Patients with pulsatile tinnitus occupy a small (approximately 4%) subset of most patients presenting with a tinnitus [
7]. Unlike subjective tinnitus, it arises from a mechanical somatosound. And consideration of differential diagnosis for pulsatile tinnitus is important because there is an array of causes that can be identifiable and treatable easily, and many of them are vascular origin problem [
8]. DAVF are abnormal arteriovenous shunts between dural arteries and venous sinuses and around the meningeal/cortical venous vessels [
8]. Although their etiology is not clearly understood, DAVF have been found in association with venous thrombosis, intracranial surgery, tumor, trauma and congenital anomalies. Depending on their location and the venous drainage pattern, it arises mostly at the transverse, sigmoid, and cavernous sinuses [
2].
Imaging modalities for the patients with pulsating tinnitus are necessary to detect the potentially treatable vascular etiologies, and they include computed tomography angiography (CTA), non-enhanced CT of the temporal bone, MRI and MRA, catheter angiogram [
9]. CTA is considered as initial imaging examination of choice, because it provides a robust assessment of various intra- and extratemporal vascular pathologies, including DAVF, as well as an excellent description of the bony anatomy of the temporal bone [
8]. Temporal bone CT may represent the transosseous vascular shapes, which are highly specific for the diagnosis of DAVF [
9]. MRI and MRA have advantages in detecting abnormal vascular pathologies associated with the dural AV malformations. CTA and MRI could provide additional findings such as asymmetric arterial supply vascultures, “shaggy” or ill-defined dural sinus margins, or dural sinus occlusion. Among these imaging modalities, catheter angiogram is useful in the configuration of negative noninvasive evaluation with highly clinical suspicious for DAVF and has a role as the golden standard for DAVF diagnosis, characterization and planning treatment method [
2]. We explained both MRA and CT angiography to the patient at first. The patient wanted to identify the brain parenchyma with MRI, therefore MRI with MRA was taken and catheter angiography was performed for confirmation. Embolization could be performed, but the fistula was not so large, so we tried to perform compression first and got good results.
The therapeutic methods include conservative monitoring, arterial embolization, transvenous occlusion, surgical excision, and radiation therapy [
10]. And in the majority of DAVF, transarterial embolization or by transvenous occlusion with coils or ballons in the presence of multiple feeding vessels offers the best chance for successful treatment [
11]. But spontaneous occlusions or occlusion by manual compression have often been reported and the pathophysiology is suggested that the frequent and repeated temporary stasis in the fistula areas will promote the increasing thrombosing of the shunt [
5,
6]. Turbulent arterial blood flow elicted by atherosclerosis probably promotes the thrombosis on the venous side of the DAVF or sinus. Spontaneous occlusions may be the result of the pressure fluctuations similar to those caused by manual compression. Because the wide shunt defects and high flow conditions do not allow thrombus formation, the compression maneuver may be only effective in low flow fistulas [
12]. The external manual compression technique for the carotid artery and jugular vein was introduced several years ago as an alternative therapy for DAVF, but the efficacy of this technique was not confirmed due to limited available data [
4]. And factors that contribute to its success and the closure and healing of these lesions remain unclear [
13]. Moreover, the compression effect, compression time per single session, the compression frequency and total compression duration are not clear. Nevertheless, this technique may lead to fistula occlusion successfully or at least improvement of the pulsating tinnitus. Halbach, et al. [
14] stated that 22% cured completely and the improvement of clinical symptoms were 33%. And Cognard, et al. [
15] recommended that the compression technique should be increased in duration from 10 to 30 seconds and the frequency should be increased from four to six times per hour. But patients may experience the neurologic deficits such as acute hemiplegia with loss of consciousness during manual compression, which makes the patient hesitate adapting his or her daily routine to the treatment. Therefore the success rate will somewhat depend on the willingness of the patient to do the frequent compressions continuously. Since the changes in fistula activity and changes in flow sound are closely related to compression techniques, the relationship between the repeated manual compression therapy in our patient and the subsequent, sustained fistula obstruction might be close [
4]. External compression technique may stimulate the vagus nerve located in the carotid sheath and thereby induce vasovagal reactions following in cardiac arrhythmias such as sinus bradycardia, atrioventricular block, systole, and syncope. Especially, elderly patients with significant atherosclerotic disease are at increased risk for neurological deficits and stroke from compromised carotid blood flow and dislodged emboli [
13]. Although the consensus about the method of the manual compression therapy is not existed, this therapy might give useful opportunity of the improvement due to its noninvasiveness and simplicity.