The safety of leflunomide

Leflunomide, alone or in combination with other antirheumatic drugs, is an effective but potent immunosuppressive drug for patients with moderate to severe rheumatoid or psoriatic arthritis.

Common adverse effects include diarrhoea, nausea, vomiting, mouth ulcers, skin rash, alopecia, minor infections, mild increase in blood pressure and elevated liver enzymes.

Major adverse effects such as significant lung injury, severe infection and cytopenia may occur, and early recognition of these is crucial.

The risk of adverse effects is increased with concurrent medications, particularly methotrexate, and patient factors such as alcohol consumption and low body weight.

Regular monitoring with clinical assessment, blood counts and liver function tests is essential.

Leflunomide is contraindicated in pregnancy and effective contraception is required for men and women during and after treatment.

Live vaccines should be avoided for at least six months after stopping treatment.

Leflunomide has a complicated pharmacological profile, including a long half-life that makes the management of toxicity difficult.

Leflunomide is an immunosuppressive drug with an anti-inflammatory action. It inhibits the biosynthesis of pyrimidine in rapidly dividing cells and is used as a disease-modifying antirheumatic drug (DMARD) in patients with rheumatoid and psoriatic arthritis. Leflunomide was listed on the Pharmaceutical Benefits Scheme (PBS) for rheumatoid arthritis in 2000 and for psoriatic arthritis in 2007. The rate of initial prescriptions was rapid due to the large number of patients with intolerance or lack of response to other DMARDs, including methotrexate. Also, there had been no other new disease-modifying drugs for rheumatoid arthritis in the preceding decade.

It is estimated that more than 15 000 patients have received PBS-subsidised leflunomide in the last decade. The pattern of use in Australia is unique, with more than 50% of patients with rheumatoid arthritis on leflunomide also taking methotrexate. This is in contrast to the USA and Europe, where rates of simultaneous prescribing are low (<10%)

While there is increasing recent evidence to support the benefits of aggressive combination DMARD therapy, the high co-prescription rate appears to have been primarily driven by the PBS eligibility requirements to access biological DMARDs, rather than by contemporary trial findings. Etanercept was first listed on the PBS for rheumatoid arthritis in August 2003. Until the modifications to eligibility criteria in August 2010, to qualify for subsidised etanercept patients had to have already had:

• weekly methotrexate (at least 20 mg),
• a combination of methotrexate and two other DMARDs, and
• leflunomide (with or without methotrexate) or cyclosporin.

For most patients leflunomide was added to pre-existing methotrexate, hence the high rate of co-prescription in Australia. There is variability in opinions amongst rheumatologists as to whether or not the doses of leflunomide and methotrexate should be reduced by 30–50% of the maximum recommended daily dose when combination treatment is started.

Leflunomide is a prodrug which is rapidly metabolised in the liver and gut wall to the active metabolite teriflunomide or A771726. Its half-life of 2–4 weeks means a loading dose was used in the clinical trials to have a clinical effect as quickly as possible – without this, it may take up to two months to take effect. However, a loading dose is not typically used in current practice because of the increased risk of adverse effects, particularly gastrointestinal intolerance.

Leflunomide's long half-life means adverse effects and drug interactions may persist for several weeks after cessation. The active metabolite undergoes extensive enterohepatic recirculation and is eliminated by biliary and renal excretion. If major toxicity or unplanned pregnancy occurs, a washout procedure is undertaken with oral cholestyramine (typically 8 g three times daily for 11 days) or activated charcoal. Teriflunomide cannot be removed by dialysis therefore haemodialysis is not a treatment approach for patients who are experiencing major toxicity or who have taken an overdose of leflunomide.

In clinical studies lasting two years, the most common adverse effects (in more than 5% of patients) included diarrhoea, nausea, vomiting, mouth ulcers, skin rash, alopecia, minor infections, mild increase in blood pressure and asymptomatic reversible liver enzyme increases. However, concerns have been raised about the risk of major adverse effects and early recognition of leflunomide's potential toxicity.²

The Adverse Drug Reactions Advisory Committee (ADRAC)* received reports of severe pulmonary disease and other serious hepatic, haematological and neurological adverse effects.³ It is unclear whether the apparently high number of spontaneously reported serious adverse effects in Australia relates to the high concurrent DMARD prescription rate, an epidemiological phenomenon, reporting bias, or a combination of these factors.

A failure to recognise the possibility of leflunomide-induced lung disease led to a review by the regulatory authorities and the recommendation for an educational update to medical practitioners.

Interstitial lung disease (including interstitial pneumonitis and pulmonary fibrosis) has been rarely reported (less than 0.38% of patients) during treatment with leflunomide.⁴ The two most common symptoms of lung injury are shortness of breath, particularly with exertion, and a dry cough. Additional symptoms may include fever, fatigue and generalised myalgia. The symptoms may occur acutely during therapy or develop insidiously. The onset of new or worsening of pre-existing respiratory symptoms such as cough or dyspnoea should prompt further investigation. Although lung injury has been associated with both leflunomide and methotrexate taken alone, the risk of interstitial pneumonitis appears to be increased when the drugs are taken concurrently.^5–7

A multivariate analysis of more than 5000 Japanese patients prescribed leflunomide identified pre-existing interstitial pneumonitis, use of a loading dose, cigarette smoking and low patient body weight as significant risk factors for the development of leflunomide-induced lung injury.⁸

When significant drug-related lung injury is suspected, the drug should be stopped immediately and washout treatment with cholestyramine is recommended.⁹ However, the evidence of its benefit is inconclusive to date. High serum C-reactive protein, low serum albumin, severe hypoxaemia and mechanical ventilation indicate poor prognosis. Peripheral blood lymphopenia often occurs in association with lung injury and a sustained low lymphocyte count portends a fatal outcome.⁴

Patients with rheumatoid arthritis have an increased risk of morbidity and mortality related to infections. This may occur due to the disease itself or as a result of drugs used to control the disease.

In the controlled trials, respiratory infection (including bronchitis and pneumonia) was observed in 15% of patients treated with leflunomide over a six-month period. Respiratory infections, especially upper respiratory tract infections, are the most commonly reported site of infection followed by urinary tract, and skin or soft tissue infections. Over two-thirds of infections were mild or moderate in severity, however, serious infections (3.3 per 100 patient-years) including pneumocystis pneumonia and tuberculosis have been reported.¹⁰

Rarely, some patients develop fulminant, and even fatal, sepsis. The risk of severe infection is greatest in patients with severe rheumatoid arthritis, or in those receiving combination DMARD therapy or continuous prior corticosteroid use (oral prednisone >5 mg/day).

As teriflunomide has a direct inhibitory effect on proliferating T-lymphocytes, there is also a risk of new or reactivated herpetic infections such as shingles and oral or genital herpes simplex infection.

Although leflunomide alone or in combination with other DMARDs is associated with elevations of liver enzymes (alanine aminotransferase and aspartate aminotransferase), most are less than two times the upper limit of normal, are transient or resolve with dose reduction. Rare cases of severe liver injury, some with fatal outcome, have been reported. Most cases occurred within six months and in patients with multiple risk factors for hepatotoxicity.

Patients should be counselled to have no or minimal alcohol (<3–4 standard drinks per week) and should avoid binge drinking. Concurrent use of other potentially hepatotoxic medications such as methotrexate should be monitored, as should over-the-counter, herbal and naturopathic medicines. Caution should be advised with current or recent hepatitis.

Up to 18% of patients with rheumatoid arthritis report paraesthesia from entrapment neuropathy, a mild distal symmetric predominantly sensory neuropathy, mononeuritis multiplex or severe sensory motor neuropathy. In addition, leflunomide-induced peripheral neuropathy has been noted with paraesthesia reported in 2.9%. Some data suggest patients who discontinue within 30 days of symptoms are more likely to improve or recover than patients who continue the drug.

Leflunomide can impair marrow function. The risk of impairment is increased in older patients and when leflunomide is combined with methotrexate. Regular monitoring will identify problems promptly. Minor depressions of white blood cells/neutrophils are common but significant abnormalities are rare.

Leflunomide-related hypertension has been noted in up to 10% of patients. It can aggravate pre-existing hypertension or induce new-onset hypertension within three months of therapy. Most experts consider hypertension to be manageable during treatment with either a maintained or reduced dose of leflunomide, and by giving antihypertensive treatment.

Dose-dependent alopecia is a common transitory adverse effect of leflunomide and occurs in 6–23% of patients. Hair loss is diffuse and often mild to moderate. It seems prudent to reduce the leflunomide dose.

Pruritus and a variety of skin conditions have been reported. These include a non-specific rash, isolated pruritus, mucosal ulcers, Stevens-Johnson syndrome, toxic epidermal necrolysis, lichenoid reaction, cutaneous vasculitis, erythema multiforme and subacute cutaneous lupus.

Diarrhoea is the most common adverse effect, occurring in 17% of trial patients. It is common when loading doses of 100 mg/day for three days are given, so Australian rheumatologists usually omit the loading dose. The majority of cases resolve with time and can be managed symptomatically. The exact mechanism of leflunomide-induced diarrhoea is not known.

All patients prescribed leflunomide must be monitored regularly during treatment. Current Australian practice follows British and American recommendations of blood pressure check, full blood count, urea and electrolyte tests, and liver function assessment before starting treatment. If blood pressure is greater than 140/90 mmHg on two consecutive readings two weeks apart, hypertension should be treated before starting leflunomide. During therapy, full blood count and liver function should be checked every month for the initial six months and if stable, 2–3 monthly thereafter. If leflunomide is co-prescribed with methotrexate or other potentially hepatotoxic medication or an immunosuppressant, monitoring should be monthly. After six months of combination treatment, patients should be monitored at least every 2–3 months thereafter. Guidelines for managing adverse events associated with leflunomide are listed in the Table.

Caution should be exercised with drugs metabolised via cytochrome P450 2C9, such as warfarin and phenytoin. Leflunomide may reduce the metabolism of warfarin, thereby increasing the INR and risk of bleeding. Because of the active metabolite's long half-life, the effects of an interaction may persist for 2–4 weeks after stopping the drug. No significant interactions between leflunomide and triphasic oral contraceptives or cimetidine have been found. Smoking increases leflunomide clearance and this may be of particular relevance as several studies have shown smoking adversely influences the severity of rheumatoid arthritis.

Influenza and pneumococcal vaccinations are recommended for patients before starting leflunomide. Hepatitis B vaccination should be given if risk factors are present and vaccination has not previously been administered. No clinical data are available on the efficacy and safety of live vaccinations during leflunomide treatment. The US Food and Drug Administration recommends not using live vaccines during treatment and for at least six months after ceasing therapy.

As animal studies with leflunomide showed an increase in teratogenicity and embryonic death, leflunomide is contraindicated in pregnancy. Conception should be excluded before commencing leflunomide treatment. Because of the long half-life, the sponsor advises ceasing leflunomide at least two years before a planned pregnancy or to use a washout procedure if inadvertent pregnancy occurs. A prospective cohort study (1999–2009) showed that 64 women with rheumatoid arthritis who were exposed to leflunomide during pregnancy had a similar incidence of major infant abnormalities (5.3% of infants) compared to 108 pregnant women with rheumatoid arthritis not treated with leflunomide (5.3% of infants) and 78 healthy pregnant women (4.2% of infants). Similarly, no particular type of birth defects was treatment related.¹¹ Breastfeeding is not recommended with leflunomide.

Men taking leflunomide should avoid getting their partner pregnant while taking leflunomide and for up to 64 days after therapy (at least one cycle of spermatogenesis).¹²

Patients should be warned about the potential for mild adverse effects with leflunomide. The possibility of serious pulmonary, hepatic, haematological and neurological adverse effects should also be discussed. The necessity for regular blood monitoring should be emphasised. Leflunomide should be ceased if there is concern about possible adverse effects.

Dr Kubler is Chair of the Editorial Executive Committee of Australian Prescriber.

* In 2010, ADRAC was replaced by the Advisory Committee on the Safety of Medicines (ACSOM)