What Stephen Deeks actually said at ICAAC about HAART and the widely reported 53% failure rate. The following was taken verbatim from a lecture that Deeks presented on December 4, at a conference in San Francisco, "Clinical Care of The AIDS patient.
Stephen Deeks, Assistant Professor of Medicine at UCSF and San Francisco General, is the one who broke into the headlines after ICAAC, when he reported the 53 % failure rate after one year of therapy on HAART treatment and has received a lot of national attention.
Why rehash old news. Because I still read the same statistics everywhere, without any reference to the three factors that Deeks found are predictive of failure. Deeks actually found that when protease inhibitors are used early and aggressively as indicated by recent guidelines, they work as they are supposed to and found a very high success rate, phenomenally high.
This lecture discusses" the study" that was incorrectly reported by the media. It is available only in real audio. Listening to the lecture requires a Real Player (free download.) hivinsite.ucsf.edu
Then click on "Establishing Control of Virologic Failures" by Stephen Deeks. This is part of Deek's lecture.
"Assume this is currently the goal of therapy, with "hit hard" therapy, you take a patient, and drive the viral replication down to zero, and as a consequence viral load drops to very low levels, zero, and the assay drops down to undetectable levels, and clearly there is convincing data that this can go on for 2-3 years. This is currently the goal of therapy and this would be a virologic success.
"How to define failure: Patient who despite being on a potent anti-retroviral regimen, who does not achieve undetectable levels, or has a viral load>500 copies, after at least 3-4 months on therapy. That is indication of viral replication.
"If you take a patient that is treatment naïve, has a reasonably low viral load, is immunologically intact, adherent, and if you take that idealized patient, and you give two nucleotide analogues and a protease inhibitor, you will achieve in 70-90% of patients, the goal of therapy, which is long term viral suppression, and that is what we have seen in clinical trials. The big problem trouble is clinics are not made up of such perfect idealized patients and actually what we have is a very heterogeneous group of patients in real world.
I'D LIKE TO SHOW THE DATA MAINLY TO UNDO SOME OF THE DAMAGE THAT CAME OUT OF THE MEDIA COVERAGE AT ICAAC.
How do these drugs actually work in the real world was a topic of a study we did this summer and got a fair amount of media attention at ICAAC.
We did a retrospective cohort study, and analyzed the medical records of all patients followed in our clinic, which is San Francisco Hospital, an urban medical clinic, who had been seen at least twice by the same physician, who devoted at least 20% of their time to primary HIV management. We looked at data FROM MARCH OF 96 TO MARCH OF 97, WHEN RITONAVIR OR INDINAVIR WERE WIDELY AVAILABLE, up UNTIL MARCH 97, when Nelfinavir (Viracept) became widely available (March 97), and prescribing patterns changed dramatically.
WE WERE NOT INTERESTED IN THE EXPERIENCE OF SAQUINAVIR BUT FOCUSED MAINLY ON THE SO CALL POTENT PROTEASE INHIBITORS. We identified 450 patients, only 135 actually received indinovir or ritonavir and stayed on those drugs for 6 months. And only 47% achieved a success rate of at least 2 viral loads < 500 copies for 6 months, most were over 5000 copies. This was not the most important part of the study, but unfortunately this was the number that got bandied about.
DOES THIS MEAN THAT A PATIENT WHO IS GOING TO MAKE A DECISION TO TAKE PROTEASE INHIBITORS HAS ONLY A 50% CHANCE OF SUCCESS? ABSOLUTELY NOT. Bob Grant went back, did a multi-variant regression analysis, to see what would predict success or failure. We went back and looked at baseline predictors of failure and this baseline is still developing:
PREDICTORS OF FAILURE:
1. Having a viral load>100,000 when protease inhibitor therapy is initiated.
2. Adding a protease inhibitor to a pre-existing nucleoside reverse transcriptase, and not switching to a new nuke reverse transcriptase inhibitor or non-nucleotide reverse transcriptase inhibitor.
3. Evidence of non-adherence
We went back and, and we found that these three factors were powerful independent predictors of a poor outcome. If you actually look at the 33% of patients who modified their nucleotides at the time they initiated protease inhibitors, we have a very high success rate, phenomenally high.
We looked at this data and we thought that this data actually supports very early and aggressive use of protease inhibitors and that in general, these drugs do what they are supposed to, if and only if, they are taken as they are supposed to according to current guidelines of ISUSA and US Dept. of Heath and Human Services, then these drugs will work."
From Clinical Care of the AIDS patient: December 4-6, 1997, San Francisco.
hivinsite.ucsf.edu
Establishing Control of Virologic Failures
Steven Deeks, MD
|
