Ego Pharmaceuticals Pty Ltd and Minister for Health and Ageing

[2010] AATA 825

Administrative Appeals Tribunal

DECISION AND REASONS FOR DECISION [2010] AATA 825

ADMINISTRATIVE APPEALS TRIBUNAL      )

)          No 2008/4472

GENERAL ADMINISTRATIVE DIVISION        )

Re      Ego Pharmaceuticals Pty Ltd

Applicant

And    Minister for Health and Ageing

Respondent

DECISION

TribunalJustice Downes, President

Dr Schafer, Member

Date26 October 2010

PlaceSydney

DecisionThe matter will be listed for directions as to its further progress and final determination.

...............[sgd].............................

Garry Downes
  President

CATCHWORDS

HEALTH AND AGEING – Therapeutic Goods - registration of pharmaceutical products – generic products – requirement of bioequivalence - Australian and European guidelines  ambiguous – appropriateness of vasoconstrictor assay – validation of assay – assay not conducted in accordance with guidelines - products not approved – further argument from parties

RELEVANT ACT/S:

Therapeutic Goods Act 1989 (Cth)

Therapeutic Goods Amendment (2009 Measures No. 3) Act 2010 (Cth)

CITATIONS

Drake v Minister for Immigration and Ethnic Affairs (1979) 46 FLR 409

Drake v Minister for Immigration (1979) 24 ALR 577

Re Drake and Minister for Immigration and Ethnic Affairs (No 2) (1979) 2 ALD 634

McKinnon v Department of Treasury (2006) 228 CLR 423

Shi v Migration Agents Registration Authority (2008) 235 CLR 286

REASONS FOR DECISION

26 October 2010

Justice Downes, President

Dr Schafer, Member

Overview

1.      Pharmaceutical products (whether prescription or non-prescription) can only be marketed if they are registered in the Australian Register of Therapeutic Goods (ARTG).  Registration will be approved when quality, safety and efficacy have been satisfactorily established. In the case of pharmaceuticals containing the same active ingredient as products already registered, the requirements of the system may be less stringent than with new products.  Equivalence must, however, be established. The new product is known as a “generic”.  This is because it is generically the same as the established product.

2.      Ego Pharmaceuticals Pty Ltd seeks approval to market such a generic, for use topically to treat dermatitis and other skin conditions. The generic is to be called Zatamil. The established product is called Elocon or Novasone. The active ingredient in both products is mometasone furoate.

3.      Although the active ingredient in both products is identical, the bases of the ointment, lotion and hydrogel in which they are carried, are not. The question which we must answer is whether the testing which has been undertaken by Ego, particularly bioequivalence testing, sufficiently demonstrates equivalence between the two products to satisfactorily establish the quality, safety and efficacy of Zatamil for the purposes for which it is to be used.

4.      The evidence does not presently satisfy us as to these matters to the requisite degree, although we are inclined to think that the products are both safe and efficacious.  There is no issue as to their quality. We will hear the parties further on how the matter should proceed. 

5.      We have approached the matter differently to the primary decision-maker and the reviewing decision-maker.  We note, however, that the evidence and the analysis of the issues before us has necessarily been much more thorough and rigorous than before the prior decision-makers.  They brought their expert knowledge to bear on the material before them.  We had the advantage of listening to evidence and submissions over five days of hearing.  We had the opportunity to explore the issues with six expert witnesses.  Two of these witnesses came to Australia from the Northern Hemisphere especially to give evidence in the case.

Introduction

6.      Ego has applied for review of a decision of the delegate of the Minister for Health and Ageing to refuse applications made pursuant to the Therapeutic Goods Act1989 (Cth) for the registration of Zatamil (mometasone furoate) ointment, lotion and hydrogel.

7.      Ego submitted three separate applications to the Therapeutic Goods Administration (TGA) for registration of the Zatamil products:

(a)      an application submitted on 7 June 2006 for Zatamil ointment;

(b)      an application submitted on 7 July 2006 for Zatamil lotion; and

(c)       a separate application, also submitted on 7 July 2006, for Zatamil hydrogel.

8.      Applications for registration of a therapeutic good are considered in the first instance by the Secretary of the Department of Health and Ageing.  Persons affected by a decision of the Secretary may seek a review of that decision by the Minister.  In the present case, Dr Ruth Lopert, the delegate of the Minister, on 12 September 2008, affirmed a decision of Dr Peter Bird, the delegate of the Secretary, of 20 June 2008, which refused each of the applications. 

9.      Section 25(1)(d) of the Act relevantly makes the following provision relating to the evaluation process:

25       Evaluation and registration of therapeutic goods

(1)       Where:

(a)an application is made for the registration of therapeutic goods in relation to a person in accordance with section 23;...

the Secretary must evaluate the goods for registration having regard to:

(d)whether the quality, safety and efficacy of the goods for the purposes for which they are to be used have been satisfactorily established.

The section is quoted in its present form.  This reflects an amendment to the Act made in 2010 which substituted “the Secretary must evaluate the goods for registration” for “the goods are to be evaluated for registration” (Therapeutic Goods Amendment (2009 Measures No. 3) Act 2010 (Cth)). The amendment was associated with other amendments relating to immunity from civil actions but does not affect the test to be applied.

10.     The Minister conceded that there is no issue as to the quality of the Zatamil products, stating in opening written submissions (at [2.3]) that the Minister "does not contest that the quality of the Zatamil products is satisfactorily established for the purposes for which they are to be used."    However, Dr Lopert confirmed the initial decision of Dr Bird rejecting Ego’s applications for registration, on the basis that the safety and efficacy of the products had not been satisfactorily established.  It is the issue of safety and efficacy which concerns us. 

11.     In the decision under review, Dr Lopert gave the following reasons:

As a prescription-only medicine it is expected that topically applied products containing mometasone furoate should meet the standards of safety and efficacy applied to other prescription only medicines.  For topical corticosteroid preparations, these standards include the demonstration of bioequivalence in accordance with the relevant guidelines.  The EMEA Note for Guidance on the Investigation of Bioavailability and Bioequivalence (CPMP/EWP/QWP/1401/98) as adopted in Australia by the TGA, states with respect to locally applied products, that whenever systemic exposure resulting from locally acting medicinal products entails a risk of systemic adverse reactions, systemic exposure should be measured.  The issue of systemic absorption of mometasone furoate following topical application of Zatamil has not been measured. 

The concerns regarding the lack of studies to demonstrate systemic absorption were conveyed to Ego by the TGA prior to submission of the applications to register the Zatamil products.  In addition, Ego was advised by the TGA that adequately designed in vitro skin absorption studies comparing the Zatamil products with the appropriate originator products were required to assist in addressing concerns regarding the potential for systemic absorption of mometasone furoate, particularly through damaged skin.  These skin absorption studies have not been provided. 

In conclusion, the applications to register Zatamil ointment, lotion and gel on the ARTG are predicated on the satisfactory establishment of the products as generic versions of Elocon®/Novasone®.  I am of the view that the studies submitted in support of these applications are insufficient to demonstrate that as the bioequivalence of Zatamil (0.1% mometasone furoate as ointment, lotion or gel) with Elocon®/Novasone® has not been adequately demonstrated, the safety and efficacy of the product cannot be considered to have satisfactorily established (sic).  I therefore concur with the original decision-maker, Dr Peter Bird, that the applications to register Zatamil (0.1% mometasone furoate) ointment/lotion/gel (Submission IDs OM-2006-2888-11; OM-2006-3046-11; OM-2006-3047-11) should be rejected. 

12.     Dr Lopert concluded generally that the studies furnished by Ego in support of the applications were “insufficient to demonstrate… safety and efficacy”.  Although her conclusion was general, she specified three particular grounds:

(a)Ego failed to measure systemic absorption from the Zatamil products (the systemic exposure issue);

(b)Ego failed to conduct in vitro skin absorption studies comparing the Zatamil products with the originator products, including studies on damaged skin (the skin absorption issue); and

(c)the studies Ego submitted in support of registration were insufficient to demonstrate the bioequivalence of the Zatamil products with Elocon/Novasone, and therefore the safety and efficacy of the product was not satisfactorily established (the bioequivalence issue).

The dispositive parts of Dr Lopert’s reasons do not expressly identify any issues with respect to the studies relied upon by Ego, but the preceding narrative of the history of the application does refer to concerns relating to the vasoconstrictor assay (VCA) relied upon by Ego.  The usefulness and the reliability of the VCA used by Ego are central to the outcome in this matter.

13.     A VCA is an assay which is based on the propensity of corticosteroids to produce blanching (whitening) or vasoconstriction (narrowing of blood vessels) of the skin.  This property has been shown to correlate with the amount of drug entering the skin and has been used, at least in the United States, as a basis for assessing the bioequivalence of topical steroid formulations.

The Relevant Guidelines

14.     The TGA has issued guidelines which set out its general requirements for the registration of medicines. These include, relevantly:

(a)the Australian regulatory guidelines for over-the-counter medicines of July 2003 (ARGOM); and

(b)the Australian regulatory guidelines for prescription medicines of June 2004 (ARGPM).

15.     Mometasone furoate is a schedule 4 (prescription only) medicine under the current Standard for the Uniform Scheduling of Drugs and Poisons, as decided by the National Drugs and Poisons Scheduling Committee and published by the TGA. However, the Zatamil products were originally considered by the TGA as over-the-counter medicines because, at the time Ego submitted its applications for registration of the Zatamil products, the down-scheduling of mometasone furoate from schedule 4 to schedule 3 (pharmacist-only or over-the-counter medicine) was being considered by the Scheduling Committee. By the time of the decision upon reconsideration by Dr Lopert, however, the Scheduling Committee had decided that mometasone furoate should remain a schedule 4 medicine.  Dr Lopert, in the passage in her decision quoted above, particularly referred to the requirements for “prescription only” medicines.  The precise basis upon which the applications were considered by Dr Lopert is not, however, particularly concerning, because the requirements, specified in the ARGPM and ARGOM, for registration of a generic prescription only medicine and a generic over-the-counter medicine, at least, both require the demonstration of bioequivalence in accordance with the relevant guidelines issued by the European Medicines Agency (EMA).

16.     The ARGOM provides (at p 6A.5), that:

[T]he TGA will accept applications to register ‘generic’ products without clinical data on the basis of data that demonstrate that the ‘generic’ and ‘innovator’ products are bioequivalent, or a justification that bioequivalence data are not required.  The requirements for bioequivalence data and justification for not providing bioequivalence data are set out in the EU guideline, Note for guidance on the investigation of bioavailability and bioequivalence (CPMP/EWP/QWP/1401/98).

The ARGOM then identifies products for which “[i]n general, bioequivalence data are not required” including:

Products for topical use provided the product is intended to act without systemic absorption when applied locally.  See Topical products in this chapter for data requirements for formulation-dependent topical products.

Circumstances in which bioequivalence data are required are then identified.  These include:

The ‘innovator’ product has been approved on the basis of a full dossier and there are no ‘generic’ products on the register that have been approved without bioequivalence data or a justification for not providing bioequivalence data.

17.     The section on Topical Products (at p 6A.8), provides that “[w]here the efficacy of the product is likely to be formulation dependant… the efficacy of the particular formulation… will need to be established”.  The section concludes:

For guidance on data requirements for locally applied/acting products, refer to the EU document Guideline on the clinical requirements for locally applied, locally acting products containing known constituents (CPMP/EWP/239/95 final).

18.     The ARGPM separately address “essentially similar medicines” (at p 51).  These are medicines where the products have “the same qualitative and quantitative composition in terms of active principles/substances”, “the same pharmaceutical form” and are “bioequivalent”.  The ARGPM relevantly deals with applications for essentially similar medicines as follows:

In lieu of safety and efficacy data, an appropriate bioavailability study (or studies) should normally be submitted.  Provided that the essentially similar product has a sufficiently similar plasma concentration/time profile to a leading brand in Australia, the two products may be considered bioequivalent.  Appendix 15 provides inter alia guidance on types of applications for which bioavailability data need not be submitted.  In other cases, the applicant must justify not providing bioavailability data.  A minimum list of points to be addressed in such a justification may be found in Appendix 15.  The comparative impurity profile of the two products will also be relevant to whether or not safety data are required on the essentially similar brand.

If the PI [Product Information] and/or CMI [Consumer Medicine Information] are different in content to those of the already registered brand, safety and/or efficacy data may be needed.

The European Medicines Agency Guidelines

19.     Both the ARGOM and the ARGPM refer to regulatory guidelines issued by the EMA.  These are guidelines which have been adopted by the TGA and form part of the guidance to sponsors in meeting the data requirements for registration.

20.     The parties agreed that the following EMA Guidelines, which have been adopted by the TGA, are relevant to this decision:

(a)Note for Guidance on the Investigation of Bioavailability and Bioequivalence, 26 July 2001 (CPMP/EWP/QWP/1401/98) (Bioequivalence Guidelines);

(b)Clinical Requirements for Locally Applied, Locally Acting Products, Containing Known Constituents, November 1995 (CPMP/239/95) (Clinical Requirements Guidelines);

(c)Clinical Investigation of Corticosteroids Intended for Use on the Skin, February 1987 (3CC26a) (Clinical Investigation Guidelines); and

(d)Note for Guidance on Non-Clinical Local Tolerance Testing of Medicinal Products, March 2001 (CPMP/SWP/2145/00) (Local Tolerance Guidelines).

The Bioequivalence Guidelines were referred to by Dr Lopert in her decision.  They are also prominently referred to in the ARGOM as quoted above. 

21.     The following Question and Answer documents, which provide clarification of the Bioequivalence Guidelines and the Clinical Investigation Guidelines are also relevant:

(a)Questions & Answers on the Bioavailability and Bioequivalence Guideline, 27 July 2006 (EMEA/CHMP/EWP/40326/2006) (Bioequivalence Q&A); and

(b)Questions and Answer (sic) on Guideline Title: Clinical Investigation of Corticosteroids Intended for Use on the Skin 16 November 2006 (CHMP/EWP/21441/2006) (Clinical Investigation Q&A).

22.     The Bioequivalence Guidelines begin by identifying problems experienced from “differences in bioavailability” of active substances.  They state that “bioavailability of an active substance from a pharmaceutical product should be known and reproducible.” They go on (section 1):

This is especially the case if one product containing one certain active substance is to be used instead of its innovator product.  In that case the product should show the same therapeutic effect in the clinical situation.  It is generally cumbersome to assess this by clinical studies.

Comparison of therapeutic performances of two medicinal products containing the same active substance is a critical means of assessing the possibility of alternative use between the innovator and any essentially similar medicinal product.  Assuming that in the same subject an essentially similar plasma concentration time course will result in essentially similar concentrations at the site of action and thus in an essentially similar effect, pharmacokinetic data instead of therapeutic results may be used to establish equivalence: bioequivalence. 

23.     The Guidelines continue by stating that “the objective of this guidance [is] to define, for products with a systemic effect, when bioavailability or bioequivalence studies are necessary and to formulate requirements for their design, conduct and evaluation.”  (s 1) The Guidelines deal briefly with locally applied products as follows:

5.1.8Locally applied products

(a)      Locally acting

For products for local use (after oral, nasal, inhalation, ocular, dermal, rectal, vaginal etc. administration) intended to act without systemic absorption the approach to determine bioequivalence based on systemic measurements is not applicable and pharmacodynamic or comparative clinical studies are in principle required.  The lack of them should be justified (see specific Note for Guidance). 

Whenever systemic exposure resulting from locally applied, locally acting medicinal products entails a risk of systemic adverse reactions, systemic exposure should be measured. 

(b)Systemically acting

For locally applied products with systemic action a bioequivalence study is always required. 

24.     The Guidelines do not appear expressly to identify what equivalence standard applies to particular circumstances.  There are, however, a number of definitions such as “pharmaceutical equivalence”, “bioequivalence”, “therapeutic equivalence” and “essentially similar products”.  Therapeutic equivalence would seem to be the highest and best test.  The text of the definition may confirm this (section 2.6):

A medicinal product is therapeutically equivalent with another product if it contains the same active substance or therapeutic moiety and, clinically, shows the same efficacy and safety as that product, whose efficacy and safety has been established. 

In practice, demonstration of bioequivalence is generally the most appropriate method of substantiating therapeutic equivalence between medicinal products, which are pharmaceutically equivalent or pharmaceutical alternatives, provided they contain excipients generally recognised as not having an influence on safety and efficacy and comply with labelling requirements with respect to excipients.

25.     “Bioequivalence” is defined as follows (s 2.4):

Two medicinal products are bioequivalent if they are pharmaceutically equivalent or pharmaceutical alternatives and if their bioavailabilities after administration in the same molar dose are similar to such degree that their effects, with respect to both efficacy and safety, will be essentially the same.

26.     The Clinical Requirements Guidelines are more specific.  They proceed on the basis that “full toxicological and clinical data would not normally be necessary” for “locally acting products… which are applied locally and are assumed to exert their effect at the site of application”.  This is so notwithstanding that “none of these products can be considered essentially similar”.  What is permitted is “abridged applications” (s 1):

Abridged applications for locally acting products should therefore be regarded as hybrid applications.  Omission of data should be justified in the expert reports.  Also in case of a type II variation, therapeutic equivalence with the original product has to be ensured.  Choice of studies and omission of data have to be justified in the expert reports. 

27.     The Guidelines further specify as follows (s 1):

Abridged applications

As is the case for systemically acting products, it is necessary to show for locally acting products that the product to be approved (either a generic or a reformulated product) is therapeutically equivalent to the product already approved i.e. that both products are “equivalent” with regard to efficacy and safety, so that data generated for the “innovator” product (i.e. a product for which the marketing authorisation has been obtained on the basis of a dossier with full documentation) apply also to the other product. 

In order to demonstrate therapeutic equivalence clinical trials are in principle necessary, especially for locally applied dermatological preparations but other models may be used or developed.  For this purpose, depending on the situation, human pharmacodynamic studies, local availability studies or where appropriate even animal or in vitro studies can be considered, provided that all studies used are adequately validated.  Moreover, safety and local tolerance have to be addressed appropriately. 

…

Other models

If instead of clinical trials another model is chosen to show therapeutic equivalence the relevance of this model must be demonstrated/justified, i.e. the model must be validated: in particularly its relationship with the therapeutic situation must be demonstrated.  It should be noted that at the moment very few models are available and most of them have not (yet) been validated sufficiently.

For locally applied products bioequivalence generally is not a suitable way to show therapeutic equivalence since plasma levels are not relevant for local efficacy, although they may play a role with regard to safety. 

28.     A “summary table” at the end of the document deals with “requirements for the clinical dossier on locally applied/locally acting products” as follows:

II      abridged/hybrid application

       a     different indication

       b     different dosing schedule

       c     different strength, but usual              dosing schedule

relevant clinical studies

relevant clinical studies

if possible pharmacodynamic studies or local availability studies; possibly in vitro studies (e.g. eye drops) or argumentation in case of minor differences.  Otherwise clinical studies.  Any safety issue has to be addressed appropriately.

       d     different formulation (e.g. cream              to ointment; aerosol to power

             for inhalation)

clinical studies to demonstrate efficacy/safety and/or if possible pharmacodynamic studies; if necessary, studies of absorption, penetration and local tolerance

       e    generic products

see II c

29.     The Clinical Investigation Guidelines deal with narrower issues, particularly the use of a VCA (s 3.1):

Corticosteroid induced vasoconstriction in man may provide a preliminary rough but useful guide to topical anti-inflammatory activity.  Vasoconstriction tests derived from McKenzie and Stoughton (Archives of Dermatology, 1962, 86, 608-10), using the new substance incorporated in its base should therefore be compared with the effects of established corticosteroid preparations, not only with those of approximately equal activity, but also with stronger and weaker preparations.

Other methods can also be used or are currently under development.  It is not considered desirable to specify particular methods of investigation as obligatory; regulatory agencies will be prepared to consider new methods provided the clinical relevance of these methods is demonstrated.  The level of activity estimated in the light of these tests will always have to be confirmed by clinical investigation.

30.     In a section dealing with Clinical Investigation the following appears:

Where the product can be regarded as a minor modification of an exiting one, it will often be possible to rely on a comparison with this latter product used as a reference, at least through an adequate vasoconstriction test.

31.     The most specific document is the Clinical Investigation Q&A.  While it is not a guideline per se, and has apparently not been formally adopted by the TGA, it is a document issued by the EMA to explain the Clinical Investigation Guidelines, which have been adopted by the TGA.  On this basis, we think that it is appropriate for us to treat it, particularly in the confusing state of the guidelines, as a further “guideline” to be treated as adopted by the TGA.  It is worth noting that it is much more recent (2006) than the other guidelines and is the only other document, apart from the Clinical Investigation Guidelines, which deals with VCAs. 

32.     It is appropriate to set out a long passage from the Clinical Investigation Q&A:

2.        PROBLEM STATEMENT

Pharmacodynamic models, e.g. the vasoconstriction assay (VCA), have been developed over the last years, and may be appropriate to reduce the need for data from clinical trials.  Clearer statements are desirable about which methods are scientifically accepted and sufficiently validated for demonstration of efficacy and local tolerance.

3.        FREQUENTLY ASKED QUESTIONS

Question 1:

In which cases can clinical data be replaced by a pharmacodynamic model(s), e.g. a vasoconstriction assay?

Answer:

The Note for Guidance on the Clinical requirements for Locally Applied, Locally Acting Products, Containing Known Constituents states that “In order to demonstrate therapeutic equivalence clinical trials are in principle necessary, especially for locally applied dermatological preparations but other models may be used or developed… Moreover, safety and local tolerance have to be addressed appropriately… If instead of clinical trials another model is chosen to show therapeutic equivalence the relevance of the model should be demonstrated/justified, i.e. the model must be validated: in particular its relationship with the therapeutic indication”.

The current wording of the Guideline on Clinical Investigation of Corticosteroids Intended for Use on the Skin states: “… Corticosteroid induced vasoconstriction in man may provide a preliminary rough but useful guide to topical activity… The level of activity will always have to be confirmed by clinical investigation…”.  However, it also states: “…When the product can be regarded as a minor modification of an existing one, it will be often be possible to rely on a comparison with this latter product used as a reference, at least through an adequate vasoconstriction test”.

The vehicle itself may have a great impact on efficacy and may modify the dermal absorption, as it is acknowledged in the Note for Guidance on the Clinical Requirements for Locally Applied Locally Acting Products Containing Known Constituents, which differentiates locally applied products in dermatology from other locally applied products: “In these products a change in formulation or in dosage form may influence the efficacy and/or safety of the product.  This may occur for instance by changing the physicochemical properties of the product or by changing the non-active ingredients and thereby the extent of penetration of the active compound.  Moreover, at least in dermatology, the vehicle itself may influence the disorder”.  Therefore, a pharmacodynamic model, e.g. the vasoconstriction assay (VCA) is sufficient only if the generic medicinal product possesses the same or a similar qualitative and quantitative composition to that of the reference product. 

Differences in excipients have to be considered case by case.  In case of only minor changes, e.g. slight differences in the quantity of the same excipients in generic applications, VCAs can be accepted instead of clinical efficacy studies.  However, qualitative changes in the composition imply the need for clinical efficacy data.  In addition, it is not possible to present an abridged application for a topical corticosteroid preparation based on a VCA between two different steroid concentrations or two different pharmaceutical forms, e.g. cream and ointment, as the potency of the corticosteroid alone (classification system to assess the strength of the corticosteroid) is not able to predict the efficacy of the entire preparation.  Two different formulations may be very different with regard to their efficacy, e.g. a corticosteroid cream vs. a corticosteroid ointment containing the same corticosteroid in the same concentration.  In this case, relevant clinical studies are required. 

Question 2:

How to use vasoconstriction assays for comparisons?

Answer:

A detailed description of how to perform vasoconstriction assays can be found in the FDA’s Guidance for Industry “topical dermatologic corticosteroids, in vivo bioequivalence”, issued June, 1995.

The following is a summary of testing principles:

For both the pre-test and pivotal study relevant inclusion and exclusion criteria have to be adhered to.

Healthy subjects with an adequate vasoconstriction to topical corticosteroids have to be included.

Test product, vehicle, reference product(s), and untreated control should be randomly assigned to application sites on the ventral forearms.

The reference products should include different potency classes (suggested equally potent, more potent and less potent) and have received a Marketing Authorisation based on a full dossier. 

For the pivotal study a minimum of 12 healthy persons have to be included.  The vasoconstriction reaction should be determined at baseline (before drug application), time of drug product removal and several times after drug product removal (e.g. 2, 4, 6, 19, 24 hours).

The time course of response should be followed until return to baseline to ensure that the maximal pharmacodynamic response is observed.  The assay must be optimised to ensure that the products are compared in the linear portion of the blanching curve.  Several application times should be tested in pre-test.  Also the lower limit of sensitivity has to be determined. 

The vasoconstriction reaction should be determined at several time points and AUC data should be generated.  A single time point for estimation of the vasoconstriction reaction is not acceptable. 

33.     Specific reference is made to a guideline issued by the United States Food and Drug Administration (FDA).  The document is titled: “Guidance – Topical Dermatological Corticosteroids : In Vivo Bioequivalence”.  It was issued on 2 June 1995.

34.     The Guidance describes its purpose as providing “recommendations to pharmaceutical sponsors on methods to document in vivo bioequivalence of topical dermatologic corticosteroids… [using] a pharmacodynamic approach, based on an update of the Stoughton-McKenzie vasoconstrictor bioassay, to assess bioequivalence of topical corticosteroids.”  It notes: “As with all bioanalytical methods, this pharmacodynamic bioassay will require careful validation on the part of pharmaceutical sponsors.” 

35.     The method of conducting a VCA is summarised in the Guidance as follows (p 3):

The Guidance suggests conducting two in vivo studies – a pilot dose duration-response study and a pivotal in vivo bioequivalence study comparing test and reference products.  The pilot study characterizes the dose duration-response relationship for the drug in terms of the E_max model (Section III) and is conducted solely with the reference listed drug (RLD).  The dose duration method as recommended in this guidance for documentation of bioequivalence is based on three dose durations: ED_50, D₁ and D₂.  The comparison of test and reference products in the pivotal study is conducted at a dose duration approximately equal to the population ED₅₀ determined in the pilot study.  Sensitivity in the pivotal study is established through dosing of the RLD calibrators at two dose durations, D₁, (the shorter dose duration calibrator) and D₂ (the longer dose duration calibrator).  The guidance recommends that D₁ equal approximately 0.5 times ED_50, and D₂ equal approximately 2 times ED₅₀ determined from the pilot study.  Because each subject becomes a ‘detector’ in the study, only the data of those subjects whose D₂/D₁ ratio of pharmacodynamic responses meets a specified minimum value may be included in the data and statistical analyses supporting in vivo bioequivalence.  The proposed methodology is more fully explained in subsequent sections of the guidance. 

36.     The “Guidance strongly encourages the performance of a pilot study by each study site for each reference listed drug (RLD) under investigation.”  Under the heading “Study Design and Analysis” it specifies a “dose duration-response study based on RLD only, with randomization of dose duration skin sites.”  It specifies that “twelve subjects” are required. 

37.     “Fitting dose duration-response data by averaging across subjects at each dose duration” is described in the Guidance as not acceptable.  The Guidance continues (p 13):

Rather, the data should be fit by using all observations of all individual subjects simultaneously.  The modeling software should provide ED₅₀ and E_max values for the data pooled from 12 subjects.  The following methods are acceptable:

i.Fitting based on the assumption of a nonlinear mixed effect model (population model) using suitable software (Figure AIII.1).  The mixed effect modeling technique accounts for within- and among- subject variability, or

ii.Fitting based on nonlinear least squares regression, pooling individual observations from all subjects (naive pooled data method).

38.     For the pivotal study the Guidance states “Forty to sixty evaluable subjects… who meet the ‘responder’ and ‘detector’ criteria…” are required.

39.     The Guidance makes the following provision relating to data analysis (p 18):

Only the data of ‘detectors’, i.e., individual subjects whose AUEC values (Table AIV.5) at D₁ and D₂ are both negative and that meet the dose duration-response criterion below, should be included in the data analysis .

40.     Finally, for present purposes, the Guidance mandates “Locke’s method” of statistical analysis and says, of the method known as “Fieller’s method”:

Previously used approximate methods, for example calculating a confidence interval for the difference between test and reference product averages, and dividing these limits by an estimate of the reference product average, are not applicable. 

Application of the Guidelines

41.     Before we discuss specific aspects of the guidelines, it is important to consider the extent to which the guidelines should be applied and what role they play in assisting sponsors to satisfy the requirements for registration.

42.     The ARGPM provide information to sponsors about the use of guidelines in Australia.  Section 1.7 of the ARGPM provides as follows:

In Australia, as in the European Union, guidelines are used as guidance for applicants. The EU use of guidelines as the basis for guidance for applicants is stated as:

The use of guidelines, which are not legally binding, rather than a formal legal instrument, such as a Directive, has been preferred in order to maintain an element of flexibility and not place legislative restraints on scientific progress. It is recognised that in some cases, as a result of scientific developments, an alternative approach may be appropriate. However, where an applicant chooses not to apply a guideline, that decision must be explained and justified in the Expert Reports submitted by the company in support of the application.

The Australian approach to guidelines is somewhat more flexible in the sense that other reasons, beyond scientific development, may be accepted as a rationale for a sponsor not to follow the guidelines rigidly. In Australia, situations such as the following may be relevant:

·a circumstance unique to the product in question can be demonstrated;

·a general guideline is not relevant in a particular instance;

·the sponsor adopts an acceptable approach which had not previously been considered by the TGA; or

·sufficient alternative studies have been conducted which, whilst not exactly what the guidelines seek, nevertheless satisfy the criteria of quality, safety and efficacy.

A sponsor has the option to argue that a particular guideline, or an aspect of a guideline, is not applicable to a given product. However, where sponsors opt to provide data that do not conform with the guideline, a justification should be provided explaining why the proposed alternative is valid.

43.     One matter we must decide, therefore, is whether the guidelines referred to above are necessary and sufficient for the decision we must make.  We note that our discretion is possibly even wider than that of the initial decision maker and the decision maker upon reconsideration.  As Bowen CJ and Deane J said in Drake v Minister for Immigration and Ethnic Affairs (1979) 46 FLR 409 at 420-421:

It is not desirable to attempt to frame any general statement of the precise part which government policy should ordinarily play in the determinations of the Tribunal. That is a matter for the Tribunal itself to determine in the context of the particular case and in the light of the need for compromise, in the interests of good government, between, on the one hand, the desirability of consistency in the treatment of citizens under the law and, on the other hand, the ideal of justice in the individual case.

44.     Similarly, in Re Drake and Minister for Immigration and Ethnic Affairs (No 2) (1979) 2 ALD 634, Brennan J said at 640-641:

[D]iscretion cannot be so truncated by a policy as to preclude consideration of the merits of specified classes of cases. A fetter of that kind would be objectionable, even though it were adopted by the Minister on his own initiative ...That is not to deny the lawfulness of adopting an appropriate policy which guides but does not control the making of decisions, a policy which is informative of the standards and values which the Minister usually applies. There is a discretion between an unlawful policy which creates a fetter purporting to limit the range of discretion conferred by a statute, and a lawful policy which leaves the range of discretion intact while guiding the exercise of the power.

Evaluation of the Guidelines

45.     It is appropriate that we should comment on the usefulness of the guidelines.  We cannot properly carry out our task without noting that there seems to us to be problems with the guidelines.  They are, in our opinion, ambiguous, confusing and even contradictory.  They are fragmented and illogical.  It is to demonstrate this that we have gone to the trouble, which we would not normally do, of setting them out in such detail.  The problems the guidelines present are so extensive that it is pointless to attempt to identify them. Doubt as to what the guidelines relevantly provided underlay most of the issues in this matter.  We think it is impossible to read the guidelines without being uncertain as to what is required.  It is not clear when bioavailability data, equivalence data or clinical data are required.  It is not even clear what is meant by equivalence and the extent to which it is more than bioequivalence. 

46.     We are aware that the guidelines are not meant to be prescriptive.  They recognise that cases for different treatment are capable of being made out.  The advantages of generality, however, should not be confused with the disadvantages of ambiguity.  Indeed, the more general and less prescriptive that guidelines are intended to be, the more they should be logical, consistent, clear and well structured.  This is because those reading the guidelines are looking for reasoning which can be carried over to a number of cases and not for a narrow rule, whether logical or not, which must be followed.

47.     We have great sympathy with the predicament of pharmaceutical companies which need to make extremely costly decisions about the trials and tests they will undertake as part of the process of seeking marketing approval.  This task is difficult enough without the difficulties being compounded by uncertainty as to what is required. 

48.     It seems to us that the guidelines are due for an overhaul.  Although it may be desirable to have different guidelines addressing different issues and sets of guidelines moving from the general to the particular, the guidelines should be written as parts of an integrated whole which can be read together.  It is not, of course, appropriate for us to say what the guidelines should contain, but we think they should be written so that they address topics in some appropriate order of reduction or specificity so that, for example, they might treat the following topics:

·Prescription or over-the-counter

·New drug or generic

·Ingestion or topical application

·Systemic or local operation

·Bioavailability data or not

·Clinical trials or not

·Permissibility of abridged application

·Permissibility of pharmacokinetic studies.

49.     Importantly, similar topics should not be treated differently in one part of the guidelines to the way they are treated elsewhere. More specific provisions should read comfortably with general provisions even though they are in different documents.

50.     We think that a particular problem arises in the present case because three sets of guidelines are relevant from three different parts of the globe.  The EMA guidelines are not so comprehensive or clear that they should continue to be adopted merely to achieve international consistency.  A 2006 Question and Answer document to explain a much older guideline, apparently updating it in the process, is not satisfactory.  The FDA Guidance is more comprehensive and clear, even if also quite old.  It may be capable of incorporation, but its status should be made clear. We do, however, think that the problems created by the EMA guidelines are as significant as the problems created by the Australian Guidelines. We think that consideration should be given to their both being replaced by a cohesive set of Australian guidelines.  

Expert Opinion

51.     During the hearing, we heard evidence from the following clinical dermatologists:

(a)Prof Ross Barnetson, who was commissioned by Ego, is Professor Emeritus of Dermatology at the University of Sydney.  In his expert report dated 20 December 2009 (Exhibit A), Prof Barnetson stated that he has over forty years experience in clinical dermatology in the United Kingdom and Australia, is a world authority on atopic dermatitis and has examined and treated thousands of individuals suffering from atopic dermatitis and other dermatoses and has considerable experience with topical corticosteroids (at [9]).

(b)Associate Prof Peter Foley, who was commissioned by the Minister, currently holds positions as Associate Professor, Department of Medicine (Dermatology) at the University of Melbourne, Visiting Dermatologist at St Vincent’s Hospital Melbourne and at Peter MacCallum Cancer Centre.  Prof Foley is concerned with clinical research involving dermatological products and has his own private dermatology practice.  According to his curriculum vitae (Exhibit 1), Prof Foley has been a clinical dermatologist for more than 17 years.

52.     Evidence was also given by the following pharmaceutical experts:

(a)Dr James Rowe, who was commissioned by Ego, currently holds positions of Director of NxGen Pharmaceuticals Pty Ltd, which is involved in the development and testing of a wide range of pharmaceutical dosage forms, of Director of CoPharm Pty Ltd, which provides consultancy services in the pharmaceutical and associated industries and of Visiting Lecturer in Pharmaceutics at the University of Sydney. Dr Rowe has been involved in the formulation of pharmaceuticals for more than forty years;

(b)Dr Lynn Pershing, who was commissioned by Ego, currently holds the position of Independent Consultant in the field of Pharmacology and is a retired Research Professor in the Department of Dermatology, University of Utah School of Medicine. Dr Pershing has worked in the field of pharmacology and toxicology for 25 years and, according to her expert report dated 13 January 2010 (Exhibit D), has had extensive experience in and is qualified in the methodology advocated in the FDA Guidance;

(c)Prof David Le Couteur, who was commissioned by the Minister, currently holds the positions of Professor of Geriatric Medicine at the University of Sydney, Director of the Centre for Education and Research on Ageing. Director of the Biogerentology Laboratory of the ANZAC Research Institute and Senior Staff Specialist at the Concord Repatriation General Hospital, Sydney. According to his curriculum vitae (Exhibit 3), Prof Le Couteur also serves on the Pharmaceutical Benefits Advisory Committee and has qualifications as a clinical pharmacologist;

(d)Professor Richard Guy, who was commissioned by the Minister, currently holds the positions of Professor of Pharmaceutical Sciences at the University of Bath and of Adjunct Professor at the University of California Department of Biopharmaceutical Sciences. According to his curriculum vitae (Exhibit 5), Professor Guy has worked in the field of pharmacology for more than thirty years and has particular expertise in transdermal technology, the understanding of the skin and the in vivo evaluation of topical bioavailability and percutaneous diffusion.

The Suitability of the Vasoconstriction Assay (VCA) for Determining Bioquivalence of the Zatamil Products and the Elocon Products

53.     In the state of uncertainty which the guidelines leave us, we must decide what use we should give to the VCA.

54.     The starting point is the provisions of the Australian guidelines.  In the way we have described above, they incorporate the EMA guidelines, which, in turn, include the Clinical Investigation Guidelines and, we consider, the Clinical Investigation Q&A.  These documents clearly show a significant acceptance of the VCA in bioequivalence studies for corticosteroids for use on the skin and, moreover, an acceptance of the FDA Guidance as an acceptable standard for conducting a VCA. 

55.     In addition, the literature refers to the VCA as:

(a)“a reliable, qualitative, comparative indicator of topical corticosteroid availability and potency”, in E.W. Smith, E. Meyer, J.M. Haigh, H.I. Maibach “The human skin blanching assay for comparing topical corticosteroid availability” 2 Journal of Dermatological Treatment  (1991) 69-72 at 69;

(b)“a highly reliable method to determine bioequivalence of generic formulations”, in R.B. Stoughton “The vasoconstrictor assay in bioequivalence testing: practical concerns and recent developments” S1 International Journal of Dermatology (October 1992) 26-28 at 26;

(c)“the most widely used bioassay to measure the potency of topical corticosteroids”, in R. Cornell “Clinical trials of topical corticosteroids in psoriasis: correlations with the vasoconstrictor assay” S1 International Journal of Dermatology (October 1992) 38-40 at 38; and

(d)“the method of choice to assess dermatologic corticosteroid products bioequivalence if dose-response is validated”, in G.J.P. Singh, W.P. Adams, L.J. Lesko, V.P. Shah, J.A. Molzon, R.L. Williams, L.K. Pershing “Development of in vivo bioequivalence methodology for dermatologic corticosteroids based on pharmacodynamic modelling” 66(4) Clinical Pharmacology & Therapeutics (October 1999) 346-357 at 346.

56.     In points of agreement prepared by the Pharmaceutical Experts (Exhibit G), it was agreed that the VCA is "accepted as an appropriate pharmacodynamic assay to determine bioequivalence of topical corticosteroid efficacy".

57.     There was considerable debate, however, as to the adequacy of the VCA to determine the bioequivalence of products that had similar, but not the same, excipient formulations.

58.     A comparison of the formulations of the Zatamil products and the Elocon products is provided in Tables 1 to 3.  The Zatamil formulation has been compared with the corresponding Elocon formulation that was used as a reference in the VCA.

Table 1:  Zatamil Ointment compared with Elocon Ointment

Ingredients	Zatamil Ointment %ww	Elocon Ointment %w/w	Purpose
Internal Phase
Hexylene glycol	12.0	12.0	Solvent
Citric acid - anhydrous	0.02	---	Alternative pH adjusters
Phosphoric acid	---	QS
Mometasone furoate	0.1	0.1	Active
Water – purified	3.0	3.0	Penetration enhancer
External Phase
Soft white paraffin	48.88	ü*	Alternative ointment bases consisting of paraffins, fats and waxes
Light liquid paraffin	25.0	---
Cetostearyl alcohol	5.0	---
Propylene glycol monostearate	---	ü
Polyethylene	5.0	---

* Qualitative information only; quantitative information not available

Table 2:  Zatamil Lotion 0.1% compared with Elocon Lotion 0.1%

Ingredients	Zatamil Lotion %w/w	Elocon Lotion %w/w	Purpose
Methylated spirit – industrial	40.0	---	Alternative alcoholic solvent systems for the mometasone furoate
Isopropyl alcohol	---	40.0
Propylene glycol	30.0	30.0	Penetration enhancer
Hypromellose	0.15	---	Alternative cellulose based thickeners
Hydroxypropyl cellulose	---	0.15
Citric acid – anhydrous	0.04	---	Alternative pH adjusters
Sodium phosphate monobasic with phosphoric acid	---	QS
Mometasone furoate	0.1	0.1	Active
Water – purified	29.71	QS to 100	Bulk vehicle

Table 3:  Zatamil Hydrogel 0.1% compared with Elocon Lotion 0.1%

Ingredients	Zatamil Hydrogel %w/w	Elocon Lotion %w/w	Purpose
Isopropyl alcohol	---	40.0	Solvent for active mometasone furoate
Propylene glycol	---	30.0	Penetration enhancer
Hexylene glycol	60.0	---	Solvent for active / penetration enhancer
Hypromellose	1.20	---	Alternative cellulose based thickeners
Hydroxypropyl cellulose	---	0.15
Mometasone furoate	0.1	0.1	Active
Citric acid – anhydrous	0.03	---	Alternative pH adjusters
Sodium phosphate monobasic with phosphoric acid	---	QS

59.     Dr Rowe gave evidence that the VCA is able to detect differences that excipients have on the bioequivalence of two formulations with the same active ingredient in the same concentration.  He stated, at Transcript 419, [26]–[42]:

Insofar as the fact that I believe that the VCA is very, very sensitive, as we have seen in the table, it can discriminate between various delivery systems containing the same active moiety.  For example, we saw mometasone being classed in two different categories [or] schools.  And this was tested for by the VCR (sic).  And, therefore, to me it is quite a very sensitive test to look at the rate and extent of absorption, i.e. bioavailability or bioequivalence – sorry, bioequivalence.  I agree with the questions that your Honours have put, and I agree with Dr Guy's answers, in the fact that various excipients can cause differences in the rate and extent of absorption.

For example, you could put a permeation enhancer and that would make it go through even quicker.  That is not the question…  If there was a difference in the excipients, if we, for example, put a penetration enhancer in there, this would manifest in the VCR (sic) test, provided it was done on the linear part of the curve, which I believe has been shown.

60.     Professor Guy conceded that the VCA tests the efficacy of the drug as long as there is correlation between the extent of the vasoconstriction response and the clinical effect.  He stated, at Transcript 400, [35] – Transcript 402, [5]:

Mr Kirk:  …..You would accept [the VCA] tests bioavailability of the drug?

Prof Guy: It certainly gives you an indication that the drug has become available in a part of the skin where it has its effect. Yes.

Mr Kirk:  It, thus, tests efficacy of the drug?

Prof Guy:  As long as you have got a correlation between the extent of the vasoconstriction response – which is not the efficacy of the drug, it is some other side effect, if you like, of the drug – and the clinical effect.  Yes.

Mr Kirk:  Yes.  And that correlation has been established as accepted in your first agreed point?

Prof Guy:  I can't dispute that either.

Mr Kirk:  Yes.  So you accept it is a test – taking account of our discussion of the last 60 seconds, you accept it is a test of efficacy?

Prof Guy:  I accept that it is accepted as a test for efficacy, yes.

Mr Kirk:  Yes.  And you referred to potency a minute ago.  You would accept that efficacy here equates to potency of the topical corticosteroid?

Prof Guy:  Certainly in a number of studies the vasoconstriction response has been correlated with the potency of the steroid formulation and the steroids tested, yes.

…

Mr Kirk:…Implicit in the answers you provided then is that the VCA is generally accepted as being sensitive enough to pick up differences in potency for topical corticosteroids?

Prof Guy:  Yes.

Mr Kirk:  And that it is sensitive enough to pick up differences in potency, even for the same drug but in different vehicles, such as Elocon in ointment or Elocon in cream?

Prof Guy:  It is certainly – I think it is pretty good at picking out the difference between, for example, a very potent steroid and one that is not considered to be very potent, a mildly potent steroid.  I think it does – there are very few, if any, examples, of where that is not the case.  When you're comparing two very similar formulations, the vasoconstriction assay is an accepted method for comparing them, no question about that.  Many regulatory authorities use the technique.  But not surprisingly, I think, as you make the things closer – the formulations more and more similar, the sensitivity you have to pick out the difference becomes less, as with any assay.

61.     It is significant that both Dr Rowe and Professor Guy stated that the VCA is sufficiently sensitive to detect any differences that the excipient formulations may have on the potency of mometasone furoate.  Relying, in part, on that evidence, we are satisfied that the VCA was an appropriate test for Ego to use to determine the bioequivalence of the Zatamil products with the Elocon products.

62.     We are cognisant that the Clinical Investigation Q&A provides that:

(a)a pharmacodynamic model such as the VCA is sufficient only if the generic medicinal product possesses the same or a similar qualitative and quantitative composition to that of the reference product; and

(b)differences in excipients have to be considered case by case.

63.     However, given that the evidence from Dr Rowe and Professor Guy was that the VCA is sufficiently sensitive to detect any effect that the excipient formulations may have on the potency of mometasone furoate, it seems to us that it is not necessary to separately consider whether Zatamil "possesses the same or a similar qualitative and quantitative composition" to that of Elocon, or to specifically consider the differences in the excipients "case by case".  In this regard, we consider that the guidelines need not be strictly followed and proceed on the basis of our expert evidence.  We accept that a properly conducted VCA is an appropriate means of determining bioequivalence.

Validation of the VCA

64.     The issue of validation is an important one. The FDA Guidance itself states that "[a]s with all bioanalytical methods, [the VCA] will require careful validation on the part of pharmaceutical sponsors" (at p 1). 

65.     Similarly, the Clinical Requirements Guidelines state in s 2:

If instead of clinical trials another model is chosen to show therapeutic equivalence the relevance of this model must be demonstrated/justified, i.e. the model must be validated: in particular its relationship with the therapeutic situation must be demonstrated. It should be noted that at the moment very few models are available and most of them have not (yet) been validated sufficiently.

66.     The Pharmaceutical Experts agreed that the key parameter of concern with respect to validation of the VCA is the test of linearity, which is established during the pilot study and requires the derivation of the dose duration figure (ED₅₀) from the linear portion of the area under the effect curve.   The ED₅₀ is the dose required to achieve 50% of the maximum effect.

67.     The Pharmaceutical Experts agreed that the VCA is able to establish bioequivalence of the Zatamil products with the Elocon products, provided that there is certainty that the response that is being observed is in the linear portion of the dose response curve and, therefore, there is correlation between the observed vasoconstriction response and drug delivery.

68.     We spent some time considering whether the VCA had been appropriately validated for the purposes of assessing the differences between the Zatamil products and Elocon products.   We considered, in particular, the evidence of Dr Pershing, who has considerable prior experience conducting and validating the VCA.

69.     The methodology described in the FDA Guidance uses a pilot study to ensure that the bioequivalence assessment is performed within the linear and, therefore, the most discriminating portion of the curvilinear dose duration response curve, which maximises the ability to detect small differences between the test and reference products.

70.     Dr Pershing gave evidence that, in her view, the bioequivalence studies conducted by Zenith Corporation were performed appropriately using doses in the linear portion of the dose duration response curve, consistent with the methodology described in the FDA Guidance.  She stated in her report:

(a)for the ointment, at paragraph 61:  "While an ED₅₀ value of 1.72 hours was fitted to Elocon® Ointment, 0.1% by the naïve pooled mean method, an ED₅₀ value of 1.24 hours was used in the Pivotal Study.  Use of a smaller ED₅₀ value (1.24 [hours] versus 1.72 hours) is valid because both ED₅₀ values lie on the linear and most sensitive portion of the Elocon® Ointment, 0.1% dose duration-response curve, and the smaller ED₅₀ is more consistent with the ED₅₀ values fitted by the population-based, mixed-effects model."

(b)for the lotion, at paragraph 83:  "In my opinion, use of the smaller ED₅₀ of 0.36 hours, rather than 0.40 hours originally fitted by Zenith Technology Corporation Limited in the Pivotal Study was appropriate, since it lies on the linear portion of the dose-duration response curve of Elocon® Lotion, 0.1%."

(c)for the hydrogel, at paragraph 103:  "An ED₅₀ value of 0.36 hours, representing the mean of the fitted data from the reference and test products was used in the Pivotal Study.  In my opinion, use of a smaller ED₅₀ value of 0.36 hours than that actually fitted for the reference product (0.40 hours) was appropriate because both values lie on the linear portion of the dose duration-response curve for Elocon® Lotion, 0.1%.

71.     We rely on the evidence of Dr Pershing that the ED₅₀ values derived for bioequivalence testing of Zatamil ointment, lotion and hydrogel were in the linear portion of the dose duration response curve.  We conclude on this basis that use of the VCA was suitable to detect any differences between the Zatamil and Elocon products.

136.   The VCA established that the Zatamil products were bioequivalent to the Elocon products.  We must, nevertheless, consider whether we are prepared to accept the VCA results in circumstances in which the test was not conducted in accordance with the FDA Guidance.

Other Testing

137.   In addition to conducting the VCA tests, Ego furnished the TGA with other material in support of its applications for registration of the Zatamil products.  These included a Vasoconstrictor Study in Healthy Male and Female volunteers of Blank Ointment, Blank Lotion and Blank Hydrogel Bases (in comparison with mometasone furoate 0.1% w/w Ointment by VCA), solubility studies and Repeat Insult Patch Test studies.  These studies undoubtedly provided further support for the case for registration.  However, the parties did not single these matters out for specific attention, either in evidence or in submissions, except to the extent that Ego relied upon them as assisting any case that there was a safety or efficacy issue arising out of the differences in excipients employed in the Elocon and Zatamil products.  The experts agreed, however, that the excipients in both products were similar and well accepted as safe.

Consideration

138.   We are, of course, exercising executive or administrative power, not judicial power.  We are not testing the lawfulness of the decision under review.  We are making the decision afresh.  In the present case we are considering what is the preferable decision, not the correct decision, to use the words of the guiding description for decision-making in the Tribunal, namely making “the correct or preferable decision” (Drake v Minister for Immigration (1979) 24 ALR 577 at 589 and 591; see also Shi v Migration Agents Registration Authority (2008) 235 CLR 286 at 309 [82], 314 [98] and 327 [140]). The decision in the present case is to be “the preferable decision” because our task involves broad discretionary powers. There is no single “correct” decision (see McKinnon v Department of Treasury (2006) 228 CLR 423 at 427 [1], per Gleeson CJ and Kirby J).

139.   Ultimately, we must decide whether the Zatamil products should be registered.  We must make this decision in circumstances in which, provided we make no error of law, decisions both to register or not to register might be equally sustainable.  The breadth of the decision-making process entrusted to us is matched by the breadth of the considerations we may take into account.  We are not simply focussing on questions of law or even whether there was other error in the decision under review.  We are, of course, assisted by all the evidence before us and by the submissions before us.  All of that material has been reviewed by us in considering our decision.  Although we have addressed many of the arguments put to us, the breadth of the field is such that we have not been able to deal with all of them in the way we might have if we had been confined to the negative characteristics of a search for error.  We must undertake a much more open-ended task.  In the end, the question is what do we consider to be the preferable decision and what are our reasons for arriving at that decision. 

140.   We undoubtedly have before us evidence that would be capable of satisfying a decision maker of the safety and efficacy of the Zatamil products.  The evidence is the documentary material we have, particularly the Zenith reports, together with the expert evidence of Dr Pershing and the other experts called by Ego.  Equally, however, we have expert evidence called on behalf of the Minister which would justify refusal of registration.  In making our determination as to what is the preferable decision we need to resolve the expert conflict. 

141.   Conflicts between evidence can often be resolved internally by reference to its content.  That is, however, difficult in this case, notwithstanding the presence of expertise in the Tribunal.  It is, of course, a mistake to consider that expert evidence is necessarily objective.  Honest and well thought out scientific evidence nevertheless contains significant subjective elements. 

142.   It is in this situation of conflicting subjective opinions that we think it is appropriate to return to the guidelines, however inadequate they may be.  If the experts were all agreed, failures to comply with guidelines might not stand in the way of a decision in accordance with the expert opinion.  Where such agreement is absent, however, the guidelines assume greater importance. 

143.   We are conscious of the fact that the experts whose evidence tends against approval themselves rely significantly on failures to comply strictly with the guidelines, particularly the evidence of Prof Le Couteur, but we do not see that as a reason not to place weight on the guidelines ourselves. 

144.   There were two significant failures to comply with the FDA Guidance.  They relate to the number of detectors issue and the ED₅₀ issue.  Both of these involve a clear failure to comply.  The former, in extent, is a significant failure.  The latter, while possibly less significant, is accepted to lead potentially to different data being derived.  The defences to both failures depend on opinions of Dr Pershing which have no rigorous scientific basis. One depends on an anecdotal report of one case of which no details were given.  The other depends upon the notion that alternatively derived ED₅₀ values are acceptable so long as they lie in the linear portion of the dose response curve. 

145.   Where applicants for registration seek to proceed on the basis of an abridged application which contains no bioavailability data and which includes no clinical trials we agree that they need to justify their reliance on the abridged data.  We accept that reliance upon a satisfactory VCA is sufficient.  It must, however, be a satisfactory VCA.  If the supporting data are abridged it is important to be able to be confident with the data supplied. 

146.   In these circumstances we have decided that the failures of the Zenith conducted VCAs to comply with the FDA Guidance raise sufficient doubt as to equivalence to make refusal of registration the preferable decision.  We cannot be satisfied as to safety or efficacy when, in circumstances of uncertainty on that issue at a subjective level, the objective requirements or guidelines have not been met. 

147.   Although the number of detectors issue applies only to the lotion and hydrogel, we have decided that the presence of the ED₅₀ issue  in the circumstances of a number of failures to comply with the FDA Guidance, which raise some general doubt about the Zenith reports, should lead to the same decision for the ointment as the decisions relating to the lotion and hydrogel. 

148.   So far as Zatamil hydrogel is concerned we also note that, because there was no Elocon hydrogel, it was compared with Elocon lotion.  This, itself, raised concerns for us because of the Clinical Investigation Q&A which preclude comparison by VCA “between two… different pharmaceutical forms, e.g. cream and ointment”.  We note, in addition, that the dose response curves for the Elocon lotion and the Zatamil hydrogel are markedly different, even though apparently satisfying the statistical requirements to establish bioequivalence. 

149.   On the other hand, the Guidance satisfies us that there is sufficient identity between the lotion and ointment to permit comparison by VCA, but only insofar as the VCA is conducted, in all respects, in accordance with the Guidance. 

Conclusion

150.   It follows that we cannot now approve the Zatamil products for registration.  We consider, however, that if satisfactory VCAs are conducted in accordance with the VCA Guidance there would then be no reason why they should not all be registered.

151.   We could simply affirm the decision under review, or we could remit the matter with directions as to what is required before registration is approved, or we could stand this matter over to enable further VCAs to be conducted.  Which course we adopt will depend upon the attitudes of the parties.  We proceed on the premise that all steps should be taken for this matter to be finalised at the earliest opportunity.  We encourage the parties to explore the ways this might be achieved.  The most satisfactory way of this occurring will be by agreement between the parties. 

152.   We will list the matter for further directions at a time suitable to the parties. 

I certify that the 152 preceding paragraphs are a true copy of the reasons for the decision herein of Justice Downes, President  and Dr Schafer, Member.

Signed:   .................[sgd]...................................................
  Alison Connor, Associate

Date/s of Hearing:  12-15, 20 April 2010
Date of Decision:  26 October 2010
Solicitor for the Applicant:                  Thompson Eslick Solicitors
Counsel for the Applicant:                 Mr Jeremy Kirk
Solicitor for the Respondent:             DLA Phillips Fox
Counsel for the Respondent:           Ms Margaret Allars