The concept of “interchangeability” of medicines was first introduced in 1997 when the government introduced therapeutic groups, which grouped together drugs that were considered to be therapeutically interchangeable on the basis of having similar safety and efficacy profiles, and health outcomes.
Therapeutic groups were used as the basis for the introduction of reference pricing, which is the principle that drugs with similar health outcomes should cost the same. Because all drugs in a therapeutic group are considered to have similar safety and efficacy profiles, they are reimbursed at the level of the cheapest drug in that group. A fundamental question, however, is how this similarity between drugs is determined for the purposes of deciding whether two drugs are, in fact, interchangeable.
Earlier concepts of Interchangeability
Prior to the 2007 PBS reforms, the similarity of medicines for the purposes of reference pricing was determined by the test of equivalence. Equivalence is a well established concept in evidencebased medicine. However, the reforms replaced the concept of equivalence with the requirement that drugs must be interchangeable on “an individual patient basis”.
Because of the important consequences that arise when a drug is placed in a therapeutic group, the pharmaceutical industry has sought on a number of occasions to obtain a clear definition of the term “interchangeability”.
What is interchangeability and how is it determined?
The term “interchangeable” is not defined in the National Health Act 1953 (Act), the legislation that underpins the regulation of pharmaceutical benefits. However, section 101(3BA) of the Act provides that if the PBAC is of the opinion that a drug should be made available as a pharmaceutical benefit, it must, in its recommendation to the Minister, specify whether the drug and another drug, should be treated as interchangeable on an individual patient basis.
The Department of Health’s website defines “interchangeable” as “brands of a pharmaceutical item with a particular strength (and brands of related pharmaceutical items) where evidence of bioequivalence or therapeutic equivalence (refer to Therapeutic Group) on an individual basis (or justification for not needing such data) has been accepted by the TGA”.1
This definition, however, confuses the concept of interchangeability with the concept of substitutability. The concept of interchangeability of medicines is quite different from a finding by the Therapeutic Goods Administration (TGA) that certain brands of a medicine are sufficiently bioequivalent to the innovator drug so as to be treated as identical and substitutable. Such drugs are listed on the PBS and subject to reference pricing by virtue of their bioequivalence and substitutability.2 The 1997 Therapeutic Group reforms extended reference pricing to drugs that are not identical but have very similar safety and efficacy profiles, and health outcomes, such that they can be treated as interchangeable for the purposes of PBS pricing policies.
It is interesting to note that the PBAC has interpreted the term ‘interchangeable on an individual patient basis’ in the following way:
Drugs within the therapeutic group are very alike - that is, they belong to the same therapeutic class and, in the vast majority of patients, would work just as well as one another. That is, in commencing a patient on any one of the drugs in a therapeutic group it would make no difference in health outcomes for the vast majority of patients.
Undoubtedly, the question as to whether two different drugs are interchangeable on an individual patient basis requires some clinical judgment. However, the main issue of concern with respect to interchangeability is not so much whether individual patients can be initiated on either therapy and essentially achieve the same health outcomes, but, rather, whether patients can be switched from one therapy to another without any difference in health outcomes.
The data that are generally submitting in support of applications for PBS listing comprise head-to-head studies involving the drug proposed for PBS listing and an appropriate comparator. The evidence presented in support of cost-minimisation submissions seems to be the principal source for determining whether a medicine is interchangeable on an individual patient basis with another medicine. However, cost-minimisation submissions typically only present data from trials that are designed to establish ‘non-inferiority’, that is, they are designed to test the hypothesis that a drug is statistically no worse clinically than the drug to which it is being compared. Strictly speaking, if such studies demonstrate that the drug in question is no worse in terms of safety and efficacy than its comparator, that would support a conclusion that the two drugs are interchangeable at a population level, not on an individual patient basis. To draw meaningful conclusions about interchangeability on an individual patient basis, studies which investigate health outcomes in patients who are switched between the drug in question and its comparator would be more appropriate.
Proceedings in the Federal Court: AstraZeneca v Minister for Health & Ors
On 12 August 2010, AstraZeneca Pty Ltd (AstraZeneca) commenced proceedings in the Federal Court against the Minister for Health and Ageing and 17 current and past members of the PBAC.
Details of the claim
AstraZeneca seeks declarations that:
- PBAC advice received by the Minister was not, for the purposes of the Act, advice as to whether atorvastatin and rosuvastatin should, or should not, be treated as interchangeable on an individual patient basis
- in the alternative, in providing advice to the Minister, the PBAC erred in law by failing to consider whether it should recommend to the Minister that, pursuant to section 84AG(2) of the Act, rosuvastatin 40mg should not be included in the Statins-HP group
- determinations made by the Minister’s delegate purporting to determine the existence of a Statins-HP therapeutic group are invalid and of no effect.
The statement of claim
AstraZeneca’s statement of claim concerns advice the Minister received in March 2009 that:
- atorvastatin and rosuvastatin should together comprise a new therapeutic group (called the Statins-HP Group)
- atorvastatin and rosuvastatin are interchangeable on an individual patient basis.
AstraZeneca has argued that although the PBAC purported to ask itself whether atorvastatin and rosuvastatin are interchangeable on an individual patient basis, it did not ask itself whether atorvastatin and rosuvastatin are, in the case of each of the particular forms in which the drugs are listed, interchangeable on an individual patient basis.
In particular, AstraZeneca claims that it had made representations to the PBAC following the March 2009 advice that rosuvastatin 40mg should not be treated as interchangeable on an individual patient basis with atorvastatin; and should be excluded from the therapeutic group containing atorvastatin and rosuvastatin as it was not safely interchangeable with any form of atorvastatin.
AstraZeneca claims that following a meeting in July 2009, the PBAC reiterated its advice to the Minister that:
- atorvastatin and rosuvastatin should form a therapeutic groups
- it was of the opinion that atorvastatin and rosuvastatin are interchangeable on an individual patient basis.
AstraZeneca claims that in providing this advice, the PBAC determined that it was not obliged by the Act to consider whether atorvastatin and rosuvastatin are, in the case of each of the particular forms in which the drugs are listed, interchangeable on an individual patient basis.
Analysis of the claim
The statement of claim does not elaborate on the reasons why rosuvastatin 40mg is not safely interchangeable with any available form of atorvastatin. In the original application for PBS listing of rosuvastatin, it was determined that the relative potency of rosuvastatin to atorvastatin was 1:3. Based on this relative potency, therefore, patients who are taking rosuvastatin 40mg would require an equivalent dose of atorvastatin 120mg. However, the currently approved product information for atorvastatin states that the maximum allowable dose of atorvastatin is 80mg, such that a patient who was switched from rosuvastatin 40mg to the equivalent dose of atorvastatin would be administered a dose that is currently unapproved in Australia and for which, importantly, safety and efficacy has not yet been adequately established. This would appear to be the main basis under which AstraZeneca is seeking to argue that rosuvastatin 40mg is “not safely interchangeable with any available form of atorvastatin”.
It is noteworthy that AstraZeneca did not in its statement of claim challenge the interchangeability of other strengths of rosuvastatin with atorvastatin. Based on a relative potency of 1:3, rosuvastatin 5mg would be equivalent to atorvastatin 15mg, but a 15mg strength of atorvastatin is not available in Australia (or, in fact, elsewhere). Thus, a patient taking rosuvastatin 5mg could not be switched to an equivalent dose of atorvastatin. The next nearest doses would either be atorvastatin 10mg (an under-dose) or atorvastatin 20mg (higher than the equivalent dose), which would provide differences in efficacy.
Similarly, a patient taking rosuvastatin 10mg would require atorvastatin 30mg as an equivalent dose. Again, a 30mg strength of atorvastatin is not available and a patient switching from rosuvastatin 10mg would require either three tablets of atorvastatin 10mg or one tablet of atorvastatin 10mg and one tablet of atorvastatin 20mg to be able to obtain an equivalent dose. In the latter case, a patient would be required to purchase two separate packs of atorvastatin (atorvastatin 10mg and atorvastatin 20mg) and, therefore, pay the patient copayment twice. Whilst this is an economic consideration which does not appear to have any bearing on PBAC determinations of interchangeability, the potential impact to patients should not be disregarded, as it may be contrary to the objective of the PBS to provide affordable access to medicines.
A patient who was switching from the 20mg strength of rosuvastatin would require atorvastatin 60mg as an equivalent dose. Again, a 60mg strength of atorvastatin is not available, but equivalent dosing may be achieved by the patient taking three tablets of atorvastatin 20mg. Although this may seem reasonable on its face, it is important to note that many patients taking cholesterol-lowering drugs have co-morbidities and are on poly-pharmacy. Requiring such patients to take three tablets instead of one may not be welcomed and may lead to compliance issues.
Where to now?
It is anticipated that the Federal Court proceedings will look purposively at PBAC determinations of interchangeability. It is concerning, prima facie, that the PBAC has admitted that while it asked itself a question as to whether atorvastatin and rosuvastatin should, or should not, be treated as interchangeable on an individual patient basis, it did not consider that question in the context of each of the particular forms in which the two drugs are listed. The Court’s views on the significance of these admissions will be of particular interest and relevance to both the government and industry. As discussed above, there are significant clinical issues concerning the interchangeability of rosuvastatin 5mg and 40mg with atorvastatin, and also economic and compliance issues concerning the interchangeability of rosuvastatin 10mg and 20mg with atorvastatin.
In the meantime, a proper definition of interchangeability and clear and accepted criteria for the determination of interchangeability remain unresolved issues which continue to have ramifications for the industry. A recent Senate inquiry into consumer access to pharmaceutical benefits and the formation of therapeutic groups, which was initiated on 25 November 2009 and concluded on 24 November 2010,3 recommended that the PBAC:
- develop agreed principles of what constitutes “interchangeable on an individual patient basis”
- develop criteria by which the “interchangeability” of a medicine will be determined and
- publish both the agreed principles and criteria.
Regardless of the outcome of the Federal Court proceedings, there is clearly a need to implement these recommendations.