

doubt on the integrity of the way trials were performed at the site generally and on the reliability

of data generated at that site.“

● In February 2015 EMA published a comprehensive list of all affected generics in the EU

(700 pharmaceutical forms and strenghts).

● In May 2015 EMA issued a statement confirming recommendations made earlier.

53

Incomplete study documentation

● Prior to introduction of Swissmedic Bio Trial Information Form (BTIF)

an incomplete study documentation was the main course of rejection

or delay in decision.

● Initially viewed as burdensome, the BTIF serves as a check list for

sponsors and guide for reviewer

● The Swissmedic BTIF is similar to the forms used by Health Canada

or WHO.

● EMA has similar approach with Appendix IV to section 2.7.1.

● Swissmedic accepts the EMA document as well.

54

Incomplete study package

● Sponsors are under an obligation to submit and to discuss

comprehensively all conducted BE studies (pilot studies, negative BE

studies) at the time of the original submission

● Sponsors occasionally submit additional study reports in response to a

list of questions (e.g. when dose strength biowaivers are in question

due to doubts about linear PK).

● Negative BE studies must not be withheld and need to be discussed in

the submission package (cover letter, clinical sections such as 2.5,

2.7.1 and 2.7.2). Possible reasons for a negative outcome (insufficient

sample size, sub-optimal blood sampling schedule) should be

discussed. Pooling of BE data from various studies is not accepted.

55

Fed studies

● Fed BE studies need to be conducted with an FDA/EMA

recommended breakfast (high-calorie, high-fat) or acceptable

alternatives.

● Fed BE studies with «normo-caloric meal» or «continental» breakfast»

are not acceptable.

● If the drug is to be taken independent of food, only a fasted study is

required (in contrast to FDA, which routinely requires fasted and fed

studies).

● Swissmedic will request a fed study in case the label requires the drug

to be taken with food.

56

Widening of acceptance criteria for Cmax

● Prior to 2010 EMA Guideline on investigation of bioequivalence,

sponsors could argue for widening of acceptance criteria for Cmax,

based on review of literature.

● Swissmedic will only accept a widening of the acceptance criteria for

Cmax in case the BE study is performed in a fully or partially replicate

design as defined insection 4.1.10 Highly variable drugs or drug

products of the EMA Guideline on the investigation of bioequivalence.

● If the BE study is not in line with these requirements, the acceptance

criteria of 80 – 125 will be applied. If these criteria are not met, the

study will be considered negative.

57

PK data in BE study vs. known PK data

58

Cmax (ng/ml) AUC0-t (ng*h/ml)

AUC0-inf (ng*h/ml)

Test Comparator Test Comparator Test Comparator

Application 1 195 (41) 183 (40) 1256 (49) 1323 (48) 1356 (48) 1404 (46)

Application 2 395 (44) 401 (50) 2949 (47) 3004 (46) 3456 (45) 3556 (45)

Literature reference

465 (44) 410 (40) 3029 (42) 2928. (43) 3505 (37) 3418 (40)

Mean (%CV) PK values – illustrative example (not real data) Note: all data derive from putative BE studies with the same design

Comment: PK data in Application 1 is very different from PK data in Application 2 and the

literature. PK data in Application 1 may not be representative. Study to study variability alone

cannot explain these differences.

Demonstration of linear PK ● Difference in dose-normalized AUC should be < 25% when comparing the dose studied

in the BE study with dose for which biowaiver is granted.

● Example:

o BE study is conducted with the highest dose (e.g. 100 mg); biowaiver is sought for

the lower dose of 25 mg

o AUC for 100 mg dose: 400 ng*h/ml

=> acceptable range for 25 mg dose: 75 to 125 ng*h/ml

o Therefore, AUC data for 25 mg dose needs to be available in public domain (source:

original data, not review articles) to cover the intended biowaiver

● Alternatively, Swissmedic accepts reference to adequate information on linear PK

contained in the Swiss label of the originator company.

59

Parent compound vs. metabolite ● In principle, evaluation of bioequivalence should be based upon measured concentrations of

the parent compound. The reason for this is that Cmax of a parent compound is usually more

sensitive to detect differences between formulations in absorption rate than Cmax of a

metabolite.

● Also for inactive prodrugs, demonstration of bioequivalence for parent compound is

recommended.

● The use of metabolite data is justified only if the bioanalytical method for the parent compound

is not sufficiently sensitive.

● Sponsors are therefore encouraged to monitor routinely the most recent developments /

improvements in bioanalytical methodology.

● EMA recommendations:

o Clopidogrel – use parent

o Losartan – use parent

o Mycophenolate mofetil – use metabolite

o Ebastine vs. Carebastine?! Both are possible but must be pre-specified in the protocol.

60

Determination of truncated AUC

● Truncated AUCs (AUC0-72h) may be used as an alternative to AUC0-t for the

comparison of the extent of exposure, which is routinely done for drugs with a

long half-life of elimination.

● However, if a 72-h sample is not available for a given subject, this subject

should be excluded from the pivotal analysis, as AUC is under-estimated.

● Extrapolations of missing 72-h concentrations are not acceptable.

● Swissmedic will request a sensitivity analysis excluding subjects with missing

72-h samples.

● If additional analysis fails to meet acceptance criteria, the BE study is

considered negative.

61

Underestimation of AUC due to a missing 72-h time point

62

Comment: Without the 72-h time point, AUC0-72h will be under-estimated. Therefore, the

subject data needs to be excluded from the PK and statistical analysis.

Lack of adverse events

63

Study A,

SD fasted

Study B, SD fed

Study C,

MD fasted

Application 1

1

4

10

Application 2

59

12

163

Literature data

35

15

115

Illustrative example Note: study designs across applications and literature are identical

Comment: The number of AEs in Application 1 is much lower than in the other

studies. Therefore, there is concern that AEs in the application 1 studies were not

collected in a GCP-compliant fashion. Subjects may feel pressured not to report

AEs for fear of exclusion from the study.

General aspects

● The use of spare subjects (e.g. 30+2) should be avoided.

● Referencing draft guideline documents released for commenting is not

acceptable in support of a BE study design or BE study package.

● Referencing the regulatory requirements at the time of the BE study conduct is

not acceptable.

● Regulatory guideline documents in force at the time of submission are

applicable to ensure equal treatment of all applications.

● Deviations from regulatory requirements, if appropriately justified, can be

accepted.

● Sponsors are encouraged to seek scientific advice in case of doubt or

questions regarding the appropriate BE strategy.

64

Exclusion of subjects

● The exclusion of completed subjects from PK and statistical analysis

should be avoided (only exception: potentially incomplete absorption due

to vomiting or diarrhea if timing of the adverse event suggests causation).

● The exclusion of completed subjects is usually biased by the knowledge

of the impact on the study results.

● Statistical outlier analyses to support the exclusion of subjects will not be

accepted by Swissmedic.

● Swissmedic will request an additional analysis including such subjects.

● If additional analysis fails to meet acceptance criteria, BE study is

considered negative.

65

Other reasons given by applicants for exclusion of completed subjects

o Missing samples (=> «incomplete PK profile»), case-by-case

decision, if deemed acceptable by Swissmedic.

o Tmax coincides with first measurement time point (=> «incomplete

PK profile» due to potentially inadequate characterization of Cmax);

applicant may decide to exclude subject data.

• However, Swissmedic will request sensitivity analysis including such

subjects.

66

Re-assays for PK reasons

● The re-assays of samples for PK reasons is not acceptable.

● Only exceptions: measurable pre-dose concentrations or extremely

low concentrations as a result of incomplete or no absorption.

● A re-assay for PK reasons is biased by the knowledge of the impact on

study results and is to be avoided.

● If performed, Swissmedic will request a sensitivity analysis with the

original values.

● If additional analysis fails to meet acceptance criteria, the BE study is

considered negative.

67

Recommendations for industry

● Avoid these deficiencies as they cause additional questions, may delay the

approval process and may even result in a rejection of the application.

● Submit all studies at the beginning of the review process.

● Data integrity is critically important. Any sign that data integrity is compromised

is a serious signal.

● Do not exclude data. Do not use a statistical outlier analysis. Do not perform

re-assays for PK reasons.

● Individually, most deficiencies are likely to be addressed satisfactorily.

● Collectively, however, deficiencies have the potential to undermine the validity

of a BE study.

68

Conclusions

● Worksharing and reliance are an integral part of the Swissmedic assessment

of generics applications.

● These approaches offer great benefits for the generics applicants and for

Swissmedic and will therefore be actively pursued and promoted by

Swissmedic.

● In general, Swissmedic is aligned with the EMA requirements for

bioequivalence studies but will also consider guideline documents developed

by other agencies.

● Topics of interest such as BCS-based biowaivers or NTIDs are rapidly evolving

and will continue to be in the focus of regulatory attention at the level of

individual agencies, multi-lateral worksharing activities and harmonization

initiatives.

69