Polyradiculopathy in the course of superficial siderosis of the CNS.

Abstract

Sirs: We have read with interest the letter by Straube et al. [7] on a patient with superficial siderosis of the CNS due to repeated bleeding from a parasellar pilocytic astrocytoma, who also had flaccid paraparesis supposed to be due to a motor axonal polyradiculopathy caused by the superficial siderosis. The authors recognize that they have no evidence for siderosis affecting the lumbar roots. Sequences able to demonstrate haemosiderin deposits were not used in lumbar MRI, which showed, however, a “cystic enlargement of the sacral dura”. In the patient, the flaccid paraparesis with sphincter abnormalities, together with the electrophysiological findings of reduced motor nerve compound muscle action potentials and polyphasic motor unit potentials, suggests involvement of the spinal cord or the ventral roots, or both. However, these data are not sufficient to demonstrate definite involvement of the roots as opposed to damage to the cell bodies in the ventral horn or to the fibres emerging from the ventral horn (intramedullary fibres). A careful study of the F-responses could help clarify this issue. The clinical and electrophysiological features may be explained by the neuropathological abnormalities reported in superficial siderosis of the CNS. In his book, Developmental Neuropathology, Friede [3] shows a picture of a spinal cord affected by superficial siderosis of the CNS, which appears dark because of the deposits of the by-products of haemoglobin, in contrast to the pure white of the roots of the cauda equina. However, the roots appear dark for about 2mm at their entry into or exit from the spinal cord, with a “very sharp boundary between the heavily stained central part of the root and its unstained distal part” (Fig. 1). Friede adds that in the spinal cord, destruction of superficial fibres may occur and pigmentation “may extend into the gray matter, occasionally even affecting neurons”. In fact, in a series of articles, Koeppen and coworkers have demonstrated that superficial siderosis, i. e. incrustation of the subpial and subependymal CNS tissue by haemosiderin, is an active intracellular process which requires the presence of the glial tissue, particularly microglia, Bergmann glia and subpial astrocytes [2, 4–6]. This explains why the olfactory bulbs and the optic nerves are obviously affected, as they are part of the CNS [1]. It also explains why the VIII cranial nerve is the only other cranial nerve characteristically affected: this nerve has a segment which may be even longer than 1 cm that is composed by “central” myelin, and, therefore, may be affected by siderosis, while the other cranial nerves demonstrate the passage from central to peripheral myelin within 1–2 mm from their exit from the brainstem. The perfect correspondence between the extension of incrustations of superficial siderosis along the cranial nerves and their staining with glial fibrillary acidic protein (GFAP), which indicates the extension of central myelin in the nerve itself, is well illustrated in the paper published by Koeppen and Dentinger in 1998 [4]. The comparison of the “black”, incrusted, VIII cranial nerve and the non-siderotic VII cranial nerve made in that paper can now be obtained with MRI (Fig. 2) and explains why the VIII cranial nerve is characteristically affected and the VII cranial nerve is not. Returning to the case reported by Straube et al. [7], we suggest that LETTER TO THE EDITORS