Metabolic Complications in HIVInfected Patients

1
Metabolic Complicationsin HIV-Infected Patients

• William G. Powderly, MDProfessor of Medicine and
  TherapeuticsUniversity College DublinMater
  University HospitalDublin, Ireland

This program is supported by an unrestricted
educational grant from
2
About These Slides

• Users are encouraged to include these slides in
  their own presentations, but we ask that content
  and attribution not be changed. Users are asked
  to honor this intent.
• These slides may not be published or posted
  online or used for any other commercial purpose
  without written permission from Clinical Care
  Options.
• We are grateful to William G. Powderly, MD,
  University College Dublin, Ireland, who aided in
  the content creation of these slides.

DisclaimerThe materials published on the
Clinical Care Options Web site reflect the views
of the authors, not those of Clinical Care
Options, LLC, the CME providers, or the companies
providing educational grants. The materials may
discuss uses and dosages for therapeutic products
that have not been approved by the United States
Food and Drug Administration. A qualified
healthcare professional should be consulted
before using any therapeutic product discussed.
Readers should verify all information and data
before treating patients or using any therapies
described in these materials.
3
Epidemiology of Metabolic Complications and
Cardiovascular Risk Factors in HIV-Infected
Patients
4
Established Risk Factors for Cardiovascular
Disease (CVD)

• Fixed and unchangeable
• Sex
• Family history
• Prior history of CVD
• Age

• Potentially changeable
• Tobacco use
• Systolic hypertension
• Obesity
• Diabetes mellitus
• Lipids
• Elevated LDL cholesterol
• Lowered HDL cholesterol
• Increased triglycerides

5
Traditional Factors Are the Biggest Contributor
to CHD in HIV Population
HIV infection
HAART
Emerging factorsLp (a), CRP, IMT, and
endothelial function
Diabetes
Metabolic syndrome. Precise contribution
unclear.
6
Prevalence of Cardiac Risk Factors in a Cohort of
HIV-Infected Pts DAD
of Cohort With Risk Factor at Baseline
BMI, body mass index HC, hypercholesterolemia.
Friis-Moller N, et al. AIDS. 2003171179-1193.
7
DAD Risk Factors for CHD in an HIV-Infected
Population
0.1
0.5
1
5
10
Multivariable Poisson model adjusted for BMI, HIV
risk, cohort, calendar year, and race.
Lundgren J, et al. CROI 2005. Abstract 62.
8
Smoking Incidence Is Increased in HIV-Infected
Pts vs General Population
APROCO Cohort (HIV)
MONICA sample (HIV-)
70
P lt .0001
60
P lt .0001
50
P NS
40
Patients ()
30
20
P lt.01
P NS
10
0
Blood Glucose 126 mg/dL (6.99 mmol/L)
Smoking
Hypertension
HDL-C lt 40 mg/dL (1.04 mmol/L)
LDL-C gt 160 mg/dL (4.14 mmol/L)

• 223 HIV men and women on PI-based regimens vs
  527 HIV- male subjects
• HIV patients had lower HDL and higher TG
• No difference in total cholesterol
• Predicted risk of CHD gt in HIV men (RR 1.2) and
  women (RR 1.6) P lt .0001

Savès M, et al. Clin Infect Dis. 200337292-298.
9
Surrogate Markers for CVD Risk

• Established markers
• Total cholesterol
• LDL-C
• HDL-C
• Hypertrigylceridemia in conjunction with
  additional risk factors1
• Emerging markers
• Apolipoprotein B (ApoB)
• High sensitive C-reactive protein (hsCRP)
• Intima-media thickness (IMT)
• Endothelial function

1. Grundy SM, et al. Circulation.
2004110227-239.
10
Indirect/NoninvasiveMarkers of Atherosclerosis

• Carotid intimal thickness
• Some studies have reported increased
  abnormalities in HIV-infected treated
  patients1-3
• Usually associated with traditional risk factors
• Coronary calcification by electron beam computed
  tomography
• No significant differences to date compared with
  HIV-negative controls2
• Endothelial dysfunction
• HIV infection may increase endothelial
  dysfunction, reversed with treatment4
• PI-treated patients more likely to have abnormal
  brachial flow than other treated patients5

1. Maggi P, et al. AIDS. 200014F123-F128. 2.
Talwani R, et al. J Acquir Immune Defic Syndr.
200230191-195. 3. Hsue P, et al. Circulation.
20041091603-1608. 4. Torriani F, et al. EACS
2005. Abstract PS5-3. 5. Stein JH, et al.
Circulation. 2001104257-262.
11
HIV Infection May Be an Independent Risk Factor
for Early Atherosclerosis

• Case-control study evaluating CVD risk factors
  including actual plaque in 292 HIV vs 1168 HIV-
  patients
• Internal carotid plaques 12.3 in HIV and 7.8
  in HIV- (P .03 in unadjusted model)
• Common carotid IMT 5 higher in HIV (P .0002)
  when adjusted for multiple risk factors
• Higher vascular events due to HIV means 4 to
  14 increase in risk of 1 additional vascular
  event in 5 years

Stephan C, et al. CROI 2006. Abstract 738.
12
Endpoints Carotid IMT

• Study evaluated progression of IMT in
• HIV, continuous PI use ? 2 years (n 44)
• HIV, PI naive (n 45)
• HIV- controls (n 45)
• Neither HIV infection nor PI exposure
  significantly affected the rate of progression of
  carotid IMT over 3 years of follow-up

• No significant difference between
• PI naive and PI treated (P .19)
• PI-naive and HIV antiretroviral therapynaive
  group (P .78)
• Combined HIV groups and HIV- controls (P .71)

Currier J, et al. CROI 2006. Abstract 145.
13
Endothelial Function in HIV Infection

• Study evaluated effect of HIV infection and
  antiretroviral therapy on endothelial vasomotor
  function in 75 HIV-infected and 223 control
  patients
• HIV-infected patients had significantly decreased
  artery flow-mediated dilation (FMD) vs
  uninfected controls (3.6 P lt .01) in
  multivariate analysis
• Among HIV-infected patients, HIV load, IDU,
  hazardous drinking, and triglyceride levels, but
  not PI use, were associated with significantly
  decreased FMD

Solages A, et al. Clin Infect Dis.
2006421325-1332.
14
ART and Endothelial Function

• Study evaluated effect of ART on
  endothelium-dependent vasodilation in 6
  HIV-seronegative males and females
• Participants were relatively young (mean age 25
  years) nonsmokers
• Forearm blood flow (FBF) determined before and 1
  month after treatment with LPV/RTV
• In response to 30 µg/min acetylcholine, the AUC
  of absolute FBF was significantly increased with
  LPV/RTV treatment (P .03) for each dose,
  absolute or percentage increases over baseline
  were nonsignificant
• Response to nitroprusside was comparably
  increased with PI treatment but did not reach
  statistical significance

Grubb JR, et al. J Infect Dis. 20061931516-1519.
15
A5152s Impact of HIV Infection and Treatment on
Endothelial Function

• First prospective, randomized, evaluation of the
  effects of ART on endothelial function in
  patients with HIV infection
• FMD improved during HIV treatment
• Difference between blinded treatment groups not
  significant
• HIV infection itself affected endothelial
  function
• Baseline FMD 3.6
• No consistent correlations between changes in FMD
  and changes in any lipids

Week 24
Week 4
3.5
3.0
2.5
2.0
Median Change in FMD From Baseline ()
1.5
1.0
0.5
0
Arm X
Arm Y
Arm Z
Overall
Torriani F, et al. EACS 2005. Abstract PS5/3.
16
Is There an Increase in Cardiovascular Events in
HIV-Infected Patients?

• There is evidence of an increased rate of MIs or
  cardiovascular events in HIV-infected patients
  compared with HIV-negative patients
• Some studies have linked this to PI therapy
• Others have shown link to HIV disease, not to
  treatment
• Some studies have not shown increased incidence
  as clearly
• Overall rate still low and may be consistent with
  background risk
• In all studies, most incidents of cardiovascular
  disease occurred in persons with known risk
  factors

17
Overall Incidence of CV Events Is Low in an
HIV-Infected Population HOPS

• gt 8000 patients followed since 1993 in HIV
  Outpatient Study (HOPS)
• 1807 selected patients from 3/1/96 to 9/30/05
  with available baseline and 1 cholesterol,
  triglycerides, and glucose value recorded (note
  possible selection bias)
• Results 84 CV events in 57 patients
• No association found for specific ARV class,
  pre-HAART CD4 cell count, time on HAART, BMI gt
  30, peak viral load, peak TC, peak LDL, or peak
  TG
• Risk of CV events reduced with lipid-lowering
  agents

Vertical bars 95 CI. Continuous variable.
Lichtenstein K, et al. CROI 2006. Abstract 735.
18
Veterans Administration No Increase in CV Events
With Use of HAART

• CVD admissions
• 8.5-year retrospective study (n 36,766 gt 1.4
  million patient-months)
• Mean 40 months on HAART
• Median exposure PI 16 months NNRTI 9 months
  NRTI 17 months
• CVD and mortality
• 1764 CVD admissions
• 2006 admissions for or deaths from CVD
• 16,731 all-cause deaths
• Overall mortality decreased following HAART
  initiation
• No increase in CVD events with use of HAART
  observed

2.5
NRTIs PIs
NRTIs NNRTIs
2.0
1.5
Admissions per 100 Patient-Years
1.0
0.5
0
No ART
0
lt 2
2-4
gt 4
ART (Years of Use)
Bozzette SA, et al. N Engl J Med.
2003348702-710.
19
DAD Incidence of MI With Exposure to
Combination ART
Duration of Combination Antiretroviral Therapy Is
Associated With a Small Increase in Incident CVD
RR per Year of ART Overall 1.17 Men 1.14 Women
1.38
10
8
6
Incidence of MI per 1000 Patient-Year
4
2
0
Exposure to ART (Years)
El-Sadr W, et al. CROI 2005. Abstract 42.
20
Contribution of Dyslipidemia to MI Risk
Adjusted for conventional risk factors (age,
sex, prior/family history of CVD, smoking) not
influenced by CART.
El-Sadr W, et al. CROI 2005. Abstract 745.
21
HAART and the Risk of MI Updated Data From DAD
RR of MI by ART Exposure

• Exposure to elements of HAART
• NNRTI 6.3 years (range 3.8-8.3)
• PI 3.0 years (range 0.5-5.4)
• NNRTI 0.9 years (range 0-3.2)
• Peak RR of MI in 2001 falling since
• Adjustment for lipids largely explains decline in
  MI
• PI exposure associated with similar increased
  risk as HAART exposure
• NNRTI exposure not associated with increased risk
  of MI
• Adjustment for NRTI exposure did not change risk
• Suggests that increased risk previously reported
  with HAART largely driven by PIs

1.8
1.6
1.4
1.2
1.0
RR of MI
0.8
0.6
0.4
0.2
0
PI
NNRTI
ART
Adjusted for NNRTI exposure. Adjusted for PI
exposure.
Friis-Moller N, et al. CROI 2006. Abstract 144.
22
French Hospital Database MI Risk Increased With
PI Use

• Study evaluated MI diagnosis among men from
  1996-1999
• Follow-up 88,029 patient-years (n 19,795
  exposed to PIs)
• 60 MIs observed, 49 cases among men taking PIs
• Risk of MI associated with use of PIs
• Relative hazard for MI with PI use 2.56 (95 CI
  1.03-6.34)

Mary-Krause M, et al. AIDS. 2003172479-2486.
23
DAD Study MI Incidence Stable or Decreasing
Over Time

• MI incidence 3.7/1000 patient-years (345 events
  in 94,469 patient-years)
• Increased risk of MI was associated with
  increased PI exposure (RR 1.17/year of exposure
  95 CI 1.12-1.23)
• No evidence of increased risk of MI with
  increased NNRTI exposure, although fewer years of
  experience (RR 1.07/year of exposure 95 CI
  1.00-1.14)

Friis-Moller, et al. CROI 2006. Abstract 144.
24
Is the Framingham Risk Estimation Valid in
HIV-Infected Patients?
Observed and Predicted MI Rates According to ART
Exposure (DAD Study)
8
7
Observed rates
6
5
Best estimate of predicted rates
4
Rates per 1000 Person-Years
3
2
1
0
lt 1
1-2
2-3
3-4
gt 4
None
Duration of cART Exposure (Years)
Law MG, et al. HIV Med. 20067218-230.
25
SMART HIV Progression With Continuous HAART vs
Interruption

• CD4-guided drug conservation strategy was
  associated with significantly greater disease
  progression or death, compared with continuous
  viral suppression RR 2.5 (95 CI 1.8-3.6 P lt
  .001)

No. of Patients With Events
Parameter
RR (95 CI)
1.5
Severe complications
114
1.4
CVD, liver, or renal deaths
31
1.5
Risk of Complications
Nonfatal CVD events
63
1.4
Nonfatal hepatic events
14
2.5
Nonfatal renal events
7
1.0
10.0
0.1
Favors TI
Favors CT
El-Sadr W, et al. CROI 2006. Abstract 106 LB.
26
10-Year Risk of MI and HIV Progression
Estimated Risk of AIDS or Death at 3 Years and of
MI at 10 Years Among Men Initiating ART
3-year risk of progression to AIDS or death is
calculated according to the method of the ART
Cohort Collaboration. In absence of ART, the
risk of AIDS or death at 3 years is 40.
According to the Framingham equation for male
patients. Based on 25 decline in total
cholesterol.
Grinspoon S and Carr A. N Engl J Med.
200535248-62.
27
Conclusions

• Traditional risk factors (age, smoking, and
  hypertension) increase risk of CVD in individuals
  with or without HIV infection
• HIV disease may confer its own increase in CVD
  risk
• There is evidence of an association between ART
  use and a small increase in the absolute risk of
  CVD although study results are presently
  inconsistent
• CVD risk should not influence the decision to
  initiate ART in most cases
• Risk-benefit considerations should influence
  decision making when selecting an initial ART
  regimen
• CVD risk should be considered in the overall care
  of patients with HIV infection

28
Metabolic Complications Associated With
Antiretroviral Therapy
29
Lipids and HIV Infection

• HIV infection causes dyslipidemia
• Decreased cholesterol
• Decrease in total cholesterol (TC), low density
  lipoprotein cholesterol (LDL-C), and high density
  lipoprotein cholesterol (HDL-C)
• Increased triglycerides (TG)
• Treatment of HIV increases TC and LDL-C
• Does not increase HDL-C as predictably
• Specific antiretroviral drugs (and drug classes)
  have differing effects on lipids

30
HIV Disease and Lipid Metabolism

• Lipids measured in 32 AIDS patients, 8
  asymptomatic HIV infection, 17 uninfected
  controls
• Mean TG and frequency of hypertriglyceridemia
  were significantly increased in patients with
  AIDS vs controls
• Mean TG of asymptomatic patients was intermediate
• No differences in cholesterol levels among the 3
  groups

Grunfeld C, et al. Am J Med. 19898627-31.
31
MACS 10-Year Prospective Assessment of Lipid
Levels
Lipid Changes

• Multicenter AIDS Cohort Study
• 50 HIV-negative men who seroconverted
• Prior to seroconversion
• Lipid levels were within normal range and similar
  to NHANES III
• After seroconversion
• Prior to HAART (n 50)
• Notable reductions in lipid levels
• 3 years after HAART (n 38)
• For most, lipid levels returned to baseline
  values (but not HDL), with expected increase due
  to aging

30
Total cholesterol
20
10
LDL-C
0
Change From Baseline (mg/dL)
-10
HDL-C
-20
-30
-40
Baseline HIV-
HIV ART Naive 7 Years
HIV HAART3 Years
Riddler S, et al. JAMA. 20032892978-2982.
32
Lipid Effects of Available Antiretroviral Agents
Important Considerations

• Limited number of prospective studies in
  treatment-naive patients with fasting lipids,
  insulin, and glucose levels
• Need to distinguish between effects of treating
  HIV and adverse effects of therapy
• Effects may be different in patients with more
  advanced disease
• Because patients receive combination therapy,
  interactions between various components of
  regimen need to be better defined
• In many earlier studies, lipid effects of NRTIs,
  especially d4T, were not recognized and all
  effects were attributed to PIs
• Single/multiple dose studies in healthy
  volunteers have shown a significant impact of ART
  on lipid parameters as well as glucose

33
DAD Prevalence of CVD Risk Factors in a Cohort
of HIV-Infected Patients
35
30
25
20
of Cohort With Risk Factor
15
10
5
0
Hypercholesterolemia
Increased Triglycerides
Fasting lipids available from 36 of patients,
24 non-fasting, and 40 missing
information. Defined as TC 6.2 mmol/L (240
mg/dL). Defined as TG 2.3 mmol/L (200 mg/dL).
Friis-Moller N, et al. AIDS. 2003171179-1193.
34
Association Between Dyslipidemia and ART
Experience DAD Baseline Data

• Stable levels after 3-6 months of antiretroviral
  therapy
• ? TC also associated with higher CD4 cell count,
  lower viral load, lipodystrophy, longer exposure
  to NNRTI and PI, and older age

Friis-Moller N, et al. AIDS. 2003171179-1193.
35
Lipid Effects of First-Line Antiretroviral
Regimens

• Swiss HIV Cohort Study (N 1065 patients
  starting therapy between April 2000 and January
  2005 follow-up 17-18 months)
• ? cholesterol with either PIs or NNRTIs
• ? TG with PIs, particularly with RTV regimens
• Patients primarily on LPV/RTV, IDV/RTV, or NFV
• ? HDL-C predominantly with NNRTIs

Young J, et al. IAS 2005. Abstract TuPe2.2B16.
36
Metabolic Effects of Non-PIs
NVP has more of an effect on HDL than EFV and
less effect on TG.
37
Lipid Effects of NRTIs Increased Dyslipidemia
With d4T or ZDV vs TDF
GS 934 96-Week Extension1 ZDV/3TC EFV vs TDF
FTC EFV
GS 903 Week 144 Data2 d4T 3TC EFV vs TDF
3TC EFV
ZDV/3TC
d4T
TDF FTC
TDF
P lt .001
P lt .001
160
40
P .115
140
35
P .067
120
30
100
25
P lt .001
Mean Change From Baseline (mg/dL)
P .022
80
20
P lt .001
60
15
P .003
40
10
20
5
0
0
TC
LDL-C
HDL-C
TG
TC
LDL-C
HDL-C
TG
1. Pozniak AL, et al. J Acquir Immune Defic
Syndr. 2006 Oct 12 Epub ahead of print.2.
Gallant JE, et al. JAMA. 2004292191-201.
38
Varying Lipid Effects of NRTIsGreater TC
Increase With ABC vs ZDV
CNA 30024 Study ABC 3TC EFV vs ZDV 3TC
EFV
Cholesterol
TG
200
200
P lt .05
P lt .05
Lipids (mg/dL)
150
150
ABC
ZDV
ABC
ZDV
100
100
BL
2
4
8
12
16
24
36
48
BL
2
4
8
12
16
24
36
48
Study Week

• Samples drawn nonfasted
• All medians below the NCEP cutoff

DeJesus E, et al. ICAAC 2003. Abstract H-446.
39
Lipid Effects of NNRTIsEfavirenz vs Nevirapine
Lipid Changes at Week 48

• 2NN Study
• Randomized 4-arm study in treatment-naive
  patients
• EFV vs NVP QD vs NVP BD vs dual NNRTI
• All patients received 3TC d4T
• Treatment allocation
• EFV (n 289)
• NVP (n 417)

60
EFV

50
NVP
40

30
Mean Change From Baseline ()
20
10

0
-10
TC
LDL-C
HDL-C
TG
TCHDL-C Ratio
P lt .05 vs nevirapine. P lt .001 vs nevirapine.
van Leth F, et al. PLoS Med. 20041e19.
40
Metabolic Effects of PIs

• RTV associated with more pronounced effect on
  lipids than other PIs

41
Protease Inhibitors and Dyslipidemia

• PIs have demonstrated association with
  dyslipidemia
• 20 to 40 of patients on PIs have significant
  increases in TC and TG
• Median increase in TC is 30 mg/dL (0.78 mmol/L)
• Specific PIs differ in their effects on plasma TG
  and cholesterol levels
• RTV has the most robust effect, especially on TG
• Certain PIs (ie, ATV) have less effect on lipids
• Studies of earlier PIs generally found no
  beneficial effect of PIs on HDL-C levels
• More recent studies suggest increased HDL with
  use of newer PIs

Reviewed in Dube MP, et al. Clin Infect Dis.
200337613-627.
42
High-Dose RTV and Lipids Effects in Healthy
Subjects
TG
Cholesterol
4
7

3
6

mmol/L
2
5
1
4
0
3
Day 14
Baseline
Baseline
Day 14
RTV (n 11)
RTV dose 300 mg, 400 mg, or 500 mg every 12
hours on Days 1, 2, and 3-14, respectively. Day
14 means are adjusted means from the analysis of
variance. P lt .001 vs baseline P lt .01 vs 14
days of placebo P lt .01 vs baseline.
?
Purnell JQ, et al. AIDS. 20001451-57.
43
RTV Affects LipidsEven at Low Doses

• RTV 100 mg BID or LPV/RTV 400/100 mg BID x 14
  days in healthy volunteers

P .01 vs baseline.
Shafran SD, et al. HIV Med. 20056421-425.
44
Metabolic Effects of PIs LPV/RTV vs NFV

• Phase III, randomized trial of treatment-naive
  patients

100
LPV/RTV
80
NFV
60
Patients ()
40
20
0
Grade 0-1 lt 6.22 lt 240
Grade 2 6.22-7.77 240-300
Grade 3 7.77-10.36 300-400
Grade 4 gt 10.36 gt 400
mmol/L mg/dL
TC at Week 48
Walmsley S, et al. N Engl J Med.
20023462039-2046.
45
Metabolic Effects of PIs LPV/RTV vs NFV

• Phase III, randomized trial of treatment-naive
  patients

LPV/RTV
NFV
TG
TC
97
100
100
1.13
2.59
90
90
1.02
2.33
80
80
0.90
2.07
70
70
0.79
1.81
60
59
Change From Baseline to Week 24 (mg/dL)
Change From Baseline to Week 24 (mg/dL)
Change From Baseline to Week 24 (mmol/L)
Change From Baseline to Week 24 (mmol/L)
53
60
60
0.68
1.55
50
50
0.56
1.29
40
40
0.45
1.03
30
30
0.34
0.78
20
20
0.23
0.52
10
10
0.11
0.26
0
0
0
0
Walmsley S, et al. N Engl J Med.
20023462039-2046.
46
Metabolic Effects of PIsLPV/RTV vs ATV/RTV

• BMS-045 randomized trial of patients with 2
  HAART failures

ATV/RTV
LPV/RTV
TC
HDL-C
LDL-C
TG
35
30
25
15
Mean Change From Baseline to Wk 48 ()
6
2
5
1
-5
-4
-8
-7
-10
-15
Fasting values. P lt .0001 vs LPV/RTV.
Johnson M, et al. AIDS. 200519-685-694.
47
Metabolic Effects of PIs ATV vs EFV

• BMS-034 randomized trial of treatment-naive
  patients

Baseline
Week 48
TC
LDL-C
HDL-C
Mean ?
2
21
1
18
13
24
240
6
200 mg/dL
200
5
160
4
130 mg/dL
Mean (mg/dL)
Mean (mmol/L)
120
3
80
2
40 mg/dL
40
1
0
0
ATV
EFV
ATV
EFV
ATV
EFV
P lt .0001, ATV vs EFV at Week 48 for each lipid
parameter.
Squires K, et al. J Acquir Immune Defic Syndr.
2004361011-1019.
48
Effect of RTV Boosting on ATV BMS-089 Study

• BMS-089 randomized trial of 100 treatment-naive
  patients
• ATV/RTV 300/100 mg vs ATV 400 mg

ATV/RTV
ATV
TC
LDL-C
HDL-C
TG
35
30
29
30
26
23
25
20
16
15
Mean Change From Baseline to Week 48 ()
15
10
6
5
0
-5
-3
P .0034
P .21
P .69
P .0004
Malan N, et al. CROI 2006. Abstract 107LB.
49
KLEAN FPV/RTV vs LPV/RTVNo Differences in
Lipids Between Arms

• KLEAN randomized trial of 878 treatment-naive
  patients
• FPV/RTV 700/100 mg BID vs LPV/RTV 400/100 mg BID

FPV/RTV at Wk 48
FPV/RTV at BL
LPV/RTV at BL
LPV/RTV at Wk 48
250
200
150
Fasting Lipids at Week 48 (mg/dL)
100
50
0
TC
LDL-C
HDL-C
TG
Eron R, et al. Lancet. 2006368476-482.
50
ALERT FPV/RTV vs ATV/RTVNo Differences in
Lipids Between Arms

• ALERT randomized trial of 106 treatment-naive
  patients
• FPV/RTV 1400/100 mg QD vs ATV/RTV 300/100 mg QD
• Comparable virologic efficacy observed at 24
  weeks
• Lipid effects also comparable between arms
• Study evaluated investigational FPV/RTV dose

Baseline
Week 24
HDL-C
LDL-C
125
99
103
95
97
100
75
45
41
38
38
50
25
0
Median Lipid Levels (mg/dL)
FPV/RTV
ATV/RTV
FPV/RTV
ATV/RTV
TC
TG
250
180
177
200
160
160
153
133
127
150
116
100
50
0
FPV/RTV
ATV/RTV
FPV/RTV
ATV/RTV
Smith K, et al. ICAAC 2006. Abstract H-1670a.
51
GEMINI SQV/RTV vs LPV/RTV in Treatment-Naive Pts
at Wk 24

• Prospective, phase IIIb, randomized, multicenter,
  open-label study (N 337)
• Significantly greater risk of hyperlipidemia with
  LPV/RTV vs SQV/RTV

P .020
P .182
88
250
P .779
29
21
200
29
P .602
150
Mean Lipid Values (mg/dL)
12
10
100
9
9
50
0
TC
LDL
HDL
TG
Slim J, et al. Intl Congress on Drug Therapy in
HIV Infection 2006. Abstract PL 2.5.
52
Genetic Polymorphisms and Metabolic Complications

• Individual susceptibility is the missing factor
  in ARV efficacy and tolerability
• Polymorphisms also affected pretreatment lipid
  levels in these studies

1. Ranade K, et al. CROI 2006. Abstract 763. 2.
Arnedo M, et al. CROI 2006. Abstract 764. 3.
DeLuca A, et al. CROI 2006. Abstract 766. 4.
Cossarizza A, et al. CROI 2006. Abstract 767. 5.
Gometz E, et al. CROI 2006. Abstract 768. 6. Tarr
PE, et al. J Infect Dis. 20051911397-1400.
53
Conclusion

• HIV infection itself has profound effects on
  lipids
• Treatment of HIV infection reverses (at least
  partially) HIV effect but introduces the
  complexity of different drug effects on lipids
• Overall impact of ART on lipids in an individual
  patient also reflects host considerations
  including lifestyle and diet, genetic
  predisposition and comorbidities (eg, insulin
  resistance and diabetes)

54
Assessment of Lipids andUse of Lipid-Lowering
Therapy
55
Assessment and Intervention forCVD Risk in
HIV-Infected Patients

• Assess cardiovascular risk
• Address lifestyle issues
• Diet
• Exercise
• Smoking
• Treat hypertension
• Treat dyslipidemia
• Treat diabetes mellitus

56
Assessment of Cardiac Risk

• Cardiac risk factors
• Increased age
• Sex (males are at higher risk)
• Smoking
• Elevated total cholesterol
• Low HDL-C
• Elevated systolic blood pressure (gt 120 mm Hg)
• Presence of diabetes
• Presence of atherosclerosis, CVD
• Family history of CVD

• How to define cardiac risk and need for
  intervention
• Combination of risk factors results in cumulative
  risk
• Risk assessment tools can be used to calculate
  CHD risk
• Smoking presents significant risk even as single
  factor
• Framingham risk calculator1 http//hp2010.nhlbi
  hin.net/atpiii/calculator.asp
• SCORE calculator2
• PROCAM calculator3

1. Wilson P, et al. Circulation.
1998971837-1847. 2. Conroy RM, et al. Eur
Heart J. 200324987-1003. 3. http//www.CHD-taskf
orce.com
57
CV Risk Assessment Algorithms Predicted 10-Year
Risk of MI
PROCAM study recruited men only.Calculators
require SBP only.
1. Wilson P, et al. Circulation.
1998971837-1847. 2. Assmann G, et al.
Circulation. 2002105310-315. 3. Conroy RM, et
al. Eur Heart J. 200324987-1003.
58
The Framingham Risk Estimation in HIV-Infected
Patients
Observed and Predicted MI Rates According to cART
Exposure (DAD Study)
8
7
6
5
MI Rate (per 1000 Person-Yrs)
4
3
2
1
0
lt 1
1-2
2-3
3-4
gt 4
None
Duration of cART Exposure (Yrs)
Law MG, et al. HIV Med. 20067218-230.
59
Estimating CV Risk Summary

• It is possible to estimate the risk for CV
  disease using mathematical algorithms based on
  outcomes of large cohorts with long-term
  follow-up
• The total estimated risk is influenced by the
  calculated non-HDL-C, age, sex, blood pressure,
  smoking status, and other factors

60
Management of Cardiovascular Risk Nonlipid Issues

• Stop smoking
• Treat hypertension
• Encourage appropriate diet and exercise

61
Smoking Is Common Among HIV-Infected Patients
DAD Baseline Data
60
50
40
of Cohort With Risk Factor
30
20
10
0
FamilyHx of CHD
PreviousHx of CHD
CurrentSmoking
BMIgt 30 mg/m2
HTN
DM
Hyper-cholesterolemia
Increased TG
Friis-Moller N, et al. AIDS. 2003171179-1193.
62
Managing Lipid Disorders CVD Risk Guidelines
for Patients on HAART
Serum TGs gt 500 mg/dL (5.65 mmol/L) FIBRATE
Dubé MP, et al. Clin Infect Dis. 200337613-627.
63
Key Lipid Issues That Increase Risk of
Cardiovascular Disease

• Established markers of increased cardiac risk
• Increased total cholesterol
• Increased LDL-C
• Decreased HDL-C
• Hypertrigylceridemia in conjunction with
  additional risk factors1

1. Grundy SM, et al. Circulation.
2004110227-239.
64
NCEP III Update LDL-C Goals and Intervention
Thresholds
Age may be an important consideration in
patients with fewer risk factors. TLC,
therapeutic lifestyle changes.
Grundy SM, et al. Circulation. 2004110227-239.
65
Increasing Plasma LDL-C Increases Relative Risk
of CHD

• Log-linear relationship demonstrates a 30 mg/dL
  (0.77mmol/L) ? in LDL equates to 30 ? relative
  risk (RR) of CV disease

Grundy SM, et al. Circulation. 2004110227-239.
66
Higher HDL-C Reduces CVD Risk Regardless of
LDL-C Framingham

• 1 mg/dL (0.02 mmol/L) ? in HDL reduces CV risk by
  2 in men and 3 in women2
• Low HDL-C cut offs lt 40 mg/dL (1.04 mmol/L) for
  men lt 50 mg/dL (1.30 mmol/L) for women2

1. Gordon T, et al. Am J Med. 197762707-714. 2.
Gordon DJ, et al. Circulation. 1989798-15.
67
Remember Other Causes of Hypertriglyceridemia

• Causes of elevated triglycerides (? 150 mg/dL
  1.70 mmol/L)
• Obesity/overweight
• Physical inactivity
• Cigarette smoking
• Excess alcohol intake
• High carbohydrate diets (gt 60 of energy intake)
• Several diseases (type 2 diabetes, chronic renal
  failure, nephrotic syndrome)
• Drugs (corticosteroids, estrogens, retinoids,
  higher doses of beta-blockers,
  testosterone/anabolic steroids)
• Genetic dyslipidemias
• Lipoatrophy

68
Lipid-Lowering Therapy Overview
Low patient acceptance.
69
Lipid-Lowering Agents and PIsDrug-Drug
Interactions
AUC ??? with DRV.
1. Fitchenbaum CJ, et al. AIDS. 200216569-577.
2. Hsue PH, et al. Antimicrob Agents Chemother.
2001453445-3450. 3. Gerber J, et al. IAS 2003.
Abstract 870. 4. Carr RA, et al. ICAAC 2000.
Abstract 1644. 5. Telzir package insert 2003.
6. Gerber JG, et al. CROI 2004. Abstract 603. 7.
Reyataz package insert 2005. 8. Aptivus
package insert 2005. 9. Prezista package
insert 2006.
70
Pharmacokinetic Interactions Between
Lopinavir/Ritonavir and Rosuvastatin

• Rosuvastatin (ROS) not metabolized by CYP 450
• HIV patients on LPV/RTV SGC (400/100 mg BID, n
  22) with TC gt 6.2 mmol/L (239 mg/dL) treated
  with ROS for 12 weeks
• ROS increase 1.5-1.9-fold, may be used with
  caution

Data on file AstraZeneca.
van der Lee M, et al. CROI 2006. Abstract 588.
71
Studies of Treatment Interventions for
Dyslipidemia
1. Fisac C, et al. IAC 2002. Abstract ThPe7354.
2. Gatell J, et al. IAC 2006. Abstract THPE0123.
3. Hollowell S, et al. ICAAC 2005. Abstract
H-338. 4. Calza L, et al. AIDS. 2005191051-1058.
72
Fenofibrate (F) vs Pravastatin (P) in HIV
Subjects ACTG 5087

• HIV and on ARV gt 6 mos (n 174)
• LDL gt 130 mg/dL (3.37 mmol/L) and TG gt 200 mg/dL
  (2.26 mmol/L)
• Randomized to
• Fenofibrate 200 mg QD
• Pravastatin 40 mg QD
• Addition of other drug at Week 12 if lipid goal
  not achieved
• Goal NCEP composite LDL, HDL, TG targets
• 4 (7) on FP achieved composite NCEP goal
• 2 (3) on PF achieved composite NCEP goal
• Sequencing F then P resulted in greater absolute
  and TG ?

Aberg J, et al. AIDS Res Human Retrovirus.
200521757-767.
73
Ezetimibe for Lipid Management

• Ezetimibe, selective inhibitor of cholesterol
  absorption
• First prospective study with ezetimibe in HIV (n
  19)
• Open-label, 24-week study of patients with LDL ?
  130 mg/dL (3.37 mmol/L), despite pravastatin
• Ezetimibe 10 mg/day added to pravastatin 20
  mg/day and current ARVs
• Significant ? in TC and LDL-C and significant ?
  in HDL-C at Week 24
• No adverse events noted
• PK substudy no changes in LPV (n 8) and NVP (n
  7) levels after Tx

Negredo E, et al. AIDS. 2006202159-2164.
74
Treatment of Hypertriglyceridemia With Fish Oil

• Randomized, controlled trial of high-dose fish
  oil vs paraffin oil
• N 122 patients with hypertriglyceridemia gt 2
  g/L on HAART
• Mean baseline TG 4.5 g/L
• Significant efficacy at Week 8
• 10 pts with TG gt 10 g/L given open-label fish
  oil 44 ? in TG after 8 weeks
• No change in LDL

ITT, intent-to-treat analysis.
De Truchis P, et al. CROI 2005. Abstract 39.
75
Greater Improvement in Lipid Levels With LLAs vs
Switching PI to NNRTI

• Open-label study n 130 60 male mean age
  39 years
• Stable on first HAART regimen with mixed
  hyperlipidemia randomized to
• A PI ? NVP (n 29)
• B PI ? EFV (n 34)
• C Add pravastatin (n 36)
• D Add bezafibrate (n 31)
• Pravastatin or bezafibrate significantly more
  effective in management of hyperlipidemia than
  switching PI to an NNRTI

Calza L, et al. AIDS. 2005191051-1058.
76
Metabolic Effects of Switch From d4T or ZDV to
ABC or TDF RAVE Study
P .003
P .04
0.3
0.21
P .34
P .12
0.2
0.09
0.07
0.1
0.01
0
Mean Change From Baseline at Week 48 (mmol/L)
-0.1
-0.11
-0.2
-0.3
-0.25
-0.33
-0.4
-0.5
-0.45
Total cholesterol
LDL-C
HDL-C
Triglycerides
Lipid Parameter
Moyle GJ, et al. AIDS. 2006202043-2050.
77
Metabolic Effects of Switch From PI to ABC, NVP,
or EFV NEFA Study

• NEFA substudy n 29 switched to ABC, n 32
  switched to EFV, n 29 switched to NVP

ABC
EFV
NVP

0.4

0.2
0
Median Change From BL to Month 24 (mmol/L)
-0.2
-0.4

P lt .001 vs BL P lt .01 vs BL P lt .05 vs BL

-0.6

-0.8
-1
TC
Non-HDL-C
HDL-C
TG
Fisac C, et al. AIDS. 200519917-925.
78
Metabolic Effects of Switching NFV to ATV
240
6.00
200
5.00
160
4.00
Lipid Levels (mg/dL)
Lipid Levels (mmol/L)
120
3.00
80
2.00
1.00
40
0
0
Total Cholesterol
LDL-C
HDL-C
TGs
Wood R, et al. J Acquir Immune Defic Syndr.
200436684-692.
79
SWAN Lipid Effects of Switch From LPV/RTV to
ATV-Based Therapy
Fasting LDL
TC
HDL-C
Non-HDL
Fasting TG
20
9
10
2
0
-3
-4
-4
-4
-6
-10
Mean Percent Changes From Baseline at Week 48
-14
-20
P NS
P NS
-17
P lt .0001
-30
P lt .0001
-40
-44
-50
P lt .00001
117
125
207
219
48
49
207
247
157
171
Mean mg/dL, B/L
112
122
181
210
48
54
105
261
131
163
Mean mg/dL, Wk
Unboosted ATV, except ATV/RTV used in patients
also receiving TDF.
Gatell J, et al. IAC 2006. Abstract THPE0123.
80
Lipid Management May Be Suboptimal in
HIV-Infected Pts vs HIV-Negative Pts

• Retrospective, age-matched study of HIV and
  HIV- veterans with dyslipidemia
• BL TC or TGs gt 200 mg/dL
• On lipid-lowering therapy gt 2 months
• HIV subjects on HAART
• HIV subjects less frequently met NCEP goals
  after 6 months of treatment for dyslipidemia
• HIV subjects may not be receiving optimal care
  for dyslipidemia
• 85 of HIV- received simvastatin vs 23 of HIV
• Difference in usage likely due to interactions
  b/w simvastatin and PIs

100
P .033
80
P .014
64
60
58
60
Patients Achieving NCEP Goal at 6 Months ()
43
40
28
25
20
11
11
0
TG
LDL
HDL
TC
Hollowell S, et al. ICAAC 2005. Abstract H-338.
81
Cardiovascular Risk Management in HIV-Infected
Patients Summary

• In HIV-positive patients, as in the general
  population, lifestyle modification should be the
  first approach smoking cessation, diet
  modification, increase exercise
• Use of lipid-lowering therapy or ART switching
  should be individualized
• Switching ART in patients with undetectable viral
  load may result in improvements in lipid
  parameters
• Impact of smoking cessation is greater than the
  impact of any other intervention

82
Clinical Management of Insulin Resistance and
Lipodystrophy
83
2 Overlapping Epidemics
Obesity
HIV infection

• Interplay of genetic, environmental, and
  therapeutic factors

84
Diabetes and Insulin Resistance

• Type 2 diabetes secondary to insulin resistance
• Usually asymptomatic
• Prevalence
• Type 2 diabetes
• By random glucose, 1 to 2
• By OGTT, 6 to 10
• Impaired glucose tolerance, extra 15 to 30
• Classical risk factors
• Abdominal obesity, physical inactivity, ? age,
  genetic factors, dyslipidemia

85
Insulin Resistance and HIV

• Classical type 2 diabetes
• risk factors
• Obesity (abdominal)
• Physical inactivity
• Genetic
• Family history
• Race
• Older age
• Dyslipidemia

• HIV-associated risk
• factors
• Peripheral lipoatrophy
• Reduced adiponectin
• Increased liver or muscle fat
• Inflammatory cytokines
• Low testosterone
• Oxidant stress
• HCV infection
• PIs

Insulin resistance
86
HIV-Infected Men Appear to Beat Increased Risk
for Diabetes

• Data from Multicenter AIDS Cohort Study (MACS)
  HIV-infected men on antiretroviral therapy
  develop prediabetes and diabetes at a higher rate
  than HIV-negative men
• Of 627 men evaluated, 71 demonstrated prediabetes
  and 28 had diabetes
• Risk in HIV-positive men on therapy vs
  HIV-negative men
• 1.4-fold higher for prediabetes
• 1.8-fold higher for diabetes

Brown TT, et al. Lipodystrophy Workshop 2003.
Abstract 43.
87
Potential Causes of Antiretroviral-Induced
Insulin Resistance

• Direct effects of ART
• Impaired cellular glucose uptake by PIs
• In vitro model demonstrated that various PIs
  selectively inhibit glucose transporter, GLUT41
• In vivo, IDV2 and LPV/RTV3 decrease glucose
  sensitivity
• Appears not to be true for unboosted ATV3 and
  APV4

• Indirect effects of ART
• Body fat changes affecting adipocyte biology
• Central obesity
• Fat deposition in muscle and liver
• Peripheral lipoatrophy

1. Murata H, et al. J Biol Chem.
200027520251-20254. 2. Palacios R, et al.
Antivir Ther. 200611529-535. 3. Noor MA, et al.
AIDS. 2004182137-2144. 4. Grunfeld C, et al.
Lipodystrophy Workshop 2005. Abstract 15.
88
Indinavir Is Associated With Insulin Resistance
in Normal Subjects

• Indinavir 800 mg BID given to 10 HIV-negative men

12
P .009
10
8
Mean Insulin-Mediated Glucose Disposal
Rate (mg/kg per min per uU/mL)
6
4
2
0
Baseline
Week 4
Noor MA, et al. AIDS. 20011511-18.
89
Impact on Insulin Sensitivity Boosted and
Unboosted ATV and LPV/RTV
Boosted ATV Study1
Unboosted ATV Study2
ATV 400 mg
LPV/RTV 400/100 mg
ATV/RTV 300/100 mg
LPV/RTV 400/100 mg
0
0
n 20
n 23
n 24
n 20
lt -1
-5
-5
P NS
- 9
-10
-10
Change in Insulin-Stimulated Glucose Disposal
Rate at Day 10 ()
P .132
Change in Insulin-Stimulated Glucose Disposal
Relative to Placebo ()
-15
-15
-20
-20
-25
-25
- 24
- 25
P lt . 001
P . 008
-30
-30
P .023
1. Noor MA, et al. Lipodystrophy Workshop 2005.
Abstract 16. 2. Noor MA, et al. AIDS.
2004182137-2144.
90
Unboosted Amprenavir ShowsNo Effect on Insulin
Sensitivity

• N 6 HIV-negative men
• Patients given single-dose APV (1200 mg) or
  placebo
• Administered 1 hr before euglycemic
  hyperinsulinemic clamp
• At peak APV levels, fasting insulin, glucose, and
  insulin resistance (HOMA) not altered

12
10
8
Mean Insulin-Mediated Glucose Disposal Rate
(mg/kgmin per uU/mL)
6
4
2
0
Placebo
APV
Grunfeld C, et al. Lipodystrophy Workshop 2005.
Abstract 15.
91
Insulin Resistance Management Summary

• Insulin resistance in HIV infection is often a
  toxicity of specific drugs, so clinicians should
  consider avoiding causative agents or
  preemptively switching from them
• There is no indication to treat insulin
  resistance in the absence of overt diabetes
  mellitus
• Diabetes mellitus in HIV-infected patients should
  be treated in the standard manner recommended for
  the HIV-negative population

92
Body Fat Changes

• 2 phenotypes occur in HIV-infected patients
• HIV-associated lipoatrophy
• Almost certainly due to toxicity of thymidine
  analogues
• Metabolic effects incompletely described but may
  contribute to insulin resistance
• Aging-associated and HIV-associated visceral
  adiposity
• Weight gain common initially with all
  antiretroviral therapy
• May be exacerbated by certain drugs (eg, some PIs)

93
MACS Increase in Waist Size in HIV Men May Be
Return to Normal

• Longitudinal measurements of BMI, circumference
  of waist, hip, thigh, and midarm in 661
  HIV-infected and 392 HIV-uninfected men enrolled
  in Multicenter AIDS Cohort Study
• NRTIs associated with BMI decrease NNRTIs and
  PIs not associated
• Cumulative NRTI exposure associated with
  significant decreases in waist, hip, thigh
  circumferences
• Significant increases in BMI, waist, and hip
  circumference observed for each year of
  follow-up, regardless of HIV status
• Waist circumference lower at baseline in
  HIV-positive men and increased more rapidly than
  in HIV-negative controls
• Suggests increased waist may reflect return to
  normal

Brown T, et al. IAC 2006. Abstract WEPE0136.
94
Predictors of Metabolic Syndrome Baseline Data
from the DAD Study
Worm S, et al. Intl Congress on Drug Therapy in
HIV Infection 2006. Abstract P124.
95
Lipoatrophy Risk Factors

• Almost certainly interrelated
• Antiretroviral therapy
• Thymidine analogue exposure (d4T gt ZDV)
• ? PI use
• Host factors
• Age
• HIV disease factors
• Duration of illness
• Severity of illness AIDS, low CD4 cell count

96
Consequences of Lipoatrophy

• Personal appearance
• Stigmatizing
• Depression
• Physical evidence of being HIV positive
• Metabolic
• Contributes to and potentially maintains insulin
  resistance

97
Reduced Adherence to HAART With the Development
of Body Fat Changes

• N 83 patients on HAART with self-reported
  morphological changes
• Self-reported adherence declined with time

120
100
100
92
82
75
80
Patients Reporting Adherence to Treatment ()
60
40
20
0
0-6 mos
6-12 mos
12-24 mos
gt 24 mos
Time Since Self-Reported Morphologic Alterations
Guaraldi G, et al. HIV Clin Trials. 2003499-106.
98
A5005s Thymidine Analogues Associated With Limb
Fat Loss

• Subjects given ZDV/3TC or ddI/d4T EFV, NFV, or
  both
• Both NRTI combinations associated with early fat
  ?
• Likely restoration of HIV-related fat loss
• By Week 48 onward, subjects on ddI/d4T had
  significantly more limb fat loss than those
  taking ZDV/3TC

Dubé MP, et al. AIDS. 2005191807-1818.
99
GS 903 and GS 934 Differential Effect of NRTIs
on Total Limb Fat
Study 903
Study 934
TDF 3TC EFV
TDF FTC EFV
d4T 3TC EFV
ZDV/3TC EFV
10
12
9
8.6
7.9
8
10
7
8
8
8
7.4
6
5.0
Mean Limb Fat (kg)
5
Total Limb Fat (kg)
6
6
6
4.5
5.5


6.0
4
4
4
4
3
P .01
P .034
2
P lt .001
2
2
2
1
P .001
0
0
0
0
48
96
144
48
96
0
Week
Week
n 128 115 n 134 117
n 51 49 n 49 44
Gallant JE, et al. JAMA. 2004292191-201. Pozniak
AL, et al. J Acquir Immun Defic Syndr. 2006 Epub.
100
Treatment of Body Fat Abnormalities IAS-USA
Panel Guidelines

• There is no consensus as to whether it is
  appropriate to treat fat distribution
  abnormalities in the absence of other metabolic
  complications. . . . There are no proven or
  approved therapies for fat distribution
  abnormalities, although a number are under
  investigation.
• Because there is considerable etiologic and
  phenotypic diversity in fat distribution
  abnormalities, it is extremely unlikely that a
  single therapeutic approach will apply in all
  cases. . . if the decision is made to intervene,
  it is crucial to identify the specific objectives
  of treatment.

Schambelan M, et al. J Acquir Immune Defic Syndr.
200231257-275.
101
Therapeutic Strategiesfor Lipodystrophy
Other factors
Other mechanisms
Antiretroviraldrugs
Metabolic adipocyteabnormalities
Body fatchanges
Drugs withmetaboliceffects
Plasticsurgery
Change/withdrawal
102
Treatment Interventionsfor Lipoatrophy
1. Carr A, et al. JAMA. 2002288207-215. 2.
Martin A, et al. AIDS. 2004181029-1036. 3.
McComsey G, et al. Clin Infect Dis.
200438263-270. 4. Moyle G, et al. CROI 2005.
Abstract 44LB. 5. Milinkovic A, et al. CROI
2005. Abstract 857. 6. Murphy R, et al. CROI
2005. Abstract 45LB. 7. Tebas P, et al. CROI
2005. Abstract 40. 8. Sutinen J , et al.
Lipodystrophy Workshop 2005. Abstract 7. 9.
Mallon P, et al. AIDS. In press. 10. Sutinen J,
et al. Antiviral Ther. 20038199-207. 11. Carr
A, et al. Lancet. 2004363429-438. 12. Hadigen
C, et al. Ann Intern Med. 2004140786-794. 13.
Cavalcanti R, et al. CROI 2005. Abstract 854. 14.
Slama L, et al. CROI 2006. Abstract 151LB.
103
MITOX Study Change in Limb Fat Mass After Switch
From d4T or ZDV to ABC
2.0
ABC
ABC from Week 24
ZDV/d4T only
1.5
1.0
Mean Change in Limb FatFrom Baseline (kg)
0.5
0
-0.5
-1.0
0
24
48
72
104
Weeks
No. patients ABC ABC from week 24 ZDV/d4T only
47 23 29
42 23 25
35 15 22
33 19 19
28 13 18
Martin A, et al. AIDS. 2004181029-1036.
104
RAVE Change in Limb Fat After Switch From d4T or
ZDV to ABC or TDF
Switch to TDF
Switch to ABC

• Median baseline limb fat, g (range)
• TDF 3000 (1200-3900)
• ABC 2900 (1000-10,800)
• No significant difference between arms in change
  in limb fat at Week 48.

P .01
600
483
500
P .0001
400
329
Mean Change in Limb Fat by DEXA (g)
300
206
200
143
100
0
24
48
Week
DEXA, dual x-ray absorptiometry.
Moyle GJ, et al. AIDS. 2006202043-2050.
105
ACTG A5125S Change in Limb Fat With Switch to
NRTI-Sparing Regimen

• Rollover study for suppressed patients receiving
  IDV 2 NRTIs NFV or EFV in ACTG 388
• 62 patients randomized
• EFV LPV/RTV or
• EFV 2 NRTIs (78 ZDV/3TC, 20 d4T/3TC, 2
  other)
• Increase in peripheral fat with NRTI-sparing
  regimen at Week 48 decrease with NRTI-containing
  regimen
• No significant differences in truncal fat,
  glucose metabolism, HOMA-IR, BMD

EFV NRTIs
EFV LPV/RTV
1250
P .002
1000
P .085
750
P .07
500
P .05
250
Fat Change (g)
0
-250
-500
Baseline 6 kg
-750
P .007
-1000
-1250
0
48
Final
Weeks
Between arms. Within arm.
Tebas P, et al. CROI 2005. Abstract 40.
106
Pharmacologic Interventions Have Limited Efficacy
and Present Risks
Adapted from Sutinen J. Curr Opin Infect Dis.
20051825-33.
107
STARs Study Recombinant Human Growth Hormone for
Fat Accumulation
Placebo
4 mg AD
4 mg DD
0
-10
VAT on CT L4-5 (cm2)
-20
P .052
-30
P lt .001
P values for change from baseline to Week 12
compared with placebo group
-40
Placebo
4 mg AD
4 mg DD
0
-0.1
-0.2
TrunkLimb Fat Ratio by DEXA
-0.3
P lt .001
-0.4
P lt .001
-0.5
AD, alternative days DD, daily dosage, VAT,
visceral adipose tissue.
Kotler DP, et al. J Acquir Immime Defic Syndr.
200435239-252.
108
Growth Hormone Effect on Visceral Abdominal Fat

• Induction therapy for 12 weeks associated with
• Significant reduction in VAT
• Reduction in trunk fat
• Improvement in cholesterol profile
• Small but statistically significant decrease in
  abdominal subcutaneous adipose tissue
• During maintenance therapy, 40.3 of patients on
  r-hGH regained gt 50 of VAT lost during induction
  therapy
• AEs mostly mild to moderate

Induction Therapy (CT Scan)
10 5 0 -5 -10 -15 -20 -25
Change From Baseline ()

VATSAT


Placebo
r-hGH 4 mg QD
P lt .001 for r-hGH vs placebo.
FDA-mandated endpoint.
Grunfeld C, et al. IAC 2006. Abstract THLB0212.
109
Rosiglitazone for Lipoatrophy

• Decreased PPAR-? mRNA expression is found in
  adipose tissue biopsies from HIV-positive
  patients with lipoatrophy
• Rosiglitazone, a PPAR-? agonist, promotes an
  increase in subcutaneous fat in HIV-negative
  subjects
• Studies of rosiglitazone in HIV-associated
  lipoatrophy have shown mixed results
• Fat aspirates were performed in a substudy of the
  Rosey study1
• Rosiglitazone-induced PPAR-? expression was
  blunted by ongoing NRTI use2
• Results provide a molecular explanation for lack
  of response to rosiglitazone in HIV-positive
  patients
• Results do not prove that rosiglitazone has
  additional effect to that of switching from
  thymidine NRTI

1. Carr A, et al. Lancet. 2004363429-438. 2.
Mallon P, et al. Lipodystrophy Workshop 2005.
Abstract 41.
110
Rosiglitazone vs Placebo in HIV-Infected
Patients Change in Limb Fat
0.25
Rosiglitazone
Placebo
0.20
0.15
Mean Change From Baseline (kg)
P .89
0.10
0.05
0.00
0
24
48
Weeks
RSG 53 52 PLA 55 53
Carr A, et al. Lancet. 2004363429-438.
111
Rosiglitazone vs Metformin in HIV-Infected
Patients

• 39 HIV-infected men with lipodystrophy randomized
  to rosiglitazone (8 mg/day) or metformin (2
  g/day) for 26 weeks
• On antiviral therapy for mean of 5.4-5.6 years
• Whereas both drugs increased insulin sensitivity,
  other effects differed between these agents

van Wijk JP, et al. ICAAC 2005. Abstract
H-339.van Wijk JP, et al. Ann Intern Med.
2005143337-346.
112
Impact of Pravastatin on Lipids and Limb Fat

• 33 HIV-infected men randomized to pravastatin 40
  mg or placebo after 4 weeks of dietary advice

Pravastatin
Placebo
TC
HDL
Non-HDL
TGs
0.2
0.8
0.72
0.06
-0.03
P .04
0.7
0
-0.03
0.6
-0.2
0.5
P .09
-0.34
-0.4
Change in Limb Fat (kg)
-0.4
Change in Lipids (mmol/L)
-0.42
0.4
-0.6
P NS
0.3
0.19
-0.8
0.2
-0.82
-1.0
0.1
P .04
-1.02
0
-1.2
P .01
Limb Fat
Change from Week 4 of dietary advice to Week 12
of treatment.
Mallon P, et al. Lipodystrophy Workshop 2005.
Abstract 23.
113
Uridine Supplementation as Potential Treatment
for Lipoatrophy
Placebo
Uridine supplement
P lt .05
P lt .01
300
2000
1918
1600
200
205
1200
100
Mean Change From Baseline to Month 3 (g)
Mean Change From Baseline to Month 3 (cm3)
800
534
0
350
-81
400
239
111
119
-100
0
Leg Fat
Arm Fat
Total Body Fat
Intra-abdominal Fat
-200
P lt .05 from baseline.
Sutinen J, et al. Lipodystrophy Workshop 2005.
Abstract 7.
114
Polylactate Acid (PLA) Injections for Facial
Lipoatrophy
N 30
P .05
Immediate injections
10
9
P lt .001
8
7
Delayed injections
6
6
Scale
4
2
1
0
Self-Perception
Anxiety Score

• Changes in Self-Assessed Visual Appearance and
  Anxiety Level at Week 12

PLA given on Week 0, 2, and 4. PLA given on
Week 12, 14, and 16.
Moyle G, et al. HIV Med. 2004582-87.
115
Lipoatrophy Management Summary

• Lipoatrophy is only partially and slowly
  reversible, so consider avoiding causative agents
  or preemptively switching from them
• Pharmacologic interventions to treat
  lipodystrophy do not appear effective and may
  introduce new issues
• Switching 1 or more antiretrovirals is a viable
  strategy to reduce or prevent toxicities without
  sacrificing efficacy
• Switching or withdrawal of thymidine analogues is
  the only antiretroviral intervention proven to
  partially restore subcutaneous fat

116
Go Online for More CCO Coverage ofMetabolic
Complications in HIV Patients

• CCO Slidesets of all the key data in 4
  free-to-download presentations
• Available in English, French, German, and Italian

clinicaloptions.com/metabolic2006