Metabolic and Virologic Responses to Different Regimens Vary by Gender and Race

  • Gender and Race Subgroup Analyses in a Prospective Study of Hyperlipidemia in ART-Naive Subjects Taking Trizivir (TZV), Combivir (COM)/Nelfinavir (NFV), or Stavudine (d4T)/Lamivudine (3TC)/NFV (Oral 57)
    Authored by K. Tashima, P. Kumar, A. Rodriguez-French, M. Thompson, P. Wannamaker, V. Williams, K. Pappa

At the 9th Conference of Retroviruses and Opportunistic Infections (CROI), 2002, Abstract 33, Dr. Kumar presented the 48-week data from a prospective study of hyperlipidemia in ART-naive subjects taking Combivir (CBV)/abacavir (ABC, Ziagen) versus Combivir/nelfinavir (NFV, Viracept) versus d4T (stavudine, Zerit)/3TC (lamivudine, Epivir)/nelfinavir.

There were two interesting observations in that study. One was that the d4T-containing arms had significant increases in LDL cholesterol, total cholesterol (TC), and triglyceride levels (TG) compared with the Combivir/nelfinavir and Combivir/abacavir arms. The Combivir/abacavir arm had minimal changes in the lipid profiles. The other major observation was that men had a more significant increase in their LDL levels in both nelfinavir-containing arms. At this IAS meeting, Dr. Karen Tashima presented the 96-week data of the same study, now with breakdowns of the metabolic parameters by both gender and race.

A gender and racially diverse population was enrolled in this study, which offers a unique opportunity to compare metabolic complications and response rates in these populations. The objectives of this study were to evaluate changes in LDL and other lipid parameters as well as the safety and efficacy in subjects treated with Trizivir (AZT [zidovudine, Retrovir] + 3TC + ABC) versus Combivir/nelfinavir versus d4T/3TC/nelfinavir (groups 1, 2 and 3, respectively). Subgroup evaluations were assessed at 96 weeks.

A total of 254 non-diabetic, ART-naive subjects from the U.S., Puerto Rico, Panama, the Dominican Republic and Guatemala with CD4 counts of more than 50 cells/mm3 were randomized 1:1:1 to receive therapy as described above. Evaluations consisted of fasting lipid profiles, HIV-1 RNA and CD4 cell counts. Results were assessed in an analysis of covariance model. Statistical significance (ss) was defined as P<0.05.

In this diverse population (50 percent female, 40 percent black, 37 percent Hispanic), the groups were comparable at baseline, with an overall mean HIV RNA and CD4 of 4.43 log10 c/mL and 355 cells/mm3, respectively. Unfortunately, the drop out rate or loss to follow-up was approximately 50 percent. Mean LDL was significantly lower in group 1 compared to groups 2 and 3 for men (93, 117, 126 mg/dl, groups 1, 2, 3; P=0.012 vs. group 2; P<0.001 vs. group 3) and women (94, 121, 134 mg/dl, groups 1, 2, 3; P<0.001 vs. groups 2 and 3). Similar responses were seen for blacks (104, 120, 142, groups 1, 2, 3; ss: 1 and 2 vs. 3) and Hispanics (89, 111, 111 mg/dl, groups 1, 2, 3; ss: 1 vs. 2 and 3). Triglycerides were significantly higher in Hispanics in group 3 (188, 211, 295 mg/dl, groups 1, 2, 3; ss: 1 and 2 vs. 3). Total cholesterol was significantly higher in group 3 for females (173, 201, 226 mg/dl, groups 1, 2, 3; ss: 1 vs. 2 and 3) and Hispanics (162, 193, 201 mg/dl, groups 1, 2, 3; ss: 1 vs. 2 and 3). Elevated lactate with d4T was seen in female and Hispanic subgroups (female ss: 1 vs. 3; Hispanic ss: 1 vs. 2 and 3). Virologic response (ITT Observed/ MEF) was similar across gender for proportions with HIV RNA less than 50 c/mL, however women tended to have less of a virologic response compared with men. The response rate for Trizivir was consistently higher in Hispanics (HIV RNA less than 400 c/mL, 65 percent, 45 percent, 44 percent; HIV RNA less than 50 c/mL, 61 percent, 39 percent, 31 percent for groups 1, 2, 3).

Dr. Tashima concluded that both metabolic and virologic responses to the different regimens appear to vary by gender and race and that further study is warranted. I spoke with Dr. Kumar afterwards and asked about differences in gender by race but this had not been analyzed. At 48 weeks, men had more lipid abnormalities on nelfinavir-containing regimens compared with women. At 96 weeks, women and blacks had significantly more increases in LDL. Men and Hispanics had more triglyceride elevations. The question is whether there is any correlation between gender and race. Would a black woman be more at risk of developing elevated LDL while on therapy compared with a Hispanic woman? Is a Hispanic man at more risk of developing hypertriglyceridemia while on therapy compared with a Hispanic woman? Given the high discontinuation rate, I do not know if the protocol team will have an adequate sample size to answer these questions.

Another point to make is how clinically applicable the Trizivir-only arm is in light of the results of ACTG 5095 (for coverage of this study, click here), which revealed that Trizivir had a significantly lower virologic response rate compared with efavirenz (EFV, Sustiva)-containing regimens. As with other studies, Trizivir has not been shown to cause significant lipid abnormalities. The question remains whether a subpopulation exists that may benefit from being on Trizivir alone. The 96-week response rates Dr. Tashima reported (HIV RNA levels less than 50 c/mL by ITT) were discouragingly low on all arms but the treatment arms also had a very large discontinuation rate. The on-treatment results were not reported. I agree with Dr. Tashima that further studies are warranted in order to evaluate whether treatment responses and complications are significantly different with respect to gender, race and ethnicity -- with that knowledge, therapies can be prescribed to achieve the maximum benefit for an individual.