Aspen Pharmacare Australia Pty Ltd and Minister for Health and Ageing
[2012] AATA 362
•15 June 2012
[2012] AATA 362
Division GENERAL ADMINISTRATIVE DIVISION File Number(s)
2012/0348
2012/0350
Re
Aspen Pharmacare Australia Pty Ltd
APPLICANT
And
Minister for Health and Ageing
RESPONDENT
Decision
Tribunal Justice Kerr, President
Dr Nicoletti, MemberDate 15 June 2012 Place Sydney Hear counsel as to remittal of the decision under review to the Respondent for further consideration pursuant to section 42D of the Administrative Appeals Tribunal Act 1975.
...................[SGD]....................................
Justice Kerr, Presiding Member
Catchwords
Cancellation of registration of products containing dextropropoxyphene (DPP) – Whether quality, safety or efficacy is unacceptable within meaning of the Therapeutic Goods Act – Balancing the needs of the individual against benefits at a population level - Whether decision under review should be remitted to Minister for Health and Ageing for further consideration.
Legislation
Administrative Appeals Tribunal Act 1975, Section 42D
Therapeutic Goods Act 1989, Sections 19, 28, 30(2)(a), 60(6)(b).
Cases
CIC Insurance Ltd v Bankstown Football Club Ltd (1997) 187 CLR 384
Australian Education Union v General Manager of Fair Work Australia (2012) 286 ALR 625
Secondary Materials
Macquarie Dictionary (5th Edition) 2009
New Shorter Oxford English Dictionary Vols 1 and 2, 1993
REASONS FOR DECISION
Justice Kerr, President
Dr Nicoletti, Member15 June 2012
introduction
Products containing dextropropoxyphene (DPP) have been available for therapeutic use in Australia since at least 1970. The therapeutic use of products containing DPP is for the relief of pain. Two such products, Di-Gesic and Doloxene, are the subject of this decision.
Subject to this Tribunal determining otherwise, both Di-Gesic and Doloxene will be cancelled from the Australian Register of Therapeutic Goods (ARTG). This follows a chain of events which began on 3 November 2011, when the delegate of the Secretary of the Department of Health and Ageing gave notice of a proposal to cancel the registration of all products containing DPP from the ARTG as at 1 March 2012. The decision was initiated pursuant to section 30(2)(a) of the Therapeutic Goods Act 1989 (the Act).
The delegate had concluded that ‘recent evidence concerning dextropropoxyphene-containing medicines’ added to existing knowledge of each of the drugs, indicated that their safety was unacceptable. A decision to deregister the products was made.
The company that markets Di-Gesic and Doloxene in Australia, Aspen Pharmacare Australia Pty Ltd (Aspen), asked the Minister to reconsider that decision.
On 23 January 2012, a delegate of the Minister affirmed the original decisions to remove both products from the ARTG. The date for their removal was unchanged.
Aspen has exercised its right pursuant to section 60(6) of the Act to seek review by the Administrative Appeals Tribunal of the Minister’s delegate’s decisions.
The parties agree, and the Tribunal accepts, that the ultimate issue is whether the correct or preferable decision is that the ‘quality, safety or efficacy’ of each drug ‘is unacceptable’ within the meaning of section 30(2)(a) of the Act.
The products AND the Stay
Di-Gesic was automatically registered in the ARTG (“grandfathered”) when the Act commenced in 1991. It has been supplied to patients in Australia since 1984. It contains 32.5mg of DPP hydrochloride and 325 mg of paracetamol and is currently supplied in blister packs of 20 tablets. Di-Gesic is indicated for the relief for mild to moderate pain with dosing recommendations of 2 tablets every four hours for pain.
Doloxene was also registered in the ARTG when the Act commenced. As currently supplied it was entered into the Register in 2009. It contains 100 mg of DPP and is currently packaged in blister packs of 10 tablets. Doloxene is indicated for as the relief of mild to moderate pain with dosing recommendations of 100mg four hourly or as needed for pain.
After initiating its application for review, Aspen sought a stay to enable it to continue to supply the affected products while it pursued its review rights before the Tribunal. On 14 February 2012 the Tribunal, constituted by Justice Downes, President, stayed both operative decisions (those which would have otherwise resulted in deregistration of Doloxene and Di-Gesic on 1 March 2012) until further order.
To obtain this stay, Aspen undertook to add significant safety warnings to the product information (PI) and consumer medicine information (CMI) for Di-Gesic and Doloxene. Both documents currently state that patients prescribed products containing DPP to manage chronic pain should attempt to replace those products with alternatives and that doctors should not initiate the use of any products containing DPP for new patients, whether for treatment of acute or chronic pain. ‘Black box’ warnings were added highlighting, inter alia, that products containing DPP had recently been associated with substantial prolongation of the QT interval. Aspen also agreed to write ‘dear doctor’ and ‘dear pharmacist’ letters, in a form approved by the Respondent, drawing attention to the additional safety warnings. Full details of what was required as a condition of the stay are provided in the reasons of Justice Downes, President dated 16 February 2012. The Tribunal was advised that Aspen has complied with those obligations.
The law
Subject to limited exceptions not presently relevant, a prescription medicine cannot be sold in Australia unless it is registered in the ARTG. The Secretary of the Department of Health and Aging has responsibility pursuant to s 25 of the Act to evaluate ‘goods’ for which registration is sought having regard to, inter alia, whether the quality, safety, and efficacy of the goods for the purposes for which they are to be used have been satisfactorily established.
Registrations can be revoked. Section 30(2)(a) of the Act provides, subject to statutory procedural fairness rights conferred by ss 30(3) and (4), that the Secretary may cancel the registration. Section 30(2)(a) provides:
…the Secretary may, by notice in writing given to a person in relation to whom therapeutic goods are included in the Register, cancel the registration or listing of the goods if: (a) it appears to the Secretary that the quality, safety or efficacy of the goods is unacceptable…’.
It was suggested by Dr Renwick SC for the Minister that the use of the expression ‘appears’ may have been chosen by Parliament to indicate that there was a lower burden of proof than would ordinarily apply in relation to a finding that the quality, safety or efficacy of therapeutic goods was unacceptable. However, no authority was cited for this proposition and the Tribunal was not taken to any extrinsic materials to that effect. Mr Lloyd SC for Aspen submitted that the language of the legislation merely indicated that the Parliament intended that the issue was not a jurisdictional fact for a court to decide. It was used to identify the person whose satisfaction was at stake. The Tribunal agrees with those latter submissions. The Tribunal is of the view that these questions should be determined by applying the normal principles of proof that apply whenever a decision maker is required to be satisfied of some conclusion, not any lower standard.
The system created by the Act to govern deregistration has three tiers:
(a)The initial decision is made by the Secretary.
(b)A person whose interests are affected by a decision to cancel the registration of therapeutic goods has the right to request the Minister to reconsider the decision (s 60(2)). As a result of that reconsideration the Minister may confirm the decision or revoke it, or may revoke it and make another decision in substitution for it.
(c)A person whose interests are affected by, and who remains dissatisfied with, the Minister’s reconsideration may then make an application to the Administrative Appeals Tribunal for independent merits review of that decision (section 60(6)(b)).
At each stage of this three tier scheme: that is, initial decision, reconsideration and review, the criteria for cancellation of registration of therapeutic goods remain identical; thus the relevant decision-maker may deregister the goods if, but only if, it appears to him or her that ’the quality, safety or efficacy of the goods is unacceptable.’
Dr Renwick SC submitted that the relevant standard or norm against which quality, safety or efficacy must be assessed is that relating to “therapeutic use” as defined by section 3 of the Act. Therapeutic use is there relevantly defined as `use in or in connection with (a) preventing, diagnosing, curing or alleviating a disease, ailment, defect or injury in persons or (b) influencing, inhibiting, or modifying a physiological process in persons’ The Tribunal accepts that analysis. This does not appear to be contentious. For drugs intended for pain relief, the provisions of sub-clause (b) would appear relevant, viz “influencing, inhibiting, or modifying” the physiological processes of pain.
The word ‘quality’ is a defined term which has an extended meaning. Section 3 of the Act states that in relation to therapeutic goods ‘quality’ ‘includes the composition, strength, potency, stability, sterility, purity, bioburden, design, construction and performance characteristics of the goods’. However, nothing turns on this in the present case. Neither at the initial decision stage nor at reconsideration has any suggestion been advanced that the relevant ‘goods’ Di-Gesic and Doloxene were, as a result of some want of quality, unacceptable for registration. No such case was put by the Respondent in these proceedings. The Tribunal is itself satisfied that there is no basis in the evidence to suggest that the ‘quality’ of the goods is unacceptable.
There is no case law known to the Tribunal that suggests that the words ‘safety or efficacy’ should be construed otherwise than as ordinary English words, albeit read in the context of the Act as a whole: see CIC Insurance Ltd v Bankstown Football Club Ltd (1997) 187 CLR 384 at 408, and more recently, Australian Education Union v General Manager of Fair Work Australia (2012) 286 ALR 625 per French CJ, Crennan and Keifel JJ at [27]. How then should the words used in the Act be understood?
Without seeking to put a gloss on the statutory language, recognising the prime importance of reading the word in the context of the Act, ‘safety’ appears to the Tribunal to have been used in the sense of referring to ‘the quality of being unlikely to cause hurt or injury; the quality of not being dangerous or presenting a risk’ (see number 4 of the meanings given by the New Shorter Oxford English Dictionary Vol 2 N-Z 1993).
Efficacy is defined by the Macquarie Dictionary (5th Edition) 2009 as the ‘capacity for serving to produce effects; effectiveness’ and by the New Shorter Oxford English Dictionary Vol 1 A-M 1993 as the ‘ability to bring about the intended result’. It is in that latter sense that the Tribunal understands the term to be used, that is, effectiveness to bring about the intended therapeutic result for which the goods have been registered, in the instant cases, to produce analgesia or, more colloquially, relief from pain.
Only if it appears to the Tribunal that the goods are ‘unacceptable’ for reasons of safety or efficacy as so understood can the correct or preferable decision be that the registration of the goods should be cancelled.
As with the other words in this statutory context, the term ‘unacceptable’ does not appear to the Tribunal to be a term of art. It is an ordinary word to be understood in the context of the Act as a whole. Dr Renwick SC referred the Tribunal to the Macquarie Dictionary meaning of ‘unacceptable’ as ‘so far from the required standard, norm, expectation, etc. as to be not allowed.’ The Tribunal agrees that this is the meaning conveyed by the language of section 30(2)(a) of the Act.
A decision-maker charged with the responsibility of forming a view as to whether the efficacy or safety of particular therapeutic goods is so far from the required standard as to be unacceptable must balance the efficacy of the therapeutic goods against the risks of the hurt or injury their use or misuse may give rise to. That the Tribunal is required to undertake such a balancing judgment was supported both by Aspen and the Minister. Dr Renwick SC submitted, and Mr Lloyd SC agreed, that safety cannot be considered independently of efficacy. Counsel agreed that if a particular therapeutic product had exceptional efficacy (hypothetically, a wonder drug that could cure cancer) then despite dangerous and potentially lethal side effects it may be appropriate to regard its safety as acceptable. However, identical side effects encountered with a less efficacious drug might lead it to be cancelled from the ARTG because its safety would be unacceptable when weighed against its efficacy. That approach to interpretation accords, as the Tribunal apprehends it, with the legislative scheme of the Act and with common sense. The Tribunal agrees that it is required to take such a ‘balancing’ approach to the critical questions of safety and efficacy.
Some observations on the Witnesses
While the evidence for the Respondent Minister was not exclusively dependent upon the testimony of Professor Buckley and Dr Halliwell, it depended substantially on the Tribunal’s acceptance of the propositions advanced by those witnesses. The Tribunal is at pains to emphasise that what it now states should be taken as no reflection on the professional integrity or eminence of either witness, but it was plain that both witnesses were highly resistant to conceding any point that might possibly be thought to assist the interests of the Applicant. This was particularly noticeable in both cases in respect of the Hawton et al evidence, which concerned the study of the impact of the withdrawal of Co-proxamol (the UK name for a drug containing DPP) on prescribing and deaths (Exhibit R3, Tab 39). The Respondent relied on the data in that study report, and particularly the data in Table 2 on page 5 of the report, to support a proposition that after the withdrawal of Co-proxamol in the United Kingdom there was a significant reduction in the rate of deaths by suicide and no commensurate increase in the number of deaths associated with other analgesics. Both witnesses were taken to Table 2 by Mr Lloyd SC and asked whether the data showed by necessary implication that there had been an increase in the number of deaths by deliberate taking of drugs other than analgesics such that the rate of suicide, through the use of therapeutic products generally, was not as greatly affected by the withdrawal of DPP from the prescribing market as they had suggested. Even when pressed, both Professor Buckley and Dr Halliwell appeared most reluctant to accept the straightforward and inescapable mathematical consequences that flowed from the published data.
Mr Lloyd SC drew the Tribunal’s attention to the fact that Professor Buckley was a member of the Advisory Committee for the Safety of Medicines (ACSM) of the Therapeutic Goods Administration (TGA) that recommended that DPP be withdrawn (and, the Tribunal notes, the author of the article which was the trigger for the Australian Drug Evaluation Committee’s 1998/99 review of DPP (T11, at page 344). Professor Buckley needed to be prompted by Dr Renwick SC to disclose his membership of the ACSM to the Tribunal.
The Tribunal accepts that Mr Lloyd SC accurately described Dr Halliwell as a witness who sought continually to work out what counsel was going to make of his evidence and craft his answers so that those answers could not be manipulated. The Tribunal had to intervene on more than one occasion to indicate to Dr Halliwell that he should simply answer the questions asked of him.
In pointing these matters out, the Tribunal casts no doubt at all on the strong and obviously firmly and honestly held views of these two witnesses, but the Tribunal gained the impression that both witnesses found it difficult to reconcile their strong views and beliefs with their overriding duty to provide impartial assistance to the Tribunal on matters relevant to their areas of knowledge and experience, and to not become advocates for a party to the proceeding, as is the requirement under the guidelines for expert witnesses giving evidence in proceedings. The Tribunal could not help but conclude, in respect of their evidence, that wherever the data were less than transfixingly clear, any ambiguity was construed by each witness in favour of the case being advanced by the Respondent Minister and where any marginal judgement call was required, their judgement would similarly be made in that direction.
Efficacy
It was contended by the Minister, and not challenged by Aspen, that the rigour of the process required to achieve registration has increased since Di-Gesic and Doloxene were first registered. Dr Renwick SC for the Minister asked the Tribunal to note that ‘...the two drugs were not the subject of evaluation under s 25 before registration, rather they were registered following the commencement of the Act through a process known as ‘grandfathering’.
Aspen, however, pointed out that both drugs were comprehensively reviewed in 1998/9 by the Australian Drug Evaluation Committee (ADEC), which was a statutory committee established by the TGA to make recommendations as to whether or not applications for registration of prescription medicines should be approved. Mr Lloyd SC referred the Tribunal to ADEC’s finding ‘that these preparations are analgesic at normal doses’. Mr Lloyd SC submitted that the 1998/9 review was comprehensive and thorough. Dr Renwick SC disagreed, noting that ADEC only considered efficacy data that were selected and submitted by the sponsor (which was not Aspen at the time) and ADEC’s view was that the data could “therefore be expected to give a relatively favourable overview of the drug’s efficacy” (T4, page 142].
The Tribunal was referred to the review documentation and having considered it accepts that the 1998/9 review by ADEC acknowledged the efficacy of both drugs (Extract from the ratified minutes from the 200th (1998/6) meeting of the Australian Drug Evaluation Committee (ADEC) held 3-4 December 1998, paragraphs 2.23.8 and 2.23.13 [Exhibit A1, Tab 11]. Mr Lloyd SC submitted the Minister had not identified or put forward any new evidence touching on efficacy since that review.
The starting point for the Tribunal must be an examination of the decisions now subject to review. The Minister’s delegate [T3 at p119] did not make a positive finding that the efficacy of Doloxene (containing only DPP) was unacceptable. Rather, the delegate concluded:
...at this point in time I have no evidence that the efficacy of goods containing [DPP] is acceptable.
That is to be contrasted with the delegate’s unambiguous conclusion that the safety of DPP was not acceptable [T3 at p120].
In the case of Di-Gesic, which is the product containing both DPP and paracetamol, the Minister’s delegate seems to have impliedly accepted it as having efficacy as an analgesic but the finding is not express. It is only to be inferred from the same statement as was included in the delegate’s reasons in respect of Doloxene: ‘it is unclear apart from anecdotal claims if the use in combination with paracetamol…provides any additional analgesic benefit to paracetamol alone’ [T2 at p103].
The parties’ openings suggested that there might be considerable disputation, or at least uncertainty, as to the efficacy of these products. However, in the event, following a meeting arranged before their giving concurrent evidence, the parties’ key experts on efficacy, Professor Christie and Dr Halliwell, both agreed that while there were limited good quality data, such evidence as did exist supported the conclusion that DPP on its own (as supplied in Doloxene) was a weak opioid analgesic and clinically effective as such. Both experts also agreed that DPP in combination with paracetamol (as supplied in Di-Gesic) was more effective than DPP on its own (Exhibit T1: Outcome of Meeting of Joint Experts Professor Christie and Dr Halliwell). The efficacy of DPP on its own or in combination with paracetamol was, from that point onwards, as a stand-alone question, never truly in issue.
In the end, it was not in dispute between the experts that both Doloxene and Di-Gesic are each more efficacious than placebo. That was their evidence and the Tribunal so concludes.
Of course, neither product is effective in all patients. No analgesic is. The Moore Cochrane review 2008 (reprinted in 2011) [Exhibit A21 Tab7] considered published articles up to December 2007. Twelve double-blind, randomised ‘single dose’ trials were considered. Six related to DPP alone and six related to DPP plus paracetamol. The relative benefit v placebo with confidence levels reported in brackets was 1.5 (1.2-1.9) for DPP alone (page 8) and 2.5 (2.0-3.2) for DPP/paracetamol (page 12). The number-needed-to-treat (NNT) for one patient to experience 50% pain relief above that seen with placebo was 7.7 (4.6-22) for DPP and 4.4 (3.5-5.6) for DPP/paracetamol. These data provide no reason to justify a conclusion that the products lack efficacy.
In summary, and notwithstanding Dr Renwick SC’s submissions that the Tribunal should not conclude that Doloxene and Di-Gesic have efficacy, the Tribunal finds that each is analgesic for mild to moderate pain (WHO Pain Relief Ladder Tier 2) where a weak opioid, either alone or in conjunction with a non-steroidal anti-inflammatory drug (NSAID) or paracetamol, would be sufficient.
That, of course, did not exhaust the differences as between the parties concerning evidence about comparable efficacy and the relationship between efficacy and safety.
Professor Christie and Dr Halliwell clearly differed in relation to whether the published data suggested that the combination of DPP with paracetamol (as in the case of Di-Gesic) provided a greater analgesic effect than would a normal therapeutic amount of paracetamol alone. In that regard, Professor Christie acknowledged that while that the data fell short of proving the additional efficacy of the DPP/paracetamol combination over a therapeutic dose of paracetamol alone at the 95% confidence level, it was suggestive of that effect. In contrast, Dr Halliwell took the position that no weight at all could be placed on any data insufficiently robust to show results at that level of confidence.
That question, and its relevance or otherwise, to the ultimate issue is addressed later in these reasons at paragraphs 62-83 inclusive.
A further area of dispute between the views of Professor Christie and Dr Halliwell and the other witnesses was with respect to the benefit: risk ratio when the issue of efficacy of these goods intersected with the question of safety. In contrast to Professor Christie, Dr Halliwell and Professor Buckley opined that DPP had significant safety risks and that there would always be equally or more efficacious alternatives that do not such that, in consequence, the ‘quality, safety or efficacy’ of products containing DPP is ‘unacceptable’. There is no doubt in the Tribunal’s view that this area of dispute is both relevant and critical to the identification of the correct or preferable decision.
Safety
Issues of safety appear to the Tribunal to be conveniently capable of being divided into three parts. The first relates to the capacity of DPP to be used for suicide by deliberate overdose. The second concerns risks of safety through accidental overdose. The third is whether drugs containing DPP can be regarded as safe in their normal therapeutic applications.
Deliberate Overdose
There appears little disagreement between the parties as to the fact that drugs containing DPP can be, have been, and will be (unless withdrawn) used for the purpose of deliberate suicide. The Respondent drew the Tribunal’s attention to the fact that the potential cancellation of the registration of DPP had been noted by Exit International [Exhibit R3, Tab 41]. However, the fact that DPP can be used for such purposes has been known for a considerable time. It is not the only drug that is so used. The study based on the NSW Coroner’s data, with reference to the cause of death, reports a total of 32 deaths that involved the primary mechanism of death by poisoning by DPP between 1 July 2000 and 30 April 2012 [Exhibit R3, Tab 40]. This can be compared with 54 deaths over the same reporting period where the primary mechanism of death by poisoning was tramadol and 141 cases in which codeine was the drug involved. Even more startling were the 121 cases in which paracetamol was the drug reported as the primary mechanism of death by poisoning over the same period. Evidence given to the Tribunal suggested that a person would need to take hundreds of paracetamol tablets to achieve that outcome, thereby suggesting at least that a proportion of those intending suicide have a very fixed determination in that regard.
Whilst the NSW Coroner’s figures cannot be conclusive as to accuracy and were contested by the witnesses for the Respondent as under-reporting deaths caused by DPP, they are useful to indicate that many other pharmaceutical products are and will continue to be used by people who seek deliberately to end their lives. In the ratified minutes of the 200th meeting of ADEC, which considered the issue of whether DPP products should remain on the market following the publication of an article, authored by Professor Buckley, drawing attention to the disturbing number of fatalities from self-poisoning which were associated with DPP, ADEC resolved as follows (T11 at p344):
2.23.8 The Committee agreed that these preparations are analgesic at normal doses, relatively safe at normal doses, but can be lethal if abused. It was not appropriate to refuse registration approval for these preparations simply on the basis that overdose can have severe adverse effects. The approach recommended by the Clinical Evaluation was endorsed. There were measures Sponsors could take to reduce the potential for inappropriate use or abuse of these preparations. These included reducing the pack size of the combination tables to 20 and of the DPP tablets to ten, as well as inclusion in the PI of warnings about adverse effects, toxicity in overdose and the enhancement of toxicity if DPP is taken with alcohol…”
The Tribunal, whilst acutely sensitive to public disquiet regarding the use of any product for the purpose of suicide, shares the conclusions that were reached by ADEC in 1998 that it is not appropriate to refuse registration approval for a preparation simply on the basis that it may be used deliberately for the purposes of self-harm. Many therapeutic preparations share that quality.
Accidental overdose
There is equally little disagreement between the experts as to the problems associated with accidental overdoses of products containing DPP. That is because there is a relatively small variation between the dose that is described as therapeutic and the amount which is sufficient to create a risk of serious adverse harm and even the risk of fatality. Whilst the evidence is not completely clear on the point, it seems to be common as between the parties that the amount of DPP contained in a single blister pack of either Doloxene or Di-Gesic, if combined with alcohol, could in a vulnerable patient be sufficient to create a risk of sudden death. Convincing examples were provided by Professor Halliwell as to the circumstances in which such accidental overdose might occur. Professor Christie appeared to share that conclusion. This testimony suggests that the number of tablets required before a risk of fatality occurs is less than the number referred to following the 1998 review by ADEC (T11, p345).
The critical conclusion of the ADEC review was that a reduction of the public health risk could be assisted by the introduction of blister packaging and decreasing the pack size to make spontaneous overdosing more difficult, together with safety-related changes to the Product Information to include a stronger statement regarding the use of DPP with alcohol and other drugs. Those conditions have been met. The Applicant has consented to retaining the even stronger warnings that have been included in the Doloxene and Di-Gesic product information for the period for which a stay has been granted, even if the Tribunal overturns the decision of the Minister that the registration of the products should be cancelled.
Despite those undertakings, the Tribunal accepts that the risk of accidental overdose, while it can be mitigated, cannot be entirely overcome. It is real and appears to be greater than was understood by the ADEC in 1998.
Safety in Therapeutic Doses
The trigger for the Secretary’s consideration of a possible deregistration of the two drugs was the publication of a Multiple Ascending Dose study (MAD) which implicated the use of DPP with QT prolongation and life-threatening arrhythmias (at T63, p938):
In November 2010 the TGA received new information about the toxicity of dextropropoxyphene-containing medicines (specifically, new information identified by the FDA concerning the potential for dextropropoxyphene-containing medicines to cause QT prolongation and life-threatening arrhythmias, based on results of a multiple ascending dose study of the effects of dextropropoxyphene on cardiac conduction).
While that study prompted regulatory action not just in Australia but in a number of other countries, Dr Renwick SC explicitly and repeatedly stated that the Minister’s case for deregistration was not linked to the Tribunal’s acceptance of the validity of that study. That appears to the Tribunal to have been a sound concession given the effective attack on the weight that should be placed on the study mounted by Professor Horowitz.
Professor Horowitz testified as follows (See Transcript 29.5.12 at p143-144):
I would now like to return to the MAD study. Now, I’ve taken exception to the MAD study on many grounds, and I would like to go through those in detail. First of all, the MAD study was what is called a thorough QT study. The idea of this was to assess the potential – usually of a new drug, one which has not yet been used in human beings – the chance that this drug will be arrhythmogenic. The sequence of events is the drug is found to have some effect on hERG channels – to block hERG channels, and it is found that this is relatively potent, as indeed could be argued with DPP when it’s metabolised, and therefore you screen to see how much QT prolongation actually occurs.
What actually happened with the MAD study is very hard to understand, because it has never been published. It is very, very unusual, almost unprecedented, that a study which purports to have a major influence on therapeutics, which has superficially led to the withdrawal of this drug in many countries in the world, has actually never been released for either presentation at a major meeting, or more importantly, has ever been through the publication process. And there are - even looking at the reports of this study superficially…there are two obvious problems with this study: (1) the study involves some errors of treatment assignation so that the 600 milligram dosing schedule had to be done a second time.
Secondly, the study is undersized, in that the 600 milligram dose has not quite reached significant QT prolongation, and yet the study was stopped, despite the fact there were no adverse events. Now, for any normal clinical trial, there are what are called stopping rules agreed beforehand, and it is not clear – I have not seen what the stopping rules were for this study, but they chose to stop it, again, very unusual.
There was then a very useful exchange between Professors Buckley and Horowitz in which the following evidence is recorded (at Transcript 29.5.12 p153):
PROFESSOR HOROWITZ: The second critical point: that this is particularly important because in the MAD study the 600 milligram dose had to be repeated because the first run of it had been fouled because of inappropriate assignment of treatment groups and at the time when the drug was stopped, even allowing for the fact that the drug had two effects – two components effects: an effect …at the QT. According to the best analysis, that summation of those effects did not quite reach statistical significance and yet it was stopped. So perhaps I can ask you, Professor Buckley, what sort of stopping rules would you say that that achieved – abided by?
PROFESSOR BUCKLEY: I would have to look at – there will be a summary of the total somewhere, won’t there? I would think that in an ascending dose study you would be looking for some margin of effects that would indicate a concern. This clearly crossed that boundary in terms of the standard guidance on interpretation of QT intervals. It’s…
PROFESSOR HOROWITZ: I would remind you that the total effect of the 600 milligram dose on their rerun did not reach statistical significance. How can that achieve a stopping rule?
PROFESSOR BUCKLEY: The purpose of the guidelines is not to indicate that you show a statistically significant difference, it’s to indicate where the confidence in the rule is my understanding, and that’s – so the purpose of these sorts of studies in a normal context is to rule out an effect like this, and quite clearly after the 900 milligram dose they could not rule out that effect.
The discussion then continued later in evidence where the two experts appeared to reach agreement (at Transcript 29.5.12 p157):
PROFESSOR BUCKLEY: Yes, so the primary object, in fact, as stated, was to determine the daily maximum tolerated dose, and they defined that as one which didn’t cause a statistically significant increase in QT.
PROFESSOR HOROWITZ: Yes, which was 600.
PROFESSOR BUCKLEY: Yes, so it determined that 600 was the maximum tolerated dose before you would achieve a statistically significant increase in QT.
PROFESSOR HOROWITZ: Yes, that’s what the data suggests.
Dr Nicoletti then asked on behalf of the Tribunal (at Transcript 29.5.12 p158):
DR NICOLETTI:…in terms of the lack of statistical significance with the 600 milligram dose, that is the maximum dose per day, is it not?
PROFESSOR HOROWITZ: Yes, but I do agree with Professor Buckley that if – that – let’s say for argument’s sake that it’s 600. First of all, let me just put a caveat in here: my view is that minor degrees of widening of the QRS complex are not dangerous, right? But, having said that, I certainly agree that if you are prescribed 600 milligrams but you happen to be a person with impaired clearance of the drugs, say by being elderly, or if you happen to take some alcohol, that 600 could easily transcend into 900 or maybe a bit more than that. I agree with that entirely, so I agree with the context of the trial. I’m happy to say that the MAD study more or less clearly demonstrates that there is some QRS widening at 900 milligrams, which is likely to be achieved by some, for example, people with slightly impaired kidney function who are at the upper age range. I think – I agree with that entirely. What I don’t think is that this is necessarily a disaster. In fact, I say there’s an absolute absence of evidence of any such thing.
The Tribunal has carefully considered the documents, witness statements and the oral testimony and debate between the expert witnesses. It acknowledges that Professor Buckley held strongly to the opinions he expressed, throughout the concurrent evidence session and cross examination, regarding the danger of unexplained sudden death that he is concerned DPP is implicated in. However, the Tribunal is not persuaded that those concerns are supported by any robust evidence. It notes that Professor Buckley found it difficult to concede any point to either Mr Lloyd SC or Professor Horowitz during the concurrent evidence session. Professor Horowitz summarised his opinion during the course of oral evidence in the following terms (at Transcript 29.5.12 page 147.39):
PROFESSOR HOROWITZ: If [DPP] were taken in ordinary doses for up to a week, by people with normal kidneys, under the age of 70, it’s my view that unless they took a huge dose of a sedative on top of it, the potential for sudden death is about zero.
The Tribunal accepts Mr Lloyd’s submission that the evidence supports a finding that Professor Horowitz’s opinion regarding the risk of “sudden death” was correct.
In reaching its conclusion, the Tribunal has not ignored the tender by Mr Renwick SC of the Adler et al case report: Propoxyphene-induced Torsades des Pointes (Exhibit R3, Tab 38). This study purportedly documented the first case of Torsades related to the use of propoxyphene (which, the Tribunal notes, was described in the study as an effective pain-killer). However, Mr Lloyd SC in his closing submissions pointed out that the following statement in the Adler et al study equally implicated other drugs (at p1953):
Even short term treatment with non-steroidal anti-inflammatory drugs has been associated with increased risk of death and recurrent myocardial infarction in patients with coronary disease.
Mr Lloyd SC also drew attention to the fact that the patient whose death was the subject of the study had a congenital defect and a genetic predisposition to Torsades. The patient was aged 74 (Exhibit R3, Tab 38). Dr Renwick SC acknowledged those points but nonetheless pressed a submission to the Tribunal that it was evidence of at least one example of an association of DPP with Torsades following ingestion. The Tribunal is sceptical, for the reasons articulated by counsel for Aspen, that any weight can be placed on the case report. The Tribunal also notes that Dr Renwick SC chose not to put the study to Professor Horowitz to comment upon despite this being directly within the area of, but apparently inconsistent with, his testimony.
Finally, the Tribunal should address the issue of the lesser side-effects of DPP. There are many contra-indications for use which are identified in the product information. These appear in the main uncontentious. Many medicines have contra-indications for particular patient groups or conditions. DPP in that regard is no different. As to therapeutic use in patients for which it is not contra-indicated, the evidence is that DPP can have significant side-effects but no greater than the range of alternative drugs for mild to moderate pain. Thus, for example, The Cochrane review by Moore et al [Exhibit A21, Tab 3] refers to the authors’ conclusions in the following terms:
Since the last version of this review, no new relevant studies have been identified. The combination of dextropropoxyphene 65mg with paracetamol 650mg shows similar efficacy to tramadol 100mg for single dose studies in postoperative pain but with a lower incidence of adverse effects. The same dose of paracetamol combined with 60mg codeine appears more effective but, with the slight overlap in the 95% CI, this conclusion is not robust. Adverse effects of both combinations were similar.
In another study which compared the reporting rate of adverse drug reactions to the French pharmacovigilance system for DPP, tramadol and codeine in combination with paracetamol, the seriousness of adverse drug reactions was significantly higher with DPP than with codeine and paracetamol, but was less with DPP when compared with tramadol.
Both the Cochrane review and the French study indicate that the safety of DPP at therapeutic doses is not sufficiently different from other weak opioids and does not support a finding that the safety of DPP is unacceptable.
ALTERNATIVES
From the discussion above it is evident that the Tribunal’s main concern regarding safety is the potential lethal consequences that can follow from accidental overdose. If the Tribunal was persuaded that there would always be one or more suitable and efficacious alternative pain relief medicines for mild to moderate pain which avoided the risk of such consequences following accidental overdose, the Tribunal would favour upholding the deregistration of Doloxene and Di-Gesic. The evidence associated with alternatives is therefore crucial in relation to the Tribunal’s findings. On the one hand the Tribunal received evidence from both Dr Halliwell and Professor Buckley that if the registrations of Doloxene and Di-Gesic are cancelled, there will always be effective alternatives available to patients and prescribers. On the other, Professor Christie argued there would still remain a group of patients, albeit small, for whom no other comparable drug was available or suitable and for whom DPP or DPP/paracetamol would be the only efficacious and appropriate drug for the treatment of their pain.
The DPP/Paracetamol Debate
Before addressing that subject, the Tribunal should deal with an issue which, despite the large attention that was given to it by counsel for both the Minister and Aspen, was, in its opinion, largely a red herring. That question is whether or not DPP in combination with paracetamol has or has not greater efficacy than a therapeutic dose of paracetamol alone.
Professor Christie and Dr Halliwell clearly differed as to whether the underlying published data suggested that the combination of DPP with paracetamol (as in the case of Di-Gesic) provided a greater analgesic effect than would a normal therapeutic amount of paracetamol alone.
Professor Christie acknowledged that while that the data fell short of proving the additional efficacy of the combination of DPP and paracetamol over a therapeutic dose of paracetamol alone at the 95% confidence level, they were suggestive of that effect.
Dr Halliwell by contrast took the position that no weight at all could be placed on any data insufficiently robust to show results at that level of confidence.
Additionally Dr Halliwell advanced the more robust proposition found in much of the prescribing literature tendered to the Tribunal to the effect that the evidence established there to be no difference in clinical effect between DPP/paracetamol and paracetamol alone. He drew the Tribunal’s attention to the data collected in the Moore review [see paragraph 37 above] as justifying that conclusion given the overlapping confidence intervals as between DPP and DPP/paracetamol. Dr Halliwell’s evidence was that if there was any overlap in the confidence intervals when different analgesics were compared in the context of a patient that had had surgery, no meaningful conclusions about differences in efficacy between those analgesics could be drawn.
During cross-examination Mr Lloyd SC referred Dr Halliwell to the results tabulated in Summary table B of pages 11-14 of the Moore review, which tabulated the results the study had judged to be reliable in all types of surgery.
Questioned in relation to these tabulated data Dr Halliwell conceded that the same proposition would have to apply to different doses of paracetamol (500 mg, 600/650 mg and 975/1000 mg), which also had overlapping confidence intervals:
MR LLOYD: If you turn back to page 14, you will see three different doses of paracetamol, so this is the same group that we were previously looking at. For all types of surgery, the NNT amounts for this, having regard to the confidence intervals, they overlap for each of those three amounts of paracetamol. And I think what your evidence was, was something to the effect of, if they overlap that means there’s no basis – there’s no scientific basis for preferring one over the other in terms of its likely effect?
DR HALLIWELL: Correct
When this proposition was explored further by the Tribunal, Dr Halliwell gave the following evidence:
DR NICOLETTI: Dr Halliwell, are you suggesting that if there’s any overlap in the confidence intervals, then you can’t draw any conclusions about any difference is that – that seems to be what you’re saying?
DR HALLIWELL: Yes, but I’m saying that that is in the context of a patient that has had surgery. You’re only giving them paracetamol. They’ve got moderate to severe pain. So if the analgesic is not terribly effective, you don’t really see much of a difference, do you?
DR NICOLETTI: So just as an example, on page 20, if there’s some overlap between paracetamol 600/650 and oxycodone 10/paracetamol 650, are you suggesting that a combination of oxycodone 10 and paracetamol 650 might not be different [to paracetamol 600/650], because that would seem to be an extraordinary proposition for a strong opioid like oxycodone?
DR HALLIWELL: Oxycodone 10 is only a small dose. Yes, I mean, the problem with looking at oxycodone, it’s not an all or none effect, it’s a graduated response.
DR NICOLETTI: I just don’t accept that from the potencies of those opioids that you could essentially say that taking 10 milligrams of oxycodone along with paracetamol would have no better effect than [a] 650 milligram dose of oxycodone (sic). Intuitively, that would seem an extraordinary proposition?
DR HALLIWELL: No, and that’s from the data presented here. But, I mean, the –
DR NICOLETTI: From the data presented here, but in clinical practice is that what you’re saying? Are you saying that in your view that would be quite logical in clinical practice?
DR HALLIWELL: That dose isn’t enough. I mean, oxycodone five milligrams provides – on its own, provides almost no pain relief.
DR NICOLETTI: What I’m asking you, Dr Halliwell, in your clinical practice, in your experience with patients, would you say that oxycodone 10 milligrams plus paracetamol 650 milligrams is equally efficacious to 650 milligrams of paracetamol?
DR HALLIWELL: No, I’m not, but we have to look at the context. This is other painful conditions; we have to see what that is. Maybe it’s not effective because it’s a painful condition where oxycodone is not enough. That’s why I’m saying graduated doses of paracetamol in someone with severe pain; you won’t see a different because it’s such a weak effect.
DR NICOLETTI: But doesn’t that same argument then apply to dextropropoxyphene and paracetamol or dextropropoxyphene in combination with paracetamol and paracetamol alone? The principles that you’re suggesting?
DR HALLIWELL: Yes, the same principle. Yes, but what we need to do is compare either paracetamol alone with paracetamol plus dextropropoxyphene, because our question is: is it worth adding on dextropropoxyphene? Or we compare dextropropoxyphene versus nothing.
DR NICOLETTI: Well, according to this, if we would apply this table here, we would say, “Well, it’s not worth adding oxycodone 10, in this context, because there’s an overlap.”?
DR HALLIWELL: In that context, yes.
Notwithstanding this exchange Dr Halliwell remained committed to his evidence that the Moore review supported the proposition that DPP/paracetamol had been proven to be no more effective than paracetamol alone. Yet if the Tribunal is to accept the proposition advanced by Dr Halliwell on that basis, then it must also accept that the Moore review supports the proposition that oxycodone 10 mg/paracetamol 650 mg is no more effective than paracetamol 650 mg alone, and paracetamol 975mg/1000 mg is no more effective than paracetamol 600/650 mg or paracetamol 500 mg. The Tribunal is not persuaded that this is plausible. Moreover, as can be seen from the evidence reproduced at paragraph 69 above, Dr Halliwell accepted that such a conclusion would be inconsistent with his own experience of patients in his clinical practice.
The Tribunal does not accept, therefore, that the Moore review provides support for the proposition that a combination of DPP and paracetamol is no more efficacious than paracetamol alone. In fact, no meaningful conclusions can be drawn from its data about the relative efficacy of many of the analgesics it compared. To use the words of Mr Lloyd SC [at Transcript 30.5.12, p206], where there were overlapping confidence levels (which was commonly the case) the data could only be relied upon to support the proposition that there is “no evidence of effect” as opposed to “evidence of no effect”.
The Tribunal further finds that despite much effort expended on this subject, no other study of analgesic products was produced that would stand as proof of the negative proposition that the combination of DPP and paracetamol was no more effective than paracetamol alone.
Dr Renwick SC referred the Tribunal to statements to that effect in the Australian Medicines Handbook and the Therapeutic Guidelines and suggested that the Tribunal would be acting ‘irrationally’ if it declined to accept the collective wisdom of the clinical community which had formed such views. However, as Mr Lloyd SC responded, clinical advice, to be reliable, must be founded on rigorous science and ‘a stream cannot rise higher than the source’. Mr Lloyd SC noted that ‘it is reasonably apparent that the observations on efficacy in those documents were based on the 1997 study of Li Wan Po’ (Exhibit A21 Tab 6). Timing and the context suggests that was the case. Dr Renwick SC did not press the contrary. The Tribunal accepts Mr Lloyd SC’s contention.
The Li Wan Po study failed to find evidence at the 95% confidence interval that DPP/paracetamol was more efficacious than paracetamol alone but it was viewed by Professor Christie as demonstrating a trend in favour of the combination, rather than the contrary. Dr Halliwell insisted that data falling just short of the 95% confidence level should be entirely disregarded as proof of an effect. However, even on that basis, absence of clear proof of an effect is not the logical equivalent of proof of no effect.
The Li Wan Po study is the high watermark of the literature from which any adverse conclusion regarding the relative efficacy of DPP/paracetamol combination as contrasted with paracetamol alone might be drawn. As the high watermark, it does not justify the more robust statements found in the prescribing literature referred to by Dr Renwick SC.
If relevant, the Tribunal accepts Dr Christie’s evidence that the Li Wan Po study demonstrates a trend in favour of the combination, rather than the contrary.
Further, Mr Lloyd SC drew the Tribunal’s attention to many older studies, single and multiple dose, a number seemingly credible albeit much critiqued by Dr Halliwell, such as Huskisson 1984 (Exhibit A21, Tab5) which suggest that patients receiving DPP with paracetamol gain greater analgesia than with paracetamol alone. Huskisson’s study was of sufficient quality to be included in the Cochrane review, but it was excluded from the Li Wan Po research because its data was not in a form usable in their analysis. The Tribunal can find nothing to justify Dr Halliwell’s suggestion (Transcript 30.05.12 at page 229.8-229.24) that the Huskisson study was excluded because it was flawed.
Counsel also drew the Tribunal’s attention to the fact that the efficacy of DPP is likely to be underestimated in single dose trials such as in the Li Wan Po study and to be more efficacious in multiple doses after it reaches a steady state (Statement of Professor Christie, Exhibit A10 at paragraph 4) and, to the apparent concession by Dr Halliwell, that at least one study validly demonstrated that young women who had had episiotomies had been shown to so respond (Transcript 30.05.12 at p228.21). The Tribunal takes the view it is entitled to regard the witness as having actually conceded the latter point given its own assessment of the circumstances in which the evidence was given and Dr Halliwell’s generally strong resistance to concessions discussed at paragraphs 25-28 above.
The Tribunal therefore finds, on the balance of probabilities, that the evidence is modestly in favour of the likelihood that DPP and paracetamol in combination is more efficacious than paracetamol alone, at least for some patient groups.
However, in the end this debate, while clearly significant to the parties, appears to the Tribunal to have been an arid detour. The Tribunal raised this concern with Dr Renwick SC when, in his concluding submissions, he was pressing the contention that the evidence before the Tribunal should be found to have established that the combination of DPP and paracetamol had no greater efficacy than paracetamol alone. The supposed link between the resolution of that question and the evaluative task actually required by the Act was never made clear. Without such a link the issue is not relevant.
In deference to the weight placed on the issue by both counsel and lest we be in error in our conclusion in that regard, the Tribunal has addressed the question and provided its reasons for its findings. However, the Tribunal is not persuaded that its task, properly understood, required it to umpire the argument about whether or not DPP in combination with paracetamol has, or does not have, greater efficacy than paracetamol alone.
Nor does a conclusion in respect of that contested question seem relevant to the Tribunal to any relevantly required balancing consideration regarding the interrelationship between the goods efficacy and safety. While it was contended by the Respondent that there would always be appropriate alternative analgesic products that could be used if DPP was withdrawn, it was not part of the Minister’s case that Doloxene or Di-Gesic could always be substituted with paracetamol alone.
No Suitable Alternative?
Given the Tribunal’s findings with respect to the efficacy and safety of Doloxene and Di-Gesic, the Tribunal concludes that the remaining relevant question is whether, notwithstanding DPP’s safety risks, there is a group of patients for whom DPP is efficacious and for whom there is no suitable alternative to analgesics containing DPP, whether supplied on its own as Doloxene or with paracetamol as Di-Gesic.
There can be no doubt that at least within the patient community that is currently prescribed DPP, some strongly believe that their pain condition cannot be effectively treated by alternative products. An example of such a view being expressed can be found in submissions given to the Minister’s Delegate (T94, page 1455), which include correspondence from a patient who explains she suffers from Fibromyalgia Syndrome (FMS). She says she takes Di-Gesic and finds it extremely helpful as part of her pain management schedule. She recounts that she has taken other medications when she has had stronger pain and they have not been as effective as Di-Gesic, while exacerbating her gut dysfunction that is a significant symptom of the syndrome. She writes:
…I have found codeine and similar opiates to be ineffective, due to the way my body processes medications.
I find that most medications I take carry an element of risk. All will probably shorten my life, compared to a healthy person not taking medications. For me the difference is between living a life of hell, with inadequate pain management, and living the best life I can within the limitations of my illness. I am sure I’m not the only person with FMS who cannot tolerate opiates. This seems like a decision made by the nanny state, instead of informing patients of the risk and allowing them to decide what level of risk they will accept to achieve some quality of life.
The response to such submissions by Professor Buckley and Dr Halliwell was to the effect that many such patients are just reporting nothing more than a placebo effect or are in fact addicted to DPP and unable to rationally appreciate that there are appropriate alternatives available to them. In their view, such patients may suffer a period of some discomfort as they shift to new products, but overall their pain relief can be maintained. This assumption is challenged by the evidence of Professor Christie. Their disagreement is captured by an exchange of testimony recorded at Transcript 29.05.12 at page 116.5 – 116.33):
DR RENWICK: …Assume Di-Gesic, so that’s DPP and paracetamol together in the one drug, is being used to treat a patient, or is being used to treat patients with some efficacy. So make that assumption for the minute. Dr Halliwell, is it right to say that your evidence is that a class of patients using Di-Gesic with some efficacy will always be able to find a clinically adequate alternative or alternatives? In a nutshell, is that it?
Dr HALLIWELL: Yes, that’s true.
DR RENWICK: All right. Professor Christie, to the extent you feel able to comment on that clinical matter, do you agree or disagree with that proposition?
PROFESSOR CHRISTIE: I agree that would be the case for most patients. I believe there will still be a subset of patients for whom that’s not true.
DR RENWICK: Right.
PROFESSOR CHRISTIE: And those are patients for whom codeine and tramadol do not work, or non-steroidals – we’re getting to, I believe, a relatively small population of people, but for whom tramadol and codeine are not suitable, or contrary indicated, for whom perhaps even oxycodone is not a suitable alternative, and I make the point that oxycodone carries very much the same risk, for example, of constipation and somnolence that codeine does. It has got exactly the same, sort of, distribution and mechanism of action.
DR RENWICK: I will come back to it – yes, all right.
PROFESSOR CHRISTIE: So, there would still be subgroups of patients that it would be very difficult to find a suitable alternative.
As background to the Tribunal’s consideration of this evidence, reference should be made to other non-contentious evidence before the Tribunal that as a matter of general clinical practice, a step-wise approach should be adopted to pain management. This requires, as a general rule, that strong opioids should not be prescribed where a weak opioid or NSAID would provide sufficient relief from pain, and a strong opioid should not be administered with a weak opioid, either alone or in combination, where an NSAID would be sufficient. In this regard, DPP may be regarded as a weak opioid in a class together with codeine and tramadol, although its advantage is that it is that it is the only weak opioid available for the small number of patients who cannot metabolise codeine or tramadol.
Professor Christie was unable to specify the exact subset of patients who would be left without adequate alternative pain relief if DPP was removed from the market. Pressed on this point by Counsel for the Minister, Professor Christie undertook what he described as a crude and difficult calculation (See Transcript 29.05.12 at p132.21:
PROFESSOR CHRISTIE: As I said, if you take all of the people - you take the 10 per cent who are CYP2D6 insufficient for codeine, for example, and the hyper metabolises together, that might make something - you know 15 per cent of the population but it’s a sloppy figure. Then you add to that people who get a reasonable effect from codeine but it’s intolerable for other reasons so you need an alternative and then to actually intersect that with all of the other alternatives is an extremely complex exercise even from first principles.
Professor Christie agreed with Dr Renwick SC that we simply do not have the evidence to precisely identify the group concerned. Pressed on the point, Professor Christie acknowledged that the number is “a small number in terms of the percentage of the Australian population”, and conceded that it might be “around one (per cent) or less”. (See Transcript 29.05.12 p118.12-17).
Mr Lloyd SC in closing accepted that the group of patients who may not be able to find a suitable alternative could be as small as one per cent of the population, but observed that even one per cent is 200,000 people, and even if it was only a fraction of one per cent it could still be a large number of people in absolute terms. Of course, Mr Lloyd SC disclaimed the suggestion that all of such people would be in pain, so not all would need pain relief, but he maintained that there would still be a significant sub-group of the population (not insignificant numbers) for whom no suitable alternative to DPP would be available.
The Tribunal accepts that the analysis by Professor Christie is persuasive. The rationale for his conclusions was clearly articulated. He readily made appropriate concessions through the concurrent evidence process or cross examination. In contrast, the Tribunal is sceptical of the confident assertions of Dr Halliwell and Professor Buckley that all patients who would be prescribed and who would use DPP would be able to find appropriate analgesic alternatives. The Tribunal accepts, at least for the present, that the weight of evidence, albeit based on calculations not susceptible to producing a numerically defined exact subset, inclines to a conclusion that there is likely to be a group of patients for whom there is unlikely to be an adequate analgesic alternative to DPP for mild to moderate pain. Further, this hypothesis appears to have been accepted by other regulators. For example, the United Kingdom regulator, when withdrawing market authorisation for DPP explicitly stated (See Exhibit R3, Tab 6):
We recognise that there is a small group of patients who are likely to find it very difficult to change from co-proxamol [DPP] or where alternatives appear not to be effective or suitable. For these patients, following cancellation of the licences at the end of 2007 there is a provision for the supply of unlicensed co-proxamol,
Indeed, the decision of the UK regulator included the statement that the manufacturer of Co-proxamol had agreed to continue to supply the drug for that purpose.
Whilst the Tribunal has no doubt that both Professor Buckley and Dr Halliwell were sincere in their testimony, their confident assertion that there will always be an alternative seemed more in keeping with advocacy than dispassionate regard for the probabilities.
INTERNATIONAL REGULATORY APPROACHES
One of the issues the Tribunal raised with the parties early in the proceedings was the status that should be given to other international regulatory decisions to restrict or remove the availability of products containing DPP from the market. Both counsel took the position that the materials relied upon and reasons given by overseas regulators could be relevant to this Tribunal’s consideration, but the making of a decision of any particular kind needed to be understood in the context of varying regulatory and legislative frameworks that could be quite different from those applicable in Australia. The Tribunal has approached its task with those caveats in mind.
There is no doubt that there have been increasing international restrictions on the supply of DPP, particularly following reviews triggered by the MAD study referred to in the discussion of safety above. A recommendation to remove products containing DPP was made by the US regulatory authorities following a 14-12 vote. That vote shows the closeness and division of views within that regulatory authority. The United Kingdom also moved to remove DPP from general availability. The European Union has followed with a similar recommendation, as has New Zealand.
Crucially, however, from the point of view of actual consequences rather than labels, it appears that both in the US and in the UK, products containing DPP can continue to be supplied to patients who are prescribed them. Sales data provided in confidence to the Tribunal show that the products continue to be supplied in those markets. The Tribunal was not taken to the reasons why continuing supply in the US was permissible, but the documents supplied by the Respondent Minister in relation to the British removal of DPP from authorised supply are quite revealing. As mentioned in the discussion above regarding alternatives, when DPP was removed from general availability in Britain, it remained possible for doctors to continue to prescribe it as an unlicensed product if they thought it clinically appropriate to do so. The Tribunal received material dated from March of this year confirming that the MHRA still permitted DPP to be prescribed ‘where the prescriber judges that the medicine is needed to meet the special needs of patients’. The manufacturer undertook to keep the product available so that it could be prescribed on that basis.
While the amount of product containing DPP sold in the UK has fallen substantially since its removal from general use as a result of the regulatory authority’s decision, it still remains prescribed in not insubstantial amounts. It is clear that some UK doctors have continued to prescribe it at least for a small cohort of patients. Dr Renwick SC suggested that that number might be no more than a few thousand patients and submitted, by reason of its lower population, that there would be even fewer Australian patients who would need the drug, such that it could safely be removed from registration with little or no concern because if would affect only a small handful of people. As he put the argument [at Transcript 01.06.12 p64.43]:
Potential implications for a hypothetical few should not trump the safety of an identified overwhelming majority.
BALANCING THE NEEDS OF THE INDIVIDUAL AGAINST THE BENEFITS AT A POPULATION LEVEL
The alternative position that appeals to the Tribunal is that the needs of individuals are not to be subjected to the larger community interest unless they are truly inconsistent. The Tribunal takes the position that a decision-maker seeking to achieve the correct or preferable decision should take the position that if the interests of the few can be reconciled with the interests of the many, then they should be, rather than overridden. In the case of New Zealand, there is evidence that when DPP was removed from the market, those responsible for the administration of the regulatory scheme in New Zealand considered alternatives that would permit such reconciliation, but concluded that none existed.
In the UK, it is plain that the legislation and regulatory framework allowed such reconciliation. Thus, in November 2007, the Medicines and Healthcare products Regulatory Agency (MHRA) issued a Drug Safety Update: Co-proxamol withdrawal: reminder to prescribers (see Exhibit R3, Tab 5):
We recognise that there is a small group of patients who are finding it difficult to change, or where there is an identified clinical need when alternatives seem to be ineffective or unsuitable. For this small group of patients, continued provision of co-proxamol through normal prescribing may continue until the cancellation of the licences at the end of 2007. After this time a provision will remain for the supply of unlicensed co-proxamol on the responsibility of the prescriber, and the Marketing Authorisation holder for the brand leader (Distalgesic) has indicated that they will continue to manufacture co-proxamol to meet clinical demand.
A similar statement was repeated by the MHRA in May 2006. Significantly, it included the following statement [at Exhibit R3, Tab 7] referring specifically to balancing individual rights against those of the majority:
The avoidable death toll from co-proxamol overdose cannot be ignored. Sometimes regulation has to balance the needs of the individual against the benefits at a population level. In this case the removal of marketing authorisations with continued use possible in exceptional circumstances is the best balance that could be achieved.
During the course of the proceedings, the Tribunal tentatively explored with the parties the possibility that the correct or preferable decision might involve a similar reconciliation of individual and public health interests in Australia if the regulatory regime in Australia (unlike New Zealand) allowed such a possibility. Those discussions identified the possibility that if the Minister’s decision was set aside, additional conditions could be imposed under section 28 of the Act which might achieve an analogous outcome to that which was obtained in Britain. That is, Doloxene and Di-Gesic might remain available with the same kind of highly visible warnings and in blister packs designed to minimise the risk of misuse that have been in use since the former President of the Tribunal granted a stay of the decision under review. On this basis, Doloxene and Di-Gesic could be prescribed for patients either finding it difficult to change or where the prescriber judged that the medicine was needed to meet the special needs of patients.
In that regard, it is relevant to note that Mr Lloyd SC, on behalf of Aspen, indicated that Aspen would be prepared to enter into contractual obligations with all pharmacies retailing either Doloxene or Di-Gesic, to the effect that these products would only be supplied on production of a signed statement from the prescriber stating that he or she had taken into account all relevant product information and warnings before prescribing the drug, and had advised the patient of all relevant risks and that the statement would have to be countersigned by the patient as evidence that those warnings had been given.
Other possible mechanisms were also explored by the parties. Dr Renwick SC directed the Tribunal’s attention to possible supply of Doloxene and Di-Gesic under the Special Access Scheme and Authorised Prescriber Scheme, which provide for the supply of an unapproved therapeutic good pursuant to section 19 of the Act. The Tribunal is, however, sceptical that the criteria and processes which authorise the use of unapproved therapeutic goods under section 19 could achieve anything resembling the reconciliation of interests that was achieved in Britain. The Special Access Scheme and Authorised Prescriber Scheme simply do not appear to be relevant to the circumstances the Tribunal has been considering in the proceedings. There was also tentative discussion about possible changes to the listing under the relevant Poisons Standard. However, the Tribunal was advised by Dr Renwick SC that there were issues of State compliance that could make the provisions of Part 6-3 of the Act of insufficient utility from the Minister’s perspective.
It therefore appears to the Tribunal that section 28, rather than any other provision of the Act, may present a means to achieve a sensible reconciliation of the needs of the individual with benefits at a population level. As Mr Lloyd SC pointed out, the population at large is not affected by a product that most will never use or be prescribed. There has already been a large reduction in the use of products containing DPP in Australia since the additional warnings were included as a condition of the stay (the Tribunal received sales data relating to Doloxene and Di-Gesic in confidence). Such evidence as exists from the UK suggests that there is unlikely to be an ongoing problem with large or irresponsible prescribing. The further undertaking by Aspen to enter into contractual arrangements with its retailers would make that even more unlikely, and would ensure that the product was less easy to prescribe in Australia than the UK where it has been formally unlicensed.
OPPORTUNITY FOR FURTHUR CONSIDERATION: POWER TO REMIT
Given the conclusions reached by the Tribunal as to the weight to be given to the various balancing factors and its thinking as to what might constitute the correct or preferable decision, the Tribunal has reached the conclusion that it would be appropriate for it to exercise its authority under section 42D of the Administrative Appeals Tribunal Act 1975. Section 42D provides for the Tribunal, at any stage in the proceedings, to remit the matter to the person who made it for reconsideration. If that were done, it would allow time for both parties to explore whether, in light of the Tribunal’s reasons, they can find an agreed solution rather than having one imposed on them. Assuming orders are made under section 42D, the Minister would then have 28 days (or such other time as specified by the Tribunal) to affirm, vary or set aside the decision and make a new decision in substitution. If Aspen remained aggrieved it would have the right to return to the Tribunal. If it did, the proceedings would continue as a review of any new decision or the re-affirmed decision, as the case may then be.
However, rather than simply announcing its decision, the Tribunal is mindful that both Mr Lloyd SC and Dr Renwick SC requested that they have the opportunity to be heard should the Tribunal be minded to make any decision other than either simply affirming or setting aside the decision on review. The Tribunal will therefore hear counsel as to any views they may have regarding the proposed substantive exercise of its power under section 42D and, if it be exercised, any ancillary orders that should be made.
Finally, the Tribunal expresses its appreciation to both Dr Renwick SC and Mr Lloyd SC for the assistance they have provided to it. The Tribunal and the parties had to deal with a number of practical difficulties during these proceedings, not the least of which was taking concurrent evidence by a sometimes frail video link from London. Only highly competent counsel would have been able to rise to such challenges unruffled, while never failing to ensure that their respective clients’ positions were fully and ably presented to the Tribunal.
I certify that the preceding 105 (one hundred and five) paragraphs are a true copy of the reasons for the decision herein of Justice Kerr, President and Dr Nicoletti, Member.
.................[SGD].........................
Associate
15 June 2012
Dates of hearing 28 May to 1 June 2012 Counsel for the Applicant Mr S Lloyd SC and Mr S Robertson Solicitor for the Applicant
King & Wood Mallesons
Counsel for the Respondent
Dr J Renwick SC and Dr H Bennett
Solicitor for the Respondent Corrs Chambers Westgarth
4
3
0