September 28-October 1, 1997
Steve Deeks and colleagues from San Francisco General reported an important study that looked at triple drug therapy in "real world" circumstances. The results are sobering. The investigators reviewed the records of all patients followed in the program who started a potent protease inhibitor between March 1996 and March 1997 (when nelfinavir was licensed.
Only patients followed by an experienced HIV provider were included. Of 196 who began therapy, 136 stayed on treatment and could be evaluated. Protease inhibitors were always combined with nucleoside. Success was defined as having a viral load less than 500 by BDNA on the 2 most recent visits.
Only 47% of patients had a successful response to HAART. Predictors of failure included higher baseline viral load, lower baseline CD4 count, not changing the nucleoside backbone at the time of starting protease inhibitors, and suspected problems adhering to the regimen. If the notes indicated that there might be problems in following the regimen, patients were more than 15 times more likely not to maintain an undetectable viral load.
Clinical trials tend to show drugs in a best case situation, and reports of small, intensive trials among extremely well motivated volunteers (such as Merck 035 -- see below) raise expectations beyond what can usually be achieved. The risk factors for treatment failure are not really surprising, but are very important. Protease inhibitors should be started with at least one, and preferably two other potent drugs. Failure to do so in this study was associated with a 14 fold increase in the risk of breakthrough. The measurement of how well patients took their medicines was very limited, but the message is clear. Another poster from a group in San Diego also reported relatively low success rates in general clinical practice. Despite the enormous benefits of protease containing triple combination regimens, they are challenging for doctors to use properly and difficult for patients to take.
Do children also benefit from protease containing regimens? Dr. Nachman reported preliminary results from pediatric ACTG study 338, which compared a ritonavir-containing 2 drug regimen, a ritonavir-containing 3 drug regimen and AZT+3TC in children with previous nucleoside treatment. Of 298 children in the trial, results are available for 162. At 12 weeks, the median viral load decline was 0.33 for the AZT+3TC group, compared to 1.81 log and 1.65 log for the two ritonavir-containing regimens. Only 14% had viral loads below 400 copies in the AZT+3TC group, compared to 61% and 57% in the two ritonavir groups. Because there is no difference yet between the d4T+ritonavir and the AZT+3TC+ritonavir group, these arms remain blinded, and the study will try and determine if there is any difference between the two.
This study clearly demonstrates that protease regimens are superior to 2 nucleosides in drug experienced children, just as in adults. PACTG 300 recently reported that AZT+3TC or AZT+ddI was superior to nucleoside monotherapy in children. The message so far seems to be that the principles of antiviral therapy learned in adult trials hold true (in general) in children. Draft guidelines for antiviral treatment have just been released, and will be published after a comment period.