Background: Hyperlipidemia is a common side effect of HAART therapy.  The effect that non-PI-containing regimens have on the development of hyperlipidemia and fat redistribution (FR) is not fully known. Our objectives were to evaluate changes in fasting metabolic parameters, clinical evaluations of FR, safety, and efficacy in subjects treated with COM/ABC vs COM/NFV vs d4T/3TC/NFV for 96 weeks. 

Methods: 258 non-diabetic, ART-naïve subjects with CD4 >50 cells/mm³ were randomized 1:1:1.  Evaluations consisted of fasting lipid profiles and other clinical evaluations of FR at baseline through 96 weeks.  Results were compared in an analysis of covariance model adjusting for baseline.

Results: A planned interim analysis at 48 weeks (50% female, 40% African American, 37% Hispanic) indicate significant increases in LDL and total cholesterol in both of the PI-containing arms compared to the COM/ABC arm.

LS Mean (SE)	COM/ABC N=86	COM/NFV N=89	D4T/3TC/NFV N=83
Baseline LDL (mg/dL)	91.5 (3.6)	95.2 (3.4)	95.3 (3.6)
LDL- ∆ from BL†‡	-2.4 (4.2) n=55	19.4 (4.1) n=55	27.1 (4.5) n=49
BL Cholesterol	154.7 (3.9)	164.3 (3.8)	160.4 (4.0)
Cholesterol- ∆ from BL†‡	9.9 (4.7) n=57	32.1 (4.6) n=57	45.7 (4.9) n=54
BL Triglycerides	128.6 (10.1)	148.4 (9.8)	142.2 (10.4)
Triglycerides- ∆ from BL‡	23.6 (13.2) n=57	31.4 (13.1) n=57	69.3 (13.8) n=54
BL Lactate	9.7 (0.7)	10.2 (0.6)	10.9 (0.7)
Lactate- ∆ from BL‡	(1.2) n=58	2.3 (1.2) n=55	7.9 (1.3) n=51
BL HIV-1 RNA  (log10) Adj. Mean (SE)	4.67 (0.06)	4.74 (0.06)	4.77 (0.06)
Subjects with <50 copies HIV-1 RNA at 48-weeks (ITT M=F) n/N (%)	41/86 (48)	42/89 (47)	32/83 (39)
BL CD4+ cell count	309 (22.9)	320 (21.9)	315 (23.2)
CD4+ ∆ from BL 48-weeks	160 (19.4) n=56	140 (19.0) n=56	159 (20.2) n=54

^† COM/ABC vs COM/NFV yielded p<0.05 ^‡ COM/ABC vs d4T/3TC/NFV yielded p<0.05 Suspected hypersensitivity reaction to abacavir occurred in 2% of subjects.

Conclusions:  These 48-week results indicate that therapy with COM/ABC has comparable virologic and immunologic efficacy with a more favorable lipid profile as compared to PI-containing regimens.  Gender differences will be presented.