Non Steroidal Anti-Inflammatory Drugs

Synonyms: NSAIDs, Cyclo-oxygenase inhibitors

NSAIDs are widely used drugs (both prescribed and over-the-counter), valued for their anti-inflammatory, analgesic and antipyretic properties. However they also cause serious side-effects, most notably gastric perforation and haemorrhage. In 1999, before the introduction of coxibs, more than 18.5 million courses of NSAIDs were prescribed in England and Wales for musculo-skeletal conditions which were responsible for 10 000 hospital admissions and 2000 deaths.¹ Over the last few years, concern has mounted regarding the vascular safety, initially of coxibs and then of all NSAIDs, leading to alterations in prescribing advice and behaviour.²

• Aspirin is the longest established of the anti-inflammatory drugs but is often considered separately to other NSAIDs since its predominant use is in low dose for CV prevention. In low dose it is a potent and irreversible inhibitor of COX-1 and has some COX-2 activity which increases with high dose. Anti-inflammatory activity is only present at very high dosage (over 3 g daily).
• Standard non-selective NSAIDs inhibit both COX-1 and COX-2 but do vary in their relative pathway selectivity.³
• They are usually differentiated by their propensity to cause side-effects with a spectrum ranging from ibuprofen (least likely to cause GI complications) to azapropazone (associated with a very high risk of GI bleeding).

  Risk of serious upper GI side-effects with non-selective NSAIDs3: High risk - azapropazone [discontinued] Intermediate to high risk - piroxicam Intermediate risk - diclofenac, indometacin, naproxen, ketoprofen Low risk - ibuprofen

• COX-2 selective NSAIDs fall into 2 categories: older NSAIDs which display COX-2 preference (etodolac, meloxicam); and the newer coxibs (celecoxib, etoricoxib, rofecoxib (withdrawn), valdecoxib (suspended)). ³

• Pain relief following acute injury or surgery
• Fever and pain in children - although current NICE guidelines stress that we should not use antipyretics solely to reduce fever in children who are otherwise well or to prevent febrile convulsions (since they are ineffective)⁷
• Post-childhood immunisation induced pyrexia
• Acute inflammatory disorders eg gout
• Chronic inflammatory disorders (rheumatoid arthritis, severe osteoarthritis, chronic back pain).
• Pericarditis
• Palliative pain relief for bony metastases, soft-tissue tumour infiltration and pleural pain
• Specialist use in neonatology to aid closure of patent ductus arteriosus

• Allergy - never prescribe with a history of previous hypersensitivity reaction (including asthma, angioedema, urticaria, and rhinitis) to aspirin or other NSAID.
• Active or GI bleeding.
• History of peptic ulcers or upper GI bleeding.
• Pregnancy as during the last trimester, chronic or high dose NSAID use can cause premature closure of the ductus arteriosus, pulmonary hypertension of the newborn and delayed onset and duration of labour. There has been a widely reported cohort study suggesting an association between the risk of miscarriage and the use of NSAIDs but further confirmatory studies are needed. ⁸
• Coagulation defects.

• Asthma
• Renal insufficiency
• CV disease or risk factors
• Hepatic and cardiac failure
• Hypertension
• The elderly
• The acutely unwell or debilitated ³
• Concurrent use of drugs that increase risk of bleeding
• Women trying to conceive due to reversible impairment of fertility
• Breast feeding - although some NSAIDs are felt to be safe as secreted in very small quantities in breast milk (ibuprofen, diclofenac)

• Differences between NSAIDs' anti-inflammatory activities are small - systematic reviews have found no important differences in their efficacy for the symptoms of musculo-skeletal disorders.⁹
• However differences in NSAIDs' tolerability are large. Non-selective NSAIDs vary in their risk of GI toxicity and possibly CV risk.
• COX-2 selective inhibitors may have a lower risk of GI bleeding compared with older NSAIDs¹⁰ but have other disadvantages (cost, CV risk).
• Assess individual's risk for GI bleeding and CV disease prior to prescribing any NSAID.

  Choice of non-selective NSAID3
  High risk of GI side-effects?	Also needs aspirin?	Cardiovascular disease present (or high risk)?	Preferred options
  No	No	No	Ibuprofen, diclofenac or naproxen
  Yes	No	No	Ibuprofen, diclofenac or naproxen PLUS gastric protection
  Yes	Yes	Yes	Naproxen PLUS gastric protection
  No	No	Yes	Ibuprofen or naproxen. Consider gastric protection on an individual basis.
  Yes	No	Yes	Ibuprofen, naproxen or diclofenac PLUS gastric protection

• Start at the lowest recommended dose as side-effects are proportional to dose. Benefits of NSAIDs increase towards a maximum value at high dose but side-effects increase linearly with dose without an apparent ceiling. ⁹
• Pain relief should commence soon after the first dose with full analgesic benefit within a week but anti-inflammatory effect can take up to 3 weeks.
• Treating chronic musculo-skeletal pain in elderly patients or patients with significant comorbidities (in particular CVD or risk factors for IHD) provides the greatest challenge. The American Heart Association has offered a stepwise approach¹¹, starting with agents of lowest reported CV risk and progressing to other agents with consideration of risk/benefit ratio at each step:
  • Step 1: paracetamol, aspirin, tramadol, narcotic analgesics (short term)
  • Step 2: Nonacetylated salicylates
  • Step 3: Non COX-2 selective NSAIDs
  • Step 4: NSAIDs with some COX-2 activity
  • Step 5: COX-2 selective NSAIDs

Enteric Coating

Enteric-coating may reduce the incidence of dyspeptic symptoms but does not reduce the risk of bleeding.

Modified Release forms

Certain NSAIDs (eg ibuprofen, diclofenac, flurbiprofen, indometacin) are available in slow-release preparations. This can provide relief from night pain where day-time NSAID use has been successful in controlling pain. However it should not be assumed to lessen risk of GI toxicity - patients receiving slow release NSAIDs (regardless of prior risk factors) were more likely to require gastroprotection and GI investigations than those receiving standard forms and were more likely to switch away from SR formulations than to them¹² suggesting poorer tolerability.

Slow release preparations are classed as premium-priced products - only to be used where there is a potential for improvement in compliance, effectiveness or side effects.

Topical preparations

Recent systematic reviews concluded that there is limited evidence that topical NSAIDs reduce acute musculoskeletal⁹ and osteoarthritic pain¹³ compared to placebo in the short term ( < 2 weeks) but longer term evidence of efficacy was lacking. Subsequently a randomised 12 week equivalence study showed some evidence of topical diclofenac having similar efficacy but less side-effects than oral diclofenac at doses of 150mg.¹⁴ Use of topical NSAIDs does avoid many of the interactions associated with systemic NSAIDs although side-effects (hypersensitivity, asthma, renal disease) have been reported when large doses are used topically.

• Routine monitoring is not normally required for NSAIDs. However, it is advisable where there is greater risk of side-effects or where side-effects could be more serious in view of co-morbidity.
• All patients on NSAIDs should have regular medication reviews (at least 6 monthly) considering:
  • The need for continuing treatment
  • Drug efficacy
  • Side-effects
  • The need for gastroprotection
  • Possible alternative drugs
• Renal Impairment - check renal function prior to treatment if anticipating longer term NSAID treatment:
  • With mild renal impairment (GFR 20-50 mL/minute, serum creatinine 150-300 micromol/litre), use NSAID in its lowest effective dose and monitor regularly for deteriorating renal function. If renal function deteriorates, the drug should be stopped.
  • With moderate to severe renal impairment (GFR > 50 ml/minute, serum creatinine > 300 micromol/litre), NSAIDs should be avoided where possible. Advice from renal physicians would be appropriate where options are limited. Beware the development of acute-on-chronic renal failure in those with concurrent acute illness.
• Hypertension/cardiac failure - consider stopping NSAID if treatment causes clinical deterioration.
• In the elderly, check for GI symptoms particularly at the outset of treatment, although the risk of complications exists throughout the course of treatment. Consider the impact of NSAID treatment on co-morbidities. Have a low index of suspicion for checking FBC and iron studies, differentiating between iron-deficiency anaemia (with NSAID related blood loss) and anaemia of chronic disease.

Gastrointestinal adverse effects

NSAIDs' most well-known and feared association is with serious GI adverse effects. The risk of serious ulcer disease requiring hospitalisation is 1 per 100 patient years of NSAID use in unselected patients.¹⁵ In those not considered to be at high risk of GI bleeding, use of NSAIDs for musculoskeletal pain increases the risk of GI bleeding five-fold.¹⁵

NSAID induced asthma

Aspirin-induced asthma is reasonably common and potentially life-threatening. A recent systematic review¹⁶ estimated its prevalence (from oral provocation testing) as 21% in adults with asthma and 5% in children with asthma. The incidence of cross-sensitivity to NSAIDs was high (98% for ibuprofen, 100% for naproxen, 93% for diclofenac) but low for paracetamol (7%). Based on this:

• Approximately 80% of asthmatics can take aspirin safely but always ask asthmatic patients about previous experiences with aspirin and other NSAIDs and warn them to stop taking them if their asthma deteriorates.
• Paracetamol should be the analgesic and antipyretic of first choice for asthmatic patients and those with high risk features for developing aspirin-induced asthma (severe asthma symptoms, nasal polyps, urticaria, and rhinitis) should be advised to avoid all products containing NSAIDs.

It is worth noting that NSAIDs are not contraindicated in those with COPD.

Renal toxicity

NSAIDs are nephrotoxic - COX enzymes are expressed in the kidney and are thought to be important in protecting the kidney from hypovolaemia.¹⁷ Consequently, those on NSAIDs are more at risk of developing acute renal failure when haemodynamically challenged and the most vulnerable are those who have existing renal/hepatic or cardiac impairment and the elderly.
Interstitial nephritis is a rare side-effect of some NSAIDs and is not dose-related.

Fluid Retention

This is associated with NSAID use and can be sufficient to precipitate congestive cardiac failure in the elderly.

Increased cardiovascular risk

Cox-2 selective inhibitors were shown to be associated with an increased thrombotic risk and concern arose that this could also be the case for non-selective NSAIDs.² A recent, large meta-analysis (138 RCTs, 145,000 patients)¹⁸ found that high doses of ibuprofen (800 mg tds) and diclofenac (75 mg bd) moderately increased risk of vascular events (mostly MIs), similar to the coxib effect. Risks increased with longer term use. The European Medicines Agency's safety review in October 2006¹⁹ reported:

• Overall risk-benefit ratios for non-selective NSAIDs remain favourable.
• Some non-selective NSAIDs may also increase vascular risk, particularly diclofenac.
• There is no evidence of increased risk from short-term, low dose ibuprofen (1200 mg daily).
• Naproxen had a lower thrombotic risk compared to coxibs and did not appear to raise the risk of MI.
• High dose ibuprofen and diclofenac should not be used for patients with established vascular disease or in those with high CV risk.

On the basis of this, current CHM advice is:

• Use the lowest effective dose of NSAIDs for the shortest possible time.
• Review the need for long therapy regularly.
• Base prescribing decisions on an individual's GI and CV risks.²⁰

Patients on low-dose aspirin should be discouraged from combining it with ibuprofen on a regular basis as this combination increase the risk of GI adverse events and concern that it may lessen aspirin's antithrombotic action, although there has been no increase in risk of death among ibuprofen and aspirin users compared with aspirin users only.²¹