• 
• 

Three Part Question

In [premature babies] do [cranial ultrasonographic findings of periventricular flares/echodensities] predict [poor neurodevelopmental outcome]?

Clinical Scenario

A 28 week preterm boy required intubation after vaginal delivery. He received a dose of surfactant and was mechanically ventilated for less than 24 hours. He was subsequently weaned off continuous positive airway pressure (CPAP) support. Serial cranial ultrasounds showed a persistence of periventricular flares. The parents want to know whether this finding is significant in regards to their child’s development.

Search Strategy

PUBMED:
Searching free text using periventricular flares
Searching free text using periventricular echodensities.
Searching free text using periventricular echodensities.

Searching free text using periventricular flares yielded 9 results
Searching free text using periventricular echodensities (PE) yielded 57 results.
Searching free text using periventricular echodensities yielded 4 results
3 Studies overlapped and were present in two searches hence a total of 67 results.
4 studies were in a foreign language and discounted.

Search Outcome

15 were relevant to the study (including a letter to the editor and a commentary)
Full articles for 12 studies were obtained and analysed.

Relevant Paper(s)

Author, date and country	Patient group	Study type (level of evidence)	Outcomes	Key results	Study Weaknesses
Resch B, Jammernegg A, Perl E, Riccabona M, Maurer U, Muller W. 2006 Austria	72 infants born between 1995 and 2000 , gestation between 25-36 weeks with ultrasound diagnosis of periventricular echodensities (PVE) in NICU of the University Hospital, Graz, Austria	Retrospective study. Infants with cystic PVL, congenital malformation, hereditary disease and PVE only on first day were excluded Cranial ultrasounds were done, on days 1,3,5 then once a week, at discharge and at corrected gestational age of 38-40 weeks using 7.5MHz transducer. Neurodevelopmental assessments were performed at corrected gestational age of 4, 8, 12, 18 and 24 months using Griffiths, then once a year using Kaufman’s.	2 and 4 infants with PVE duration of 7-14 days and > 14 days respectively had cerebral palsy	P<0.002, 95% CI 1.017- 61.661	Not all the children had neurodevelopmental assessment at 2 years of age
Kutschera J, Tomaselli J, Maurer U, Pichler, Schwantzer G, Urlesberger B. 2006 Austria	23 infants below 1500g born in University Hospital of Graz between January 1992 and December 2000 with transient periventricular echodensities (TPE)	Retrospective study:Case control Infants with possible risk factors for developmental delay, small for gestational age,microcephaly, apgar<7 at 5 min, umbilical artery pH<7.15, intraventricular haemorrhages grades III & IV , major surgical interventions and malformations were excluded 26 infants met the criteria however only 23 were included in the study. Parents of 2 excluded did not want to participate and 1 child move away. 23 controls- no TPE, born in the same gestational week and within 12 months Cranial ultrasounds were performed every other day during the first 8 days and weekly from day 8 to discharge Children were examined by a neuropaediatrician and psychologist	Transient periventricular echodensities seems to affect cognitive development and cause minor neurological function	P value calculated using Wilcoxon signed rank test.	Children’s age not corrected for gestation at time of examination.
Pisani F, Leali L, Moretti S,Turco E, Volante E, Bevilacqua G. 2006 Italy	164 pretermss admitted to NICU, University of Parma between January 2001 and December 2002	Retrospective study: Case control Group 1: 78- normal ultrasound Group 2: 50- TPE persisting up to 2 weeks without subsequent ventricular dilatation Group 3: 36 – persistent echodensities corresponding to periventricular leucomalacia Scans done within first 72 hours, then at least 2 scans between the third and 14th postnatal day and weekly until 40weeks postconceptional age using 7MHz. All infants with other forms of sonographic abnormalities and systemic disease were excluded. Developmental outcome was assessed at 44 weeks postconceptional age, 1 month after discharge and at 3,6,9 and 12 months using Griffiths’ mental developmental scale. Developmental quotient (DQ) ≥80 was classified as normal and abnormal if below 80. Neurological assessment was classified as favourable(normal development, mild muscle tone and reflex abnormalities) or adverse(death, cerebral palsy, developmental delay, blindness or deafness)	94% of infants with TPE persisting up to 2 weeks without ventricular dilatation had normal development , 4% developed cerebral palsy and 1 died	No significant difference in neurological outcome in infants with normal or TPE	Mean gestational ages and weights differed for the different groups which may have had an impact on the outcome.
Bos AF, Martijn A, Okken A, Prechtl HFR. 1998 Netherlands	27 preterm infants, gestation 26-34 weeks , Netherlands between September 1992 and May 1994	Prospective study. Infants with periventricular echodensities as well as those with normal ultrasounds were included. Infants were excluded if they had major congenital anomalies, developed cystic periventricular leucomalacia or grade 3 or 4 intracranial haemorrhage. Cranial ultrasounds using 7.5MHz transducer were performed within first week of life after the second postnatal day and weekly thereafter until one or two normal scans were obtained. Additional scans were done at term I hour video tape recording s were made at weekly intervals till term starting within the first week of life. All recordings were made at the end of morning or beginning of afternoon. Periods of crying were excluded. Follow up in outpatient at term and corrected ages of 3, 6, 12, 18 and 24 months. Mental and motor development was tested between 18 and 24 months using Bayley scales of infant development. Developmental quotient >85 was classified as normal, 50-84: minor neurological abnormality and <50 abnormal	Echodensities persisting beyond 14 days were associated with abnormal general movement	Developmental course of movement quality was correlated to neurological outcome	Quality of movements were assessed by 2 observers who were unaware of clinical history however there is potential for inter-observer variation
Lai F F, Tsou KY 1999 Taiwan	69 preterm infants weighing ≤1500g admitted to NICU of National Taiwan University between January 1995 and June 1997	Retrospective:Case control Three groups: Group 1: normal ultrasound(22)- control Group2 : TPE(32) Group 3: grade I – III IVH(15) Cranial ultrasound were done on first, third, seventh and fourteenth days and then if necessary every 2-4 weeks until discharge Developmental assessment was done at 6 months and 1 year of corrected age. This was done by neonatologist using infant neurological international battery. Bayley scales for infant were used. A child was considered to have normal development if score was between 85 and 114, mild developmental delay if between 70 and 84 and significant delay if less than 69.	Infants with TPE had a significantly higher incidence of minor motor and mental developmental delays	P =0.001	Not clear who performed the ultrasound
Jongmans M, Henderson S, De Vries L, Dubowitz L 1993 United Kingdom	106 infants <34 weeks admitted to NICU, Hammersmith Hospital between January 1984 and February 1986	Prospective A minimum of 3 consecutive ultrasound scans at least 24 hours apart in the neonatal period At least one follow up neurodevelopmental evaluation at either 40 weeks postmenstrual age or in infancy at 6,9,12,18 and 24 months No congenital abnormalities Cranial ultrasounds were performed daily during the first week and twice a week thereafter using 5 or 7.5MHz. Formal assessments of neurological (Touwen’s examination), cognitive functions (British ability scales) and perceptuomotor (movement assessment battery for children). The battery test had two components:movement ABC test and checklist to be completed by the teacher.	Motor assessment showed that performance decreased significantly with duration of flares	P<0.001 for gross motor function
Ringelberg J, Van de Bor M 1993 The Netherlands	24 infants<32 weeks admitted to NICU, University Hospital , Leiden, The Netherlands from 1987-1990	Prospective Routine cranial ultrasound done: daily in the first week then twice a week thereafter till discharge. Neurodevelopmental outcome of children with TPE and periventricular leucomalacia was compared with the outcome of controls. Neurodevelopmental assessment (Gesell test) was done at 2 years corrected age. This was performed by developmental paediatrician blinded to ultrasound findings. Developmental quotient (DQ) > 90 was classified as normal, 80-90: minor handicap and less than 80 major handicap	Children with transient periventricular echodensities(TPE) had significantly more minor handicaps than children without	Χ2 6.0; p,0.05 for TPE vs control	Unclear who performed the cranial ultrasound Handicap not clarified
Bennet FC, Silver G, Leung EJ, Mack LA 1990 United States of America	48 infants 25-35 weeks born between January 1 1983 and June 30 1984	Prospective Newborns with suspected intrauterine infection or multiple congenital anomalies were excluded. Cranial ultrasounds were done if neonate was ≤1500g, gestational age≤32weeks or clinical symptomatology suggesting intracranial pathology. Scans were done with 7.5MHz transducer, in the first week of life, second week of life and at least one additional time beyond 2 weeks of age. Hence all subjects had at least 3 scans. All children were observed in an interdisciplinary neurodevelopmental follow up programme to a mean age of 18 months (12-24 months. Neurodevelopmental outcome was assessed by means of neurological examination and the Bayley scales of infant development. Outcome was classified as normal, minor or major abnormalities	Wide range of outcomes after detection of periventricular echodensities	Χ2 and two tailed t test	Arbitrary categorization- flares were graded and outcome were divided into minor and major with varied disabilities
Appleton RE, Lee REJ, Hey EN. 1990 United Kingdom	15 infants with TPE between 1984 and 1988 in Princess May Maternity Hospital, Newcastle Upon Tyne	Retrospective. Babies with germinal matrix or intraventricular haemorrhage were excluded Sans were done in all babies less than 34 weeks gestation, any infant requiring ventilator support irrespective of gestation and any infant with abnormal behaviour or neurological signs or both. Scans were done with HP using 5MHz probe. Scans were interpreted by a paediatric radiologist Infants were scanned within 24 hours of birth and thereafter twice or thrice weekly till discharge or transferred to another hospital. Review of clinical records and follow up (neurological examination and Denver development assessment corrected for gestational age) were undertaken by a single observer.	4 infants had neurological abnormalities including spastic diplegia
Pidcock FS, Graziani LJ, Stanley C, Mitchell G, Merton D 1990 United States of America	127 preterm <33 weeks admitted to NICU, Thomas Jefferson Hospital, USA between January 1982 and December 1986 with periventriclar echodensities (PVE)	Prospective Cranial ultrasound was obtained within 3 days of life then weekly until 1 month then at least 2-4 weeks until discharge using 5 or 7.5MHz transducer.Scans were reviewed by the authors who had no knowledge of clinical course. Neurologic examinations performed at 3 to 6 months intervals until 18-24months	Mild and moderate to severe PVE without cyst formation were not associated with spastic cerebral palsy	negative predictive values for cerebral palsy of 69% and 76% respectively
Guzetta F, Shackelford GD, Volpe S, Perlman JM, Volpe JJ 1986 United States of America	75 infants with weight ,2kg delivered at St Louis Children’s Hospital between August 1980 and March 1986	Retrospective Cranial ultrasounds were performed using a 5MHz transducer. All infants were scanned on the first or second postnatal day and approximately 3 times weekly. Neurologic evaluation was performed by ?standard techniques by two of the researchers. Cognitive function was evaluated using the Bayley scales, the Stanford binet test or Wechsler preschool and primary scale of intelligence.	31 survived, 22 of whom were neurologically evaluated		This study focuses on relationship between intraparenchymal echodensities and intraventricular haemorrhages: of 75 studied, only 3 were not associated with IVH
De Vries LS, Regev R, Pennock JM, Wigglesworth JS, Dubowitz LMS 1988 United Kingdom	59 babies with periventricular densities not evolving into extensive cystic lesion and 107 infants with normal scan admitted to Hammersmith Hospital ≤ 34 weeks between January 1984 and January 1986	Prospective Babies were scanned with 3-5-7.5MHz transducers, 7.5MHz transducer was preferred. Scans were performed daily in the first week, 2 weekly until discharge and in clinic if fontanelle remained open. Infants were followed up at 40 weeks postmenstrual age, 6, 9, 12, 18 and 24 months chronological age and assessed with Griffiths Mental Developmental scale.. Developmental quotient (DQ) uncorrected for prematurity was used and they were grouped in 4 categories: A: no abnormal neurological signs, DQ>80 B: some dystonia on at least 2 separate occasions C: definite neurological handicap, DQ>70 D: severe cerebral palsy or mental retardation with DQ<70	Echodensities persisting for >10 days seemed to be related to high incidence of dystonia. Most prognostic factor was densities in the trigone	P<0.001

Comment(s)

Cranial ultrasound scans are routinely done on preterm babies and neonatologists are more often than not asked to comment on their findings and possible implications. Some findings are obvious, however flares/increased echogenecity are commonly reported but their significance is controversial. A flare is diagnosed when the affected parenchyma appears as bright as the choroid. The appearances need to be considered and contrasted with normal periventricular blush which occurs in the peritrigonal region of the brain. The appearance should be confirmed in two planes. In general it is considered that a flare appears 24-48hours after an insult; once seen it can persist from 2 days to more than 14 days. There are various nomenclatures: periventricular echodensities, periventricular densities, transient periventricular echodensities, intracerebral echodensities and intraparenchymal echodensities with variation in the definition of ‘transient’ with respect to the periventricular echodensity or ‘flare’. Some studies graded echodensities, however this is still subjective. 12 studies were reviewed. None of the studies were randomised control trials; 5 were prospective and 7 of the studies had controls. The study numbers however were small (15-164) These studies have demonstrated a correlation between persistence of periventricular echodensities beyond 7 days and subsequent development of neurological/motor abnormalities which could be mild or moderate. Resch et al showed that 6/25 (24%) of infants with duration of periventricular echodensities (PVE) of 7-14 days had adverse neurological outcome compared to 6/14(43%) with PVE duration greater than 14 days. The duration and not the grading correlated with abnormal neurodevelopmental outcome irrespective of gestational age. Kutschera et al also identified poor cognitive development in children age between 3 and 11years who had transient periventricular echodensities in the neonatal period. Bos et al demonstrated that echodensities in the parieto-occipital white matter persisting more than 14 days affected the quality of movement which identified infants at risk of developmental problems. They have also noted that duration as well as localization of the echodensities was important. Jongmans et al looked at the impact of neonatal flares in preterms on neurological status and motor competence at 6 years of age. Their results of motor assessments showed that performance decreased significantly with increasing duration of flares. This trend was stronger in measures assessing lower limbs function than upper limbs .Their study showed no effects on cognitive abilities. Ringelberg et al’s prospective study also showed more minor motor handicaps when compared to controls. Bennet et al showed a wide range of neurodevelopmental outcomes and have stated that long term neurodevelopmental prognostication remain complex and imperfect. Appleton et al have shown that transient intracerebral echodensities are not always benign and infants should be carefully followed up with guarded prognosis. They have suggested that these flares may represent mild leucomalacia. Guzetta et al demonstrated that with extensive intraparenchymal echodensities (IPE) there is little or no chance for survival with normal neurological and cognitive outcome however with localized IPE major deficits are common and an appreciable proportion of infants have cognitive function in normal range. De Vries et al showed that the persistence of the densities for more than 10 days and the presence of densities in the trigone were especially related with subsequent problems. Postmortem findings in 2 infants and MRI studies in 6 infants suggested that periventricular densities represent the milder end of the spectrum of periventricular leucomalacia. These studies have therefore demonstrated that echodensities are not always benign. Magnetic Resonance Imaging (MRI) Increased echogenecity on ultrasound may be due to venous congestion, microscopic haemorrhage, necrosis or simply immaturity, thus the spectrum of white matter changes and subsequent neurological development is wide. Magnetic resonance imaging on preterm infants have demonstrated a range of lesions which are not easily detected using ultrasound but are associated with abnormal neurodevelopmental sequalae.

Clinical Bottom Line

Persistence of periventricular flares/echodensities rather than flares per se is associated with abnormal neurodevelopmental outcome. However, ultrasound diagnosis of flares can be subjective and may miss more subtle yet clinically significant lesions; in this respect MRI may be superior in identifying infants at risk of white matter injury.

References

1. Resch B, Jammernegg A, Perl E, Riccabona M, Maurer U, Muller W. Correlation of grading and duration of periventricular echodensities with neurodevelopmental outcome in preterm infants Pediatr Radiol. 2006;-36:-810-815
2. Kutschera J, Tomaselli J, Maurer U, Pichler, Schwantzer G, Urlesberger B. Minor neurological dysfunction, cognitive development and somatic development at the age of 3 to 11 years in very low birth weight infants with transient periventricular echodensities Acta Paediatrica 2006;-95:-1577-1581.
3. Pisani F, Leali L, Moretti S,Turco E, Volante E, Bevilacqua G. Transient periventricular echodensities in preterms and neurodevelopmental Outcome. J Child Neurol 2006; - 21:-230-235
4. Bos AF, Martijn A, Okken A, Prechtl HFR. Quality of general movements in preterm infants with transient periventricular echodensities. Acta Paediatr 1998;-87:-328-35
5. Lai F F, Tsou KY Transient periventricular echodensities and developmental outcome in preterm Infants. Pediatr Neurol 1999; - 21:-797-80.
6. Jongmans M, Henderson S, De Vries L, Dubowitz L Duration of periventricular densities in preterm infants and neurological outcome at 6 years of age. Arch Dis Child 1993;- 69:- 9-13
7. Ringelberg J, Van de Bor M Outcome of transient periventricular echodensities in preterm infants. Neuropaediatrics 1993 ;- 24:- 269-73
8. Bennet FC, Silver G, Leung EJ, Mack LA Periventricular echodensities detected by cranial ultrasonography: Usefulness in predicting neurodevelopmental outcome in low birthweight preterm infants. Pediatrics 1990 ;-85:-400-4
9. Appleton RE, Lee REJ, Hey EN. Neurodevelopmental outcome of transient intracerebral echodensities Arch Dis Child 1990;- 65:-27-9
10. Pidcock FS, Graziani LJ, Stanley C, Mitchell G, Merton D Neurosonographic features of periventricular echodensities associated with cerebral palsy in preterm infants. J Pediatr 1990;- 116:-417-22
11. Guzetta F, Shackelford GD, Volpe S, Perlman JM, Volpe JJ Intraparenchymal echodensities in the premature newborn: Critical determinant of neurologic outcome. Pediatrics 1986; - 78:- 995-1006
12. De Vries LS, Regev R, Pennock JM, Wigglesworth JS, Dubowitz LMS Ultrasound evolution and later outcome of infants with periventricular densities Early Hum Dev 1988; - 16:- 225-33