DISCOVERIES FROM THREE LARGE COHORT STUDIES OF PREMATURES

1
DISCOVERIES FROM THREE LARGE COHORT STUDIES OF
PREMATURES

• Nigel Paneth MD MPH
• Michigan State University
• http//www.epi.msu.edu/faculty/paneth.htm
• Neonatology 2006
• Miami, Nov 10th, 2006

2
THE THREE STUDIES

• Neonatal Brain Hemorrhage Study (NBH)
• Developmental Epidemiology Network Study (DEN)
• Extremely Low Gestational Age Newborn Study
  (ELGAN) aka Molecular Antecedents of Brain Damage

3
CHARACTERISTICS OF THE STUDIES
4
FOCUS OF ALL THREE STUDIES

• Assessing brain injury via cranial ultrasound
• Better understanding of the nature of brain
  injury in preterm infants
• Examining antecedents of brain injury
• Examining sequelae of brain injury

5
DIFFERENCES AMONG THE THREE STUDIES

• NBH has the longest duration of follow-up (16
  years)
• DEN paid special attention to placental findings
• ELGAN obtains multiple measurements of proteins
  on study subjects using nanotechnologies

6
TOPICS STUDIED

• IN BOTH NBH AND DEN
• Nature of brain injury
• Thyroid hormones
• Hypocapnia
• Labor and delivery factors
• MgSO4

• ONLY IN ONE STUDY
• Brain injury and late sequelae (NBH)
• Antenatal steroids (DEN)
• Placental findings (DEN)

7
THE NATURE OF BRAIN INJURY
8
Mac Keith Press 1994
9
KEY OBSERVATIONS ABOUT BRAIN INJURY

• There is really no such thing as a Grade IV
  hemorrhage
• Ventricular enlargement is usually a component of
  white matter damage
• Germinal matrix hemorrhage is not restricted to
  the periventricular region
• Cranial ultrasound does not image many forms of
  brain damage e.g. pontosubicular necrosis,
  cerebellar hemorrhage
• Paneth N Classifying Brain Damage J Pediatrics
  1999134527-9

10
CLASSIFYING BRAIN DAMAGEPaneth N J Pediatrics
1999134527-9.

• PAPILE CLASSIFICATION
• Grade I GMH alone
• Grade II uncomplicated GMH/IVH
• Grade III IVH with ventricular enlargement
• Grade IV IVH with parenchymal extension

• NBH/DEN CLASSIFICATION
• No lesion
• GMH or IVH without ventricular enlargement (VE)
• Parenchymal echodensity or lucency or ventricular
  enlargement (greater than mild)
• (DEN prioritized looking at EL)

11
FACTOR ANALYSIS PRODUCES FOUR CLUSTERS OF
HISTOLOGIC ABNORMALITIES DEN - Gilles et al J
Neuropathol Exp Pathol 1998 571026

• NON-HEMORRHAGIC LESIONS
• Small, dense, asymmetrical lesions hypertrophic
  astrocytes, macrophages, coagulative necrosis in
  white and gray matter
• Large, diffuse symmetrical lesions hypertrophic
  astrocytes and amphophilic globules
• HEMORRHAGIC LESIONS
• Intraventricular, subarachnoid and parenchymal
  hemorrhage
• Hemorrhagic necrosis

12
THE IMPORTANCE OF THYROID HORMONES
13
ODDS OF CP IN RELATION TO SEVERE HT BEFORE AND
AFTER STATISTICAL ADJUSTMENTS
14
ODDS RATIOS FOR ECHOLUCENCIES IN LOWEST QUARTILE
OF THYROXINE LEVEL (DEN Leviton et al J Pediatr
1999134706)
15
THESE RESULTS LED TO THE THOP TRIAL

• Phase 1/2 (dosage) trial funded by NINDS,
  conducted in Westchester County, NY Madrid,
  Spain Amsterdam Holland.
• Goal is to find optimum dosage and administration
  method to maintain optimum thyroid hormone and
  TSH pattern
• Six-arm study
• Placebo
• Iodine only
• 4 arms of T4 treatment
• 4 µg vs 8 µg/day
• Bolus vs Continuous
• Aiming for 24 in each group (144 total)
• Enrollment as of 11/8 is 130
• We are planning a submission for Phase 3 trial
  with many centers this spring. If interested,
  please contact Ed LaGamma. (Edmund_lagamma_at_nymc.ed
  u)

16
HYPOCAPNIA
17
FOUR VENTILATORY RISK FACTORS

• Mechanical ventilation (MV)
• Requiring mechanical ventilatory assistance
• Prolonged ventilation (P)
• Duration of ventilation longer than
  expected for GA
• Hypocapnia (C)
• Lowest quintile of cumulative PCO2 levels
• Hyperoxia (O)
• Highest quintile of cumulative PO2 levels

18
ODDS RATIOS FOR DISABLING CP BY VENTILATORY RISK
FACTORSchildren with up to two risk factors
19
ODDS RATIOS FOR DISABLING CP BY VENTILATORY RISK
FACTORSChildren with up to four risk factors
20
HYPOCARBIA IN DEN

• Definition of hypocarbia (hypocapnia) lowest
  quartile of PCO2 for gestational age on day 1
• Adjusted OR for echolucencies 1.7
• Adjustment for propensity score
• But larger adjusted ORs in some lower risk
  subsets
• gt 26 weeks OR 2.6
• Thyroxine not low OR 3.1
• No prenatal antibiotics OR 2.7

21
LABOR AND DELIVERY
22
IN DEN, FETAL VASCULITIS AND MEMBRANE RUPTURE
CONFOUND THE EFFECT OF VAGINAL DELIVERYLeviton
et al AJOG 1999 1811007
23
IN NBH, ACTIVE LABOR, AND NOT VAGINAL DELIVERY,
WAS ASSOCIATED WITH WHITE MATTER DAMAGEQiu et al
AJOG 20031891143

• In NBH, no effect of vaginal delivery, even
  unadjusted
• Active labor (onset of active phase recorded, or
  cervix gt 4 cm)
• GM/IVH Adjusted OR 1.3
• PEL/VE Adjusted OR 2.3
• Neonatal death Adjusted OR 1.8
• Disabling CP Adjusted OR 1.6

24
MAGNESIUM SULFATE
25
MAGNESIUM SULFATELeviton et al Pediatrics
199799/4/e2 and Paneth et al Pediatrics
199799/5/e1

• Leviton et al Adjusted OR
• (adjusted for GA, BW Z-score, antenatal
  steroids, PE/PIH, delivery route, labor)
• Parenchymal echodensity 1.0
• Hypoechoic image 1.2
• Ventriculomegaly (VE) 1.1
• Paneth et al Adjusted OR
• (adjusted for GA, FGR, gender, multiple birth,
  PE/PIH, delivery route, labor, amnionitis,
  hypertension)
• GM/IVH 0.89
• Parenchmal lesions/VE 0.94
• Disabling CP 0.63

26
PREDICTION OF OUTCOME FROM ULTRASONOGRAPHIC
EVIDENCE OF BRAIN INJURY
27
White Matter Damage and Cerebral Palsy
Percent with CP
Pinto-Martin et al. Pediatrics 1995
95249
28
White Matter Damage and Mental Retardation
Percent with MR
Whitaker et al. Pediatrics 1996 98719.
29
White Matter Damage and Attention Deficit
Hyperactivity Disorder
Percent with ADHD
Whitaker et al. Arch Gen Psychiatry 1997
54847
30
PLACENTAL INFLAMMATION
31
DEN WMD AND PLACENTAL VASCULITIS

• Risk of echolucency is increased with fetal
  vasculitis (adjusted OR 10.8) when membranes
  are not ruptured gt 1 hr prior to delivery.
• Membrane inflammation alone was not associated
  with echolucency
• OR adjusted for
• GA, BW Z-score, maternal antibiotic, antenatal
  steroids, fever, white count.
• Leviton et al Ped Research 1999 46566

32
ANTENATAL STEROIDS
33
ANTENATAL STEROIDSDEN Leviton et al AJOG
19991811007

• Overall effects on brain lesions modest, except
  for VE
• Adjusted ORs
• IVH 0.7 (full course) 0.8 (partial course)
• EL 0.7 (full course) 0.8 (partial course)
• VE 0.5 (full course) 0.3 (partial course)

34
SUMMARY OF 5 KEY FINDINGS ABOUT PREMATURE INFANTS
IN DEN AND NBH

• Brain damage is widespread in infants who die,
  and can be diffuse or focal, but white matter is
  the tissue most affected.
• US evidence of white matter injury
  (echodense/echolucent parenchymal lesions and/or
  persistent ventricular enlargement, especially
  with shunts) is the most important determinant of
  long-term outcome.
• Thyroid hormone is the single most predictive
  measure of outcome obtainable on serum in the
  first week of life
• Hypocapnia (PCO2 lt 25) and perhaps hyperoxia (PO2
  gt 60) should be avoided.
• The finding in the placenta most predictive of
  brain injury is fetal vasculitis membrane
  inflammation alone is not associated with brain
  injury

35
SUMMARY OF LESS CERTAIN FINDINGS

• Antenatal steroids may prevent brain damage
• Magnesium sulfate probably does not prevent brain
  injury
• Vaginal delivery predisposes to brain damage, but
  this may be because of its role as a marker of
  placental inflammation

36
RESEARCH AIMS FOR THE FUTURE. MOLECULAR
EPIDEMIOLOGY IN THE ELGAN STUDY

• AIMS OF THE STUDY
• Identify antecedents of cerebral white matter
  damage
• Initiators of the damage promotion process
• Damage promoters
• Identify inflammation modulators and protectors
• Maintain contact with parents for follow-up
• Create specimen bank placenta, umbilical cord,
  blood spots

37
ELGAN STUDY HUBS AND HOSPITALS
38
ELGAN STUDY

• Studied in NBH and DEN
• characteristics of the mother, pregnancy,
  delivery and the newborn from interview and
  medical records
• protocol ultrasonography with duplicate readings
  and consensus resolution
• Studied in DEN only
• placental pathology
• ELGAN records all of the above plus
• organisms identified in the placenta
• placental levels of growth factors
• blood levels of cytokines, hormones, growth
  factors

39
NIH- FUNDED T-32 TRAINING PROGRAM IN PERINATAL
EPIDEMIOLOGY AT MICHIGAN STATE UNIVERSITY

• Accepting applications for 2007-8 until May 2007
• Would love to have a neonatologist in the
  program!
• If interested, email cv to paneth_at_msu.edu