Medsafe Logo
<% Dim q q = request.form("q") If len(q) > 0 Then Response.redirect "/searchResults.asp?q=" & q End if %>
Hide menus
Show menus


Updated 31 December 2012

Minutes of the 104th Medicines Adverse Reactions Committee Meeting - 7 December 2000


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Descriptions of unpublished case reports have not been included to protect the privacy of those involved.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



Associate Professor TJB Maling (Chairperson)
Dr M Tatley (Medical Assessor)
Dr R Savage (Medical Assessor)
Dr DM Coulter
Dr J Goldsmith
Dr H Kingston
Dr S Jessamine
Dr F McClure
Dr M Rademaker
Professor DCG Skegg
Dr N Rafter (Medical Advisor, Medsafe)
Dr K Ronaldson (Secretary)


Dr S Martindale
Ms S Von Afehlt


1.1 Welcome and apologies

Apologies were received from Professor Ellis. The Chairperson welcomed Ms S Von Afehlt, newly appointed Senior Advisor (Pharmacy), and Editor of Prescriber Update and other Medsafe publications to the meeting.

1.2 The Chairperson's comments

The Chairperson began the meeting with some introductory comments. He thanked members for the hard work during the year, and noted that CARM was progressing well. He notified members of his intention next year to deal with the outstanding issues raised in last year's pharmacovigilance review and to discuss subjects such as hospital admissions, medicine errors, appropriate prescribing and polypharmacy. The CIOMS document on benefit-risk analysis should be discussed. Plus he would like to see Medsafe conducting some evaluation of the impact of communication recommended by the Committee.

In addition, the Chairperson asked for an update of progress towards harmonisation with Australia. Dr Jessamine advised the meeting that both governments are considering a joint agency. New legislation would be required in both countries and a product licensing system would be likely. The licensing fee would then fully cover the cost of pharmacovigilance. At present in the pharmacovigilance area there is an open invitation for MARC members to attend ADRAC and Dr John McEwen from ADRAC would be attending some MARC meetings next year.

The Ministry of Health is setting up a system for controlled drug prescriptions to be picked up electronically, and it may be possible to pick up prescriptions for IMMP medicines in the same way.

With regard to the impact of messages sent out to general practitioners by Medsafe, there was a possibility of using PreMec as a means to obtain feedback.

1.3 In remembrance of Professor Kellaway

The Chairperson advised the meeting that the previous Chairperson, Professor Kellaway had died recently. Members asked that he pass on the condolences of the Committee to Professor Kellaway's widow and express to her the appreciation of the Committee for his contribution.


Members requested the following amendments to the minutes:

In addition, a member commented that the cisapride indication for children should state "where the diagnosis has been made or confirmed by a paediatrician" not a "specialist physician." Members agreed that the minutes represented the discussion at the meeting. The advice of the change in the indications had already be sent out to medical practitioners with a dear doctor letter. While the wording was not ideal, it was agreed that corrective action was not necessary, as specialist physician includes paediatrician.

3. DATES of the meetings for 2001

The following were confirmed as the meeting dates for 2001:

14 March (Wed), 14 June (Thu), 5 September (Wed), 6 December (Thu).


4.1 Conflict of interest

- Conflict of interest response form (current one)
- Draft guidelines for declaration of competing interests - MARC members
- Draft form for declaration of competing interests - MARC members
- Australian Drug Evaluation Committee competing interest guidelines, Feb 1999

The Secretary had drafted updated conflict of interest guidelines and response form as a consequence of the discussion at the previous meeting. The guidelines proposed that members submit conflict of interest forms when they return the draft minutes with their corrections. The conflicts of interest would be tabulated and sent to the Chairperson for a decision on how the conflict should be managed.

After some discussion members accepted the proposal that conflict of interest forms be returned with the draft minutes with the proviso that it be possible to notify conflicts at any time including at a meeting. The Committee considered that the decision on the management of any conflict of interest should be made by the Committee as a whole, not exclusively by the Chairperson.

Regarding the statements about conflicts of interest to be included in the minutes, members agreed that neither the member nor the specific conflict should be named, but the method by which a conflict was managed should be stated. Members did not wish the guidelines to outline how different kinds of conflict would be handled. However, it was agreed that it would be appropriate to revisit this issue in a year's time to see whether the Committee considered they would like the guidelines to include this aspect after accumulating some precedent.

Members had only one difficulty with the conflict of interest form. They asked that the line requesting declaration of "a personal or religious belief that may influence your view of issues discussed" be deleted.

Members asked that the guidelines and form for declaration of competing interests be revised according the members' comments and presented to the next meeting for ratification.

4.2 Publications

Dr Jessamine advised the Committee that the previous week Medsafe had held a meeting to discuss the future shape of its publications. It had been agreed that Prescriber Update would be published according to a strict schedule, with 3 issues in 2001, and probably 4 issues in the following year. The focus would be on medicine safety issues and reviews of treatment options for a particular condition would no longer be published. In addition, to make it more readable, the number of articles in each issue and the length of each article would be reduced. In general, the articles would not provide the background to an issue. They would briefly outline the nature of the issue and then present clinical advice. References would be provided for those who wished to do further research and the full article, including background would be published on the web site. Members accepted this proposal without comment.


5.1 Report on the recommendations

- K Ronaldson. Secretary's report
- Report to the Minister's delegate on the minutes and recommendations of the 103rd meeting of MARC, 16 Nov 2000
- Recommendations of the 103rd meeting of the MARC

5.1.1 MARC minutes on the web site (4.2, Sep 2000)

- P Ellis. Minutes. Email 31 Oct 2000
- Minutes of 103rd meeting prepared for the web site

Members discussed the potential for misunderstanding associated with the listing of deaths at the beginning of the discussion of case reports where the individual cases are discussed under the therapeutic group. It was considered that some people seeing the list would think the deaths had not been discussed. Members recommended that a hyperlink from the death as listed at the beginning of the section be made to the discussion of the case, and that consideration be given to providing some statement to the effect that comment on each of the deaths can be accessed by double clicking on each of the entries.

In addition, there was a discussion about naming persons who were designated, or to be approached, to write an article or for some other reason. Members agreed that in this case names should be deleted from the version of the minutes to go on the web site, although it was appropriate for the person to be named in the full minutes.

5.1.2 Doxazosin and the ALLHAT study (5.1.3, Sep 2000)

- N Rafter. Doxazosin and the ALLHAT study. Published on web site, Nov 2000

Members noted the final version of the article, as it was published on the web site.

5.1.3 Minocycline and benign intracranial hypertension (6.4.1, Sep 2000)

- H Kingston. Minocycline and benign intracranial hypertension. Draft Prescriber Update article.

Dr Kingston indicated that she intended to shorten the article a little. Members agreed that it suited the intended purpose well.

It was noted that while it is more effective than doxycycline in the treatment of acne, the use of minocycline is declining because it is no longer funded. It is now more expensive than isotretinoin. Furthermore it is cheaper for a patient to have (unfunded) isotretinoin prescribed by a general practitioner, than to pay for a visit to a specialist and have the prescribed isotretinoin funded.

Members asked that the safety and usage of isotretinoin be included on the agenda of the next meeting.

5.1.4 Nefazodone (8.2, Sep 2000)

- K Ronaldson. Safety issues with Serzone (nefazodone). Letter to Bristol-Myers Squibb. 2 Oct 2000

Members noted that a letter had been sent to Bristol-Myers Squibb asking the company to update the data sheet for hepatic, withdrawal and anticholinergic effects. No reply had as yet been received.

5.1.5 Cisapride and cardiac arrhythmias (12.1, Sep 2000)

- Report for the Delegate of the Director General on the s.36 notice concerning the safety of Prepulsid (cisapride). 4 Oct 2000
- K Ronaldson. Indication changes for cisapride. Published on web site Nov 2000

Members noted the action that had been taken regarding the changes in the indications recommended at the September meeting. At the end of November the sponsor for Prepulsid had sent out dear doctor, dear pharmacist letters with the updated data sheet and consumer medicine information (CMI), and the updated data sheet and CMI had been put on the Medsafe web site along with an article by Medsafe. The Committee asked that the safety and availability of cisapride be brought up again for review by the Committee in a year's time.

5.1.6 Risk factors project for venous thromboembolism with oral contraceptives

- C Webber, Executive Officer to the CEO, Royal New Zealand College of General Practitioners. Risk factors for venous thromboembolism (VTE) with oral contraceptives. Letter 27 Sep 2000
- G Greer, Executive Director, Family Planning Association. Letter 6 Oct 2000
- JR Doig, Chairman, NZ Committee, Royal Australian and New Zealand College of Obstetricians and Gynaecologists. Letter 12 Oct 2000
- J Patterson, Manager Policy & Support, New Zealand Medical Association. Letter 8 Nov 2000

The Secretary had written to the Royal New Zealand College of General Practitioners (RNZCGP), Family Planning Association (FPA), Royal Australian and New Zealand College of Obstetricians and Gynaecologists (RANZCOG) and the New Zealand Medical Association (NZMA) enclosing a copy of the report by Associate Professor ... and Ms ... on risk factors for venous thromboembolism (VTE) with oral contraceptives. The Secretary had asked whether these organisations considered that the material in the report should be disseminated to their members. All four organisations were in favour of the material being disseminated and FPA and NZMA suggested using the summary as a basis for an article with a brief explanation regarding each risk factor.

The Committee recommended:

5.1.7 Safety of sildenafil (12.6, Sep 2000)

Dr ... discussed with the Committee her intentions regarding a paragraph on the 24 hour post sildenafil contraindication to the use of nitrates, and the treatment of angina in this period. Members considered that advice should include a recommendation to ask a man presenting with angina whether he has used sildenafil recently, particularly in the last 24 hours. In such cases nitrates should be avoided, the patient treated with morphine and admitted to hospital.

Members recommended that:

5.1.8 Hyperprolactinaemia and antipsychotics (5.1.5, Mar 2000)

- N Rafter. Hyperprolactinaemia with antipsychotics. Draft article for Prescriber Update

Members considered that the article covered the subject well. It was noted that it is to be sent to Professor ... for peer review.

5.1.9 Phentermine and amfepramone (12.6, Mar 2000)

- Report for the Delegate of the Director-General of Health on the final action following s.36 notices for Tenuate Dospan, Umine and Duromine. 27 Oct 2000
-.... Safety and efficacy of amfepramone and phentermine. Report for Medsafe. Oct 2000
- P Kopelman. Prescribing for obesity. Comment on the Royal College of Physicians' working party report on clinical management of overweight and obese patients with particular reference to drugs. J R Coll Physicians Lond 33:31-2, 1999

The Committee noted that Dr ... who had been asked to evaluate material submitted by the pharmaceutical companies concerning the safety and efficacy of amfepramone and phentermine, had considered that there was no reason to withdraw either medicine from the market because of either safety or efficacy issues. He had made recommendations concerning the indications, contraindications and warnings and precautions. Medsafe had asked the companies to change their prescribing advice for Tenuate Dospan, Umine and Duromine accordingly, but responses had not yet been received.

5.1.10 The treatment of fever in children (5.3, Mar 2000)

- .... A fleshy problem. Overheads for a presentation, 23 Nov 2000

Members noted the material used in the presentation on the treatment of fever in children and considered it to be a good basis for an article. The material cautioned against routinely treating children for fever, because mild fever can be beneficial in the body's fight against the infection, and there may be adverse consequences from the analgesics used. Members, however, wanted the article to state more clearly that paracetamol can be used appropriately to treat pain.

Members asked Dr ... to draft the material into an article. It was not yet clear where it would be appropriate for the article to be published.

5.1.11 Amiodarone and pain (6.9.1, Mar 2000)

Dr ... advised that she was seeking advice from a cardiologist on the article about amiodarone and pain. She had been told that similar symptoms of pain can occur following injections of protamine.

5.1.12 Request from Pharmac for advice

- L Li Bassi, PTAC Secretary. Desogestrel 75mcg (Cerazette). Letter 27 Oct 2000
- K Ronaldson. Desogestrel 75 mcg (Cerazette). Letter 31 Oct 2000

Members noted that Pharmac had requested advice from the MARC concerning the risk of VTE with the progestagen-only pill, desogestrel. The Secretary had responded by supplying the evaluator's report and a minute item from MARC (March 2000) discussing 2 small studies of the risk of VTE with progestogen used alone.

5.1.13 Comment from the Coroner's Council

- R McElrea, Chairman Coroners' Council. Coroners' Findings, Release under the Official Information Act. Letter 1 Nov 2000
- Coroners Act 1988, s.44(2)
- Health Information Privacy Code 1994 Amendment No.4 (see 4(n))
- Law Commission Report on Coroners #321-5
- National Coroner Information System (Australia), Brochure
- Health Information Privacy Code, article in The New Zealand Coroner Aug 2000

Members noted the comment of the Chairman of the Coroner's Council to the effect that Coroner's Findings are public documents. However, it was also noted that Coroner's Findings are now given mention in the Health Information Privacy Code and the consequences of this are not yet clear.

The Committee asked that a letter of thanks be sent to Mr McElrea.

5.1.14 Action sheet

Members noted that, except for the comments of the Secretary in her reports, they were not provided with any means to keep a track of action on all recommendations. Members asked that they be provided at each meeting with a list of recommendations made, an indication of who was responsible, and an indication of whether action had been completed.

6. General Reporting

- Adverse reaction reports reviewed by MARC since Nov 97, by medicine class
- Adverse reaction reports reviewed by MARC since Nov 97, by recommendation
- Glossary of terms

There was some discussion of the presentation of the adverse reaction reports considered at previous meetings. Members asked that the following changes be made:

In addition members asked that the CIOMS definitions of the frequency terms be provided each time with the 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:

  • comment about causality;
  • information about similar suspected adverse reactions reported with the same or related medicines;
  • prescribing advice;
  • advice related to the care of the patient, including information that may assist the practitioner to make a risk/benefit assessment for future treatment; and
  • any specific action being taken by the Centre, including entry of the reaction into the National Health Index against the patient's name, presenting the reaction to the MARC, including it in an article to be published, etc.

6.1 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.

Tramadol and hepatitis, gastrointestinal haemorrhage, 45531
Celecoxib and aortic aneurysm rupture, 44840
Enoxaparin and haematoma, haemorrhage, 45661
Mercilon and pulmonary embolism, 45866
Tetracosactrin and nausea, tachycardia, hypertension, cerebral haemorrhage, atrial fibrillation, 45706

6.2 Analgesics

6.2.1 Tramadol and hepatitis, gastrointestinal haemorrhage (death), 45531

Comment: The WHO database holds 37 reports of gastrointestinal haemorrhage and 73 of hepatitis with tramadol. The latter is listed as an adverse reaction in the data sheet, but not the former. Because of the disagreement between the information supplied from different sources, members asked that the pathologist be asked on what basis he or she believed the patient had been taking NSAIAs. It was considered that until further information had been received the report should be designated unclassified. No further action was recommended.

6.3 Antiepileptics

6.3.1 Carbamazepine and Stevens Johnson syndrome, liver enzymes raised, 45648

6.3.2 Carbamazepine and rash, peripheral oedema, 45849

6.3.3 Comment

The CARM database holds 71 reports of rash, 10 of Stevens Johnson syndrome, 12 of oedema and 60 of hepatic events with carbamazepine. The data sheet for Tegretol (carbamazepine) describes rash as very common and Stevens Johnson syndrome as very rare. Members suggested that the first case was probably not Stevens Johnson syndrome, but hypersensitivity syndrome, and suggested that CARM consider changing the reaction term for the skin reaction. Both of these adverse reactions were considered to be possibly related to the use of carbamazepine. It was noted that skin reactions with carbamazepine had recently been removed from the list of adverse reactions of current concern and that an article had been published about this subject in Prescriber Update in June 1999. No additional action was recommended.

6.4 Antiinfectives

6.4.1 Norfloxacin and vomiting, dizziness, myositis, myalgia, creatine phosphokinase increased, 45705

Comment: The CARM database holds 1 report of myalgia with norfloxacin and the WHO database 74 reports of myalgia and 2 of myositis. Neither the data sheet nor the literature refer to myositis or creatine phosphokinase elevations with norfloxacin. Members questioned whether all of the patient's symptoms may have been associated with pyelonephritis, arising from complication of the urinary tract infection, or whether the symptoms may have been of a hypersensitivity reaction. All of the events were classified possible. No further action was recommended.

6.5 Antiinflammatories

6.5.1 Celecoxib Celecoxib and aortic aneurysm rupture (death), 44840

Comment: Members considered it to be unlikely that there was a causal association between celecoxib and aortic aneurysm rupture in this patient. Celecoxib/amiodarone and melaena, gastritis, anaemia, hypotension, 45519

Comment: The CARM database holds 4 reports of melaena with celecoxib and 5 reports of hypotension with amiodarone. Since the patient was already taking omeprazole there probably was an underlying predisposition to gastrointestinal bleeding. Members agreed that the causal association should be designated possible in this case Celecoxib/amitriptyline and nausea, 45283

Comment: Celecoxib may inhibit the metabolism of amitriptyline by inhibiting cytochrome P450 2D6, but celecoxib by itself may cause nausea, which is likely to be dose-related. The association with celecoxib was designated probable, but the interaction was not classified. Comment

For more on celecoxib see section 12.5. The meeting of June 2000 had recommended publication of an article on the safety profiles of celecoxib and rofecoxib, and preparation of this article was still in progress. No additional action was recommended by the Committee.

6.5.2 Diclofenac and amelia, undescended testicle, 45279

Comment: There are no reports of amelia or undescended testes in an infant exposed in utero to diclofenac, ergotamine, caffeine or cyclizine in the CARM database, the international literature or the product data sheets. Members considered the associations to be unlikely, especially considering the timing of medicine use relative to conception, and recommended no further action.

6.6 Antithrombotics

6.6.1 Enoxaparin and haematoma, haemorrhage (death), 45661

Comment: Haematoma and haemorrhage were designated probable, and the outcome was given as "died - drug may be contributory" for this very ill patient who suffered many complications. CARM now holds 12 reports of haematoma with enoxaparin. No additional action was recommended.

6.7 Cardiovascular

6.7.1 Flecainide and pulmonary fibrosis, 45644

Comment: The CARM database holds no reports of pulmonary fibrosis with flecainide, but there are 7 reports in the WHO database. The Tambocor (flecainide) data sheet indicates that extremely rare cases of pneumonitis have been reported, but these may not be causal. The case was designated unclassified, because this adverse reaction is not known to be caused by flecainide. The Committee recommended no further action on this report.

6.8 Hormones

6.8.1 Oral contraceptives Loette (ethinyloestradiol 20µg/levonorgestrel 100µg) and DVT, 45067

Comment: Members comment that the original pain in the woman's knee may have been from DVT rather than a sprain. This is the first New Zealand report of DVT with this very low oestrogen dose second generation oral contraceptive. Marvelon and DVT, 45824 Mercilon and DVT, 44785 Mercilon and pulmonary embolism, DVT, 45209

Comment: Members considered that the woman's suspected asthma may have been the first symptoms of pulmonary emboli. Mercilon and pulmonary embolism (death), 45866 Comment

There was some discussion about the causality designations of these cases. CARM has historically designated them possible. Some members considered they should all be rated probable, because of the rarity of these events in healthy women not taking oral contraceptives. Others suggested that those for which there were no other possible risk factors should be rated probable and the others possible. But others considered that the presence of risk factors should not be a reason to change the designation, because the risk factors did not change the strong likelihood of the oral contraceptive being a causative factor. Members agreed that it was important that the designations were consistent, and left it up to the CARM staff to decide how such cases are designated.

No action was recommended on the basis of these reports further to the action recommended under 5.1.6. See also 12.1.

6.8.2 Hormone replacement therapies (HRTs) Estrofem (oestradiol) and DVT, 45141

Comment: This was the first New Zealand report of VTE with this product, although this reaction has been reported with other HRTs. It was noted that the dose appeared to be rather high, but there may have been a transcription error in writing the report. Members asked that the reporter be contacted to clarify the dosage used. Kliogest (oestradiol/norethisterone continuous) and pulmonary embolism, pulmonary hypertension, right cardiac failure, DVT, 45130

Comment: Members observed that a previous report of pulmonary hypertension and pulmonary embolism with a combination of oestrogen plus progestagen (the OC Marvelon, case no. 41693) had been received by CARM in 1999. The woman had not recovered at the time of reporting, although the initial event had occurred 4 years previously. Members commented that pulmonary hypertension may accompany all cases of pulmonary embolism. Both cases had been designated possible. No further action was recommended.

6.8.3 Other hormones Tetracosactrin and nausea, tachycardia, hypertension, cerebral haemorrhage, atrial fibrillation (death), 45706

Comment: The Committee considered that there were possible post-operative complications that may have initiated the string of events leading to this patient's death. However, nausea and hypertension are events listed in the data sheet for Synacthen (tetracosactrin), and the causality designation for these events should remain possible, but the other events should be designated unclassified. No further action was recommended. Somatropin and vision abnormal, brain neoplasm malignant, 45640

Comment: The WHO database holds 14 reports of brain neoplasm with somatropin, and neoplasm has been reported in the literature. The data sheet refers to intracranial hypertension and relapsed malignancy. The causality was listed as possible. No additional action was recommended.

6.9 Psychiatric

6.9.1 Citalopram and ventricular septal defect, 45804

Comment: The WHO database holds 3 reports of foetal heart malformation (unspecified) with citalopram and 1 with trimethoprim. The literature appears to have no reports of similar events. The data sheet states that animal studies have shown no evidence of teratogenic potential. Members commented that ventricular septal defect is a self-healing malformation and the defect in this baby was very small. The causality was designated possible, and the Committee recommended no further action.

6.9.2 Paroxetine/amoxycillin and hepatic enzymes increased, hepatorenal syndrome, anorexia, fall, 45269

Comment: The CARM database holds 4 reports of hepatic reactions with paroxetine and 10 for amoxycillin. There are no reports of hepatorenal syndrome for either on the CARM database, but there are 2 reports for paroxetine on the WHO database. Members commented that the patient may have had abnormal liver function for some time. They were doubtful of the designation, hepatorenal syndrome, and asked that the diagnosis be verified, if possible, with the reporter. The events were considered to be unclassifiable. No further action was recommended.

6.10 Alternative medicines

6.10.1 Ginger/warfarin and epistaxis, 45460

Comment: Ginger (Zingiber officianale) has been reported to have platelet aggregation inhibition activity. The CARM database holds 3 previous reports of epistaxis with warfarin. Members commented that either warfarin or ginger extract or both could be causative, and considered that the designation should be possible. The Committee recommended no further action.

6.10.2 Raspberry leaf tea and threatened abortion, 45313

Comment: Raspberry leaf tea can stimulate and coordinate uterine contractions. The causal association in this case was regarded as possible. No further action was recommended.

6.10.3 St Johns wort and cataract, 45867

Comment: Members considered the association in this case to be unlikely, and recommended no further action.

6.10.4 St Johns Wort/sumatriptan and amnesia, 45481

Comment: Members observed that a possible interaction between sumatriptan and St Johns wort was mentioned in the material sent out by Medsafe on interactions with St Johns wort, but the effects anticipated were serotonergic in nature and would not include short term memory loss. Amnesia is not listed in the data sheet as an adverse reaction of sumatriptan. Members questioned whether stopping St Johns wort may have been a factor. However, the adverse reaction was listed as possibly associated with St Johns wort and sumatriptan. No additional action was recommended.

6.10.5 Thyromax and TSH decreased, 45674

Comment: Thyromax was thought to be a thyroid extract and the association was considered to be possible. Earlier this year the FDA put out an alert about a product which contained a precursor of T3. At the time there were not thought to be any similar products available in New Zealand. Members noted that the Compliance Team would be asked to check the source of this product and investigate its continued availability.

7. Vaccines

- Vaccine adverse reaction reports reviewed by MARC since Nov 97

7.1 New Zealand adverse reaction reports

7.1.1 Influenza vaccine and Guillain Barré syndrome, 45240

Comment: CARM holds reports of 4 previous cases of Guillain Barré syndrome. An epidemiological study conducted in the US found the excess (over background) incidence of cases of Guillain Barré syndrome with influenza vaccine to be slightly over 1 per million.1 The study considered cases to be vaccine related if they occurred within 6 weeks of immunisation. The causality for this case was designated possible.

7.1.2 Influenza vaccine and vision blurred, dizziness, hypoaethesia, slurred speech, headache, 45873

Comment: The CARM database holds a huge diversity of adverse reactions with influenza vaccine, including reports of the symptoms described in this case. Some of these reports may describe coincidental symptoms, especially considering the widespread use of the vaccine. The Committee considered the events in this case to be unclassifiable.

7.1.3 Comment

No additional action was recommended in relation to these reports.

7.2 Articles from the literature

7.2.1 Hypnotic hyporesponsive episodes with pertussis-containing vaccines

- T S DuVernoy, M Braun, and the VAERS Working Group. Hypotonic-hyporesponsive episodes reported to the vaccine adverse event reporting system (VAERS), 1996-1998. Pediatrics 106, Oct 2000

DuVernoy, et al have conducted a study of reports of hypotonic hyporesponsive episode (HHE) reported to the US vaccine adverse event reporting system (VAERS) during 1996 to 1998. The identification of cases involved interview of a caregiver to ascertain whether the case fitted the case definition. The total number of cases was 215. All were non-fatal and 53% were in females. The age range was 1.1-107 (median 4.0) months. 67% of children had received whole-cell pertussis-containing vaccines and 26% received diphtheria-tetanus-acellular pertussis and 7% received a vaccine without a pertussis component.

The time to onset of the event was 1 minute to 48 hours, with 82% occurring within 12 hours of vaccination. For whole-cell pertussis vaccine the peak for occurrence was at 3-4 hours after vaccination, but for acellular pertussis-containing vaccine it occurred within the first hour. 98.6% returned to their prevaccination state, and median time for recovery was 6 hours. Only 3 of the children were reported to be not fully recovered and these developed conditions (autism, complex partial epilepsy and developmental delay) not known to be causally associated with vaccination. 26.7% had not had any subsequent vaccinations withheld and only 1 was reported to have developed a repeat episode of HHE. This event followed hepatitis B vaccine.

In the 12 cases aged > 2 years the median time to onset was 5 minutes, compared with 215 minutes in the younger children. The shorter median time to onset suggests that a larger proportion of the cases in the older children were vasovagal reactions.

Because of the switch to acellular pertussis vaccines a reduction in rate of reporting of HHE was observed with time and a concomitant increase in the proportion of cases associated with vaccines containing acellular pertussis.

Dr Coulter advised that the CARM reports of HHE had a median onset time of 3 hours (180 minutes) which is similar to the median observed in this study.

Members observed that the data confirmed that subsequent vaccination need not be withheld following an HHE reaction. Advice given about subsequent vaccination in a Prescriber Update article (Apr 1998) about HHE had included continuing with the vaccination schedule as an option. During 2000 the pertussis-containing vaccine in the immunisation schedule had been changed to an acellular vaccine.

Prior to the switch to acellular vaccine, free acellular pertussis was made available for subsequent vaccinations for infants who experienced specified severe adverse reactions following immunisation. During the period (8.5 months, 6 Dec 1999 to 13 Aug 2000) of this arrangement 30 reports of HHE were received. Since the switch to acellular vaccine the number of adverse reaction reports had fallen almost to zero, but one report of HHE with the new vaccine had been received by the date of the meeting.

7.2.2 Adverse reactions reported with acellular pertussis vaccines

- M M Braun, et al. Infant immunization with acellular pertussis vaccines in the United States: assessment of the first two years' data from the vaccine adverse event reporting system (VAERS). Pediatrics 106, Oct 2000

Braun et al conducted a study using the US VAERS looking at the impact on adverse reaction reporting of the switch to acellular pertussis vaccine. The rate of reporting of adverse reactions fell from 2071 in 1995 to 491 in the first half of 1998. Rates of reporting of both serious and less serious events fell, but the rate of reporting of deaths was not significantly changed. No events were found to be reported at a clinically significantly higher frequency with DTaP than with DTP or DTPH. No reports associated with DTaP mentioned anaphylaxis. There were 92 reports of urticaria with the acellular vaccine, and 2 of these included respiratory symptoms, but in neither case was the infant hospitalised. Febrile seizures were reported in 39 cases in 1995 and in 20 cases in the first half of 1998. The study did not indicate total numbers for the acellular vaccine, and the authors were not able to conclude that the rate of febrile seizures was lower with the acellular vaccine than with DTP or DTPH.

Members recommended that the articles about HHE and rates of reporting of adverse reactions with acellular vaccine be passed on to the immunisation committee and to those in the Public Health Directorate who have responsibility for the immunisation programme.

7.2.3 Hepatitis B and Hib vaccines and diabetes

- F Destefano, et al. Hepatitis B and Hib vaccines are not associated with increased risk of type 1 diabetes. 40th Interscience conference on Antimicrobial Agents and Chemotherapy, 17 Sep 2000. Reactions abstract

Destefano found no association between administration of hepatitis B or Hib vaccines and type 1 diabetes in a study including 260 children with diabetes and 795 matched controls. No evidence was found that age at administration of hepatitis B vaccine was a factor.

It was noted that these results were in line with the article about vaccination and type 1 diabetes mellitus published in the February 2000 issue of Prescriber Update.


8.1 Montelukast

- D Coulter. Montelukast. Comments on interim monitoring summary

Dr Coulter commented that most patients are taking montelukast for asthma and/or chronic obstructive respiratory disease, but it also is being used by small numbers of patients for a range allergic conditions including rhinitis, eczema and urticaria. To date 34% of withdrawals have been due to an inadequate therapeutic response. Age-wise the highest proportion of users is in the age range 40-59 years and only 12% are children.

8.2 Nefazodone

- D Coulter. Nefazodone. Comments on interim monitoring study

More than 3000 patients are recorded in the IMMP database as nefazodone recipients. Despite the restrictions on its use, patient numbers are increasing at a greater rate than for fluoxetine when it first became available.

The 2 most commonly reported types of events are those of hypotension and disturbance of liver function. Visual disturbances also feature prominently. Of these events, only hypotension occurs in the top 8 adverse reactions for moclobemide, and none feature for fluoxetine. Treatment had ceased in 8.4% of patients. Of this group, 38 % were lost to follow-up and 21% had an inadequate therapeutic response.

8.3 Diane 35

- DM Coulter. Proposal to monitor Diane 35 on the IMMP

Dr Coulter had put together a proposal for the monitoring of Diane 35 on the IMMP in order to obtain an estimate of the incidence of VTE. The cost for the most desirable option was estimated at $170,000 per annum and monitoring would probably need to be continued for 5 years. The study would be without controls which would reduce its usefulness. Members commented that a cohort study such as the IMMP is an inefficient method for studying an event as rare as VTE with Diane 35.

The Committee did not support proceeding with the proposed IMMP monitoring. It was suggested that when record linkage between prescriptions and hospital data becomes a possibility, the feasibility of a study of VTE and Diane 35 could be examined.

The Committee recommended setting up a project team to evaluate possible studies using record linkage.

9. Other NZ activities

9.1 Minutes of PTAC (PHARMAC)

- Minutes of Pharmacology and Therapeutics Advisory Committee meeting, 17 Aug 2000

The Committee raised the issue of cross-membership of committees, and recommended that Medsafe explore with Pharmac the possibility of having a member of MARC on PTAC or visa versa, or having mutual observer status.

9.2 CARM report on development and activities

- DM Coulter, M Tatley. CARM report, November 2000

Members noted the positive tone of the report, and the progress made.

9.3 ACC medical error levy

- P MacKay. Medical misadventure revisited. NZMA Newsletter 24 Nov 2000

The Committee expressed some concern about the proposal by ACC to initiate a levy on medical practitioners that would cover all adverse outcomes including adverse outcomes resulting from medicines properly prescribed and administered. Members were concerned that such a levy may have an affect on adverse reaction reporting, because of the implied culpability.

The Committee recommended writing to ACC to ask what it intends to do and to find out what would be an appropriate process by which a submission could be made by the Committee.

10. Other countries

10.1 Canada

- Canadian Adv Drug React Newsletter 10(4), Oct 2000

10.2 United Kingdom

- Current Problems in Pharmacovigilance 26, Sep 2000

10.3 Overseas conference attendance

10.3.1 Consultation at Wake Forest University 5-6 Oct 2000

- DM Coulter. Developing an optimal approach to drug safety. Report on conference
- DM Coulter. Intensive Medicines Monitoring Programme.
- JJ Mohr. A systems approach to reducing medication errors.
- P Honig. AERS and regulatory perspectives on the value of spontaneous adverse event reports.
- CERT's role in improving the safe use of medications

The impact of medication errors and ways to reduce them were discussed at the meeting at Wake Forest University. Specialised units with no association with any function that may impute blame are required to work in the area of the reduction in medication errors.

The Committee asked that papers be gathered on this subject and that it be placed on the agenda for a meeting in 2001.

The editor of the New England Journal of Medicine addressed the need to convey information in a way that results in change. In his view data sheets are an inadequate means of communicating messages on risk for both medical practitioners and patients.

10.3.2 Drug Safety Research Unit, Southampton

- DM Coulter. Visit to the Drug Safety Research Unit, Southampton, 8-9 Nov 2000
- LV Wilton. Prescription event monitoring. Presentation DM Coulter. IMMP (not included but similar to presentation at WFU)

10.3.3 WHO meeting in Tunis, 11-14 Nov 2000

- DM Coulter. WHO Drug Monitoring Programme. Report on attendance at the 23rd annual meeting of national centres.
- IR Edwards. UMC Director's annual report
- E Vinge. Comment on rofecoxib and cardiac events
- Sten Olsson, Uppsala Monitoring Centre. Rofecoxib and cardiac adverse reactions. Email 20 Nov 2000
- New study confirms the GI safety advantage of rofecoxib. MI differences in the VIGOR study. Pharm J 264:835, 3 Jun 2000
- DM Coulter. Prescription event monitoring. The New Zealand Intensive Medicines Monitoring Programme.
- DM Coulter. Sumatriptan and cerebrovascular events
- DM Coulter. Cardiac failure/oedema with omeprazole

Dr Coulter mentioned issues that had been discussed at the WHO meeting. Bupropion and seizure

Bupropion (Zyban) which is now approved in many countries as an antismoking aid, but was originally available as an antidepressant was a concern for some countries. In Canada about 10% of reports for bupropion are of seizures. Dr Jessamine advised that bupropion was not approved as an antidepressant in New Zealand. Medsafe received a new medicine application for the indication of antismoking aid, and it was considered as such by the MAAC. The dossier included analysis of the risk of seizure. It is included as an adverse reaction in the entry for Zyban in New Ethicals Catalogue. CARM reports for bupropion have mostly been of nervous system (8) or psychiatric (7) events. Rofecoxib and cardiac events

In the Netherlands, 54 reports of adverse reactions with rofecoxib had been received. Five of the events were cardiac and all of these occurred within 3 hours of taking rofecoxib. The outcome was fatal in 2 cases, including a death which occurred 2 hours after the first tablet in a woman aged 69 years with a history of angina.

In one study (VIGOR) which has not yet been published rofecoxib was associated with a higher rate of myocardial infarction than naproxen (0.4% vs 0.1%, in a trial with 4047 vs 4029 patients). It has been proposed that the difference reflects an antiplatelet and cardioprotective effect of naproxen. If this is so the protective effect is greater than that seen with aspirin (75% reduction).

10.3.4 European Society of Pharmacovigilance, Verona, 21-23 Sep 2000

- N Rafter. Report for Medsafe on attendance at the meeting of European Society of Pharmacovigilance, Verona, Italy, 21-23 Sep 2000-11-29

11. Publications

Coulter DM. The NZ IMMP in pro-active safety surveillance. Pharmacoepidemiology and Drug Safety 2000; 9:273-280.

Clark D, Morgan A, Hananeia L, Coulter D, Olds R. Drug Metabolism Genotypes and their Association with Adverse Drug Reactions in selected Populations: a Pilot Study of Methodology. Pharmacoepidemiology and Drug Safety 2000; 9:393-400.

12 Current topics

12.1 The safety of oral contraceptives

- K Ronaldson. Oral contraceptives. Report for MARC 24 November 2000

12.1.1 Oral contraceptives and venous thromboembolism Perioperative use of Ocs and HRT

- Drugs in the peri-operative period: hormonal contraceptives and hormone replacement therapy. Drug and Therapeutics Bulletin 37:78-80, Oct 1999

Members noted the article about use of OCs and HRT in the perioperative period. The Committee recommended writing to the editor of the Drug and Therapeutics Bulletin and seeking permission to publish the article. Comment in New Zealand publications

- P Egermeyer. Correspondence. New Ethicals Journal p.78, Nov 2000
- S Jessamine. Correspondence. New Ethicals Journal p.78 & 28, Nov 2000
- F Goodyear-Smith, B Arroll. The pill scare: publicising remote risks exposes women to other dangers. NZ Family Phys 27:33-6, Oct 2000
- D Skegg. Oral contraceptives and venous thromboembolism. NZMJ 385, 8 Sep 2000
- P MacKay. The great pill debate - Is there a generation gap. NZMA Newsletter No.244, 8 Sep 2000

The correspondence published in the New Ethicals Journal was also noted. It was agreed that no further action was required.

The article in New Zealand Family Physician by F Goodyear-Smith and B Arroll about the history of the issue of VTE with second and third generation oral contraceptives in New Zealand was considered to be factually inaccurate in some places.

Dr P MacKay in the NZMA Newsletter was critical of the MARC for not having sought advice from the NZMA or RANCOG concerning the project to advise of the risk of VTE with oral contraceptives. It was noted that at the time of Dr MacKay's comment no action had been taken on this project. However, letters had since been sent to NZMA, FPA, RANZCOG and RNZCOGP on the recommendation of the September meeting (see 5.1.6).

Members considered that it was inappropriate for the NZMA Newsletter to be included in the New Zealand Medical Journal which is a peer reviewed publication. The Committee asked that the Chairperson write to the editor of the publication suggesting that the newsletter no longer be included. Studies in the international literature

- H Jick, et al. Risk of venous thromboembolism among users of third generation oral contraceptives compared with users of oral contraceptives with levonorgestrel before and after 1995: cohort and case-control analysis. BMJ 321:1190-5, 11 Nov 2000
- DCG Skegg. Pitfalls of pharmacoepidemiology. Oral contraceptives show a need for caution with databases. BMJ 321:1171-2, 11 Nov 2000
- R Farmer, T Williams, A Nightingale. Pitfalls of pharmacoepidemiology. BMJ 321:1352, 25 Nov 2000

Jick et al have conducted a rigorous study of the risk of idiopathic VTE with third generation oral contraceptives and with those containing levonorgestrel. Their study used the United Kingdom General Practice Research Database (GPRD) and spanned the period January 1993 to December 1999, i.e. before and after the announcement in October 1995 in the UK about the higher rate of VTE with third generation pills. Furthermore, their source of data was the same as that for the Farmer study2 (considered at the September 2000 meeting) which found no significant change in the incidence of VTE before and after the UK announcement, despite the sharp decline in the use of the third generation pills (from 53% to 14%).

The Jick study found the adjusted incidence ratio for VTE for the period prior to the announcement was 2.0 (1.2-3.4) and after the announcement it was 1.7 (0.8-3.5) for third versus second generation OCs. A case-control analysis included in the same paper (106 cases and 569 controls) found a relative risk of 2.3 (1.3-3.9) for third compared with second generation OCs. The higher odds ratio in the case-control study is because this design allows for better control for confounding factors (BMI, smoking, duration of use and those related to calendar time of diagnosis).

Jick et al estimated that, in their study population which was only a fraction of users of OCs in the UK, 9 fewer cases of VTE had occurred after October 1995 than would have occurred had women not switched to an OC containing levonorgestrel. This is a 26% reduction in number of cases. The authors made the following observations that would contribute towards the reduction in incidence for the later period not being larger:

In his editorial Professor Skegg noted a further factor: the predominant reduction in use of third generation pills occurred among younger women who are subject to a lesser background risk of VTE and hence the reduction in incidence would not be so marked in this group as in an older group.

Jick et al individually reviewed clinical information on each case and identified > 12 cases for exclusion because of the presence of proximate causes (the presence of these cases causes results to tend towards a finding of no difference in risk3 and excluded further cases because the diagnosis was not confirmed. Jick claimed that Farmer did not conduct this manual review of cases, but Farmer in his response said records were reviewed individually by the same method used by Jick et al.

Jick's analysis was restricted to third generation OCs and to those containing levonorgestrel, and was not obscured as Farmer's was by inclusion of all OCs, including some for which the risk of VTE is uncertain, and may be greater than that for third generation OCs (e.g. Diane 35). Furthermore, from many studies using the GPRD, Jick has excluded those practices which do not have high quality data.

The editorial note by the editor of BMJ commented that the journal's statistician was concerned by the adequacy of the age adjustment in the Farmer study and by the power of the study to detect a difference in risk as large as 50%. Note that the Jick study found a reduction in incidence of only 26%.

Members commented that the Jick study was probably the most rigorously conducted of all the studies examining the relative risk of VTE for third and second generation OCs.

The Committee noted that the results of the Jick supported the view of the Committee at previous meetings and the view promulgated by means of articles in Prescriber Update and Dear Doctor letters. The Committee agreed that there was no need to change its advice and it did not wish any other action to be taken at present, on the basis of these new data.

12.1.2 Oral contraceptives and haemostatic effects on coagulation

- S Middeldorp, et al. Effects on coagulation of levonorgestrel- and desogestrel-containing low dose oral contraceptives: a cross-over study. Thromb Haemost 84:4-8, 2000
- J C M Meijers, et al. Increased fibrinolytic activity during use of oral contraceptives is counteracted by an enhanced factor XI-independent down regulation of fibrinolysis. Thromb Haemost 84:9-14, 2000
- G Tans, et al. A randomized cross-over study on the effects of levonorgestrel- and desogestrel-containing oral contraceptives on the anticoagulant pathways. Thromb Haemost 84:15-21, 2000

Three randomised controlled crossover studies have been conducted evaluating the effect of combined OCs containing levonorgestrel or desogestrel on clotting factors. All 3 found that an OC containing desogestrel significantly increased levels of factors promoting clotting or inhibited endogenous fibrinolysis. The results of each of the studies were consistent with a greater risk of clotting with combined OCs containing desogestrel compared with those containing levonorgestrel. These studies provide a possible biological basis for a greater risk of VTE with third generation OCs, but they do not link these findings with clinical outcomes.

12.1.3 Oral contraceptives and breast cancer

- D M Grabrick, et al. Risk of breast cancer with oral contraceptive use in women with a family history of breast cancer. JAMA 284:1791-98, Oct 2000
- W Burke. Oral contraceptives and breast cancer. A note of caution for high-risk women. JAMA 284:1837-38, Oct 2000

Grabrick et al have conducted a study looking at the relative risk of breast cancer in women taking oral contraceptives who have a strong family history of breast cancer. They used a study cohort of 3396 women with blood relatives with breast cancer, and 2754 women who married into families with a history of breast cancer. 51% of each group had used oral contraceptives (OCs) at some time (p = .99).

The relative risk of breast cancer associated with use of OCs was 1.4 (95% CI 1.0-2.0) for the entire cohort. For sisters and daughters of women with breast cancer the relative risk was 3.3 (1.6-6.7) with use of OCs. The relative risk did not change with duration of use or time since last use, or other similar parameters. When the analysis was limited to families with at least 3 blood relatives with breast cancer the relative risk for sisters and daughters was 4.6 (2.0-10.7), and for families with 5 blood relatives it was 11.4 (2.3-56.4). The risk was not elevated for more distant relatives or for marry-ins. No association was found between OC use after 1975 and breast cancer, but after 1975 there were only 2 cases of breast cancer among women with first degree relatives with breast cancer. All oral contraceptives introduced after 1975 contained 50µg or less of ethinyloestradiol.

The authors commented that the lack of apparent elevated risk in women using OCs after 1975 may be associated with the shorter period of follow-up and younger age of these women. They plan to screen women for the presence of BRCA1 and BRCA2 to ascertain whether the presence of these genes is responsible for the modifying effect of family history on risk of breast cancer.

In his editorial, Burke points out the dilemma raised by this study: women with a high risk of ovarian cancer have been advised to take OCs to reduce their risk. This advice is based on the results of population-based studies, and relates particularly to women with the BRCA1 and BRCA2 mutations. However, the present study suggests that this same group of women may have an elevated risk of breast cancer with the use of OCs.

Members observed that the Oxford collaborative reanalysis4 and the New Zealand study5 which had larger numbers of cases with a family history of breast cancer found no independent increase in risk with use of OCs. It may be that the increased risk is not associated with use of low dose pills.

The Committee asked that any new data on breast cancer risk with OC use in women with a family history of breast cancer, or women with the BRCA1 and BRCA2 mutations be brought to its attention. No other action was recommended at this time.

12.1.4 Oral contraceptives and cardiovascular events

- P Hannaford. Cardiovascular events associated with different combined oral contraceptives. A review of current data. Drug Safety 22:361-71, May 2000

Members observed that the review by Dr Hannaford covering myocardial infarction, ischaemic and haemorrhagic stroke and VTE was an excellent summary of the area.

The Committee recommended writing to the editor of New Ethicals Journal to suggest that the article be republished there.

12.2 Inhaled corticosteroids and growth retardation

- N Rafter. Inhaled corticosteroids and growth suppression. Report for MARC, 6 Nov 2000
- The Childhood Asthma Management Program Research Group. Long-term effects of budesonide or nedocromil in children with asthma. NEJM 343:1054-1063, Oct 2000
- L Agertoft, S Pedersen. Effect of long-term treatment with inhaled budesonide on adult height in children with asthma. NEJM 343:1064-1069, Oct 2000
- M E B Wohl, J A Majzoub. Asthma, steroids, and growth. NEJM 343:1113:1114, Oct 2000

The Childhood Asthma Management Program Research Group trial was designed to evaluate whether continuous long-term treatment (4-6 years) with either inhaled corticosteroids (budesonide) or inhaled non-corticosteroids (nedocromil) safely produces an improvement in lung function, compared with placebo. All children received treatment for symptoms with salbultamol and oral prednisone as required. The effect of treatment on growth in stature was studied as a secondary outcome. Between 1993 and 1995 1041 children, aged from 5-12 years, were enrolled from 8 centres. In 1999 the study medication was discontinued and the children returned after 2-4 months for final spirometry and methacholine challenge. Mean age at baseline was around 9 years with a diagnosis of asthma made 5 years prior. Height at baseline was greatest in the budesonide group (56.8 percentile) and lowest in the placebo group (55.3 percentile).

At the end of the treatment period the increase in height in the budesonide group was 1.1cm less than placebo (change in height 22.7 vs. 23.8cm p=0.005). Nedocromil and placebo had similar height increases. Differences in rate of growth between the budesonide group and the other 2 groups occurred within the first year of treatment. All groups had similar growth velocity by the end of treatment and similar change in bone density. All groups by the end of treatment had the same bone age, projected final (adult) height and Tanner stage.

In the study by Agertoft and Pedersen, 332 children with asthma, which was controlled without continuous inhaled corticosteroids, were offered enrolment in the trial and a change to treatment with budesonide. 270 accepted and the rest acted as controls. Healthy siblings of the budesonide group were added to the controls. Average treatment duration was 9.2 years.

The primary outcome was measured adult height in relation to target adult height. Height was measured by Harpenden stadiometer. Any other asthma medication that was required was added. 30 of the original control group changed to budesonide. If a child received prednisone for longer than an average of 2 weeks per year they were excluded from the analysis of adult height.

More than 95% of all children attained an adult height that was within 9cm above or below their target adult height. The mean differences between the measured and target adult heights were +0.3cm for the budesonide group, -0.2cm for the controls with asthma, and +0.9cm for the healthy siblings. However, budesonide was associated with a significant change in growth rate during the first years of treatment. The initial growth retardation was significantly correlated with age (p=0.04), with a more pronounced reduction in younger children. The adult height depended significantly (p<0.001) on the child's height before budesonide treatment. There was no dose-response relationship.

Members commented that these are the first long term studies of any size looking at the effect of inhaled corticosteroids on growth velocity and adult height. It was noted that the main effect occurred in the first year of use, and that some, mainly younger children, appeared to be more sensitive.

The Committee had already placed a watching brief on this topic (12.4.1, Jun 2000). Members recommended no further action on this topic at this time.

12.3 The safety of antipsychotic medication

12.3.1 VTE with antipsychotic medication

- G L Zornberg, H Jick. Antipsychotic drug use and risk of first-time idiopathic venous thromboembolism: a case-control study. Lancet 356:1219-23, Oct 2000

Zornberg and Jick have conducted a case-control study using the GPRD to ascertain whether there is an increased risk of VTE with antipsychotics. Their study included 42 cases and 168 matched controls who were younger than 60 years and had used a conventional or atypical antipsychotic at some time. The study found a relative risk of 7.1 (2.3-21.9) for VTE in current users (14 cases) compared with non-users of conventional antipsychotics. The risk was greatest in the first 3 months of therapy with 10 of the cases occurring in this period. Use of oestrogens and hypertension were independent risk factors.

Members commented that pulmonary embolism is notoriously difficult to diagnose. Psychiatric illness is associated with higher than the background rate of sudden death. Some of these patients may have died of undiagnosed pulmonary embolism.

The Committee recommended gathering more information about this subject to ascertain whether it would be appropriate to put out some advice.

12.3.2 Clozapine and interstitial nephritis

- D Fraser and M Jibani. An unexpected and serious complication of treatment with the atypical antipsychotic drug clozapine. Clin Nephrology 54:78-80, 2000
- K E Southall, S N Fernando. A case of interstitial nephritis on clozapine. Aust NZ J Psychiatry 34:697-698, Aug 2000
- Clozaril data sheet, extract from adverse effects.

Two cases of interstitial nephritis with clozapine had been reported. One case had occurred in New Zealand. In one case the patient had no evidence of renal insufficiency. The diagnosis was made on the basis of abnormal urinary microscopy. The data sheet for clozaril notes that isolated cases of interstitial nephritis have been reported in association with clozapine. The Committee recommended no action.

12.4 The safety of anticoagulants

- A T Gerlach, et al. Enoxaparin and bleeding complications: a review in patients with and without renal insufficiency. Pharmacotherapy 20:771-5, 2000
- Clexane data sheet. Extract on treatment in renal insufficiency
- C Pohl, et al. Neurological complications in heparin-induced thrombocytopenia. Lancet 353:1678-79, May 1999
- L M Brass, et al. Intracranial hemorrhage associated with thrombolytic therapy for elderly patients with acute myocardial infarction. Results from the cooperative cardiovascular project. Stroke 31:1802-11, Aug 2000

Gerlach et al retrospectively examined the rate of bleeding complications with enoxaparin treatment in patients with (n = 53) and without (n = 50) renal impairment. 22% of those with normal renal function and 51% with renal impairment developed bleeding complications (p<0.01). Major bleeds occurred in 2% versus 30 % (p<0.001). The rate of bleeding complications appeared to be unrelated to dose.

The data sheet for Clexane (enoxaparin) states that no dosage adjustment is required in renal impairment in doses up to 60mg daily.

Members considered it was not possible to make a recommendation on the basis of these data.

Pohl et al identified 105 consecutive patients with heparin-induced thrombocytopenia. Ten of these (9.5%) had neurological complications. Five of this group died, and 2 had residual deficits. All those with neurological complications were on unfractionated heparin.

Brass et al found a rate of intracranial haemorrhage of 1.43% (455 of 31,732). The following were found to be independent risk factors: age ≥ 75 years, female, black race, prior stroke, blood pressure ≥ 160mm Hg, use of tissue plasminogen activator, excessive anticoagulation, and less than median weight.

No action was recommended on these issues at present.

12.5 Celecoxib

- K Ronaldson. Celecoxib. Report for MARC, 22 Nov 2000

A member advised the Chairperson of a conflict of interest, but was permitted to participate in the discussion and decision making.

12.5.1 Upper gastrointestinal ulcer

- J L Goldstein, et al. Reduced risk of upper gastrointestinal ulcer complications with celecoxib, a novel COX-2 inhibitor. Am J Gastro-enterology 95:1681-1690, July 2000 - FE Silverstein, et al. Gastrointestinal toxicity with celecoxib vs non-steroidal anti-inflammatory drugs for osteoarthritis and rheumatoid arthritis. JAMA 284:1247-55, 13 Sep 2000 - DR Lichtenstein, MM Wolfe. COX-2-selective NSAIDs. JAMA 284:1297-9, 13 Sep 2000

Two large studies of celecoxib and upper gastrointestinal ulcer have been published. Both have industry funding, and both are co-authored by Jay Goldstein and Fred Silverstein, plus others.

The study by Goldstein et al was a pooled analysis of 14 multicentre double-blind randomised controlled trials (total celecoxib recipients 6,376, placebo 1,864 and NSAIAs 2,768) and an open-label extension for up to 2 years' treatment with celecoxib (5,155 patients) in patients with osteoarthritis (OA) or rheumatoid arthritis (RA). None of the patients had known active GI disease, but around 2% had a history of gastrointestinal bleeding and 12% of ulcer disease. About 12% were taking aspirin concurrently.

A total of 215 potential ulcer complications were reported, and 20 were judged to meet the study criteria for upper GI ulcer complications. In the randomised trials, no cases occurred among those treated with placebo, 2 occurred with celecoxib (annualised incidence 0.20%) and 9 occurred with NSAIAs (1.68%; p = 0.002). The remaining 9 cases occurred in the open label trial, at an annualised incidence of 0.18%. The rates observed with NSAIAs are similar to those found in other trials. The incidence of GI ulcer complications found with celecoxib is similar to that observed in other studies of patients with similar age and disease characteristics who were not taking an NSAIA.

The study by Silverstein et al was a large scale randomised controlled trial including 2376 celecoxib recipients and 2197 NSAIA recipients all with RA or OA who completed 6 months' treatment. No patients had symptomatic ulcer or ulcer complications at entry. An unusually high number of patients were taking aspirin (20.9% in the celecoxib group and 20.4% in the NSAIA group). About 1.6% had a history of GI bleeding and 8.2% a history of GI ulcer. Celecoxib was taken at 400mg twice daily (twice the maximum recommended dose in New Zealand). Ibuprofen and diclofenac were taken at standard doses.

The annualised incidence of upper GI ulcer complications was 0.76% in the celecoxib group and 1.45% in the NSAIA group (relative risk 0.53; 95% CI 0.26-1.11). This difference was not statistically significant. The relative risk of upper GI ulcer complication in the celecoxib group with aspirin versus no aspirin was 4.5 (6 events among aspirin users and 5 among non-aspirin users; p = 0.01). Upper GI ulcer complication incidence was not so strongly affected by aspirin use in the NSAIA group (RR 1.7). The annualised incidence in non-aspirin users was 0.44% in the celecoxib group and 1.27% in the NSAIA group (p = 0.04).

This study also recorded the incidence of other adverse effects and found that only rash and pruritus were significantly more frequent with celecoxib than with the NSAIAs. No serious rashes occurred with celecoxib. At 6 months the rate of clinically significant reductions in haematocrit and/or haemoglobin with celecoxib was approximately half that with NSAIAs (1.3% vs 3.4% in patients not taking aspirin). The difference in rates of elevated liver function enzymes was even more striking (ALT ≥ 3 ULN: 0.19% for celecoxib and 1.7% for NSAIAs; p < 0.001).

The authors commented that this study does not address the adverse reactions that may occur over longer periods of treatment than 6 months nor those that increase in frequency with co-morbid conditions.

The editorial by Lichtenstein and Wolf discussed possible adverse effects of COX-2 inhibitors which have a theoretical basis but which have not yet been seen clinically: e.g. delayed ulcer healing with pre-existing GI ulcer, thromboembolic events and exacerbation of pre-existing inflammatory bowel disease.

Members observed that in the open label arm of the first study, those who developed adverse reactions in the first phase would have dropped out, so that the incidence calculated would be low.

12.5.2 Interaction with warfarin?

- JD Linder, et al. Cyclooxygenase-2 inhibitor celecoxib: a possible cause of gastropathy and hypoprothrombinaemia. South Med J 93:930-2, Sep 2000

Linder et al describe a case of gastropathy and hypoprothrombinaemia in a 71-year-old man who began celecoxib 7 days previously and who was taking a range of medications including warfarin (INR within therapeutic range for previous 12 months). He had black tarry loose stools and an INR of 4.00. The authors suggested there may have been a metabolic interaction between celecoxib and warfarin due to the range of medication the man was taking, but it was not possible to be certain that he did not have pre-existing ulcer disease. Although a study in healthy volunteers found no interaction between warfarin and celecoxib, the situation may be different in patients taking warfarin for a therapeutic effect, particularly if they are on other medication. The authors recommend using other medicines for pain control in patients taking warfarin.

This case is similar to one reported to CARM (haematuria and prothrombin time 3.1; other medication sotalol and paracetamol; case no. 43355).

12.5.3 Blood dyscrasias

- Celecoxib and aplastic anaemia. ADRAC minutes, item 10.7, 28 July 2000

The ADRAC data suggest that pancytopenia and related blood dyscrasias may occur with celecoxib, but causality assessment would be assisted by comparative data.

12.5.4 Renal failure

- I W Boyd, et al. COX-2 inhibitors and renal failure: the triple whammy revisited. MJA 173:274, Sept 2000

The article by Boyd et al (also from ADRAC) describes 11 cases of acute renal failure or worsened chronic renal failure with COX-2 inhibitors. Six of the patients were also taking a diuretic and an ACE inhibitor. Two additional reports mentioned one or other of these and in another report the patient was on a diuretic and an angiotensin-receptor antagonist. NSAIAs, diuretics and ACE inhibitors are individually and in combination linked to acute renal failure. The authors warned against use of COX-2 inhibitors in renal impairment, and commented that concomitant use of a diuretic or an ACE inhibitor increases the risk of renal complications.

12.5.5 Recommendations

The Committee asked that

12.6 Tamoxifen and endometrial cancer

- L Bergman, et al. Risk and prognosis of endometrial cancer after tamoxifen for breast cancer. Lancet 356:881-7, 9 Sep 2000
- K Gelmon. One step forward or one step back with tamoxifen? Lancet 356:868-9, 9 Sep 2000
- L Deigdisch, et al. Endometrial histopathology in 700 patients treated with tamoxifen for breast cancer. Gynecologic Oncology 78:181-6, Aug 2000. Reactions abstract
- K Ronaldson. Tamoxifen and endometrial cancer. Report for MARC, 23 Nov 2000

Bergman et al have conducted a case-control study of endometrial cancer in Dutch women with breast cancer who have taken tamoxifen. Each of the 309 cases had developed endometrial cancer at least 3 months after being diagnosed with breast cancer. Controls had a diagnosis of breast cancer and intact uterus but no diagnosis of endometrial cancer. They were matched with the cases for date of birth and date of diagnosis of breast cancer, and they must have survived at least as long after breast cancer diagnosis as the matching case. Cases were more likely than controls to have advanced and metastatic breast tumours (16.4% vs 9.2%; p = 0.02) and to have received other hormone therapy for breast cancer (p = 0.04).

The unadjusted relative risk of endometrial cancer with recent use of tamoxifen (≤ 12 months since last use) was 1.5 (95% CI 1.1-2.1), and the risk increased with duration of use (p<0.001). The risk was similar for recent and ex-users, but only 5% of cases had used tamoxifen > 2 years previously. The relative risk associated with 2-5 years use of tamoxifen and diagnosis 2-5 years after breast cancer diagnosis was 2.4 (1.2-4.8).

Endometrial cancer of FIGO stage III and IV was more frequent among long term (≥ 2 years) users of tamoxifen than non-users (17.4% vs 5.4%; p=0.006). Users were also more likely to present with malignant mixed mesodermal tumours or sarcoma of the endometrium than non-users (10.6% vs 2.9%), and the risk increased with duration of use. Furthermore, oestrogen-receptor status was more likely to be negative in users of tamoxifen (p≤0.001); 26.2% of non users had endometrial cancer with no oestrogen receptors vs 60.8% of those who had used tamoxifen for ≥ 2 years.

13.9% of deaths among non-users who were case patients were from endometrial cancer compared with 23.4% of those who used tamoxifen (33.3% for duration ≥ 2 years). Death from breast cancer was also higher among tamoxifen users (40.4% vs 25.0%) and overall death rate was higher (42.7% vs 36.2%). However, tamoxifen users generally had more advanced breast tumours.

It is likely that the matching for time in the selection process precluded detecting any change in risk with time since last use of tamoxifen. There was no evaluation of whether other factors associated with use of tamoxifen may have contributed to an increased rate of endometrial cancer with poor prognosis.

The authors considered that surveillance bias had not artificially elevated the rate of endometrial cancer in tamoxifen users because association with tamoxifen use was not found in the first year after diagnosis of breast cancer, nor among women who had used tamoxifen for < 2 years. Also the high prevalence of high grade tumours in long-term users is not consistent with surveillance bias.

In contrast to the results of the present study, all 36 invasive endometrial cancers that occurred in the Breast Cancer Prevention Trial (BCPT) were FIGO grade I, and most of these were diagnosed after ≥ 2 years' tamoxifen6. In the earlier study in the NSABP series, which was the first to confirm the association of tamoxifen with endometrial cancer in a randomised clinical trial, 88% of 24 endometrial cancers were localised and 78% of 23 were of grades I and II7. However, in this study there were only 2 cases of endometrial cancer in the placebo group and these 2 patients had taken tamoxifen. Hence, the seriousness of endometrial cancer cannot be compared with that in non-users.

Dr Gelmon, in her commentary, suggests that the results of the Dutch study may not be entirely relevant because many of the cases occurred before there was an awareness of an association between tamoxifen and endometrial cancer (breast cancer diagnosis 1972-1997). She contrasted these results to the low grade of invasive endometrial cancers found in the BCPT. But she also observed that surveillance is more rigorous in trials than in clinical practice and many of the endometrial cancers identified in the Dutch study would not be associated with clinical features to assist early diagnosis.

Members commented that the study results would appear to be preliminary in nature and require confirmation from other studies, because of variance from other studies and the oldness of the some of the data. No action was recommended.

12.7 Sedation with non-sedating antihistamines

- RD Mann, et al. Sedation with "non-sedating" antihistamines: four prescription-event monitoring studies in general practice. BMJ 320:1184-7, 29 Apr 2000
- JG Ramaekers, A Vermeeren. All antihistamines cross blood-brain barrier. BMJ 321:572, 2 Sep 2000
- K Ronaldson. Non-sedating antihistamines and sedation. Report for MARC, 24 Nov 2000
- Non-sedating anti-histamines and sedation. Minute of MARC, Mar 1995

12.7.1 Recent data

The Southampton Drug Safety Research Unit (Mann et al) has published the results of a study comparing the sedative effects of acrivastine, cetirizine, fexofenadine and loratadine using prescription event monitoring. The events were reported by the prescribing general practitioner. Recording of events depended on the patient mentioning them to the prescriber. Numbers of patients ranged from 7863 for acrivastine to 16,638 for fexofenadine. Acrivastine and cetirizine were associated with higher incidences of sedation with 7.9 and 8.5 events per 1000 patients, respectively, in the first month of treatment compared with 2.6 and 3.1 events per 1000 for loratadine and fexofenadine, respectively. Using loratadine as the comparator agent the adjusted (for age and sex) relative risk of sedation and drowsiness with cetirizine was 3.53 (p < 0.0001) and for acrivastine it was 2.79 (p < 0.0001). The number of events declined after the first month, but data permitting examination of reasons for this were not supplied (probably attrition of affected individuals, but could also reflect tolerance). The study found that the increased incidence of sedative effects with cetirizine and acrivastine was not associated with an increased rate of accident or injury.

The data collection for fexofenadine occurred 7 years after collection for the other medicines, but the authors commented that they were not aware of any publicity that would have affected the rate of reporting of sedation in these patients.

No events indicative of any cardiotoxicity were recorded for any of the patients.

The authors concluded:

Our findings suggest that in situations in which even very infrequent reports of sedation are undesirable loratadine or fexofenadine are preferable to acrivastine or cetirizine.

At a rate of 8.5/1000 (for cetirizine), 1 in 118 individuals would be affected by sedation.

Ramaekers and Vermeeren commented that all of the second generation antihistamines cross the blood brain barrier but to lesser degrees than the earlier antihistamines. In therapeutic doses terfenadine produces about 17% occupancy of frontal lobe histamine H1-receptors compared with 77% for chlorpheniramine. In driving tests which tracked deviation of lateral position, acrivastine and cetirizine affected driving at therapeutic doses. Fexofenadine and loratadine had measurable effects at twice the recommended dose.

12.7.2 History and NZ situation

The MARC considered a study by Ramaekers et al8 of the effects of antihistamines on psychomotor performance in March 1995. The authors of that study had concluded that the majority of patients taking a standard dose of a non-sedating antihistamine will not have impaired driving performance, but a small fraction of individuals will be affected and all should receive a warning.

The Medicines Regulations require a sedation warning on the labelling of all antihistamines unless the Director-General issues an exemption. At the meeting in March 1995 the MARC recommended use of the following or words of similar meaning for second generation antihistamines:

Although the medicine is unlikely to affect your ability to drive or operate machinery, a few people may be impaired and care should be taken.

This statement is listed in the Regulatory Guidelines. We require it to go on the package insert of all the second generation antihistamines, or on the package if there is no insert.

Since the launch of Zyrtec (cetirizine), Schering-Plough (loratadine; Claratyne) has contacted Medsafe repeatedly, the most recent occasion being this year and in connection with the Mann article, regarding the labelling of Zyrtec. Their request is that deletion of "non-sedating" be required from the carton of Zyrtec.

This topic was dealt with by postal consultation. All members agreed to the continued use of the above warning statement for all of the second generation antihistamines, but some members considered that some differentiation in labelling between the products would be appropriate. Hence, this topic was put on the agenda of the next meeting for discussion.

12.8 Items of interest

The following items were noted by members but were not discussed at the meeting.

12.8.1 Interactions

- J A Carrillo, J Benitez. Clinically significant pharmacokinetic interactions between dietary caffeine and medications. Drug Interactions 39:127-153, Aug 2000
- J Lill, et al. Cyclosporine-drug interactions and the influence of patient age. Am J Health-Syst Pharm 57:1579-84, Sep 2000

12.8.2 Result of paracetamol restriction in the UK

- D Robinson, et al. Severity of overdose after restriction of paracetamol availability: retrospective study. BMJ 321:926-7, Oct 2000

12.8.3 Minocycline and liver damag

- R A Lawrenson, et al. Liver damage associated with minocycline use in acne. Drug Safety 23:333-49, Oct 2000
- Minomycin data sheet. New Ethicals Compendium 7th edition, 2000

12.8.4 Nitrovasodilators and upper gastrointestinal bleeding

- A Lanas, et al. Nitrovasodilators, low-dose aspirin, other nonsteroidal antiinflammatory drugs and the risk of upper gastrointestinal bleeding. NEJM 343:834-839, Sept 2000

12.8.5 Adverse events associated with sedation in pediatrics

- C J Cote, et al. Adverse sedation events in pediatrics: analysis of medications used for sedation. Pediatrics 106:633-644, Oct 2000

12.8.6 Liver disease and antidiabetic agents

- S S Jick, et al. Frequency of liver disease in type 2 diabetic patients treated with oral antidiabetic agents. Diabetes Care 22:2067-2071, Dec 1999

12.8.7 Calcium channel blockers

- H T Sørensen, et al. Cancer risk and mortality in users of calcium channel blockers. A cohort study. Cancer 89:165-170, July 2000
- Furberg attacks calcium antagonists again. Scrip No.2570-1, p.26, 30 Aug/1 Sep 2000
- Experts condemn Furberg's meta-analysis. Scrip No.2574, p.22, 13 Sep 2000

12.8.8 Adverse medicine reactions in nursing homes

- J H Gurwitz, et al. Incidence and preventability of adverse drug events in nursing homes. Am J Med 109:87-94, Aug 2000

12.8.9 Doxazosin and sildenafil

- C Bankhead. Doxazosin and sildenafil: reassuring results emerge in recent trials. Reactions no.821, p.3-4, 30 Sep 2000

The meeting closed at 5pm.

0 1 2 4 5 6 7 9 [ /