Impact of Specific NRTI and PI Exposure on the Risk of Myocardial Infarction A Case-Control Study Nested within the French Hospital Database on HIV ANRS CO4

1
Impact of Specific NRTI and PI Exposure on the
Risk of Myocardial InfarctionA Case-Control
Study Nested within the French Hospital Database
on HIVANRS CO4

• S Lang, M Mary-Krause, L Cotte, J Gilquin,
  M Partisani, A Simon, F Boccara, D Costagliola

Unité 943
2
Background - I
Risk of myocardial infarction and exposure to
protease inhibitors
RR (95 CI) 1.16 (1.10 1.23)
Cumulative exposure to protease inhibitors (Pis)
has been associated with an increased risk of
myocardial infarction but the role of specific
protease inhibitors has not been reported
Mary-Krause M et al., AIDS 2003 17 2479 -
2486 DAD Study Group, Friis-Møller N et al., N
Engl J Med 2007 3561723-1735
3
Background - II
Risk of myocardial infarction according to
exposure to abacavir
DAD Study Patients with MI n 517 Recent
exposure to abacavir was associated with a
higher risk of MI RR (95 CI) 1.94 (1.48 2.55)
SMART Study Patients with MI n 19 Current use
of abacavir was associated with an increased risk
of MI HR (95 CI) 4.3 (1.4 13.0)
Interactions with CVD risk factors were in the
opposite direction in the two studies
In these 2 studies, exposure to abacavir showed
signals which were not completely concordant
DAD study group, Sabin CA, et al., Lancet 2008
371 1417-26 The SMART/Insight and the DAD
study groups, Lundgren JD et al., AIDS 2008 22
F17-F24
4
Objectives

• In a nested case-control study within the French
  Hospital Database on HIV, to evaluate the
  association between the risk of myocardial
  infarction (MI) and
• cumulative to specific NRTIs
• recent (current or within last 6 months) and past
  exposure (gt6 months ago) to specific NRTIs
• cumulative exposure to specific PIs

5
Cases

• Over 115000 HIV-infected patients have been
  enrolled into the FHDH between 1989 and 2006
• Patients with a first MI prospectively reported
  between January 2000 and December 2006 were
  included
• Only definite or probable MI cases validated by a
  cardiologist (FB) according to the ASC/ESC
  criteria were eligible
• Out of the 418 cases identified, 129 were
  excluded
• 45 had incomplete medical records
• 36 MIs occurred before the study period
• 2 cases of MI were undated
• 4 cases of MI occurred before the diagnosis of
  HIV infection
• 6 cases had a MI before being enrolled in the
  cohort
• 36 cases did not have a confirmed MI

Luepker R et al., Circulation 2003 108
2543-2549
6
Controls

• HIV-infected patients with no history of MI,
  followed at the time of MI diagnosis of the
  corresponding case
• Matched for
• Age at diagnosis of MI 3 years
• Sex
• Clinical center
• Matching based on these factors yields similar
  results in a nested case-control study to those
  obtained with the cohort approach used in our
  first study on the risk of MI
• For each validated case, up to five matched
  controls randomly selected with replacement from
  the database
• Cases eligible as control up to the time of the
  diagnosis of MI
• 3 cases with 1 control, 11 with 2, 246 with 3, 24
  with 4 and 5 with 5 controls

Guiguet M et al., Pharmacoepid Drug Saf, 2008
17 468-474
7
Methods

• Data collected for cases and controls
• Cardiovascular risk factors
• Smoking, family history, hypertension,
  hyperlipidemia, diabetes
• Treatments for lipid, metabolic and hypertensive
  disorders
• BMI, current IV drug use
• Validation of HIV data recorded in the database
• CD4 cell count, current (within 3 months of MI)
  and nadir
• CD4/CD8 ratio (within 3 months of MI)
• Plasma HIV-1 RNA (within 3 months of MI)
• ART treatment history
• Stage C (AIDS) before MI

8
Analyses - I

• Several conditional logistic regression models
  were constructed
• A first model including cumulative exposure to
  each ART
• A second model including cumulative exposure to
  each ART and exposure to each NRTI as a
  three-class variable
• no exposure
• last use gt 6 months (past)
• ongoing exposure or interruption lt 6 months
  (current/recent)
• In these models, potential confounders which
  affected the association between any ART and the
  risk of MI by at least 10 in any of the models
  were included from
• Age, smoking, family history of CHD, BMI,
  hypertension, intravenous drug use
• CD4 cell nadir, plasma HIV-1 RNA, CD4 cell count,
  CD4/CD8 cells ratio within 3 months before MI and
  AIDS before MI

9
Analyses - II

• Odds Ratios (OR) are reported only for NRTIs and
  PIs with at least 100 exposed patients
• AZT, ddI, ddC, d4T, 3TC, ABC, TNF
• SQV, IDV, NFV, LPV, APV/fAPV
• although cumulative exposures to FTC, EFV, NVP,
  ATV and TPV were also accounted for in the
  analyses

10
Characteristics
Cases (n 289) Controls (n 884)
Sex, male, n ( ) 257 (88.9) 788 (89.1)
Age, years, median (IQR ) 46.9 (40.7 54.1) 46.3 (40.2 53.7)
Hypertension, n () 59 (20.6) 102 (11.8)
Smoking, n () 210 (72.7) 388 (43.9)
Family history of CHD, n () 53 (18.5) 58 (6.7)
Hypercholesterolemia, n () 148 (51.7) 282 (32.6)
Intravenous drug use, n () 38 (13.3) 83 (9.5)
Number of CV risk factors 0, n () 1-2 n, () 3 n, () 3 (1.0) 172 (59.5) 114 (39.4) 163 (18.4) 553 (62.6) 168 (19.0)
Viral load, copies/mL, median (IQR) 127 (50 - 3900) 50 (50 1368)
Viral load lt50 copies/mL, n () 125 (43.3) 457 (51.7)
CD4 count, cells/mm3 median (IQR) 427 (256 - 638) 451 (291 634)
CD4/CD8 ratio ? 1, n () 19 (6.6) 116 (13.1)
No treatment before MI, n () 11 (3.8) 55 (6.2)
1st treatment after inclusion in FHDH, n () 210 (72.7) 677 (76.6)
11
Exposure to abacavir and risk of MI - I
N exposed N exposed cases OR 95 CI p value
Cum exposure to abacavir 410 127 0.97 0.86 - 1.10 0.651
Model 1
12
Exposure to abacavir and risk of MI - II
N exposed N exposed cases OR 95 CI p value
Cumulative exp to abacavir 410 127 0.97 0.86 - 1.10 0.651
Model 1
Cumulative exp to abacavir 410 127 0.88 0.75 - 1.04 0.138
No exposure Current/Recent exposure Past exposure 763 290 120 162 88 39 1 1.57 1.59 - 0.91 - 2.72 0.89 - 2.83 - 0.107 0.116
Model 2
For abacavir, there was evidence of an
interaction between recent/past and cumulative
exposure, while no such effect was observed for
any other NRTI

• A final model including exposure to abacavir as a
  five-class variable
• and cumulative exposure to all other ART was
  constructed
• no exposure
• exposure lt 1 year and last use lt 6 months prior
  to the MI (current/recent)
• exposure lt 1 year and last use gt 6 months prior
  to the MI (past)
• exposure gt 1 year and last use lt 6 months prior
  to the MI (current/recent)
• exposure gt 1 year and last use gt 6 months prior
  to the MI (past)

13
Exposure to abacavir and other NRTIs and risk of
MI - III
N exposed N exposed cases OR 95 CI p value
No exposure Expo lt 1 year, current/recent Expo lt 1 year, past Expo gt 1 year, current/recent Expo gt 1 year, past 763 72 76 218 44 162 31 24 57 15 1 1.97 1.31 1.05 1.42 - 1.09 - 3.56 0.68 - 2.52 0.65 - 1.69 0.60 - 3.35 - 0.025 0.415 0.844 0.420
Cum exp to zidovudine 998 256 1.08 0.99 - 1.18 0.086
Cum exp to didanosine 691 186 0.91 0.82 - 1.01 0.071
Cum exp to zalcitabine 314 92 0.99 0.81 - 1.21 0.924
Cum exp to stavudine 718 199 1.09 0.98 - 1.22 0.132
Cum exp to lamivudine 1043 269 0.95 0.85 - 1.07 0.387
Cum exp to tenofovir 238 65 0.97 0.75 - 1.24 0.785
Final model
No interaction was found between exposure to
abacavir and numbers of CV risk factors on the
risk of MI (p 0.384) Similar results were
observed when restricting the analysis to
patients with first ART after inclusion in the
cohort
14
Exposure to PIs and risk of MI
Cumulative exposure (per additional year) N exposed N exposed cases OR 95 CI p value
Saquinavir /-r 324 92 0.96 0.80 1.15 0.669
Indinavir /-r 497 146 1.10 0.98 1.24 0.117
Nelfinavir 453 131 1.12 0.98 1.28 0.110
Lopinavir/r 290 94 1.37 1.09 1.72 0.006
Amprenavir/fos-amp /-r 117 46 1.52 1.19 1.95 0.001
Final model
Final model combining all PIs but SQV
Cumulative exposure (per additional year) N exposed N exposed cases OR 95 CI p value
PI /-r 864 239 1.16 1.07 1.26 lt0.001
Saquinavir /-r 324 92 0.95 0.83 1.10 0.502
Similar results were observed when restricting
the analysis to patients with first ART after
inclusion in the cohort
15
Interpretation - IExposure to abacavir and risk
of MI

• We found a signal slightly different from those
  of the DAD study and of the SMART Study
• only early exposure to abacavir was associated
  with an increased risk of MI
• no interaction between exposure to abacavir and
  CV risk factors on the risk of MI

16
Interpretation - IIExposure to other NRTIs and
risk of MI

• Trends towards an increased risk of MI by
  cumulative exposure to AZT and to d4T were
  evidenced
• In line with the original hypothesis in the DAD
  study
• These associations deserve additional evaluations
  in independent studies
• No signal was evidenced for the other NRTIs,
  including ddI and TNF

17
Interpretation - III Exposure to PIs and risk of
MI

• In our study the association between the risk of
  MI and cumulative exposure to PI was in
  concordance with that observed in the DAD study
• Increased risk for all studied PIs, but
  saquinavir
• Significant in specific analyses for lopinavir/r
  and amprenavir/fos-amprenavir /-r
• Unlikely explained by selection biases and
  confounding
• After 10 years of exposure, the risk would be
  increased by 4.4

18
Acknowledgments - I

• We thank the study team without whom it would
  have been impossible to complete the study in
  time
• Lydie Béniguel, Sandra Firmin, Sophie
  Pakianather, Serge Rodrigues, Selma Trabelsi,
  Sarah William-Faltaos
• We are grateful to the following colleagues who
  read and provided comments on the analysis plan
• S Evans, M Hernán, C Sabin and I Weller
• A special thank to Rob Murphy
• The study was funded by ANRS

19
Acknowledgments - II

• Clinical Epidemiology Group of the FHDH
• Scientific committee S Abgrall, F Barin, M
  Bentata, E Billaud, F Boué, C Burty, A Cabié,
  D Costagliola, L Cotte, P De Truchis, X Duval, C
  Duvivier, P Enel, L Fredouille-Heripret,
  J Gasnault, C Gaud, J Gilquin, S Grabar, C
  Katlama, MA Khuong, JM Lang, AS Lascaux,
  O Launay, A Mahamat, M Mary-Krause, S Matheron,
  JL Meynard, J Pavie, G Pialoux, F Pilorgé, I
  Poizot-Martin, C Pradier, J Reynes, E Rouveix, A
  Simon, P Tattevin, H Tissot-Dupont, JP Viard, N
  Viget
• DMI2 coordinating centre French Ministry of
  Health (V Salomon), Technical Hospitalisation
  Information Agency, ATIH (N Jacquemet)
• Statistical analysis centre U943 INSERM and UPMC
  (S Abgrall, D Costagliola, S Grabar, M Guiguet, E
  Lanoy, L Lièvre, M Mary-Krause,
  H Selinger-Leneman), INSERM-Transfert (JM
  Lacombe, V Potard)
• Clinical centres
• Paris area Ambroise Paré, Antoine Béclère,
  Avicenne, Bichat-Claude Bernard, Cochin, Henri
  Mondor, HEGP, Jean Verdier, Kremlin Bicêtre,
  Laennec, Lariboisière, Louis Mourier,
  Necker-adultes, Pasteur, Paul Brousse, Pitié
  Salpêtrière, Raymond Poincaré, Rothschild,
  Saint-Antoine, Saint-Denis, Saint-Joseph,
  Saint-Louis, Tenon
• Outside Paris area Aix en Provence, Antibes,
  Arles, Avignon, Belfort, Besançon, Caen,
  Clermont-Ferrand, Digne les Bains, Dijon, Gap,
  Grenoble, Lyon , Marseille, Martigues,
  Montpellier, Mulhouse, Nancy, Nantes, Nice,
  Nîmes, Reims, Rennes, Rouen, Saint-Etienne,
  Strasbourg, Toulon, Toulouse, Tourcoing, Tours
• Overseas Guadeloupe, Guyane, La Réunion,
  Martinique, Saint-Martin