Etoposide as a virocidal anticytomegalovirus therapy: intravitreal toxicology and pharmacology in rabbits.

Abstract

Although Cytomegalovirus (CMV) retinitis is now a common intraocular infection, current therapy is only virostatic so ongoing treatment is required. Etoposide was found to be virocidal for CMV in laboratory experiments and it might prove to be beneficial clinically. We investigated the toxicity and intraocular concentration of etoposide (VP16) and its new analogue etoposide-phosphate (VP16P) following intravitreal injections in rabbit eyes. First a sequential dose-response was assessed with flash electroretinogram for both eyes of light- and dark-adapted rabbits (n = 7; one rabbit for each dose) over a range of light intensities before and after intravitreal injection of VP16 or VP16P to one eye; the other eye was injected with normal saline as a control. A multidose study was then performed on four rabbits. A non-toxic dose of VP16P (50 or 75 g) was injected into the vitreous of one eye on four occasions 1 week apart. A photopic electroretinogram was performed before the first injection and 6 weeks after the last injection. All the eyes from the electroretinogram studies were fixed in formalin, placed in paraffin, then stained with haematoxylin and eosin and examined under a light microscope. To determine the time-course of the intraocular concentrations of VP16P a sequential pharmacokinetic study was performed using a further 12 rabbits. Each rabbit was injected with 50 g VP16P to one eye and 75 g VP16P to the other eye. Three of these rabbits were killed at 1, 3, 6 and 9 h after injection. Samples of vitreous were assayed for both VP16 and VP16P using HPLC. An in vitro dose response assay was performed using third-passage bovine retinal pigment epithelial (RPE) cells cultured in Dulbecco's modified Eagles medium with fetal calf serum. The effect of a log-dose increment of VP16P on the RPE cell proliferation was assessed using tritiated thymidine incorporation. The electroretinogram studies suggested that VP16 was toxic even with the 10 g dose. For VP16P a toxic effect was noted following injection of a single dose greater than 100 g. Multiple injections of 50 or 75 g VP16P did not produce a toxic response. Histological examination demonstrated significant abnormality only with the 500 g dose of VP16 or VP16P. VP16P was rapidly metabolized to VP16 in the eye, producing concentrations of 2.0 g/mL or more for up to 9 h following a 75-microg dose. This suggests that the electroretinogram findings following VP16 injections were confounded by a toxic effect of the ethanol solvent (which is absent from the VP16P preparation). VP16P was quite potent, the ID50 was about 0.1 g/mL for bovine RPE cells in the in vitro assay. These results indicate that multiple 75-gVP16P intravitreal injections were not toxic to the rabbit eye and provide a therapeutic intraocular concentration for up to 9 h after the injection.