November 19, 1999
What are the best combinations for a treatment naïve patient? What is the best way to sequence nucleoside reverse transcriptase inhibitors? Are non-nucleoside reverse transcriptase inhibitors as potent as protease inhibitors in an initial combination? How active is a triple nucleoside combination? Answers to some of these questions were provided in several reports at this meeting. We've already seen the data from these studies at previous meetings. However, at this meeting data of one year's duration or more was presenting providing the opportunity to judge the durability of responses seen in earlier reports. I will report on three studies, Dan Kuritzkes will discuss others.
Joe Eron presented the one-year data from the START II trial (Abstract 14), a study comparing two nucleoside reverse transcriptase inhibitor combinations, stavudine plus didanosine (d4T + ddI) or zidovudine plus lamivudine (ZDV + 3TC), together with indinavir (IDV) in antiretroviral naïve patients with CD4+ counts greater than 200 cells/mm3 and viral loads greater than 10,000 copies/mL. Approximately 100 patients were enrolled into each arm of the study. Baseline CD4+ counts were in the low 400's in each group; the baseline viral loads of the d4T/ddI/indinavir and ZDV/3TC/indinavir groups were 29,512 copies/mL and 38,905, respectively.
In the as-treated analysis that just considers those patients continuing in the trial, after 48 weeks, 83% of patients on the d4T/ddI/IDV arm had viral loads < 500 copies/mL and 69% had viral loads < 50 copies/mL, compared to 79% and 68% of patients receiving ZDV /3TC/IDV. The discontinuation rate after one year was 38% among the d4T/ddI recipients and 44% among the ZDV /3TC recipients; 16% of subjects in each arm discontinued study medication because of toxicity. As a result, in the intent-to-treat analysis, which considers all patients and counts those who discontinue as treatment failures, after 48 weeks, only 41% of the d4T/ddI/IDV group and 35% of the ZDV/ 3TC/IDV group had viral loads < 50 copies/mL. Therefore, d4T + ddI proved as active as ZDV + 3TC.
Differences between the two treatment arms were observed with respect to CD4+ count changes and the types of side effects patients experienced. At week 48, mean CD4+ counts had increased by 214 cells/mm3 in the group that got d4T + ddI versus a 142 cell/mm3 increase in the group that got ZDV + 3TC (p = 0.026). However, there was no difference in mean increase in CD4+ T-cell percentages between the two groups. Patients who got d4T + ddI were more likely to experience diarrhea, headache, fever, and increases in hepatic transaminases and triglyceride levels. While those receiving ZDV + 3TC were more likely to complain of nausea and have lower absolute neutrophil counts.
This study demonstrates that d4T + ddI and ZDV + 3TC are comparable dual nucleoside options together with a protease inhibitor when used in a treatment naïve patient population.
The 72 week data from the DuPont 006 study was presented by Karen Tashima. (Abstract 366) This study is the trial that demonstrated 3-drug combinations with efavirenz (EFV) are at least as potent, and better tolerated, than 3-drug combinations with indinavir, and moved us into an antiretroviral therapy era in which protease-sparing strategies deserve serious consideration.
In this trial of predominantly treatment naïve patients, individuals were randomized to receive open-label zidovudine, lamivudine and efavirenz, zidovudine, lamivudine and indinavir, or the 2-drug combination of efavirenz and indinavir. After 72 weeks, 7% of patients had discontinued medication due to side effects in the two efavirenz arms, compared to 23% of patients who discontinued due to side effects in the ZDV/3TC/IDV arm. Virologic failure (viral load >400 copies/mL) occurred in 1% of patients on the 3-drug combination with efavirenz and 6% in the 3-drug combination with indinavir. In the intent-to-treat analysis which considered all reasons for drop out as a treatment failure, 60% of the ZDV/3TC/EFV recipients had viral loads < 50 copies/mL compared to 40% of the ZDV/3TC/IDV recipients. Among those patients with baseline viral loads greater than 100,000 copies/mL (43 - 48 patients in each arm had baseline viral loads above this threshold), 58% of patients on the ZDV/3TC/EFV arm had viral loads < 50 copies/mL after 72 weeks, compared to 30% of patients on ZDV/3TC/IDV. Therefore, the triple combination with efavirenz was as equally active in all patients, irrespective of their baseline viral load, while the triple combination with indinavir was not as active in patients with high viral loads as it was in patients with viral loads < 100,000 copies/mL. Lastly, in an analysis that just considered those patients able to achieve a viral load < 400 copies/mL, patients who received ZDV/3TC/IDV were more likely to experience a viral load rebound to > 400 copies/mL than those who received ZDV/3TC/EFV. This data shows that the superior efficacy of an efavirenz based regimen compared to an indinavir based regimen continues over a 72 week follow-up period.
Chuck Hicks presented the 48 week data from the CNA3003 trial (Abstract 314), a trial comparing ZDV + 3TC with and without abacavir (ABC) in treatment naïve patients. After 16 weeks, patients with viral loads > 400 copies/mL had the option of adding open-label abacavir, and adding or switching other licensed antiretrovirals. This report provided additional details on the original ZDV + 3TC group. After 48 weeks, approximately half of the original ZDV/3TC/ABC group have a viral load < 400 copies/mL by an intent-to-treat analysis. At week 16, less than 40% of patients originally assigned to ZDV + 3TC had a viral load < 400 copies/mL and 92% of patients tested had a codon 184V mutation associated with 3TC resistance. However, following the addition of abacavir with or without the addition of other antiretroviral agents, the proportion of patients with viral loads < 400 copies/mL in these patients was similar to those originally treated with the 3-drug combination. This observation calls into question the significance of the codon 184V mutation, and whether 3TC should be continued or stopped in patients who develop it. In either case, patients experiencing a viral load rebound who have a codon 184V mutation need to have additional antiretroviral agents added to their combination. Whether only a single agent or multiple drugs should be added is a subject for future investigations. However, this study indicates that abacavir remains effective in patients experiencing virologic failure with a codon 184V mutation.
Although not discussed in this presentation, the triple combination of ZDV/3TC/ABC was not as efficacious in patients whose baseline viral loads were greater than 100,000 copies/mL.
There was a fair amount of gastrointestinal toxicity observed in this study with 47% reporting nausea, 23% experiencing diarrhea, and 15% with vomiting. Most of these side effects were mild to moderate in severity and occurred during the first several months on study. Only 5 of 221 patients (2.3%) who received abacavir developed the hypersensitivity reaction associated with that drug.
When these three presentations are considered together, several conclusions can be made. The backbone of antiretroviral therapy for the treatment naïve patient remains dual nucleoside therapy. The selection of which two nucleosides to use in combination is more appropriately made with respect to the potential for side effects and CD4+ count changes, rather than differences in activity. Protease-sparing antiretroviral therapy can be as potent and less toxic than combinations that include protease inhibitors. The remaining variable in this equation is resistance. We know a fair amount about what mutations are associated with the development of resistance in patients with virologic failure (more on this subject below). However, less data is available to our choice of agents following an initial treatment failure. Abacavir is effective in patients with virus a single 184V mutation. But, what is the best way to sequence drugs to achieve the most durable suppression of viral replication? Strategic trials that contain pre-selected combinations for patients failing on their initial regimen will hopefully provide these answers.