Protease inhibitors linked to drop in CMV retinitis

Big reduction in cases linked to immune restoration

Several studies report an added and welcomed benefit of protease inhibitor therapy: fewer cases of cytomegalovirus (CMV) retinitis, which is one of the most common and costly opportunistic infections in AIDS patients. At the same time, however, the studies warn clinicians that patients who have subclinical CMV may become symptomatic once their immune system is restored by the new therapies.

"It is exciting because we have a lot of data on what the clinical course of CMV retinitis has been, and now we have indications that it has changed dramatically," says Scott Whitcup, MD, clinical director of the National Eye Institute, part of the Bethesda, MD-based National Institutes of Health.

Although there is no official surveillance of CMV nationwide, Whitcup says the cases of CMV retinitis in AIDS patients seen at the institute have declined as much as 50%. Anecdotal reports indicate a similar phenomenon across the country as more patients are put on potent anti-HIV "cocktail" regimens that include protease inhibitors, Whitcup tells AIDS Alert.

A letter published recently in the Journal of the American Medical Association details Whitcup’s findings among four patients on protease inhibitors who were able to discontinue CMV treatment without their infection progressing to disease.1 CMV retinitis, which afflicts about 40% of AIDS patients, mostly strikes patients in the later stages of AIDS — usually after their CD4 counts have dropped below 50. Even with costly and toxic intravenous drugs, the disease often progresses after several months of treatment, causing blindness in many.

Restored immune system fights CMV

While it is possible that protease inhibitors alone are responsible for the findings, Whitcup and other researchers attribute them to the restoration of the immune system that results from these powerful drug cocktails.

"Our findings suggest that the combination of protease inhibitors and other anti-HIV medications can restore the immune system to a point where it can fight the CMV retinitis infection on its own," he says. "The use of anti-HIV drug combinations, when accompanied by an increase in CD4 cell counts, may allow us to change standard treatment for CMV retinitis."

These patients’ results are all the more striking when considering that even on treatment, most patients progress to the disease within several months. One patient was about to receive an eye implant with ganciclovir, but decided against it. Even after a year without treatment for CMV, the patient is free of the disease, Whitcup says. He also has observed what appear to be healed CMV lesions in patients on combination therapy.

"There are a number of people who have seen cases like this, and because they didn’t see the patients when it was active, they can’t say for sure," he says. "But there are few other things that look like healed CMV."

The current method of treatment for the disease is intermittent intravenous ganciclovir. This treatment can cause serious liver toxicity, and cost can range between $50,000 and $100,000 a year. Some patients with low CD4 counts also are being put on oral ganciclovir for CMV prophylaxis, which costs about $13,000 a year.

Whitcup cautions that without further study, it is premature to recommend that patients can discontinue CMV retinitis therapy once their CD4 counts reach a certain level. The National Eye Institute is recruiting patients for a larger clinical study that should help determine not only what is responsible for the findings, but also the most reliable point at which therapy can be stopped.

"At this point, we don’t recommend that people should stop therapy," he says. "I am hoping that within the next year or two we will have data to make some recommendations, but so far I am fairly optimistic that patients will do well off medication as long as their counts are up."

Therapy may increase CMV as well

Research from other CMV experts also corroborates Whitcup’s findings. However, at San Francisco General Hospital, researchers recently reported that AIDS patients who are given protease inhibitors may actually develop CMV retinitis. The findings seems to contradict the institute’s study, but on closer look the two studies report the same phenomenon.

Mark Jacobsen, MD, associate professor of medicine at the University of California-San Francisco AIDS Program, reports on five patients who had taken protease inhibitors and subsequently developed CMV retinitis. The findings, published in Lancet, are unusual because most patients have developed CMV with CD4 counts below 50 cells, while these five patients developed the disease after their immune system was considerably restored, to nearly 200 cells.2

"Now we have found that patients can have counts well over 100 and still develop CMV retinitis, and that it might be triggered by taking protease inhibitors," Jacobsen says. "This is a speculative time, but something is going on, big changes are happening, and we need to communicate our observations."

Jacobsen’s observations, however, have been misinterpreted to mean that protease inhibitors themselves are responsible for the CMV cases. Like Whitcup, however, Jacobsen doesn’t believe the phenomenon is drug-specific.

"The article says CMV retinitis is occurring in patients with higher CD4 counts and that is absolutely true, that is what we observed," he clarifies for AIDS Alert. "But what does that mean? It could mean that protease inhibitors are increasing CD4 cells but aren’t protecting people against the disease, but the observation in our five cases actually suggests those people had subclinical CMV infection before starting protease inhibitors and the inflammatory response of having an increased immune system response unmasked the retinitis."

Whitcup concurs with Jacobsen’s assessment, adding that in patients with low CD4 counts, the immune system doesn’t have the cells to fight infection — a response that is manifested as tell-tale floaters in the eye. Once the patients’ immune systems improved, they became symptomatic and were recognized as having the early stages of the disease. Whitcup notes that at least half of the patients at the Institute are asymptomatic when their CMV diagnosis is made. Both researchers also point out that the CMV was detected shortly after initiation of treatment. Had it developed more than six months or so after treatment in patients who had no prior evidence of CMV, then they might have reached a different conclusion.

"You don’t see it [CMV retinitis] downstream in people who had 50 T cells and then went up to 200 and stayed there and 10 months later developed CMV retinitis," Jacobsen says. "These cases are described soon after, which leads you to believe it was probably already there."

The take-home message from his study, says Jacobsen, is that clinicians should not be lulled into thinking that just because patients’ CD4 counts are rising, they are no longer at risk for developing CMV retinitis. "If you give protease inhibitors to a patient with low CD4 cells, you must be vigilant," he says. "Clinicians should be aware that patients with CD4 cell counts well over 100 can be affected, and that an increase in CD4 cells following protease inhibitor treatment is not necessarily protective."

That is not to say there may not be some unknown medical effect that protease inhibitors have on CMV retinitis. Although no data suggest that is the case, the phenomenon needs more study, Whitcup says. While the new antiretroviral drugs are able to boost the immune system, HIV causes a decline not only in the number of cells but in their diversity as well. Recent research has shown that some cells are not restored, and those cells may include those that fight against CMV, he adds. Further study may show that patients’ CMV outcomes depend on how low their CD4 counts are at initiation of combination therapy, or how high their CD4 counts rise with treatment, he adds.

Toward that end, the National Eye Institute’s study will enroll patients on protease inhibitors whose CD4 counts are above 150 cells, whose eyesight is not immediately threatened by CMV retinitis, and who are still on CMV treatment. Once CMV treatment is discontinued, they will be monitored to see if they progress or remain free of CMV disease. If most patients do well without their therapy, the trial will be expanded, and at some point guidelines for CMV treatment may be rewritten, Whitcup adds.

The results of the study could also affect standards for CMV prophylaxis. Hoffmann-La Roche of Nutley, NJ, manufacturer of ganciclovir, recently received FDA approval for the drug in capsule form and for use as preventive therapy. The rationale for CMV prophylaxis was based on the risk of disease prior to the advent of protease inhibitors. A significant decreased incidence in CMV retinitis could make preventive therapy less attractive. Currently, the U.S. Public Health Service recommends that preventive therapy for opportunistic infections be continued even if CD4 counts rise above the recommended threshold levels.

Indeed, even using previous estimates, the cost and risk of side effects weigh against the use of oral ganciclovir, according to a study published in the June issue of AIDS. Based on conflicting CMV preventive therapy studies, the authors conclude that preventive therapy is not effective, considering that it costs $39 a day, it is relatively toxic, and the expected gain in life expectancy is so low.3

"Our cost-effectiveness analysis found that even using the most optimistic assumption that it is 49% effective, oral ganciclovir preventive therapy still results in very small gains in life expectancy (five days) at a great cost," the authors write. "The resulting cost-effectiveness ratio, $1,762,517 to extend life by one year, is unacceptably high and far greater than other accepted prevention interventions."


1. Whitcup S, et al. Therapeutic effect of combination antiretroviral therapy of cytomegalovirus retinitis. JAMA 1997; 277:1519.

2. Jacobsen M, Zegans P, et al. Cytomegalovirus retinitis after initiation of highly active antiretroviral therapy. Lancet 1997; 349:1443-1445.

3. Rose D, Sacks H. Cost-effectiveness of cytomegalovirus disease prevention in patients with AIDS: Oral ganciclovir and CMV polymerase chain reaction testing. AIDS 1997; 11:883-887.