TORONTOPotent new anti-HIV three-drug cocktails are far more likely to fail in the real world of primary care than in carefully controlled clinical trials, researchers reported at the American Society for Microbiologys 37th Interscience Conference on Antimicrobial Agents and Chemotherapy (ICAAC). However, it appears that such failures are much less likely if patients are started on the three agents concurrently rather than sequentially.
State-of-the-art treatment for HIV infection currently calls for a regimen known as highly active antiretroviral therapy (HAART). This strategy combines one of the four currently approved protease inhibitorsritonavir (Norvir), indinavir(Drug information on indinavir) (Crixivan), saquinavir(Drug information on saquinavir) (Invir-ase), or nelfinavir(Drug information on nelfinavir) (Viracept)with two nucleoside analog reverse transcriptase inhibitors, usually zidovudine(Drug information on zidovudine) (AZT, Retrovir) and lamivudine(Drug information on lamivudine) (3TC, Epivir).
In clinical trials, HAART resulted in sustained reduction of viral load to undetectable levels in more than 80% of patients. But data from the public health clinic at San Francisco General Hospital show that most patients do not achieve such dramatic responses to protease inhibitors. This is of great concern, as even modest viral loads predict the outgrowth of HIV strains resistant to all protease inhibitors and signal disease progression.
Virologic failure to indinavir and ritonavir(Drug information on ritonavir) is more common in the primary care setting than that reported in controlled clinical trials, said Steven Deeks, MD, of the AIDS Program at the University of California, San Francisco.
Dr. Deeks and his colleagues performed a retrospective chart review of all patients at San Francisco General Hospital who received ritonavir or indinavir from March 1996 to March 1997 under the regular care of an experienced AIDS clinician.
They defined treatment success as a viral load of less than 500 HIV copies/mL on the two most recent viral-load measurements using a branched DNA assay.
Of the 137 patients who completed at least 24 weeks of treatment, more than half (53%) had evidence of failure. The success rate fell from 47% to 33% in a more rigorous intent-to-treat analysis.
Predictors of failure were low baseline CD4 count, high baseline viral load, prior nucleoside analog therapy, failure to change nucleoside analog therapy when indinavir or ritonavir was initiated, prior saquinavir therapy, and evidence of noncompliance or dose reduction, Dr. Deeks reported.
Prior Nucleoside Analog Therapy
Prior nucleoside analog therapy increased the risk of failure by a factor of 10.1 (P = .007). A baseline CD4 count of less than 200 cells/mL increased failure risk by a factor of 7.4 (P = .009).
A dramatic finding is that patients on at least 8 weeks of suboptimal therapy were more likely to fail, Dr. Deeks said. He also noted a high failure rate in patients with a clinical suspicion of noncompliance in following the rigid regimen of more than 20 pills a day.
Dr. Deeks suggested that earlier use of aggressive treatment strategies would increase the success rate, and noted that his findings essentially confirm data reported earlier in the ICAAC meeting by New York University researcher Roy Gulick, MD.
The Gulick et al data came from a clinical trial in which 97 zidovudine (AZT)-experienced patients (2.5 year median duration of AZT therapy) received either concurrent or sequential triple therapy with indinavir, AZT, and lamivudine (3TC); that is, they either received the three-drug regimen from the outset or were switched to the triple combination after 24 to 52 weeks treatment with AZT/3TC or indinavir alone.
Like the patients in Dr. Deeks study, patients who began treatment with suboptimal therapy had a disappointing success rate: Only about 40% of those originally assigned to the AZT/3TC or indinavir monotherapy groups achieved a viral load of less than 500 copies/mL after a switch to indinavir/AZT/3TC.
In contrast, 80% to 85% of patients originally receiving the three-drug regimen achieved a viral load of less than 500 copies/mL, and this success was sustained to the end of the two-year study period.
There is a striking difference between concurrent and sequential triple-drug therapy, Dr. Gulick said. Delayed, sequential triple-drug therapy achieves more modest antiretroviral effects.
Patients who originally received indinavir/AZT/3TC had continually rising CD4 cell counts; after 100 weeks of treatment, this reached a median increase from baseline of 230 cells/mL. Patients with delayed triple-drug therapy had a smaller increase over baseline CD4about 100 cells/mLthat reached a plateau and did not continue to rise.
Dr. Gulick concluded that for patients who have received prior therapy with nucleoside-based reverse-transcriptase inhibitors, it is not enough simply to add a protease inhibitor to existing regimens. At the time a protease inhibitor is added, he said, at least one potent new reverse-transcriptase inhibitor should be added as well.
Adding two potent drugs is important, not just adding two drugs, Dr. Gulick said. It has only recently been appreciated that you have to start the two new drugs at the same time.