Two studies presented at the 10th CROI investigated use of structured treatment interruption (STI) in populations of patients with very advanced HIV disease and long treatment histories. Typically, these individuals are very difficult to treat, having developed resistance to most available antiretroviral drugs in each therapeutic class (multiple drug resistance -- MDR). Because of the potential for serious consequences due to disease progression while off therapy, especially in these advanced patients, clinical use of the technique of STI has been controversial. One theoretical rationale for its use in a salvage setting, however, is to allow drug-resistant HIV species to become overgrown by drug-susceptible wild-type virus, which might then be suppressed by available drugs. Another rationale for STI in this population is to allow individuals suffering from drug toxicity a period of drug-free time in which to recover. Although two trials investigating these issues were discussed in Boston, they showed contradictory results and aren't likely to settle the controversy. Nonetheless, the larger of the studies produced compelling evidence of the dangers of unsupervised STI for individuals with advanced HIV disease.
After nearly a year of follow up 16 patients had died, with 8 deaths occurring in each comparison arm. Overall, 34 patients experienced clinical disease progression or died while on the study, with 22 of those events occurring in the interruption group and 12 in the continuing group. After noting that patients in the interruption group were not likely to be protected from disease progression, a Data Safety Monitoring Board overseeing the study halted new enrollments into CPCRA 064 in June of 2002.
In addition to experiencing more clinical events, patients who interrupted treatment before changing their regimens also had poorer CD4 cell count responses and higher HIV RNA viral loads than those who switched immediately, although these trends were diminishing during a follow-up period out to one year. There were no benefits for treatment adherence or quality of life with either strategy. The study investigators recommend that patients with multiple drug resistant HIV should be maintained on an optimized antiretroviral regimen and should not undertake an interruption before switching. (Abstract 67, 10th CROI)
Another trial also investigated the strategy of using treatment interruption in people with advanced HIV disease and multiple drug resistant virus.
Christine Katlama, a clinical investigator from the Hospital Pitie-Salpetriere in Paris, reported on a study of an intensive HIV regimen called GIGHAART in highly treatment experienced patients. Sixty-eight participants were randomized to either receive the GIGHAART regimen (containing 6 to 8 drugs) immediately or to wait for 8 weeks before starting therapy. The median CD4 cell count of study participants was 26 cells/mm3, indicating the advanced stage of HIV disease in this group. At the time of study entry, participants' HIV viral loads were not being controlled by therapy and all had previously received multiple drugs from each therapeutic class.
Twelve weeks after starting the GIGHAART regimen, patients in both study arms had experienced reductions in HIV viral load, although there was significantly improved reduction among those who had interrupted treatment as compared to those who began their new regimen immediately. The median decrease in plasma HIV RNA in the delayed treatment arm at week 12 of therapy was -1.91 log copies compared to -0.37 in the immediate arm. While only 15 percent of those starting GIGHAART without delay had undetectable viral load after 12 weeks of therapy, 38 percent of the deferred treatment group were undetectable.
The interruption strategy produced favorable results in other study parameters as well. After nearly a year of follow-up, CD4 cell counts were up by 69 cells/mm3 in the deferred arm compared to a median increase of 7 cells/mm3 for those who started immediately. Overall, in this difficult-to-treat population of people with advanced HIV disease, an 8-week interruption before starting an intensive antiretroviral regimen was associated with improved virological and immunological results that were sustained out to one year.
During a discussion following her talk, Dr. Katlama speculated that the superior potency of the GIGHAART regimen might explain why these results differed from those in CPCRA 064. Commenting on the shorter duration of STI in her study, Katlama also said she didn't believe that it was necessary to wait for the wild-type virus to come back completely as long as drugs could be found to keep the virus down. (Abstract 68, 10th CROI)
A paper recently published in the journal AIDS by Steven Deeks of the University of California, San Francisco, also addressed the role of STI in salvage therapy. In a non-randomized observational study, 24 patients who were experiencing virological failure despite remaining on HAART elected to stop all ARV medications for at least 12 weeks. Following re-initiation of treatment, patients' viral genotype, phenotypic drug susceptibility, viral load and CD4 counts were monitored. In a previous study, Deeks had shown that during a treatment interruption in patients with MDR virus, drug-susceptible wild-type virus usually eventually outgrew the less replication competent virus population that had been selected by drug pressure. In an extension of that investigation, this study looked at the long term effects of what happened when patients restarted their various ARV regimens after STI.
Fifteen subjects (64%) maintained viral load below 200 copies for up to 109 weeks of follow-up after restarting therapy. Twenty of the subjects had a shift in viral phenotype from resistant to susceptible during the period off drugs. After restarting therapy, 13 of these 20 patients were able to suppress and maintain viral load below 200 copies within the follow-up period. This durable viral suppression did not occur in any of the 6 patients who restarted therapy without adding at least one drug to which their baseline virus was phenotypically sensitive. In contrast, all 9 patients who started a regimen with at least one drug to which their baseline viral population was susceptible achieved viral suppression. For 5 of the 9, the new drug was an NNRTI. Finally, 4 of the 5 who restarted with a regimen containing at least 2 drugs active against their baseline viral population also sustained successful suppression through follow-up. (AIDS 2003; 17(3):361-370)
The theory behind this approach in the MDR patient population is that after stopping drugs, the less-fit MDR virus will be overgrown by a drug-susceptible wild-type virus that has been waiting quietly in the viral archives. Once the drug-resistant strain has been sent to the archive and the WT reestablished, therapy including several recycled drugs is restarted and the dominant drug-susceptible population is suppressed. As Deeks has demonstrated in previous studies, if only recycled drugs are used, the MDR virus soon bounces back and viremia blooms within weeks. But the new study suggests that, if only one new drug is added along with the recycled meds, then this may be sufficient to keep the less replication-capable MDR strain from establishing a foothold. The key is the STI, which allows the MDR strain to retreat to archival levels. If the MDR strain were still actively replicating as the dominant strain when treatment was switched, there might be a temporary lowering of viral load, but resistance to the single highly active new drug would quickly appear and the evolved virus would be all that tougher to treat.
The conclusion that adding only a single new drug may provide durable viral suppression after an STI could be welcome news for people with MDR virus who perhaps can't access more than one drug that their virus is sensitive to. With Fuzeon waiting in the wings, this finding, if confirmed, may have profound implications for the outlook for salvage therapy. Conventional wisdom during the past couple of years has held that to forestall resistance after changing regimens, patients must add at least two new drugs to which their virus is sensitive. This may still be true, especially when an STI is not feasible or too risky, since, as the CPCRA trial warns, taking an STI may not be such a good idea for people lacking the immunological cushion to keep them from getting sick while off therapy. Finally, if no new drug is accessible, earlier work by Deeks demonstrating that drug resistant virus may be less replication competent and perhaps less pathogenic suggests that it may be advantageous to remain on a failing regimen for as long as tolerable until newer drugs come along.
In an extension of his salvage STI paper in the journal AIDS, Dr. Deeks also presented a poster at CROI on the selective interruption of only one component of HAART. This non-randomized study in 20 patients applied a finer scalpel to the interruption strategy by halting only the PIs and continuing NRTIs (or vice versa). It also introduced a new acronym to the literature, PTI, for partial treatment interruption. The study participants' median CD4 count was 336, the median viral load was 3.9 log copies/mL, and all had been experiencing persistent viremia despite good adherence. The decision whether to halt PIs or NRTIs was based on each individual's toxicity profile.
For 15 subjects who interrupted all PIs and continued NRTIs, viral load and CD4 counts remained stable, while triglycerides and cholesterol were significantly reduced by week 12 of the intervention (TG by -90mg/dL; non-HDL-C by -30mg/dL). Genotypic and phenotypic resistance remained stable past week 16, although in 2 patients, drug susceptible PI mutations began to dominate by week 24 and viremia increased as viral replicative capacity improved. Overall, stopping PIs improved lipid values and staying on failing NRTIs continued to provide some virologic benefit. Larger, randomized studies are needed to confirm this.
The 5 patients who stopped NRTIs but continued their PIs did not fare as well, experiencing immediate and sustained viral load increases of about 0.03 log copies per week. Three of the 5 eventually had their M184V 3TC resistance mutation disappear, which was accompanied by an up-tick in viral replication fitness. (Abstract 640, 10th CROI)
Complete or partial treatment interruption as a tool to guide viral resistance properties may have some promise in settings where close monitoring is assured, but at this stage it remains a risky undertaking best reserved as a subject for additional research.
|Comparison of CD4 and Viral Load Changes in STI Studies for People With Multi-drug Resistant Virus|
|# of pts on STI||135||34||24|
|Baseline CD4 (cells/mm3)||180||28||218|
|Baseline VL (log10 copies/mL)||5.0||5.4||4.6|
|Duration off therapy||16 weeks, fixed||8 weeks, fixed||Median 20 wks pt. choice|
|End of STI:|
|CD4 change from baseline||-53||-10||-84|
|VL change from baseline||+0.31||+0.16||+0.76|
|48 Weeks After Restarting Treatment:|
|CD4 change from baseline||+7||+69 (40 wks)||-3|
|VL change from baseline||-0.76||-0.79||-2.00|