Topical administration of chloramphenicol can induce acute hepatitis

We report a case of drug induced hepatitis attributable to conjunctival administration of chloramphenicol. Although hepatobiliary complications resulting from systemic chloramphenicol have been described,1 such reactions after topical application have not been reported in adults.

A 37 year old male engineer was treated for conjunctivitis with 0.5% chloramphenicol eye drops (one drop in each eye every two hours for 24 hours followed by one drop every six hours for four days) as suggested by the BNF.2 He complained of lethargy, pruritus, dark urine, and icteric sclera seven days after completing the course of eye drops. His only other concurrent medication was a regular proton pump inhibitor (rabeprazole 20 mg once daily), which was started three years before he developed symptoms and was not stopped during the course of his illness. He had no history of using over the counter antihistamines or analgesics, taking chloramphenicol previously, or chemical exposure at work. His alcohol intake was 10 units spread evenly over the week.

At initial outpatient review, liver function tests were widely deranged with predominant transaminitis: aspartate aminotransferase concentration was 868 (normal range 13-42) IU/l, bilirubin 32 (0-22) μmol/l, and alkaline phosphatase 224 (30-130) IU/l. An urgent ultrasound scan was reported as normal, and blood tests for a liver screen were negative for common viruses (hepatitis A, B, C) and for less likely viruses (cytomegalovirus, Epstein-Barr virus, toxoplasmosis, and enteric cytopathic human orphan (ECHO) viruses); autoantibodies (anti-smooth muscle antibodies), antimitochondrial autoantibodies, antineutrophil cytoplasmic antibodies, anti-dsDNA antibodies, liver-kidney microsomal autoantibodies); and heavy metals (copper and ferritin). A liver biopsy performed six weeks after stopping the drug when the aspartate aminotransferase was 152 IU/l showed interface hepatitis (inflammation around the portal limiting plate or areas) and an inflammatory infiltrate with a high eosinophil load and thus highly suggestive of drug induced hepatitis. His liver function test returned to normal within 10 months of stopping the treatment. A repeat liver biopsy or challenge with chloramphenicol could not be justified on ethical grounds, and lymphocyte transformation testing could not be done owing to unavailability (regional immunology laboratory, personal communication).3 However, causality scores (7 on the Naranjo scale (scores 6-8 = “probable”)4; and 6 on the RUCAM (Roussel Uclaf Causality Assessment Method) scale (5-8 = “probable”)5) suggested that topical administration of chloramphenicol was the probable cause of this hepatitis.

The Committee on Safety of Medicines has been notified of two possible cases of hepatitis associated with chloramphenicol, one of which was with topical eye administration (but this was in an infant).