Rom J Ophthalmol. 2026 Jul-Sep;70(3):381-392. doi: 10.22336/rjo.2026.53.
ABSTRACT
OBJECTIVE: Benzalkonium chloride (BAK) is the most commonly used preservative in multidose ophthalmic solutions, and experimental studies indicate that it damages corneal nerves in a concentration- and duration-dependent manner. No randomized trial has examined whether BAK-preserved artificial tears alter the corneal subbasal nerve plexus in dry eye disease. This analysis asks whether six months of exposure to BAK at the concentration used in artificial tears is associated with detectable alterations in corneal innervation.
METHODS: In this substudy of a six-month, randomized, single-masked trial, imaging was a protocol-defined secondary objective. We compared 0.15% sodium hyaluronate preserved with 0.005% BAK (BAK HA) with the preservative-free formulation of the same product (PF HA), both instilled pro re nata. Thirty-two of the 43 trial completers underwent in vivo confocal microscopy of the central cornea at baseline and at six months, with 19 receiving PF HA and 13 receiving BAK HA. We analyzed seven corneal nerve parameters using mixed-effects models that accounted for both eyes.
RESULTS: Baseline characteristics and all baseline nerve parameters were comparable. Over six months, no parameter showed a statistically significant between-group difference in change. For corneal nerve fiber density, the change in the BAK HA group exceeded that in the PF HA group by 3.52 fibers/mm² (95% confidence interval, -2.32 to 9.37; P = 0.238).
DISCUSSION: Our design removes the active drug that confounds the glaucoma literature, but the estimates are imprecise, and the design cannot establish equivalence.
CONCLUSIONS: This substudy detected no between-group difference in corneal subbasal nerve parameters.
PMID:43001155 | PMC:PMC13744673 | DOI:10.22336/rjo.2026.53
