Send to

Choose Destination
Ann Pharmacother. 2010 Dec;44(12):1914-21. doi: 10.1345/aph.1P268. Epub 2010 Nov 30.

The role of benzalkonium chloride in the occurrence of punctate keratitis: a meta-analysis of randomized, controlled clinical trials.

Author information

Ocular Surface Center and Refractive Surgery, Department of Ophthalmology and Visual Sciences, Case Western Reserve University and University Hospitals Eye Institute, Cleveland, OH, USA.



Benzalkonium chloride (BAK), the most commonly used preservative in topical ophthalmic solutions, has undergone considerable criticism in recent years, principally based on in vitro and in vivo studies. Relevance to the clinical setting has not been confirmed.


To determine whether administration of twice the amount of BAK was associated with an increased incidence of punctate keratitis in long-term, double-masked trials comparing latanoprost ophthalmic solution and vehicle with timolol ophthalmic solution in patients with glaucoma or ocular hypertension.


A meta-analysis of the double-masked phases of 7 prospective, controlled clinical trials compared the incidence of punctate keratitis among patients assigned to treatment with latanoprost or timolol. In all studies, the amount of BAK administered daily in the latanoprost arms was approximately twice the amount used in the timolol arms. All reports of punctate keratitis either as a finding or an adverse event were included. A fixed-effect model was used because the heterogeneity was small and not statistically significant. Sensitivity analyses were conducted. Funnel plots were provided to address potential publication bias.


Of the 1694 patients enrolled in the double-masked portion of the trials (latanoprost, n = 892; timolol, n = 802), the overall incidence of punctate keratitis was 6.3% (106/1694). The incidence in latanoprost-treated patients was 6.5% and in timolol-treated patients was 6.0%. The risk difference for punctate keratitis of latanoprost versus timolol was 0.005 (95% CI -0.011 to 0.020; p = 0.574), and the risk ratio of latanoprost versus timolol was 1.084 (95% CI 0.739 to 1.589; p = 0.680).


These results indicate that BAK does not produce significant corneal toxicity in the vast majority of patients with glaucoma or ocular hypertension at the concentrations used in these studies.

[Indexed for MEDLINE]

Supplemental Content

Full text links

Icon for Atypon
Loading ...
Support Center