Roche Protease (Proteinase) Combination Trial: Triple Combination Results

A study sponsored by the U.S. National Institute of Allergy and Infectious Diseases found that a three-drug combination (the Hoffmann-La Roche protease (also called proteinase) inhibitor saquinavir [formerly called Ro 31-8959], plus AZT, plus ddC) worked significantly better than two separate two-drug combinations against which it was compared (saquinavir plus AZT, and AZT plus ddC). This trial, in 302 patients, looked at changes in blood tests, such as T-helper count, and viral measures, over 24 weeks. This study was not designed to look for differences in clinical events, such as different rates of progression to AIDS, or death. And there were too few such events in this trial to allow any meaningful comparison. As a result, the researchers concluded that "standard clinical patient management should not be affected by this study."

The improvement in blood work was modest, but still important. For example, the mean T-helper count at baseline was 156 in the triple-combination group; this increased to 195 after four weeks of treatment, 200 after eight weeks, 202 after 12 weeks, 189 after 15 weeks, and 180 after 24 weeks. In other words, the average T-helper count increase with the triple therapy was, at its maximum, 46; this average then began to fall, and at week 24 (the latest point at which data is now available), the average T-helper increase was only 24.

All patients had been taking AZT for at least four months before beginning this trial -- and three quarters of them had been taking AZT for over a year. Therefore, AZT alone would not be expected to have accounted for a T-helper count increase. [Note: A new trial of saquinavir, a large, international study which could begin in a couple months, will test the combination in patients who have had little or no prior use of AZT. If AZT resistance limited the effectiveness of saquinavir in the current trial, which seems likely, the new study might show better results.]

Viral load data, including HIV RNA, is still being analyzed. Some viral culture data is available, however; and it also suggests that the triple-drug treatment may be clearly better than the two-drug treatments. But even the three-drug combination never quite achieved a 10-fold reduction in viral load in the tests which have so far been run -- and early indications are that a hundred-fold or greater reduction may be necessary before dramatic benefits may occur (see "Antivirals and Immune Recovery," AIDS TREATMENT NEWS #200, June 3, 1994).

This study also looked at safety, and found that, "The type, severity, and frequency of adverse experiences were the same in the three treatment regimens. The small number of AIDS- defining events and two deaths resulted from complications of advanced HIV infection and did not appear to be associated with the study therapies." (All patients had T-helper counts between 50 and 300 at entry to the study, in order to qualify for enrollment. About 52 percent were symptomatic but did not have AIDS; about 36 percent were asymptomatic; and about 11 percent had AIDS when starting the study.)

The doses used were fairly standard: AZT 200 mg three times a day (a total of 600 mg per day), ddC 0.75 mg three times a day, and saquinavir 600 mg three times a day (probably the most common dose now used in trials).

More data will become available from this trial, and results will be further analyzed for publication in a peer-reviewed journal.

Comment

This study represents a modest advance in learning how to use a new drug. The result does not have immediate practical use, however, because saquinavir is not widely available; according to Hoffmann-La Roche, it is very difficult and expensive to manufacture.

No one knows how long it will take for saquinavir to be approved, even if future trials continue to look good. The big question is whether the FDA will allow accelerated approval -- approval based on blood tests, but on condition that the company complete the much longer lasting trials required for statistical proof that there is less death or progression to AIDS with the new treatment than with standard treatment. The FDA may be reluctant to grant accelerated approval, because saquinavir is a new kind of drug (a protease inhibitor), and the FDA may be cautious about trusting the blood tests in this case, since there are more unknowns. Also, there has been much unhappiness with how Roche handled the followup trials after its previous accelerated approval for ddC.

Hopefully the new tests for viral load will provide a way out of this kind of dilemma. If they are found to be reliable as indicators of future patient outcome, in a wide variety of different trials and different kinds of patients, they might be trusted enough to be the main basis of drug approval, even without statistical proof of reduced death and progression to AIDS. (Fortunately, such validation of the tests can often be done quickly, without needing to wait for years to see if the predicted disease outcomes occur, by using frozen blood samples saved years ago in clinical trials.) And, alternatively, if a treatment were powerful enough to clearly restore many people to health when they were chronically ill, then this would be a "clinical" benefit which could be shown very quickly, leading to approval without the need to wait for rare "events."

It is possible that saquinavir would work better if the dose were increased; a small trial at Stanford is now trying the drug at doses two and four times the one used in the recent trial. Unless manufacturing can be improved, however -- or unless the drug can be formulated to be more orally bioavailable, reducing the amount of it required -- there is concern that saquinavir may be too expensive to become widely used.