Relationship of intracranial, optic-nerve sheath and retinal hemorrhage

Retinal hemorrhages in patients with epidural hematomas

PURPOSE

To present a series of infants found to have sparse retinal hemorrhages associated with isolated epidural hematomas. None of these infants had other intracranial or subdural hemorrhages present.

METHODS

Children less than 3 years of age admitted to an urban children's hospital between January 1998 and December 2002 with radiographic evidence of an isolated epidural hematoma at the time of presentation were eligible for this study.

RESULTS

During the time period of the study, 15 children were admitted with traumatic epidural hematomas. Nine patients had an ophthalmologic examination, of which five had evidence of sparse retinal hemorrhages. All five patients with retinal hemorrhages were less than 8 months of age and all required surgical evacuation of the epidural hematomas. Of the five patients with retinal hemorrhages, four were unilateral, one was bilateral, and in all cases, the retinal hemorrhages were superficial, few in number, and confined to the posterior pole. There were no deep retinal or subretinal hemorrhages present. The institutional child protection team evaluated all five patients with retinal hemorrhages and each case was felt to be consistent with the history provided and no history of shaking was elicited.

CONCLUSIONS

We found sparse retinal hemorrhages in five of nine patients who presented to our hospital with isolated epidural hematomas and who had had an ophthalmologic examination. All nine patients were evaluated by the institutional child protection team, who did not feel that there was sufficient evidence to be suspicious of nonaccidental trauma.

Retinal findings in children with intracranial hemorrhage

PURPOSE

To identify the incidence of Terson's syndrome in children.

DESIGN

Prospective, observational case series.

PARTICIPANTS

Fifty-seven consecutive children with known intracranial hemorrhage from nonabuse causes.

METHODS

Dilated fundus examination to detect intraretinal hemorrhages or other abnormalities.

MAIN OUTCOME MEASURES

Presence or absence of intraretinal hemorrhages or other abnormalities.

RESULTS

Fifty-five patients (96%) had no evidence of intraretinal or vitreous hemorrhage. Two patients had abnormal retinal examinations. One patient had a single dot hemorrhage associated with presumed infectious white retinal lesions. The second patient had three flame and two deeper dot intraretinal hemorrhages after a motor vehicle accident (1.5% incidence of retinal hemorrhage).

CONCLUSIONS

Retinal hemorrhage is uncommon in children with intracranial hemorrhage not resulting from shaken baby syndrome. The maximal incidence of intraretinal hemorrhage in children with nonabuse intracranial hemorrhage is 8%.

Benign idiopathic haemorrhagic retinopathy

PURPOSE

To describe a new condition characterised by an unusual unilateral idiopathic haemorrhagic retinopathy.

METHODS

A review is presented of patient histories from 5 patients with acute-onset unilateral idiopathic haemorrhagic retinopathy, including results of ophthalmological, haematological and fluorescein angiographic examinations.

RESULTS

All patients had an extensive deep blot haemorrhagic retinopathy without significant vascular signs or abnormal optic discs. In 4 cases the haemorrhage was sufficiently severe to break through into the vitreous. Fluorescein angiography demonstrated normal arteriovenous flow, without capillary non-perfusion, vessel or disc leakage. Disc swelling, macular oedema and cotton wool spots were not seen at any stage in these patients. All patients recovered the visual acuity in the affected eye by 4 months. Systemic examination in all cases was unremarkable.

CONCLUSION

This distinct and rare form of retinopathy is important to define since it has a good prognosis without treatment.

Bilateral subinternal limiting membrane hemorrhage with Terson syndrome

PURPOSE

To report the anatomic location of bilateral dome-shaped posterior pole hemorrhages in a patient with Terson syndrome.

METHODS

Case report. We performed bilateral vitrectomy for vitreous hemorrhage in a patient with Terson syndrome. After removal of vitreous hemorrhage, the tissue overlying a large discrete hemorrhage in the posterior pole was removed, and the tissue from one eye was examined histologically.

RESULT

The discrete dome-shaped hemorrhage in the posterior pole was confined to the retina anteriorly by the internal limiting membrane.

CONCLUSION

Large dome-shaped retinal hemorrhages with Terson syndrome can be located beneath the internal limiting membrane of the retina.

Vitreous Hemorrhage

The incidence of spontaneous vitreous hemorrhage is approximately 7 cases per 100,000 population. Proliferative diabetic retinopathy (32%), retinal tear (30%), proliferative retinopathy after retinal vein occlusion (11%) and posterior vitreous detachment without retinal tear (8%) are the most common causes of spontaneous vitreous hemorrhage. Vitreous hemorrhage can be caused by the pathologic mechanisms of disruption of normal retinal vessels, bleeding from diseased retinal vessels or abnormal new vessels, and extension of hemorrhage through the retina from other sources. Hemorrhage into the vitreous gel results in rapid clot formation and is followed by slow clearance of approximately 1% per day. The cellular response to vitreous hemorrhage is unusual with regard to hemorrhage in any tissue outside the vitreous cavity and has been compared to a "low-turnover" granuloma. Unique clinicopathologic features of long-standing vitreous hemorrhage include cholesterolosis bulbi (synchysis scintillans), hemoglobin spherulosis, and vitreous cylinders. Complications of nonclearing vitreous hemorrhage are hemosiderosis bulbi and glaucoma. Ghost cell glaucoma, hemolytic glaucoma, and hemosiderotic glaucoma may result from vitreous hemorrhage. The established treatment option for nonclearing vitreous hemorrhage is pars plana vitrectomy. Experimental nonsurgical treatment options involve improvement of physiologic clearance mechanisms in order to accelerate fibrinolysis, liquefaction, hemolysis and phagocytosis.

Dense vitreous hemorrhages predict poor visual and neurological prognosis in infants with shaken baby syndrome

OBJECTIVE

A retrospective study was performed to examine the relationship between the severity of vitreous hemorrhages and the severity of neurological injury in infants with shaken baby syndrome, and the result of early vitrectomy in these infants.

METHODS

Five infants, ages 3 to 8 months, with confirmed child abuse underwent ocular examination and neurological testing to identify varying degrees of vitreoretinal changes and neurological dysfunction. Pars plana vitrectomy was performed on one or both eyes in each infant to remove the vitreous hemorrhage.

RESULTS

The three infants with bilateral dense vitreous hemorrhage and multiple subarachnoid hemorrhage and cerebral contusions had poor postoperative ocular and neurological outcomes including light perception (LP) or no light perception (NLP) vision, large retinal holes or tears, retinal ischemia, and severe encephalopathy. The other two infants presented with bilateral subhyaloid and retinal hemorrhages without dense vitreous involvement, along with mild subarachnoid hemorrhage and minimal neurological findings. Following vitrectomy, visual acuities in these infants improved to or remained at the fix and follow state, and both infants improved neurologically.

CONCLUSIONS

The presence of dense vitreous hemorrhage in infants with shaken baby syndrome has a poor visual prognosis, due to the frequent concomitant occurrence of significant retinal and visual cortical pathology. In contrast, infants with only intraretinal or subhyaloid hemorrhage tend to have a much better prognosis, with less disruption of both intraocular and intracranial structures.

Terson's syndrome: a reversible cause of blindness following subarachnoid hemorrhage

Terson's syndrome refers to the occurrence of vitreous hemorrhage with subarachnoid hemorrhage (SAH), usually due to a ruptured cerebral aneurysm. Although it is a well-described entity in the ophthalmological literature, it has been only rarely commented upon in the neurosurgical discussion of SAH. Fundus findings are reported in a prospective study of 22 consecutive patients with a computerized tomography- or lumbar puncture-proven diagnosis of SAH. Six of these patients had intraocular hemorrhage on initial examination. In four patients vitreous hemorrhage was evident on presentation (six of eight eyes). In the subsequent 12 days, vitreous hemorrhage developed in the additional two patients (three of four eyes) due to breakthrough bleeding from the original subhyaloid hemorrhages. The initial amount of intraocular hemorrhage did not correlate with the severity of SAH. Two of the six patients with intraocular hemorrhage died, whereas five of the 16 remaining SAH patients without intraocular hemorrhage died. Of the four survivors with intraocular hemorrhage, three showed gradual but significant improvement in their visual acuity by 6 months. The fourth underwent vitrectomy at 8 months after presentation and had a good visual result. With modern and aggressive medical and microsurgical management, Terson's syndrome should be recognized as an important reversible cause of blindness in patients surviving SAH.

Long-term visual outcome in Terson syndrome

The presentation and long-term visual outcome in 30 eyes with Terson syndrome is evaluated. In 25 of 30 eyes (83%), visual acuity of 20/50 or better was attained. This occurred in 12 of 16 eyes (75%) managed by observation alone and 12 of 14 eyes (86%) treated by pars plana vitrectomy. The most common long-term sequelae in all eyes studied was the formation of an epiretinal membrane. These occurred in 14 of 18 eyes (78%) followed for 3 or more years but accounted for significant visual loss in only 2 eyes. There was no difference in final visual outcome between those patients undergoing vitrectomy and those managed conservatively. However, visual recovery was more rapid in eyes undergoing vitrectomy despite the fact that vitrectomy was reserved for eyes with more dense vitreous hemorrhage.

Sequential cranial computed tomography in infants with retinal hemorrhages

Five whiplash-shaken infants presented initially with lethargy, vomiting, irritability, and intraocular hemorrhage and had sequential cranial computed tomography (CT). In three of the five infants, results of the initial CT scans of the head were either normal or demonstrated subtle abnormalities. However, significant subdural hemorrhages which were demonstrated by subdural aspiration developed subsequently in all three infants. Results of repeat CTs of the head showed progressive ventricular dilation and symmetrical bilateral fronto-parietal subdural hematomas. The retinal hemorrhages preceded both the clinical and radiologic recognition of subdural hematoma. Results of the initial head CT of the remaining two children showed intracerebral hemorrhages, cerebral edema and subsequently subdural hematoma. The finding of unexplained retinal hemorrhages in some infants, even though the initial CT head scan may show only minimal changes, warrants monitoring of the head circumference, fundus, and repeat head CT as indicated.

Terson's syndrome. Clinicopathologic correlations

In seven cases of Terson's syndrome, an elevated, dome-shaped, membrane was detected in the posterior pole by ophthalmoscopy, echography, or during pars plana vitrectomy. Light and electron microscopic examination of two additional eyes obtained postmortem from an acute case of Terson's syndrome revealed that the posterior vitreous face was elevated by blood and that the internal limiting membrane of the retina was intact and in its normal position. In two chronic cases, a dome-shaped epiretinal membrane was excised from the macula during vitrectomy. Light and ultrastructural studies demonstrated that the membranes consisted of glial cells and basement membrane material. The dome-shaped structure observed in eyes with Terson's syndrome is due to the formation of a subhyaloid hemorrhage. The partially detached posterior hyaloid face created by this hemorrhage provides a scaffold for cellular proliferation and the development of an elevated epiretinal membrane in long-standing cases. Echographically, this membrane resembles a retinal detachment in B-scans, but can be clearly distinguished from retina with standardized A-scan.

Ophthalmoscopic findings in occult child abuse

Two previously healthy infants initially presented with vomiting and lethargy. There was a history of whiplash shaken infant syndrome in one infant, but neither child had external or skeletal survey evidence of abuse. However, ophthalmoscopy revealed bilateral retinal hemorrhages, papilledema and vitreous hemorrhage. Subdural aspirations revealed occult hematomas. Both children required ventricular peritoneal shunts to control intracranial pressure. Ophthalmologists, pediatricians, and primary physicians should be aware of the possibility of child abuse in the differential of diagnosis of Terson's syndrome, especially in infants since this may be the only physical finding.

Vitreous hemorrhage after intracranial hemorrhage

Eight patients, aged 2 months to 55 years, developed vitreous hemorrhages as a result of subarachnoid or subdural bleeding. Subhyaloid hemorrhages were associated with, or preceded, hemorrhage into the vitreous cavity in four cases. Most vitreous hemorrhages cleared spontaneously, several months later, without major visual sequelae. In one patient, intravitreal blood persisted after 28 months. Vitreous hemorrhage can be a serious complication in patients surviving subarachnoid or subdural hemorrhages and, though uncommon, probably occurs with greater frequency than previously acknowledged. While vitrectomy may be a reasonable therapeutic approach in selected cases, in most instances vitreous hemorrhage following intracranial hemorrhage should be treated conservatively.