Article Text
PDF
Abstract
We discuss possible aetiological factors, MRI evolution of injury and neuro-developmental outcomes of neonatal encephalopathy (NE). Thirty-six consecutive infants diagnosed with NE were included. In this cohort, four infants (11%) were identified with injury predominantly in the deep white matter on MRI who were significantly of younger gestation, lower birthweight with higher Apgars at one and five minutes compared to controls. Placental high grade villitis of unknown aetiology (VUA) was identified in all four of these infants. Our hypothesis states VUA may induce white matter injury by causing a local inflammatory response and/or oxidative stress during the perinatal period. We underline the importance of continued close and systematic evaluation of all cases of NE, including examination of the placenta, in order to come to a better understanding of the clinical presentation, the patterns of brain injury and the underlying pathophysiological processes.
Statistics from Altmetric.com
Introduction
White matter injury (WMI) is well recognised as a characteristic pattern of brain injury in premature infants and is increasingly being reported in term infants.1 Li et al reported on focal non-cystic WMI in term newborns with neonatal encephalopathy (NE) and concluded that WMI in term newborns could be acquired near birth. We report four further newborns with NE who were identified with focal non-cystic WMI on early MRI, and discuss possible aetiological factors, MRI evolution of the injury and neurodevelopmental outcomes.
Methods
Our institution is a tertiary-level centre for maternity and neonatal care. All infants were ≥36 weeks gestation and presented with NE within 72 h of birth. NE was defined as abnormalities in two of the following: level of consciousness, respiration, feeding and tone/reflexes. Infants also had one of the following: (1) evidence of fetal distress on fetal monitoring or meconium staining, (2) delayed onset of respiration, (3) Apgar score ≤7 at 5 min and (4) metabolic acidosis with a cord pH ≤7.15 or a base excess of ≥12. Infants with congenital malformations/infections or genetic abnormalities were excluded. Detailed neurological examinations were performed on all infants and NE was graded according to the Sarnat scoring system.2 A standard Bayley neuropsychological assessment was performed at 2 years of age. Placentae underwent detailed histopathological examination and findings were reported in a systematic fashion.3 Villitis of unknown aetiology (VUA) was defined as the presence of lymphocytes and/or histiocytes in villi, graded 1–4, according to the criteria of Knox and Fox.4 Cases in which plasma cells were identified were not included in the diagnosis of VUA, as plasma cells are often seen in cases of villitis associated with infection. Grades 1–2 villitis were considered low grade (figure 1A) and grades 3–4 villitis were considered high grade (figure 1B).
(A) Low grade villitis with adjacent uninvolved villi (×20). (B) High grade villitis with involvement of greater than half a low-power field (×10). (C) Diffusion weighted imaging (DWI) from case 2 demonstrating multiple areas of patchy restricted diffusion bilaterally. This finding was demonstrated in all four cases on early MRI. (D) Apparent diffusion coefficient (ADC) maps of case 2 confirm the DWI as acute ischaemia or infarction. This finding also demonstrated in all four cases. (E) Decreased T2 signal in the areas of white matter injury (WMI) demonstrated in case 2 and found in all four cases. (F) Increased T1 signal in the areas of WMI demonstrated in case 2 and found in all four cases. (G) Gradient-echo imaging suggests that the diffusion abnormality does not represent haemorrhage in the areas of DWI. The areas of haemorrhage identified were not abnormal on diffusion imaging. (H) T2 Flair Sequence demonstrates angular configuration of the lateral ventricles on 2 year MRI in case 2. (I) T2 Flair Sequence demonstrates deep white matter increased signal on 2 year MRI in case 2. (J) T1 Flair Sequence shows an angular appearance of the lateral ventricles and significant white matter volume loss bilaterally on 2 year MRI in case 2. The combination of (F), (G) and (H) are features routinely seen in patients with periventricular leukomalacia.
MRI was performed on a General Electric (United Kingdom) 1.5 Tesla system. Sequences employed included axial diffusion weighted imaging, sagittal Spin Echo T1 weighted and axial Fast Spin Echo T2 weighted. Apparent diffusion coefficients maps were computed. 2D Gradient-echo sequence was performed on three infants. WMI was defined as areas of restricted diffusion, abnormal T1 hyperintensity and T2 hypointensity in the white matter. The severity was scored as minimal (≤3 lesions of ≤2 mm), moderate (>3 lesions or lesions >2 mm but involving <5% of the hemisphere) or severe (≥5% of the hemisphere), according to the method previously reported.1
Statistical analysis was performed using SPSS V.18 with a significance defined as p=0.05. Measurements (from a Gaussian population) were compared using a two-tailed t test, and rank measurements (from a non-Gaussian population) were compared using Mann–Whitney test. Categorical variables were reported as percentages. The study was approved by the Institution's Ethics Board.
Results
Over a 2 year period, a total of 15 695 neonates were born. Thirty-six consecutive infants diagnosed with NE were included. Of this cohort, four infants (11%) were identified with abnormality predominantly in the deep white matter on MRI. Table 1 outlines the clinical characteristics of these four infants compared with the remainder of the infants with NE (n=32). Table 2 reports on the MRI findings and long term outcomes. MRI images are shown in figure 1C–J.
Patient characteristics of infants with NE associated with or without specific WMI
Patient early and late MRI findings and neurodevelopmental outcomes
Discussion
The historical assumption that all WMI in term infants was prenatal in origin was not unreasonable given that there is often a documented absence of intrapartum asphyxia in infants with NE. Our infants demonstrated patchy restricted diffusion in the deep white matter bilaterally on diffusion weighted imaging (DWI), confirmed as acute ischaemia/infarction on apparent diffusion coefficient (ADC) maps, strongly suggesting that this injury occurred in the perinatal period.
The exact nature of these white matter lesions has been debated as histological specimens are usually not available. Li et al suggest that the failure to detect WMI on CT, despite observing subdural/intraventricular haemorrhage, and the lack of T2 change associated with these lesions in most newborns, indicate that the WMI is largely non-haemorrhagic.1 Our findings support this theory. While the pattern of high T1 and low T2 signal abnormality found in our infants is suggestive of haemorrhage, a gradient-echo sequence performed on three patients showed no evidence of signal dephasing to indicate the presence of blood products in the areas of restricted diffusion. Gradient-echo imaging has a limited role in the evaluation of the presence or absence of haemorrhage but is an adjunct to the signal characteristics on T1 and T2 sequences. Susceptibility weighted imaging is more sensitive but was not available at the time of initial imaging of our infants. The persistence of the white matter signal abnormality and/or the finding of white matter loss in all four cases on follow-up MRI also support the suggestion that these lesions represent focal areas of WMI/infarction.
Placental pathology can identify clinically silent, pathophysiological processes that may directly cause central nervous system damage, decrease the threshold for neurological injury or serve as a marker for a deleterious in utero environment.3 ,5 ,6 High grade VUA has been identified in only 3% of control placentas. The finding of high grade (grades 3–4) VUA in all four of our cases has never been previously reported. A unique feature of VUA is that fetal tissues—the chorioamniotic membranes and chorionic villi—are infiltrated by maternal T cells. This process could result in disruption of endothelial integrity leading to inflammation, thrombosis and obliteration of the fetal vessel. The relationship among chronic chorioamnionitis, antenatal inflammation and WMI in preterm animals is established in the recent experimental literature.7 Our hypothesis is that VUA may also induce white matter damage by causing a local inflammatory response and/or oxidative stress during the perinatal period.
Conclusions
Isolated non-cystic WMI should not be assumed to be antenatal in origin. The retrospective nature of our cohort study and the associated risk of selection bias must be acknowledged. Our study underlines the importance of continued close and systematic evaluation of all cases of NE, including a detailed examination of the placenta, in order to come to a better understanding of the clinical presentation, the patterns of brain injury and the underlying pathophysiological processes.
Footnotes
-
Contributors MB: Writing of the paper, evidence based search, coordinator and reviewer. Trainee paediatrician; VO'D: Writing of the paper, evidence based search, analysis of radiology images and review of paper drafts. Consultant radiology lead; EM: Contributed to the Histopathology section of the paper. Reviewer of the drafts of the paper. Consultant Histopathology lead; MS: The paediatric psychologist who assessed each child and contributed to the developmental aspects of the paper; TP: Identification of the original cohort of patients. Analysis of the radiological images with the radiology colleagues. ET: reviewer of the radiology, reviewer and contributor to the text relating to radiological images; SR: reviewer of radiology. Reviewer and contributor to the entire text; EL: reviewer of the radiology, reviewer and contributor to the text relating to radiological images; AT: the primary neonatologist who identified the cohort. MB's supervisor and principle co-author.
-
Competing interests None.
-
Provenance and peer review Not commissioned; externally peer reviewed.