Eye implants shown superior to IV for retinitis

Prophylaxis may not be needed for some

New research on the efficacy of ganciclovir eye implants for cytomegalovirus retinitis shows conclusively that fewer patients progress to disease with the implant compared to those who are given intravenous ganciclovir. The picture is less clear, however, on how new antiretroviral treatment will affect implant protocols and the need for secondary prophylaxis, say leading ophthalmologists.

In a randomized trial, 188 patients with AIDS and newly diagnosed CMV retinitis were treated with either IV ganciclovir or sustained-release ganciclovir implant. The median time to progression to retinitis was 221 days with the implant, compared to 75 days with IV ganciclovir. That translates into a threefold higher risk of progression for IV ganciclovir, say the authors of the study, which was published recently in the New England Journal of Medicine.1

"In terms of controlling retinitis, it is a no-brainer as to what is best for the eye, and that is what this study showed," co-author Daniel Martin, MD, associate professor medicine at Emory School of Medicine in Atlanta, tells AIDS Alert.

Implants face obstacles

Controlling retinitis, however, is not the only consideration in choosing a therapy. While the implants have become the standard of care for many ophthalmologists since they were approved nearly two years ago, they have had to overcome several disadvantages. First, the implants protect only one eye, and because CMV is a systemic disease, the other eye as well as other organs can become infected. Second, some studies have raised concerns that retinal detachment increases with implants. And third, the cost of implants, which needed regular replacement, can run up to $30,000 a year.

New research, however, has shed light on all three issues, says Martin. And in the final analysis, eye implants are the more attractive choice.

Initially, Martin and others reported what seemed to be increased incidence of retinal detachment with implant therapy. Although IV ganciclovir also increases the risk of detachment, the proximity of the implant to the retina made it appear that the procedure heightened the risk. Further study, however, has clarified the issue, Martin says.

"There is no statistically significant difference between the rate of retinal detachment in eyes treated with implants vs. those treated with IV ganciclovir," he says. "With implants, you probably have a slightly increased risk of detachment between one and two months, but the long-term risk is the same if not less."

A leading authority on CMV retinitis, Douglas Jabs, MD, professor of medicine and professor of ophthalmology at Johns Hopkins School of Medicine, agrees. "My clinical impression is the detachment rate is no greater and may be lower in the long term, although not the short term," he tells AIDS Alert.

The second issue, and a more serious one, is the risk that the second eye will likely develop CMV retinitis, and that other organs could also become infected. "What is clear is the risk for fellow-eye involvement is greater, and that is a serious consideration," Martin says.

At Emory and Johns Hopkins, ophthalmologists have tackled the problem by putting implant patients on oral ganciclovir (Cytovene) as maintenance therapy. By using this newly approved drug, patients avoid the costly and cumbersome daily infusions required for IV ganciclovir, as well as the high risk of infection from catheters. How well this combination treatment protects patients from reactivation of CMV or infection outside the eye is not known. Hoffmann-La Roche Inc. of Nutley, NJ, the manufacturer of Cytovene, has funded a study evaluating that combination, and data could be presented this fall.

"A lot of people feel that the combination of implant plus oral ganciclovir is the logical marriage. Even though this trial is not done, it is the standard of care for many of us," Martin explains.

The alternative — treating CMV retinitis patients with IV ganciclovir and implants — is untenable because it decreases a patient’s quality of life and escalates treatment costs, Martin says. Indeed, there is no good rationale for using IV ganciclovir because the blood level barrier allows so little of the drug to reach the eye. The high doses required during the induction phase of treatment are strongly toxic, while the reduced levels during maintenance therapy are inadequate to prevent reactivation, he explains.

While oral ganciclovir may solve the problem of systemic involvement with CMV, it is not an ideal solution, Martin notes. As many as 18 pills are required daily to reach adequate absorption; the drug can cause neutropenia; it is difficult to take with AZT; and it costs as much as $20,000 a year.

While the combination of oral ganciclovir and an implant would seem prohibitive costwise, it turns out to be comparable to IV ganciclovir treatment, Martin says. A cost analysis he developed with an unbiased third party showed that one implant plus oral ganciclovir costs about $30,000 a year. The most conservative estimate for a year of IV ganciclovir treatment, which requires home care costing about $200 a day, is $35,000. For third-party payers, the cost can rise above $50,000, he notes.

Jabs also estimates that the costs of the two treatments are on a par, even when accounting for a second eye implant with co-lateral infection and the replacement of the implants every six months.

Just as the standard of care for CMV retinitis has become more concrete, the impact of protease inhibitors has confounded treatment decisions. An estimated 25% to 45% of AIDS patients develop CMV disease. With the initiation of triple-drug therapy, patients have experienced CD4 count increases of as much as 150 to 200 cells. Such a dramatic improvement in the immune system has, at least temporarily, put many patients beyond the fear of CMV retinitis. As a result, the number of cases seen at Johns Hopkins dropped 55% during the past three years. Other centers report similar declines, Jabs says. The decline cannot be attributed to primary prophylaxis with oral ganciclovir because it is has not received widespread use, he adds.

Protease inhibitors may reduce risk

A study in the August issue of the American Journal of Ophthalmology reports the impact of protease inhibitors on patients with CMV retinitis and how the treatment appears to be healing retinal lesions in some patients.2 Whether the improvements are a direct result of the drugs or of increased CD4 counts is not known, but the phenomenon has raised the question of whether prophylaxis or implant replacements are needed in these patients.

"We are in a transient period, and for the next few years we are going to be feeling our way along," Martin says. "The guess is that with protease inhibitors we have shifted the curve to the right, and for people who had CD4 counts of 2 and now have 200, their risk of reactivation is much less. Maybe they don’t need therapy at all."

At the same time, however, there have been reports of patients experiencing CMV retinitis despite higher CD4 counts. Earlier this year, Mark Jacobson and his colleagues reported that prior to protease inhibitor therapy, only 4% of study patients with CMV retinitis had CD4 counts above 50. With the advent of combination therapy, that number rose to 29%, with 14% having CD4 counts above 100. The findings suggest that specific immunity to CMV may not be fully restored despite higher CD4 counts, Jabs notes.

Prophylaxis remains optional

Taking the conservative approach, the U.S. Public Health Service has chosen not to change its preventive therapy guidelines for CMV retinitis. The updated guidelines, published recently in the Morbidity and Mortality Weekly Report, state that prophylaxis against the first episodes of CMV retinitis remains optional. However, if preventive therapy is chosen, clinicians should base its use on the patient’s lowest CD4 count, the guidelines state.3

"What the PHS said is that in the absence of data one should be conservative, rather than allowing disease to reactivate left and right in large numbers of people," Jabs explains.

As more patients become experienced with protease inhibitors and more resistance develops, researchers expect that the levels of CMV will rise again, although not to the levels seen before the advent of combination therapy. In fact, the number of new cases at Johns Hopkins for 1997 is about the same as in 1996, indicating that incidence may have reached its nadir, says Jabs.

Martin also has seen a change in patients seeking CMV treatment this year. The majority have been on protease inhibitors for a year or more and have failed, he says. For most patients, the same rules for treatment will still apply, he adds.


1. Much D, Martin D, Gordon J, et al. Treatment of cytomegalovirus retinitis with a sustained-release ganciclovir implant. N Engl J Med 1997; 337:83-90.

2. Reed J, Schwab I, Gordon J, et al. Regression of cytomegalovirus retinitis associated with protease inhibitor treatment in patients with AIDS. Am J Ophthalmol 1997; 124:199-205.

3. USPHS/IDSA Prevention of Opportunistic Infections Working Group. 1997 USPHS/IDSA guidelines for the prevention of opportunistic infections in persons infected with HIV: A summary. MMWR 1997; 46(RR-12):1-46.