Updated 31 December 2012

Minutes of the 106th Medicines Adverse Reactions Committee Meeting - 14 June 2001


The material listed as being considered on an issue is not intended to be exhaustive.

Descriptions of unpublished case reports have not been included to protect the privacy of those involved.

Names of individuals have been deleted where the contribution is not in the public domain and will not shortly be so. For example the names of those to be approached to write an article are deleted, but not usually the names of those who have contributed a draft article. Names are not usually deleted when a contribution has been made in an official capacity.

The recommendations of the committee are in bold type face.


Note relevant to these minutes: All of the recommendations of this meeting were accepted by the Delegate of the Minister of Health except the final one (12.9) about requiring all ingredients to be listed in the labelling of topical medicines. The feasibility of this recommendation is being investigated in the face of international harmonisation issues, New Zealand consumer legislation and space constraints on packaging.


Associate Professor TJB Maling (Chairperson)
Dr M Tatley (Medical Assessor)
Dr R Savage (Medical Assessor)
Professor P Ellis
Dr J Goldsmith
Dr S Jessamine
Dr M Rademaker
Professor DCG Skegg
Dr N Rafter (Medical Advisor, Medsafe)
Dr K Ronaldson (Secretary)


Ms S Von Afehlt


Apologies were received from Dr Coulter, Dr McClure and Dr Kingston.


The minutes of the 105th meeting and of the teleconference on thioridazine (held 20 April) were signed by the Chairperson to indicate that they were true and accurate records of the meetings.


The dates 5 September (Wed), 6 December (Thu) were confirmed for the next meetings


4.1 Conflict of interest

- Declaration of competing interests - MARC members
- Form for declaration of competing interests

4.2 Prescriber Update

- Contents of June 2001 issue
- Origin of Prescriber Update articles, designated by logo
- Print copy of article on thioridazine and arrhythmia
- Time line for publication of Prescriber Update
- Authors and peer reviewers for pending articles
- Process for Prescriber Update articles
- Medsafe editorial guidelines for Prescriber Update articles

Ms Von Afehlt advised that she had set up a new system for classifying Prescriber Update articles, involving logos to designate those that were adverse reaction updates and those that were specifically the prescribing advice of the Committee such as the thioridazine article in the June 2001 issue. She also proposed placing a purple triangle containing the words "important please read" at the top of articles containing important prescribing advice.

Ms Von Afehlt pointed out that, for each recommended article, the Committee needs to state clearly the purpose of the article and the key messages to be conveyed. She drew the Committee's attention to the process for Prescriber Update articles recommended by the MARC. In particular, she mentioned that a member should be nominated to peer review each article recommended. This would allow articles to progress independently of committee meetings, when necessary to meet publication deadlines. The Chairperson would sign off each MARC-recommended article before the final review by the author and the sign-off by Medsafe.

Members were also alerted to the long lead-time involved in Prescriber Update articles. There is ordinarily 3 months between the date at which all articles must be submitted and publication of the issue.

Members suggested that in order to improve the standing of Prescriber Update, articles should be peer reviewed by specialists or experts not proposed by the author, and the publication should adopt the volume number, issue number format, rather than the current issue number in series system. With these changes an author of an article would consider that he or she could include the article in a curriculum vitae.

In addition, members suggested that the guidelines should refer to the Vancouver style for referencing and to the guidelines for authors of articles in the New Zealand Medical Journal.

Members also commented that authors need to see peer reviewers' comments, since this will help them to appreciate the need to change the article, and why.

.. advised that she was willing to peer review the article on gastrointestinal haemorrhage with aspirin.

4.3 Point of order - written opinions of absent members

.. raised the fact that his written opinion on thioridazine, presented in his absence from the meeting, had not been either listed in the documents considered or referred to in the discussion. The Chairperson ruled that this omission was appropriate. He considered that even contrary opinions of absent members should not be included in the minutes since if the member had been present for the discussion, the opinion may have changed. The situation would be different if the member had been formally requested to provide a written report.

In summary: the Chairperson ruled that opinions of absent members would not be recorded in minutes unless they had been formally asked to prepare a report on the topic in question.


5.1 Report on the recommendations

- K Ronaldson. Secretary's report to the members of the MARC for the 106th meeting
- Report to the Delegate of the Minister of Health, 23 May 2001
- Recommendations of the 105th meeting of the MARC
- Recommendations tables
- S Gardiner, E Begg. Drug safety in lactation. Prescriber Update June 2001
- Medsafe Editorial Team. Viagra reminder: Ask! Don't by shy! Prescriber Update June 2001
- R Savage. NSAIAs can cause lower GIT damage. Prescriber Update June 2001
- Updated HRT prescribing guideline released. Prescriber Update June 2001

Members noted the articles to be published in the June issue of Prescriber Update.

5.1.1 Minutes on the web site (4.3, Mar 2001)

The Secretary had identified that people accessing the minutes on the web site need to have available to them the causality definitions so they can understand their use in the text. With the agreement of the Committee, the Secretary undertook to ensure that the causality definitions were included when the March minutes were published.

5.1.2 Cross-communication between MARC and PTAC (5.1.2, Mar 2001)

Dr Jessamine and Dr Rafter reported that not a lot of progress had been made negotiating with Pharmac over cross-membership or a suitable alternative arrangement. The Committee asked the Chairperson to speak to the Chief Executive of Pharmac.

5.1.3 Inhaled/intranasal corticosteroids and growth retardation (5.1.3, Mar 2001)

- N Rafter. Growth retardation with inhaled and intranasal corticosteroids. Statement from MARC to be published in Prescriber Update
- Inhaled corticosteroids and growth retardation. Item 12.2, minutes of 7 Dec 2000 meeting of MARC

Members noted Dr Rafter's draft article on inhaled/intranasal corticosteroids and growth retardation. The Committee recommended making the following alterations to the article:

5.1.4 Perioperative use of OCs and HRT (5.2, Mar 2001)

- Use of COCs and HRT in the perioperative period. Letter from the Australian and New Zealand College of Anaesthetists. 28 May 2001
- ... Use of COCs and HRT in the perioperative period. Letter on behalf of the Royal Australasian College of Surgeons. 11 June 2001
- .., Orthopaedic Registrar. Perioperative use of HRT/OCs. Correspondence. 2 Mar, 11 May 2001

The previous meeting had recommended contacting several professional colleges for their comments regarding the perioperative use of combined oral contraceptives (COCs) and hormone replacement therapy (HRT) and publishing a paragraph in Prescriber Update based on the recommendations of the New Zealand Guidelines Group.

The Australian and New Zealand College of Anaesthetists (ANZCA) advised that it has no guidelines or recommendations concerning the use of COCs and HRT in the perioperative period. Because of the practical difficulties around ensuring a woman stops hormone treatment 30 days before surgery and the risk of pregnancy if she is on a COC, the College considered that the emphasis in any guidelines should be on thromboembolic prophylaxis.

The College also recommended contacting for comment the Royal Australian and New Zealand College of Obstetricians and Gynaecologists and the Royal Australas-ian College of Physicians (Haematology subspeciality). The Committee recommended following this advice.

The Royal Australasian College of Surgeons (RACS) had a similar view to the ANZCA regarding the emphasis of the guidelines. Both Colleges were interested in having the opportunity to conduct a prepublication review of the paragraphs to be published in Prescriber Update. The Committee recommended replying to .. of the RACS to comment on the fact that he did not differentiate between COCs and HRT in his advice and to ask him what evidence he had that thromboembolic prophylaxis in women undergoing surgery and taking COCs and HRT would be safe and effective.

The Committee also recommended asking .. to peer review the paragraphs for Prescriber Update. .. said he was willing to be the MARC peer reviewer.

5.1.5 Rofecoxib and cardiovascular events (8.1.1, Mar 2001)

- E Barr. Expert opinion report on the cardiovascular findings of the VIGOR study. Report for MSD. 21 Dec 2000.
- Rofecoxib tablets and oral suspension. Periodic Safety Update Report for 1 Aug 2000 to 31 Jan 2001. Merck Research Laboratories. Excerpts on cardiovascular disease and the overall conclusion.

The Committee did not have the time to discuss this material. Members asked that it be brought back to the next meeting and be included on the agenda as a separate item.

5.1.6 Mirena and benign intracranial hypertension (8.1.1, Mar 2001)

- Mirena and benign intracranial hypertension. Comment from Schering. 4 Aug 2000 and 30 May 2001

As a result of a New Zealand report of benign intracranial hypertension (BIH) in a woman using Mirena, comment had been sought from Schering. The Company had one other possible case. However, while this patient had had BIH, it appeared to have occurred prior to the use of Mirena, and "worsened headache" was considered the correct term for the adverse event which occurred in association with Mirena.

The Committee agreed that there was insufficient information to establish a causal association between Mirena and BIH, and that the matter should not be taken any further at present.

5.1.7 ACC medical misadventure levy (9.2, Mar 2001)

- K Ronaldson. Proposed ACC medical misadventure levy. Indicative comment from the Medicines Adverse Reactions Committee, April 2001
- R Savage. Medical Misadventure Levy. Submission from the Centre for Adverse Reaction Monitoring, April 2001

The Committee noted the comments prepared on behalf of the Committee and CARM concerning the proposed ACC medical misadventure levy. The Secretary reported that she had contacted the Department of Labour in the previous few days and found that there was no move foreseen in the near future to reinitiate investigation of the possibility of an ACC medical misadventure levy.

.. reported that he had discussed the levy with Mr Ron Paterson, Health and Disability Commissioner, who said, having raised the matter, he was not going to pursue it further.

5.1.8 Thioridazine and QT-prolongation (12.1, Mar 2001)

- Medsafe Editorial Team. Thioridazine and arrhythmia: prescribing changes. Prescriber Update June 2001
- Changes to the prescribing information for thioridazine recommended by the Medicines Adverse Reactions Committee, 20 April 2001
- Draft minutes of the teleconference of the MARC concerning thioridazine, 20 April 2001.
- S Evans. Risk of arrhythmias results in Melleril/Aldazine (thioridazine HCl) safety update. Dear Doctor/Pharmacist letter from Novartis (also on behalf of Pacific). 13 June 2000

Members noted the documents on thioridazine and that a combined dear doctor/pharmacist letter from Novartis and Pacific was being sent out to coincide with publication of the June issue of Prescriber Update. An abstract of the article was also being published in New Zealand General Practitioner and Pharmacy Today.

5.1.9 Isotretinoin - safety and usage (12.2, Mar 2001)

- K Ronaldson. The funding and availability of isotretinoin. Letter to Peter Moodie, Medical Director, Pharmac. 8 April 2001
- R Smart/K Ronaldson. Oratane (isotretinoin) and suicide/depression. Correspondence of 8 May and 25 May 2001

Douglas and Roche had been asked to update the information in their patient information for Oratane and Roaccutane, respectively, with further detail on the risk of depression, suicidal ideation and suicide with isotretinoin. Douglas had initially refused, but had agreed following correspondence from the Secretary.

5.1.10 Propofol, excessive dose for prolonged sedation and cardiac failure (12.3, Mar 2001)

- K Ronaldson. Dosage instructions for propofol infusion in prolonged sedation. Letter to .., College of Anaesthetists.
- AstraZeneca letter re dosage instruction for propofol infusion in prolonged sedation. 2 May 2001
- DF Kelly et al. Propofol in the treatment of moderate and severe head injury: a randomised, prospective double-blinded pilot trial. J Neurosurg 90:1042-52, June 1999
- JE Risdall, et al. Propofol and cerebral autoregulation following head injury. J Neurosurgical Anaesthesiology 9:404(Abs 4), 1997
- OL Cremer et al. Long term propofol infusion and cardiac failure in adult head-injured patients. Lancet 357:117-8, 13 Jan 2001
- Propofol. Item 8.1 ADRAC 250th meeting, 16 Feb 2001
- R Freebairn. Propofol sedation for intensive care patients. Comment on behalf of the ANZCOA. 31 May 2001

The previous meeting had recommended that companies marketing products containing propofol in New Zealand be asked to include in the data sheets for their products a maximum infusion rate of 4 mg/kg/h for prolonged sedation (> 24 hours). AstraZeneca had objected, providing with their objection two studies in which high infusion rates (> 5 mg/kg/h) were considered to be advantageous in head injury patients, and had resulted in an index of cerebral autoregulation that is close to 1.0 (i.e. normal).

The Secretary had sought the advice of the Australian and New Zealand College of Anaesthetists, Faculty of Intensive Care. A reply received from the College advised that it was not in favour of being prescriptive about a maximum infusion rate of propofol, since doing so may not improve overall survival. However, the College did consider it appropriate to warn about the potential hazards of prolonged use of propofol at a dose rate > 5 mg/kg/h. Unexplained acidosis or cardiovascular compromise should warrant a review of the sedation regime. The College had sent its advice to intensive care units and intensive care specialists.

The Committee recommended that the advice of the College of Anaesthetists also be sent to neurosurgeons and emergency departments. The Committee also concurred with the suggestion of the College that its advice be distilled into a form suitable for the data sheets of products containing propofol and recommended that the sponsor companies be asked to update their data sheets accordingly.

5.1.11 Dosage of Estrofem, case no. 45141 (, Dec 2000)

- Estrofem (oestradiol) and deep vein thrombosis (update), 45141

A case considered at the meeting in December 2000 was of deep vein thrombosis with Estrofem for hirsutism. The Committee had queried the dosage which was 16mg daily. .. reported that she had received confirmation of this dosage from the reporter who was not the prescriber.

5.1.12 Antipyretic medicine use in children (5.3, Mar 2000)

- ... Is the use of antipyretic medicines necessary or safe in children with fever? Draft article

.. had written an article arguing that use of antipyretics in the treatment of fever in children was unnecessary, and because of the potential for adverse reactions the benefit:risk ratio was unfavourable.

Members agreed that the article was not suitable for Prescriber Update but that it would be appropriate to have it published in the New Zealand Medical Journal where it could become the subject of debate. The Committee suggested that .. submit the article with a covering letter saying that it had arisen out of MARC discussion.

6. General Reporting

- Case reports considered since Nov 1997, by medicine class
- Case reports considered since Nov 1997, by recommendation
- Glossary of terms

All spontaneous reports presented to the MARC meeting have been assessed by CARM and replies have been sent to the reporters. The purpose of these responses is to assist the practitioner to discharge his/her responsibility to patients. These individual replies include as appropriate:

6.1 CARM Quarterly report

- Quarterly reports as at 31 March 2001

During the 3 months to March 2001, CARM received 3 or more reports of adverse events following brand switches for pantoprazole (5), tamoxifen (3), diclofenac (3), verapamil (3) and fluoxetine (15).

The Committee recommended that Medsafe contact Pharmac to add to CARM's request that it pass on to CARM cases of adverse events following brand switch which have been reported to it.

There have been 6 reports of hypotonic hyporesponsive episodes (HHE) with DTaP vaccine. While these cases matched the case definition, they were less severe than cases encountered with whole cell pertussis vaccines. Similarly, 11 cases of abnormal crying with the acellular vaccine were less profound than cases with whole cell pertussis. Dr Tatley advised the Committee that he planned to conduct a critical analysis of the cases of HHE with DtaP vaccine.

Contrast media feature prominently in the list of adverse reactions occurring 3 or more times, because a cumulated batch of these adverse reactions was sent in one package.

Dr Tatley indicated that he would like to present more in the way of trend analysis, but it was difficult to know what trends would be useful to track. It would be good to identify trends that would serve as performance indicators, such as trends showing that the reporting rate was increasing in a specific group, or adverse reaction figures for a particular medicine indicating that a safety message was being observed. Dr Tatley indicated that he would welcome suggestions from members.

6.2 Deaths

Note: These deaths were discussed under the therapeutic group of the medicine involved. Each of the deaths has an associated report number which is also a hyperlink to the comment on that case. To access the comment click on the hyperlink.

Carbamazepine et al and hyponatraemia, cardiac failure, oedema, 46205
Clozapine and sudden death, 46625
Cyproterone acetate and jaundice, myocardial infarction, 46520
Donepezil and angina, cardiac failure, myocardial infarction, 46870
Flecainide and sudden death, 46833
Flutamide and hepatic failure, encephalopathy, 46665
Paroxetine and suicide, 47069

Note: In the comment associated with each report the case has been given a causality designation using terms and definitions developed by the WHO. The precise definitions are available on the web site of the WHO Collaborating Centre These designations, certain, probable, possible, unlikely, unclassified and unclassifiable, refer to the degree of certainty about the relationship between the medicine and the adverse event. The terms should not be understood literally. So "certain" means the appropriate elements are present to match the international definition. It does not mean there is absolute certainty that the medicine caused the adverse event.

Explanations of the terms used by CARM and MARC can be accessed by hyperlink at each causality designation, certain, probable, possible, unlikely, unclassified and unclassifiable.

6.3 Analgesics

6.3.1 Diclofenac and necrotising fasciitis, 46774

Comment: The WHO database holds 6 reports of necrotising fasciitis in association with diclofenac. The causal association for this case was designated "possible"; the open lesion and diabetes were considered to be other risk factors. Serious soft tissue infection with NSAIAs are an adverse reaction of current concern. An article about this possible association was published in the February 2001 issue of Prescriber Update. The Committee recommended not taking any further action.

6.4 Cardiovascular

6.4.1 Flecainide and sudden death (death), 46833

Comment: CARM holds 38 reports of adverse reactions with flecainide including one previous report of sudden death with flecainide, one of arrhythmia resulting in death and a third of cardiac arrest (patient recovered). Members noted that flecainide is a pro-arrhythmic and can cause ventricular tachycardia and ventricular fibrillation. Members considered that the patient may have received an overdose, either intentionally or unintentionally. Because of the deficiency of detail, this case was designated "unclassified". The pharmaceutical company which reported the case had been asked to supply the coroner's report when it becomes available. The Committee recommended that no further action be taken at present.

6.5 Hormones

6.5.1 Cyproterone acetate and jaundice, myocardial infarction (death), 46520

Comment: Hepatotoxicity and hepatic failure are known adverse reactions of cyproterone acetate. CARM holds a total of 9 reports of hepatic reactions with this medicine. Atenolol, cilazapril and frusemide have rarely been associated with liver compromise of varying degrees. The events in the present case were designated "unclassified" because of the lack of information. Further details had been requested from the reporter. The Committee recommended that no additional action be taken.

6.5.2 Diane 35 and venous thromboembolism Diane 35 and DVT, multiple pulmonary emboli, 47130

Comment: The fracture, surgery and the associated immobilisation would have added to the risk of venous thromboembolism (VTE) occurring. The causal relationship was designated "possible" in this case. Diane 35 / cyproterone acetate and DVT, 47208

Comment: Members considered this case should be designated "possible" for Diane 35. Although there was a 6-week period between discontinuation of Diane 35 and the event, the coagulant effect of the oestrogen would persist for several weeks after withdrawal of the medication. Diane 35 and pulmonary embolism, 47246

Comment: CARM had written to the reporter to ask if the patient had thrombophilia or any other risk factors for venous thromboembolism (VTE). The causal relationship with Diane 35 was considered to be "possible". There is not known to be an association between VTE and insulin use. Comment

Diane 35 is known to be associated with VTE. The CARM database holds a total of 11 reports of pulmonary embolism with Diane 35 and 6 of DVT for a total of 15 cases. Internationally, there have been very rare reports of VTE with cyproterone acetate, but a causal association has not been established. VTE with Diane 35 is an adverse reaction of current concern. The Committee recommended that no further action be taken.

6.5.3 Femulen (ethynodiol) and pulmonary embolism, 47230

The WHO database has 3 reports of pulmonary embolism with ethynodiol (total of 190 reactions). Although there appears to be an added risk with high dose progestogen, the increase in risk of VTE with progestogen-only oral contraceptives is very small or zero.1 The association in this case was designated "unlikely", because the woman had a number of risk factors independent of the oral contraceptive use which would have increased the risk of VTE. The Committee recommended taking no additional action related to this case.

Members commented that, according to some of the data sheets, progestogen-only oral contraceptives are contraindicated in women with a history of VTE. When a report of VTE in a woman taking a COC is received by CARM, the reply to the reporter advises that the woman should avoid oestrogen-containing oral contraceptives in future. The woman is then left with few options if progestogen-only pills are contraindicated. It was considered appropriate that CARM continue to recommend avoidance of oestrogen-containing medicines in women who develop VTE.

6.5.4 Flutamide and hepatic failure, encephalopathy (death), 46665

Comment: Hepatic failure is a known adverse reaction of flutamide and 7 reports of hepatic reactions have been recorded in the CARM database and 51 reports of hepatic failure in the WHO database. In this case the adverse events were rated "possible". The Committee recommended that no further action be taken.

6.5.7 Marvelon and pulmonary embolism, 47229

Comment: 13 previous cases of pulmonary embolism and 15 cases of DVT have been reported with Marvelon. Although with a positive antinuclear antibody test there were other possible aetiologies, the association of the woman's pulmonary embolism with Marvelon was designated "possible". VTE with oral contraceptives are an adverse reaction of current concern. The Committee recommended taking no additional action.

6.6 Psychiatric

6.6.1 Carbamazepine and hyponatraemia, cardiac failure, oedema (death), 46205

Comment: Hyponatraemia is a known adverse reaction of carbamazepine, and 8 previous reports of this event have been received by CARM. In this case, members considered that there may have been other causes for the patient's heart failure. Flupenthixol may also have been causal. The causal designation given for each of the events was "unclassified" because of the lack of data. CARM had requested further information from the reporter. The Committee recommended no other action.

6.6.2 Clozapine and constipation, toxic megacolon, septicaemia, 46498

Comment: CARM holds 2 previous reports of constipation and 2 of paralytic ileus with clozapine. The possibility of constipation is mentioned in the data sheet for Clozaril and it recommends that patients be questioned about their bowel habits, because of the risk of complications from delayed diagnosis. Constipation with clozapine was mentioned as a potentially life-threatening adverse reaction in a Prescriber Update article published February 2001. In the present case, it was considered that benztropine may also have contributed to the patient's serious constipation, and the adverse reaction was designated "possible" for clozapine, trifluoperazine and benztropine. Members considered that the risk of constipation and complications as a result was adequately dealt with in the data sheet and the article was a suitable alert for medical practitioners. The Committee recommended taking no additional action.

6.6.3 Clozapine and sudden death (death), 46625

Comment: CARM had requested further details, including post mortem report and relevant history, from the reporter. CARM has received 3 previous reports of sudden death with clozapine. The present case was designated "unclassified" because of the lack of detail provided. The Committee recommended that no further action be taken.

6.6.4 Clozapine / sodium valproate and hypersensitivity syndrome, drug interaction, 46804

Comment: The CARM database holds a total of 182 reports with clozapine including 6 previous reports of rash, 3 of hypotension, 13 of fever, 1 of diarrhoea, and 6 of cardiac rhythm disorders. Members observed that the time to onset (14 days) was consistent with a hypersensitivity reaction to clozapine. The events associated with hypersensitivity were designated "probable". The sodium valproate may have raised the clozapine levels, and the interaction was designated "possible". The Committee recommended that no further action be taken.

6.6.5 Donepezil and angina, cardiac failure, myocardial infarction (death), 46870

Comment: The WHO database holds 26 reports of myocardial infarction with donepezil. Bradycardia and syncope are listed in the data sheet. In the present case the events were designated "unlikely" and the death was said to be unrelated to the medicine, because of the presence of atherosclerotic coronary vascular disease. The Committee recommended taking no further action.

6.6.6 Migril/diclofenac and ectromelia one limb, skeletal malformation, 46824

Comment: Neither ergotamine nor diclofenac are thought to cause problems in the first trimester of pregnancy according to the booklet of the Australian Drug Evaluation Committee Prescribing Medicines in Pregnancy. Members commented that asymmetrical anomalies, as in the present case, are odd and unusual. It was agreed that the causality designation for ergotamine and diclofenac should be "possible". The Committee recommended that no additional action be taken.

6.6.7 Paroxetine and suicide (death), 47069

Comment: The CARM database holds one previous case of suicide and one of attempted suicide with paroxetine. The event was designated "possible" for causation. See the discussion on 12.7 and the recommendation under that item.

6.7 Urological

6.7.1 Sildenafil and myocardial infarction, chest pain, 47235

Comment: This is the first case of myocardial infarction following sildenafil reported to CARM. Of 17168 adverse reactions with sildenafil in the WHO database, 513 are myocardial infarction. The data sheet for Viagra states that of patients experiencing myocardial infarction after sildenafil, most if not all, have had cardiovascular risk factors. The association in the present case was designated "possible." It was noted that the June 2001 issue of Prescriber Update includes an article warning against the use of nitrates in patients experiencing an angina attack after taking sildenafil. The Committee recommended that no further action be taken at present.

6.8 Alternative medicines

- Table of case reports for alternative therapies

6.8.1 Nature Bees and diarrhoea, diarrhoea bloody, 46809

Comment: Diarrhoea has previously been reported to CARM with bees wax pills, but not with any other bee products. In the present case the diarrhoea was designated "certain" because of the positive results of rechallenge. The bloody diarrhoea was denoted "probable". The Committee recommended taking no further action.

6.8.2 Thompson's Restful Sleep and INR increased, 46632

Comment: This is the first report CARM has received in association with this product. A literature search in herbal textbooks yielded no information on bleeding changes with any of the components of Thompson's Restful Sleep. There may alternatively have been interference with the metabolism of warfarin. The increase in INR was designated "probable". CARM had written to the reporter to ask for further information. The Committee recommended no further action.

7. Vaccines

- Vaccine adverse reaction reports reviewed by MARC since Nov 97

7.1 New Zealand adverse reaction reports

7.1.1 Hepatitis B vaccine and thrombocytopenic purpura, GI haemorrhage, melaena, haematemesis, epistaxis, 46471

The CARM database holds 3 other reports of thrombocytopenia /thrombocytopenic purpura with hepatitis B vaccine. The WHO database has 45 reports of thrombocytopenic purpura. In the present case the causality of the reaction was considered to be "probable".

It was noted that of the varieties of hepatitis B vaccine available in New Zealand, the data sheet for Twinrix lists thrombocytopenic purpura as an adverse reaction, but those for Engerix B and Infanrix-Hep B do not. A member drew to the Committee's attention a letter that had been circulated in 1988 by Professor IR Edwards who was then Medical Assessor asking for reports of thrombocytopenia with hepatitis B vaccine.

Members noted that one previous case of thrombocytopenic purpura with hepatitis B vaccine and 2 cases with MMR vaccine (see 7.2.2) had been designated "unlikely" for causality in the committee database for previously considered adverse reactions.

The Committee asked that the designations for these cases be changed to "probable" in both the committee and CARM databases.

The Committee also asked that thrombocytopenic purpura with hepatitis B vaccine be put on the agenda of the next meeting, with material to be considered including the previous case reports, published articles and data from the pharmaceutical companies.

7.2 Articles from the literature

- K Ronaldson. Vaccine adverse reactions. Report for MARC, 8 June 2001
- Extract from MMR II vaccine data sheet

7.2.1 Hepatitis B vaccine and demyelinating disease

- T Jefferson, H Heijbel. Demyelinating disease and Hepatitis B vaccination. Drug Safety 24:249-254, 2001
- A Ascherio, et al. Hepatitis B vaccination and the risk of multiple sclerosis. NEJM 344:327-332, 1 Feb 2001
- C Confavreux, S Suissa et al. Vaccinations and the risk of relapse in multiple sclerosis. NEJM 344:319-26, 1 Feb 2001
- BG Gellin, W Schaffner. The risk of vaccination(the importance of "negative" studies. NEJM 344:372-3, 1 Feb 2001

Because of a suggested link between multiple sclerosis and hepatitis B vaccine (evidence: several cases developing within a few weeks of immunisation; and 2 case-control studies reporting a non-significant increase in risk of multiple sclerosis, and published only in abstract form), the French government temporarily suspended the hepatitis B vaccination programme in schools (Jefferson and Heijbel).

Hepatitis B vaccine is one of the least reactogenic vaccines, and has not been associated with any major adverse reaction scares in New Zealand. It provides protection against a disease which is fatal in around 20% of carriers (350 million worldwide). Up to 31 March 2001 there have been 444 reports of adverse reactions (thrombocytopenia and arthralgia, etc have both featured) associated with hepatitis B vaccine in New Zealand.

A study published in 2000 and conducted in adolescents in British Columbia found no association between hepatitis B vaccine and multiple sclerosis.2 The study compared incidence before the vaccination programme was initiated with that after initiation. (7.2.3, Mar 2000)

A further study (Ascherio et al) has examined the possibility of an association between hepatitis B vaccine and multiple sclerosis, and another (Confavreux et al) between vaccination with any vaccine (including hepatitis B) and relapse of multiple sclerosis. Neither found any association.

Ascherio et al conducted a case-control study using the US Nurses' Health Study I and II, including > 230,000 women in the total cohort. Cases of multiple sclerosis were identified through the biannual questionnaires sent to the women included in the study. Diagnoses were verified by a neurologist in 93% of cases. The final number of cases included in the study was 192 women with multiple sclerosis, matched to 534 healthy controls and to 111 women with breast cancer. Women reporting never having been vaccinated were considered unvaccinated. For those who reported being vaccinated, vaccination records were checked.

The age-adjusted relative risk of multiple sclerosis in vaccinated versus unvaccinated women was 0.9 (95% CI 0.5-1.6) using healthy controls and 1.2 (0.5-2.9) using the controls with breast cancer. The corresponding figures for women vaccinated within two years before the index date were 0.7 (0.3-1.7) and 1.0 (0.3-1.7). Adjustment for other covariates made little difference to the relative risks.

The authors concluded that their results do not demonstrate an association between hepatitis B vaccine and multiple sclerosis.

The second study conducted in Europe looked at the risk of relapse following vaccination with any vaccine in patients with multiple sclerosis. 643 patients were selected. Patients had experienced at least one relapse in the study period, and the index relapse was confirmed by a neurologist (77% of relapses were rated definite) and preceded by a relapse-free period of at least 12 months. They were interviewed for history of vaccination and information about reported vaccination was confirmed on the basis of medical documentation. Non-vaccination was also confirmed in a sample of participants and no vaccinations were discovered in this group. Patients served as their own controls, with exposure in the interval immediately preceding the relapse being compared with exposure to vaccination in the earlier intervals in the 12-month period preceding the relapse.

The proportions of patients vaccinated in any of the four 2-month control periods were between 2.8 and 4.0% and the proportion in the 2-month period prior to the index relapse was 2.3%. The relative risk of relapse associated with vaccination with any vaccine was 0.71 (0.40-1.26). The specific relative risk for hepatitis B was 0.67 (0.20-2.17). The relative risk remained stable in each of the following instances: if the risk period prior to relapse was changed (1 or 3 months); if only confirmed vaccinations were included; and if the analysis was restricted to patients with a definite relapse.

The authors concluded that their results were consistent with vaccination not being associated with relapse in multiple sclerosis patients who have been relapse-free for the previous 12 months.

The review by Jefferson and Heijbel listed data requirements for better evaluation of the safety of vaccines.

Members noted that for all analyses in these two studies the 95% confidence included 1.0 meaning that they provided evidence for neither a causal nor protective effect of hepatitis B vaccine on multiple sclerosis or relapse thereof. Both studies were very well designed with sufficiently large numbers of cases, diagnoses confirmed by neurologists to strict criteria and vaccination status also confirmed. In the study by Confavreux et al patients serving as self-controls reduced bias and confounding and using different at-risk intervals added weight to the results.

7.2.2 MMR vaccine and idiopathic thrombocytopenic purpura

- E Miller, et al. Idiopathic thrombocytopenic purpura and MMR vaccine. Arch Dis Child 84:227-229, 2001
- V Vlacha, et al. Recurrent thrombocytopenic purpura after repeated measles-mumps-rubella vaccination. Paediatrics 97:738-9, May 1996
- AP Jonville-Béra, et al. Thrombocytopenic purpura after measles, mumps and rubella vaccination: a retrospective survey by the French Regional Pharmacovigilance Centres and Pasteur-Mérieux Sérums et Vaccins. Pediatr Infect Dis J 15:44-8, 1996

Acute idiopathic thrombocytopenic purpura (ITP) is a childhood disorder occurring at an annual incidence of 3-4 cases per 100,000 children aged < 15 years and characterised by an increase in platelet production and decrease in platelet lifespan. ITP has an immunological cause and 60-80% of cases occur within 1-3 weeks of a viral infection. It is also known to occur within 6 weeks of vaccination with MMR vaccine or measles or rubella vaccine together or separately. It was recognised in the Institute of Medicine review to be an adverse reaction of MMR vaccine.

The CARM database holds 2 reports of thrombocytopenia with MMR vaccine.

Miller et al conducted a study of the association between MMR vaccine and ITP with a total number of doses of MMR vaccine administered to the cohort during the study period of 290,300. The cases were children aged 12-23 months admitted to hospitals in the South East Thames region with an ICD 9 discharge code of 278.3.

The study included 35 children with MMR vaccination histories with a total 44 admissions for ITP in the second year of life. 13 of the episodes occurred in the 6 weeks following MMR vaccination. The control period was that period outside of the at-risk period (6 weeks following vaccination) during the second year of life. A relative risk of 3.27 (1.49-7.16) was calculated for the at-risk period. Nine of the 13 cases were considered to be attributable to MMR vaccine; the other 4 were considered to be background cases. The highest rate of onset in the 6-week post-vaccination period was found to be between 15 and 28 days. The study calculated an attributable risk of 1 in 32,000.

Of the 14 children who had an episode of ITP prior to vaccination, none had a repeat episode in the six weeks following vaccination, although three of these had a further episode unrelated to vaccination. The vaccine-associated cases tended to be milder with a shorter duration of hospital stay and higher platelet count than those not associated with vaccination.

Vlacha et al published a case of a 9-year-old boy who developed a repeat episode of ITP 23 days after a third dose of MMR vaccine. After the first dose (12 months of age) he had apparently experienced no adverse reaction. After the second dose at 4 years ITP developed 19 days post dose. The minimum platelet concentration was 10,000/mm³ on the first occurrence of ITP and 9000/mm³ on the second.

No systematic study has been conducted of the consequences of repeating MMR vaccination after ITP has followed an earlier dose.

Jonville-Béra et al conducted a study of cases of ITP reported to the French pharmacovigilance centres and occurring following one of the seven available measles, mumps and/or rubella vaccines. Sixty cases (age range 12-72 months) of ITP were reported during the 8-year study period. The symptoms occurred 2-45 days after vaccination. The most frequent sign found in 57 cases were petechiae and purpura. At the time of diagnosis mean platelet count was 8000/mm³ (range 400-70,000). The immediate outcome was provided in 57 cases, and 51 of these recovered completely, 2 had incomplete normalisation and 4 experienced a relapse. The incidence calculated in this study was 0.95/100,000 doses of MMR vaccine. The authors comment that no deaths have been reported after vaccine-associated ITP. They concluded that the risk of ITP does not modify the benefit:risk ratio for MMR vaccine.

Members noted the evidence in the Miller study for a causal association between MMR vaccine and ITP with an attributable rate of about twice the background rate. Members also noted the evidence that cases associated with vaccination were generally less serious than those triggered by other causes. Measles and rubella infection may also trigger ITP, and do so at a higher rate than the vaccine. Furthermore, measles, mumps and rubella infections are associated with other more serious consequences in rare cases. It was also observed that it appeared to be safe to administer MMR vaccine to infants who have previously developed non-vaccine-related ITP.

With the annual New Zealand birth cohort of around 50,000 children, no more than 2 cases of ITP would be expected in association with vaccination of the entire cohort.

The Committee recommended asking the sponsor company of Triviraten to include mention of ITP in the data sheet.

7.2.3 MMR vaccine and autism

- L Dales, et al. Time trends in autism and in MMR immunization coverage in California. JAMA 285:1183-1185, March 2001
- M Edwards, et al, L Dales et al. MMR immunization and autism. Corres-pondence. JAMA 285:2852-3, 13 June 2001

Dales et al conducted a time trend analysis of uptake of MMR vaccine and diagnosis of autism in California. This analysis is similar to, but not as detailed as, that conducted by Kaye et al using United Kingdom data and considered at the March 2001 MARC meeting.4

The Californian study used data from the annual survey of vaccination uptake by the Californian Department of Health Services (a sample of children surveyed), and statistics for diagnosis of autism from the 21 regional centres for persons with developmental difficulties. The authors found that the uptake of MMR vaccine increased by a small amount for the 1988 birth cohort from about 75 to 82% at age 24 months, followed by a plateau for subsequent birth cohorts (increase over the period 1980-1994 was from 72% to 82%), but the number of cases of autism rose steeply from 176 in the 1980 birth cohort to 1182 cases in the 1994 cohort (572%). If the increase was measured as a rate per 100,000 births the relative increase was 373%.

The authors' conclusion was similar to that of Kaye, et al: The increase in MMR immunisation rate among young children could not account for the increase in the rate of diagnosis of autism in the same group.

The group headed by Wakefield et al5 who first postulated the association between MMR vaccine and autism had claimed that the trend in California was in support of the association.

The diagnosis of autism is also on the increase in New Zealand. The Committee did not have available to it data which would permit a comparison of the trends for uptake of MMR vaccine and diagnosis of autism.

7.2.4 MMR vaccine and Guillain-Barré syndrome

- A Patja, et al. Risk of Guillain-Barré syndrome after measles-mumps-rubella vaccination. J Pediatrics 138:250-254, February 2001

Patja, et al conducted a study in Finland in which they identified all the hospitalised cases of polyradiculitis during a 4-year period. The cases were reviewed by a neurologist to exclude inappropriately coded cases. The dates of MMR immunisation of cases were obtained from the patients' personal vaccination cards. 20 evaluable children and 4 evaluable male patients aged 21-25 years were identified. Four of the patients had not received MMR vaccination at the time of hospitalisation for GBS. None of the patients had received MMR vaccine in the 6 weeks prior to onset of GBS (6 weeks is the usually recognised post-vaccination at-risk period). The shortest time lapse was 80 days. Six individuals received their first dose of MMR vaccine and 11 received a repeat dose after the illness without relapse.

The authors concluded that their study found no increased risk of GBS after MMR vaccination in children.

Members observed that the assumption in this study that all cases of GBS would be hospitalised might not have been valid.

7.2.5 Childhood vaccination and SIDS

- P J Fleming, et al. The UK accelerated immunisation programme and sudden unexpected death in infancy: case-control study. BMJ papers 322:822-825, April 2001

The United Kingdom study by Fleming et al aimed to include all cases of sudden unexpected death (explained and unexplained) in infants aged 1 week to 1 year from a study population of 17.7 million (470,000 births in the study period). These deaths were matched for age, locality and time of sleep to four controls. Parental interviews were conducted and a "reference sleep" was identified to be the subject of the interview for controls. The interview was conducted within a week of the death. The reference sleep occurred in the period of the day in which the index baby had died and was in the 24 hour period prior to the interview. Of the evaluable cases, 303 deaths were attributed to SIDS and 65 to explained causes.

The immunisation programme was DTPH vaccine and oral polio vaccine given at 2, 3 and 4 months.

Just under half of the SIDS infants had begun or completed the immunisation programme compared with two thirds of the controls: odds ratio 0.48 (0.37-0.63). The significance of this odds ratio was increased when adjusted for matching (0.23 (0.14-0.37)). However, the difference became non-significant when adjustment was made for the infants' sleeping environment (0.67 (0.31-1.43)). Of the infants who died of explained causes, 54% had begun or completed the immunisation programme, compared with 61% of the controls.

5% of the vaccinated cases had died within 48 hours of immunisation and 5% of the vaccinated controls had had their reference sleep within 48 hours of immunisation. The corresponding figures for the two week period were 21% and 27%.

Symptoms shown by the babies who died of SIDS suggested that 21% required medical advice in the 24 hours before death, compared with 7% of the controls. The need for medical attention was greater among those who died of explained causes (65%).

The authors concluded that the immunisation programme in the United Kingdom is not associated with sudden unexpected death in infancy whether the death is explained or unexplained.

Members commented that the results of this study should not be interpreted as meaning that DTPH and oral polio vaccines afforded protection against SIDS. Nevertheless, the results were consistent with there being no association.

7.2.6 Influenza vaccine and myelopathy

- A J Larner, S F Farmer. Myelopathy following influenza vaccination in inflammatory CNS disorder treated with chronic immunosuppression. Neurology 7:731-3, Nov 2000. (Reactions abstract)

This abstract described a case of myelitis involving right-sided weakness and difficulty with walking following influenza vaccination. The 42-year-old patient had a history of inflammatory CNS disease.

7.2.7 Rubella vaccine in pregnancy

- D Josefson. Rubella vaccine may be safe in early pregnancy. BMJ 322:695, March 2001

This article was a news item describing a retrospective study in which pregnancy outcome was compared in 94 pregnancies inadvertently exposed to rubella vaccine and 94 not so exposed. The rate of foetal anomalies was said to be similar for the two groups and hearing test results were said to be equivalent.

Members observed that even if the results of this study were accurately presented, they were not grounds for removing the contraindication on the administration of rubella vaccine in pregnancy, but if accurately presented they did provide reason not to automatically recommend termination in exposed cases.

7.2.8 Comment and recommendation

Members considered the studies under 7.2.1 to 7.2.5 to be very helpful in elucidating the safety of vaccination with regard to certain adverse effects with various vaccines. The Committee recommended summarising the conclusions of these studies and other studies on vaccination recently considered by the Committee in a Prescriber Update article. .. agreed to peer review the article for the MARC, and members suggested that .. be approached to be an external peer reviewer.

Members discussed the fact that there is no immunisation database in New Zealand, and no grand plan for assessing and dealing with vaccination safety issues. Dr Rafter advised that she and Dr Jessamine were attending a meeting with interested groups (including IMAC) on vaccination on Thursday, 21 June. Dr Tatley said he was also attending the meeting. He had been asked to do a presentation on CARM and its proposals for the intensive monitoring of new vaccines. It was to be expected that some of these issues would be covered.

The Committee asked the Chairperson to write to the Minister and the Director-General of Health expressing concern at the absence of an immunisation database and asking each to support the set up of such a database and a coordinated approach to the safety monitoring of vaccines. The content of the letter would depend on the outcome of the above immunisation meeting.


9. Other NZ activities

9.1 Medical Error

- C Vincent, et al. Adverse events in British hospitals: preliminary retrospective record review. BMJ 322:517-519, March 2001
- K G M M Alberti. Medical errors: a common problem. BMJ 322:501-502, March 2001
- P Davis, et al. Adverse Events Regional Feasibility Study: methodological results. NZMJ 114:200-2, 11 May 2001
- P Davis, et al. Adverse Events Regional Feasibility Study: indicative findings. NZMJ 114:203-5, 11 May 2001

The British pilot study found that 10.8% of 110 hospitalised patients experienced an adverse event. About half of these were judged preventable. The paper did not indicate how preventability was determined. The review by Alberti recommended prescribing databases which provide warnings if incompatible medicines are prescribed as a means of reducing medicine-related error.

Members asked that the study conducted in Birmingham and authored by Nightingale et al be provided for the next meeting (BMJ 320:750-3, 2000).

The New Zealand study by Davis et al found that 10.7% of 1326 cases from screened medical records documented adverse effects arising as a consequence of health care management. In 60% of adverse effects preventability was low or non-existent. 20.4% of the adverse effects were medicine-related. These were more likely to have occurred out of hospital and to result in acute admission.

The Committee asked that the Chairperson write to the Minister and Director-General of Health expressing the importance of maintaining a no-blame culture and establishing a national system for the reporting of medical errors and adverse events occurring in hospitals or resulting in hospital admissions so that preventative measures can then be introduced.

9.2 Consumer reporting of adverse reactions

- Consensus document adopted at the First International Conference on Consumer Reports on Medicines. September 29 - October 1, 2000. Sigtuna, Sweden
- A Consumer adverse drug event reporting system for Australia. Briefing for the Australian Council for Safety and Quality in Health Care. 2001

Members were provided with a consensus document on consumer reporting of adverse reactions prepared following a significant conference with high-level participation by regulatory authorities, pharmaceutical companies and the World Health Organisation as well as representatives of the consumer movement.

The report argued that much additional information, including earlier accumulation of signals, can be obtained from consumer reporting. Members commented, however, that consumers reporting adverse events are having to deal with the personal consequences as well as supply an objective description of the event. The consensus document proposed that consumer reports be handled by a separate infrastructure than reports by medical practitioners and other health professionals. It proposed a consumer watchdog type role as part of the function. Members considered this emphasis would not assist the scientific goal of adverse reaction reporting.

Members noted that consumers may report to CARM adverse reactions to medicines in New Zealand. The view of CARM and of the Committee is that it is best that consumers report adverse reactions to their health professional, since professional attention, advice or a change in medication may be required, and the professional is better qualified to report the event. However, in circumstances under which the consumer or caregiver is reluctant to refer to the health professional or the health professional does not wish to report the event, consumers are free to report.

The Committee recommended writing to Mr David Russell of the Consumers' Institute, including the above comments of the Committee and:

9.3 Harmonisation with Australia

- Report on stakeholder comment. Consultation paper between Medsafe, Ministry of Health New Zealand and the Australian Regulatory Reform Taskforce, October 2000

9.4 Adverse reaction reporting at Dunedin Hospital

- K Chuang, et al. A lot of difficulties in reporting drug side-effects. Adverse drug reactions reporting in Dunedin Hospital. Trainee Intern Health Care Evaluation Project. Department of Preventive and Social Medicine, Otago Medical School. April 2001

Several trainee interns had conducted interviews of 90 doctors and 5 pharmacists at Dunedin Hospital regarding adverse reaction reporting. Around 50% of doctors had not heard of CARM. The reasons for not reporting adverse reactions were not enough time (66%), uncertainty about diagnosis (61%), uncertainty about reporting criteria (51%), unavailability of reporting cards (48%) and uncertainty about reporting process (45%).

Members commented that the results of the survey reinforce the need for ongoing promotion of CARM and its activities in the hospital environment where there is constant staff mobility. Dr Tatley advised that CARM will also continue to consider new approaches and strategies aimed at maintaining awareness and in particular will be discussing the findings with Dunedin Hospital Management in an attempt to develop a joint plan to improve local awareness and reporting rates which could have wider application in other settings.

10 Other countries

10.1 Australia

- ADRAC review of case reports of anaphylaxis with royal jelly, May 2001

10.1.1 Australian review of causality assessment

Members mentioned the review of causality assessment conducted by the Australian Adverse Drug Reactions Advisory Committee. It was suggested that it might be appropriate to conduct a similar review in New Zealand. Members agreed that CARM staff should decide whether they would like it reviewed. The Committee asked CARM staff to consider this and bring suitable material to a subsequent MARC meeting.

10.2 Canada

- Canadian Adv Drug React Newsletter 11(2), April 2001

10.3 United Kingdom

- Current Problems in Pharmacovigilance 27(1), Feb 2001

11. Publications

- DM Coulter. The development of prescription event monitoring in the New Zealand Intensive Medicines Monitoring Programme. Chapter for a book edited by RD Mann, and a colleague on pharmacovigilance.

Members found the chapter a very helpful and succinct description of the IMMP.

12 Current topics

12.1 The safety of bupropion

- M Tatley. Zyban (bupropion hydrochoride) - A review of safety issues in its use in smoking cessation. Report for MARC. 29 May 2001
- Bupropion. CARM Type B report, 25 May 2001
- Bupropion. A review of its use in the management of smoking cessation. Drugs 59:1007-24, Apr 2000
- A Breckenridge, Chair of CSM. Important safety message: Safety of Zyban. 26 Apr 2001
- Zyban safety update, MCA. 31 Mar 2001
- ADRAC Update on bupropion (Zyban). Article for next issue of ADRAC bulletin
- Deaths from myocardial infarction in Australia in 1997 compared with deaths from all causes reported with bupropion
- Summary details of Zyban fatal outcomes, Australian data
- Zyban update. Comment provided by GlaxoSmithKline to doctors with anti-smoking interests and to their legal advisers. Apr or May 2001
- Zyban safety. Letter from GSK to the editor, NZMJ prepublication copy.
- Zyban data sheet
- Documents from the UK
- Documents submitted by GSK to the TGA

In his report Dr Tatley noted that up to 29 May 2001, 59 adverse reactions with bupropion had been reported to CARM. Three of these have been serious, including 2 cases of convulsions and 1 of myocardial infarction. In addition, a death by suicide in a patient with a history of alcohol abuse and depression treated with fluoxetine at the time of death had been reported to the sponsor. The cases of convulsions do not appear to have been associated with risk factors, but the patient having myocardial infarction had a history of coronary artery disease.

ADRAC had received 780 reports (at 11 May) and the CSM 4986 (at 31 Mar). The profile of reports (i.e. types of events being reported) is similar in each of the three countries. There have been 35 reports of death in the UK and 9 in Australia. Whether bupropion was a causative factor in these deaths is not known. ADRAC has commented that there is no factor in all or most of the deaths that would point to a common cause. As stated in the ADRAC article: "The death of a patient may be caused by a drug or may be coincidental."

According to the review article by Holm and Spencer, sustained bupropion for smoking cessation was very well tolerated in the clinical trials with only insomnia and dry mouth occurring more frequently with bupropion than with placebo.

107 reports of seizure with bupropion have been reported in the UK (0.1% of users). In approximately half of these the patient had either a past history of seizure or risk factors for seizure. Bupropion is contraindicated in both of these groups of patients.

The UK has recently changed the day for increasing the daily dose from 150mg to 300mg (150mg twice daily) from day 4 to day 7. The reason for this change is not known, but it may be to reduce the risk of seizures. GlaxoSmithKline (GSK) has produced time to onset analysis of cases of seizure with bupropion. Of 545 reports, time to onset data has been supplied for 238 cases. While rates are generally highest in the first week of use (daily range first week 9 to 23, mean 14; mean for second week 9), with 23 cases on day 1 and 18 on day 4. The median time to onset was 9 days, with a range of day 1 to > day 28 (14.7% occurred after day 28). From these data, there appears to be no reason to change the day for increasing the dose in order to reduce the risk of seizure.

The direct to consumer advertising (on television) was discussed and it was noted that the advertisement states that Zyban (bupropion) is a prescription medicine, is effective smoking cessation therapy and is not suitable for all individuals.

Members observed that the fact that smoking is a major underlying cause of death in New Zealand needs to be taken into considered when evaluating the risks versus benefits of bupropion. Further, there is no evidence to date that any of the deaths have been caused by bupropion.

Members noted that GSK is updating the New Zealand data sheet to strengthen the warnings and precautions.

The Committee asked the Chairperson to write to Professor Breckenridge, Chairperson of the UK CSM to ask him why the day for increasing the dose had been changed from day 4 to day 7.

The Committee recommended considering whether to request the same change to the day for increasing the dose in New Zealand following receipt of the reply from Professor Breckenridge.

The Committee recommended that:

The Committee asked that the watching brief placed on bupropion at the meeting in March 2001 be continued.

12.2 COX-2 inhibitors

12.2.1 Warfarin-COX-2 inhibitor interaction

- K Ronaldson. Interaction between warfarin and COX-2 inhibitors
- Interaction of celecoxib and warfarin. Aust Adv Drug Reactions Bull Feb 2001
- DC O'Donnell, JS Hooper. Increased international normalised ratio in a patient taking warfarin and celecoxib. J Pharmacy Technology 17:3-5, Jan-Feb 2001. (Reactions abstract)
- Celebrex and Vioxx data sheet extracts

Celecoxib-warfarin interaction
At 18 May 2001, ADRAC had 37 reports of suspected interaction between warfarin and celecoxib. 16 of these cases described an increase in INR, 3 a decrease in prothrombin activity and 1 an increase in warfarin levels without any clinical event. The other 17 reports listed a range of bleeding events including purpura (2), melaena, haematuria (2), haematemesis, haemorrhage NOS (3), rectal haemorrhage, intracranial haemorrhage and gastrointestinal haemorrhage (4). In several of these cases (both short and long onset cases) the haemorrhagic event was linked to a rise in INR.

Of the 24 of the cases for which the duration to onset was provided, 16 patients showed evidence of the interaction within 2 weeks. In 3 patients the evidence was detected within 2-3 days. Onset took more than 2 months in 6 cases. None of the patients died.

Most of the patients were aged > 60 years, but 6 were aged 40-60 years and 1 was aged less than 40 years.

One report clearly traced the effect of adding celecoxib to the regimen of a 38-year-old woman stabilised on warfarin. Prior to commencing celecoxib the woman's INR had been 1.8, and three days after celecoxib was started the INR had risen to 4.2. Celecoxib and warfarin were discontinued on day 7 when the INR was 6.4. Two days later the INR was in the therapeutic range at 2.4. Warfarin was recommenced 9 days after it was discontinued, when the INR was 2.7.

O'Donnell and Hooper described a case in which the dose of warfarin was reduced in order to address the problem. In this case the woman's INR rose from a stable 2-3 to 3.5 three weeks after commencing celecoxib. The woman's weekly warfarin dose was reduced from 26.25 mg/week to 23.75 mg/week.

CARM has received one case of possible interaction between celecoxib and warfarin in a 59-year-old man. He developed haematuria with an INR of 3.1. The pre-celecoxib INR was not stated.

Rofecoxib-warfarin interaction
The ADRAC database also holds four reports of interaction between rofecoxib and warfarin: cerebral haemorrhage (fatal); epistaxis, bleeding gums and purpura (INR >11); and two cases on increased INR (to 3.8 in one case) without any resultant clinical event.

Comment and recommendations
Members agreed that although most of these cases were not well documented, these data indicate that celecoxib may increase the INR in patients taking warfarin. The evidence is less clear for an interaction between rofecoxib and warfarin, but there probably is a class effect. The mechanism of this interaction is not certain, but it may occur predominantly through celecoxib inhibiting the metabolism of warfarin by CYP 2C9.

The data sheets for Celebrex and Vioxx mention interaction with warfarin as a possibility. The Celebrex data sheet cites a study of healthy subjects in whom no interaction was measured. Members agreed that this statement was unfortunate because it has the effect of suggesting that interaction is not a problem. The Committee recommended asking the sponsor of Celebrex to remove the statement about the study in healthy volunteers, and agreed that otherwise the statements in the data sheets were acceptable.

The Committee also recommended:

12.2.2 Rofecoxib and gastrointestinal effects

- D J Watson, et al. Gastrointestinal tolerability of the selective cyclooxygenase-2 (COX-2) inhibitor rofecoxib compared with nonselective COX-1 and COX-2 inhibitors in osteoarthritis. Arch Intern Med 160:2998-3003, October 2000

Watson et al conducted an analysis of 8 studies comparing rofecoxib with ibuprofen, diclofenac or nabutamone in the treatment of osteoarthritis. They found that the number of treatment discontinuations due to gastrointestinal symptoms during 12 months was lower with rofecoxib than with the comparative NSAIA (8.2 vs 12.0 per 100 patient-years; relative risk 0.70, 95% CI 0.52-0.94). The study was funded by Merck.

Members commented that the differences in risk of gastrointestinal effects found with rofecoxib (i.e. the protective effect) compared with NSAIAs in this study and in other studies with rofecoxib and celecoxib are not as great as would be expected from the marketing of the COX-2 inhibitors.

The Committee noted that .. is preparing an article for publication in Prescriber Update on the adverse effects profile of the selective COX-2 inhibitors.

12.2.3 Celecoxib in aspirin intolerance

- B Dahlén, et al. Celecoxib in patients with asthma and aspirin intolerance. N Engl J Med 344:142, Jan 2001

Dahlén et al performed a study in which 27 patients with asthma and aspirin intolerance were challenged with celecoxib. In no case did the challenge induce bronchospasm. Despite this result the authors cautioned against the use of selective COX-2 inhibitors in this group of patients, except in the context of research. Members agreed with this conclusion.

12.3 The safety of oral contraceptives

12.3.1 Risk factors for venous thromboembolism

- ... Risk factors for venous thromboembolism in women using oral contraceptives. Report to Medsafe, Ministry of Health, October 1999, revised April 2001

.. had prepared an update of their earlier report on risk factors for VTE in women using oral contraceptives. Endorsement of this report was to be sought from the Royal New Zealand College of General Practitioners, the Royal Australian and New Zealand College of Obstetricians and Gynaecologists, the New Zealand Family Planning Association and the New Zealand Medical Association. It is then to be published in full on the Medsafe web site and an abbreviated version both in Prescriber Update and on the web site. Publication in the New Zealand Medical Journal is also to be investigated.

Members commented that there was some significant material, namely the review by Vandenbroucke et al (see 12.3.4) and updated guidelines from the WHO, which had been published since .. update had been completed.

The Committee recommended asking .. to include the new material in the report, before proceeding. The Committee also recommended asking .. to peer review the report.

12.3.2 Use of oral contraceptives in migraine

- .., Family Planning Association. Use of oestrogen-containing oral contraceptives in women with a history of migraine. Letter 3 April 2001
- EA MacGregor, J Guillebaud. Combined oral contraceptives, migraine and ischaemic stroke. Brit J Family Planning 24:53-60, 1998
- C Tzourio, et al. Migraine and risk of ischaemic stroke: a case-control study. BMJ 307:289-92, 31 Jul 1993
- SM Schwartz, et al. Stroke and use of low-dose oral contraceptives in young women. A pooled analysis of two US studies. Stroke 29:2277-84, 1998
- CL Change, et al. Migraine and stroke in young women: case-control study. BMJ 318:13-8, 2 Jan 1999

.. of the Family Planning Association (FPA) had asked the Committee for an opinion on the use of COCs in women with migraine. The FPA has regarded COCs to be contraindicated in women with migraine without aura with more than one risk factor for stroke and in women with migraine with aura. Its clinical practice follows the guidelines published by MacGregor and Guillebaud. The Association understands that some neurologists do not agree with its recommendations.

Each of the three studies listed above found that the use of oral contraceptives increases the risk of stroke in women with migraine (increase non-significant in Tzourio study). Chang et al commented that in women with migraine the use of oral contraceptives increased the risk of stroke in a greater than multiplicative fashion, but they did not distinguish between migraine with and without aura in this analysis.

.. advised that the WHO had included recommendations about the use of COCs in migrainous patients with and without aura in the recently updated guidelines. These guidelines are to be used as a basis on which regulatory authorities and other bodies can develop their own advice.

The Committee recommended:

12.3.3 Recent published comment on oral contraceptives

- P Egermayer. Obesity, oral contraceptives, and fatal pulmonary embolism. NZMJ 114:170-171, April 2001
- R Paterson. Prescribing the pill: obligations of health professionals. NZMJ 114:188, April 2001

Dr Egermeyer's comment emphasised that obesity is a risk factor for VTE with COCs and recommended that women with obesity take non-thrombogenic alternatives.

Mr Paterson, Health and Disability Commissioner, had summarised his opinion on a case in which no family history had been taken or advice given regarding the risk of VTE in a woman seeking a COC. The opinion had been that although a patient would not normally expect to be informed of risks of the order of 2 in 10,000, fuller disclosure was required in this instance because of the publicity surrounding the issue.

12.3.4 Review article

- JP Vandenbroucke, et al. Oral contraceptives and the risk of venous thrombosis. NEJ M 344:1527-35, 17 May 2001

The review by Vandenbroucke, et al focussed particularly on the influence of various kinds of thrombophilia on the risk of VTE.

It was noted that the authors had cited the Prescriber Update article on deaths with third generation oral contraceptives (Dec 1998), with reference to the observation of the article that in most cases the prescriber could not have foreseen the risk.

12.4 Oral corticosteroids and fracture risk

- T P van Staa, et al. Oral corticosteroids and fracture risk: relationship to daily and cummulative doses. Rheumatology 39:1383-1389, 2000

Van Staa et al used the UK General Practice Research Database to identify 244,235 oral corticosteroid users and 244,235 controls. They found that higher daily doses (( 7.5mg of prednisolone or equivalent) was associated with an increased risk of non-vertebral fracture (1.44; 1.34-1.54), hip fracture (2.21; 1.85-2.64) and vertebral fracture (2.83; 2.35-2.40) relative to use of oral corticosteroids at lower daily doses. Fracture risk was also elevated at higher cumulative dose, but this effect was almost completely removed with controlling for daily dose, age, gender and other confounding factors.

Members commented that 7.5mg daily is not usually regarded as a "high" dose; ( 20mg daily would usually be regarded as "high". This study found that the demineralising effect of corticosteroids is an acute effect.

The Committee recommended publishing a paragraph in Prescriber Update on the results of this study with a recommendation to use the lowest effective dose. The article is to be written by .., and peer reviewed by .. and by a rheumatologist.

12.5 Hydroxyethylstarch in sepsis and renal function

- F Schortgen, et al. Effects of hydroxyethylstarch and gelatin on renal function in severe sepsis: a multicentre randomised study. Lancet 357:911-916, 2001

Schortgen et al randomised 129 patients with severe sepsis or septic shock to receive 6% hydroxyethylstarch or 3% fluid modified gelatin. Severity of illness and serum creatinine were similar in the two groups at baseline. Acute renal failure and oliguria were significantly higher in the group given hydroxyethylstarch than in the gelatin group and peak serum creatinine rose to a higher level in the former group.

The Committee recommended that comment be sought from the College of Anaesthetists, Faculty of Intensive Care.

12.6 Anticonvulsants and teratogenicity

- L B Holmes, et al. The teratogenicity of anticonvulsant drugs. NEJM 344:1132-1138, April 2001

Holmes et al screened 128,049 pregnancies at delivery for exposure to anticonvulsants, maternal history of seizures without anticonvulsant therapy and no maternal history of seizures. Embryopathy was higher in 223 infants exposed to anticonvulsants than in 508 control infants with no maternal history of seizures (odds ratio 2.8; 1.1-9.7). Exposure to two or more anticonvulsants increased the rate further (4.2; 1.1-5.1). The infants (98) whose mothers had a history of epilepsy, but took no anticonvulsants did not have a frequency of embryopathy above that in the control group.

The Committee noted the clear result of the study, that congenital abnormalities in infants of mothers with epilepsy are associated with the medication not the disease.

The Committee recommended that no action be taken.

12.7 Fluoxetine and suicide/suicidal ideation

- ... Fluoxetine and suicide. Report for Medsafe. May 2001
- GM Anderson, et al. Serotonin and suicidality: The impact of fluoxetine administration. II: Acute neurobiological effects. Isr J Psychiatry Relat Sci 32:44-50, 1995
- MH Teicher, et al. Antidepressant Drugs and the emergence of suicidal tendencies. Drug Safety 8(3):186-212, 1993
- D Healy. Emergence of antidepressant induced suicidality. Primary Care Psychiatry 6:23-8, 2000
- MJ Tueth. Revisiting fluoxetine (Prozac) and suicidal preoccupations. J Emergency Med 12:685-7, 1994
- M Fux, et al. Emergence of depressive symptoms during treatment for panic disorder with specific 5-hydroxytryptophan reuptake inhibitors. Acta Psychiatr Scand 88:235-7, 1993
- SS Jick, et al. Antidepressants and suicide. BMJ 310:215-8, 28 Jan 1995
- Fluoxetine and suicide. Item 8(5), MARC minutes Nov 1995.
- A Jackson. Two years' jail for antidepressant killer. Sydney Morning Herald 24 May 2001
- SSRI data sheet extracts on suicide, emergent psychiatric symptoms for Zoloft, Serzone, Cipramil.

A member who declared that he had a conflict of interest was permitted to participate in the discussion and decision-making.

Members also noted that .. who had prepared the report for Medsafe had a conflict of interest.

.. noted that treatment-emergent suicidality with fluoxetine was first recorded in 1990 by Teicher et al,6 in an article reporting 6 cases. A subsequent review by Teicher et al concluded that despite the efficacy of fluoxetine in reducing suicidal ideation in the majority of patients when this symptom is present prior to the commencement of treatment, fluoxetine is associated with treatment-emergent suicidality in 4.1% of users.

.. cited the study using the UK General Practice Research Database by Jick et al which was based on 143 patients who completed suicide out of a cohort of 172,598 who received at least one prescription of an antidepressant. Compared with dothiepin, fluoxetine was found to be associated with a relative risk of suicide of 3.8 (1.7-8.6). When the analysis was limited to those prescribed only one antidepressant and having no history of suicidality the relative risk fell to 2.1 (0.6-7.9).

He noted that a study by Warshaw and Keller7 found that suicidal attempts were more than twice as frequent among patients with anxiety disorders but not depression taking fluoxetine (4/118, 3.6%) compared with others not taking this medication (6/370, 1.5%).

A study which .. considered to be key was one by Healy in which 20 healthy volunteers were randomised to sertraline or reboxetine in crossover fashion. Two of those taking sertraline became suicidal in association with a combination of akathisia, dysphoria and disinhibition.

Suicidality with fluoxetine has been observed to be of rapid onset, compared with the longer lag time (several weeks) for a beneficial change in mood to occur. Some side effects, such as akathisia, agitation, dysphoria and disinhibition can develop rapidly and might lead to suicidality. .. favoured this hypothesis and proposed that the co-prescription of benzodiazepines might relieve the agitation and akathisia. He considered that this effect (treatment-emergent suicidality) was a class effect applying to all SSRIs and all of these medicines need to carry a warning that these effects might occur in a small number of patients. In addition, prescribers should be advised to carefully monitor any patients showing signs of restlessness, agitation, akathisia or disinhibition.

Members agreed with .. that treatment-emergent suicidality does occur with fluoxetine, and other SSRIs. However, members understood that suicidal preoccupation often emerges as the first symptom rather than occurring following akathisia and agitation. Hence prescription of benzodiazepines would not help.

Members agreed that the key message to convey was that there have been reports of increased or emergent suicidal preoccupation with fluoxetine and other SSRIs. If these symptoms develop the medication should be stopped immediately. The national health guidelines recommend that patients with depression should be seen once a week, and those with severe depression should be seen twice a week. Members realised that these recommendations were unrealistic. An alternative useful recommendation would be to advise medical practitioners to say to those to whom they prescribe an SSRI "If you are feeling a lot worse ring me." Prescribers should also warn patients of the possibility of agitation, and advise them to make contact if these symptoms develop.

The Committee recommended:

12.8 Items of interest

12.8.1 Orlistat and diabetic ketoacidosis

- S T Azar, MS Zantout. Diabetic ketoacidosis associated with orlistat treatment. Diabetes Care 24:602, Mar 2001. (Reactions abstract)

Azar and Zantout described a case of an 18-year-old woman with type 1 diabetes mellitus who developed diabetic ketoacidosis a month after commencing orlistat therapy. Five days after withdrawal of orlistat her insulin requirements returned to normal. This adverse reaction is of interest because it is of a systemic adverse reaction in association with a medicine which is not thought to be systemically absorbed to a significant degree.

12.8.2 NSAIAs and miscarriage

- G L Nielsen, et al. Risk of adverse birth outcome and miscarriage in pregnant users of non-steroidal anti-inflammatory drugs: population based observational study and case-control study. BMJ 322:266-270, February 2001
- LY Chan et al, P Kristensen, GL Nielsen, et al. Risk of miscarriage in pregnant users of NSAIDs. Correspondence. BMJ 322:1365, 2 June 2001

Nielsen et al found a small non-significant increase in the risk of congenital abnormality in infants exposed to NSAIAs during pregnancy, and a significant association between taking up a prescription for an NSAIA in the previous weeks and miscarriage (6.99; 2.75-17.74).

The Committee recommended no action on this subject, but asked that the issue of NSAIA exposure during pregnancy and pulmonary hypertension in infants be on the agenda at the next meeting.

12.8.3 Oestrogen replacement therapy and ovarian cancer

- C Rodriguez, et al. Estrogen replacement therapy and ovarian cancer mortality in a large prospective study of US women. JAMA 285:1460-1465, March 2001

12.8.4 Sildenafil and cardiovascular events

- S A W Shakir, et al. Cardiovascular events in users of sildenafil: results from first phase of prescription event monitoring in England. BMJ papers 322:651-652, March 2001

Using the prescription event monitoring (PEM), Shakir et al found 10 fatal cases of ischaemic heart disease or myocardial infarction in men with a prescription for sildenafil. Details of temporal association to administration were not available. The number of deaths per year in a cohort of English men from the general population matched for age was 74/1000 for 1998. The calculated number of deaths from the PEM cohort was 29/1000/year.

Members noted that while the study result was interesting, the study had many limitations, such as potential differences between this cohort and the general population of English men.

12.8.5 Nasal fluticasone and benign intracranial hypertension

- D W Bond, et al. Benign intracranial hypertension secondary to nasal fluticasone propionate. BMJ 322:897, 14 April 2001

12.8.6 Safety issues with itraconazole and terbinafine

- FDA issues health advisory regarding the safety of Sporanox products and Lamisil tablets to treat fungal nail infections. FDA Talk Paper 9 May 2001

12.9 Eczema creams and peanut allergy

- Eczema creams in peanut allergy link. News item 7 June 2001

The news item proposed an association between peanut oil in eczema creams and creams for other skin conditions (such as nappy rash) and peanut allergy. Members commented that it was also possible that the association was the presence of the atopic gene which predisposed to both eczema and peanut allergy. Whatever the association in this particular case, members considered it to be important that the ingredients of topical preparations be on the labelling, so that those with allergies or at risk of developing them can avoid products containing allergens.

The Committee recommended requiring that all ingredients, including preservatives, of topical medicines be listed in the labelling.

The meeting closed at 5.15pm.