I guess I am missing something but I haven't been able to see anything "eye-popping" about the Sustiva data except that it has managed to fool the majority. It's not hard to give a positive spin to the data, emphasizing the "On Treatment" analysis, which only observes those patients remaining at the end. That is what "On Treatment" analysis is. I listened to Schlomo Staszewski's lecture available at: webcast.aids98.org
Dr. S.'s explanation of the 3 kinds of analysis:
1. Observed Data or On treatment analysis.
Looks exclusively at the patients being treated, and missing data is NOT accounted for.
2. "Intent to treat: There are two types.
a. Intent to treat: LOCF: Last Observation Carried Forward: where missing data is assumed to be the same as the last available data. So the viral load would be assumed to be the same at 20 weeks as 16 weeks if measurements were not available.
b. Intent to treat: Non-completer=failure the most rigorous analysis.
All missing data is assumed to be a failure or anyone who does not complete the study is counted as a failure.
That "94.5% of subjects below 400 copies at 24 weeks" that Dr. Eron refers to as "eye-popping" used "On Treatment" analysis .
Comparing the three methods, the results were: 94.5%; 84% and 75% for
"On Treatment" vs "Intent to Treat:LOCF" vs "Intent to treat" NC=F, where non-completer=failure, respectively for patients who were <400 ml at 24 weeks.
Unfortunately the more important data, the number of patients who were <50 copies/ml is not given. If someone knows the numbers for patients<50 ml, please post them. All we know is that in this "Landmark 006 study" 19% more patients in the Sustiva arm than in the Crixivan arm were <50 copies.
Dr. Eron does say: "What becomes clear is that in studies that use a pure intent-to-treat analysis plan, on-treatment rates may greatly effect the interpretation of the results."
healthcg.com
In study 005: EFV+AZT+3TC the results showed that 11/11 patients or 100% were <400 ml at 36 weeks, another "mind-boggling figure." Until you see that 23 other patients in the same study either dropped out or were removed for various reasons which are unclear, but include adverse reactions, viral failure etc.
Using "Intent to treat-LOFC" analysis the numbers are: 34.6% <50 copies ( 9/26)
Using "Intent to Treat-NC" or 26% are <50 copies at 36 weeks (11/34)
Yet using ON TREATMENT; 81%<50 copies (9/11)
Does anyone know if the Amplicor 400 copy test is approved by the FDA now? A report written
1/21/98 at the NATAP site, natap.org says that the Amplicor 400 copy test, used in these Dupont's studies, is not FDA approved because it is not sensitive or reliable.
>>Investigators used the Amplicor 400 copy test to see if there was any virus signal. They found that they were unable to detect virus signal from 80% of the individuals in the efavirenz+indinavir arm using the 400 copy test. This does NOT mean viral load was not present in the blood. Using the test for this purpose is not FDA approved because the results are not considered reliable enough. ..In fact, the 400 copy test is not as sensitive as the Roche Ultrasensitive test."
<<At Lake Maggiori a researcher said that viral load should be reduced to zero, that cells containing latent virus do decay and that eradication MAY be possible after 7-8 years of continual viral load suppression to zero or very low levels. Over time viral load may continue to decline even after initially reaching <50 copies/ml.>>
Resistance develops quickly in NNRTI's, that is one reason they are used together with a protease inhibitor, to help suppress resistance and that is why Dupont used this high dose of 600 mg to try to prevent resistance, even though there were more side effects at this dosage than at lower doses.
Re Dr. Eron's statement: "The ease and administration and tolerability of the EFV/3TC/ZDV regimen cannot be underestimated."
Then why is the discontinuation rate so high in that "Landmark 006 study. In the EFV+3TC+AZT arm, 20.8% discontinued at 24 weeks, and 38% discontinued in the Crixivan arm. Both rates are very high.
Another quote: "Bravo to Dr Staszewski and the DuPont-Pharma team for going head-to-head against a "gold-standard" regimen"
Crixivan+AZT+3TC a gold-standard ??? Crixivan?? Maybe last year. If it still is, why did Crixivan's market share fall from ~60% in March 1997 to 33% in June 1998 if it is still the
"gold-standard." while Viracept's market shared increased to ~32% from 3/97 when it was approved.
>>> Don't know how this data squares with the information Margie posted about changes in patients dosing regimens while on study ... that's not apparent from this presentation of the data
- preliminary data from a small study shows Viracept a little less potent in combo than Sustiva (may be just noise in the numbers), and both together more potent than either alone (Rick's point )
I listened to Margaret Albrecht's presentation of ACTG 34 on the AIDS98Webcast site. Sustiva does raise the level of Viracept by 21-26%, but obviously that doesn't mean the results will be 20% better than without Sustiva.
All the patients in this trial were heavily pre treated with nucleosides and had completed two previous trials of nucleoside treatment. You do have to compare apples to apples. You are comparing patients heavily pretreated with nucleosides, with higher baseline viral loads and lower CD4 cell counts to those in Dupont's studies who were mainly treatment naive subjects with no previous 3TC, no PI's, no NNRTI's and had lower baseline viral loads and higher baseline CD4 counts.
In ACTG 364, the three arms were: Nelfinavir (NFV)+2 nukes; Efavirenz (EFV) +2 nukes; and EFV+NFV plus one nuke. All the patients in ACTG 364 were rolled over from two other trials, ACTG 175, where they received either AZT or ddi monotherapy and then they were put in ACTG Trial 302 or 303 where they received d4t monotherapy or AZT + 3tc, AND THEN they were put in 364. Dr. Albrecht commented on the surprisingly low discontinuation rate. Dr. Albrecht also commented that the nelfinavir arms had the highest baseline viral loads of the three arms: The baseline viral loads were: NFV viral loads ranged from <500 to 1,284,874 (mean~ 8700) vs EFV of <500 to 421,811 (mean-7600)and the NFV+EFV arm of <500 to 348,000 copies/ml (mean-6770).
At 16 weeks 81% of NFV/EFV, 64% of NFV, 69% of EFV were<500 VL at 16 weeks and the viral loads were getting lower with time. I didn't see data for <50 copies. In the Sustiva+2 nuke arm in 005, it looked like the Viral loads were increasing with time and the CD4 cell counts decreasing. So although the data looks a little weak at first glance, the higher initial viral loads in the NFV arm and the history of heavy pretreatment with nucleosides in all patients, make this data more impressive, not less impressive.
Skowron used nelfinavir+nevirapine+d4t in patients who were PI na‹ve and had less than 6 months of d4t. At 21 weeks, 87% <400 copies and 73% were <50 copies in 25 patients.
As far as Sustiva lowering the levels of Crixivan, it has been addressed, indirectly. In Dupont 005 and 006, when IDV (indinavir) was used with EFV, the dose of Indinavir was increased to 1000 mg q8hr. The normal dose is 800 mg/q8hr.
A great site for all HIV drug interactions is at the HIVINSITE:
hivinsite.ucsf.edu
It even says now not to take Fortovase with Sustiva although most physicians who prescribe it are doubling the fortovase dose. The list of drug interactions for Norvir has grown since I looked last. Norvir has the most drug interactions of all the protease inhibitors..
As far as the denouement of protease inhibitors, I guess some of the well respected physicians in HIV research haven't heard, or don't read the WSJ, like Dr. Charles Flexner, Pedersen, Dr. Cassy Workman, Professor Danner or why would they have recommended a few days ago that initial treatment consist of dual protease inhibitor use- Ritonavir and Saquinavir.
Professor Danner concluded, ''The main benefits of dual-protease therapy are strong viral suppression, more convenient dosing, and the need for patients to take less medication (Ritonavir+Fortovase) which may improve patient adherence and therapeutic outcomes. For these reasons, dual-protease therapy should be considered one of the promising options for first-line treatment for both protease-experienced and naive patients.''
A study presented by Dr. Flexner evaluated 20 protease-inhibitor naive patients who were evenly randomized in a dose-comparison trial of ritonavir 400mg/nelfinavir 500mg vs. ritonavir 400mg/nelfinavir 750mg. All medications were administered twice daily.
At 32 weeks in the low-dose nelfinavir group, six patients had viral load suppressed to less than 400 copies/ml while eight patients in the high-dose nelfinavir group had viral suppression of less than 400 copies/ml. The BID dosage of Nelfinavir is 1250 twice a day, so maybe the results would have been better with a higher dose of Nelfinavir.
''The combination of ritonavir and nelfinavir is a potent and promising regimen in the management of patients with HIV,'' said Dr. Flexner. |
