Antiretroviral Drug Toxicity in the Context of Prevention of Mother to Child Transmission of HIV in

1
Antiretroviral Drug Toxicity in the Context of
Prevention of Mother to Child Transmission of
HIV in the Era of PEPFAR and WHO 3x5
Lynne M. Mofenson, M.D. Pediatric, Adolescent and
Maternal AIDS Branch National Institute of Child
Health and Human Development National Institutes
of Health Department of Health and Human Services
2
Background Safety of ARVs in Pregnant Women and
Their Infants

• Several short-course ARV regimens, including AZT,
  AZT/3TC and single-dose NVP, alone or in
  combination, have been proven effective and safe
  for PMTCT.
• With PEPFAR and WHO 3x5, there is increasing
  HAART use in resource-poor settings for treatment
  of childbearing-aged women who may become
  pregnant, as well as by pregnant women needing
  treatment.
• For women who dont require ARV treatment for
  their own health, use of HAART for PMTCT is also
  increasingly being considered in resource-poor
  settings.

3
Background Safety of ARVs in Pregnant Women and
Their Infants

• Safety information on HAART use for pregnant
  women and their infants in resource-poor settings
  is limited to date most data is extrapolated
  from information from resource-rich countries.
• Risks of ARV exposure to the infant varies by
  timing and duration of exposure, and type/number
  of drugs is exposed to.
• The longest (and most complex) regimens have the
  greatest efficacy for PMTCT but also the greatest
  exposure (and hence potential risk).

4
Overview What is New on ARV Safety?

• Issues related to ARV use in non-pregnant women
  of childbearing potential
• Teratogenicity, developmental toxicity
• Issues related to ARV use in pregnant women
• ARV PK in pregnancy
• NVP toxicity
• Pregnancy complications/outcome
• Issues related to fetal/infant safety
• Mitochondrial toxicity
• Issues related to use of ARVs for prevention of
  breastfeeding transmission

5
ARV Use in Women of Childbearing Age ARVs and
Congenital Abnormalities/Developmental Toxicity
6
Antiretroviral Pregnancy Registry (APR) Update

• AZT and 3TC have sufficient prospective data in
  the APR to rule out a 1.5-fold or greater
  increase in overall birth defects with 1st
  trimester exposures.
• ABC, ddI, d4T, NFV, and NVP have sufficient
  prospective data in the APR to rule out 2-fold or
  greater increase in common birth defects with 1st
  trimester exposure.
• However, higher rates in ddI than others (no
  specific pattern), continue to monitor.
• Other drugs insufficient data to date to draw
  conclusions.
• In retrospective reports, only EFV has concern
  related to specific birth defect.

7
Antiretroviral Pregnancy Registry
(http//www.APRegistry.com) Interim Report -
Prospective Cases 1/89-1/05
8
Potential for Teratogenicity with Efavirenz

• EFV is teratogenic in primates at drug exposures
  equivalent to humans receiving therapeutic doses
  3/20 monkeys with anencephaly, an- or
  micro-ophthalmia.
• Human data
• Prospective reports from APR, 5 defects/206
  pregnancies (2.4), no pattern.
• However, retrospective APR reports 4 cases
  human infants with significant CNS defects with
  1st trimester exposure (meningomyelocoele,
  Dandy-Walker).
• In 2005, US FDA changed Pregnancy Classification
  to FDA Pregnancy Class D (positive evidence of
  human fetal risk).

9
Tenofovir and Developmental Bone Toxicity

• Fetus (Tarantal. JAIDS 199920323-33) Bone
  marker/ density changes in primate fetuses with
  in utero exposure 25x gt than human exposure
• No gross congenital abnormalities.
• Significantly lower fetal insulin-like growth
  factor-1 (regulator linear growth), high BP-3.
• Reduction fetal bone porosity.
• Chronic TFV in immature animals of multiple
  species results in reversible bone changes is
  dose-, exposure-, age-, species-specific.
• Unknown effect with human fetal exposure 1
  defect (renal)/96 1st trimester exposure (APR).
• Ped Rx (Gafni. Pediatr Res 200455329A, abs
  1873) ARV-experienced children 6-16 yrs
  baseline BMD lt normal, decrease with 24 wks TFV.

10
PIs and Infant Hyperbilirubinemia

• Indinavir, Atazanavir
• Associated with indirect hyperbilirubin-emia.
• Whether might be a potential effect on neonatal
  hyperbilirubinemia is unclear.
• Primates given IDV 3rd trimester pregnancy
• Poor transplacental transfer with in utero
  exposure, fetal levels were only 1-2 maternal
  plasma levels at 1 hour post maternal dosing.
• No exacerbation physiologic hyperbilirubinemia
  seen in infant monkeys with only in utero
  exposure.

11
Summary Women of Childbearing Age

• Contraception is critical part of care of
  HIV-infected women potential for 1st trimester
  exposure before pregnancy recognized.
• Concern related to birth defects with 1st
  trimester exposure primarily relate to EFV.
• If effective contraception not available, avoid
  use in women of childbearing age.
• If become pregnant on EFV and recognize in 1st
  trimester, change to NVP (?) or PI.
• Time-limited use 2nd/3rd trimester (eg, TB).
• Effect of TFV on fetal bone development is
  unclear however, if woman needs therapy, benefit
  likely exceeds theoretical risk.
• If woman on IDV or ATV, closer monitoring of
  infant bilirubin might be considered.

12
ARV Use in Pregnant WomenARV Pharmacokinetics
in Pregnancy
13
Antiretroviral PK in Pregnancy

• NRTI/NNRTI studied need no dose adjustment.
• Problems maintaining therapeutic levels with
  unboosted PIs late pregnancy (hepatic metabolism
  change?), potential for resistance.
• SQV RTV boost results in adequate levels.
• IDV RTV boost may give adequate levels (Tubiana
  12th CROI 2005 abs 810).
• LPV/RTV levels low 3rd trimester.
• NFV dosing conflicting data
• PACTG 353 (N14) Kosel et al AIDS 2003 (N9)
  1250 bid okay levels, but variable.
• Van Heeswijk Clin Pharm Ther 2004 (N11) Neelen
  CID 2004 (N27) 1250 bid lower levels pregnancy,
  more subtherapeutic.

14
LPV/RTV PK in Pregnant WomenStek A et al.
Bangkok AIDS Conf 2004 (Abs. LBOrB08)

• Lower systemic LPV exposure in 3rd trimester with
  standard 400/100 bid dosing, 17 women
• Mean AUC half and mean trough a third that of
  non-pregnant pts.
• 82 were lt10ile AUC and all were lt50ile for
  non-pregnant adults.
• By 6-12 wks PP, AUC and trough increased.
• Now studying LPV/r 533/133 BID in 3rd trimester.
  12 women studied
• Median AUC similar to that in non-pregnant
  adults.
• Fewer had lt10ile AUC (25).
• Clinical relevance of transient decrease in LPV
  levels for lt12 weeks unclear does it warrant the
  difficulty of temporarily changing dose?

15
ARV Use in Pregnant WomenNevirapine Toxicity
16
Hepatic Toxicity with Chronic NVP Therapy

• FDA label language on hepatic toxicity 2005
• 12-fold higher risk early symptomatic hepatitis
  in treatment-naïve women with CD4 gt250 when
  initiating NVP.
• Hepatic injury may progress after stop NVP.
• US guidelines In women with CD4 gt250 or men
  with CD4 gt400, NVP should not be initiated unless
  benefit clearly outweighs risk.
• Unknowns
• Risk for women with CD4 250-350?
• Risk if switch to NVP post immune re-constitution
  (cases were naïve patients)?
• Risk in women in resource-poor settings?

17
Symptomatic Hepatic Events in 1st 6 Weeks of NVP
Therapy by Baseline CD4 Count and Gender
(Boehringer-Ingelheim)
Baseline CD4
18
Fatal Acute Hepatic Events with NVP in Controlled
Trials by Baseline CD4 Count and Gender
(Boehringer-Ingelheim)
Baseline CD4
19
NVP-Associated Hepatotoxicity in Non-Pregnant
Women S. Africa FTC vs 3TC StudySanne I et al.
JID 2005191825-9

• Blinded study 3TC vs FTCd4TNNRTI (NVP if RNA
  lt105) in 468 pts with CD4 gt200 60 women, 78
  black, mean CD4 398.
• 82 (385/468) received NVP, others EFV.
• Overall rate gtgrade 3 LFT 17 NVP, 0 EFV.
• Onset within first 12 weeks in 53/66 (80).
• 2 deaths due to hepatic failure in women (0.7)
  both black, non-pregnant, normal baseline LFT, no
  HBV/HCV, occurred at wk 4/5 of therapy, died
  despite NVP withdrawal.

20
Risk Factors for Early NVP HepatotoxicitySouth
Africa FTC vs 3TC StudySanne I et al. JID
2005191825-9

• Risk factors for hepatic toxicity female sex,
  body mass index lt18.5, low serum albumin, MCV
  gt85, RNA lt20,000, LDH lt164, entry AST lt75.
• Frequency 20.1 women, 12.8 in men.

21
NVP and NFV Side Effects More Common in Pregnant
than Non-Pregnant Women?Timmermans S et al.
AIDS 200519795-9
Compared toxicity in 186 pregnant and 186
non- pregnant Dutch HIV-infected women from 15
centers.
22
NVP Toxicity Pregnant Women, U.S. P1022Hitti J
et al. JAIDS 200436772-6

• Compared AZT/3TC NVP vs NFV in ARV-naïve
  pregnant women stopped early.
• Treatment limiting toxicity seen in 1/21 (5) NFV
  women and 5/17 (29) NVP women, including 1
  hepatic death (14).
• NVP arm 53 black, 41 hispanic.
• Among women with CD4 gt250, 0/14 receiving NFV had
  toxicity vs 5/14 (36) receiving NVP.
• NVP toxicity included SJS (1), increased ALT (2/3
  symptomatic), fulminant hepatic failure and death
  despite stopping drug (1).
• All normal ALT entry and no HBV/HCV.

23
NVP, Pregnant Women, Resource-Poor Settings

• Kisumu, Kenya (HAART for PMTCT pilot) N24,
  Thomas et al. 2005 CROI, abs 809
• Gr 3/4 hepatic toxicity 3 no CD4 difference. No
  deaths.
• Thailand (HAART for Rx and PMTCT) N250,
  Phanuphak et al. 2005 CROI, abs 21
• Hepatic toxicity 4 in pregnant vs 2 in
  non-pregnant women and 1 in men slight CD4
  difference (CD4 lt vs gt250). No deaths.
• Mozambique (HAART for Rx and PMTCT) N606,
  Palombi et al. 2005 CROI, abs 67
• Hepatic toxicity 6.3 no CD4 difference. No
  deaths.
• Malawi (HAART, CD4 gt200 for PMTCT trial with
  intensive lab monitoring) N34, CDC meeting 4/05,
  early interim analysis
• Possible hepatic SAE 11.8. No deaths.

24
ARV Use in Pregnant WomenPregnancy
Complications and Pregnancy Outcome with HAART
25
Impaired Glucose Tolerance/Gestational Diabetes
and ARV Use in Pregnancy

• Several studies suggest may be increase in
  glucose intolerance in pregnant women on PIs but
  extent unclear gestational DM in general
  population 2-3 .
• Chmait R (J Perinatol 200222370) US, N64
  Retro-spective, 30 impaired GTT on PI vs 13
  AZT.
• Watts H (Am J Obstet Gynecol 2004190506) US,
  N1,407 Trial (P316) 5 rate of gestational DM
  if got PI therapy from early pregnancy.
• Tuomala R (JAIDS 200538449-73) US, N2,543
  Pro-spective, 3.5x gtgestational DM if late PI
  use.
• El Beitune (Diab Metab Res Rev 2005online 7/05)
  Brazil, N57 Prospective, higher glycemia AUC
  in women on AZT/3TC/NFV vs AZT/no ARV.

26
ARV Use in Pregnancy and Pre-Eclampsia

• Some studies suggest lower rate pre-eclampsia
  with HIV infection, with increase in women on
  HAART, but extent unclear.
• Wimalasundera (Lancet 20023601152-4) UK, N214
  
• Preeclampsia 4 HIV vs 6 in HIV-.
• HIV no/1-2 ARVs 1 with HAART 11.
• Frank (Obstet Gynecol 2004104238-42) S Africa,
  N704
• Preeclampsia 6 HIV (no ARV) vs 5 HIV-.
• Coll (11th CROI 2004 Abs 921) Spain, N472
  retro, N82 prosp
• Retrospective (85-03) increase pre-eclampsia
  gt2001 7 HAART vs 2 all HIV.
• Prospective Pre-eclampsia 11 in HIV (most
  HAART) vs 3 HIV-.
• Associated duration HIV, duration HAART.

27
Adverse Pregnancy Outcome and HAART

• Controversial if HAART associated with pre-term
  delivery data differ US and Europe.
• Lorenzi (AIDS 199812F241-7) Swiss, N30
• 33 preterm in women on combo ARV.
• ECS (AIDS 2000142913) Europe, N2,819
• OR 2.6 preterm if combo PI, 1.8 if combo no PI vs
  no ARV no increase with 1 ARV.
• Risk highest if combo ARV started pre-pregnancy
  than if started 2nd/3rd trimester.
• Tuomala (NEJM 20023461863) US, N1,143
• Tuomala (JAIDS 200538449-73) US, N2,543
• No increase preterm with PI or combo.

28
Increase in Prematurity (GA lt37 Weeks) in
HIV-Infected Women Over TimeEuropean
Collaborative Study. AIDS 2004182337-9
1 ARV (N704) 16.8 2 ARVs (N254)
13.4 HAART (N1,075) 25.5
1985- 1989
1990- 1994
1995- 1999
2000- 2004
29
Pregnancy Outcome and Maternal ARV Use, Latin
American (NISDI)Szyld E et al. 12th CROI,
Boston MA 2005 (abs 806)

• 516 pregnant women followed in NISDI study
  9/02-6/04 with singleton infant and known ARV
  use

Difference not statistically significant Grp 2
vs Grp 1
30
Summary I ARV Pharmacokinetics in Pregnancy and
Pregnancy Complications

• PI levels during pregnancy may be low, use of RTV
  boosting recommended.
• LPV/RTV may need increased dosing in third
  trimester.
• PI use may predispose to increase in
  pregnancy-associated hyperglycemia.
• Unclear association with pre-eclampsia.
• Unclear association of combination ARV use and
  adverse pregnancy outcome further data needed,
  particularly in resource-limited settings.

31
Summary II Chronic NVP Toxicity in Pregnant
Women

• NVP toxicity is more common in women with higher
  CD4 but threshold not clear.
• Risk in women with CD4 250-350 important to
  define, as ARV initiation considered when CD4
  lt350.
• NVP toxicity may be more common in pregnant women
  but need more data.
• Data from resource-poor settings has generally
  lower rates hepatic toxicity (but lower rate of
  monitoring?).
• Use of NVP HAART solely for PMTCT (CD4 gt350) of
  concern cautious use for treatment of women with
  CD4 250-350 may be okay.

32
Fetal/Infant SafetyMitochondrial Toxicityin
Uninfected but ARV-Exposed Infants
33
Possible Mitochondrial Dysfunction and Perinatal
Exposure to Nucleoside Analogues

• Blanche. Lancet 19993541084-9 Barrett. AIDS
  2003171769-85 French Perinatal Cohort Study
  Grp. Lancet 2002359583-4
• French Perinatal Cohort has reported 12 cases (2
  deaths) mitochondrial dysfunction in cohort of
  2,644 uninfected children with ARV exposure.
• Primarily neurologic symptoms
• May have hyperlactatemia
• Abnormalities respiratory chain function
• 18 mo incidence 0.26 (95 CI, 0.10-0.54).
• 18 mo mortality 0.07 (2 of 2,644).
• Also reported elevated risk of first febrile
  seizure in uninfected ARV-exposed children.

34
Clinically Significant Mitochondrial Disease with
ARV Exposure, Other Cohorts

• Perinatal Safety Review Work Grp. JAIDS
  200025261-8
• US Deaths,1985-99, in 16,313 uninfected children
  lt5 yrs gt50 ARV-exposed 30 deaths, none with
  mitochondrial symptoms.
• Bulterys M et al. Ann NY Aca Sci 2000918212-21
• US, PACTS 1,954 uninfected children, 35
  ARV-exposed 3 with possible mitochondrial
  symptoms (develop delay/sz, transient sx lt12 mos)
  but none of the 3 exposed to ARVs.
• European Collab Study. JAIDS 200332380-7
• Europe 2,414 uninfected children, 42
  ARV-exposed ARV exposure not associated with
  febrile seizures, moderate/severe clinical
  symptoms, malignancy or mortality.

35
Possible Effect of Perinatal Exposure to
Antiretroviral Drugs on Hematopoiesis

• Clinically asymptomatic heme abnormalities seen
  in uninfected children with in utero ARV exposure
  vs no exposure in several cohorts.
• Le Chenadec (AIDS 2003172053-61) France
• ANC, lymphs, platelets slightly but signif lower
  through 18 mo in ARV-exposed combogt mono ART
  exposure.
• ECS, Bunders (AIDS 2004182009-17
  2005191071-9) Europe.
• ANC, TLC, and CD8 slightly but signif lower
  through 8 yr in ARV-exposed (blackgtwhite)
  combogtmono ART exposure.
• Pacheco (Bangkok AIDS Conf 2004) US
• TLC, CD4/8 slightly but signif lower to 24 mo in
  ARV-exposed ANC normalize by 6 mo/o.

36
Elevated Lactic Acid in Uninfected Infants with
Perinatal Antiretroviral Exposure

• Transient hyperlactatemia may occur with ARV
  exposure (normal infant has mild lactate
  elevation at birth, rarely gt5 mmol/L, normalizes
  within days and reaches adult levels by 1 wk).
• Giaquinto (AIDS 2001151074-5) Italy, N20
• 85 ARV-exposed infants had gt1 lactate gt2.5
  mmol/L, returned to normal no clinical signs.
• Alimenti (Pediatr Infect Dis 200322782-8)
  Canada, N38
• 92 ARV-exposed infants had abnormal lactate 94
  normal by 28 wks. 10 (26) gt5 mmol/L 2 had
  transient symptoms, resolved.
• (Pediatrics 2004114e598-603) US, N127
• 50 ARV-exposed infants had gt1 elevated lactate
  3 had transient neurologic symptoms.

37
Conclusions Mitochondrial Dysfunction and ARV
Exposure

• In utero ARV exposure may rarely be associated
  with clinically symptomatic symptoms of
  mitochondrial dysfunction (0.3 - 3 per 1,000),
  that can very rarely (0.07 - 7 per 10,000) be
  fatal.
• Mild, clinically asymptomatic, but persistent,
  hematologic abnormalities may be associated with
  in utero ARV exposure.
• Risk more with combogtmono exposure.
• Transient elevations in lactic acid may be common
  with in utero exposure but usually asymptomatic
  and resolves within months.

38
Breastfeeding Issues
39
Issues Related to Maternal Combination ARV When
Used Solely for BF PMTCT

• Choice of regimen problematic NVP toxicity in
  women with higher CD4, and expense/cold chain
  requirement with PI makes choice difficult.
• Penetration of drugs into breast milk
• If penetrate but in subtherapeutic levels?
• If one penetrates but others do not?
• May end up with resistant virus in milk (eg, NVP
  resistance higher in milk than plasma).
• Infant exposure Shapiro R et al. IDSA, Boston
  2004 (Abs. LB-1) Found 3TC, NVP levels in
  breastfeeding infants of moms on HAART higher
  than expected NVP 971 ng/mL, 3TC 28 ng/mL.
• Infant exposure gives potential protection but
  also exposes to potential toxicity.

40
Transfer of ARVs into Breast Milk
41
Conclusions

• Short course ARV regimens for PMTCT have been
  shown to be effective and safe.
• Benefits of HAART clearly outweigh risks for
  woman who needs ART for own health.
• HAART is used solely for PMTCT in women with
  detectable virus in resource-rich countries, but
  intensive monitoring is possible.
• Increasing HAART use in resource-poor countries
  requires increased surveillance for adverse
  effects on women/infants
• Issues of concern include birth defects, drug
  (especially NVP) safety, complications of
  pregnancy, preterm delivery, mitochondrial
  dysfunction.