Fisher’s exact test, p = 0.010). These non-favourable patterns were associated with motor developmental delay in the surviving patients (χ2 test, p = 0.001). For neurodevelopment, non-favourable patterns showed a trend towards delay (χ2 test, p = 0.100).Eighteen patients (36.7%) did not survive the neonatal period. Thirteen patients died within 72 h after birth, mainly due to acute multiple organ failure. The other 5 patients died following a decision to redirect care due to the expected serious long-term sequelae, based on clinical parameters (e.g. therapy-resistant convulsions) and/or extensive cerebral damage documented by MRI. The de-ceased patients had a significantly higher need for respi-ratory and circulatory support compared to the living pa-tients, but there was no difference in the presence of sei-zure activity. Imaging was performed in 25 patients between day 2 and day 8 after birth (median: day 6). Five of the patients died, and MRI imaging was performed post-mortem in 3 of the patients. DWI images were available for 18 infants imaged during life. The clinical characteristics of infants with an MRI were not significantly different compared to the total study population. In 3 deceased patients, MRI imaging was obtained c d post-mortem. In 1 deceased patient, DWI (MRI obtained during life on day 6) showed few abnormalities in con- Fig. 1. MRI imaging. a, b MRI day 7: term born male with Apgar trast to the conventional images that were clearly abnor- scores 5 and 6 at 1 and 5 min and an initial Hb concentration of mal, possibly due to antenatal onset of the anaemia. 1.9 mmol/l [axial T2-weighted image at the level of the basal nuclei (a); axial DWI (b)]. Mild abnormalities in occipital white matter, White Matter Injury not seen as increased signal intensity on DWI. The patient did not come to the follow-up appointments, but had a favourable out-In 16 of the 25 patients (64%), some form of white matter come according to the general practitioner at the age of 2 years.After injury was found, both in surviving and deceased pa- c, d MRI day 2: term born female, Apgar scores 4 and 8 at 1 and 5 Neurodevelopmental test results were available in 20 of the 31 surviving patients (median age: 19 months, range: 14–35). The mean Z-score for neurodevelop-ment was 0.43 (SD 0.71). Only 1 patient (5%) was mild-ly delayed (Z-score: –1.08). The mean Z-score for mo-tor outcome was 0.58 (SD 0.81). One patient (5%) was mildly delayed (Z-score: –1.10). Unilateral spastic cere-bral palsy was diagnosed in 1 patient with haemor-rhagic cortical infarction. Of the 11 patients who were not formally tested, information was retrieved from the paediatrician, general practitioner or the parents. Cere-bral palsy was not diagnosed in any of these patients, but behavioural problems were reported in 1 patient. Information about outcome could not be retrieved in 2 patients. Thirteen of the 20 survivors with available MRI data were tested. The median age of testing was 24 months (range: 15–35). The mean Z-score for neurodevelopment was 0.37 (SD 0.76) and the mean Z-score for motor out-come was 0.67 (SD 0.88). No differences were found in neurodevelopmental and motor outcome between pa-tients with no/mild white matter abnormalities and pa-tients with moderate-to-severe white matter abnormal-ities. There were no differences between the two Hb con-centration groups for the neurodevelopmental outcome In this retrospective study we were able to show that lesions in the basal ganglia and thalami, and especially in the white matter, are common findings in infants with severe neonatal anaemia. Cerebral damage has been suggested to be caused by several pathophysiological mechanisms. In full-term in-fants, the deep grey matter nuclei are probably affected after an acute hypoxic-ischaemic insult due to changes in brain maturation and increased metabolic demands [1]. White matter injury is also found in full-term infants and is considered to be due to more prolonged and repetitive hypoxic-ischaemic events. A combination of deep grey matter damage and white matter lesions can also be found [2]. We hypothesize that the mechanisms that cause bas-al ganglia and white matter injury following severe anae-mia might be similar to the mechanisms responsible for cerebral injury in full-term infants with perinatal asphyx-ia due to other causes. White matter injury has been de-scribed in animals as well as in humans after moderate prolonged foetal or neonatal asphyxia, as well as after neonatal hypoglycaemia [3, 4]. It is important to note that a substantial percentage (31%) of our anaemic patients also had a period of hypoglycaemia. The mechanism of injury due to hypoglycaemia is still unclear, although it has been hypothesized that it is caused by increased re-gional cerebral blood flow during hypoglycaemia with a subsequent reduction in regional glucose uptake [3]. Even more interesting are the clinical implications of these findings. MRI is a well-established method to as-sess brain injury in infants suffering from perinatal as-phyxia, and these findings are closely related to later neurodevelopmental outcome [5–11]. Data on the long-term outcome of patients with severe acute anaemia combined with a less severe component of perinatal as-phyxia are still limited. It is of interest that early neuro-developmental and motor outcomes in the first 2 years of life in the survivors is favourable. In the absence of damage in the thalamus and basal ganglia, neonatal anaemia itself does not have a significant effect on out-come at 2 years of age [12]. However, the children are still young and their outcome should be re-assessed at school age. We also found that the severity of anaemia does not influence the need for intensive care treatment in the neonatal period. Moreover, we also found that there was no significant difference in survival between patients with an initial Hb concentration ≤3.0 mmol/l versus patients with an initial Hb concentration between 3.1 and 6.0 mmol/l. There was also no increased need for respiratory or circulatory support in the infants with more severe anaemia. However, the degree of organ fail-ure and especially the degree of encephalopathy scored clinically or with aEEG and the presence of seizure activ-ity and severe injury on MRI were correlated with mor-tality and/or redirection of care. Limitations This retrospective study has several limitations.