Medullary vein thrombosis
Venous infarct locations
- primary white matter haemorrhage
- terminal vein infarction
- primary medullary vein haemorrhage
- inferior ventricle vein infarction
Congestion
In infants with a vascular anomaly (like vein of Galen malformation) or with deep cerebral vein thrombosis (ICV or basal vein or tributaries thereof) it is common for lateral ependymal collector veins to be congested, and this may lead to medullary vein congestion, subsequently also micro- or macrohaemorrhage; but there is than evidence of thrombosis of large distal veins and often also IVH (e.g. due to leakage from choroidal veins). The imaging changes mainly affect white matter near the ventricles. Haemorrhage associated with polycythaemia is rarely described. Its characteristics are therefore unknown. Several cases are reported of sinus thrombosis in that context.Superior sagittal sinus thrombosis
Polycythaemia
Medullary veins
Medullary veins from just below the subcortex drain into deep veins, gradually coalescing into larger trunks (Okudera et al. 1999). This high drainage with a peculiar organisation of confluence explains the feathered appearance of the outer (subcortical) contour of a medullary venous infarct. It also explains the triangular shape as veins fan out from the ependymal collectors near the GMH.IVV inferior ventricle vein
LCV longitudinal caudate vein (anterior terminal vein)
SV striatal vein
TV terminal vein
Transmedullary vein
Transcerebral or transmedullary veins cross the white matter from ventricle to pia mater; anastomosing medullary veins join a pial draining vein to an ependymal draining vein, creating a transmantle connection based on three vessels; in third trimester, transmedullary veins are common in the temporo-occipital area.Venous congestion
Medullary vein "thrombosis"
The entity was recently picked up by fetal and neonatal imaging (both CUS and MRI) and is now referred to by radiologists as primary medullary vein thrombosis (Arrigoni et al. 2011, Benninger et al. 2021, Khalatbari et al. 2021, Alves Fonseca et al. 2021). Although this may on occasion be the mechanism (e.g. in a fetus with ATIII deficiency described by Konanki et al. 2015), thrombosis may not be the primary event in all. Seventy-five published cases of neonatal DMV “thrombosis" were reviewed by Pin et al. 2023. Respiratory distress, resuscitation or need for inotropes were present in 45%. Signs and symptoms at presentation included seizures (48%), apnoea (36%), lethargy or irritability (35%). At MRI, fan-shaped (in coronal scans) linear T2 hypointense lesions were documented in all. All had ischaemic injuries, often involving the frontal (84%) and parietal lobes (76%). Signs of haemorrhagic infarction were present in 53/54 (98%). A proportion of patients developed sequelae (intellectual disability in 37% and epilepsy in 18%). Term, emergency caesarean, seizures; developed spastic quadriplegia with epilepsy; no mechanism found. The lesions are MR typical haemorrhages: hyperintense on T1 (bright) in the acute stage, hypointense on T2 (dark) and very hypointense in SWI sequences. Some are accompanied by perivenous cytotoxicity, suspected by the presence of ADC reduction. They can be unilateral and bilateral, some areas lead to frank bleeding, others remain petechial. With a stretch of imagination they have been called “iris”like and “brush”like, mainly due to the irregular feathered character of the pial side and due to the fan shape in coronal CUS or MRI.Asymmetric venous injury
Inflammation ?
GA 37w, SGA, induction of labour at 37w; fever day 4 with seizures; no intrapartum asphyxia; CSF clearly elevated leukocytes and protein, cultures negative; metabolic screen negative; antibiotics for 21 days; recurrent fever week later with still elevated CSF leukocytes and protein; steroids started for “inflammation” (courtesy dr Naulaers, UZ Leuven).Courville 1960
Extensive bilateral haemorrhagic change in cerebral white matter, without clear injury to grey matter, is rare but regularly reported in recent years. The initial descriptions (Courville 1960: Central hemorrhagic encephalopathy of early infancy) framed the event as a sequel to asphyxia (during or after birth). But medullary vein thrombosis was not obligatory, although some fibrin deposition was observed here and there. Three term infants were described, died within hours, two suspected of asphyxia during labour, one after birth due to a mucus plug. Characteristic were multiple petechial haemorrhages around cerebral white matter veins, variable neuronal necrosis of the cortex, no gross lesions to deep grey matter; multiple veins were congested, no thrombi were found in SSS or ICV.Case example
A one day old girl presented to evaluate a systolic murmur and oxygen need (below 30 %). She was one of twins, born by caesarean section because of breech presentation. There was meconium staining of the amniotic fluid. Apgar scores were 8 and 9 at 1 and 5 minutes. Birthweight was 2380 gr (< 2sd), head circumference 33 cm (-1.5sd). Findings: mild tachypnoea, a systolic murmur heard in the fourth left intercostal space and an enlarged liver (3 cm below ribcage). Echocardiogram showed an open duct and mild tricuspid valve insufficiency. Capillary and venous haematocrit measured 78 and 70% respectively, at 26 and 28 hours after birth. Viscosity was not measured. A partial exchange transfusion was performed 40 hrs after birth via an umbilical arterial catheter and a peripheral venous line. Biochemical evaluations were normal with the exception of a mild thrombocytopenia (nadir 68 x 109/L at 28 h) and mild hypoglycaemia (nadir 1.7 mmol/L) corrected with a maximum suppletion of 12.5 mg/kg/min. The first brain US, at 6 days of age, revealed symmetrical hyperechoic change in white matter mainly in the frontal regions anterior to striatum, in a structurally normal brain without IVH and with patent superior sagittal sinus and internal cerebral veins. MRI on day 16 confirmed these findings and documented their haemorrhagic nature. Clinical recovery was uneventful, seizures were not observed. At 1 year of age, mild hypotonia and delay in gross motor development were observed (she was not yet able to sit alone).References to germinolysis
References- Alves Fonseca AP, Melo de Carvalho R, Padula M, de Cassia Maciel Pincerato R. Deep medullary vein thrombosis in a neonate: a peculiar MRI pattern. Neurology. 2021 Mar 9;96(10):492–3.
- Arrigoni F, Parazzini C, Righini A, Doneda C, Ramenghi LA, Lista G, et al. Deep medullary vein involvement in neonates with brain damage: an MR imaging study. AJNR Am J Neuroradiol. 2011 Dec;32(11):2030–6.
- Benninger KL, Benninger TL, Moore-Clingenpeel M, Ruess L, Rusin JA, Maitre NL. Deep medullary vein white matter injury global severity score predicts neurodevelopmental impairment. J Child Neurol. 2021 Mar; 36(4):253–61.
- Courville CB (1960) Central hemorrhagic encephalopathy of early infancy. Neurology. Jan;10:70-80.
- Khalatbari H, Wright JN, Ishak GE, Perez FA, Amlie-Lefond CM, Shaw DWW. Deep medullary vein engorgement and superficial medullary vein engorgement: two patterns of perinatal venous stroke. Pediatr Radiol. 2021 May;51(5):675–85.
- Konanki R, Varma DR, Ratha C, Lingappa L, Shah N. Teaching NeuroImages: Fetal deep medullary vein thrombosis presenting as progressive intracerebral hemorrhage. Neurology. 2015 Jul 7;85(1):e5-6.
- Okudera T, Huang YP, Fukusumi A, Nakamura Y, Hatazawa J, Uemura K (1999) Micro-angiographical studies of the medullary venous system of the cerebral hemisphere. Neuropathology 19:93-111.
- Pin JN, Leonardi L, Nosadini M, Pelizza MF, Capato L, Piretti L, Cavicchiolo ME, Simioni P, Baraldi E, Perilongo G, Luciani M, Sartori S. Deep Medullary Vein Thrombosis in Newborns: A Systematic Literature Review. Neonatology. 2023;120(5):539-547.
- Ramenghi LA, Fumagalli M, Righini A, Triulzi F, Kustermann A, Mosca F. Thrombophilia and fetal germinal matrix-intraventricular hemorrhage: does it matter? Ultrasound Obstet Gynecol. 2005;26(5):574-6.
- Takashima S, Tanaka K (1978) Microangiography and vascular permeability of the subependymal matrix in the premature infant. Can J Neurol Sci 5(1):45-50.