Treatment of rhinocerebral mucormycosis with intravenous interstitial, and cerebrospinal fluid administration of amphotericin B: case report.

IMPORTANCE: Rhinocerebral mucormycosis is extremely difficult to treat. Approximately 70% of patients are poorly controlled diabetics, and many of the remainder are immunocompromised as a consequence of cytotoxic drugs, burn injuries, or end-stage renal disease. Despite standard treatment consisting of surgical debridement and the intravenous administration of amphotericin B, rhinocerebral mucormycosis is usually a fatal disease. CLINICAL PRESENTATION: We describe the case of a 16-year-old male patient with juvenile onset diabetes mellitus who presented with fever, right-sided hemiparesis, and dysarthria. Axial view computed tomography revealed abscess formation in the left basal ganglia and frontal lobe, which was proven by stereotactic biopsy to contain Rhizopus oryzae. INTERVENTION: Intravenous administration of amphotericin B (30-280 mg/dose) was begun on the day of admission. On hospital Day 20, after the occurrence of frank abscess formation, the lesion was aggressively debrided. Despite these therapies, there was neurological deterioration characterized by the development of hemiplegia and aphasia. Sequential computed tomographic scans enhanced with contrast medium demonstrated progressively enlarging lesions. Ommaya reservoirs were placed into the abscess cavity and the frontal horn of the contralateral lateral ventricle. The patient was then treated with intracavitary/interstitial injections of amphotericin B during the course of 80 days and three doses of intraventricular amphotericin B. Clinical and radiographic improvement was achieved after treatment. Two years after the initial diagnosis, magnetic resonance imaging of the brain showed no evidence of disease and an examination revealed a neurologically intact and fully functional patient. CONCLUSION: We conclude that with an infection as morbid as rhinocerebral mucormycosis, it is advisable to use surgical debridement and all available routes for delivering amphotericin B to infected cerebral parenchyma, which include intravenous, intracavitary/interstitial, and cerebrospinal fluid perfusion pathways.

Anatomical basis of syringomyelia occurring with hindbrain lesions.

Hindbrain lesions that distort or compress the cervicomedullary junction are commonly associated with syringomyelia. As a basis for discussing pathogenetic mechanisms, the upper end of the central canal of the spinal cord was examined histologically in six aborted fetuses and 14 adults dying of natural causes; the results were correlated with magnetic resonance images in 40 normal subjects. The central canal of the medulla, which extends from the cervicomedullary junction to the fourth ventricle, was found to migrate dorsally, elongate in dorsoventral diameter, and dilate beneath the tip of the obex to form a large, everted aperture. This opening communicates directly with the subarachnoid space through the foramen of Magendie and is indirectly continuous with the main body of the fourth ventricle. In adults, the aperture of the central canal is located approximately 1.0 cm below the tela choroidea inferior and 3.5 cm below the midpoint of the fourth ventricle. Analysis of magnetic resonance imaging scans in 45 patients with syringomyelia and simple hindbrain lesions revealed two patterns of cavity formation: 1) lesions that obstructed the upper end of the central canal or its continuity with the subarachnoid space produced a noncommunicating type of syringomyelia; and 2) lesions that obstructed the basilar cisterns or the foraminal outlets of the fourth ventricle produced a communicating type of syringomyelia (hydromyelia) in association with hydrocephalus. Evidence is presented that syrinxes occurring with hindbrain lesions are not caused by a caudal flow of cerebrospinal fluid from the fourth ventricle into the central canal of the spinal cord.

Histopathology of experimental hematomyelia.

The pathology of hematomyelia was examined in 35 rats following the stereotactic injection of 2 microliters blood into the dorsal columns of the thoracic spinal cord. This experimental model produced a small ball-hemorrhage without associated neurological deficits or significant tissue injury. Histological sections of the whole spinal cord were studied at intervals ranging from 2 hours to 4 months after injection. In acute experiments (2 to 6 hours postinjection), blood was sometimes seen within the lumen of the central canal extending rostrally to the level of the fourth ventricle. Between 24 hours and 3 days, the parenchymal hematoma became consolidated and there was an intense proliferation of microglial cells at the perimeter of the lesion. The cells invaded the hematoma, infiltrated its core, and removed erythrocytes by phagocytosis. Rostral to the lesion, the lumen of the central canal was found to contain varying amounts of fibrin, proteinaceous material, and cellular debris for up to 15 days. These findings were much less prominent in the segments of the canal caudal to the lesion. Healing of the parenchymal hematoma was usually complete within 4 to 6 weeks except for residual hemosiderin-laden microglial cells and focal gliosis at the lesion site. It is concluded that the clearance of atraumatic hematomyelia probably involves two primary mechanisms: 1) phagocytosis of the focal hemorrhage by microglial cells; and 2) drainage of blood products in a rostral direction through the central canal of the spinal cord.