The Women's Health Initiative (WHI) study, a large randomised trial comparing combination hormone replacement therapy (HRT) to placebo, found that the increase in risk associated with long term therapy exceeded the benefits.¹ In particular the promise of cardiovascular benefit from HRT was unfulfilled and therapy was found to increase the risk of cardiovascular events (see Table below).

A further surprising result of the WHI study was an increase in the incidence of dementia with HRT using oestrogen plus progestogen, and a failure to enhance cognitive function.² The WHI study did, however, demonstrate protection against fracture with oestrogen plus progestogen,^1,3 but this protection, even in women with the highest risk of fracture in the study, did not outweigh the other negative effects.³

The Million Women Study (MWS), which had a prospective observational design, further emphasised the risks of HRT, indicating a higher risk of breast cancer with oestrogen plus progestogen than with oestrogen alone (see Table).⁴ The results indicated that the increase in the incidence of breast cancer with oestrogen plus progestogen (compared to oestrogen alone) was greater than the reduction in incidence of endometrial cancer associated with adding progestogen to oestrogen therapy.⁴ The MWS also reported a significant increase in the incidence of breast cancer with tibolone and with implanted and transdermal oestrogen-only preparations.

Following its comprehensive review of these studies and other available data, the Australian Drug Evaluation Committee has recommended:⁵

"The use of HRT for any long term disease prevention cannot be generally justified as the potential harm may out weigh potential benefits. This concern also applies to the use of HRT to prevent osteoporosis.

"HRT has an established place in the short term management of symptoms of the menopause. For treatment of established osteoporosis, the selection of HRT by the patient and doctor should be based on a careful consideration and discussion of risks and benefits for that individual."

In addition, ADRAC advises that HRT use be for as short a time as practical, and be reviewed regularly.

References

1. Writing Group for the Women's Health Initiative Investigators. Risks and benefits of estrogen plus progestin in healthy postmenopausal women. Principal results from the Women's Health Initiative randomized controlled trial. JAMA 2002;288:321-33
2. Shumaker SA, Legault C, Rapp SR et al. Estrogen plus progestin and the incidence of dementia and mild cognitive impairment in postmenopausal women. The Women's Health Initiative Memory Study: a randomized controlled trial. JAMA 2003;289:2651-62
3. Cauley JA, Robbins J, Chen Z et al. Effects of estrogen plus progestin on risk of fracture and bone mineral density. The Women's Health Initiative randomized trial. JAMA 2003;290:1729-38
4. Million Women Study Collaborators. Breast cancer and hormone-replacement therapy in the Million Women Study. Lancet 2003;362:419-27
5. Australian Drug Evaluation Committee. Update to ADEC statement on use of hormone replacement therapy. 17 Oct 2003. http://www.health.gov.au/docs/html/hrtadec2.htm

Abbreviations: WHI, Women's Health Initiative Study; MWS, Million Women Study; O, oestrogen; P, progestogen.
Note: These event rates are age-dependent: mean age in WHI 63 years; mean age at recruitment for MWS 56 years

ADRAC has received reports of interactions between warfarin and all four macrolide antibiotics (azithromycin, clarithromycin, erythromycin, and roxithromycin) (see Table).¹ Although most cases were asymptomatic, some reports documented substantial increases in INR. The haemorrhagic complications reported included haematoma, haemoptysis, haematuria, melaena, and retroperitoneal haemorrhage

The one fatality reported involved a 79 year old woman who was started on warfarin and roxithromycin simultaneously. By day 8 of treatment, her INR had risen to 11.6. She subsequently died from widespread bleeding, including subdural haemorrhage and haemopericardium.

Nearly all reactions occurred in the first week after starting the antibiotic; reactions reported later than this may reflect the frequency of INR monitoring.

Table: ADRAC reports of macrolide-warfarin interaction

* metronidazole was another potentially interacting agent in 2 cases.
# onset > 1 year in a further patient.

Close attention should be paid to INR monitoring in patients taking warfarin who are commenced on a macrolide antibiotic. Consideration could also be given to the use of an alternative antibiotic if possible. Azithromycin has a particularly long half-life (around 68 hours), so that an interaction with warfarin may theoretically persist for some days after azithromycin has been ceased.

Reference

1. Interaction of warfarin with macrolide antibiotics. Aust Adv Drug Reactions Bull 1995;14:11.

A recent report to ADRAC described an elderly male with low bone mineral density in the hip who developed uveitis three weeks after beginning to take risedronate 35 mg once weekly. He later restarted risedronate but again developed eye pain. He then switched to alendronate 70 mg once weekly but eye pain occurred again.

The bisphosphonates have become important in the treatment and prevention of osteoporosis and are also used in the treatment of Paget's disease, hypercalcaemia and skeletal metastases.

Predominant adverse effects of the bisphosphonates are those affecting the gastro-intestinal system such as nausea, dyspepsia, abdominal pain and oesophageal disorders, and musculoskeletal effects such as arthralgia and myalgia. A small number of reports describe eye disorders, some of which were serious including uveitis (13 reports), iritis (6), scleritis/episcleritis (7), haemorrhage (4), optic neuritis (2), visual field defect (2) and one case each of scotoma, glaucoma, blindness and macular degeneration. These reports were associated with alendronate and pamidronate in 18 cases each, and with risedronate and zoledronic acid in one case each.

Of particular interest are the 28 reports that describe inflammatory reactions such as uveitis, iritis, scleritis, episcleritis and optic neuritis. Time to onset ranged from 2 days after the drug was commenced to over 3 years, with a median time of 3 weeks. As might be expected, most of the patients were female and almost half were elderly (range 48-79; median 63 years). Details of outcome were documented for 21 patients, of whom 15 had recovered at the time the report was submitted. Four of the six who had not recovered were reported to be improving although one of them required a trabeculectomy. Another patient had reduced visual acuity.

Inflammatory ocular disorders have been reported to ADRAC and in the literature only in association with alendronate, pamidronate, risedronate and zoledronic acid.^1,2 The risk may be higher with bisphosphonates administered intravenously such as pamidronate and zoledronic acid but otherwise the number of reports probably relates to usage. Minor reactions such as blurred or abnormal vision or conjunctivitis have been reported to ADRAC or overseas with etidronate, clodronate and tiludronate.^1,2

Inflammatory ocular disorders appear to be a rare effect of all bisphosphonates and prescribers should be aware that eye pain, redness and abnormal vision may be indicators of such disorders.

References

1. Fraunfelder FW, Fraunfelder FT. Bisphosphonates and ocular inflammation. N Engl J Med 2003;348:1187-8.
2. Uppsala Monitoring Centre. Bisphosphonates and ocular side effects. WHO Signal 2003;Dec:15-22.

Four statins (HMG CoA inhibitors) are available in Australia for the treatment of hypercholesterolaemia: simvastatin, atorvastatin, pravastatin and fluvastatin. Each of the statins may cause myalgia or rhabdomyolysis. Cerivastatin was removed from the market worldwide because of an unacceptably high rate of rhabdomyolysis, including fatal cases, particularly when used with gemfibrozil.¹

The rates of muscle disorders observed in clinical trials of statins have not been significantly different from those with placebo,² but wider clinical use involves individuals having multiple disease states or taking potentially interacting medication. Recent reviews indicate that factors which increase the plasma concentrations of statins are associated with an increase in the risk of myalgia, myopathy and, particularly, rhabdomy-olysis.^3,4 For simvastatin and atorvastatin which are metabolised by the liver enzyme CYP3A4 these factors are presented in Table 1.

ADRAC has received 91 reports of rhabdomyolysis with simvastatin and 26 with atorvastatin, as well as many reports of myalgia, myopathy or creatine kinase (CK) increase. Table 2 (top section) shows the percentage of cases with identified risk factors, as defined in Table 1. For simvastatin the factors listed most commonly in reports describing rhabdomyolysis were age ≥ 70 years (40 reports) dose ≥ 40 mg (33), cyclosporin (19), gemfibrozil (21), diltiazem (20) and diabetes (15). Over half of the simvastatin cases with rhabdomyolysis had more than one identified risk factor. Individuals with several risk factors may be at risk of developing rhabdomyolysis, rather than a less serious muscle disorder.

A feature of the cases of rhabdomyolysis is that long term statin therapy was well tolerated until after a change in medication (e.g. increase in the dose of statin, or addition of clarithromycin or diltiazem).

Table 1: Factors increasing the risk of muscle disorders with simvastatin and atorvastatin

Pravastatin and fluvastatin are not metabolised by CYP3A4 and are less subject to increases in plasma concentration by interaction with other drugs. ADRAC reports of muscle disorders with these statins are shown in Table 2 (bottom section). The dominant risk factors for pravastatin and fluvastatin were advanced age and high dose. The lower number of cases of rhabdomyolysis with these statins is probably associated with the lesser likelihood of drug interaction, but is also related to the lower usage in Australia (From 1992 to November 2003, 85% of statin prescriptions have been for simvastatin or atorvastatin).

High doses of statins should be used with caution in the elderly, in patients with renal or hepatic insufficiency, hypothyroidism or diabetes. Particular caution should be observed in patients taking simvastatin or atorvastatin with these conditions, if gemfibrozil, cyclosporin or diltiazem are being taken concomitantly. Consideration should be given to temporary discontinuation of simvastatin or atorvastatin, if short-term macrolide antibiotic or azole antifungal therapy is required. Patients should be advised to report to their doctor if muscle aches, pains or weakness develop.

References

1. Cerivastatin and rhabdomyolysis - avoid gemfibrozil. Aust Adv Drug Reactions Bull 2001;20(1):3.
2. Gotto AM. Safety and statin therapy. Arch Intern Med 2003;163:657-9.
3. Thompson PD, Clarkson P, Karas RH. Statin-associated myopathy. JAMA 2003;289:1681-90.
4. Ballantyne CM, Corsini A, Davidson MH et al. Risk for myopathy with statin therapy in high risk patients. Arch Intern Med 2003;163:553-64.

High dose cyproterone (50mg, 100mg; Androcur, Androcur-100) is used predominantly for advanced prostate carcinoma. For the year ending June 2003, 59,000 prescriptions were dispensed for the 50mg or 100mg tablets and 97% of patients prescribed these tablets were male.

Over the years, ADRAC has received 105 reports implicating high-dose cyproterone. The most common adverse reactions are related to the liver with 32 reports. Other more commonly reported reactions include fatigue, dyspnoea, asthenia, confusion, depression and deep vein thrombosis.

All except one of the hepatic reactions involved male patients being treated for prostate cancer, whose ages ranged from 56 to 92 (median: 77) years. Time to onset of liver dysfunction ranged from 4 days to 4 years (median: 4-5 months); only 4 cases had a time to onset under a month.

Where liver function test results were available, the majority indicated the presence of cholestatic hepatitis, but some showed marked elevation of hepatocellular enzymes (AST, ALT). Most patients were jaundiced. Ten patients died, nine due to hepatic failure. One patient with early stage pro-state carcinoma received a liver transplant. Eleven patients had recovered at the time of reporting.

While cyproterone-induced hepatotoxicity is rare, it can be fatal or life-threatening. Severe hepatic reactions have not been reported to ADRAC with the use of low dose cyproterone (1-2mg) in combination with ethinyloestradiol (Brenda, Diane, Juliet) or oestradiol (Climen).

Although the value of monitoring is not clear, it would be reasonable to monitor liver function tests intermittently in patients taking long-term high-dose cyproterone.

Serotonin syndrome is caused by excessive central nervous system and peripheral serotonergic activity. It most commonly occurs with a combination of serotonergic agents, but may also occur with a single agent. A combination of agents increasing serotonin by different mechanisms, such as by inhibition of serotonin uptake and serotonin metabolism, is associated with a high risk of the syndrome.¹ Table 1 lists agents which have been associated with serotonin syndrome.

Serotonin syndrome is a clinical triad of cognitive-behavioural changes, autonomic dysfunction and neuromuscular dysfunction. At least three of the features listed in Table 2 must be present.^1,2 There is no laboratory test to aid diagnosis. The syndrome often occurs within a day of a change in treatment (increase in dose or addition of another serotonergic agent) and the evolution of symptoms is rapid. It should not be confused with neuroleptic malignant syndrome which is clinically similar, but is an idiosyncratic response to neuroleptic agents, usually occurs after longer periods of treatment and develops over a period of days or weeks.¹

Table 1: Agents causing serotonin syndrome

ADRAC has received 161 reports of serotonin syndrome. The majority describe the syndrome in association with the concomitant use of 2 or more serotonergic agents, in particular SSRIs (68), tramadol (29), moclobemide (23), venlafaxine (18), tricyclic antidepressants (18) and St John's wort (8). In 61 reports, the serotonin syndrome developed in association with a single agent: SSRIs (40), moclobemide (5), venlafaxine (5) and tramadol (5). Serotonin syndrome with tramadol was the subject of an earlier Bulletin article.³

Table 2: Clinical features of serotonin syndrome

Serotonin syndrome is potentially serious. Reports to ADRAC have described confusion (31), convulsions (23), hypertension (22), hallucinations (12) and delirium (7). In the majority of reports, the signs and symptoms developed within 24 hours of the addition of another serotonergic agent or an increase in dose of an agent. Patients responded to withdrawal of the serotonergic agent(s) and appropriate treatment. Recovery was documented in 85% of the cases where the outcome was known and the remainder of patients had not recovered at the time of reporting.

Health professionals should note the drugs that may cause serotonin syndrome, alone or in combination with other serotonergic agents, and be alert to the features of serotonin syndrome. Patients should be informed of the risk and symptoms of serotonin syndrome when serotonergic agents are prescribed.

References

1. Langford N.J. Serotonin Syndrome. Adverse Drug Reaction Bulletin December 2002, No. 217
2. Sternbach H. The serotonin syndrome. Am J Psychiatry 1991;148:705-13.
3. Tramadol and serotonin syndrome. Aust Adv Drug Reactions Bull 1991;21:14

Many of the patients with peptic ulcer had known risk factors: they were aged = 60 years (73% for celecoxib vs 97% for rofecoxib), they had a history of peptic ulcer (18% vs 0%) or they were taking other medication which increased their risk (45% vs 81%). However, 16 of those who developed peptic ulcer with celecoxib (none with rofecoxib) were aged < 60 years and had no stated risk factors. In five of these cases the ulcer was diagnosed within 4 weeks of initiation of celecoxib. In nine of the 16 cases, the diagnosis was confirmed by endoscopy, radiology, or during surgery.

Of the reports of GI haemorrhagic events (in the absence of a diagnosis of peptic ulcer), 16 cases with celecoxib and five with rofecoxib involved patients aged < 60 years, with no stated history of GI ulcer and no concurrent use of another NSAID. Those reports mentioning alcohol as a possible factor were excluded. For these 21 cases, time to onset ranged from 1 day to 8 months (median 13 days). In four cases the reaction occurred after a single dose, and in one of these cases the patient's haematemesis recurred following a single dose a week later.

Initial results from clinical trials indicated a rate of upper GI ulceration with celecoxib or rofecoxib of around 2 per 100 patient-years during 6-9 months' treatment, significantly lower than with the nonselective NSAIDs.^1,2 However, while a pivotal study suggested that there may be a long-term advantage of rofecoxib over the nonselective NSAIDs for upper GI ulceration,¹ results after 12 months' usage of celecoxib indicated similar rates of ulcer complications to diclofenac and ibuprofen.³ The differences between celecoxib and rofecoxib apparent in the ADRAC data may reflect the differences seen in the clinical trials and/or they may relate to differences between the populations of users. Whatever the absolute rates of peptic ulcer may be with celecoxib and rofecoxib, the serious events reported to ADRAC suggest that selective COX-2 inhibitors should be treated with similar caution to other NSAIDs.

ADRAC has received 26 reports of neonates with symptoms attributed to withdrawal effects due to maternal third trimester ingestion of SSRIs (paroxetine 10, sertraline 7, fluoxetine 7, citalopram 2). The table presents the most frequently reported reactions. Other reactions included convulsions, tremor, fever and respiratory disorders (respiratory depression, apnoea, tachypnoea). Two babies had marked extensor posturing with back-arching. The usual day of onset, if reported, was the day of birth, but ranged from 0 to 4 days of age. The symptoms resolved in 2-3 days in most cases.

* In one case the symptoms may have been from breast-milk transfer.

In addition, 13 reports have been received of neonatal adverse effects probably resulting from breast-milk transfer of an SSRI (sertraline 9, paroxetine 2, fluoxetine 2). There was some overlap of the symptoms resulting from drug transfer into breast-milk and from drug withdrawal (see table). However, sleepiness was reported only with breast-milk transfer, and in two cases the baby slept for prolonged periods.

One study found that 12 (22%) of 55 neonates exposed to maternal paroxetine in the third trimester required prolonged hospitalisation for neonatal complications.¹ The most common problem was respiratory distress (9), but two neonates had hypoglycaemia and one each had bradycardia, tachycardia, jaundice and feeding problems. None had underlying pathology and all recovered following a brief period of intensive intervention. In the same study, exposure to paroxetine through breastfeeding caused symptoms in 8 (22%) of 36 infants, with alertness (6), sleepiness (1) and irritability (1).

In adult users, withdrawal effects following paroxetine appear to be more likely than following use of other SSRIs, and hence neonatal withdrawal may be more likely with paroxetine, but this is yet to be demonstrated in comparative studies.² However, paroxetine may have an advantage in breastfeeding since breast-milk transfer is proportionately lower than with fluoxetine or citalopram.³ One study in 11 infants detected sertraline in breast-milk but there were no adverse effects associated with exposure.⁴

It is probable that neonatal withdrawal effects would be minimised by using the lowest effective maternal dose, while breast-milk transfer can be treated by stopping or reducing the dose of SSRI, or by using formula milk.

More specifically, the combined use of ACE inhibitors, diuretics and NSAIDs, termed the "triple whammy", is implicated in a significant number of reports to ADRAC of drug-induced renal failure.² This effect is also seen with COX-2 inhibitors and angiotensin receptor antagonists ("sartans").³ In 2002, 28 of the 129 reports to ADRAC of acute renal failure implicated one of these combinations. Most reports to ADRAC of drug-induced renal failure relate to elderly patients, and this applies as well to renal failure associated with the triple therapy (median age 76 years). The fatality rate for ADRAC cases of renal failure with the "triple whammy" is 10%.

The use of ACE inhibitors and angiotensin receptor antagonists is increasing, as is the use of these agents in combination products with a diuretic. Episodes of renal failure appear to be precipitated by mild stress (e.g. diarrhoea, dehydration) in a patient taking the triple combination or by the addition of a third drug (usually an NSAID) to the stable use of the other two. ADRAC suspects that the risk of acute renal failure is underestimated and the syndrome underrecognised.

ADRAC wishes to remind prescribers that the combination of ACE inhibitors (or angiotensin receptor antagonists), diuretics and NSAIDs (including COX-2 inhibitors) should be avoided if possible, and great care should be taken with ACE inhibitors and NSAIDs in patients with renal impairment.

Patent Blue V (occasionally referred to as Sulphan Blue) is a dye used to colour lymph vessels. In lymphangiography, Patent Blue V is injected subcutaneously so it can enter the lymphatic vessels, which are then visible through the skin and can be injected with a radiological contrast agent. Patent Blue V is being increasingly used for a number of purposes including breast surgery, where it is injected into breast tissue in order to locate draining lymph nodes.
Over the years ADRAC has received 42 reports of reactions to Patent Blue V, including six reports of anaphylaxis. Five of these have been reported since October 2000 and were in women aged 37-54 years, undergoing breast surgery. In four cases, the anaphylaxis was described as severe; two patients required admission to an Intensive Care Unit. Past exposure to Patent Blue V was generally not recorded, but is probably unlikely. One patient had a history of cold urticaria. Most recovered without sequelae.
Surgeons and anaesthetists should be aware of the potential for severe allergic reactions to Patent Blue V. The Product Information recommends testing for hypersensitivity by injecting a small volume of solution initially, then waiting a short time to see if an allergic reaction develops.

Prescribers are reminded that the drug is contraindicated in men for whom sexual intercourse is inadvisable such as those with severe cardiovascular disease, established cardiac failure and unstable angina pectoris. The possibility of undiagnosed cardiovascular disorders in men with erectile dysfunction should be considered before prescribing. Sildenafil has been shown to potentiate the hypotensive effects of both acute and chronic nitrate administration and, therefore, its coadministration with nitrates in any form, either regularly or intermittently, is contraindicated.

ADRAC has received 354 reports in association with tramadol. The most common reactions include nausea, vomiting, sweating, dizziness, rash, tremor and headache. The more serious adverse reactions reported are presented in Table 1.

For the cases of convulsions, the median time to onset was 2 (range 1-19) days. Tramadol was the only suspected drug in 11 cases, but in 14 other cases the patient was taking additional drugs which may lower the seizure threshold, including propofol, bupropion, hydrocortisone, morphine, and