Morbidity and Mortality in the HAART era

1
Morbidity and Mortality in the HAART era

• Andrew Phillips
• Royal Free University College Medical School
• London

2
Death in the HAART era rates and reasons
3
Trends in death rate HOPS
4
Trends in death rate over calendar time in UK
Rate per 100 people
Number of deaths in year Number seen for care
in year
96 97 98 99 00 01
02 03 04 05 06
Year
1481 749 514 472 484 477
520 572 495 539 497
Deaths
15 16 18 20 22 26
32 36 41 46 52
Seen for care(thousands,rounded)
Source HPA
5
Breakdown of causes of death France 2005
AIDS
Cancer Hepatitis C
CVD Suicide Non-AIDS
infection Accident
Hepatitis B Liver disease OD /
drug abuse neurologic
renal pulmonary
digestive iatrogenic
metabolic psychiatric
other unknown
N 937 deaths
Lewden et al, CROI 2007
ANRS EN19 Mortalité 2005
Percent
6
Audit of 397 deaths in UK 2005 Scenario leading
to AIDS-related deaths
Scenario of AIDS deaths Diagnosed too late
for effective treatment 40 Under care,
but with untreatable complication
29 Treatment ineffective due to poor adherence
12 Chose not to receive treatment
8 Known HIV, not under regular care,
6 re-presented too late MDR HIV, ran out of
options 5
BHIVA Audit Johnson et al 2006
7
Incidence of non-AIDS death 1994-2004
(excluding death from unknown causes)
Rate per 100 person years
Test for trend p lt 0.0001
95 96 97 98 99 00
01 02 03 04
Year
EuroSIDA Mocroft, Lundgren et al, personal
communication
8

• Might HIV increase the risk of serious
• non-AIDS conditions and non-AIDS
• death ?

9

• Might HIV increase the risk of serious non-
• AIDS conditions and non-AIDS death ?
• Incidence of, and death from
• - Non-AIDS malignancies
• - End stage renal disease
• - Cardiovascular events
• - Liver cirrhosis
• - Deaths from other non-AIDS causes

• Not focussing on adverse effects of ART

10

• Possible mechanisms General
• - Very early loss of CD4 T cells in
  gastrointestinal tract
• Loss of immunological and epithelial integrity
  of the mucosal
• barrier leading to microbial translocation
• Generalized immune activation
• - Fibrosis of lymphatic tissue

Veazey et al, Science 1998 Brenchley et al,
Nature Med 2006 Brenchley, J Exp Med
2004 Schacker et al, Clin Vacc Immunol 2006
11
Possible mechanisms Non-AIDS malignancies

• Immunodeficiency, leading to
• - reduced control of oncogenic pathogens
• - damage due to infections and resulting
• chronic inflammation
• - loss of ability to identify transformed
  cells

Littman et al. Cancer Epidemiol Biomarkers Prev
2005
12
Possible mechanisms Kidney disease

• HIV associated nephropathy (HIVAN)
• (viral nephritis reversed by ART)
• Link with other kidney pathologies (e.g.
• immune complex glomerulonephritis)
• High prevalence of proteinuria, associated
• with HIV RNA level and CD4 count
• HIV RNA and CD4 count predict raised
• creatinine levels
• proteinurea elevated creatinine associated
• with all cause mortality in HIV patients

Szczech et al, Kidney International 2002 Lucas
et al, AIDS 2004 Szczech et al, Kidney
International 2004 Kimmel et al, Ann Intern Med
2003
13
Possible mechanisms Cardiovascular disease

• Association of HIV-infection with adverse
• changes in known or potential biomarkers for
• CVD.
• HDL-cholesterol depletion
• Inflammation (raised IL-6, C-reactive protein)
• Endothelial activation/dysfunction (VCAM, ICAM)
• Activation of coagulation (D-dimer)
• Several of the changes appear to be at least
• partially reversed by ART

Riddler JAMA 2003 de Larranaga et al, Blood
Coag. Fibrinolys 2003 Lau et al, Arch Intern
Med 2006 Wolf et al, J Infect Dis 2002
14
Possible mechanisms Liver disease

• immunodeficiency linked to more
• rapid progression of liver fibrosis in
• HBV and HCV infected people
• affect CD4 and CD8 response to
• HBV / HCV
• - alter HBV / HCV quasi-species
• increased hepatocyte apoptosis

Tan et al, Current HIV research 2006 Thio et al,
Lancet 2002 Eyster et al, JAIDS 1993 Soto et
al, J Hepatol 1997
15

• Might HIV increase the risk of serious non-
• AIDS conditions and non-AIDS death ?
• Types of evidence
• comparison of risk of serious
  non-AIDS events
• between HIV-infected and HIV-uninfected
  people
• studies of the association between CD4
  count
• (and HIV RNA) and risk of serious non-AIDS
• events
• randomized trials of the impact on serious
  non-
• AIDS events of reduction in HIV RNA level
  and
• increase in CD4 count with ART

16
Comparison of risk of events between
HIV- infected and HIV-uninfected people
limitations

• HIV -ve comparison group will differ from
  HIV-infected
• group in more ways than just the HIV infection
  (eg smoking)
• Adjustment for such confounding bias may not
  be possible.
• Each non-AIDS condition has its own set of
  risk factors
• which could act differently in HIV-infected
  people.
• HIV infected subjects often mixture of those on
  ART and
• ART-naïve, so not possible in all studies to
  distinguish
• effect of HIV from effect of ART.

17
HIV and risk of non-AIDS malignancies
Meta-analysis 444,172 people with HIV, 31,977
transplant patients
For 20 / 28 cancers examined there was
significantly increased incidence in both groups
strongly suggesting a link with
immunodeficiency Standardized
Incidence Ratio HIV/AIDS
Transplant Lung 2.7 2.2 Leukaemia
3.2 2.4 Kidney 1.5 6.8 Oesophagus
1.6 3.1 Stomach 1.9 2.0
Grulich et al, Lancet 2007
18
HIV and risk of lung cancer, independent of
smoking
CID 2007
AIDS 2007
19
HIV and risk of End Stage Renal disease
U.S. Veterans without diabetes
Hazard ratio for End Stage Renal Disease
people ESRD Hazard ratio White HIV
-ve 1,201,870 3991 1.0 HIV ve 6,139
13 0.8 (0.5 1.3) Black HIV -ve
206,636 1425 2.0 (1.9 2.2) HIV ve
6,816 129 4.6 (3.4 6.1)
Adjusted for age, sex, baseline eGFR category,
CAD, HTN, heart failure, COPD, PVD, HCV
infection, cerebrovascular disease, and
SES. Little effect of HIV in diabetics
Choi et al J Am Soc Nephr 2007
20
HIV and Cardiovascular Disease
Subject source N CVD Risk in HIV cases
in vs. HIV ve HIV Klein Administr
ative 72 Increased clinical
management database Mary-Krause HIV cohort
/ 60 Increased in those general
population with gt 18 m PI use Currier Adminstr
ative 1360 Increased at database younger
ages Triant Patient Data 189 Increased R
egistry
Klein et al, JAIDS 2002 Mary-Krause et al,
AIDS 2003 Currier et al, JAIDS 2003 Triant et
al, J Clin Endocrin Metab 2007
21
HIV and Liver disease
4865 men and boys with haemophilia (and probable
HCV infection), of whom 1218 HIV-infected

• HIV (and haemophilia) status 25 year
  cumulative risk
• of liver death
• Severe haemophilia, not HIV 1.4 (0.7 3.0)
• Moderate / mild haemophilia, not HIV 1.2 (0.5
  2.6)
• HIV-infected (all haemophilia severities) 6.5
  (4.5 9.5)

Darby et al, Lancet 1997
Similarly for HBV in MACS Thio et al, Lancet
2002
22
All cause death rates in ART-naïve patients with
CD4 count gt 350 /mm3, compared with the general
population
Abstract N-264 Wednesday 10.30 - Lodwick et
al
23

• Might HIV increase the risk of serious non-
• AIDS conditions and non-AIDS death ?
• Types of evidence
• comparison of risk of serious non-AIDS
  events
• between HIV-infected and HIV-uninfected
  people
• studies of the association between CD4
  count
• (and HIV RNA) and risk of serious non-AIDS
• events
• randomized trials of the impact on serious
  non-
• AIDS events of reduction in HIV RNA level
  and
• increase in CD4 count with ART

24
CD4 count and risk of death DAD and CASCADE
1.6
1.2
CASCADE (ART-naïve)
0.8
Rate / 100 person years 95 CI
0.4
0.0
Non-AIDS causes
All causes
1.6
1.2
DAD
0.8
0.4
0.0
200 350 gt 500 349 499
200 350 gt 500 349 499
CD4 count (/mm3)
Weber at al Marin et al
25
Hospitalization events according to cause and
CD4 count Aquitaine cohort, 2000-2004
3863 patients
Number of hospitalizations Number (mean
per patient) during 2000-2004 CD4 count of
patients gt 500 2442 16 (0.01) 335
(0.14) 351 (0.14) 200-499 2922 60 (0.02) 581
(0.20) 641 (0.22) lt 200 1229 260 (0.21) 439
(0.36) 699 (0.57)
AIDS non-AIDS All
p lt 0.001
p lt 0.001
Bonnet et al, HIV Medicine 2007
ANRS C03 Aquitaine Cohort
26
HIV RNA and risk of serious non-AIDS events SMART
All serious non-AIDS
Non-AIDS malignancy
Renal
CVD
Liver
Other non-AIDS death
1.0
1.5
0.5
0.2
Adjusted hazard ratio lt 400 vs. gt 400
copies/mL
Adjusted for age, gender, prior AIDS, hep B/C,
smoking, latest CD4 count
SMART, unpublished
27

• Might HIV increase the risk of serious non-
• AIDS conditions and non-AIDS death ?
• Types of evidence
• comparison of risk of serious non-AIDS
  events
• between HIV-infected and HIV-uninfected
  people
• studies of the association between CD4
  count
• (and HIV RNA) and risk of serious non-AIDS
• events
• randomized trials of the impact on serious
  non-
• AIDS events of reduction in HIV RNA level
  and
• increase in CD4 count with ART

28
SMART Study
Participants with CD4 count gt 350
84 on ART, 16 off ART
Randomization
n 2752
n 2720
Intermittent ART Stop or defer ART when CD4 count
gt 350, restart or start ART when CD4 count lt 250
Continuous ART
Follow-up
33 on ART 96 CD4 gt 200
94 on ART 99 CD4 gt 200
N Engl J Med 2006
29
Risk of serious non-AIDS events in SMART
Number of events Intermittent Continuous
ART ART
113 73
All serious non-AIDS
Non-AIDS malignancy 27 24

Renal 9
2
CVD 48
31
Liver 10
7
Other non-AIDS death 30
16
1
2
0.5
3
5
10
Of the 85 deaths that occurred in SMART, only 7
(8) were from AIDS diseases
Hazard ratio Intermittent ART
vs. Continuous ART
SMART Study Group, NEJM 2006 Neaton et al,
Current Opinion in HIV/AIDS 2008
30
Risk of serious non-AIDS events in SMART
patients ART naïve or off ART for gt 6 months
N 477 patients
Number of events Hazard ratio Deferred
vs. Deferred Immediate Immediate
ART ART ART (95 CI)
p-value 12 2
7.02 (1.57 31.4) 0.01
Emery et al, JID (in press)
31

• Inflammatory and coagulation markers in SMART
• Abstract D-60 Wednesday 10.00 Kuller et al
• Illustrates value of biomarker studies based on
  stored
• samples from a randomized trial with clinical
  endpoints

32

• Might HIV increase the risk of serious non-AIDS
  conditions and non-AIDS death ?
• Summary
• On balance, evidence suggests HIV may well play
  a role in
• several serious non-AIDS defining events.
• In the upper CD4 count range, while overall
  risk of any disease is
• relatively low, non-AIDS events are much more
  common than
• AIDS events.
• Given the associations with latest level of CD4
  count / HIV RNA
• and the results from SMART use of ART may well
  reduce risk of
• some serious non-AIDS events.

33
What steps can we take towards
further reduction in morbidity and
mortality ?
34

• What steps can we take towards further reduction
  in morbidity and mortality ?
• Continued efforts to diagnose HIV as early as
  possible
• Research into prediction of non-AIDS events in
  context of HIV
• - ART-naïve and ART-treated
• - standardize diagnostic criteria and
  data collection methods
• Trial of ART initiation in people with CD4
  count gt 500 /mm3
• - non-AIDS diseases relatively common
  at higher CD4 count
• - SMART suggests risk / benefits of ART
  favour benefit
• - durable virological benefit of
  current ART
• - cost-effectiveness / reduction in
  transmission risk
• - basis for identifying biomarkers that
  mediate raised
• risk, providing insights into
  mechanisms (also beyond HIV)

35

• Conclusions
• The study of serious non-AIDS conditions is an
  important
• emerging area for HIV research
• Research is needed to provide a basis for
  defining models
• of care for people with HIV which take into
  account the risk
• of all serious conditions
• Research into mechanisms by which HIV affects
  risk of non-
• AIDS conditions is needed, and it may help us
  understand
• more about the causes of such conditions
  outside HIV
• - The possibility that ART should be initiated
  much earlier
• should be investigated in a randomized trial.
  Such a trial
• will form a key resource for this new
  research area.

36
Acknowledgements Jens Lundgren, Jim Neaton HIV
Epidemiology Biostatistics Group, Royal Free,
UCL Caroline Sabin, Amanda Mocroft, Fiona Lampe,
Alessandro Cozz-Lepri, Colette Smith, Zoe Fox,
Wendy Bannister, Loveleen Bansi, Rebecca Lodwick,
Joanne Reekie DAD (Aquitaine, Nice, CPCRA,
EuroSIDA, ICONA,SHCS, Brussels, BASS,
AHOD, ATHENA, HivBivus) EuroSIDA SMART
FIRST CASCADE HOPS Extra analyses Jacquie
Neuhaus (SMART), Grace Peng, Jason Baker
(FIRST), Benoit Marin, Genevieve Chene, Abdel
Babiker (CASCADE), Colette Smith, Caroline Sabin
(DAD), Amanda Mocroft (EuroSIDA)