Intraspinal bleeding in the hemophilic patients is rare. De Tezanos Pinto et al  reported only two cases of intraspinal bleeding among his patients (1410 hemophiliacs) between 1960 and 1991. In another 11-year study from 1965 to 1976, there were 6 patients with intraspinal bleeding in a population of 2500 hemophiliacs . The incidence of intracerebral bleeding was 65 out of these 2500. Among these, spinal subdural hematoma is rare. We could find only 1 case of spinal subdural hematoma in the hemophilic patients reported on Medline . The source of intraspinal bleeding in the hemophiliacs is rarely found. But, in patients who come to operation or autopsy, the majority of bleeds are found to be extramedullary, and only a small number are intramedullary . Spinal epidural bleeding is presumed to be due to the rupture of the epidural veins. Lack of valves and minor resistance of these vessels can result in bleeding if a sudden increase in intra-abdominal or intrathoracic pressure occurs. Spinal subdural hematomas cannot be explained through this pathogenesis. Vinters et al  have postulated that in many cases, spinal subdural hematomas may be due to a primary lesion that eventually dissects into the subdural space. As the subarachnoid blood is dissipated by CSF, the subdural clot is all that remains. Our patient's clear CSF study doesn't support this mechanism.
According to the works of Haines et al  and Morandi et al , under normal conditions there is no evidence of a naturally occuring space being extant at the dura-arachnoid junction. A space may appear at this point subsequent to pathological/traumatic processes that result in tissue damage with a cleaving along the structurally weakest plane in the meninges through the dural border cell layer. Furthermore, when a space does appear, it is not "subdural" in location but rather within a morphologically distinct cell layer. So according to works of these authors, the so-called spinal subdural hematoma could be viewed as a spinal dural border hematoma.
Spinal subdural hematoma usually presents with sudden local neck or back pain followed by neurological deficit. Sometimes the onset is chronic with no pain .
In our presented case, the clinical course of the disease could not be reliably ascertained. The patient has not reported acute pain. The child's deficit gradually worsened. This gradual delayed neurological deterioration, apparently occurs because the processes of primary hemostasis are only temporarily effective. Delayed bleeding has been reported to occur several hours or even days later . Since the lumbar puncture level had been considerably below the lowest limit of the hematoma, the hematoma itself cannot be attributed to this procedure, but lumbar puncture may be one of the reasons for further neurological deterioration of the patient.
We obtained spinal MRI and CT scan on the operation day, which was six days post trauma. In his thoracic axial CT, the hematoma was hyperdense, circumferential and mainly anteriorly located. In the sagittal T2-weighted cervical MRI, the hematoma from C5 to midthoracic level, was hyperintense and mainly posteriorly located in the thoracic and thoracolumbar level. This is compatible with the timing of the imaging (6th day after trauma and hemorrhage)[9, 10] and is postulated to occur because of methemoglobin formation . In the T1-weighted axial MRI scan of the lower thoracic region, areas of hyperintensity in the periphery of the hematoma were seen, suggesting methemoglobin formation. It has been said that visualization of the dura mater and preservation of the fat pad may favor of the intradural location of the hematoma . Sagittal T2-weighted image of the lumbar region, showed mixed signals posteriorly.
Since the trauma to our patient was not of a degree to generally cause neural injuries, we can assume that the compressive effect of the hematoma was the main cause of neurological deterioration of the patient. In our case, a clear preoperative CSF study, the tension of the dura during operation and normal looking arachnoid and cord underlines the significance of compression as the pathophysiologic mechanism. We postulate that after initial control of the hemorrhage, minor rebleeding or expansion of the hematoma resulted in the deteriorating neurological status of the patient. Recent rebleeding as the main reason of neurological deterioration of our patient cannot be proved through our images.
The limited available literature mainly addresses the management of spinal epidural hematoma in the hemophiliacs [13–15, 9]. Schmitz et al  has reported his survey of treatment and outcome in hemophilic patients with spinal epidural hematoma from 1977 to 1998. In his survey of eleven cases of such patients (4 hemophilia B and 7 hemophilia A), 3 were operated on and one of the three operated cases had complete recovery. The rest (8 patients), were conservatively managed with (75%) recovery. The cases are not completely comparable, but the author concluded that considering the fact that a significant number of the hemophilic patients suffering from intraspinal bleeding are children, and that there is a high risk of spinal deformity after decompressive laminectomy, conservative management should be preferred as much as possible. Although surgery has been associated with high morbidity and mortality, early surgical intervention is always indicated when the patient's neurological status progressively deteriorates. We operated on our patient after his neurological deterioration. Although our patient had been operated on late after the appearance of first neurological deficits, he recovered completely. This good functional outcome is not the rule. Review of the literature on this pathology indicates clearly that delay of surgical decompression for several hours after the appearance of an important sensorimotor deficit increases the risk of a poor functional outcome [17, 18, 7]. Considering his spinal level of neurological deterioration, we only approached the thoracolumbar hematoma. But the patient improved postoperatively and in his MRI taken 6 months later, the resorption of the hematoma in the cervicothoracic region is clearly seen. In fact the thoracolumbar subdural hematoma of this patient has been managed operatively and its cervicothoracic portion, conservatively.