Resistance/Intolerance to Treatment:

Development of ganciclovir-resistant CMV infection has been reported to occur in approximately 10% of AIDS patients receiving ganciclovir for > 3 months. Most CMV strains resistant to ganciclovir are susceptible to foscarnet, which may be useful for patients who cannot tolerate or are clinically resistant to ganciclovir. Jacobson et al. enrolled 156 patients with CMV retinitis in ACTG 093, a study of foscarnet as salvage therapy. 72 patients were categorized as ganciclovir-intolerant and 84 as ganciclovir-resistant. The median interval between initiating ganciclovir therapy and switching to foscarnet was 180 days. The first 87 patients were randomized to foscarnet either 60 or 90 mg/kg/day and the remaining 69 patients received 120 mg/kg/day. Median time to retinitis progression for all patients was 8, 9 and 9 weeks for the 60, 90 and 120 mg/kg/day groups, respectively. Overall, moderate or severe adverse effects to foscarnet (renal toxicity, seizure, hypocalcemia, nausea, ataxia and altered mental status) were observed in 95% of patients with a trend toward dose-limiting renal toxicity in the 90 and 120/mg/kg/day groups. There was no survival difference among the three treatment groups. Overall median survival was 25 weeks.

The manufacturer of HPMPC (cidofovir), a cytosine-derived nucleotide analogue with a long intracellular half-life, has submitted a New Drug Application (NDA) to the Food and Drug Administration seeking approval of intravenous cidofovir for patients intolerant or failing ganciclovir or foscarnet. HPMPC must be administered with probenecid to protect against renal toxicities.

Lalezari et al. recently presented data from a study comparing two maintenance doses of HPMPC (3 mg/kg vs. 5 mg/kg every other week) in 60 patients who failed induction therapy with either ganciclovir or foscarnet. AR patients received weekly HPMPC (5 mg/kg) for two weeks as induction therapy. No statistically significant difference was seen between the two maintenance doses; a median of 115 days to progression was reported in the 5 mg/kg group and a median of 49 days to progression in the 3 mg/kg group. Adverse reactions due to concomitant of probenecid were reported in 29/60 (48%) patients enrolled.

Lalezari et al. have reported results from a phase II/III immediate versus deferred trial of intravenous HPMPC. In this study, 40 patients were randomized to receive either immediate treatment with HPMPC (5 mg/kg every week for two weeks) or placebo until disease progressed. Time to progression in the immediate treatment group was 120 days, whereas time to progression in the deferred treatment arm was 22 days. Eleven patients (27%) reported renal toxicities. Toxicities attributed to concomitant probenecid use were reported in 38% of patients enrolled.

Phase I/II dose-escalating studies of intravenous HPMPC have been reported in subjects with CMV viruria. Polis et al. administered HPMPC 0.5, 1.5 or 5 mg/kg weekly or twice weekly, with or without probenecid, to 20 patients. CMV urine cultures became negative in 4/6 evaluable patients receiving the highest doses of HPMPC. The maximum tolerated dose was determined to be about 5 mg/kg weekly, with proteinuria, glycosuria, and creatinine elevations being the dose limiting toxicities. Concomitant probenecid administration appeared to protect against renal toxicity.

Two recently reported small case studies (Kirsch et al.) of intravitreal HPMPC (cidofovir) suggest that prolonged control of CMV retinitis may be achieved with a single injection. In the first study, patients who had failed or were non- compliant with standard therapy received one 10, 20, 40 or 100 µg injection with concomitant oral probenecid. Patients were assigned to doses in a non- randomized fashion. The median progression time reported in those patients who received a single 20 µg injection was 64 days; data from other diseases were invaluable. No cases of endophthalmitis were reported. In the second case study, 17 patients (24 eyes) received a single dose of intravitreal HPMPC (20 µg) with oral probenecid. The median time to progression after the initial injection was 55 days. Disease progression was reported in 8/24 eyes treated. These eyes were retreated; median time to progression in retreated eyes was 64 days.

ISIS 2922, an antisense compound for the intravitreal treatment of CMV retinitis in patients who have failed standard therapy, is currently under investigation. In May 1995, the manufacturer suspended patient recruitment because of safety concerns, including the development of decreased peripheral vision with retinal pigment epithelial cell stippling in 20% of patients receiving intravitreal injections at the highest dose. Two phase III trials are currently enrolling. For patients With CMV intolerant or failing standard therapy who do not qualify for the two trials, a compassionate use program is available.

A phase I/II dose-ranging study of ISIS 2922 has been reported. The manufacturer reports that 10 patients with CMV retinitis were treated intravitreally with ISIS 2922 once a week for 4 weeks and every two weeks thereafter. Two patients received 75 µg (low-dose), 3 received 150 µg (middle- dose) and 5 received 300 µg (high-dose). Responses were observed in 0/2, 2/3, and 5/5 in the low, middle, and high dose groups, respectively. Side effects observed were vitreal inflammation, temporary loss of color vision and ocular discomfort.