Renal disease and use of topical non-steroidal: anti-inflammatory drugs

BMJ 1994; 308 doi: (Published 08 January 1994) Cite this as: BMJ 1994;308:110
  1. C A O'Callaghan,
  2. P A Andrews,
  3. C G Ogg
  1. Renal Unit, Guy's Hospital, London SE1 9RT
  1. Correspondence to: Dr O'Callaghan
  • Accepted 22 September 1993

Renal impairment is a recognised complication of oral, rectal, and intramuscular non-steroidal anti-inflammatory drugs. We report cases of renal disease associated with a topical preparation.

Case report

Case 1

A 74 year old woman presented with a six week history of breathlessness and oedema. Investigations showed proteinuria (+ + + by dipstick), a serum albumin concentration of 17 g/l, and a creatinine concentration of 169 mumol/l. Her urine contained no leucocytes but grew Escherichia coli in culture and she was prescribed trimethoprim, frusemide, and prophylactic heparin. Her renal function deteriorated, and three days later she was transferred to our care. She was not volume depleted but remained grossly nephrotic with proteinuria of 18 g/day. There was no eosinophilia or eosinophiluria, and results of renal phlebography and ultrasonography were normal. Renal biopsy showed a florid interstitial nephritis with normal glomeruli.


Serum creatinine concentration of patient (case 1) rose until start of dialysis. During this time topical treatment with piroxicam stopped, and renal function rapidly improved

The combination of nephrotic syndrome and interstitial nephritis was highly suggestive of use of non-steroidal anti-inflammatory drugs, but this was denied by the patient, her relatives, and her general practitioner. All drugs were therefore changed or stopped, and she was given high doses of steroids. Her renal function worsened, however, and she started haemodialysis. Shortly afterwards she became confused, possibly because of her uraemia, steroid treatment, or dialysis. A tube of piroxicam gel was then discovered in her locker, which she had been applying regularly to her shoulder and back for musculoskeletal pains. Over six weeks she had used three 60 g tubes of 0.5% piroxicam and had been applying it in the ward bathroom at least twice daily until her confusion. After its removal her renal function rapidly recovered so that 10 days later she stopped dialysis and three weeks later her oedema was reduced with proteinuria only +, serum albumin concentration 32 g/l, and creatinine concentration 110 mu mol/l. The figure shows the changes in creatinine concentration over time.

Case 2

A 57 year old woman had been using a topical cream of 3% benzydamine hydrochloride for four months and had used a total of 400 g of cream. She was referred for investigation of plasma concentrations of creatinine and urea of 137 mumol and 13.2 mmol/l respectively. When the drug was stopped these concentrations fell to 96 mumol/l and 6.5 mmol respectively, results consistent with the drug causing a substantial reduction in glomerular filtration rates. No other cause was found.


About 5-18% of outpatients taking non-steroidal anti-inflammatory drugs have renal impairment.1 Case-control studies suggest that use of these drugs doubles the risk of renal disease; in men aged over 65 the risk increases tenfold.2 Risk factors include age, volume depletion, pre-existing renal impairment, and other high renin states such as heart failure and cirrhosis. Inhibition of tonic prostaglandin dependent renal vasodilatation reduces the glomerular filtration rate and may lead to acute renal failure. Other effects include sodium and water retention, hyperkalaemia, hypertension, and medullary ischaemia, sometimes sufficient to cause papillary necrosis.

Nephrotic syndrome can occur with or without interstitial nephritis, but the combination is characteristic of the use of non-steroidal anti-inflammatory drugs.3 Allergic features such as rash, eosinophilia, or eosinophiluria are usually absent. Prognosis is good if the drug is stopped, and in a review of 56 cases steroids gave no clear benefit.3 The development of nephrotic syndrome does not seem dose dependent, and most non-steroidal anti-inflammatory drugs, including piroxicam, have been implicated.4 In healthy volunteers daily topical treatment with 2 g of 0.5% piroxicam gel for 14 days gave a blood concentration that was 5% of that achieved by the standard oral dose.5 Elderly patients might achieve higher blood concentrations because of reduced drug clearance, thin skin, and frequent application of the drug over large areas.

The first case demonstrates that topical non-steroidal anti-inflammatory drugs may cause considerable systemic side effects. The renal lesion seen is highly specific for non-steroidal anti-inflammatory drugs and strongly suggests that the piroxicam was responsible. A complete drug history must include questions on the use of topical preparations.


View Abstract