Dural sinus malformation (DSM) is a rare congenital malformation which contains a dilated dural sinus pouch that communicates with the other sinuses and drains cerebral veins [3]. It belongs to the group of dural arteriovenous shunts, accounting for less than 2% of congenital intracranial vascular anomalies. Prenatal diagnosis has only been reported in about 50 cases thus far in literature.
They have been classified anatomically into a midline type and a lateral type, primarily based on the location. The former involves the posterior sinus with or without the torcula, with giant dural lakes and slow flow mural arteriovenous shunting, associated with spontaneous thrombosis, hemorrhagic infarction and a poorer prognosis. The latter involves the jugular bulb with otherwise normal sinuses and an associated high flow sigmoid sinus arteriovenous fistula. It has a good prognosis due to normal contralateral drainage pathway [4].
The etiopathological mechanism of development of DSM is debatable and many hypotheses have been put forth. The more commonly accepted theory considers that it results from excessive and disorganised development of the posterior sinuses, even when they should be decreasing in size [1, 5]. Thrombosis is proposed to be a secondary effect rather than a causative phenomenon, due to possible local factors like immaturity of sinuses, disturbance of blood flow, and modification of the sinus endothelial lining [6]. However, recently it has been postulated that the normal gestational remodelling of the ballooned sinuses is impeded by the high venous pressure due to the presence of dural arteriovenous fistulas and thus suggests a common mechanism for both antenatal and postnatal DSM [7].
Prenatal diagnosis is possible and sonography is the initial imaging modality of choice as well as for subsequent regular follow-up. Typical sonographic finding of DSM includes a huge anechoic cystic structure in the posterior fossa attached to the dura mater [8]. On colour Doppler, vascularity at the lateral margins of the lesion has been described with no detectable flow in the centre [9]. The marginal flow is postulated to represent multiple mural arteriovenous shunts usually associated with the lesion [4]. Failure to detect blood flow within the lumen has been attributed to its very low velocity [4], and in such cases altering the usual Doppler setting to low pulse repetition frequency and wall filters may aid in detecting flow within the lesion. Also, modern techniques such as 3D/4D colour Doppler through transvaginal approach have been employed to detect vascularity and obtain additional structural data about the lesion [10].
There will be an increase in the possibility of missing this diagnosis if one is not aware of such an entity. Any such suspicious cystic focus should be investigated further with fetal MRI.
Fetal MRI is an essential imaging modality, to characterise and confirm the diagnosis of DSM and also identify any intracranial complications. Commonly, black blood techniques are used in fetal imaging. Profoundly hypointense signal is given by DSM on fast spin-echo T2 weighted sequences, as seen in our case [9]. However, if thrombosis is present, the signal can be iso-hyperintense on T1 and iso-hypointense on T2 depending upon the age of the thrombus. Resultant mass effect over the cerebellum, fourth ventricle and associated hydrocephalus has also been described in a few cases [2].
Bright blood techniques, mainly time of flight angiography (TOF), are widely used in postnatal imaging and hence their features on prenatal studies have not been described in literature. However, in our case, fetal 3D gradient recalled echo Dixon-based MRA was attempted and proved useful as it revealed arterial supply to the DSM.
Characteristic location and profound T2 hypointensity of venous pouch of DSM differentiates it from the more common fetal posterior cranial fossa lesions like Dandy–Walker malformation and retrocerebellar arachnoid cysts; vascular malformations, such as vein of Galen aneurysmal malformation; and congenital tumours like teratomas. While they may appear similar on ultrasound, fetal MRI easily distinguishes them [9].
According to the literature, the prognosis of patients with dural sinus malformation is still uncertain, and no prognostic criteria could be suggested for the fetus. The overall mortality rate is 15%, with a favourable outcome seen in 87% of the surviving patients [11]. Factors in favour of a good neurological outcome were lateral type of DSM and antenatally detected thrombosis which either showed progressive decrease in size or spontaneous resolution [7, 11,12,13] .Presence of parenchymal infarctions, haemorrhage, hydrocephalus, congestive cardiac failure and non-resolution of thrombus or postnatal onset of thrombosis all contributed to poor prognosis resulting in either fetal/postnatal demise or a poor neurological outcome in the neonate [7, 11].
Diffusion-weighted imaging (DWI) helps in prognostication by demonstrating intralesional thrombosis and excluding potentially devastating intracerebral complications [14]. In our case, though no intrasinus thrombosis was seen, a small focus of restricted diffusion was seen in the right hemipons, likely representing an acute ischemic insult. There has been no literature that has reported the presence of an ischemic lesion on prenatal DWI in the absence of DSM thrombosis. Hence this finding is likely of inconclusive significance which needs follow-up.
Treatment strategies primarily aim at preserving venous drainage of the brain and depend on the angioarchitecture of the lesion. Multistage transarterial or transvenous endovascular embolisation with glue or coils is the primary therapeutic method [7, 13, 15]. In addition, medical treatment with heparin helps to prevent dural sinus thrombosis [15]. Surgical treatment in combination with embolisation has been shown to have positive clinical outcomes in neonates [15]. Information provided by MRI has a potential role in adjusting the treatment strategy to the postulated pathophysiology of symptoms and spontaneous vascular changes [9].